principles and practice of pediatric sleep medicine

PRINCIPLES AND PRACTICE OF PEDIATRIC SLEEP MEDICINE ISBN 0–7216–9458–6Copyright © 2005 by Elsevier Inc.

All rights reserved. No part of this publication may be reproduced or transmitted in any form or by anymeans, electronic or mechanical, including photocopy, recording, or any information storage and retrievalsystem, without permission in writing from the publisher.

Library of Congress Cataloging-in-Publication Data

Principles and practice of pediatric sleep medicine/Stephen H. Sheldon . . . [et al.].– 1st ed. p. cm.

Includes bibliographical references.ISBN 0-7216-9458-61. Sleep disorders in children. 2. Children–Sleep. I. Sheldon, Stephen H.

RJ506.S55P746 2005618.92′8498–dc22 2004066249

Acquisitions Editor: Dolores MeloniProject Manager: Joan NikelskyDesign Coordinator: Ellen Zanolle

Printed in the United States of America.

Last digit is the print number: 9 8 7 6 5 4 3 2 1

Notice

Medicine is an ever-changing field. Standard safety precautions must be followed, but as new researchand clinical experience broaden our knowledge, changes in treatment and drug therapy may become nec-essary or appropriate. Readers are advised to check the most current product information provided bythe manufacturer of each drug to be administered to verify the recommended dose, the method and dura-tion of administration, and contraindications. It is the responsibility of the treating physician, relying onexperience and knowledge of the patient, to determine dosages and the best treatment for each individ-ual patient. Neither the Publisher nor the editor assumes any liability for any injury and/or damage topersons or property arising from this publication.

The Publisher

Prelims.qxd 1/25/05 3:03 AM Page ii

We dedicate this book, first to our families and, second, to all of the children and theirfamilies who have had the courage to work with us and help us achieve a better

understanding of how to recognize and treat sleep problems in the child.

Prelims.qxd 1/25/05 3:03 AM Page iii

v

Roberto L. Barretto, MDClinical Assistant Professor, Department ofOtolaryngology—Head and Neck Surgery,University of California, Irvine, School ofMedicine, Irvine; Attending PediatricOtolaryngologist, Children’s Hospital of OrangeCounty, Orange, California

Otolaryngologic Management of Sleep-Related BreathingDisorders

Lee J. Brooks, MDClinical Associate Professor of Pediatrics,University of Pennsylvania School of Medicine;Attending Physician, Pulmonary Division, TheChildren’s Hospital of Philadelphia, Philadelphia,Pennsylvania

Obstructive Sleep Apnea Syndrome in Infants and Children:Clinical Features and Pathophysiology; Enuresis inChildren with Sleep Apnea

Ronald D. Chervin, MD, MSAssociate Professor of Neurology, University ofMichigan Medical School; Director, SleepDisorders Center/Michael S. Aldrich SleepDisorders Laboratory, Department of Neurology,University Hospital, Ann Arbor, Michigan

Attention Deficit, Hyperactivity, and Sleep Disorders

Richard Ferber, MDAssociate Professor of Neurology, HarvardMedical School; Director, Center for PediatricSleep Disorders, Children’s Hospital Boston,Boston, Massachusetts

Mark E. Gerber, MD, FACS, FAAPAssistant Professor, Department ofOtolaryngology—Head and Neck Surgery,Northwestern University Feinberg School ofMedicine, Chicago; Attending, Children’sMemorial Hospital, Chicago, and EvanstonNorthwestern Healthcare (Evanston, Glenbrook,and Highland Park Hospitals), Evanston, Illinois


Daniel G. Glaze, MDAssociate Professor of Pediatrics and Neurology,Baylor College of Medicine; Medical Director, TheTexas Children’s Hospital, the Children’s Sleep

Center; Medical Director, The Methodist HospitalSleep Disorders Center, Houston, Texas

Sleep in Neurologic Disorders

David Gozal, MDChildren’s Foundation Chair of Pediatric Researchand Professor of Pediatrics, Pharmacology andToxicology, and Psychology and Brain Science,University of Louisville School of Medicine;Director, Comprehensive Sleep Medicine andApnea Center, Kosair Children’s Hospital; ViceChairman for Research and Director, KosairChildren’s Hospital Research Institute, Louisville,Kentucky

Consequences of Obstructive Sleep Apnea Syndrome

John H. Herman, PhD, FCCP, Dipl ABSMProfessor of Psychiatry and Director, SleepMedicine Fellowship Program, UT SouthwesternMedical Center School of Medicine; Director,Sleep Disorders Center for Children, Children’sMedical Center of Dallas, Dallas, Texas

Chronobiology of Sleep in Children; Circadian RhythmDisorders: Diagnosis and Treatment; Pharmacology ofSleep Disorders in Children

Eliot S. Katz, MDInstructor, Harvard Medical School; Attending,Division of Respiratory Diseases, Children’sHospital Boston, Boston, Massachusetts

Diagnosis of Obstructive Sleep Apnea Syndrome in Infantsand Children

Michael Kohrman, MDAssociate Professor of Pediatrics and Neurology,University of Chicago Pritzker School ofMedicine; Director, Pediatric ClinicalNeurophysiology, University of ChicagoChildren’s Hospital, Chicago, Illinois

Idiopathic Hypersomnia

Suresh Kotagal, MDProfessor and Chair, Division of Child andAdolescent Neurology, Department of Neurology,Mayo Medical School; Consultant, SleepDisorders Center, Mayo Clinic, Rochester,Minnesota

Narcolepsy in Childhood

Contributors

Prelims.qxd 1/25/05 3:03 AM Page v

Meir H. Kryger, MD, FRCPCProfessor of Medicine, University of ManitobaFaculty of Medicine; Director, Sleep DisordersCentre, St. Boniface Hospital Research Center, St.Boniface General Hospital, Winnipeg, Manitoba,Canada

Differential Diagnosis of Pediatric Sleep Disorders

Mark W. Mahowald, MDProfessor of Neurology, University of MinnesotaSchool of Medicine; Director, Minnesota RegionalSleep Disorders Center, Department of Neurology,Hennepin County Medical Center, Minneapolis,Minnesota

Disorders of Arousal in Children

Carole L. Marcus, MBBChProfessor of Pediatrics, University of PennsylvaniaSchool of Medicine; Director, Pediatric SleepCenter, The Children’s Hospital of Philadelphia,Philadelphia, Pennsylvania

Diagnosis of Obstructive Sleep Apnea Syndrome in Infantsand Children; Treatment of Obstructive Sleep ApneaSyndrome in Children

Susanna A. McColley, MDAssociate Professor of Pediatrics, Department ofPediatrics, Northwestern University FeinbergSchool of Medicine; Division Head, PulmonaryMedicine, Children’s Memorial Hospital, Chicago,Illinois

Primary Snoring in Children

Louise Margaret O’Brien, PhDResearch Fellow, Department of Pediatrics,University of Louisville School of Medicine,Louisville, Kentucky

Consequences of Obstructive Sleep Apnea Syndrome

Judith Owens, MD, MPHAssociate Professor of Pediatrics, Division ofAmbulatory Pediatrics, Brown Medical School;Director, Pediatric Sleep Disorders Clinic, HasbroChildren’s Hospital, Providence, Rhode Island

Epidemiology of Sleep Disorders during Childhood

Gerald M. Rosen, MD, MPHAssociate Professor, Department of Pediatrics,University of Minnesota School of Medicine;Medical Director, Pediatric Sleep Center,Children’s Hospitals and Clinics, St. Paul,Minnesota; Pediatric Sleep Specialist, MinnesotaRegional Sleep Disorders Center, HennepinCounty Medical Center, Minneapolis, Minnesota

Case-Based Analysis of Sleep Problems in Children;Disorders of Arousal in Children

Stephen H. Sheldon, DO, FAAPAssociate Professor of Pediatrics, NorthwesternUniversity, Feinberg School of Medicine;Director, Sleep Medicine Center, Children’sMemorial Hospital, Chicago, Illinois

Introduction to Pediatric Sleep Medicine; Anatomy ofSleep; Polysomnography in Infants and Children;Physiologic Variations during Sleep in Children;Disorders of Initiating and Maintaining Sleep; Post-Traumatic Hypersomnia; Kleine-Levin Syndrome andRecurrent Hypersomnias; Sleep in NeurologicDisorders; The Parasomnias; Sleep-Related Enuresis;Pharmacology of Sleep Disorders in Children

Julie L. Wei, MDAssistant Professor, Department ofOtolaryngology—Head and Neck Surgery,University of Kansas School of Medicine, KansasCity, Kansas; Attending Physician, PediatricOtolaryngology, Children’s Mercy Hospital,Kansas City, Missouri


Marc Weissbluth, MDProfessor of Clinical Pediatrics, NorthwesternUniversity Feinberg School of Medicine; ActiveAttending, Children’s Memorial Hospital,Chicago, Illinois

Sleep and Colic

vi Contributors

Prelims.qxd 1/25/05 3:03 AM Page vi

vii

Sleep medicine is a unique medical discipline,the one area of specialization that focuses onpatients’ health and disease during the hours ofsleep. Pediatric sleep medicine holds special sig-nificance in that it is concerned with about halfof the child’s life during the important earlyyears of development. Because sleep constitutessuch a large proportion of the child’s day, andbecause the development of the neurologicstructures responsible for sleep, in patterns ofincreasing complexity and predictability, occursso rapidly during the early months and years oflife, the maturation of these systems is clearlyamong the most important of neurodevelopmen-tal milestones to be met. Early disruption ofthese vital processes has major cognitive andpossibly emotional consequences that are onlyrecently becoming clear. Studies have repeatedlyconfirmed the importance of sleep for the acqui-sition of new knowledge and for processing andmaintaining information previously gained.Knowledge of primary and secondary sleep-related disorders also has been shown to beessential to the understanding of many otherchildhood disorders and in the development ofproper treatment.

Principles and Practice of Sleep Medicine in theChild, first published as a separate volume fromPrinciples and Practice of Sleep Medicine (2nd edi-tion) in 1995, helped move the practice of pedi-atric sleep medicine from the sphere of adultmedicine to one that is overseen by professionalswho have dedicated their careers to the healthand well-being of the pediatric patient. As statedin the Preface to that volume, “In recent years, arobust scientifically based body of knowledgehas emerged, and the tools to diagnose and effec-tively treat children with sleep disorders are nowavailable.” Principles and Practice of PediatricSleep Medicine represents another step in thedevelopment of pediatric sleep medicine as a dis-tinct discipline.

Diagnosis and management of sleep disordersin children may hold unique significance in atleast three very important ways:

1. Primary sleep-related pathology may directlycause daytime symptoms, and only throughtreatment of the sleep disorder is resolutionpossible.

2. Sleep-related pathology may be a co-morbidcondition contributing to daytime symptoma-tology, and through treatment of the sleep-related pathology, the patient becomes moreresponsive to treatment of the co-existing dis-orders.

3. A child’s sleep difficulties may have greaterimpact on other family members than on theaffected child, causing a caretaker, for exam-ple, to be sleep deprived, with medical andemotional compromises (including impairedability to care for the child). Treating theseproblems can improve the lives of child andfamily members alike.

As we continue to gain better understanding ofthe effect of sleep disorders on children, majoradvances will be made, with great significancefor all three of these scenarios, including insightinto the often unclear area of cause and effect—namely, whether disordered sleep is a cause or aresult of a specific disease or disorder.

Sleep disorders in infants and children reflectan interplay among many factors, includingcentral nervous system function, parental-childinteraction, social stress, patient needs, andother medical conditions. Comprehensiveknowledge of these interactions is essential forall child-care professionals who want to deliveroptimal management. This book providesresources for sleep medicine specialists as well asprimary care practitioners to use to deliver thebest possible care to their pediatric patientsthroughout the 24-hour day.

Stephen H. Sheldon, Chicago

Richard Ferber, Boston

Meir Kryger, Winnipeg

Preface

Prelims.qxd 1/25/05 3:03 AM Page vii

ix

The editors wish to thank the entire editorial staffat Saunders/Elsevier for their outstanding effortsin the creation of this book. Specifically, we wishto single out Joan Nikelsky and Delores Melonifor their patience and persistence, without whichthis project could have never been completed.

We also would like to thank all of the coura-geous health care professionals who care for chil-dren during sleep—for their dedication and

foresight to understand the importance of sleep inhealth and disease in the pediatric patient.Beginning with the scientific foundation laid bypioneers Nathaniel Kleitman, Arthur Parmelee,Jr., and Heinz F.R. Prechtl, many other childhealth care practitioners have begun to follow intheir footsteps, delving into the life of children atnight. These researchers and practitioners of pedi-atric sleep medicine truly are leading the way.

Acknowledgments

Prelims.qxd 1/25/05 3:03 AM Page ix

FUNCTION OF SLEEP

Although sleep has been studied subjectively andobjectively for more than 100 years, the exactfunction of sleep and its components remainselusive. Historically, physicians have recom-mended sleep for the treatment of many disor-ders. This prescription has been based on theassumption that sleep provides a unique restora-tive purpose. Although the relationship betweenthe immune system and sleep may be importantwith many clinical implications, no study hasdocumented that sleep cures anything.1

Circadian rhythms of various biologic processes,for example the immune system, appear to bemodulated by sleep, and lymphocyte functionsare dramatically altered at sleep onset and duringsleep.2 Specific pokeweed mitogen response andnatural killer cell activity are altered by sleep inhealthy young men. Interleukin-1–like activitiesare followed by interleukin 2–like activities dur-ing sleep, and interleukins-1 and -2 are disruptedby sleep deprivation.3 Insomnia has been shownto be associated with nocturnal sympatheticarousal and declines in natural immunity, includ-ing a decrease in natural killer cell response.3a

Narcoleptic patients present disordered diurnalpatterns of immune function;4 however, the clin-ical effects that these changes produce or howthey may be therapeutically modified areunknown.

Theories of sleep function fall into severalmajor categories, with many overlaps. An under-standing of these hypotheses provides a basis forcomprehension of the varied effects that disor-dered sleep may have on health and disease.

Restoration TheoryIn 1946 Sherington suggested that sleep was astate required for enhanced tissue growth and

repair.5 This hypothesis proposes that certainsomatic and/or cerebral deficits occur as a resultof wakefulness, and sleep allows or promotesphysiologic processes to repair or restore thesedeficits thereby assuring normal daytime func-tioning.6-8 Special focus has been placed on bothrestoration of somatic function and the centralnervous system (CNS) function. Non–rapid eyemovement (NREM) sleep is thought to functionin reparation of body tissue and REM sleep inrestoration of brain tissue. Supporting evidence,however, is indirect. Napping, when its timingand duration are designed properly, has thepotential to improve waking function. Even briefnapping of less than 30 minutes in durationcounteracts decreased alertness and performanceunder conditions of sleep deprivation.8a Therole of NREM sleep in repair of somatic tissuecomes from investigations that have shown thefollowing:

1. Slow-wave sleep (SWS) increases followingsleep deprivation.9

2. The percentage of SWS is increased duringthe developmental years.10

3. Total sleep duration increases with bodymass.11

4. Release of growth hormone occurs at sleeponset and peak levels occur during SWS inprepubertal children.12

5. The release of many endogenous anabolicsteroids (prolactin, testosterone, and luteiniz-ing hormone) occurs in relation to a sleep-dependent cycle.13,14

6. The nadir of catabolic steroid release, such ascorticosteroids, occurs during the first hoursof sleep, coincident with the largest percent-age of SWS.15

7. Increased mitosis of lymphocytes andincreased rate of bone growth occur duringsleep.16

Introduction to Pediatric SleepMedicine

Stephen H. Sheldon

1

1

Ch01.qxd 1/25/05 3:05 AM Page 1

8. There is a gradual increase in SWS percentageof total sleep time in response to a gradedincrease in physical exercise.17

However, other observations might suggestan influence of sleep on a physiologic processrather than a direct effect. For example, whilepeak rates of cell division occur during sleep,they do not appear to be due to sleep itself.Increased mitosis is demonstrable after a nightwithout sleep and is positively influenced by oralglucose load and negatively influenced by corti-sol secretion.18 Similarly, in adolescents andadults, somatomedin levels are the highest dur-ing wakefulness, not during sleep as it is in pre-pubertal children.19

On the other hand, REM sleep—characterizedby intense CNS activation—has been thought tofunction in the restoration of CNS function. Itmay have evolved in order to “reprogram” innatebehaviors and to incorporate learned behaviorsand knowledge acquired during wakefulness.20

The synthesis of CNS proteins is increased duringREM sleep.21 This sleep state also appears in sig-nificantly higher proportions in the fetus andnewborn, gradually decreasing over the first fewyears of life. Increased protein synthesis duringREM sleep may be critical in the development ofthe CNS.

Evolutionary and AdaptiveTheoriesDevelopment of many physiologic functions fol-lows an orderly progression that mirrors phylo-genetic development. It has been suggested thatthe development of sleep in the human organismalso follows this same phylogenetic pattern.Evidence for this theory is scant. Animals sleepin many different ways, often influenced moreby the environment and lifestyle than by evolu-tion of the species.22 SWS and REM sleeprebound are characteristic features seen aftersleep deprivation in the dog, cat, rabbit, andhuman.23 Definitive REM sleep, however, hasnever been documented in the dolphin.Dolphins do not have a pulmonary reflex tohypoxemia; therefore, they have complete, vol-untary control of breathing, and sleep wouldpresumably be associated with impaired neu-rorespiratory control. Actually, dolphins appearto exhibit hemispheric sleep. That is to say, when

dolphins appear to sleep, slow-wave patterns areseen over a single hemisphere at a time, whilethe other hemisphere shows waking rhythm.24 Ifthe evolutionary theory of sleep is true, animalswith highly complex CNS function, such as thedolphin, should follow this pattern. It stands toreason that if the dolphin slept in the same man-ner as the dog, cat, and human, survival in itsaquatic environment would be impossible.Skeletal muscle atonia during REM sleep (as cur-rently understood) would result in drowning.Therefore, the lifestyle and environment of thedolphin play a much more significant role in thepattern of sleep development in these speciesthan phylogeny.

In some species, sleep may function toenhance survival. Animals that graze for foodtend to sleep in bursts over a short period oftime, a behavior that may provide the timeneeded for sufficient food-seeking while protect-ing the animal from predators.25 Carnivorousanimals who do not require large amounts oftime for foraging and who are relatively safe frompredation tend to sleep for long periods of time.

Sleep may also be an instinctive behavior, apatterned response to stimuli that conservesenergy, prevents maladaptive behaviors, and pro-motes survival.26 According to the evolutionarytheory of sleep, cortical activation in REM sleepmay perform additional survival functions.27

Energy Conservation TheorySleep may function to conserve energy. In fact,mammalian species exhibit a high correlationbetween metabolic rate and total sleep time.28

This view states that energy reduction is greaterduring sleep than during periods of quiet wake-fulness and that sleep provides periods ofenforced rest, barring the animal from activityfor extended periods of time. Endothermic ani-mals exhibit SWS. During NREM sleep, endoge-nous thermoregulation continues, thoughfunctioning at levels below those of wakefulness.Poikilothermic species, on the other hand, donot exhibit clear SWS patterns. It is doubtful,however, that this theory explains the functionof sleep in humans. Reduction in metabolismthat occurs during sleep is minimal. Althoughthe hypothesis is intriguing, energy conservationtheory has been disputed and evidence existsthat an increase in sleep time does not correlate

2 Chapter 1 Introduction to Pediatric Sleep Medicine

Ch01.qxd 1/25/05 3:05 AM Page 2

with increased metabolic rate.29,30 It has beenshown that there is only an approximately 8 to10% reduction in metabolic rate during sleepwhen compared with relaxed wakefulness. Thiswould be insignificant when considering anadult human’s basal metabolic expenditure.

Learning TheoryA particularly interesting theory of the functionof sleep centers on the role of sleep in the processof learning and memory. A significant body ofknowledge exists which suggests that retentionof new information depends on activation ofsome brain function that occurs at a criticalperiod after the registration of this informa-tion.31,32 Two pivotal phases appear to exist. Thefirst one is “consolidation.” Medication thatcauses stimulation of the reticular activating sys-tem and cortical excitation during the first 90seconds after acquisition of new informationappears to enhance memory and increase reten-tion. Although the consolidation phase of learn-ing is important, it cannot be considereddefinitive for fixation of information, since pro-cessing continues for a long period of time.

The second critical phase of information pro-cessing seems to occur during sleep, specificallyREM sleep. Two theories have been proposed,one based on the passive hypothesis (unlearningtheory) and the other on the active hypothesis,which suggests that there are active consolida-tion mechanisms. An active process is supportedby the following facts: Considerable brain activ-ity occurs during this phase of sleep, in whichbrain oxygen consumption increases, there is anincrease in cerebral blood flow, and there isintense activity of cortical and reticular neuronsindicating an active, functional process.

Over the past 50 years, the beneficial effectsof sleep on the retention of memories acquiredduring wakefulness have been documented.33,34

REM sleep appears to have special significance.Despite evidence from animal and human stud-ies, the exact function of REM sleep in child-hood development and learning remainsunknown. Diverse reasons have been proposedfor children’s learning difficulties, but no singlefactor appears to be consistent for all individu-als. Most diagnostic and treatment protocolshave empirically focused on the child’s daytimecapabilities.

Theories on the function of REM sleep anddreaming, with which it has a contingent rela-tionship, remain diverse. Facilitation of memorystorage, reverse learning, anatomic and func-tional brain maturation, catecholamine restora-tion, and others have been postulated.35

Growing evidence supports the idea that sleepfollowing a learning experience is critical tomemory formation. Studies suggest that infor-mation acquired during wakefulness is reacti-vated and possibly consolidated duringsubsequent REM sleep.36 Several brain areashave been shown to be activated duringsequence learning when awake and during sub-sequent REM sleep.37 This activation suggeststhat REM sleep participates in the reprocessingof recent memory traces.38 Regional cerebralreactivation during post-training REM sleep isnot related simply to the acquisition of basicvisuomotor skills during prior practice of a serialreaction time task but rather to the implicitacquisition of the stimulus sequence. REM sleepis deeply involved in the reprocessing and opti-mization of high-order information contained inthe material to be learned. Additionally, the levelof acquisition of probabilistic rules attainedbefore sleep correlates with an increase inregional cerebral blood flow during subsequentREM sleep. This suggests that post-training cere-bral reactivation is modulated by the strength ofthe memory traces developed during the learn-ing episode.38

Interestingly, learning of some perceptualskills has been shown to depend on the plasticityof the visual cortex and to require post-trainingnocturnal sleep.39 Sleep-dependent learning of atexture discrimination task can be accomplishedin human subjects having only brief (60- to 90-minute) naps containing both NREM sleep andREM sleep.

Newborn infants preferentially orient to face-like patterns at birth, but months of experiencewith facial observations are required for full face-processing abilities to develop.40 Models gener-ally assume that the brain areas responsible fornewborn orienting responses are not capable oflearning and are physically separate from thosethat later learn from real faces. Newborn face-orienting may be the result of prenatal exposureof a learning system to internally generatedinput patterns, such as those found in ponto-geniculo-occipital (PGO) waves during active

Introduction to Pediatric Sleep Medicine Chapter 1 3

Ch01.qxd 1/25/05 3:05 AM Page 3

(REM) sleep. Neonates spend about 18 hoursasleep per day, at term, and about 50% of thistime is spent in active / REM sleep. Preterminfants spend considerably more time in activesleep. A combination of learning and internalpatterns is an efficient way to specify anddevelop circuitry for facial perceptions.40 Thisprenatal learning can account for the newbornpreferences for schematic and photographicimages of faces, providing a computationalexplanation for how genetic influences interactwith experiences to construct a complex adap-tive system.

Sleep loss adversely affects certain types ofcognitive processing, particularly associativememory.41 Long-term potentiation represents aputative cellular basis for learning and memoryconsolidation. The influence of sleep depriva-tion has been shown to result in the delay ofmaximal induction, and the degradation of themaintenance phase of long-term potentiationmay represent the sleep deprivation–inducedimpairment of the underlying neurochemicalmechanisms normally responsible for memoryacquisition.

Studies relating sleep states to memoryprocesses typically present learning material toparticipants and then examine recall abilityafter intervening sleep or sleep deprivation.42

Most studies have utilized either sleep record-ings or sleep deprivation after a learning task.Cueing and positron emission tomography havealso been utilized. Data strongly suggest thatREM sleep is involved in the efficient memoryprocessing of cognitive procedural material butnot declarative material. There are some data,however, which support the contention thatSWS or NREM sleep is necessary for declarativememory consolidation. Additionally, the lengthof the NREM-REM cycle may also be importantfor declarative memory. Stage 2 NREM sleepmay also be involved in memory for motor pro-cedural but not cognitive procedural tasks.42

After declarative learning tasks, the density ofsleep spindles is significantly higher than thatwith a nonlearning control task and is greatestduring the first 90 minutes of sleep.43 Spindledensity also correlates with recall performanceboth before and after sleep. These findings indi-cate that spindle activity during NREM sleep isquite sensitive to previous learning experi-ences.44

Flexible cognitive processes are regarded asfundamental to problem solving and creativity.REM sleep has been shown to be associated withcreative processes and abstract reasoning, withincreased strength of weak associations in cogni-tive networks.43 When early and late REM andNREM awakenings are assessed, a dissociationbecomes evident, with NREM-awakening taskperformance becoming more REM-like later inthe night but REM-awakening performanceremaining constant. The neurophysiology of REMsleep, therefore, represents a brain state moreamenable to flexible cognitive processing thanthat of NREM sleep.

The role of sleep in the development of theCNS and neural circuitry may be extremelyimportant. Kisley and colleagues showed thatnormal, young infants exhibited significantresponse suppression.45 A correlation betweenincreasing age and stronger response suppres-sion was also uncovered, even within an agerange restricted to 1 to 4 months. These datasuggest that neuronal circuits underlying sen-sory gating are functional during very early post-natal development.

Minor neurologic and EEG abnormalities havebeen described in children with “hyperactivity”syndrome.46 These abnormalities have been asso-ciated with specific or global learning difficulties,and the syndrome has previously been describedas “minimal cerebral dysfunction.” However, neu-rologic and EEG abnormalities associated withthis hyperactivity syndrome have been shown tobe nonspecific and variable,47 resulting in a changeof the name of the syndrome to attention deficithyperactivity disorder (ADHD).

It is noteworthy that of the 15 reading-dis-abled (dyslexic) children studied by Levinson,97% revealed evidence of cerebellar-vestibular(C-V) dysfunction. Ninety-six percent of 22blinded neurologic examinations and 90% of 70completed electronystagmograms indicated sim-ilar C-V dysfunction.48 Ottenbacher and associ-ates explored the relationship between vestibularfunction, as measured by duration of postrota-tory nystagmus, and human figure–drawing abil-ity in 40 children labeled as learning disabled.49

Chronologic age and postrotatory nystagmusdurations shared significant amounts of variancewith human figure–drawing. The variables of IQand sex were not significant. DeQuiros andSchrager have also identified vestibular dysfunc-


Ch01.qxd 1/25/05 3:05 AM Page 4

tion in some learning-disabled children50 anddescribed another related syndrome termed“vestibular-oculomotor split,” which results inimpaired ocular fixation, reduced scanning abil-ity, and poor eye-head coordination.

Despite the evidence that some children withlearning disabilities display soft or nonfocal neu-rologic signs51 and low scores on tests of visuo-motor integration, reading achievement, andocular scanning,52 contrary evidence of normalvestibular responses to rotation in dyslexic chil-dren has been published. Brown and coworkersmeasured eye movements provoked by sinu-soidal rotation of the subjects at low frequen-cies.53 Gain, phase, and asymmetry of theresponses were calculated from the eye velocityand stimulus velocity waveforms. There were nodifferences between the groups in any of themeasurements. These results led to the conclu-sion that “there are no clinically measurable dif-ferences in this aspect of vestibular function” intheir carefully selected population of dyslexicand control children. However, their conclusionswere based on evidence obtained during thewaking state. Vestibular nuclei play a major rolein the control of eye movements when awakeand asleep. If these nuclei are destroyed, eyemovements during REM sleep are absent. In apilot study of four reading-disabled children, asignificant difference was found in the meanangular velocity of eye movements during REMsleep when compared with three normal-readingcontrols.54

Correlations exist that associate the phasicevents of REMs with C-V control. Pompeianohas shown that lesions of the medial anddescending vestibular nuclei in the cat elimi-nated all phasic inhibition of sensory input,spinal reflexes, and all motor output associatedwith phasic REM bursts, including eye move-ments themselves.55,56 They have also demon-strated that intense, spontaneous dischargesfrom neurons of the vestibular nuclei occur syn-chronous with the ocular activity of REM sleep.Nystagmus evoked by rotation can be most read-ily induced during sleep at the time of phasicevents of REM sleep57 and in Wernicke-Korsakoff disease, in which the vestibular nucleiare often damaged and eye movements areabsent during REM sleep.58 These observationspartially confirm the influences of vestibularmechanisms on the phasic activity of REM sleep.

An age-related development of phasic inhibi-tion of auditory evoked potentials during theocular activity of REM sleep and an age-relatedincrease in the duration of the REM bursts havebeen described in normal subjects.59 It seemsthat central vestibular influences underlie theseevents and that vestibular control of phasicactivity follows a developmental maturationschedule.

Considering this evidence, there may be arelationship between the C-V control of REMphasic events and the importance of phasic REMsleep in learning and memory. However, thisrelationship currently remains a mystery.

A significant body of literature exists support-ing a relationship among REM sleep, phasicREM activity, and learning. Sleep patterns inhyperkinetic and normal children were studiedby Busby and colleagues.60 Analysis of sleep pat-tern variables revealed a significantly longerREM onset latency and greater absolute and rel-ative amounts of movement time for the hyper-kinetic group than controls. No other sleepparameter differentiated the groups. Clinicalobservations of autistic children have suggestedthat fundamental symptoms of the syndrome ofchildhood autism involve disturbances of motil-ity and perception. The nature of these distur-bances indicates a maturational delay in thedevelopment of complex motor patterns and themodulation of sensory input.59 Sleep studieshave provided some evidence for a maturationaldelay in the differentiation of REM sleep patternsand the development of phasic excitatory andinhibitory mechanisms during REM sleep inthese children. These findings implicate a failureof central vestibular control over sensory trans-mission and motor output during REM sleep.The notation that there is a dysfunction of cen-tral vestibular mechanisms underlying thedelayed organization and differentiation of theREM sleep state is supported by observations ofaltered vestibular nystagmus in the waking statein autistic children.59

Studies in animals and humans support theimportance of REM sleep in learning. Luceroconducted experiments that showed a signifi-cant increase in REM sleep duration with respectto controls, a nonsignificant increase in totalsleep time, and no changes in SWS duration inanimals subjected to consecutive learning expe-riences.61 An increase in REM sleep time


Ch01.qxd 1/25/05 3:05 AM Page 5

observed after incremental learning suggests thatREM sleep might be involved in the processingof information acquired during wakefulness. Ithas been postulated that such processing mightconsist of the transformation of a “labile” pro-gram acquired in the learning session into amore “stable” program devoid of superfluousinformation.

Major evidence regarding the importance ofREM sleep in facilitating recall of complex asso-ciative information has been documented byScrima.62 The beneficial effects of isolated REMand isolated NREM sleep on recall were tested in10 narcoleptic subjects. The results of complexassociative tasks indicated significant differencesamong three conditions for free recall. Recallwas significantly better after isolated REM thanafter isolated NREM sleep or wakefulness andwas significantly better after NREM sleep thanafter wakefulness. It was concluded that theresults were consistent with the proposed neu-ronal activity correlation theory of Emmons andSimon that REM sleep actively consolidatesand/or integrates complex associative informa-tion and that NREM sleep passively preventsretroactive interference of recently acquiredcomplex associative information.63

Newborn animals and human infants show agreater proportion of REM sleep with respect tototal sleep time than adults64,65 and a progressivedecrease in that proportion as growing continuesthat is paralleled by a decrease in learning abil-ity.24 Fishbein has shown that REM sleep depri-vation, both before and after learning, disruptsprimarily long-term memory processes.66

Evidence has been provided that learninginduces a protracted augmentation of paradoxi-cal sleep time, lasting for at least 24 hours.67

This work, together with previous work, suggeststhat REM sleep augmentation may be a neurobio-logic expression of the long-term process of mem-ory consolidation. Fishbein was able to augmentREM sleep using behavioral techniques of learn-ing. Therefore, one psychobiologic function ofREM sleep may be to process and maintain infor-mation during wakefulness.

Results obtained from nondeprivation studiesof animals provide consistent support for thehypothesis that REM sleep is functionally relatedto learning. Results of studies that haveemployed multiple training sessions may be

interpreted to suggest that either prior REM sleepprepares the organism for subsequent learning orthat REM sleep facilitates consolidation andretrieval of prior learning. Given the equivocalityof prior REM deprivation literature, the secondinterpretation seems more reasonable.68

Impaired cognitive functioning has been doc-umented in studies on sleep-deprived physicians.In one investigation, cognitive functioning inacutely and chronically sleep-deprived houseofficers was evaluated.69 Analysis of data revealedsignificant deficits in primary mental tasksinvolving basic rote memory and language andnumerical skills, as well as in tasks requiringhigh-order cognitive functioning and intellectualabilities.

Acquisition of many simple learning tasks inanimals is followed by augmentation of REMsleep without any modification of NREMsleep.31 Augmentation of REM sleep after learn-ing has also been described in human infants.70

Sleep may be particularly important for RNA andDNA synthesis linked to memory processes.There is some evidence that during sleep, RNA ismore actively synthesized, less rapidly degraded,and more slowly transported into the cyto-plasm.71

As in infants, there is some evidence thatREM sleep may increase following learning inolder children and adults. Hartman has demon-strated an increase in REM sleep time occurringafter days of increased learning, mental stress,and especially demanding events.72 If learningdoes cause an increase in REM sleep, brain-dam-aged patients who are improving should have ahigher proportion of REM sleep than those whoshow no improvement. Following up a group of9 patients with severe traumatic brain damage,Ron and associates found a correlation betweencognition and REM sleep improvement in 7.73

Greenberg and Dewan compared the percentageof REM sleep in improving and nonimprovingaphasic patients and found that the nonimprov-ing patients did, in fact, have less REM.74 In 32patients with Down syndrome, phenylketonuria,and other forms of brain damage, Feinberg founda positive correlation between the amount of eyemovement during REM sleep and estimates ofintellectual function,10 while in a comparison of38 normal individuals and 15 brain-damagedpatients, it was shown that mentally retarded


Ch01.qxd 1/25/05 3:05 AM Page 6

patients had less REM sleep.75 Linkage betweenREM sleep and development of the visual systemhas been further supported by a recent study byOksenberg and coworkers.76 They reported sig-nificant adverse changes in the microscopicanatomy of the visual cortex in REMsleep–deprived cats.

In spite of evidence from human and animalstudies, the exact function of sleep in the processof learning, memory, and child development isstill speculative.

Unlearning TheoryAn antithetical hypothesis for the function ofREM sleep in learning and memory involves aprocess of unlearning. No single memory centerappears to exist in the brain.77 Many parts of theCNS participate in the representation of a singleevent. However, localization of memory for asingle event generally involves a limited numberof neural pathways, and those collections of neu-rons within which a memory is equivalently rep-resented probably contain a set of no more thana thousand neurons. These interconnectedassemblies of cells could store associations.78,79

If the cells involved in the memory of an eventform mutual synapses, regeneration of the activ-ity of the entire neuronal set would occur whenpart of that event is encountered again.Therefore, Crick and Mitchison proposed thatthe function of REM sleep is to remove certainundesirable modes of interaction in networks ofcells in the cerebral cortex.80 This would beaccomplished during REM sleep by a “reverselearning” mechanism so that the trace in thebrain of the unconscious dream can be weak-ened rather than strengthened by the activity ofdreaming. A mathematical and computer modelof a network of 30 to 1000 neurons has beendeveloped by Hopfield and colleagues.81 Theirmodel network has a content-addressable mem-ory or “associative memory” that allows it tolearn and store many memories. A particularmemory can be evoked in its entirety when thenetwork is stimulated by an adequately sizedsubpart of the information of that memory.When memories are learned, spurious informa-tion is created and can also be evoked. Applyingan unlearning process, similar to the learningprocess but with a reversed sign and starting

from a noise input, enhanced the performance ofthe network in accessing real memories andminimizing spurious ones.

Sleep (in particular REM sleep) may thenfunction to reduce or prevent unwanted, unno-ticed, or spurious material acquired duringwakefulness. Isolation of the cortex from envi-ronmental stimuli may be a necessary feature ofthe removal of inhibiting and /or competitivestimuli, inappropriate behavior patterns, andoverloading of neuronal networks, thus permit-ting reprogramming and consolidation of morevital material.

Other HypothesesThe presence of circulating hypnotoxin(s) as pos-sible causes of sleep has received some attentionin the literature. These theories hold that duringwakefulness there is accumulation of sleep-pro-ducing “toxin(s)” which stimulates or results insleep. Once the toxin(s) has been modified duringsleep, awakening occurs. Investigation of manysubstances, which are claimed to be released sub-sequent to sleep deprivation, and thalamic stimu-lation have failed to produce convincing evidenceof the accuracy of this theory.82 Studies of cran-iopagus and thoracopagus twins reveal independ-ent sleep–wake cycles despite the sharing ofcirculatory and /or nervous systems.83

Sleep may be required for the normal func-tioning of the motor system and /or skeletal mus-culature. Muscle aching and change in skeletalmuscle enzyme activity have been reported fol-lowing NREM sleep deprivation.84 Although theability to work is not drastically affected by sleepdeprivation, physical performance follows defi-nite circadian rhythmicity.85 Sleep also has a pro-found effect on certain diseases that affect themotor system, including Parkinson’s disease86

and hereditary progressive dystonia.87

Although all theories on the function of sleephave had significant support, there is much con-flicting evidence for each theory. The function ofsleep may be explained very simply or may beone of the more complex biologic mechanismsknown. A unitary explanation of the function ofsleep is probably unrealistic. The exact functionand purpose of sleep may prove to be a combi-nation or a series of integrations of all proposedhypotheses.


Ch01.qxd 1/25/05 3:05 AM Page 7

SLEEP DEPRIVATION

Experiments conducted to unravel the mean-ing and functions of various physiologicprocesses have involved modification of thesesystems followed by observations of the conse-quences of each modification. For many years,research has focused on total sleep deprivation,partial sleep deprivation, and deprivation ofvarious sleep stages in an attempt to identifyrepercussions.

Total sleep deprivation studies in animals haveshown significant deleterious effects. In earlyexperiments, in order to keep animals awake forprolonged periods of time, constant activity wasnecessary, confounding the results. In 1983,Rechtschaffen and coworkers described an ingen-ious experimental method that controlled for thestimuli used to keep the animals awake.88

Experimental and control rats were placed on aplatform that rotated and caused awakening onlywhen the experimental animal fell asleep. Theactivity of the experimental and control animalswas kept constant. Experimental animals sufferedsevere pathologic changes from the sleep depriva-tion. Changes ranged from a severely debilitatedappearance (ungroomed and yellowed fur) tointense neurologic abnormalities (ataxia andmotor weakness) and death. Interestingly, therewas a loss of EEG amplitude to less than half ofnormal waking values before the morbidity ormortality of the experimental animals. Necropsyfindings included pulmonary edema, atelectasis,gastric ulcerations, gastrointestinal hemorrhage,edema of the limbs, testicular atrophy, scrotaldamage, bladder enlargement, hypoplasia of theliver and spleen, and hyperplasia of the adrenalglands (indicating a significant stress response).Body weight decreased in both experimentaland control animals but was significantlygreater in the experimental, sleep-deprived ani-mals. It is also interesting to note that the ani-mals that ate the most lost the most weight.There was a surprisingly high correlationbetween the amount of paradoxical (REM) sleepobtained by the experimental animals and thesurvival time.

In comparison, the effects of total sleep depri-vation on human subjects have been remarkablyfew. Although brief psychotic episodes havebeen reported in some subjects, long-term psy-chological effects do not appear to result.89 The

only certain and reproducible effect of total sleepdeprivation has been sleepiness.

Fatigue, decline in perceptual, cognitive,and psychomotor capabilities, and increasingtransient ego-disruptive episodes have beenreported by Kales and associates during sleepdeprivation.90 Reality testing was impairedand regressive behavior was noted as theexperiment continued. Tests for thought dis-orders showed shifts in thought processes to amore childlike level of cognition; however,there was no obvious evidence of schizo-phrenic thinking.

Performance may also be impaired by sleepdeprivation. Vigilance and performance on reac-tion time tests have been shown to be signifi-cantly impaired by the loss of as little as onenight of sleep.91 In a study of 44 men participat-ing in a strenuous combat course in Norway, sig-nificant impairment was observed in vigilance,reaction time, code testing, and profile of moodstate after 24 hours of sleep deprivation.Complaints of symptoms occurred first, and dis-turbances of senses and behavior followed.Horne and coworkers have reported that after 60hours of continuous wakefulness, an inherentcapacity for signal detection exhibited a stepwisedecline during deprivation, falling sharply dur-ing the usual sleep time period and leveling outduring the daytime.92 A clear circadian rhythmoverlaid the decline due to deprivation. It wasconcluded that changes in this inherent capacityseem to be consistent with a brain “restitutive”role for sleep.

Significant physiologic changes after totalsleep deprivation have been reported. Reboundof stage 4 sleep on the first recovery day andREM rebound on the second and third recoverydays were documented by Berger and Oswald in1962 and Williams and colleagues in 1964.9,93

After 205 hours of sleep deprivation, Kales andassociates reported significant increases in stage4 sleep and REM sleep and significant decreasesin stage 2 sleep.90 On the first two recoverynights, alterations in REM sleep were noted.There was an increase in REM percentage,appearance of sleep-onset REM (SOREM) peri-ods, a decrease in REM latency, and a decrease inintervals between REM sleep periods. Theseoccurred most dramatically in those subjectswho had the greatest psychological disturbancesduring the deprivation period.


Ch01.qxd 1/25/05 3:05 AM Page 8

Significant changes in performance and sleepphysiology have been documented in subjectsonly partially deprived of sleep. In 1974, Webband Agnew reported the results of an experimentconducted on 15 subjects restricted to a regimenof 51⁄2 hours of sleep a night for a period of 60days.94 The initial effect was an increase in thetotal amount of stage 4 sleep. By the fifth weekof the experiment, the amount of stage 4 sleepreturned to the baseline level. The initial effecton REM sleep was a sharp reduction when com-pared with the baseline level. During the entirecourse of the experiment, there was a reductionin REM sleep by 25%. Latency to the onset ofstage 4 sleep and latency to the onset of REMsleep were also reduced. Behaviorally, only theWilkinson Vigilance Task showed a decline inperformance associated with continued sleeprestriction. Initially, the subjects experienced dif-ficulty with arousal in the morning and feltdrowsy during the day, but this did not continuethroughout the entire experiment. Mood scalesshowed no significant changes. It was concludedthat the chronic loss of as much as 2.5 hours ofsleep per night is not likely to result in majorbehavioral consequences. However, significantphysiologic effects were documented (especiallyin REM sleep) polysomnographically with par-tial restriction. Restricting sleep by early morn-ing awakenings generally deprives the subject ofREM and stage 2 sleep, usually leaving NREMstage 4 sleep intact. However, recovery nightshows a substantial increase in the amount ofNREM stage 4 sleep, suggesting that the amountof high-voltage SWS is due to some extent to thepreceding total sleep time.95

Children differ from adults in their responseto acute restriction in sleep. When sleep hasbeen restricted by 4 hours or more, there is adecrease in all stages of sleep (except SWS), areduction in sleep onset latency, stage 4 latency,and REM latency, and a reduction of wakefulnessduring the sleep period. Carskadon and cowork-ers studied the effect of acute, partial restrictionof sleep in nine children between the ages of 11and 13.2 years.96 These children were permittedto sleep 10 hours on baseline and recoverynights and 4 hours on a single restricted night.No significant differences were found on anyperformance test. The lack of demonstrableeffects may have been because the tests werebrief. Wilkinson emphasized that task duration

is a major factor in determining a test’s sensitiv-ity to sleep loss.97 On the other hand, significantchanges did occur on objective sleep measure-ments. The multiple sleep latency test showed asignificant increase in daytime sleepiness, whichpersisted into the morning after sleep restriction.This suggested that children were more severelyaffected by sleep restriction than adults. Onpolysomnography, the findings were comparableto those in adults, but children did not showrecovery rebound of SWS and REM sleep similarto that reported for adults. Although childrenappear to be able to tolerate a single night ofrestricted sleep without a decrement in perform-ance of brief tasks, perhaps more prolongedrestriction and prolonged tasks similar to thoserequired in school would show decrements.Children seem to require more time to recuper-ate fully from nocturnal sleep restriction thanadults. The extent of daytime sleepiness thatoccurs is not trivial. With additional nights ofpartial sleep deprivation, cumulative sleepinessmight rapidly become a significant problem. Theimportance of sleep restriction, daytime sleepi-ness, and performance of children in school andon their behavior may be greater than previouslyrealized.

In 1972, Dement98 powerfully described possi-ble outcomes of restricted sleep on wakefulness:

“After an excessively long period of wakefulness, thestate of sleep becomes preemptive. When we enforcewakefulness, we are probably preventing or minimiz-ing activity in the neural systems that subserve sleepinduction and maintenance. As the potency of thesesystems increase [sic] during the period of theirinduced inactivity, they may begin to intrude uponwakefulness in an ever more aggressive manner.

“The notion of total sleep deprivation could besomewhat illusory, and could result merely in a redis-tribution of activity in sleep and arousal systems inwhich NREM sleep would occur in the form of hun-dreds of microsleeps.”

Microsleep episodes have been well docu-mented in both human and animal studies.99-101

Kales and colleagues have described disorienta-tion and misperceptions during sleep depriva-tion that seemed to be associated with “lapses”that become more frequent as deprivation con-tinues.90 Armington and Mitnick found thatsleep deprivation eventually produced brainwave patterns that were more or less continu-ously at the NREM stage 1 level in subjects who


Ch01.qxd 1/25/05 3:05 AM Page 9

appeared to be behaviorally awake.99 The mostconsistent result of modest amounts of sleep lossin humans is the occurrence of these microsleepperiods100 which increase in frequency through-out periods of sleep deprivation. Since grosswaking behavior is not affected, these lapses mayhave significant consequences on performanceand its assessment, especially for the school-aged child whose consistent attentiveness isrequired for success in school. Since the percep-tual shutdown can occur before EEG changes areapparent at the outset of sleep under ordinarycircumstances, there may be many more suchepisodes in sleep-deprived subjects than EEGpatterns alone would suggest.98

PHYLOGENETIC CONSIDERATIONS

Understanding sleep in humans requires somereflection on sleep in other species. Althoughperiods of sluggish activity can be documentedin reptiles, it does not appear that physiologicsleep occurs.102 Although only a relatively smallnumber of mammalian species have been stud-ied, it appears that most, if not all, birds andmammals sleep. Quiescent periods, intervals ofreduced responsiveness to environmental stim-uli, rapid reversibility of state, specific postures,and characteristic EEG changes have beenobserved.103 However, all of these criteria neednot be present concomitantly, and quiescence isnot always equivalent to inactivity. Ritualisticpresleep activity and behaviors occur in manyspecies, including humans. The timing of sleepvaries: Some species consolidate sleep into a sin-gle period of time, while others distribute sleepthroughout a 24-hour continuum.

Sleep in birds is remarkably similar to sleep inmammals. Both demonstrate two distinct typesof sleep, with comparable electrophysiologicactivity. Major differences appear to be in thepattern of sleep and the greater number of sleepstates observed in avian species.104

Zepelin and Rechtschaffen studied availablesleep data on more than 50 mammalianspecies.28 Sleeping patterns were correlated withmetabolic rates, gestational periods, and brainweights. Animals with lower metabolic ratestend to sleep less than those with higher meta-bolic rates. Species that have longer sleep peri-ods tend to exhibit shorter life spans and aresmaller in size. Meddis replicated this study in a

sample of 65 species and obtained similarresults.105

NREM-REM cycling appears to be the basicorganization of sleep in most species studied.Although the quality and quantity of NREM andREM sleep vary considerably, a regularly pat-terned series of state changes occurs, withdemonstrable slowing of the EEG and the pres-ence of spindling activity.103 Paradoxical sleephas been recorded in almost all mammalianspecies studied. Characteristics of this sleep stateinclude desynchronization (activation) of CNSelectrical activity, skeletal muscle atonia, peri-odic twitching, and physiologic instability (espe-cially of the cardiovascular and respiratorysystems). Changes in thermoregulation and higharousal thresholds are present. Rhythmic thetaactivity and PGO spikes are typically seen onEEG during mammalian paradoxical sleep.

Phylogenetic development of REM sleep hasbeen studied by Allison and Van Twyver.106 Itappears to have developed approximately 130million years ago. Allison and Cicchetti con-cluded that the volume of REM sleep correlatedwith lifestyle, risk of predation, and degree ofexposure of the sleeping environment.25

REM sleep is the preponderant state early inlife in most mammals (including humans).Though considered to be ontogenetically primi-tive, the role of REM sleep in the development ofthe CNS may be significant. Premature newbornhumans spend approximately 90% of their totalsleep time in active sleep. This falls rapidly toabout 50% by term. A gradual decrease contin-ues throughout the first few years of life to alevel of about 20% to 25%. This level remainsremarkably constant throughout the remainderof the life cycle.65

Jouvet-Mounier and associates have studiedthe ontogenetic development of sleep in infantcats, rats, and guinea pigs.107 More than 70 ani-mals underwent electrocortical, electro-ocular,electromyographic, and behavioral monitoringfrom birth to 50 days of age. REM sleep was thepreponderant form of sleep in all of them. Eachspecies varied significantly in the degree ofdevelopment at birth, with rat pups the mostimmature, kittens intermediate, and guinea pigsthe most mature. Degree of immaturity at birthhighly correlated with the amount of paradoxi-cal sleep recorded during the perinatal period.Rat pups exhibited 70% paradoxical sleep at


Ch01.qxd 1/25/05 3:05 AM Page 10

birth, which decreased rapidly almost to adultlevels by 30 days of life. A decrease of paradoxi-cal sleep in kittens was considerably slower.Guinea pigs showed the lowest volume of para-doxical sleep (7%); however, this was stillapproximately double the volume seen in theadult animal. Maturation of SWS is late in com-parison with paradoxical sleep, and the timespent in paradoxical sleep and SWS varies dur-ing the first postnatal month. These variationsare different among species. Newborn kittenshave a more highly developed cortex than ratpups.108 Cortical neurons mature very rapidlyand reach the histologic characteristics of adultcortical neurons by the 12th postnatal day, con-comitant with the appearance of SWS. In contrast,the cortex of the newborn guinea pig appears his-tologically the same as that of the adult.109

Sleeping dolphins and porpoises are fascinat-ing and of particular ontogenetic interestbecause of the complexity of the Cetacean CNS.Mukhametov has studied the neurophysiologyof sleep in the bottle-nosed dolphin (Tursiopstruncatus) and the porpoise (Phocoena pho-coena)110 and showed that the main characteris-tics of sleep in these marine mammals areunihemispheric SWS and the apparent absenceof paradoxical sleep. EEG characteristics weresimilar to those typical for the mammalian brain,and three distinct stages can be identified:desynchronization; intermediate synchroniza-tion, with sleep spindles, theta activity, and deltawaves; and maximal synchronization, with slowwaves constituting more than 50% of eachrecording period. In all dolphins studied, uni-hemispheric SWS was the main type of sleeprecorded. Interestingly, this type of sleep is notfound in other mammals. Synchronization of theEEG occurs in one hemisphere, while the oppo-site hemisphere reveals desynchronization.These cycles of synchrony and desynchronyappear to be independent. Each hemisphereexhibits different amounts of SWS, and depriva-tion of SWS in one hemisphere does not result incontralateral rebound. Ipsilateral rebound isnoted in the SWS-deprived hemisphere only.Mukhametov attempted to identify neurophysio-logic and behavioral correlates of paradoxicalsleeping in 30 animals of two species and con-cluded that paradoxical sleep does not appear tobe present in the dolphin or porpoise. However,it is very difficult to prove the complete absence

of paradoxical sleep, since testing of dolphinfetuses and calves has not yet been possible. It isunknown whether these characteristics representa phylogenetic, developmental, or adaptive phe-nomenon. Unraveling the mystery has fascinat-ing teleologic implications.

BEHAVIORAL AND PHYSIOLOGICCONSIDERATIONS

At first glance, sleep appears to be a simpleprocess, a required part of our 24-hour life cycle.Little attention is paid to the sleeping state,because human lifestyles focus primarily oninteractions with the environment and dailyfragments of disengagement seem of secondaryimportance. Time spent in activities not relatedto goal attainment, pursuit of sustenance, fulfill-ment, happiness, or success appears to consist ofintrusive, unwelcome gaps. However, the impor-tance of these gaps in permitting the individualto function and appropriately interact with theenvironment during the waking state has onlyrecently been discovered.

Any definition of sleep is complex, both frombehavioral and physiologic perspectives. In thesimplest terms, it is a reversible disengagementwith and unresponsiveness to the external envi-ronment, regularly alternating in a circadianmanner with engagement and responsiveness. Itis now known that this definition is significantlyincomplete and simplistic, since sleep is a highlyactive and complex state.

It seems easy to determine when an individ-ual is sleeping. Behavioral correlates include arecumbent position, closure of the eyelids, qui-escence, and diminished responsiveness toexternal stimuli. These behaviors are fairly con-sistent among individuals. However, sleep onsetrequires complex interactions of learned behav-iors and physiologic processes. Absence of sud-den external stimuli; a suitable, safe, comfortableenvironment; relaxation of postural muscles;and learned stereotypical behaviors associatedwith bedtime are required.111 There is some evi-dence that rhythmic, monotonous sensory stim-ulation helps promote sleep.112 Whether this isbehavioral, physiologic, or a combination isspeculative.

Physiologically, sleep onset and maintenanceare not passive processes. Isolation of the cere-brum from the brainstem and spinal cord


Ch01.qxd 1/25/05 3:05 AM Page 11

(cerveau isolé) produces a state indistinguishablefrom physiologic sleep.113 A series of exquisiteexperiments identified neurons of the reticularformation that received collateral input fromsomatic, visceral, and special sensory pathwaysand sent ascending projections dorsally and ven-trally to the basal forebrain.114-116 These collec-tions of neurons constitute what is called thereticular activating system. Complex projectionsof neurons from the reticular formation to theposterior hypothalamus-subthalamus, to the basalforebrain, and then to the cortex are responsiblefor the maintenance of wakefulness.117

Although it was initially thought that sleepwas the result of a decrease in the activity of thissystem, brainstem transection experimentsresulting in diminished sleep suggested thatsleep inducing structures must also be present inthe CNS.118 This sleep inducing structureappeared to be located in the lower brainstem,specifically the dorsal medullary reticular forma-tion and nucleus of the solitary tract. Lesions inthis area produced EEG activation in a sleepinganimal.119,120 A sleep facilitation center appearsto be present in the rostral hypothalamus,121 andcortical synchrony can be elicited by stimulationof the midline thalamus.122 Sleep-inducing neu-rons are also found in the preoptic area and basalforebrain. Gamma-aminobutyric acid neuronslocated in the cortex, as well as neurons locatedin the hypothalamus and basal forebrain, arevital for slow-wave production.117

Therefore, sleep onset results from a complexseries of events involving changes in levels ofsomatic, visceral, and special sensory input;active inhibition of neuronal networks that pro-duce cortical desynchronization; and activestimulation of neuronal systems and pathwaysresponsible for cortical synchrony. In addition,the rhythmic organization of these activities isextremely complex and appears to be controlledby neurons located in the suprachiasmaticnucleus.123 Jouvet described a separate system ofneurons located in the upper pons that con-trolled the induction and manifestations of REMsleep.124 This system was under the influence ofan “oscillator” that was separate (though linkedto) that which controlled the rhythmicity of thesleep–wake cycle. According to Hobson, thecholinergic, “REM-on” system of neurons islocated primarily in the mesencephalic, medullary,and pontine gigantocellular tegmental fields but

may be widespread.125 Discharges from theseneurons are responsible for REM sleep epiphe-nomena of cortical desynchronization, conjugateeye movements, a decrease in muscle tone byactive inhibition of alpha motor neurons, muscu-lar twitching, and cardiorespiratory irregularities.It has also been shown that a self-inhibitory,aminergic, “REM-off” system of neurons, locatedin the dorsal raphe nuclei, locus coeruleus, andthe nucleus parabrachialis lateralis, interacts withthe opposing system, resulting in alternationsbetween NREM and REM sleep.

It is clear that the behavioral, neurochemical,and neurophysiologic mechanisms of thesleep–wake cycle and electrophysiologic cyclesduring sleep itself are complex and intenselyintegrated. Not all of sleep’s characteristics havebeen elucidated. Further research is needed.Their biological substrate may prove to be sim-ple or reflect extraordinarily complex physio-logic processes. The implications for fetal andchildhood development may be more significantthan our wildest dreams.

References

1. Rechtschaffen A: The function of sleep:Methodological issues. In Drucker-Colin R,Shkurovich M, Sterman MB (eds): The Functionsof Sleep. New York, Academic Press, 1979.

2. Moldofsky H: Immunology and sleep. Paper pre-sented at the Association of Professional SleepSocieties, First Annual Meeting, Columbus, Ohio,June 15-20, 1986.

3. Moldofsky H, Lue FA, Davidson JR, GorczynskiR: The effect of 40 hours of wakefulness onimmune functions in humans: II. Interleukins-1-and -2-like activities. Sleep Res 1988;17:34.

3a. Irwin M: Effects of sleep and sleep loss on immu-nity and cytokines. Brain Behav Immun2002;16:503.

4. Moldofsky H, Lue FA, Davidson JR, GorczynskiR: Disordered diurnal patterns of immune func-tions in three patients with narcolepsy-cataplexy.Sleep Res 1988;17:223.

5. Sherington CS: Man on His Nature. Cambridge,Cambridge University Press, 1946, p 413.

6. Oswald I: Sleep. Harmondsworth, Middlesex,Penguin Books, 1974.

7. Hartmann E, Orzack MH, Branconnier R: Deficitsproduced by sleep deprivation: Reversal by d- andl-amphetamine. Sleep Res 1974;3:151.

8. Adam K, Oswald I: Sleep is for tissue restoration.J R Coll Physicians Lond 1977;11:376.


Ch01.qxd 1/25/05 3:05 AM Page 12

8a. Takahashi M, Arito H: Maintenance of alertnessand performance by a brief nap after lunch underprior sleep deficit. Sleep 2000;23:813.

9. Berger RJ, Oswald I: Effects of sleep deprivationon behaviour, subsequent sleep, and dreaming.J Ment Sci 1962;108:457.

10. Feinberg I: The ontogenesis of human sleep andthe relationship of sleep variables to intellectualfunction in the aged. Comp Psychiatry1968;9:138.

11. Adam K: Body weight correlates with REM sleep.Br Med J 1977;1:813.

12. Sassin JF, Parker DC, Mace JW, et al: Humangrowth hormone release relation to slow-wavesleep and sleep-waking cycles. Science1969;165:513.

13. Sassin JF, Frantz AG, Kapen S, et al: The noctur-nal rise of human prolactin is dependent on sleep.J Clin Endocrinol Metab 1973;37:436.

14. Boyar RM, Rosenfeld RS, Kapen S, et al: Humanpuberty: Simultaneous augmented secretion ofluteinizing hormone and testosterone duringsleep. J Clin Invest 1974;54:609.

15. Weitzman ED, Hellman L: Temporal organizationof the 24-hour pattern of the hypothalamic-pituitary axis. In Ferin M, Halberg F, Richart RM,Vanderviele RL (eds): Biorhythms and HumanReproduction. New York, John Wiley, 1974.

16. Valk IM, van der Bosch JSG: Intra-daily variationof the human ulnar length and short termgrowth—a longitudinal study in eleven boys.Growth 1974;42:107.

17. Griffin SJ, Trinder J: Physical fitness, exercise, andhuman sleep. Psychophysiology 1978;15:447.

18. Fisher LB: The diurnal mitotic rhythm in thehuman epidermis. Br J Dermatol 1968;80:75.

19. Finkelstein JW, Roffwarg HP, Boyar RM, et al:Age-related change in the twenty-four hour spon-taneous secretion of growth hormone. J ClinEndocrinol Metab 1972;35:665.

20. Jouvet M: The function of dreaming: A neuro-physiologist’s point of view. In Gazzaniga MS,Blakemore C (eds): Handbook of Psychobiology.New York, Academic Press, 1975.

21. Giuditta A, Neugebauer-Vitale A, Grassi-ZucconiG, et al: Synthesis of brain RNA and DNA duringsleep. In Borbely A, Valtax JL (eds): SleepMechanisms. Berlin, Springer-Verlag, 1984.

22. Bert J: Sleep in primates under natural conditionsand in the laboratory. In Koella WP, Levin P (eds):Sleep 1976: Third European Congress of SleepResearch. Basel, S Karger, 1977.

23. Webb WB: Sleep stage responses of older andyounger subjects after sleep deprivation.Electroencephalogr Clin Neurophysiol 1981;52:368.

24. Kovalzon VM: Brain Temperature Variations inECoG in Free-Swimming Bottle-Nose Dolphins:Sleep 1976: Third European Congress of SleepResearch. Basel, S Karger, 1977.

25. Allison T, Cicchetti DV: Sleep in mammals:Ecological and constitutional correlates. Science1976;194:732.

26. McGinty DJ, Harper TM, Fairbanks MK:Neuronal unit activity and the control of sleepstates. In Weitzman E (ed): Advances in SleepResearch. New York, Spectrum, 1974.

27. Snyder F: Toward an evolutionary theory ofdreaming. Am J Psychiatry 1966;123:121.

28. Zepelin H, Rechtschaffen A: Mammalian sleep,longevity, and energy conservation. Brain BehavEvol 1974;10:425.

29. Carpenter AC, Timiras PS: Sleep organization inhypo- and hyperthyroid rats. Neuroendocrinology1982;34:438.

30. Eastman CI, Rechtschaffen A: Effects of thyroxineon sleep in the rat. Sleep 1979;2:215.

31. Block V, Hennevin E, Leconte P: The phenome-non of paradoxical sleep augmentation afterlearning: Experimental studies of its characteris-tics and significance. In Fishbein W (ed): Sleep,Dreams and Memory. Jamaica, NY, Spectrum,1981.

32. Smith C, Butler S: Paradoxical sleep at selectivetimes following training is necessary for learning.Physiol Behav 1982;29:469.

33. Jenkins J, Dallenbach K: Oblivescence duringsleep and waking. Am J Psychol 1924;35:605.

34. VanOrmer EG: Retention after intervals of sleepand waking. Arch Psychol 1932;137:5.

35. Staunton H: The function of dreaming. RevNeurosci 2001;12:365.

36. Wetzel W, Wagner T, Balschun D: REM sleepenhancement induced by different proceduresimproves memory retention in rats. Eur JNeurosci 2003;18:2611.

37. Maquet P, Laureys S, Peigneux P., et al:Experience-dependent changes in cerebral activa-tion during human sleep. Nat Neurosci2000;3:831.

38. Peigneux P, Laureys S, Fuchs S, et al: Learnedmaterial content and acquisition level modulatecerebral reactivation during post-training rapid-eye-movement sleep. Neuroimage 2003;20:125.

39. Mednick S, Nakayama K, Stickgold R: Sleep-dependent learning: A nap is as good as a night.Nat Neurosci 2003;6:697.

40. Bednar JA, Mikkulainen R: Learning innate facepreferences. Neural Comput 2003;15:1525.

41. Davis CJ, Harding JW, Wright JW: REM sleep dep-rivation-induced deficits in the latency-to-peakinduction and maintenance of long-term


Ch01.qxd 1/25/05 3:05 AM Page 13

potentiation within the CA1 region of the hip-pocampus. Brain Res 2003;973:293.

42. Smith C: Sleep states and memory processes inhumans: Procedural versus declarative memorysystems. Sleep Med Rev 2001;5:491.

43. Walker MP, Liston C, Hobson JA, Stickgold R:Cognitive flexibility across the sleep-wake cycle:REM-sleep enhancement of anagram problemsolving. Brain Res Cogn Brain Res 2002;14:317.

44. Gais S, Molle M, Helms K, Born J: Learning-dependent increases in sleep spindle density.J Neurosci 2002;22:6830.

45. Kisley MA, Polk SD, Ross RG, et al: Early postna-tal development of sensory gating. Neuroreport2003;14:693.

46. Carter S, Gold A: The syndrome of minimal cere-bral dysfunction. In Barnett MH, Einhorn A (eds):Pediatrics. New York, Appleton-Century-Crofts,1972.

47. Dykman RA, Ackerman PT, Clements S, et al:Specific learning disabilities: An attentionaldeficit syndrome. In Myklebust HR (ed): Progressin Learning Disabilities. New York, Grune andStratton, 1971.

48. Levinson HN: Dyslexia: A Solution to the Riddle.New York, Springer-Verlag, 1980.

49. Ottenbacher K, Abbott C, Haley D, et al: Humanfigure drawing ability and vestibular processing inlearning-disabled children. J Clin Psychol1984;40:1084.

50. deQuiros JB, Schrager OL: NeuropsychologicalFundamentals in Learning Disabilities. SanRafael, Calif, Academic Therapy Press, 1978.

51. Steinberg M, Rendle-Short J: Vestibular dysfunc-tion in young children with minor neurologicalimpairment. Dev Med Child Neurol 1977;19:639.

52. Ottenbacher K, Watson PJ, Short MA, et al:Nystagmus and ocular fixation difficulties inlearning-disabled children. Am J Occup Ther1979;33:717.

53. Brown B, Haegerstrom-Portnoy G, Yingling CD,et al: Dyslexic children have normal vestibularresponses to rotation. Arch Neurol 1983;40:370.

54. Sheldon SH, Spire JP, Levy HB: REM sleep eyemovements in reading disabled children. SleepRes 1990;19:34.

55. Pompeiano O: The neurophysiological mechanismsof the postural and motor events during desyn-chronized sleep. In Kety SS, Evarts EV, Williams HL(eds): Sleep and Altered States of Consciousness.Baltimore, Williams & Wilkins, 1967.

56. Pompeiano O: Mechanisms of sensorimotor integra-tion during sleep. Prog Physiol Psychol 1970;3:1.

57. Reding GR, Fernandez C: Effects of vestibularstimulation during sleep. ElectroencephalogrClin Neurophysiol 1968;24:75.

58. Appenzeller O, Fischer AP Jr: Disturbances ofrapid eye movements during sleep in patients withlesions of the nervous system. ElectroencephalogrClin Neurophysiol 1968; 25:29.

59. Ornitz EM: Development of sleep patterns inautistic children. In Clemente CD, Purpura DP,Mayer FE (eds): Sleep and the Maturing NervousSystem. New York, Academic Press, 1972.

60. Busby K, Firestone P, Pivik RT: Sleep patterns inhyperkinetic and normal children. Sleep1981;4:366.

61. Lucero MA: Lengthening of REM sleep durationconsecutive to learning in the rat. Brain Res1979;20:319.

62. Scrima L: Isolated REM sleep facilitates recall ofcomplex associative information. Psychophysi-ology 1982;19:252.

63. Simon CW, Emmons WH: Responses to materialpresented during various levels of sleep. J ExpPsychol 1956;51:89.

64. Parmelee AH Jr, Schulz HR, Disbrow MA: Sleeppatterns of the newborn. J Pediatr 1961;58:241.

65. Roffwarg HP, Dement WC, Fisher C: Preliminaryobservations on the sleep patterns in neonates,infants, children, and adults. In Harms E (ed):Problems of Sleep and Dreams in Children.London, Pergamon Press, 1963.

66. Fishbein W: Disruptive effects of rapid eye move-ment sleep deprivation on long-term memory.Physiol Behav 1971;6:279.

67. Fishbein W, Kastaniotis C, Chattman D:Paradoxical sleep: Prolonged augmentation fol-lowing learning. Brain Res 1974;71:61.

68. McGrath MJ, Cohen DB: REM sleep facilitation ofadaptive waking behavior: A review of the litera-ture. Psychol Bull 1978;85:24.

69. Hawkins MR, Vichick DA, Silsby HD, et al: Sleepand nutritional deprivation and performance ofhouse officers. J Med Educ 1985;60:530.

70. Paul K, Dittrichova J: Sleep patterns followinglearning in infants. In Levin P, Koella U (eds):Sleep 1974. Basel, S Karger, 1975, p 388.

71. Parkes JD: Sleep and Its Disorders. London, WBSaunders, 1985.

72. Hartman E: The Functions of Sleep. New Haven,Yale University Press, 1976.

73. Ron S, Algom D, Hary D, et al: Time-relatedchanges in the distribution of sleep stages in braininjured patients. Electroencephalogr ClinNeurophysiol 1980;48:432.

74. Greenberg R, Dewan EM: Aphasia and rapid eyemovement sleep. Nature 1969;223:183.

75. Feinberg I: Eye movement activity during sleepand intellectual function in mental retardation.Science 1968;159:1256.


Ch01.qxd 1/25/05 3:05 AM Page 14

76. Oksenberg A, Marks J, Farberk K, et al: Effect ofREM sleep deprivation during the critical periodof neuroanatomical development of the cat visualsystem. Sleep Res 1986;15:53.

77. Squire LR: Memory and the Brain. New York,Oxford University Press, 1987.

78. Kohonen T: Associative Memory. New York,Springer-Verlag, 1977.

79. Palm G: Neural Assemblies: An AlternativeApproach to Artificial Intelligence. New York,Springer-Verlag, 1982.

80. Crick F, Mitchison G: The function of dreamsleep. Nature 1983;304:111.

81. Hopfield JJ, Feinstein DI, Palmer RG: ‘Unlearning’has a stabilizing effect in collective memories.Nature 1983;304:158.

82. Kleitman N: Sleep and Wakefulness. Chicago,University of Chicago Press, 1963.

83. Webb WB: The sleep of conjoined twins. Sleep1978;1:205.

84. Moldofsky H, Scarisbrick P: Induction of neuras-thenic musculoskeletal pain syndrome by selec-tive sleep stage deprivation. Psychosom Med1976;38:35.

85. Nicholson AN, Marks J: Insomnia. Lancaster,MTP, 1983.

86. Marsden CD: ‘On-off’ phenomenon in Parkinson’sdisease. In Rinne UK, Klinger M, Stamm G (eds):Parkinson’s Disease—Current Progress, Problemsand Management. Amsterdam, Elsevier/NorthHolland, 1980.

87. Segawa M, Hosaka A, Miyagawa F, et al:Hereditary progressive dystonia with markeddiurnal fluctuation. Adv Neurol 1976;14:215.

88. Rechtschaffen A, Gilliland MA, Bergmann BM,et al: Physiological correlates of prolonged sleepdeprivation in rats. Science 1983;221:182.

89. Passouant P, Popoviciu L, Velok G, et al:[Polygraphic study of narcolepsy during thenycthemeral period.] Rev Neurol (Paris)1968;118:431.

90. Kales A, Tan TL, Kollar EG, et al: Sleep patternsfollowing 205 hours of sleep deprivation.Psychosom Med 1970;32:189.

91. Glenville M, Broughton R, Wing AM, et al: Effectsof sleep deprivation on short duration perform-ance measures compared to the Wilkinson audi-tory vigilance task. Sleep 1978;1:169.

92. Horne JA, Anderson NR, Wilkinson RT: Effects ofsleep deprivation on signal detection measures ofvigilance: Implications for sleep function. Sleep1983;6:347.

93. Williams HL, Hammack JT, Daly RL, et al:Responses to auditory stimulation, sleep loss andthe EEG stages of sleep. Electroencephalogr ClinNeurophysiol 1964;16:269.

94. Webb WB, Agnew HW Jr: The effect of a chroniclimitation of sleep length. Psychophysiology1974;11:265.

95. Dement W, Greenberg S: Changes in totalamount of stage four sleep as a function of par-tial sleep deprivation. Electroencephalogr ClinNeurophysiol 1966;20:523.

96. Carskadon MA, Harvey K, Dement WC: Acuterestriction of nocturnal sleep in children.Percept Motor Skills 1981;53:103.

97. Wilkinson RT: Sleep Deprivation: Performancetests for partial and selective sleep deprivation.In Apt LE, Riess BF (eds): Progress in ClinicalPsychology, vol 8. New York, Grune & Stratton,1968, pp 28-33.

98. Dement WC: Sleep deprivation and organizationof the behavioral states. In Clemente C, PurpuraD, Mayer F (eds): Sleep and the Maturing NervousSystem. New York, Academic Press, 1972.

99. Armington JC, Mitnick LL: Electroencephalo-gram and sleep deprivation. J Appl Physiol1959;14:247.

100. Williams H, Lubin A, Goodnow J: Impaired per-formance with acute sleep loss. Psychol Monog1959;73:1.

101. Friedman L, Bergmann BM, Rechtschaffen A:Effects of sleep deprivation on sleepiness, sleepintensity, and subsequent sleep in the rat. Sleep1979;1:369.

102. Cartwright RD: A Primer on Sleep andDreaming. Reading, MA, Addison-Wesley, 1978.

103. Zepelin H: Mammalian sleep. In Kryger MH, RothT, Dement WC (eds): Principles and Practice ofSleep Medicine (3rd ed), Philadelphia, WBSaunders Company, 2000, pp 82-92.

104. Tobler I: Phylogeny of sleep regulation. In KrygerMH, Roth T, Dement WC (eds): Principles andPractice of Sleep Medicine (3rd ed), Philadelphia,WB Saunders Company, 2000, pp 72-81.

105. Meddis R: The evolution of sleep. In Mayes A(ed): Sleep Mechanisms and Function inHumans and Animals: An EvolutionaryPerspective. Berkshire, England, Van NostrandReinhold (UK), 1983.

106. Allison T, Van Twyver H: The evolution of sleep.Nat Hist 1970;79:56.

107. Jouvet-Mounier D, Astic L, Lacote D:Ontogenesis of the states of sleep in rat, cat, andguinea pig during the first postnatal month. DevPsychobiol 1970;2:216.

108. Noback CR, Purpura DP: Postnatal ontogenesisof neurons in cat neocortex. J Comp Neurol1961;117:291.

109. Peters HG, Bademan H: The form and growth ofstellate cells in the cortex of the guinea pig.J Anat 1963;97:111-117.


Ch01.qxd 1/25/05 3:05 AM Page 15

110. Mukhametov LM: Sleep in marine mammals.Exp Brain Res 1984;8:227.

111. Konorski J: Integrative Action of the Brain.Chicago, University of Chicago Press, 1967.

112. Gastaut H, Bert B: Electroencephalographicdetection of sleep induced by repetitive sensorystimuli. In Wolstenholme GEW, O’Connor M(eds): On the Nature of Sleep. London,Churchill, 1961, p 260.

113. Bremer F: Quelques proprietes de l’activite elec-trique du cortex cerebral “isole.” CR Soc Biol(Paris) 1935;118:1241.

114. French JD, Magoun HW: Effects of chroniclesions in central cephalic brain stem of mon-keys. Arch Neurol Psychiatry 1952;69:591.

115. Lindsley DB, Bowden JW, Magoun HW: Effectupon the EEG of acute injury to the brain stemactivating system. Electroencephalogr ClinNeurophysiol 1949;1:475.

116. Moruzzi G: The sleep-waking cycle. ErgebPhysiol 1972;64:1.

117. Jones BE: Basic mechanisms of sleep-wakestates. In Kryger MH, Roth T, Dement WC:Principles and Practice of Sleep Medicine.Philadelphia, WB Saunders, 1989, p 121.

118. Barini C, et al: Effects of complete pontine tran-sections of the sleep-wakefulness rhythm: Themidpontine pretrigeminal preparation. Arch ItalBiol 1959;97:1.

119. Freemon FR, Salinas-Garcia RF, Ward JW: Sleeppatterns in a patient with a brain stem infarctioninvolving the raphe nucleus. ElectroencephalogrClin Neurophysiol 1974;36:657.

120. Westmoreland BF, Klass DW, Sharbrough FW,et al: Alpha-coma. Electroencephalographic,clinical, pathologic, and etiologic correlations.Arch Neurol 1975;32:713.

121. Nauta WJH: Hypothalamic regulation of sleep inrats: An experimental study. J Neurophysiol1946;9:285.

122. Morison RS, Dempsey EW: A study of thalamo-cortical relations J Physiol 1942;135:281.

123. Hanada Y, Kawamura H: Sleep-waking electro-corticographic rhythms in chronic cerveau isolérats. Physiol Behav 1981;26:725.

124. Jouvet M: Paradoxical sleep: A study of itsnature and mechanisms. In Himwich WA,Schade JP (eds): Sleep Mechanisms: Progress inBrain Research. Amsterdam, Elsevier, 1965.

125. Hobson JA: The cellular basis of sleep cycle con-trol. Adv Sleep Res 1974;1:217.


Ch01.qxd 1/25/05 3:05 AM Page 16

Children do not usually complain of sleep prob-lems. An adult, most often a parent or anothercaregiver, usually initiates the medical evalua-tion. The major presentations of sleep disordersare insomnia, hypersomnia, and abnormalbehaviors during sleep. This chapter will reviewthe differential diagnosis of sleep disorders inchildren, using as a framework the InternationalClassification of Sleep Disorders.1

Insomnia, hypersomnia, and abnormal move-ments are not simply features of sleep disordersbut may provide a clue that an unsuspectedmedical problem is present. For example, insom-nia may be an important clue to the presenceof pinworm2 or iron deficiency3. Hypersomniacan be a clue that the tonsils and adenoids areenlarged,4-6 which may be the cause of cardiacfailure.7

What Is the Problem? Whose Problem Is It?Children with sleep disorders seldom bring theirdisorder to the attention of clinicians directly.Sleep disorders in children are usually suspectedbecause of an observation (for example, of snoring,apnea, behavioral changes) by others. Therefore,the evaluation of a child’s sleep must include aninterview with the caregiver and a review of thesymptoms and behaviors as they affect the patientand the family unit.

Frequently, there is no medical problem at all,but the expectations of the parent or caregiverare inappropriate. For example, a sleepless cou-ple might bring in their 3-month-old child won-dering why the child does not sleep through thenight and explaining that this is interfering withtheir own nocturnal sleep. The problem here is

that the parents are not aware of what is normal.Thus, clinicians must know what is normal andwhat is expected, and they must transmit thisto the parents. What is normal in the child canresult in a problem for the parents when theirwork quality suffers as a result of their baby wak-ing up at night.

At times, sleeplessness in a child may be theresult of problems in the family (a severe illnessin a parent, or marital problems). Thus, in eval-uating a child for a sleep disorder, the clinicianmust ascertain the following: What is the sleepproblem? Whose problem is it? What is the causeof the problem?

Children present to clinicians with threetypes of sleep-related problems. First are theinsomnias—disorders of initiating and main-taining sleep; second are the hypersomnias—disorders in which the child sleeps at the wrongtime and the wrong place and has excessivesleepiness; and third is abnormal activity orbehaviors during sleep. A child may have symp-toms that fall into more than one category. Thischapter focuses on the most commonly seendisorders.

INSOMNIA

Insomnia is a symptom that reflects a perception,by the patient or the caregiver, that it takes toolong to fall asleep or that it is difficult to maintainsleep. Terms such as sleeplessness, insomnia, andrestlessness may be the actual words used. Differentpeople may notice different clinical features. Forexample, the parent of a child may comment onthe child’s nocturnal difficulties, whereas theschool teacher might notice that the sleep-deprived child falls asleep in class or demon-strates features of hyperactivity.

Children can have disorders in any of themajor categories listed in Table 2–1. The clinician

Differential Diagnosis of Pediatric Sleep Disorders

Meir H. Kryger

2

17

This work was supported by National Institutes of Healthgrant R01 HL63342-01A1.

Ch02.qxd 2/12/05 1:03 PM Page 17

18 Chapter 2 Differential Diagnosis of Pediatric Sleep Disorders

ICD-9-CM ICD-10-CM Sleep Disorder (Alternate or Other Commonly Used Name)Insomnia: Problems Falling Asleep and Staying Asleep*307.42 F51.01 Psychophysiological Insomnia (conditioned insomnia)307.42 F51.02 Paradoxical Insomnia (sleep state misperception)307.41 F51.03 Adjustment Sleep Disorder (transient insomnia)307.41 F51.04 Inadequate Sleep Hygiene780.52 G47.01 Idiopathic Insomnia (childhood onset insomnia)307.42 F51.05 Behavioral Insomnia of Childhood (sleep onset association disorder)780.52 G47.09 Insomnia due to other known condition291.89 F10-19 Insomnia due to substance abuse995.2 T36-50 Insomnia due to adverse drug effect307.42 F51.09 Insomnia due to psychiatric/behavioral condition

Sleep-Related Breathing Disorder: EDS and/or Restless Sleep†

780.51 G47.31 Primary Central Sleep Apnea770.81 P28.3 Primary Sleep Apnea of Infancy (infant sleep apnea)780.53 G47.32 Obstructive Sleep Apnea, Pediatric780.53 G47.0 Sleep-Related Breathing Disorder due to cardiorespiratory disease780.57 G47.34 Sleep Related Non-obstructive Alveolar Hypoventilation (central

alveolar hypoventilation)780.57 G47.35 Congenital Central Alveolar Hypoventilation Syndrome

Hypersomnia Not Due to a Sleep-Related Breathing Disorder: EDS‡

347.01 G47.41 Narcolepsy with Cataplexy347.00 G47.42 Narcolepsy without Cataplexy780.54 G47.11 Recurrent Hypersomnia (Kleine-Levin syndrome)780.54 G47.12 Idiopathic Hypersomnia with long sleep time780.54 G47.13 Idiopathic Hypersomnia without long sleep time307.44 F51.11 Behaviorally-Induced Insufficient Sleep Syndrome995.2 F10-19 Other Hypersomnia due to substance abuse995.2 T36-50 Other Hypersomnia due to the adverse effects of a drug307.44 F51.19 Hypersomnia due to psychiatric/behavioral condition

Circadian Rhythm Sleep Disorder: EDS and/or Problems Falling Asleep and Staying Asleep§

PRIMARY

780.55 G47.21 Delayed sleep phase type (delayed sleep phase syndrome)780 55 G47.22 Advanced sleep phase type (advanced sleep phase syndrome)780.55 G47.23 Irregular sleep-wake type780.55 G47.24 Non-entrained type780.55 G47.20 Other Primary Circadian Rhythm Sleep Disorder due to a known

physiological conditionBEHAVIORALLY INDUCED

307.45 F51.21 Jet-lag type307.45 F51.22 Shift-work type291.89 F10-19 Circadian Rhythm Disorders due to substance abuse995.2 T36-50 Circadian Rhythm Sleep Disorders due to adverse drug effect

Parasomnia: History of Abnormal Sleep-Related Behavior¶¶

780.56 G47.51 Confusional Arousals307.46 F51.3 Sleepwalking307.46 F51.4 Sleep Terrors780.56 G47.52 REM Sleep Behavior Disorder, including Parasomnia Overlap Disorder780 56 G47.53 Recurrent Isolated Sleep Paralysis307.47 F51.5 Nightmare Disorder

Table 2–1. Sleep Disorders in Children

Ch02.qxd 2/12/05 1:03 PM Page 18

decides which ones are most likely to be encoun-tered and pursues a diagnosis.

Behavioral and PsychophysiologicDisordersSome behavioral and psychophysiologic disor-ders occur in children but not in adults. Inlimit-setting sleep disorders, for example, care-givers may complain that children make persist-ent “curtain calls” or refuse to go to sleep. Insleep-onset association disorders, caregiversmay complain that the child cries during sleepunless the adult rocks the child or sleeps withthe child. The children are either too young toarticulate the problem or do not see the behav-ior as a problem.

Psychiatric DisordersChildren can have almost any of the major psy-chiatric disorders. Insomnia may be caused bythe psychiatric disorder itself 8 or by treatmentfor the disorder.9 For example, antidepressantsmay lead to difficulty in falling asleep or toincreased movements during sleep (see “DrugUse and Abuse,” later).

Although psychiatric disorders may indeedcause sleep problems, equally important is thepotential misdiagnosis of a sleep disorder as apsychiatric disorder10,11 or even as epilepsy.12

For example, patients who were initially diag-nosed as having schizophrenia have been subse-quently found to have narcolepsy.11 Their veryvivid and sometimes frightening hypnagogic

Differential Diagnosis of Pediatric Sleep Disorders Chapter 2 19

300.15 F44.9 Nocturnal Dissociative Disorder788.36 N39.44 Sleep-related Enuresis780.56 G47.59 Parasomnia due to substance abuse307.47 F51.39 Parasomnia due to psychiatric disorders

Sleep-Related Movement Disorder: Problems Initiating or Maintaining Sleep, Excessive SleepMovements, or EDS¶

333.99 G47.61 Restless Legs Syndrome780.58 G47.62 Periodic Limb Movement Disorder780.58 G47.63 Sleep-Related Leg Cramps780.58 G47.64 Sleep-Related Bruxism780.58 G47.65 Sleep-Related Rhythmic Movement Disorder (body rocking, head

banging)

Isolated Symptoms, Apparently Normal Variants, and Unresolved Issues307.49 R29.81 Long Sleeper307.49 R29.81 Short Sleeper786.09 R06.5 Snoring307.49 R29.81 Sleeptalking781.01 R25.8 Sleep Starts, Hypnagogic Jerks781.01 R25.8 Benign Sleep Myoclonus of Infancy

*See Chapter 11.†See Chapters 17 to 23.‡See Chapters 13 to 16.§See Chapters 8 and 9.¶¶See Chapters 25 to 27.¶See Chapter 26.EDS, excessive daytime sleepiness; ICD-9-CM, International Classification of Diseases, Ninth Edition, Clinical

Modification; ICD-10-CM: International Classification of Diseases, Tenth Edition, Clinical Modification.

Table 2–1. Sleep Disorders in Children—Cont’d

Ch02.qxd 2/12/05 1:03 PM Page 19

hallucinations were misinterpreted to be psy-chotic hallucinations. Adults with narcolepsycan usually differentiate dreamlike imagery asbeing “unreal,” but children with narcolepsy maybe unable to determine whether the hypnagogichallucinations were “real” (i.e., a hallucinationthat appeared to be real) or dreams. Similarly,children may be diagnosed as having depressionwhen in fact they have sleepiness that resultsfrom a sleep disorder such as one of the move-ment disorders.

Drug Use and AbuseA great number of children are being treatedwith medications that can cause insomnia. Thesemedications might include stimulants (as usedfor the treatment of attention-deficit/hyperactiv-ity disorder13) or antidepressants.9 The use ofalerting illicit drugs must also be assessed.14 Theclinician seeing teenagers should be aware ofthe illicit drugs being used in the communitythat might lead to insomnia and other sleepcomplaints. Thus, if a parent brings a teenagerfor assessment who has had a personality changeand perhaps weight loss, and who has to haveclothes laundered frequently because of obviousand increased perspiration, use of drugs such as3,4-methylenedioxymethamphetamine (MDMA,“Ecstasy”) should be considered. Such childrenmay also have restless legs syndrome andinsomnia.15

Sleep-Induced RespiratoryImpairmentInsomnia appears to be a rare manifestation ofsleep breathing disorders in children. Caregiversmay complain that the child is “restless” andmoving a great deal, but insomnia per se isunusual. On the other hand, children with dis-orders such as asthma, or those with chronicconditions such as cystic fibrosis who maycough a great deal, may have difficulty in initiat-ing and maintaining sleep because of theirunderlying respiratory disorder.

Movement DisordersMovement disorders such as restless legs syn-drome and periodic movements in sleep occurin children and may cause severe insomnia. As in

the adult, the disorder may be secondary—forexample, caused by a reduction in iron stores.3

Surprisingly, other movement disorders—forexample, head banging or body rocking (rhyth-mic sleep disorders)—are less likely to appearwith insomnia than might be expected. Patientswith the rhythmic sleep disorders are oftenbrought to the physician’s attention by a care-giver who is quite concerned about the hazardsthey believe might be caused by the sometimesvigorous movements.

When a child is suspected or proven to havea movement disorder, especially restless legssyndrome, it is worthwhile to determine whethera reduction in iron stores exists.3 Serum ferritin,iron, and iron-binding capacity should be meas-ured. A screening complete blood count is notadequate to exclude reduction in iron stores. Ifreduced levels of iron stores are confirmed, thenthe source of the iron loss should be determined.Poor diet, blood loss from menses, or blood lossfrom the gastrointestinal tract should be explored.If gastrointestinal symptoms are present, the cli-nician should screen for disorders that maycause malabsorption, such as celiac disease.

Disorders of Timing in the Sleep–Wake CycleParents of very young children may believe thata sleep–wake cycle timing problem is present,but frequently this results from unreasonableexpectations about the children’s sleep—in par-ticular, about when they should be able to sleepthrough the night.

Teenagers may develop the delayed sleepphase disorder. Patients with this disorder go tobed but are not able to fall asleep until very late.When allowed to sleep in, they will sleep in untillunchtime or later. They have a great deal of dif-ficulty getting out of bed in the morning—forexample, to go to school. This diagnosis is reallyone of exclusion, and it is important that the cli-nician determine that other causes of difficultyin falling asleep are not present—for example,a movement disorder, the use of stimulants, orplaying video games and listening to music lateinto the night. One should also exclude depres-sion and a pattern of school avoidance.

Children with blindness or neurologic dam-age may have a timing problem related to theirnot being able to use the retinohypothalamic


Ch02.qxd 2/12/05 1:03 PM Page 20

pathways to synchronize their sleep–wake cycleto the 24-hour day.16

ParasomniasSleepwalking, confusional arousals, sleep terrors,and sleeptalking do not necessarily, by them-selves, cause insomnia. In fact, they usually donot. However, a child may develop a fear of fallingasleep after such an episode has occurred. Thiscan be quite distressing. For example, a childwho is known to sleepwalk a great deal may beunable to fall asleep when at a friend’s house forfear of having a sleepwalking episode.

Central Nervous System DisordersDiseases involving the central nervous systemmay cause insomnia. It is probably prudent toinclude in this category disorders such as autism,which may result in the child often being awakeand active at night.17,18 As is often the casefor childhood sleeping disorders, it is seldomthe child who complains of the problem but thecaregiver.

HYPERSOMNIA

Disorders of hypersomnia result in excessivesleepiness when the child should be wide awakeand alert. Sleepiness has many faces in children,ranging from the inability to stay awake (evenfalling asleep in school) to what can best bedescribed as hyperactivity and the inability to sitstill and focus in school. In some cases, sleepi-ness is apparent because the child falls asleep atthe wrong time and in the wrong place. In othercases, the teacher complains that the child isdaydreaming and cannot focus.

Two disorders of excessive daytime sleepiness(EDS) require special attention: those that leadto sleep breathing disorders and those that affectthe central nervous system. These disorders arefrequently missed by clinicians. Several categoriesof disorders that cause daytime sleepiness areidentical to categories of disorders that are asso-ciated with insomnia.

Behavioral and PsychophysiologicDisordersAny of the disorders that cause insomnia, if theylead to a sufficient loss of nocturnal sleep, can

result in daytime sleepiness. Thus, the teenagerwho falls asleep at 4 AM because of a delayedsleep phase is likely to fall asleep in class andmay present with a clinical problem of excessivedaytime sleepiness.

Some disorders, however, result in daytimesleepiness even though the child has apparentlybeen in bed and asleep for an acceptable numberof hours. Thus, the clinician and the parentshould know the normal nocturnal sleep timesfor children of various ages. The author has eval-uated 10-year-old children for EDS whose par-ents believed it was normal for a child of that ageto sleep 7 hours per night, whereas 9 to 11 hoursis more appropriate.

Psychiatric DisordersSeveral major psychiatric disorders, or the drugsused to treat them, can cause daytime sleepiness.In particular, depression has this effect.

Drug DependencyThe use of licit and illicit drugs must be consid-ered in children with hypersomnias. The clini-cian should be aware of what drugs are beingused in the community.

Sleep-Induced RespiratoryImpairmentObstructive sleep apnea syndrome and its vari-ants lead to excessive daytime sleepiness becauseof the increased number of arousals during thenight. Several issues in children are worth atten-tion. In adults, the main cause of sleep breathingdisorders is obesity, but in children it is enlargedtonsils and adenoids. Obese children, however,can develop obstructive sleep apnea syndromethat is indistinguishable from that in adults.19,20

Arterial hypertension may be caused by sleepapnea in a child.21 Loud and disruptive snoringin the child with documented daytime sleepinessshould be evaluated as a probable case of a sleepbreathing disorder. In children, retrognathiaand micrognathia can also lead to sleep breathingdisorders. Thus, an examination of the upperairway is absolutely critical in the child withexcessive daytime sleepiness. Sleepiness in chil-dren with obstructive sleep apnea syndrome hasbeen misinterpreted as a feature of depression.Sleep apnea can occur in children of any age.


Ch02.qxd 2/12/05 1:03 PM Page 21

Sleep breathing abnormalities should be sus-pected in children if they have the symptomstypically seen in adults (snoring, sleepiness), andit should also be suspected in children with awakehypoventilation, growth retardation, and certaincongenital disorders.

Hypoventilation

Children with congenital central hypoventila-tion syndrome have symptoms that may be sim-ilar to those in children with obstructive sleepapnea syndrome.22-24 Obstructive sleep apneacaused by adenotonsillar hypertrophy may com-plicate congenital central hypoventilation syn-drome24 and chronic lung diseases. Any sleepbreathing disorder that causes hypoventilationmay result in pulmonary hypertension, cor pul-monale, and polycythemia.25,26

Growth Retardation

Growth retardation may be present in newbornswhose mothers had a sleep breathing disorder.27,28

Growth retardation is a feature of sleep breathingdisorders in children of any age.6,29-32

Congenital Disorders

Congenital diseases including myopathies,33,34

muscular dystrophies,34 neurologic disordersthat might affect the system controlling breath-ing, and disorders resulting in facial skeletalabnormalities or soft tissue abnormalities of theupper airway can all result in sleep breathingdisorders.14,15,35-38

Movement DisordersChildren with movement disorders may be quitesleepy during the daytime. This is in contrast toadults, in whom excessive daytime sleepinessmay not be a prominent symptom with move-ment disorders.

Disorders of Timing of the Sleep–Wake PatternA patient who has non–24-hour sleep–wake syn-drome, or the child who is asleep almost all thetime or who has a completely random sleep–wake

schedule, may have a mass lesion in the centralnervous system.

Central Nervous SystemAbnormalityJust as abnormalities of the central nervous sys-tem can lead to insomnia, some also lead toexcessive daytime sleepiness. Narcolepsy is adisorder that is frequently missed because mostclinicians do not take the time to ask about apatient’s sleep. In fact, in some series, it tookover 15 years from the onset of symptoms beforethe diagnosis of narcolepsy was first made.39

Many patients who are ultimately diagnosed withnarcolepsy have previously been diagnosedwith a psychiatric condition, frequently depres-sion. Similarly, patients with the obstructivesleep apnea syndrome are often diagnosed withdepression.

Some disorders that result in excessivedaytime sleepiness because of involvement ofthe hypothalamic areas of the nervous systemalso lead to excessive weight gain, which maythen cause sleepiness (e.g., obstructive sleepapnea). This may be the situation, for example,in Asperger’s syndrome or Kleine-Levin syn-drome.

ABNORMAL SLEEP-RELATEDBEHAVIORS

Children with abnormal movements and behav-iors seldom complain about them. It is usuallyothers who notice the abnormalities, whichrange from the merely annoying (the chipmunk-like noises of patients with sleep bruxism) toepisodes of sleep terrors that are frighteningto the observer. Some of these disorders can causethe child a great deal of embarrassment (e.g.,sleep enuresis). Although some of the disordersin this category are not serious, they can causethe patient a great deal of distress.

Behavioral or PsychophysiologicDisordersThe parent may complain about a child’s wan-dering at night, making frequent curtain calls,or wanting to sleep in the parental bed. These“disorders” can be disruptive to the family.


Ch02.qxd 2/12/05 1:03 PM Page 22

Movement DisordersRhythmic movement disorders (e.g., head-banging) are surprisingly common in children.These disorders are brought to the clinician’sattention usually not because of any symptombut because the parent assumes that the some-times violent movements must be associated withabnormal pathology.

ParasomniasParasomnias are extremely common in children.Disorders such as night terrors, sleepwalking,and enuresis can be quite disruptive to the fam-ily and cause a great deal of embarrassmentfor the patient. It is important to differentiatesuch movements from sleep-related epilepsy (seelater).

Central Nervous System DisordersSevere disorders of the neurologic system,including seizures, may result in abnormalsleep-related behaviors. A sleep behavior thatis very stereotypic and almost always the sameshould alert the clinician to the possibility of aseizure disorder.

SLEEP DISORDERS IN OTHERCLINICAL SYNDROMES

When patients have complex clinical syndromes,a sleep disorder is sometimes overlooked. Table2–2 gives examples of some complex disordersthat have been associated with sleep disorders.Sometimes, treating the sleep disorder can result

in quite substantial clinical improvement. Someclinicians are reluctant to consider a sleep-relateddiagnosis in patients with, for example, Downsyndrome, because of their belief that nothingcan be done for sleep disorders in such patients.However, some patients with Down syndromecan do extremely well with treatment for theirsleep disorders.

CONCLUSION

Children referred for sleep disorders may nothave a sleep disorder at all: the problem may bethat the caregiver does not understand normalsleep in children. Some children who are referredfor sleep disorders may have a serious condition(e.g., obstructive sleep apnea) that can betreated. If they have a movement disorder causedby an iron deficiency, they can be cured. Childrenwith narcolepsy, on the other hand, will requirelifelong management to function normally.Children with sleep disorders require a detailedreview of clinical features and a detailed assess-ment to optimize their management.

References

1. Diagnostic Classification Steering Committee:International Classification of Sleep Disorders:Diagnostic and Coding Manual. Chicago, Ill,American Academy of Sleep Medicine, 2004.

2. St Georgiev V: Chemotherapy of enterobiasis(oxyuriasis). Expert Opin Pharmacother 2001;2:267-275.

3. Kryger MH, Otake K, Foerster J: Low body storesof iron and restless legs syndrome: A correctablecause of insomnia in adolescents and teenagers.Sleep Med 2002;3:127-132.


Syndrome or Disease Sleep AbnormalityAchondroplasia Sleep breathing disorders40,41

Asperger’s syndrome Hypersomnia, sleep breathing disorders42-44

Autism Insomnia17,18

Chiari malformation Sleep breathing disorder45

Down syndrome Sleep breathing disorders26,46-49

Hirschsprung’s disease Sleep breathing disorders50

Mucopolysaccharide storage disorders Sleep breathing disorders51,52

Pierre Robin Sleep breathing disorder36-38,53

Prader-Willi Sleep Breathing disorder54

Syringomyelia Sleep breathing disorder55

Table 2–2. Examples of Complex Syndromes Associated with Sleep Disorders

Ch02.qxd 2/12/05 1:03 PM Page 23

4. Lind MG, Lundell BP: Tonsillar hyperplasia inchildren: A cause of obstructive sleep apneas, CO2retention, and retarded growth. Arch Otolaryngol1982;108:650-654.

5. Mangat D, Orr WC, Smith RO: Sleep apnea, hyper-somnolence, and upper airway obstruction sec-ondary to adenotonsillar enlargement. ArchOtolaryngol 1977;103:383-386.

6. Nimubona L, Jokic M, Moreau S, et al: Obstructivesleep apnea syndrome and hypertrophic tonsils ininfants [in French]. Arch Pediatr 2000;7:961-964.

7. Ramakrishna S, Ingle VS, Patel S, et al: Reversiblecardio-pulmonary changes due to adeno-tonsillarhypertrophy. Int J Pediatr Otorhinolaryngol 2000;55:203-206.

8. Doghramji PP: Detection of insomnia in primarycare. J Clin Psychiatry 2001;62(Suppl 10):18-26.

9. Rosenberg DR, Stewart CM, Fitzgerald KD, et al:Paroxetine open-label treatment of pediatricoutpatients with obsessive-compulsive disorder.J Am Acad Child Adolesc Psychiatry 1999;38:1180-1185.

10. Daniels E, King MA, Smith IE, Shneerson JM:Health-related quality of life in narcolepsy. J SleepRes 2001;10:75-81.

11. Douglass AB, Hays P, Pazderka F, Russell JM:Florid refractory schizophrenias that turn out tobe treatable variants of HLA-associated nar-colepsy. J Nerv Ment Dis 1991;179:12-17.

12. Zeman A, Douglas N, Aylward R: Lesson of theweek: Narcolepsy mistaken for epilepsy. BMJ 2001;322:216-218.

13. Efron D, Jarman F, Barker M: Side effects ofmethylphenidate and dexamphetamine in chil-dren with attention deficit hyperactivity disorder:A double-blind, crossover trial. Pediatrics 1997;100:662-666.

14. Gahlinger PM: Club drugs: MDMA, gamma-hydroxybutyrate (GHB), Rohypnol, and keta-mine. Am Fam Physician 2004;69:2619-2626.

15. Vollenweider FX, Gamma A, Liechti M, Huber T:Psychological and cardiovascular effects andshort-term sequelae of MDMA (“ecstasy”) inMDMA-naive healthy volunteers. Neuropsycho-pharmacology 1998;19:241-251.

16. Lapierre O, Dumont M: Melatonin treatment of anon-24-hour sleep-wake cycle in a blind retardedchild. Biol Psychiatry 1995;38:119-122.

17. Posey DJ, McDougle CJ: The pharmacotherapy oftarget symptoms associated with autistic disorderand other pervasive developmental disorders.Harv Rev Psychiatry 2000;8:45-63.

18. Richdale AL, Prior MR: The sleep/wake rhythm inchildren with autism. Eur Child Adolesc Psychiatry1995;4:175-186.

19. Daniels SR: Obesity in the pediatric patient: Cardio-vascular complications. Prog Pediatr Cardiol2001;12:161-167.

20. Slyper AH: Childhood obesity, adipose tissue dis-tribution, and the pediatric practitioner. Pediatrics1998;102:e4.

21. Marcus CL, Greene MG, Carroll JL: Blood pres-sure in children with obstructive sleep apnea. AmJ Respir Crit Care Med 1998;157:1098-1103.

22. Silvestri JM, Weese-Mayer DE, Nelson MN:Neuropsychologic abnormalities in children withcongenital central hypoventilation syndrome.J Pediatr 1992;120:388-393.

23. Seid AB, Martin PJ, Pransky SM, Kearns DB:Surgical therapy of obstructive sleep apnea in chil-dren with severe mental insufficiency. Laryngo-scope 1990;100:507-510.

24. Kurz H, Sterniste W, Dremsek P: Resolution ofobstructive sleep apnea syndrome after ade-noidectomy in congenital central hypoventilationsyndrome. Pediatr Pulmonol 1999;27:341-346.

25. Freed D: Pulmonary hypertension in childrenwho snore. Br Med J (Clin Res Ed) 1981;283:231.

26. Levine OR, Simpser M: Alveolar hypoventilationand cor pulmonale associated with chronic air-way obstruction in infants with Down syndrome.Clin Pediatr (Phila) 1982;21:25-29.

27. Harding SM: Complications and consequences ofobstructive sleep apnea. Curr Opin Pulm Med2000;6:485-489.

28. Franklin KA, Holmgren PA, Jonsson F, et al:Snoring, pregnancy-induced hypertension, andgrowth retardation of the fetus. Chest 2000;117:137-141.

29. Trachtenbarg DE, Golemon TB: Office care of thepremature infant: Part II. Common medical andsurgical problems. Am Fam Physician 1998;57:2383-2390, 2400-2402.

30. Commare MC, Francois B, Estournet B, Barois A:Ondine’s curse: A discussion of five cases. Neuro-pediatrics 1993;24:313-318.

31. Breton D, Morisseau-Durand MP, Cheron G:Growth retardation and obstructive sleep apneain infants. Arch Fr Pediatr 1993;50:493-496.

32. Stradling JR, Thomas G, Warley AR, et al: Effectof adenotonsillectomy on nocturnal hypoxaemia,sleep disturbance, and symptoms in snoring chil-dren. Lancet 1990;335:249-253.

33. Sasaki M, Takeda M, Kobayashi K, Nonaka I:Respiratory failure in nemaline myopathy. PediatrNeurol 1997;16:344-346.

34. Khan Y, Heckmatt JZ, Dubowitz V: Sleep studiesand supportive ventilatory treatment in patientswith congenital muscle disorders. Arch Dis Child1996;74:195-200.

35. Cohen SR, Ross DA, Burstein FD, et al: Skeletalexpansion combined with soft-tissue reduction inthe treatment of obstructive sleep apnea in chil-dren: Physiologic results. Otolaryngol Head NeckSurg 1998;119:476-485.


Ch02.qxd 2/12/05 1:03 PM Page 24

36. Kiely JL, Deegan PC, McNicholas WT: Resolutionof obstructive sleep apnoea with growth in theRobin sequence. Eur Respir J 1998;12:499-501.

37. Tomaski SM, Zalzal GH, Saal HM: Airwayobstruction in the Pierre Robin sequence.Laryngoscope 1995;105:111-114.

38. Reimao R, Papaiz EG, Papaiz LF: Pierre Robinsequence and obstructive sleep apnea. Arq Neuro-psiquiatr 1994;52:554-559.

39. Broughton RJ, Fleming JA, George CF, et al:Randomized, double-blind, placebo-controlledcrossover trial of modafinil in the treatment ofexcessive daytime sleepiness in narcolepsy.Neurology 1997;49:444-451.

40. Tasker RC, Dundas I, Laverty A, et al: Distinct pat-terns of respiratory difficulty in young childrenwith achondroplasia: A clinical, sleep, and lungfunction study. Arch Dis Child 1998;79:99-108.

41. Mogayzel PJ Jr, Carroll JL, Loughlin GM, et al:Sleep-disordered breathing in children withachondroplasia. J Pediatr 1998;132:667-671.

42. Godbout R, Bergeron C, Limoges E, et al: A labo-ratory study of sleep in Asperger’s syndrome.Neuroreport 2000;11:127-130.

43. Patzold LM, Richdale AL, Tonge BJ: An investiga-tion into sleep characteristics of children withautism and Asperger’s disorder. J Paediatr ChildHealth 1998;34:528-533.

44. Berthier ML, Santamaria J, Encabo H, Tolosa ES:Recurrent hypersomnia in two adolescent maleswith Asperger’s syndrome. J Am Acad ChildAdolesc Psychiatry 1992;31:735-738.

45. Frim DM, Jones D, Goumnerova L: Developmentof symptomatic Chiari malformation in a childwith craniofacial dysmorphism. Pediatr Neurosurg1990-91;16:228-231.

46. Jacobs IN, Gray RF, Todd NW: Upper airwayobstruction in children with Down syndrome. ArchOtolaryngol Head Neck Surg 1996;122:945-950.

47. Hultcrantz E, Svanholm H: Down syndrome andsleep apnea: A therapeutic challenge. Int J PediatrOtorhinolaryngol 1991;21:263-268.

48. Kasian GF, Duncan WJ, Tyrrell MJ, Oman-GanesLA: Elective oro-tracheal intubation to diagnosesleep apnea syndrome in children with Down’ssyndrome and ventricular septal defect. Can JCardiol 1987;3:2-5.

49. Hoch B, Barth H: Cheyne-Stokes respiration as anadditional risk factor for pulmonary hypertensionin a boy with trisomy 21 and atrioventricular sep-tal defect. Pediatr Pulmonol 2001;31:261-264.

50. Stern M, Hellwege HH, Gravinghoff L, LambrechtW: Total aganglionosis of the colon (Hirschsprung’sdisease) and congenital failure of automatic con-trol of ventilation (Ondine’s curse). Acta PaediatrScand 1981;70:121-124.

51. Kurihara M, Kumagai K, Goto K, et al: Severe typeHunter’s syndrome: Polysomnographic and neu-ropathological study. Neuropediatrics 1992;23:248-256.

52. Semenza GL, Pyeritz RE: Respiratory complica-tions of mucopolysaccharide storage disorders.Medicine (Baltimore) 1988;67:209-219.

53. Clift S, Dahlitz M, Parkes JD: Sleep apnoea in thePrader-Willi syndrome. J Sleep Res 1994;3:121-126.

54. Doshi A, Udwadia Z: Prader-Willi syndrome withsleep disordered breathing: Effect of two years’nocturnal CPAP. Indian J Chest Dis Allied Sci2001;43:51-53.

55. Pasterkamp H, Cardoso ER, Booth FA: Obstructivesleep apnea leading to increased intracranial pres-sure in a patient with hydrocephalus andsyringomyelia. Chest 1989;95:1064-1067.


Ch02.qxd 2/12/05 1:03 PM Page 25

NORMAL SLEEP PATTERNSAND BEHAVIOR

Relatively few large-scale epidemiologic studieshave been performed to systematically define nor-mal sleep and wakefulness patterns and sleepduration in infants, children, and adolescents. Ofthose that have examined the issue of what con-stitutes normal sleep in children in order to betterdefine abnormal or insufficient sleep, most haveutilized subjective, parent-report, retrospective,cross-sectional surveys in selected populations.There are even more limited data from studies uti-lizing more objective methods of measuring sleepquality and duration, such as polysomnographyand actigraphy, and many of these studies wereconducted prior to the establishment of acceptedsleep monitoring and scoring standards.

Descriptions of sleep phenotypes and defini-tions of normal sleep patterns and requirementsmust necessarily incorporate the wide range ofnormal developmental and physical matura-tional changes across childhood and adoles-cence. Although cross-sectional studies yieldimportant information regarding sleep in dis-crete age groups, they do not describe the evolu-tion and persistence of sleep–wake patterns overtime, nor do they help to elucidate the complexreciprocal relationship between sleep and cogni-tive/emotional development from the prenatalperiod through adolescence.

In general, large epidemiologic studies do sup-port several general trends in normal sleep pat-terns across childhood:

1. Decreased average 24-hour sleep durationfrom infancy through adolescence, whichinvolves a decline in both diurnal and noctur-nal sleep amounts.1 In particular, there isa dramatic decline in daytime sleep (sched-uled napping) between 18 months and 5 years,with a less marked and more gradual contin-

ued decrease in nocturnal sleep amounts intolate adolescence.

2. Increasingly irregular sleep–wake patternswith larger discrepancies between schoolnight and non–school night bedtimes and waketimes and increased weekend oversleep frommiddle childhood through adolescence.

3. A gradual shift to a later bedtime and sleeponset time that begins in middle childhoodand accelerates in early to mid adolescence.As with most sleep behaviors, these trendsreflect the physiologic/chronobiologic, devel-opmental, and social/environmental changesthat occur across childhood.

Also, some evidence suggests that sleep pat-terns and behaviors in children and adolescentstoday have changed somewhat from those presentin previous generations over time. Several studieshave shown not only that average sleep durationdecreases across middle childhood and adoles-cence (in contrast to sleep needs, which do notdramatically decline) but that sleep duration inequivalent age groups has declined over a periodof time. This trend appears to be related in school-aged children largely to later bedtimes, and, inadolescents, to earlier sleep offset as well as latersleep onset times.

It is also important to consider the culturaland family context within which sleep behaviorsin children occur. For example, cosleeping ofinfants and parents is a common and acceptedpractice in many ethnic groups, includingAfrican Americans, Hispanics, and SoutheastAsians, both in their countries of origin and inthe United States. Therefore, the developmentalgoal of independent self-soothing in infants atbedtime and after night wakings, although clearlyassociated in a number of studies with fewersubsequent sleep problems in young children,may not be shared by all families. In many tradi-tional societies, sleep is heavily embedded in

Epidemiology of Sleep Disorders during Childhood

Judith Owens

3

27

Ch03.qxd 01/28/05 10:27 AM Page 27

social practices, and both the sleeping environ-ment and the positioning of sleep periods withinthe context of other activities is much less soli-tary and less rigid than in more westernized cul-tures. The relative value and importance of sleepas a health behavior, the interpretation of prob-lematic versus normal sleep by parents, and therelative acceptability of various treatment strate-gies (e.g., the cry-it-out approach) for sleepproblems are just a few additional examples ofsleep issues that are impacted by cultural andfamily values and practices.

NOSOLOGY

For a variety of reasons, it is often a challenge tooperationally define problematic sleep in chil-dren. The range of sleep behaviors that may beconsidered normal or pathologic is wide and thedefinitions are often highly subjective. Researchershave taken a number of approaches to this issue:some use a priori definitions of disturbed or poorsleep (e.g., waking for longer than 30 minutesmore than three times a week), whereas othershave relied on comparison to normative popula-tions or have based the definition of sleep prob-lems on what the parent subjectively identifiesas problematic. Other authors have attempted toincorporate more concrete evidence of daytimesequelae (mood, behavior, academic perform-ance) to define significant sleep problems. Somestudies have included reports from otherobservers, including teachers, to avoid depend-ing solely on parents’ expectations for, and toler-ance, awareness, and interpretation of, sleep anddaytime behaviors. In any event, it is clear thata common nosology in terms of research defini-tions of sleep disorders in children is lacking andneeds to be developed and evaluated to identifyand target populations at risk for sleep problemsfor intervention.

On the other hand, clinically significant sleepproblems, like many behavioral problems in child-hood, may best be viewed as more loosely occur-ring along a severity and chronicity continuumthat ranges from a transient and self-limited dis-turbance to a disorder that meets specific diag-nostic criteria as outlined in the InternationalClassification of Sleep Disorders (ICSD) and/orthe Diagnostic and Statistical Manual of MentalDisorders (DSM-IV). Unlike strict research defi-nitions of sleep problems, the validity of parental

concerns and opinions regarding their child’ssleep patterns and behaviors, and the resultingstress on the family, must be considered in defin-ing sleep disturbances in the clinical context.Furthermore, successful treatment of pediatricsleep problems is highly dependent on identifi-cation of parental concerns, clarification of mutu-ally acceptable treatment goals, active explorationof opportunities and obstacles, and ongoingcommunication of issues and concerns.

VARIABLES AFFECTING SLEEPPATTERNS AND BEHAVIORS

The relative prevalence and the various types ofsleep problems that occur throughout childhoodmust also be understood in the context of nor-mal physical and cognitive/emotional phenom-ena that are occurring at different developmentalstages. For example, brief regressions in sleepbehaviors often accompany the achievement ofmotor and cognitive milestones in the first yearof life, and increased night-time fears and nightwakings in toddlers may be a temporary mani-festation of developmentally normal separationanxiety peaking during that stage. Parental recog-nition and reporting of sleep problems in chil-dren also varies across childhood, with parentsof infants and toddlers more likely to be aware ofsleep concerns than those of school-aged childrenand adolescents. The very definition of a sleepproblem by parents is often highly subjectiveand is frequently determined by the amount ofdisruption caused to parents’ sleep.

In addition to considering sleep disturbanceswithin a developmental context, a number ofother important child, parental, and environ-mental variables affect the type, relative preva-lence, chronicity, and severity of sleep problems.Child variables that may significantly impactsleep include temperament and behavioral style,individual variations in circadian preference, cog-nitive and language delays, and the presence ofcomorbid medical and psychiatric conditions.Parental variables include parenting and disci-pline styles, parents’ education level and knowl-edge of child development, mental health issues(such as maternal depression), family stress, andquality and quantity of parents’ sleep. Envi-ronmental variables include the physical envi-ronment (space, noise, perceived environmentalthreats to safety, and room and bed sharing),

28 Chapter 3 Epidemiology of Sleep Disorders during Childhood

Ch03.qxd 01/28/05 10:27 AM Page 28

family composition (number, ages, and health sta-tus of siblings and extended family members), andlifestyle issues (parental work status and compet-ing priorities for parents’ and children’s time).

Finally, although many sleep problems ininfants and children are transient and self-limitedin nature, the common wisdom that all childrengrow out of sleep problems is not an accurateperception. Certain intrinsic and extrinsic riskfactors (such as difficult temperament, chronicillness, neurodevelopmental delays, maternaldepression, and family stress) may predisposea given child toward a more chronic sleep dis-turbance. The persistence of infant sleep prob-lems into early childhood has been documentedin a number of studies.

PREVALENCE: SURVEYS OFNORMAL POPULATIONS

The following discussion largely focuses ongeneral sleep problems in children and on non-specific presenting sleep complaints, such asbedtime resistance and night wakings; preva-lence data regarding specific sleep disorderssuch as obstructive sleep apnea syndrome arefound in other chapters. A number of studies haveexamined the prevalence of parent- and child-reported sleep complaints in large samples ofchildren and adolescents; many of these studieshave also attempted to further delineate theassociation between disrupted sleep and behav-ioral concerns. Most of these studies have uti-lized broad-based parent-report sleep surveysto assess for a variety of sleep problems, rangingfrom bedtime resistance to prolonged night wak-ings to parasomnias.2,3 There are limitationsto these data, including varying definitions ofproblem sleep, difficulty in identifying daytimesleepiness–related behaviors (especially inyounger children), and limited informationregarding possible confounding factors (e.g.,comorbid medical conditions and medicationuse). Thus, parent reports may over- or underes-timate the prevalence of sleep problems.Furthermore, it should be emphasized thatparental descriptions of sleep problems are notequivalent to diagnoses of sleep disorders madeduring a clinical evaluation.

Approximately 25% of all children experiencesome type of sleep problem at some point duringchildhood. Such problems range from short-term

difficulties in falling asleep and night wakingsto more serious primary sleep disorders such asobstructive sleep apnea. Studies have reportedan overall prevalence of parent-reported sleepproblems ranging from 25% to 50% in samplesof preschool-aged children to 37% in a commu-nity sample of 4 to 10 year olds.4,5 A more recentstudy of over 14,000 school-aged children6 foundsleep problems in 20% of 5 year olds and 6% of11 year olds. Another study found a prevalenceof sleeping difficulties in 43% of 8 to 10 yearolds.7 Many studies have examined the self-reported prevalence of sleep problems in adoles-cents: upwards of 40% of adolescents also havesignificant sleep complaints,8 and 12% identifiedthemselves as “chronic poor sleepers.”9

Studies that have used self-reports of sleepproblems from older children have suggestedthat there may be a discrepancy between parentaland child reports, with parents less likely thanthe children to report sleep onset delays andnight wakings. Finally, in studies that have askedhealth care providers to identify the prevalenceof sleep problems seen in their practices, rates ofsignificant difficulties initiating or maintainingsleep across age groups vary, in one recent studymaking up on average about 3% of all practicevisits.10 In the IMS (Intercontinental MedicalStatistics) Health National Disease and Thera-peutic Index (NDTI) survey of 2930 office-basedpractices in the continental United States, whichuses diagnostic codes, 0.05% of all pediatric visitswere for sleep problems and 0.01% were specifi-cally for insomnia. However, practitioner-identified sleep problems may also underestimateprevalence11; for example, in the study by Ronaet al.,6 less than 25% of the school children withsleep problems had consulted a physician.

IMPACT OF SLEEP PROBLEMS

Although a detailed description of the impact ofdisrupted or insufficient sleep on children’shealth and well-being is beyond the scope ofthis chapter, any discussion of the epidemiologyof pediatric sleep must underscore the impor-tance of the relationships between sleep prob-lems and mood, performance, and behavior.Many of the studies that have examined thesecomplex relationships have sought to determinehow much sleep is needed by infants, children,and teenagers for healthy functioning, and how

Epidemiology of Sleep Disorders during Childhood Chapter 3 29

Ch03.qxd 01/28/05 10:27 AM Page 29

patterns of growth and development from infancyto adolescence are negatively impacted by insuf-ficient sleep. In general, studies that have exam-ined these relationships have used one of a numberof different methodologies:

1. Assessment of effects of experimental sleeprestriction on mood, neuropsychological testperformance, and observed behavior

2. Evaluation of mood, behavioral, and academicproblems in children with clinical sleep disor-ders (e.g., obstructive sleep apnea syndrome,restless legs syndrome/periodic limb move-ments during sleep)

3. Examination of the impact of treatment ofsleep disorders (pharmacologic, surgical, behav-ioral) on neurobehavioral measures

4. Identification of behavioral and academicdysfunction in “naturalistic settings,” of chil-dren identified as poor sleepers or as havinginadequate sleep

5. Identification of sleep problems in popula-tions of children with behavioral and academicproblems (e.g., attention deficit hyperactivitydisorder) compared to normal controls

A number of studies have demonstrated thatchildren for whom parents report sleep com-plaints are more likely to manifest daytime sleepi-ness, moodiness, behavioral problems, and schooland learning problems.12 In one recent surveyof sleep problems in elementary school–agedchildren, teachers reported behavioral evidenceof significant daytime sleepiness in the class-room setting in 10% of their students.4 Anotherstudy found a significant correlation betweenearly rise times and self-reported difficulties inattention and concentration13 in fifth graders.Because of their cross-sectional design, however,and the reliance on parental (or self-) report fordescription of both sleep and behavioral vari-ables in many of these studies, the results mustbe interpreted cautiously insofar as sleep is con-sidered to be a causal factor.

In similar survey studies, adolescents whoreported disturbed or inadequate sleep were alsomore likely to report subjective sleepiness, mooddisturbances, and performance deficits in bothsocial and academic spheres.14,15 Adolescentsmay be at increased risk for sleep disturbancesand inadequate sleep for a number of biologic,environmental, and psychosocial reasons. Theresultant decrease in total sleep time in adoles-

cents has been associated in several studies withpoorer grades in school as well as depressedmood, and anxiety. One study16 that found anoverall prevalence of significant sleep problemsin about one third of the adolescents surveyed,also reported an increased level of self- and parent-reported externalizing behavioral problems inthe sleep-disturbed sample.

Sleep problems are also a significant sourceof distress for families and may be one of theprimary reasons for caregiver stress in families withchildren who have chronic medical illnessesor severe neurodevelopment delays. Furthermore,the impact of childhood sleep problems is intensi-fied by their direct relationship to the quality andquantity of parents’ sleep, particularly if disruptedsleep results in daytime fatigue or mood distur-bances and leads to a decreased level of effectiveparenting. Poor parental sleep has even been impli-cated as a risk factor for child physical abuse.Successful intervention not only has the effect ofimproving the sleep of the entire family but mayhelp parents develop behavioral strategies that maygeneralize for use with daytime behavior problems.

Unlike in adults, however, daytime sleepinessin children may not be characterized by suchovert behaviors as yawning and complainingabout fatigue but may rather be associated witha host of more subtle or even paradoxical behav-ioral manifestations (such as increased activity).These range from emotional lability and lowfrustration tolerance (internalizing behaviors)to neurocognitive deficits to behavioral disinhi-bition (externalizing behaviors such as increasedaggression).12 In turn, functional deficits inmood, attention, cognition, and behavior maylead to performance deficits in the home, school,and social setting. Objective, reliable, and cost-effective measures of sleepiness and alertness inchildren are lacking—particularly measures thatcould be applied to large epidemiologic samples.In addition, subjective self-report data regardingsleepiness are largely unavailable in children,and behavioral manifestations of sleepiness notonly vary with age and developmental level butare often not reliably interpreted by parents andother caretakers. Empirical studies involvingboth normal and sleep-deprived pediatric popu-lations (children with sleep disorders, adoles-cents) have begun to describe the extent of, andthe consequences of, inadequate or disruptedsleep in children.


Ch03.qxd 01/28/05 10:27 AM Page 30

Studies of sleep in children with primarybehavioral and learning problems have furthersupported an association between sleep andperformance impairments. Empirical evidenceindicates that children experience significantdaytime sleepiness as a result of disturbed orinadequate sleep, and most studies suggesta strong link between sleep disturbance andbehavioral problems.

EPIDEMIOLOGY OF SLEEPIN SPECIAL POPULATIONS

Because of the multiple manifestations of poorand insufficient sleep, the clinical symptomsof any primary medical, developmental, or psy-chiatric disorder are likely to be exacerbated bycomorbid sleep problems. Furthermore, sleepproblems themselves tend to be more commonin those children and adolescents with chronicmedical and psychiatric conditions. Improvingsleep has the benefit of improving clinical out-come as well.

Sleep disturbances in pediatric special needspopulations are extremely common. Significantsleep problems occur in 30% to 80% of childrenwith severe mental retardation and in at least halfof children with less severe cognitive impairment;similar estimates in children with autism or per-vasive developmental delay are in the 50% to 70%range.17,18 Significant problems with initiationand maintenance of sleep, shortened sleep dura-tion, irregular sleeping patterns, and early morn-ing waking have been reported in a variety ofdifferent neurodevelopmental disorders, includ-ing Asperger’s syndrome, Angelman’s syndrome,Rett’s syndrome, Smith-Magenis syndrome, andWilliams syndrome. Other studies have suggestedthat similar rates of sleep problems also occur inboth younger and older blind children, with diffi-culty falling asleep, night wakings, and restlesssleep being the most common concerns. Thesechildren with pre-existing neurologic abnormali-ties are also more likely to have sleep problemsthat are severe, and multiple sleep disorders arelikely to occur simultaneously.

Virtually all psychiatric disorders in childrenmay be associated with sleep disruption.Epidemiologic and clinical studies indicatethat psychiatric disorders are the most commoncause of chronic insomnia. Psychiatric disor-ders can also be associated with daytime sleepi-

ness, fatigue, abnormal circadian sleep patterns,disturbing dreams and nightmares, and move-ment disorders during sleep. Growing evidencesuggests that primary insomnia (i.e., insomniawith no concurrent psychiatric disorder) isa risk factor for later developing psychiatricconditions, particularly depressive and anxietydisorders.19

Several studies have evaluated the prevalenceof sleep problems in samples of children andadolescents with a variety of psychiatric disor-ders.16,20 The results suggest an increase in a widerange of reported sleep disturbances in thesemixed clinical populations, including parasom-nias such as nightmares and night terrors, diffi-culty falling asleep and frequent and prolongednight wakings, sleep-related anxiety symptoms(e.g., fear of the dark), restless sleep, and subjec-tive poor quality of sleep with associated day-time fatigue. Similarly, an association has beenreported between DSM psychiatric disorders(including affective disorders, attention deficithyperactivity disorder, and conduct disorder)and sleep problems in surveys of children andadolescents from the general population.4

Studies of children with major depressive disor-der, for example, have reported a prevalence ofinsomnia of up to 75% and of severe insomniaof 30%, and sleep onset delay in a third ofdepressed adolescents, although it should benoted that objective data (e.g., with polysomnog-raphy) does not always support these subjectivecomplaints.

Finally, reports of sleep complaints, especiallybedtime resistance, refusal to sleep alone,increased night-time fears, and nightmares arecommon in children who have experiencedseverely traumatic events (including physicaland sexual abuse),21 and these complaints mayalso be associated with less dramatic butnonetheless stressful life events such as briefseparation from a parent, birth of a sibling, orschool transitions.22 Sleep problems are not uni-versally found in all children experiencing vary-ing degrees of stress, and some authors21 havesuggested that such variables as level of expo-sure and physical proximity to the traumaticsituation, previous exposure, and the opportu-nity for habituation to the stress may playimportant roles in either mitigating or exacer-bating associated sleep disturbances. Other con-siderations, such as the age and temperament


Ch03.qxd 01/28/05 10:27 AM Page 31

of the child, and the presence of parental psy-chopathology, also clearly have an importantinfluence.

Little is currently known about the interac-tion between sleep disorders and both acute andchronic health conditions (such as asthma, dia-betes, and juvenile rheumatoid arthritis), oneither a pathophysiologic or behavioral level,although, particularly in chronic pain condi-tions, these interactions are likely to significantlyimpact on morbidity and quality of life.23-26 Muchof the information currently available regardingthe types of sleep problems that occur in thesechildren comes from studies of adults withchronic medical conditions or from clinicalobservations. A few recent studies have begunto examine the role of sleep disturbances insickle cell disease27 and asthma, two disordersparticularly common in high-risk and minoritypopulations. The interaction between sleep andphysical and emotional dysfunction in acuteand chronic pain conditions, such as burns andjuvenile rheumatoid arthritis, has also begunto be explored. Factors such as the impact ofhospitalization, family dynamics, underlyingdisease processes, and concurrent medicationsare also clearly important to consider in assess-ing the bidirectional relationship of insomniaand chronic illness in children.

Specific medical conditions that may alsohave an increased risk of sleep problems includethe following:

Allergies: Chronic allergy-mediated rhinitiswith nasal congestion and cough, and chronicor frequent sinusitis may be associated with dif-ficulties initiating and maintaining sleep. Atopicdermatitis may cause sleep disruption and frag-mentation, difficulty falling asleep, night awak-enings, and decreased sleep duration, as a resultof frequent scratching and discomfort.

Asthma: Asthma in children has been associatedwith poorer subjective sleep quality, decreasedsleep duration, increased nocturnal wakingswith decreased sleep efficiency, and greater day-time sleepiness. The prevalence of sleep prob-lems in children with asthma has been reportedto be 60%. Asthma medications may also haveadverse effects on sleep; for example, both theo-phylline and oral corticosteroids may cause sig-nificant sleep disruption.

Burns: Children who are severely burned are atparticularly high risk for sleep problems, which in

turn may have significant adverse effects on theirrecovery. Difficulty falling asleep, increasedarousals, nightmares, and increased daytimesleepiness are common in the acute phase andmay persist for up to 1 year after the event.

Headaches: Migraine and tension headacheshave been associated with sleep onset and sleepmaintenance problems, and migraines have beenassociated with an increased incidence of sleep-walking. Furthermore, sleep deprivation maytrigger headaches in susceptible individuals.

Rheumatologic disorders: Children with juve-nile rheumatoid arthritis and with fibromyalgiahave been shown to have more frequent nightwakings and sleep fragmentation that may beassociated with daytime sleepiness.

References

1. Iglowstein I, Jenni O, Molinari L, Largo R: Sleepduration from infancy to adolescence: Referencevalues and generational trends. Pediatrics 2003;111:302-307.

2. Owens J, Nobile C, McGuinn M, Spirito A: Thechildren’s sleep habits questionnaire: Constructionand validation of a sleep survey for school-agedchildren. Sleep 2000;23:1043-1051.

3. Bruni O, Ottaviano S, Guidetti MR, et al: The sleepdisturbance scale for children: Construction andvalidation of an instrument to evaluate sleep dis-turbance in childhood and adolescence. J SleepRes 1996;5:251-261.

4. Blader JC, Koplewicz HS, Abikoff H, Foley C: Sleepproblems of elementary school children. A com-munity study. Arch Pediatr Adolesc Med 1997;151;473-480.

5. Owens J, Spirito A, McGuinn M, Nobile C: Sleephabits and sleep disturbance in school-aged chil-dren. J Dev Behav Pediatr 2000;21:27-36.

6. Rona RJ, Li L, Gulliford MC, Chinn S: Disturbedsleep: Effects of sociocultural factors and illness.Arch Dis Child 1998;78:20-25.

7. Kahn A, Van de Merckt C, Rebuffat E, et al: Sleepproblems in healthy preadolescents. Pediatrics1989;84:542-546.

8. Vignau J, Bailly D, Duhamel A, et al: Epidemiologicstudy of sleep quality and troubles in French sec-ondary school adolescents. J Adolesc Health 1997;21(5):343-50.

9. Levy D, Gray-Donald K, Leech J, Zvagulis I, PlessIB: Sleep patterns and problems in adolescents.J Adolesc Health Care 1986;7:386-389.

10. Owens J, Rosen C, Mindell J: Medication use inthe treatment of pediatric insomnia: Results of asurvey of community-based pediatricians. Pediatrics2003;111:e628-e635.


Ch03.qxd 01/28/05 10:27 AM Page 32

11. Chervin R, Archbold K, Panahi P, Pituch K: Sleepproblems seldom addressed in two general pedi-atric clinics. Pediatrics 2001;107:1375-1380.

12. Fallone G, Owens J, Deane J: Sleepiness in childrenand adolescents: Clinical implications. Sleep MedRev 2002;6:287-306.

13. Epstein R, Hillag N, Lavie P: Starting times ofschool: Effects on daytime function of fifth-gradestudents in Israel. Sleep 1998;21:250-256.

14. Wolfson AR, Carskadon MA: Sleep schedules anddaytime functioning in adolescents. Child Dev1998;69:875-887.

15. Giannotti F, Cortesi F: Sleep patterns and daytimefunctions in adolescents: An epidemiological sur-vey of Italian high-school student population. InCarskadon MA (ed): Adolescent Sleep Patterns:Biological, Social, and Psychological Influences.New York: Cambridge University Press, 2003.

16. Morrison DN, McGee R, Stanton WR: Sleep prob-lems in adolescence. J Am Acad Child AdolPsychiatry 1992;31:94-99.

17. Stores G, Wiggs L: Sleep disturbance in children andadolescents with disorders of development: its sig-nificance and management. New York, CambridgeUniversity Press, 2003.

18. Johnson CR: Sleep problems in children with men-tal retardation and autism. Child Adolesc PsychiatrClin N Am 1996;5:673-681.

19. Dahl RE, Ryan ND, Matty MK, et al: Sleep onsetabnormalities in depressed adolescents. BiolPsychiatry 1996;39:400-410.

20. Price VA, Coates TJ, Thoresen CE: Prevalence andcorrelates of poor sleep among adolescents. Am JDis Child 1978;132:583-586.

21. Sadeh A: Stress, trauma, and sleep in children.Child Adolesc Psychiatr Clin N Am 1996;6:685.

22. Field T: Peer separation of children attending newschools. Dev Psychol 1984;20:786.

23. Sadeh A, Horowitz I, Wolach-Benodis L, WolachB: Sleep and pulmonary function in children withwell-controlled stable asthma. Sleep 1998;21:379-384.

24. Rose M, Sanford A, Thomas C, Opp M: Factorsaltering the sleep of burned children. Sleep 2001;24:45-51.

25. Bloom B, Owens J, McGuinn M, et al: Sleep andits relationship to pain, dysfunction and diseaseactivity in juvenile rheumatoid arthritis. J Rheumatol2002;29:169-173.

26. Lewin D, Dahl R: Importance of sleep in the man-agement of pediatric pain. J Dev Behav Pediatr1999;20:244-252.

27. Samuels M, Stebbens V, Davies S, et al: Sleep-related upper airway obstruction and hypoxaemiain sickle cell disease. Arch Dis Child 1992;67:925-929.


Ch03.qxd 01/28/05 10:27 AM Page 33

The sleeping and waking states depend on activ-ity in the entire brain, not in particular or inde-pendent regions. In general, neural activity inthe brainstem diencephalic ascending reticularactivating system is responsible for the mainte-nance of alertness and the waking state. Sleep isa complex, active phenomenon. It is composedof physiologically distinct (but highly synchro-nized) states, and it cycles at regular intervals.Sleep is not simply a reduction in activity of thereticular activating system; rather, it is generatedby many different regions of the central nervoussystem working together.

This chapter focuses on the various anatomiccenters of the central nervous system and theirfunctions in the sleep–wake cycle. Most of the datadelineating these areas have been derived fromanimal lesion experiments. Other data come fromobservations of signs and symptoms in humanswho have suffered from disease or dysfunction ofthese particular topologic regions of the brain.

THE CORTEX

Sleeping and waking behavior occurs in theabsence of the cortex. This has been well docu-mented in animal studies1-4 and in observationsof human newborns with anencephaly and holo-prosencephaly.5 The cortex is involved with theregulation and maintenance of wakefulness bystimulating the ascending reticular activatingsystem, which, in turn, results in a generalarousal response (Fig. 4–1). There is some evi-dence that this cortex–reticular activating sys-tem–cortex activation loop may be responsiblefor the hyperactivity and motor restlessness seenin many children with attention deficit hyperac-tivity disorder or with syndromes associatedwith excessive daytime sleepiness. Evidence ofincreased sleepiness has been documented in agroup of children with attention span problems,

hyperactivity, motor restlessness, behavioral diffi-culties, and school failure.6 Short mean sleep onsetlatencies, frequent sleep onsets, numerousmicrosleep episodes, and pathologically short sleeponset latencies in one or several multiple sleeplatency test naps have been documented. Excessivesleep pressure might explain the effect of stimulantmedication on decreasing activity, motor restless-ness, and improving attention span deficits in somechildren with attention deficit hyperactivity disor-der. By reducing sleepiness, reducing microsleepepisodes, and stimulating the cortex, thecortex–reticular activating system–cortex feedbackloop may be broken, resulting in improvement ofsymptoms. Figures 4–2 and 4–3 are schematic rep-resentations of the mechanisms involved in slow-wave and rapid eye movement sleep.

Behavioral components of sleep onset mayalso be under the control of cortical activity.Habits and rituals associated with sleep onsetcreate a relaxed and habitual situation that per-mits physiologic sleep-onset mechanisms to per-form their function. Absence of appropriate andhabitual sleep onset associations results in sig-nificant sleep onset and maintenance difficultiesduring early childhood.

Although sleep and wake cycling may be pre-served in the absence of the cortex, there is evi-dence that sleep and wake are dysfunctional inmany patients with generalized cortical disease.Some recognizable characteristics remain (e.g.,sleep spindles, vertex sharp transients), butarchitecture may be severely disrupted and, insome cases, no recognizable stages of sleep canbe seen.7-9

THE CEREBELLUM

The cerebellum is mainly responsible for the con-trol of skeletal musculature, posture, and move-ment. During sleep, cerebellar activity and

Anatomy of SleepStephen H. Sheldon

4

35

Ch04.qxd 01/28/05 10:27 AM Page 35

36 Chapter 4 Anatomy of Sleep

Figure 4–1. Schematic mechanism for generation of wakefulness. Solid circles A represent the neu-rons of the reticular formation. Their major ascending projections into the forebrain proceed along twomajor routes. The dorsal route terminates in the nonspecific thalamic nuclei, which in turn project ina widespread manner to the cerebral cortex. The ventral route passes through the subthalamus andhypothalamus and continues into the basal forebrain and septum, where neurons in turn project in awidespread manner to the cerebral cortex and hippocampus. Solid squares B represent catecholamineneurons of the lower brainstem and locus coeruleus (dorsal pons), which contain norepinephrine, andof the substantia nigra and ventral tegmental area (ventral midbrain), which contain dopamine. Thenorepinephrine neurons, mainly implicated in processes of cortical activation, project directly and dif-fusely to the cerebral cortex, as well as to the subcortical way stations. The dopamine neurons, whichare predominantly implicated in processes of behavioral activity and responsiveness, project heavilyinto the basal ganglia and frontal cortex. Solid triangles C represent acetylcholine neurons of the brain-stem reticular formation (including the laterodorsal and pedunculopontine tegmental nuclei in thedorsal pons and midbrain) and basal forebrain (substantia innominata, diagonal band nuclei, and sep-tum). Cholinergic neurons are implicated in cortical activation and from the brainstem project pre-dominantly to subcortical way stations, including the thalamus, the subthalamus, the hypothalamus,and the basal forebrain and septum. The cholinergic basal forebrain neurons project in a widespreadmanner to the cerebral cortex and hippocampus. Not shown are other neuronal systems implicated inwakefulness, including histamine neurons located in the posterior hypothalamus, which also projectdirectly to the cerebral cortex. Glutamate neurons located through subcortical structures and in thecerebral cortex are important in the processes of cortical activation and wakefulness. Multiple peptides,such as substance P, corticotropin-releasing factor, thyrotropin-releasing factor, and vasoactive intes-tinal polypeptide, may be involved in wakefulness and are often colocalized with one of the other pri-mary neurotransmitters, such as norepinephrine and acetylcholine. The neuronal systems implicatedin the maintenance of wakefulness may be involved in primary processes of sensory transmission andattention, motor response and activity, and orthosympathetic and neuroendocrine (particularlyadrenocorticotropic hormone and thyrotropin-releasing hormone) responses and regulation, by whichthey may enhance and prolong vigilance and arousal. (Modified and reproduced with permission fromJones BE: Basic mechanisms of sleep-wake states. In Kryger M, Roth T, Dement WC: Principles and Practiceof Sleep Medicine. Philadelphia, WB Saunders, 1989, p 122.)

Ch04.qxd 01/28/05 10:27 AM Page 36

feedback may be involved with postural adjust-ments and body movements, accessory respira-tory muscle movements, extraocular musclemovements, and alterations in muscle tone.10

There is some contribution of the cerebellum tothe generation of muscle atonia during sleep,11,12

but it seems to have little effect on the sleep–wakecycle itself. The cerebellar vermis may also beinvolved in the generation (or suppression) ofsleep spindles, as lesions of this area result in theproduction of numerous sleep spindles.13

THE BASAL GANGLIA

The role of the basal ganglia in the sleep–wakecycle appears to be peripheral. Dopaminergicneurons in the substantia nigra innervate withthousands of synaptic contacts in the striatum.14

There appears to be little change in the activityof the neurons during wakefulness and duringthe various stages of sleep (including rapid eyemovement sleep).

THE THALAMUS

Animal experiments using low frequency stimu-lation of the midline thalamic nuclei haveresulted in presleep behavior followed bysleep.15,16 The region that causes this response isvery limited, however. Stimulation of the anteriorthalamic nuclei has resulted in similar results.The same is not true, however, in humans. Thethalamus does not appear to be essential forsleep,17,18 but it is essential for the production ofsleep spindles.19 Therefore, a thalamic pace-maker may exist that drives cortical neurons andcontrols electroencephalographic synchroniza-tion during non–rapid eye movement sleep.20

Periodic bursts of high amplitude cortical wavesand spindles become more frequent as deeperstages of sleep appear, and they seem to be gen-erated from nonspecific thalamic nuclei throughdiffuse thalamocortical projections. Areas respon-sible for driving spindle production appear to bemidline and are closely associated to the regionwhere sleep can be promoted in animals by lowfrequency stimulation.21

Sleep spindles and K-complexes are charac-teristic of the sleeping state and are never seenduring wakefulness.22 The functions of spindlesand K-complexes are unknown, but they arethought to have an important function in regu-lating slow-wave sleep.23

THE HYPOTHALAMUS

The posterior, lateral, and medial hypothalamusis the cephalad continuation of the brainstemreticular activating system.24 Transection at thelevel of the posterior hypothalamus completelyabolishes wakefulness.25 Posterior hypothalamiclesions generally produce lethargy and sleepi-ness.26 Hypersomnia, coma, or changes in circa-dian rhythmicity are noted in humans, dependingon the exact site and extent of the hypothalamiclesion.

FOREBRAIN AND ANTERIORHYPOTHALAMUS

Basal forebrain areas appear to contain powerfulinhibitory and sleep-promoting centers.27,28

Lesions of the preoptic region have an effect thatis opposite to those of the posterior hypothala-mus.29 Insomnia and terminal sleeplessness30

have resulted from lesions of the anterior hypo-thalamus and basal forebrain, respectively.36

When stimulated, the basal forebrain pro-duces a short sleep onset latency, encephalo-graphic synchronization, and involuntarysleep.31-33 Lesions of this same area produceinsomnia in animals. It is difficult to clearlyidentify the cell bodies responsible for this phe-nomenon because this area is quite dense andcontains numerous transecting fiber tracts.34

High frequency stimulation of the thalamus,posterior hypothalamus, and brainstem pro-duces arousal, whereas similar stimulation of theanterior hypothalamus produces rapid sleeponset. With high frequency stimulation of thebasal forebrain and low frequency stimulation ofthe thalamus, presleep behavior, rather thansleep itself, occurs prior to encephalographicsynchronization.35

BRAINSTEM RETICULARACTIVATING SYSTEM

Rather than a single, isolated region in the brain-stem, the reticular activating system appears tobe diffusely distributed in the medulla and pons,and it extends into the posterior, medial, and lat-eral hypothalamic regions. Classic experimentsby Moruzzi and Magoun30 revealed that electri-cal stimulation of a portion of the brainstemresulted in changes on the encephalogram thatwere identical to those seen upon awakening

Anatomy of Sleep Chapter 4 37

Ch04.qxd 01/28/05 10:27 AM Page 37


Figure 4–2. Schematic representation of mechanisms generating slow-wave sleep. The Solid circlerepresents neurons of the solitary tract nucleus and adjacent tegmentum, implicated in slow-wavesleep regulation, which project forward into the visceral-limbic forebrain. Solid squares represent sero-tonergic neurons of the brainstem raphe nuclei, which may facilitate the onset of slow-wave sleep andwhich project forward into the rostral tegmentum, thalamus, subthalamus, hypothalamus, and basalforebrain and also from the midbrain directly to the cortex and hippocampus. The Solid triangle rep-resents gamma-aminobutyric acid (GABA) neurons of the hypothalamus and basal forebrain-septum,which project in a widespread manner to the cerebral cortex and hippocampus. These GABAergic neu-rons are also located in the cerebral cortex, where they are maximally active during slow-wave sleep.Not shown are other neuronal systems implicated in slow wave sleep, including adenosine neuronslocated in the hypothalamus. Multiple peptides, such as the opiates, alpha-melanocyte–stimulatinghormone, and somatostatin, may be involved in slow-wave sleep generation and are often colocalizedwith one of the other primary neurotransmitters, such as serotonin or GABA. The neuronal systemsimplicated in the maintenance of slow-wave sleep may be involved in primary processes of sensoryinhibition and analgesia, behavioral inhibition, and parasympathetic and neuroendocrine (notablygrowth hormone) responses and regulation, by which they may facilitate the onset and maintenanceof slow wave sleep. (Modified and reproduced with permission from Jones BE: Basic mechanisms of sleep-wake states. In Kryger M, Roth T, Dement WC: Principles and Practice of Sleep Medicine. Philadelphia, WBSaunders, 1989, p 125.)

Ch04.qxd 01/28/05 10:27 AM Page 38


Figure 4–3. Schematic mechanism for the generation of rapid eye movement (REM) sleep. Aminergic(REM-off ) neurons located in the dorsal raphe nuclei (serotonergic) and the locus coeruleus (nora-drenergic) produce inhibitory postsynaptic potentials on cholinergic (REM-on) neurons located in themesencephalic, medullary, and pontine gigantocellular tegmental fields. These two systems of neuronscontinuously interact to produce the alternation between non-REM and REM sleep. Neurons in thegigantocellular tegmental fields produce stimulating postsynaptic potentials to produce the epiphe-nomena seen during REM sleep. Positive postsynaptic potentials result in conjugate extraocular mus-cle movement through stimulation of the nuclei of the oculomotor, trochlear, and abducens nuclei.Cortical desynchronization occurs via projections to the thalamus and cortex through the ascendingreticular activating system. Muscle atonia is produced by direct stimulation of neurons that produceinhibitory postsynaptic potentials at the level of the anterior horn cell. Intermittent muscle twitchesoccur through direct stimulation of the anterior horn cells. Not shown are interconnections with theautonomic nervous system that result in the cardiac and respiratory irregularities noted during REMsleep. AH, anterior horn cells; am, aminergic neurons; ARAS, ascending reticular activating system;C, cortex; ch, acetylcholine; DRN, dorsal raphe nuclei; GTF, gigantocellular tegmental fields; IRF,inhibitory reticular formation; LC, locus coeruleus; na, noradrenalin; s, serotonin; T, thalamus; +, stim-ulating postsynaptic potential; −, inhibitory postsynaptic potential. (Modified and reproduced with per-mission from Hobson JA: The cellular basis of sleep cycle control. Adv Sleep Res 1974;1:217-250, and HauriP: Current Concepts: The Sleep Disorders. Kalamazoo, Mich, Upjohn Company, 1982, p 13.)

Ch04.qxd 01/28/05 10:27 AM Page 39

from sleep or during an arousal or an alertingreaction. The area from which encephalographicsynchronization occurs runs throughout most ofthe length of the brainstem in its central core.Electrical stimulation in these regions also resultsin changes in activity of cortical and thalamicneurons, abolishes thalamic recruiting potentials,has a marked effect on cortical evoked potentials,and significantly affects perception.11

Evidence exists that many reticular formationneurons have both cephalad and caudad projec-tions.33,37 This complex system also appears tohave many interneuronal connections with thebrainstem deactivating system located in themedullary raphe nuclei.

References

1. Barrett R, Merritt HH, Wolf A: Depression of con-sciousness as a result of cerebral lesions. Res PublAssoc Res Nerv Ment Dis 1967;45:241-276.

2. Kleitman N: Sleep and Wakefulness. Chicago,University of Chicago Press, 1963, pp. 241-242.

3. Oswald I: Sleeping and Waking: Physiology andPsychology. Amsterdam, Elsevier, 1962, p 232.

4. Nielson JM, Sedgwick RP: Instincts and emotionsin an anencephalic monster. J Nerv Ment Dis1949;110:387-394.

5. Puech P, Guilly P, Fishgold H, Bounes G: Un casd’anencephalie hydrocephalique: Étude electroen-cephalographique. Rev Neurol 1947;79:116-124.

6. Wilkus RJ, Farrell DF: Electrophysiological observa-tions in the classical form of Pelizaeus-Merzbacherdisease. Neurology 1976;26:1042-1045.

7. Karacan I, Schneck L, Hinterbuchner LP: Thesleep-dream pattern in Tay-Sachs disease (prelim-inary observations). In Aronson SM, Volk BW(eds): Inborn Errors of Sphingolipid Metabolism.Elmsford, NY, Pergamon Press, 1967, pp 413-421.

8. Parkes JD: Sleep and Its Disorders. Eastbourne, EastSussex, England, WB Saunders, 1985, pp 73-118.

9. Guglielmino S, Strata P: Cerebellum and atonia ofthe desynchronized phase of sleep. Arch Ital Biol1971;109:210-217.

10. Marchesi GF, Strata P: Climbing fibers of rat cere-bellum: Modulation of activity during sleep. BrainRes 1970;17:145-148.

11. Marchesi GF, Scarpino O, Mauro AM: Studiopoligrafico del sonno notturno in pazienti consindrome cerebellare. Arch Psicol Neurol Psichiat1977;4:455-472.

12. Anden NE, Fuxe K, Hamberger B, Hokfelt T: Aquantitative study on the nigro-striatal dopamineneuron system in the rat. Acta Physiol Scand1966;67:306-312.

13. Koella WA: Sleep: Its Nature and PhysiologicalOrganization. Springfield, Ill, Charles C Thomas,1967, p 199.

14. Naquet R, Denavit M, Albe-Fessard D: Com-paraison entre le role du subthalamus et celui desdifferentes structures bulbo-mesencephaliquesdans le maintien de la vigilance. ElectroencephalogrClin Neurophysiol 1966;20:149-164.

15. Bricolo A: Sleep abnormalities following tha-lamic stereotactic lesions in man. In Gastaut H,Lugaresi E, Berti-Ceroni G, Coccagna G (eds):The Abnormalities of Sleep in Man.Proceedings of the XVth European Meeting onElectrophysiology, Bologna, 1967, Bologna,Auto Gaggi, 1968, pp 135-138.

16. Purpura DP, Yahr MD (eds): The Thalamus. NewYork, Colombia University Press, 1966, p 438.

17. Eyzaguirre C, Fidone SJ: Physiology of theNervous System. Chicago, Year Book Publishers,1975, pp 343-371.

18. Naitoh P, Antony-Bass V, Muzet A, Ehrhart J:Dynamic relation of sleep spindles and K-com-plexes to spontaneous phasic arousal in sleepinghuman subjects. Sleep 1982;5:58-72.

19. Whitlock DG, Aruini A, Moruzzi G: Microelectrodeanalysis of pyramidal system during transitionfrom sleep to wakefulness. J Neurophysiol 1953;16:414-429.

20. Rossi GF: Electrophysiology of sleep. In GastautH, Lugaresi E, Berti-Ceroni G, Coccagna G (eds):The Abnormalities of Sleep in Man. Proceedingsof the XVth European Meeting on Electro-physiology, Bologna, 1967, Bologna, Auto Gaggi,1968, pp 13-23.

21. Ranson SW: Somnolence caused by hypothalamiclesions in the monkey. Arch Neurol Psychiatry1939;41:1-23.

22. Mitler MM: Toward an animal model of nar-colepsy-cataplexy. In Guilleminault C, DementWC, Passouant P (eds): New York, Spectrum1976, pp 387-409.

23. von Economo C: Sleep as a problem of localiza-tion. J Nerv Ment Dis 1930;71:249-259.

24. Moruzzi G: The sleep-waking cycle. ErgebPhysiol 1972;64:1-165.

25. Sterman M, Fairchild M: Modification of locomo-tor performance by reticular formation and basalforebrain stimulation in the cat: Evidence forreciprocal systems. Brain Res 1966;2:205-217.

26. Sterman MB, Clemente CD: Forebrain inhibitorymechanisms: Cortical synchronization inducedby basal forebrain stimulation. Exp Neurol1962;6:91-102.

27. McGinty D, Sterman M: Sleep suppression afterbasal forebrain lesions in the cat. Science1968;160:1253-1255.


Ch04.qxd 01/28/05 10:27 AM Page 40

28. Clemente CD, Sterman MB, Wyrwicka W:Forebrain inhibitory mechanisms: Conditioningof basal forebrain induced EEG synchronizationand sleep. Exp Neurol 1963;7:404-417.

29. Bremer F: Cerveau isole et physiologie du som-meil. C R Soc Biol 1935;118:1235-1241.

30. Moruzzi G, Magoun HW: Brain stem reticular for-mation and activation of EEG. Electroen-cephalogr Clin Neurophysiol 1949;1:455-473.

31. McKeough DM: The Coloring Review of Neuroscience. Boston, Little, Brown, 1982, pp30-32.

32. Magni F, Willis WD: Identification of reticular for-mation neurons by intracellular recording. ArchItal Biol 1963;101:681-702.

33. Jasper HH: Diffuse projection systems: The inte-grative action of the thalamic reticular system.Electroencephalogr Clin Neurophysiol 1949;1:405-409.

34. Olszewski J, Baxter D: The Cytoarchitecture ofthe Human Brain Stem. Philadelphia, Lippincott,1954.

35. Cairns H: Disturbances of consciousness withlesions of the brain stem and diencephalon. Brain1952;75:109-146.

36. French JD: Brain lesions associated with pro-longed unconsciousness. Arch Neurol Psychiatry1952;68:727-740.

37. Jefferson M: Altered consciousness associatedwith brain stem lesions. Brain 1952;75:55-67.


Ch04.qxd 01/28/05 10:27 AM Page 41

Children with sleep problems will present withvariations of four sleep symptoms: difficultieswith sleep onset, problems that disrupt sleep, aninability to awaken from sleep at the desiredtime, and daytime sleepiness. The sleep historyis the most important tool for gathering infor-mation on sleep symptoms; and BEARS1 (Bed-time, Excessive daytime sleepiness, Awakenings,Regularity, Snoring) is an easy-to-remembermnemonic device for gathering a history of sleepsymptoms. There are many causes for each ofthese sleep symptoms, and an effective treatmentstrategy for sleep problems is usually based on acareful analysis of the fundamental causes of thesleep problem, not on the sleep symptom. Thereare nine major factors important in the controlof sleep–wake regulation: circadian, homeosta-tic, ultradian, developmental, cardiorespiratory,neurologic, psychiatric/behavioral, drugs/alco-hol, and other medical problems. These factorsform the conceptual foundation necessary tounderstand sleep problems and to develop effec-tive treatment strategies for solving them. Eachof these processes is addressed in depth in achapter or section (noted in italics and in paren-theses) of Principles and Practice of PediatricSleep Medicine,2 so only a brief summary will beincluded here.

Circadian (Section 4) rhythms are self-sustain-ing, nearly 24-hour rhythms, present in all livingorganisms from single-celled algae to mammals,that allow the organism to anticipate the light–darkcycle, not merely respond to it. In mammals thecircadian pacemaker is located in the suprachias-matic nucleus located in the anterior hypothala-mus. The output of the circadian pacemakermodulates the level of alertness; the periodicity ofthe pacemaker in humans is about 24.2 hours, so itmust be corrected each day to remain synchro-nized with the 24-hour day. This “phase shift-ing” occurs through numerous zeitgebers, or time

cues, which facilitate entrainment. These zeitge-bers include clocks, social interactions, meals,and, most importantly, light—the most powerfulcircadian time cue. The sleep–wake cycle is themost obvious facet of the circadian rhythm, butthe output of most other physiologic systemsshow a circadian pattern of predictable changeover the 24-hour day, including: hormone levels(thyroid-stimulating hormone, follicle-stimulatinghormone, lutein-stimulating hormone, mela-tonin, cortisol, antidiuretic hormone), body tem-perature, gastric acid, intestinal motility, drugmetabolism, cardiorespiratory, cellular immunity,and propensity to enter rapid eye movement(REM) and non-rapid eye movement (NREM)sleep. There is a circadian variation in level ofalertness, which is independent of the amount ofsleep an individual had the night before.Different individuals vary in their circadian phasepreferences, preferred sleep time, amplitude ofthe circadian signal, and duration of the circadiansleep period. Morning types prefer to go to sleepearly and awaken early, whereas evening typesprefer to go to sleep late and awaken late. Neitheris better, just different. An individual’s preferredsleep patterns often declare themselves duringchildhood. During adolescence, whatever thepreferred sleep pattern is, there will typically be aphase shift to a later preferred sleep time.

Sleep is also homeostatically (Section 4) reg-ulated based on the duration of prior wakeful-ness. EEG delta power (slow-wave activity0.5-4.0 Hz) is a physiologic index of the sleephomeostatic drive. Sleep duration and timing arelargely the result of the interaction and synchro-nization of the circadian and homeostaticprocesses. Optimal sleep and wakefulness occuronly when these two processes are synchronized.

The ultradian (Section 1) rhythm is definedby the predictable alternation of wakefulness,NREM sleep, and REM sleep during the sleep

Case-Based Analysis of SleepProblems in Children

Gerald M. Rosen, MD

5

43

Ch05.qxd 1/25/05 3:12 AM Page 43

period. The ultradian cycles (NREM-REM) areabout 60 minutes in early infancy and increaseto 90 minutes during early childhood. Sincemany sleep problems are state specific (i.e.,related to wakefulness, REM sleep, NREMsleep), knowing when the specific sleep statesare likely to occur during the night and in whatintervals is often helpful to understanding whatmay otherwise be a confusing sleep history.Examples are sleep terrors (slow-wave sleep),obstructive sleep apnea (worse in REM), condi-tioned arousals in infants (during normal awak-enings), and REM sleep behavior disorder (REMsleep).

Important developmental (Section 1) changesoccur in sleep over an individual’s lifetime inquantity, percent distribution in various sleepstages, circadian rhythmicity and timing, andconsolidation of sleep. At birth, sleep accountsfor about one third of the infant’s life and doesnot have a clear circadian pattern. Sleep in theinfant is described as quiet (NREM precursor),active (REM precursor), or indeterminate.Active sleep accounts for 50% of the infant’ssleep and is the state into which the infant tran-sitions from wakefulness. The ultradian cycles ofinfants are shorter (see ultradian rhythm ear-lier). By 3 months of life, a circadian rhythm inthe distribution of sleep develops such that theinfant begins to sleep more at night and has anestablished daytime nap schedule; total sleeptime decreases to about 15 hours a night; REMand NREM can be scored using adult sleep stag-ing criteria; and REM sleep decreases to about30% of total sleep time. Total sleep time contin-ues to decrease during childhood as nap dura-tion and frequency decrease, with most childrengiving up their daytime naps by 5 years of age.Slow-wave sleep is very prominent throughoutchildhood, and this is likely a factor in someof the common childhood parasomnias, such assleepwalking and confusional arousals. Duringadolescence, sleep requirements probablyincrease, and so does sleepiness. Slow-wave sleepdecreases to the adult norms, and REM sleepaccounts for about 20% of total sleep time.

Alcohol (Sections 5 and 14) and the manyother psychoactive prescription and nonpre-scription drugs can have an important impact onwakefulness and sleep, both when they are beingtaken and while the body is withdrawing fromthem. Caffeine is a ubiquitous food additivefound in many beverages and some foods.

The cardiorespiratory (Section 3) systemundergoes many important changes duringsleep: Upper airway resistance increases; centralrespiratory drive decreases; minute ventilationdecreases; during REM sleep, the accessory res-piratory muscles—both those that hold the air-way open (pharyngeal dilators) and those thatcan expand the chest (intercostals and scalenes)—are hypotonic; hypoxic and hypercapnic ventila-tory responses decrease in NREM sleep andfurther in REM sleep; cardiac output decreasesduring sleep; bronchoconstriction increases dur-ing sleep in everyone but to a greater extent inpatients with asthma, contributing to the com-mon symptom of nocturnal cough; and func-tional residual capacity decreases (especially inobese individuals asleep in the supine position).

Sleep is fundamentally a neurologic (Section2) process controlled by and exclusively for thebenefit of the brain, although its precise functionis not known. Sleep has withstood the evolu-tionary test of time and is present as a distinctneurologic process in mammals and birds and asa distinct behavioral process in fish, amphibians,and invertebrates. An increase in parasympa-thetic tone and a decrease in sympathetic tone(except for phasic increases during REM sleep)are included among the clinically importantchanges in the neurologic systems of mammalsduring sleep. Some seizures occur exclusively dur-ing sleep (generally NREM). Somatic muscles havedecreased tone in NREM sleep and are atonicduring REM sleep, which can lead to profoundhypoventilation in individuals with neuromus-cular diseases associated with muscle weakness.Sleepiness may be a primary neurologic symp-tom, as it is in narcolepsy, idiopathic hypersom-nolence, and Kleine-Levin syndrome, or asecondary neurologic symptom associated withtrauma, infection, or tumors of the central nerv-ous system. The dyskinesia of restless legs (typi-cally occurring at bedtime and relieved bymovement) is another primary neurologic symp-tom affecting sleep.

Sleep involves transitions from wakefulnessthat occur not just at bedtime but also normallyat multiple times during the sleep period.Behavioral (Section 9) issues and psychiatricdisease that affect an individual’s level of arousal(i.e., anxiety, depression, stress, and condition-ing factors) can have a profound impact on achild’s ability to make these transitions smoothlyand quietly.

44 Chapter 5 Case-Based Analysis of Sleep Problems in Children

Ch05.qxd 1/25/05 3:12 AM Page 44

Many other medical problems (Section 13)may be associated with sleep disturbance. Anymedical problem that leads to pain, pruritis, orinflammation may affect sleep.

The nine physiologic factors just discussedform the pathophysiologic underpinning for allsleep symptoms. The sleep process matrix (Table5–1) is a tool for organizing and analyzing howthese fundamental sleep processes interact witheach other and lead to specific sleep symptoms.

As the history of the sleep problem unfolds,the clinician develops hypotheses within each ofthe nine domains: circadian, homeostatic, ultra-dian, developmental, cardiorespiratory, neuro-logic, psychiatric/behavioral, drugs/alcohol, andother medical. A hypothesis is a theory that couldexplain the problem at hand. At the beginning ofan interview, there are usually many hypothesesthat may explain the problem. As the patient’sstory becomes clear, the hypotheses are devel-oped, and questions are formulated, the answersto which will confirm or refute the hypotheses.This process ultimately leads to an analysis anda synthesis of the data, the development of adiagnostic impression, and a treatment plan.This stepwise, sequential process allows compli-cated clinical cases to be broken down intosmaller, more manageable chunks. This model isbased on an approach to clinical problem solvingdescribed by Dr. Vincent Fulginiti in the bookPediatric Clinical Problem Solving.3

The following case illustrates the use of thiscase-based analysis paradigm.

Noah is a 2-year-old boy with frequent night-time awakenings. He has awakened two to fivetimes a night for the past year, and now that hisparents are expecting the birth of their secondchild, they are bringing Noah in for an evaluation.

This much of the history can be gathered in thefirst 2 minutes of the interview in response to thequestion, “What brings you in today?” Theremainder of the 1-hour consultation is spent try-ing to understand what the cause of the problem isand what to do about it. Based on just this amountof information, the clinician develops a number ofhypotheses that might explain the problem, asillustrated in the second row of Table 5–1. At thispoint in the interview, all of these hypotheses areviable explanations for the problem at hand.Clearly, some are more likely than others.

The remainder of the interview is spent con-firming or refuting the various hypotheses bygathering the appropriate history. The data gath-ered in this manner are analyzed and synthe-sized to form a tentative diagnosis, which leadsto a provisional treatment plan. Often, varioussleep processes interact with each other to leadto specific sleep symptoms. In Noah’s case, theevaluative process led to the diagnosis of a bed-time that is too early and a sleep onset associa-tion disorder. However, the story could haveended in several other ways.

This case demonstrates how the case-basedsleep process matrix can be used to describe howthe various fundamental sleep processes can leadto specific sleep symptoms.

Case-Based Analysis of Sleep Problems in Children Chapter 5 45

Ch05.qxd 1/25/05 3:12 AM Page 45

Ho

meo

-D

evel

op

-C

ard

io-

Psyc

hia

tric

/D

rug

s/O

ther

C

irca

dia

nst

atic

Ult

rad

ian

men

tal

resp

irat

ory

Neu

rolo

gic

Beh

avio

ral

Alc

oh

ol

Med

ical

Hyp

o-

1.D

elay

ed-

1.Sh

ort

1.N

orm

al

Sep

arat

ion

Slee

p a

pne

a1.

Seiz

ures

1.

Slee

p o

nset

1.M

edic

atio

n1.

Ecz

ema

thes

essl

eep

pha

se

slee

per

aw

aken

ings

anxi

ety

asso

ciat

ion

side

effe

ct

diso

rder

2.

Inap

pro

pria

te

2.La

te

2.C

ondi

tione

d 2.

RLS/

PLM

S2.

Anx

iety

2.

Caf

fein

e2.

GER

bedt

ime

PMna

paw

aken

ings

di

sord

er

3.Ir

regu

lar

slee

p

3.Pa

rtia

l3.

Fam

ily

3.O

titis

sc

hedu

lear

ousa

lsst

ress

med

ia

4.Pa

inD

ata

to

1.Ti

min

g of

1.

Slee

p

1.Be

havi

or

Oth

er

1.Sn

orin

g1.

Neu

rolo

gic

1.Be

havi

or

Med

icat

ions

ROS

Gat

her

slee

p o

nset

s,

dura

tion

durin

g se

par

a-hi

stor

y aw

aken

ings

, (t

otal

) aw

aken

ings

tio

nsna

ps

2.Li

ght

2.N

ap

2.Ti

min

g2.

Slee

p

2.G

row

ing

2.D

evel

op-

exp

osur

e sc

hedu

leap

nea

pai

ns

men

tin

AM

3.Se

ason

al

3.D

evel

op-

3.Pe

ers

varia

tion

men

t4.

Fam

ily h

isto

ry

4.H

ow is

of

pha

se d

elay

pro

blem

ha

ndle

d an

d by

w

hom

? 5.

Mar

ital

stre

ssA

sses

s-D

elay

ed-s

leep

pha

se w

ith t

oo-e

arly

bed

time/

slee

p o

nset

ass

ocia

tion

diso

rder

men

tTr

eat-

1.Ed

ucat

e p

aren

ts a

bout

circ

adia

n in

fluen

ces

and

norm

al a

wak

enin

gs a

t ni

ght

and

the

imp

orta

nce

of s

leep

ons

et a

ssoc

iatio

ns.

men

t2.

Dis

cuss

par

ents

’ pre

ferr

ed s

leep

tim

e fo

r ch

ild,

give

n th

at t

he c

hild

sle

eps

abou

t 11

{1/2

} ho

urs

per

nig

ht.

Sche

dule

wak

e-up

tim

e at

7:

30 A

M,

nap

with

in a

win

dow

of

1-3

PM,

wak

e up

aft

er 3

PM

rega

rdle

ss o

f ho

w lo

ng t

he c

hild

has

sle

pt.

3.M

ove

bedt

ime

to la

ter.

At

first

beg

in a

t 10

:30

PM,

with

bot

tle a

nd r

ocki

ng.

Put

child

in t

he c

rib (

awak

e bu

t tir

ed)

and

follo

w

deco

nditi

onin

g p

roto

col s

o th

at c

hild

can

ulti

mat

ely

tran

sitio

n to

sle

ep o

n hi

s ow

n.

GER

, ga

stro

esop

hage

al r

eflu

x; R

LS/P

LMS,

res

tless

legs

syn

drom

e/p

erio

dic

limb

mov

emen

ts d

urin

g sl

eep

; RO

S, r

evie

w o

f sy

stem

s.

Tab

le 5

–1.

Slee

p P

roce

ss M

atri

x f

or

a 2-Y

ear-

Old

Bo

y w

ith

Sle

ep D

iffi

cult

y

46

Ch05.qxd 1/25/05 3:12 AM Page 46

Case-Based Analysis of Sleep Problems in Children Chapter 5 47

References

1. Mindell JA, Owens JA: A Clinical Guide toPediatric Sleep: Diagnosis and Management ofSleep Problems. Philadelphia, Lippincott, Williams& Wilkins, 2003.

2. Kryger MH, Roth T, Dement WC: Principles andPractice of Pediatric Sleep Medicine, 3rd ed.Philadelphia, Saunders, 2000.

3. Fulginiti V: Pediatric Clinical Problem Solving.Baltimore, Williams & Wilkins, 1981.

Ch05.qxd 1/25/05 3:12 AM Page 47

Polysomnography is the term used to describe aprocedure of objective, simultaneous recordingof many different physiologic parameters duringsleep. Practitioners who treat children with sleepproblems should be familiar with the generalmethods and techniques used in the sleep labo-ratory. Practical issues such as patient prepara-tion before the study, various methods andsystems used to monitor physiologic parameters,and various components of the assessment of thepolysomnogram and communication of the resultsrequire understanding.

PHYSIOLOGIC PARAMETERSMONITORED DURING SLEEP

ElectroencephalogramInformation from the electroencephalogram(EEG) forms the basis for differentiating thestages of sleep. Continuous monitoring of theEEG throughout the night also provides ongoinginformation about the development and integrityof the central nervous system. Therefore, it isnecessary to use standard methods for accurateand reliable recording.

Development of ElectroencephalographicActivity in the Fetus

The development of the diencephalon is essen-tial for the establishment of centers that are fun-damentally responsible for the control of thesleep–wake cycle and cycling within the sleepstate. The diencephalon is prominent during thesecond month of development; however, itbecomes concealed by the greater expansion ofadjacent parts of the developing brain. All neu-ronal impulses that eventually reach the cortexpass through the diencephalon, with the excep-tion of those originating from olfaction. The third

ventricle lies within the diencephalon. A smallarea in the caudal wall of the third ventriclebecomes evaginated during the seventh week ofdevelopment and forms the pineal body, whichin turn eventually becomes conical, solid, andglandular. Melatonin will eventually be secretedby this structure.

After week 7 of gestation, three main regionsof the diencephalon can be identified: theepithalamus dorsally, the thalamus laterally,and the hypothalamus ventrally. The thalamusrapidly outgrows the epithalamus, which ulti-mately becomes a synaptic region for olfactoryimpulses. Neuronal fibers separate the gray mat-ter of the walls of the thalamus into the variousthalamic nuclei. Similarly, the walls of the hypo-thalamus contain hypothalamic nuclei, the opticchiasm, suprachiasmatic nuclei, and the neurallobe of the stalk and body of the pituitary gland.The hypothalamus eventually becomes theexecutive region for the regulation of all auto-nomic activity.

The telencephalon is the most rostral subdivi-sion of the brain. Two lateral outpouchings(evaginations) form the cerebral hemispheres,each containing a lateral ventricle. The cerebralhemispheres first begin to become prominentduring week 6 of conceptional age and expandrapidly until they overgrow the diencephalonand mesencephalon during midgestation.

Development of Brain Electrical Activity(Electroencephalogram)

Neuronal electrical activity is essential for cellu-lar migration, dendritic branching, and synapticfacilitation. The development of the EEG con-sists of the recording of summation potentials atthe skin surface from underlying brain tissue.The underlying activity is influenced by distantportions of the central nervous system. The

Polysomnography in Infants and Children

Stephen H. Sheldon

6

49

Ch06.qxd 1/25/05 3:14 AM Page 49

development of the EEG in the fetus, newborn,and child is rapidly changing. Abrupt and strik-ing variations occur between 24 weeks of gesta-tion and 3 months post-term. The appearanceof the surface EEG of the premature infant isdependent on postconceptional age.

WEEKS 24 TO 28 OF CONCEPTIONAL AGE

Although some neuronal electrical activity ispresent before 24 weeks, wakefulness and sleepcannot be clearly differentiated by behavioral orEEG characteristics. At this time, the EEG revealsa discontinuous pattern (Fig. 6–1).

Periods of limited electrical activity lasting upto 3 minutes are separated by bursts of activitylasting up to 20 seconds. This activity is typicallysymmetrical and rhythmic and may be composedof alpha, theta, and delta frequencies. Sharpwaves are common at these conceptional ages

and are normal when they are frontal in locationor are sporadic in any location. Sharp waves andspikes that are sporadic may be normal up toshortly after 40 weeks of gestation.1,2 Abnor-malities should be considered if the spikes orsharp transients are repetitive or unilateral oroccur during the quiescent period of the discon-tinuous pattern.


The previously noted discontinuous activity pat-tern continues. However, the bursts of activity, aswell as the intervals between them, are short-ened. Rhythmic delta waves predominate duringthe bursts. Frequently, 10 to 20 Hz of activity issuperimposed on these slow waves and persistsuntil the conceptional term. This fast, superim-posed activity comprises what is referred to asdelta brushes (Fig. 6–2).

50 Chapter 6 Polysomnography in Infants and Children

Figure 6–1. Recording of a premature infant with a discontinuous EEG pattern. This 30-secondpolysomnographic segment represents this pattern, which consists of intermittent bursts of high-volt-age, slow-wave activity with somewhat prolonged periods of electrical quiescence between the bursts.The presence of delta brushes depends on the youngster’s age. During the periods of relative electri-cal quiescence, there is a paucity of observable electrical activity.

Ch06.qxd 1/25/05 3:14 AM Page 50


Presently, two patterns of EEG may be seen atdifferent times during the recording. The first isthe typical discontinuous pattern of slow wavesin the range of one to two cycles per second,similar to the previous pattern of a tracé alter-nant. This pattern is characteristic of quiet sleep.The second EEG pattern is seen during wakeful-ness and active sleep and consists mainly of con-tinuous, synchronous, rhythmic, and generalizedwaves of 1 to 2 Hz.


As the conceptional term approaches, three EEGpatterns can be identified. A quiet sleep pattern ofdiscontinuous activity continues, but the relativelyquiescent intervals and slow-wave bursts are per-sistently shortened. A second pattern consists ofcontinuous irregular waves in the theta and deltafrequencies appearing during wakefulness andactive sleep. The delta brushes tend to disappear.

A pattern of diffuse, irregular slow waves withan amplitude of less than 50 μV is an alternativeto the previous pattern defining wakefulness andactive sleep. At the conceptional term, four EEGpatterns can be identified.

CONCEPTIONAL TERM TO 3 MONTHS OF AGE

At the conceptional term, the trace-alternantpattern defines quiet sleep (Fig. 6–3). This isa modification of the previous discontinuouspattern and consists of bilateral bursts of high-amplitude slow waves lasting 4 to 5 seconds alter-nating with periods of low-amplitude activity ofa similar duration. During the bursts of slowwaves, frontal spikes may appear. The trace-alternant EEG pattern of quiet sleep generallydisappears between weeks 3 and 4 of the post-conceptional term.

During this developmental period, a new pat-tern emerges during quiet sleep that appears to rep-resent the continued development of non–rapid

Polysomnography in Infants and Children Chapter 6 51

Figure 6–2. Recording of a patient with a discontinuous EEG pattern. Again, this pattern shows high-voltage, slow-wave activity separated by periods of relative electrical quiescence. The electrical activ-ity between bursts is somewhat increased; delta brushes can be seen, and they are of relatively lowvoltage (10–20 Hz) superimposed on the slow waves during bursts of this activity.

Ch06.qxd 1/25/05 3:14 AM Page 51

eye movement (NREM) sleep. Continuous high-amplitude slow waves of 0.5 to 2 Hz appear (Fig.6–4). They are seen most often in the posteriorregion of the head. Significantly faster activity of18 to 25 Hz is often superimposed on this slow-wave activity.

Continuous medium activity is the most com-mon EEG pattern seen during active sleep andwakefulness (Fig. 6–5). This pattern consistsmainly of fairly rhythmic waves of 4 to 8 Hz andless than 50 μV. In addition, another patternoccurs during active sleep that consists of con-tinuous low-amplitude, mixed slow waves withsuperimposed larger delta activity that is eitherintermittent or continuous.

Sleep–wake transitions are generally rapidduring this maturational period. Certain behav-iors may continue into sleep (e.g., sucking andswallowing [Fig. 6–6]).

The most consistent behavior correlated withsleep in the newborn is persistent eye closure.Infants primarily enter sleep actively. Sleep-onsetREM begins to shift to sleep onset through NREMat approximately 3 months of chronologic age.However, sleep-onset REM may continue up toabout 6 months of age. Active and quiet sleepperiods have cycles of about 45 to 60 minutes.During the conceptional term, approximatelyhalf of the total sleep time consists of active sleepand half quiet sleep.

Before 40 weeks of gestation, smooth alter-ations between active and quiet sleep may be dif-ficult to identify. Mixed-state characteristics arepresent, and this state is called indeterminate sleep.

At this time sleep spindles are poorly definedand shifting, central sharp waves may begin toappear. Synchrony is variable; however, it beginsto increase between bursts of the discontinuous


Figure 6–3. Recording of a trace alternant EEG pattern. A trace alternant EEG pattern is a continua-tion of the development of this discontinuous pattern into a more mature form. There are still burstsof high-voltage, slow-wave electrical activity separated by periods of lower-voltage activity (somewhatrelative electrical quiescence compared with the bursts). This EEG pattern is characteristic of quiet sleep(equivalent to non–rapid eye movement [NREM] sleep) in the infant and lasts from about 36 weeksafter conceptional term to 44 weeks after conception. Note that respiration during this state is quitemonotonous and regular. Heart rate is also very stable and regular, although normal beat-to-beat vari-ability is present.

Ch06.qxd 1/25/05 3:14 AM Page 52

sleep pattern and is nearly completely synchro-nous by the conceptional term.

During wakefulness and sleep, EEG phenom-ena may be present that might be consideredabnormal in older children. Nonfocal sporadicspikes, frontal sharp transients, anterior slow-waveactivity, and transient asymmetries are not consid-ered abnormal in the neonate. However, persistentphenomena (e.g., frequent focal spikes, persist-ent asymmetry of activity) may be abnormal, espe-cially if they are associated with a history ofpotential abnormality or physical/behavioral find-ings that suggest abnormality.

THREE TO 12 MONTHS OF AGE

This period of development of the infant revealsstriking maturational changes, as indicated onthe EEG. A clear pattern of wakeful activitybegins to emerge. Generalized delta and thetaactivity appears and is often more prominent inthe posterior EEG channels. During the first year

of life, slow-wave activity during wakefulnessbecomes more rhythmic and diminishes consid-erably. From 3 to 5 months of age, theta activityincreases in prominence in the central and pos-terior regions and becomes the dominant wakingrhythm after 5 months. The absence of this thetaactivity at this stage of development may be con-sidered abnormal.1,2 At 3 months of age, a rela-tively high voltage (50 to 100 μV or greater) of3 to 4 Hz of occipital activity is present. By5 months of age, the frequency increases toabout 5 Hz and continues to increase throughoutthe first year to about 6 to 8 Hz at 12 months ofage. The voltage of this activity also decreases byabout 25% over this period of time.

During this period of development, sleep-wake transitions are relatively smooth. As thetransition begins, the amplitude of waveformsincreases and the frequency decreases. Rhythmicand synchronous activity of 75 to 200 μV and3 to 5 Hz becomes predominant.


Figure 6–4. Recording of continuous high-voltage activity during quiet sleep. At approximately 44weeks after conception, the discontinuous pattern begins to disappear and high-voltage, slow-waveactivity is noted throughout the record. Again, note that respiration is quite monotonous, which is alsocharacteristic of quiet sleep.

Ch06.qxd 1/25/05 3:14 AM Page 53

Major changes in the sleep EEG occur overthe first year of life. The trace alternant in thenewborn is gradually replaced by a continuousslow-wave pattern by 3 months. Clear sleepspindles consisting of waxing and waning andmedium-amplitude synchronous waves of about12 to 14 Hz appear at about 6 to 8 weeks of age.They first appear in the central regions but thenexpand in distribution. Between 2 and 6 months,sleep spindles assume mature characteristicsand appear almost continuously throughoutNREM sleep. The complete absence of sleepspindles between 3 and 6 months is most likelyabnormal. Before 6 months, sleep spindles maybe asymmetrical over the two hemispheres.They slowly become synchronous between6 and 8 months. The persistence of asymmetricalspindles after 12 months of age may representa unilateral decrease in the electrical activity ofthe brain.

Sleep spindles are a ubiquitous phenomenonin the sleep of older children and adults.However, their physiology and the effects of neu-rologic disorders on their frequency and ampli-tude are incompletely understood. The differencesin spindle frequency may be due to underlyingencephalopathy and physiologic differencesbetween partial and generalized epilepsy as wellas the possible residual effects of a variety ofanticonvulsant medications.3,4

Spindle patterns (density, duration, frequency,amplitude, asymmetry, and asynchrony) developquite rapidly between 1.5 and 3 months of age.This rapid development most likely reflects thedevelopmental changes in thalamocortical struc-tures.5,6 Spindle expression varies in relation toascending reticular activating tone, constitutinga functionally inhibitory thalamocortical responseto neurophysiologic conditions that promotecentral activation.1,7 A density of 12 to 14 Hz of


Figure 6–5. Recording of active sleep in the neonate. EEG voltage is moderately high compared witholder children and adults, and the frequency is somewhat slower. Note the characteristic morphologyof the EEG activity. There is normal beat-to-beat variability in heart rate, and respiration during thisstate is unstable and variable, with associated brief periods of erratic activity including but not limitedto apnea, hypopnea, and tachypnea.

Ch06.qxd 1/25/05 3:14 AM Page 54

activity is greater in stage 2 than slow-wavesleep.5,6 Three months of age appears to be a sig-nificant juncture in the maturational process.Sleep spindle evolution seems to be an accuratereflection of the significant maturation of centralnervous system processes and the resultingbehaviors that occur at 10 to 12 weeks of age andin the development of NREM states and slow-wave sleep. Concordance between the quantita-tive aspects and nocturnal organization ofindividualization of slow-wave sleep in infantsoccurs from about 4.5 months of age.5,6

At 3 months of age, clear differentiation ofNREM sleep states by EEG criteria is quite diffi-cult. However, by 6 months, NREM sleep can typ-ically be differentiated into three distinct states

(stages 1 and 2 and slow-wave sleep), and theEEG takes on a more mature pattern. Althoughrudimentary-vertex sharp transients and K-com-plexes can be identified in the neonatal period,they typically appear for the first time between5 and 6 months of age. Vertex waves are generallyseen during the lighter stages of sleep (stages1 and 2), are of high amplitude (up to 250 μV)and negative polarity, and last less than 200 msec.K-complexes are similar to vertex waves but areconsiderably slower (lasting at least 0.5 second).K-complexes are often followed by a sleep spindleand have a wide distribution about the vertex.Both vertex waves and K-complexes can occur inshort bursts or appear spontaneously or mayoccur in response to a sudden sensory stimulus.


Figure 6–6. Recording of a sucking artifact. Non-nutritive sucking is common in infants and smallchildren. The sucking artifact is often seen on the chin muscle electromyogram (EMG) when theyoungster is sucking on a pacifier or bottle. This sucking behavior can persist into transitional sleep andmay be seen in other sleep stages. Chin muscle EMG in this 30-second epoch reveals periodic andepisodic crescendos of muscle activity that at times are rhythmic. There may appear to be changes inairflow during this time, since the sensor (either a thermistor or cannula) can be moved by the flangeof the pacifier either deeper into the air stream or out of the air stream. In the assessment of nasal air-flow, care must be taken when the baby is sucking on a pacifier. A physiologic consequence of thechange in airflow must be apparent for an abnormal respiratory event to be considered in the pres-ence of a sucking artifact.

Ch06.qxd 1/25/05 3:14 AM Page 55

The EEG during REM sleep also undergoessignificant changes, with a decrease in the ampli-tude and a slight increase in and mixture of fre-quencies. Sawtooth waves appear, and theelectrical pattern gradually begins to resemble amore mature, relatively low-voltage, mixed-frequency pattern. By 3 to 5 months of age thepercentage of REM sleep decreases, constitutingabout 40% of the total sleep time. By the end ofthe first year of life, REM sleep equals about 30%of the total sleep time.

Changes During Early and Middle Childhood

Slow, consistent, and continuous changes occurduring this period of development. However,these changes are more subtle, evolve over alonger period of time, and become more consis-tent and reproducible.

Waking rhythms with the eyes closed increasegradually from about 5 to 7 Hz of relatively high-voltage activity to the typical sinusoidal 8- to 12-Hz frequency of low–to–medium voltage activitycharacteristic of the adult pattern. They are mostprominent over the occipital regions of the head.These developing alpha rhythms are characteris-tic of relaxed wakefulness with the eyes closed.It is easily attenuated with eye opening, focusingattention, or increased vigilance. Well developedalpha activity is present in most normal childrenby 8 years of age. The amplitude of alpha activ-ity gradually increases during early and middlechildhood but remains at a relatively low tomoderate amplitude during puberty, adoles-cence, and adulthood.1,2

Transitional sleep patterns of drowsiness canbe identified after 1 to 2 years of age. This sleep-wake transition becomes more mature and alphaactivity diffuses and becomes admixed, withslower, mixed-frequency activity. Brief microsleepepisodes occur and become more frequent dur-ing drowsiness and transitional sleep (stage 1).These microsleep episodes become longer andconsolidate, alpha activity disappears from theEEG, muscle tone may decrease slightly, andslow-rolling eye movements are clearly evident.

Slow, high-voltage activity, which is veryprominent during infancy and early childhood,decreases significantly throughout early andmiddle childhood. Diffuse, synchronous thetaactivity is very prominent between the ages of

1 and 4 years. This activity begins to diminishafter the age of 4 years, and by 5 to 6 years of agealpha activity is about equally prominent. After6 years of age, alpha activity becomes the pre-dominant waking rhythm. However, hypersyn-chronous theta activity (Fig. 6–7) during stage1 and early stage 2 sleep is very common duringthis phase of development. Indeed, even atheta–delta pattern (Fig. 6–8) can frequentlybe identified during slow-wave sleep in somechildren.

All stages of sleep are easily discernible dur-ing middle childhood. NREM sleep states can beclearly differentiated and become more similarto adult stages 1 to 4. Although the frequenciesare somewhat slower in children and graduallyincrease to those in adults, the most strikingdifference is in the higher amplitude of wave-forms at all frequencies until puberty, wherea more adult pattern is notable. From earlythrough middle childhood, the structure of sleepas recorded on the EEG also assumes a moremature adult characteristic.3,8 During laterinfancy, sleep cycles last about 40 to 50 minutes.This cycle length increases gradually to about60 minutes by 18 to 24 months and to 90 min-utes by 5 years of age. The EEG activity of REMsleep also assumes adult characteristics andis cycled regularly with NREM sleep. After 3months of age, most infants will shift from enter-ing sleep through the REM pattern to enteringsleep through the NREM pattern. This transitionis generally complete by about 6 to 8 monthsof age. The percentage of REM sleep also gradu-ally continues to decrease, from 50% at theconceptional term to 30% by 1 to 2 years andto 20 to 25% by 3 to 5 years of age.

The standard International 10-20 Systemallows for symmetrical, reproducible lead place-ment; comparison of EEGs from the same patientand from different patients; and recordings fromthe same or different laboratories (Fig. 6–9, A-C).Improper electrode placement on the scalp canlead to problems with the overall validity, evalua-tion, and scoring of sleep stages.

Sleep stage scoring has been standardizedusing a limited, referential EEG montage. In mostsleep laboratories, a central recording site (C3 orC4) is coupled with a referential site (usually theopposite earlobe or mastoid process). Althoughonly one channel of the EEG is necessary foridentification of the NREM stages in adults,


Ch06.qxd 1/25/05 3:14 AM Page 56

many laboratories use more comprehensivederivations to obtain a more complete EEGassessment. The addition of an occipital record-ing channel is helpful in demonstrating alphaactivity in older children. A coronal or parasag-ittal bipolar montage may be helpful in identi-fying nocturnal seizure disorders. A completeEEG montage and recording strategy may beused to assist in the localization of abnormalEEG activity.

Electro-OculogramThe recording of eye movements during sleepprovides important information regarding theidentification of a sleep state. Fixation of theeyes on exogenous stimuli follows a clear devel-opmental course. Visual fixation on a single pat-

terned stimulus has been recorded in prematurenewborns as early as week 30 of conceptionalage.2,5 The corneal reflection technique haspermitted determination of the waking states(crying, quiet wakefulness, and drowsiness).Younger infants seem to spend more time indrowsiness, whereas older infants spend moretime in quiet wakefulness as determined by fix-ation. Fixation time during quiet wakefulnessincreases with increasing gestational age, andfixation occurs more frequently in quiet wake-fulness than drowsiness.

Measuring eye movements to determine thestate of wakefulness has been overshadowedby the electro-oculogram (EOG) in determin-ing a sleep state. The evaluation of eye move-ments during sleep is clearly required for thedetermination of the REM sleep state.


Figure 6–7. Recording of hypersynchronous theta activity in stage 2 sleep. This is a 30-second epochrecorded from a 3-year-old child. Note the predominance of theta activity in this EEG. A slow back-ground and theta preponderance can be seen in toddlers, making the scoring of these epochs for sleepstage difficult. However, in this epoch, transients that suggest K-complexes and slow spindle-like activ-ity can be seen. Therefore, this epoch was scored as stage 2 sleep. Coincidentally, there is a mild snor-ing artifact noted on the chin muscle EMG, and the end-tidal CO2 is quite elevated. The persistenceof this pattern of respiration might be indicative of alveolar hypoventilation, which may be character-istically obstructive. Full assessment of the complete polysomnogram would be required.

Ch06.qxd 1/25/05 3:14 AM Page 57


Figure 6–8. Recording of theta-delta sleep. Theta activity in tod-dlers and small children canpredominate and be seen through-out the record. This 30-secondepoch depicts theta–delta sleep, inwhich there is considerable thetaactivity superimposed on deltaactivity during slow-wave sleep.

Figure 6–9. A, International 10-20 System of EEG electrode placement. This system allows for sym-metrical, reproducible lead placement within and between subjects. It provides for the comparison ofEEGs from the same patient and from different patients and the comparison of recordings from thesame or different laboratories. From top to bottom, the standard EEG layout begins on the left sideand progresses to the right side and begins at the front of the head and progresses to the back of thehead. L, left; R, right. B, Bipolar transcoronal electrode array. The montage consists of A1-T3, T3-C3,C3-CZ, CZ-C4, C4-T4, and T4-A2. There is also an occipital channel (CZ-O2). This electrode array isstandard in some laboratories. It provides greater information than the traditional referential layout ofa polysomnogram, since it can differentiate between sharp waves and spikes that may be epileptic innature and the vertices of sharp waves. It can also provide information regarding hemispheric differ-ences in electrical activity. L, left; R, right. C, Expanded transcoronal and parasagittal electrode array.It consists of the transcoronal array and additional electrodes over the right and left hemispheres thatcan more accurately localize questionable electrical activity as anterior or posterior. Nonetheless, thegold standard for the EEG is a complete International 10-20 System electrode array that can accuratelyidentify and diagnose abnormal electrical activity.

Ch06.qxd 1/25/05 3:14 AM Page 58

The eye is a functional dipole, with the corneabeing 75 to 100 μV positive compared with theretina. By using EOG techniques, conjugate andnonconjugate eye movements can be recordedwith the eyelids closed. Recording electrodes areconventionally placed at the outer canthus ofeach eye and offset from the horizontal plane byapproximately 0.5 to 1.0 cm. These distancesmay be modified depending on the size of theinfant. Offsetting the electrodes from the hori-zontal plane (one above and one below) is cru-cial for the identification of vertical and obliqueeye movements.

Eye movements occur in saccades duringwakefulness and in saccadic bursts during activesleep. Slow-rolling and dysconjugate eye move-ments can occur during quiet wakefulness,drowsiness, and sleep–wake transitions and incertain medical/surgical abnormalities (e.g.,extraocular muscle paralysis, brainstem anatomicabnormalities, disorders that affect the cranialnerves). Since retinal development in prematureinfants involves migration of the vasculature fromthe region surrounding the optic disc toward theperiphery as well as the development of centralcontrol of eye movements, the EOG may revealboth saccadic and dysconjugate eye movementsin younger premature infants. Since the charac-teristics of the electromyogram (EMG) may notbe as reliable as the EEG and EOG, the identi-fication of REM sleep requires a characteristicEEG pattern and the presence of at least somesaccadic eye movements.

The development of clear conjugate eyemovements during wakefulness depends to a greatextent on the stage of retinal development andthe maturational level of fixation. However, thesaccadic eye movements during sleep do notrequire retinal development or conscious fixa-tion. Control may be mediated through a num-ber of different but interrelated central neurologicmechanisms that include, but are not limited to,cerebral and cerebellar cortical and subcorticaldevelopment, oculomotor/trochlear/abducensnuclei development, and the maturation of cen-tral vestibular nuclei. The development of thevestibular system is important for fixation dur-ing wakefulness and ocular saccades duringactive sleep.1,8

Ocular movements have been identified bytraditional visual inspection of saccades beneaththe closed eyelid. This technique may have spe-

cific limitations, especially in the sleep labora-tory. EOG is considerably more reliable and cost-effective, and eye movements can be identifiedby post hoc inspection of the polysomnographicrecording. Unfortunately, the specific limitationsof EOG are present in smaller infants. The eyesof premature neonates present weaker dipoles,and there may be considerable difficulty withsubjective analysis. In an attempt to resolvethese obstacles, automated methods for theanalysis of REMs during active sleep in infantshas been described.3,9 Computer analysis ofslope and amplitude threshold criteria have beenperformed. Digital filtering is used to improvethe effectiveness, identification, and quantifica-tion of ocular movements. The method has beenshown to be highly reliable and useful in differ-entiating between tonic and phasic states duringactive sleep. This differentiation of REM sleepinto two distinct states by the presence or absenceof phasic muscle and extraocular muscle activitymay hold specific significance for the evaluationof studies of sleep-related cardiorespiratoryphysiology during infancy. In healthy adults, thedifferent physiologic characteristics of tonic andphasic REMs can be clearly identified.5,10 Thismay also be true of infants.

Dream-related and non–dream-related ocularmovements occur in both humans and animals.7,11

It appears that only high-amplitude REMs andeye movements occurring in bursts have anypossibility of corresponding to visual images.Intense REMs during sleep have been investi-gated as a possible indication of a delay in theneural development of infants.8,12 Becker andThoman evaluated the occurrence of “REMstorms” in first-born infants from the secondthrough the fifth postnatal week and again at3, 6, and 12 months of age. The amount of REMswithin each 10-second interval of active sleepwas rated on a scale based on the frequency andintensity of eye movements. Bayley scales of men-tal development were administered to the cohortof infants at 12 months of age. A significant neg-ative correlation was found between the fre-quency of REM storms and Bayley scores. By6 months of age, REM storms seemed to expressdysfunction or delay in the development of cen-tral inhibitory feedback control for sleep organi-zation and phasic sleep-related activities.

Although few studies exist in infants and chil-dren, eye movement density has been shown to


Ch06.qxd 1/25/05 3:14 AM Page 59

be significantly decreased during the second andthird REM periods after a night of complete andpartial sleep deprivation. The rebound of slow-wave sleep is generally confined to the firstNREM sleep period and increases in activityamplitude of 0.05 to 3 Hz.

Piezoelectric strain gauge transducers havealso been used to evaluate eye movements dur-ing sleep.10,13 These sensors have been shown tobe highly sensitive to fine microtremor activityduring wakefulness. This micronystagmoidactivity diminishes during NREM sleep andincreases during REM sleep. It is very intriguingto note that microtremor ocular activity alsoincreases after the presentation of an auditorystimulus to subjects during NREM sleep. Similarincreased microtremor activity also occurs withthe appearance of K-complexes in the EEG.

Electromyogram and Movement Activity During SleepAlthough EMG activity during sleep can beassessed using any number of skeletal musclegroups, it has become customary to record mus-cle tone from the chin. Certain sleep disorders(e.g., bruxism, REM sleep behavior disorder)may present with unusual muscle activity thatcan be recorded by surface EMG. Additionalinformation may be provided regarding thepatient’s sleep–wake behavior, permitting quan-tification of arousal responses and movementsduring sleep.

In general, three recording electrodes areapplied to the chin, although a recording isobtained from only two. The extra electrodeis placed to provide a backup in the event anelectrode becomes displaced during the record-ing. Electrodes filled with electroconductivecream are taped to the skin in the submentalregion. This procedure effectively records theactivity of superficial muscles as well as thegenioglossus muscle.

EMG activity from other muscles (e.g., ante-rior tibial) may be simultaneously recorded forthe evaluation of abnormal or paroxysmal move-ments during sleep.

Body movements are characteristic of thesleep state. Both short-wave sleep and REM sleepgenerally end with a body movement or abruptincrease in chin muscle EMG activity. The nextsleep cycle follows.

Intermittent body movements frequentlyoccur during sleep without a significant changein the state or beginning of a new sleep cycle.These movements without a state change involveboth brief phasic activity lasting less than 6 sec-onds and gross slower body movements (squirm-ing) lasting longer than 6 seconds. Maturationof this skeletal muscle activity also follows a reg-ular and predictable pattern. During wakeful-ness, gross motor and fine motor activities formthe basis for half of the developmental screeningtest. Maturation of gross motor, fine motor, andphasic muscle movements during sleep may alsobe a sensitive marker of neurologic development.

During quiet wakefulness, random muscleactivity usually does not occur. However, a widerange of elementary and complex motor activi-ties are known to occur during sleep. Minute,random electrical activity constitutes the basicphysiologic condition of the skeletal musclesduring sleep.11 In NREM sleep, minute, randommotor activity decreases considerably whencompared with the quiet waking state. DuringREM sleep, there is a sudden increase in isolatedmotor unit action potentials. Particular struc-tural features of the anterior tibial muscle makeit seemingly the most active muscle duringsleep,11 and monitoring of the anterior tibialEMG activity during sleep is recommended dur-ing nocturnal polysomnography in all neonates,infants, children, adolescents, and adults.

Although a significant decrease in the numberof body position shifts occurs during sleepacross the life cycle,12,15 in middle childhoodthere do not appear to be differences in the num-ber and frequency of major body movementsand position shifts in normal youngsters. Duringearly childhood, there appears to be an increasein the duration of maintenance of body positionand in the number of periods of more than 30minutes of positional immobility. In middlechildhood, prone, supine, and lateral positionsoccupy equal proportions of the sleep period.Sleeping in the prone position is characteristi-cally abandoned as maturation continues.

Characteristic body and muscle movementsin premature newborns and term neonates dur-ing quiet sleep appear similar to startle (Moro’s)reflexes, generalized phasic movements,14 and atonic increase in submental muscle activity. Incontrast to the total simultaneous pattern ofmotor activity seen during quiet sleep, active


Ch06.qxd 1/25/05 3:14 AM Page 60

sleep is characterized by a more uncoordinatedand localized pattern of movements such asgeneralized phasic movements, localized tonicactivity, localized phasic activity, and briefclonic muscle activity. With increasing concep-tional age, there is a decrease in the occurrenceof the startle reflex, generalized phasic move-ments, and localized phasic movements. Local-ized tonic activity tends to remain constant withincreasing conceptional age. The differences indecreasing tendencies among these movementsmay indicate the variation in maturationalchanges for different parts of the central ner-vous system.13

The degree of phasic activity during sleepmay reflect the maturity of the developing brain-stem. Dissociation of sleep-related phasic motorphenomena may hold special clinical signifi-cance for some infants.14,16 On the other hand,the types of skeletal muscle movement showspecific developmental relationships withrespect to conceptional age. Fukumoto and col-leagues17 described the ontogenetic evolution ofthree types of movement: gross, localized , andphasic (twitches lasting less than 0.5 second).All body movement parameters decreased in fre-quency with maturation. Interestingly, each typeof movement behavior showed a fairly specificand individual time course. The earliest decreasebetween week 30 of conceptional age and 18months of chronologic age occurs in phasicactivity, which takes place relatively early in thedevelopment of the infant. Localized movementsdecrease in frequency next, and gross movementcontinues unchanged until a basal level isreached at approximately 9 to 13 months ofpostnatal age. The number of epochs withoutbody movements increases steadily until about8 months of age.

Gross, localized, and phasic movements arecontrolled by the central nervous system at dif-ferent organizational levels. Phasic activity ismore ontogenetically simple and decreases earlyduring maturation. Gross movements are quitecomplex and require a greater degree of centralnervous system integration. Therefore, thesetypes of movement are correlated with the mat-urational processes of the central nervous sys-tem and provide an additional window andindicator of normal and abnormal development.The evaluation of these types of movement dur-ing polysomnography can provide an additional

assessment when coupled with traditional devel-opmental appraisals.

The evolution of movements and specificspontaneous behaviors during sleep in neonateshas been reported by Meyers.18 The frequencydistribution of time intervals between sponta-neous behaviors showed that movements weretypically spaced closely in time. However, noneof these spontaneous motor activities occurwith a set interval between successive behaviors.Spontaneous muscle activity in the sleepingneonate conforms to a single pattern of temporalorganization regardless of the specific movementdisplayed or the state of sleep. The pattern approx-imates a systematic alternation of periods ofincreasing and decreasing intervals between suc-cessive behaviors.

When clear NREM states can be identified,body movements reveal a clear state-dependentrelationship.19 A continuum of movements canbe demonstrated with decreasing frequency,respectively, among wakefulness, stage 1 (transi-tional sleep), REM sleep, stage 2 (NREM), andslow-wave sleep. The relative frequency of bodymovements seems to be regulated by state-dependent mechanisms, and body movementsmay be a reliable measure of the developmentand organizational maturation of sleep state dif-ferentiation, particularly if the time base is longenough. When the maturational progression ofthese states does not follow the expected devel-opmental pattern, support for the diagnosis of adelay in maturation of the central nervous sys-tem may be considered. Prolonged, uninter-rupted sleep states without body movements orposition change might indicate abnormal devel-opment of the arousal response and constitutea subtle indication of an underlying central nerv-ous system control abnormality.

Body position may have a significant effecton sleep state organization and might suggestthat the central vestibular system contributes tothe development and organization of a state.Hashimoto and coworkers studied the contribu-tion of the prone and supine body positions toboth physiologic and behavioral correlates dur-ing sleep in neonates.16,20 Quiet sleep occupieda greater percentage of the total sleep timewhen the newborn was in the prone rather thanthe supine position. Gross movements and phasicmuscle activity were also less frequently observedin the prone position, although there was no


Ch06.qxd 1/25/05 3:14 AM Page 61

difference in localized movements between thetwo body positions. Predictably, since there wasan increase in quiet sleep, respiration was moreregular in the prone position. Interestingly, thepulse rate during quiet sleep was higher in thisposition.

The type and frequency of body movementsduring sleep may provide significantly usefulinformation regarding the integrity of the centralnervous system in term newborns. Hakamodaand associates17,21 studied body movements interm newborns with significant illnesses or mal-formations such as perinatal asphyxia, purulentmeningitis, meconium aspiration syndrome, gas-trointestinal bleeding, porencephaly, and hydran-encephaly. Newborns who had recovered fromtransient vomiting were also evaluated and usedas a comparison group. Generalized body move-ments, localized tonic movements, and general-ized phasic movements were evaluated. Patientswith minimally depressed background EEGactivity showed an increase in generalized move-ments and localized tonic movements duringquiet sleep. On the other hand, hydrancephalicinfants showed an increase in generalized phasicmovements in active and quiet sleep and a pro-found decrease in generalized movements dur-ing sleep when EEG abnormalities weremarkedly severe. The absence of or a significantdecrease in generalized body movements or anincrease in generalized phasic muscle activitymight indicate a very poor prognosis for particu-lar infants. On the other hand, the presence oflocalized tonic movements (even in smallamounts) suggests the preservation of corticalfunction.

Spontaneous body movements and behaviorsappear to be controlled by an interaction ofendogenous rhythms that do not appear to runindependently.18,22 Instead, a system of interac-tion between endogenous oscillations exists. By2 weeks of age, an ultradian 3- to 4-hour cyclecan easily be identified. This cycle is typicallyrelated to feeding patterns and may be controlledby hypothalamic mechanisms and metabolicrequirements. A 50-minute basic rest and activ-ity cycle (BRAC), originally described byKleitman,23 can be identified, and a circadiancycle is present at about 8 to 12 weeks of age.

The BRAC appears to trigger sensory andmotor mechanisms characterizing both of thephases of enhanced stereotypical motor activity

during the day and night in children with devel-opmental disabilities and associated stereotypi-cal behaviors during wakefulness and sleep.18,22

The peak frequency of stereotypical behaviorsseems to follow the same mean REM-to-REMinterval on consecutive nights.

New methods of the digital and statisticalanalysis of EEG, EMG, and EOG activity havethe potential to describe the structural and tem-poral characteristics of tonic electrophysiologicactivity during sleep.20,24 Three components canbe identified: slow-fast, hemispheric shift, andocular movement activation. Cycles of tonicelectrophysiologic activity, which are both slowerand faster than the typical 90-minute ultradianrhythm of sleep states, can be identified by theuse of these components in older children andadults.

Activity monitoring accelerometry can alsoprovide significant information regarding dis-turbed sleep in infants and children. First, themonitoring of sleep states by actigraphy is gen-erally consistent with polysomnography whenassessing certain parameters of sleep and thesleep–wake cycle in children. Sadeh and col-leagues have described actigraphic sleep–wakescoring in sleep-disturbed children comparedwith a control group of healthy children.25

Movements measured actigraphically that arecharacteristic of wakefulness lasting longer than5 minutes were significantly greater in the sleep-disturbed group of children, clearly showingpoorer sleep quality in this group (12 to 18months of age). Sleep measurements showed sig-nificant night-to-night stability in both groups.The stability of specific measurements andtheir age trends were different between thegroups. These data clearly showed that the meas-urement of body movements during sleepat home could discriminate between sleep-disturbed and healthy children with a highlycorrect assignment rate.

Other motor disorders of childhood havesignificant sleep-related correlates and motorbehaviors that can be monitored and assessedpolysomnographically. Gadoth and coworkersstudied sleep-related body movements and peri-odic limb movements in unrelated patients withL-dopa–responsive, hereditary, progressive dys-tonia. Also studied were their unaffected familymembers.26 All patients with dystonia had anincrease in major body movements during REM


Ch06.qxd 1/25/05 3:14 AM Page 62

sleep. Most unaffected parents and siblings hadsimilar REM-related body movements, periodiclimb movements, or both. Therefore, a commonmechanism for the dystonia, body movements,and periodic limb movements may exist, and acausative relationship between these two motorphenomena during sleep seems to be implied.

Electrocardiogram ActivityAt least one channel of the polysomnogram isdedicated to recording the patient’s electro-cardiogram (ECG). Cardiac activity is generallymonitored for rate and rhythm, since only a lim-ited, single channel is used. Only relative infor-mation about cardiac electrical function isprovided by polysomnography, and clinical con-clusions regarding a patient’s ECG is generallynot possible. If a comprehensive evaluation ofcardiac activity is necessary, Holter monitoringmay be added to the recording strategy.

The anatomy and function of the fetal cardio-vascular system differ profoundly from those ofextrauterine life. Although the right and leftheart function in parallel during fetal life, imme-diately after birth the right and left sides func-tion in series. Anatomic and physiologic changesin the respiratory and cardiac systems occurimmediately at birth. Alveolar fluid within thelung is expressed and absorbed, being replacedimmediately with air. Pressure changes and oxy-genation result in closure of the ductus arteriosusand ventricular and atrioseptal communicatingchannels, isolating the pulmonary and systemiccirculation.

Heart rate generally decreases during naturalsleep and is about 4 to 8 beats per minute slowerwhen the heart rate during quiet wakefulness iscompared with quiet sleep.3,27 During tonic peri-ods of REM sleep, a decrease in rate of approxi-mately 8% less than that for quiet wakefulnesshas been reported. A reduction in heart rate dur-ing sleep appears to parallel sleep-related alter-ations in blood pressure.5,28 In addition, heartrate variability is clearly greater in REM sleepthan in slow-wave sleep.7,29 Although there is asignificant decrease in heart rate during tonicperiods of REM sleep, heart rate is also signifi-cantly influenced by phasic activity during REMsleep.8,30 During the phasic activity of REM sleep,there are clear episodes of short-lasting tachy-cardia followed occasionally by a brief rebound

bradycardia before a return to baseline levels.Variations in heart rate differ significantlyaccording to the sleep state. Before week 37 ofgestational age, the sequential curves of heartrate show periodic variations present in bothactive and quiet sleep.9,31 After 37 weeks, slowperiodic variations are still present in activesleep but superimposed by fast variations syn-chronous with respiratory cycles. Fast variationsprevail in quiet sleep.

In newborns suffering from pathologic condi-tions (medical or surgical), less variability asso-ciated with respiration is seen with prematurity,young age at recording, and hypercapnia.However, this diminished respiratory-relatedvariation in heart rate can be very transient.Pronounced variations similar to those in babieswithout abnormalities are observed in two thirdsof the patients with or without ventilatory assis-tance.9,31 The periodicity of heart rate and respi-ration varies during sleep in newborn babies.10,32

The period durations obtained by power spectralanalysis showed two maximums. In many casesthe cycles of respiration and heart rate were notidentical. It has been shown that newborns havetwo separate cycles with different period dura-tions. The shorter cycle probably originates fromfetal life; the longer one represents the develop-ment of a more mature periodicity. In studyingrespiration and heart rate variation in normalinfants during quiet sleep over the first year oflife, Litscher and associates showed that both therespiratory rate during quiet sleep and respira-tory variability decrease with age.11,33 A compar-ison of the cardiorespiratory data from the firstand last quiet sleep periods showed no signifi-cant differences within any of the age groupsstudied.

VanGeijn and colleagues studied heart rate asan indicator of the sleep state in newborninfants.13,34 During quiet sleep, the R-R intervallength was longer, the long-term irregularityindex lower, and the interval-difference indexhigher than those found during the immediatelypreceding or following active sleep. For noncon-secutive quiet and active sleep states, a maxi-mum separation was obtained and, withdiscriminant function analysis, correctly classi-fied in 93% of quiet and active sleep epochs.These data have specific implications for the iden-tification of the sleep state in the fetus withoutthe need for antepartum invasive monitoring.


Ch06.qxd 1/25/05 3:14 AM Page 63

Heart rate and its variability can be affectedby medications. In one study by Gabriel andAlbani, it was demonstrated that the amount ofactive sleep, as well as the incidence of apneaand/or cardiac slowing occurring predominantlyduring active sleep, was decreased at therapeuticlevels of phenobarbital.35 With declining serumdrug levels, active sleep showed a rebound effect;at the same time, apnea and/or cardiac slowingrelapsed. These findings tend to confirm the factthat neonatal apnea is facilitated by active sleep-inhibitory brain mechanisms and these mecha-nisms have significant effects not only on therespiratory system but also on the cardiovascularsystem.

Beat-to-beat variability in heart rate has beenused as an index of the integrity of the auto-nomic nervous system in early infancy. Mazzaand coworkers have shown that the beat-to-beatheart rate variability during active sleep andquiet sleep correlated very well with the instan-taneous heart rate (R-R interval).36 The correla-tion coefficient was .49 to .92 in quiet sleep and.50 to .03 in active sleep. Regression analysissupported a linear approximation of the beat-to-beat heart rate variability to the instantaneousheart rate over the range investigated. The slopesof these linear functions were similar in bothactive sleep and quiet sleep for infants from birthto 4 months of age.15,36

The assessment of vagal tone may be a signif-icant method of assessing the periodic variationin heart rate associated with respiration (respira-tory sinus arrhythmia). Arendt and associatesstudied vagal tone in 20 full-term infants andfound that the heart period and its variabilitywere highly correlated with vagal tone.37

Variation in vagal tone among sleep states wasalso detected. Repeated assessments revealedthat the average vagal tone values collected forthe same sleep state were not significantly corre-lated across successive days. This short-termvariability both among and within individualsdoes not support the notion that a single assess-ment of vagal tone can be used by itself to iden-tify infants at risk for sudden unexpected deathduring sleep or predict a neurodevelopmentaloutcome. However, successive assessmentsmight provide a greater degree of information.

In older infants and children, heart ratereveals significant respiratory modulation. Thisis referred to as a normal sinus arrhythmia (see

Fig. 6–2). However, in newborn infants the res-piratory modulation of heart rate is variable.

Particular types of heart rate variation areenhanced during periods of slow heart rate anddiminished when heart rate is high. Shechtmanand Harper have shown that the maturationalpatterns of heart rate, based on the correlation ofheart rate variations, were strongly influenced bythe sleep–wake state and were dissimilar to thosepreviously reported for the correlation betweencardiac and respiratory measures.38 Their find-ings suggest dissimilar developmental patternsfor autonomic and somatic motor systems andinclude a discontinuity in autonomic develop-ment at approximately 1 month of age. Theyspeculated that these trends reflected a change inthe nature of sleep states as forebrain connectionsdevelop.

State-dependent variations also occur overthe first year of life. Litscher and colleagues per-formed spectral analysis of breathing and heartrate patterns during the first and last episodes ofquiet sleep recorded over an 8-hour all-nightperiod on 19 infants at 6 weeks, 6 months, and1 year of age.33 A total of 43 recordings were ana-lyzed. Their results demonstrated that the respi-ratory rate decreases during quiet sleep and therespiratory variability decreases with age.Calculations of heart rate in beats per minuteand heart rate variability (%) revealed a slowingof heart rate and an increase in variability. A com-parison of the cardiorespiratory data from the firstand last quiet sleep periods showed no significantdifferences within any one of these age groups.

When Haddad and coworkers studied thestate dependence of the QT interval in normalinfants at 2 weeks, 1 month, 2 months, 3 months,and 4 months of life, they found that the QTindex (defined as the QT interval divided by thesquare root of the R-R interval) was significantlygreater during quiet sleep than REM sleep. Thissignificant difference was present in all age groupsstudied.19,39

Cardiac output is also coupled to heart rateduring wakefulness. However, cardiac output isonly slightly decreased during slow-wavesleep.3,18,27,40 A decrease in cardiac output ismore significantly pronounced during the tonicREM state, its average being about 9% less thanthat during quiet wakefulness. Changes in car-diac output are not accompanied by changes instroke volume, which tends to remain constant


Ch06.qxd 1/25/05 3:14 AM Page 64

during quiet wakefulness, slow-wave sleep, andREM sleep.

Blood pressure also varies significantly betweenquiet wakefulness and sleep. There is a relativelymodest decrease in blood pressure during NREMsleep (approximately 14 mm Hg).7,29 It decreasesto an even greater extent during tonic REMsleep. Blood pressure variation during REM sleepis complex and notable. Periods of relativehypotension are interrupted by brief, sharpincreases in mean arterial pressure.8,30 Oscil-lations in blood pressure appear to be related tothe phasic phenomena of REM sleep. Blood pres-sure increases appear to occur shortly beforeor after the onset of bursts of phasic activity(e.g., eye movements, phasic muscle twitches).

Abnormalities and intercurrent medical con-ditions can affect the variability and state depend-ence of heart rate and contribute to arrhythmias.Fagioli and colleagues have shown that there is ahigh frequency of sinus pauses in malnourishedinfants.19,41 This arrhythmia was identifiedmainly during quiet sleep. After the nutritionalrehabilitation of the infants studied, there was adramatic reduction in the frequency of thesesinus pauses. They concluded that the largenumber of sinus pauses seen in infants sufferingfrom severe malnutrition may reflect a distur-bance of neurovegetative regulation that isintensified by sleep-related physiologic changes.

Cardiac and other cardiovascular variationsduring sleep in newborns, infants, and childrenare complex and appear to hold particular andconsequential clinical relevance. Phasic excita-tion of the heart and vascular changes in thecoronary, systemic, and cerebral circulationsmay profoundly influence the course of othermedical conditions in the fragile child.

Respiratory ActivityThree respiratory parameters are typically moni-tored during polysomnographic assessment:nasal/oral airflow, respiratory effort, and oxygensaturation. The recording of nasal and oral air-flow is most commonly accomplished by theplacement of thermistors or thermocouples inthe stream of air. This method of recording air-flow is simple, comfortable, and highly reliablefor most patients but can be imprecise due topositional factors and may require the technicianto frequently adjust the sensors and alter the

amplifier sensitivity. Pneumotachograph record-ing is less commonly used, requires the patientto wear a large facemask during sleep, and maybe impractical for many patients.

Respiratory effort may be measured withstrain gauges, chest wall impedance, intercostalEMG, inductance plethysmography, and pneu-matic transducers. All of these methods providereliable information regarding the patient’sbreathing by monitoring movement of the chestand abdomen. In children and adults the leastrestrictive methods are preferable, largely becausethey minimize discomfort and promote compli-ance with the all-night procedure.

Pulse oximetry is the standard for noninva-sive, continuous monitoring of arterial oxygensaturation. The probe may be placed on an ear-lobe, finger, toe, or foot (in the younger infant).Using spectrophotoelectric principles for record-ing, the pulse oximeter circumvents the compli-cations associated with the transcutaneousmeasurement of the partial pressure of oxygen orthe use of an indwelling arterial catheter. It pro-vides reliable information regarding respiratoryfunction and highly correlates with simultane-ous blood gas determinations.

Anterior Tibial EMG Activity

If periodic leg movements during sleep or noc-turnal myoclonus is suspected, EMGs from theleft and right anterior tibial muscles are recorded.Two surface electrodes are taped approximately3 cm apart on each leg to monitor leg muscleactivity during the sleep period. Leg muscle EMGcan also provide additional information aboutlimb and body movements during the recordingperiod.

Audio-Video Monitoring

Continuous audio and video monitoring andrecording of the patient during the sleep periodcan provide significant details about underlyingsleep-related pathology. Somnambulistic episodescan be chronicled, seizure activity can be docu-mented, and some symptoms of sleep apnea(e.g., loud snoring) can be recorded. Continuousmonitoring also provides a necessary safetyfunction as well as detailed observational dataabout the patient’s sleeping positions at the timeof abnormal physiologic events.


Ch06.qxd 1/25/05 3:14 AM Page 65

POLYSOMNOGRAPHICTECHNIQUES

Neonatal and Infant MonitoringIndications for Study and Common Usage● Suspected apnea of prematurity● Suspected apnea of infancy● Suspected severe gastroesophageal reflux with

or without aspiration● Evaluation of some children who have suffered

apparently life-threatening events● Suspected seizure disorder● Presence of major morphologic abnormalities,

especially those that involve congenital mal-formations of the head, face, mouth, tongue,neck, and/or chest (e.g., Pierre Robin syn-drome, Treacher Collins syndrome, Beckwith’ssyndrome)

● Suspected central hypoventilation syndrome● Congenital neuromuscular disorders associ-

ated with generalized hypotonia● Metabolic/genetic disorders associated with

hypotonia and/or abnormalities of the headand neck

● Postoperative evaluation of infants who haveundergone surgery of the face, mouth, and/orneck (e.g., status/post–cleft palate repair)

● Other (individualized indications that dependon the infant’s presenting problem)

Potential Future Usage● Evaluation of development and/or maturation

of the central nervous system.● Evaluation of long-term prognosis for the

“new morbidity” (New morbidity includes enti-ties such as learning disabilities, behavioralabnormalities, and isolated developmentaldelays.)

● Evaluation of the acute neurophysiologic sta-tus of infants suffering from intrauterine drugexposure

● Prognosis of the intrauterine drug–exposedinfant

● Evaluation of the developmental and neuro-logic outcomes of many other congenital andacquired abnormalities (e.g., infants who arestatus/post-intraventricular hemorrhage)

● Objective neurophysiologic evaluation ofinfants with low 5-minute APGAR scores

Equipment Recommended:Standard/ElectiveStandard Polygraph and Standard EEGElectrode Array

EEG: CHOICE OF MONTAGE

Raw data monopolar recordings are standard forall patients. Two montages might be elected inthe evaluation of the neonate. A single scoringchannel is inadequate for comprehensive poly-somnographic analysis of the neonate or child.The two standard electrode arrays are shown inTable 6–1.

The standard transcoronal electrode array isused in children under 2 years of age unless oth-erwise indicated by physician order. In small andpremature infants, a double-distance electrodearray may be used. The standard-distance trans-coronal and parasaggital electrode array is usedin all children over 2 years of age unless other-wise indicated.

EMG

Standard chin muscle EMG should be used foridentification of the sleep state. In addition, theoptional limb-surface EMG (e.g., bilateral, upperand lower extremities) might be used to identifyphasic activity and limb movements.

EOG

Standard EOG activity may be monitored usingtwo electrodes placed 1 cm lateral to the outercanthi of each eye and offset from the horizontalby 1 cm. Eye movement during the neonatalperiod is often nonconjugate. This must be takeninto consideration in the evaluation of the EOG.

Wrapping the infant’s head in a soft, self-adhering bandage is usually recommended. Thisprotects the electrodes against displacement. Ifthe head is wrapped loosely, a sweat artifact isusually avoided.

Recording of RespiratoryMovements and FunctionRespiratory Effort

Respiratory effort should be measured and/ormonitored by inductive plethysmography orpiezocrystal belts. Chest effort and abdominal


Ch06.qxd 1/25/05 3:14 AM Page 66

movement are recorded on separate channelsand calibrated in phase when the child is awake,assuming there is no respiratory distress duringthe waking state. Measuring respiratory effortby means of only the chest wall and abdomi-nal impedance has poor sensitivity and speci-ficity and should be avoided. In addition tothe assessment of continuous changes in chestand abdominal circumference (or volume withplethysmography), intercostal EMG is also usedto evaluate respiratory effort. This is an indirectmeasurement of diaphragmatic muscle activity.Electrodes are placed in the fifth intercostalspace in the anterior axillary line and calibratedto record diaphragmatic as well as intercostalmuscle activity.

Nasal/Oral Airflow

Airflow is best measured and/or monitored bynasal pressure transduction and continuousmonitoring of the capnography waveform.Thermistor application has been the standardtechnique in adult laboratories but might not beas sensitive a measure of airflow as pressure is inchildren.

Although the phenomenon is not universal,newborn infants are typically obligate nasalbreathers. The measurement of nasal airflowalone may be adequate for most studies. Themeasurement of end-tidal CO2 is a highly accu-rate method of determining airflow at the noseand mouth. Side-stream analysis of end-tidalCO2 is conducted on all patients. Waveforms arerecorded, and a running 10-second average isalso recorded on all patients. A split-lumen can-nula is used for measuring both nasal/oral pres-sure and end-tidal CO2. When supplemental

oxygen is required, the nasal pressure lumen isabandoned and supplemental oxygen provided,leaving the end-tidal CO2 lumen for recordingthe capnography waveform. There is a 3-secondsampling delay when measuring side-streamend-tidal CO2; therefore, analysis and the cate-gorization of occlusive and partially occlusiverespiratory events, as well as central respiratorypause, must take this sampling delay intoaccount.

Hemoglobin-Oxygen Saturation

Continuous monitoring of SaO2 by pulse oxime-try is standard in all neonatal polysomnographicprocedures. In the sick neonate and/or the pre-mature infant, the infant’s blood pH and temper-ature should be known in order to accuratelyassess oxygen saturation during the study. Inaddition, it is important to be aware of the P50(position of the hemoglobin-oxygen dissociationcurve) for fetal hemoglobin for accurate moni-toring and assessment.

Optional Parameters

TcPO2 and TcPCO2 may be continuously or inter-mittently monitored during the study. Theseparameters will be individually ordered whenneeded. However, the complication rate (burns)from the heated probes is significantly greaterthan that seen with continuous monitoringof SaO2. Therefore, these methods should bereserved only for certain circumstances and indi-vidualized for each patient. The probe siteshould be relocated every hour when continuousmonitoring is required. Transesophageal pressuremay also be monitored and is highly sensitive


Standard-Distance Standard-Distance Transcoronal Montage Bitemporal MontageA1-T3 1. FP1-F7 9. A1-T3T3-C3 2. F7-T3 10. T3-C3C3-CZ 3. T3-T5 11. C3-CZCZ-C4 4. T5-O1 12. CZ-C4C4-T4 5. FP2-F8 13. C4-T4T4-A2 6. F8-T4 14. T4-A2CZ-O2 7. T4-T6 15. CZ-O2C4-A1 or C3-A2 8. T6-O2

Table 6–1. Standard Electrode Arrays

Ch06.qxd 1/25/05 3:14 AM Page 67

and specific for increasing nadirs of negativeintrathoracic pressure in the high upper airwayresistance syndrome.

Measurement of Heart Rate and RhythmContinuous monitoring of heart rate and rhythmmay be accomplished by using standard ECGlead II placement. Alternatively, an electrodemay be placed in the middle of the chest (mid-sternal region) and referenced to A1 or A2. Therate should be determined by the frequency ofQRS complexes and the fast Fourier transforma-tion of the R-R interval, and beat-to-beat vari-ability should be evaluated by the comparison ofR-R intervals and assessed during wakefulness,quiet sleep, and active sleep.

MovementMovement may be monitored by limb EMG(upper and lower); direct observations bythe technician, with his/her notations on therecord; and/or video recording. Actigraphymay also be considered for monitoring activity.Behavioral observations and notations are sig-nificant and important during neonatal andinfant polysomnography.

Standard Duration of the StudyPolysomnograms should last a minimum of 6hours but optimally 8 hours. The timing of thestudies is also important. They should be con-ducted during the late evening and early morn-ing hours (e.g., 10:00 PM to 6:00 AM).

Description

A polysomnogram consists of the continuousnocturnal monitoring of the EEG (bipolartranscoronal and parasaggital electrode array),EOG (left outer and right outer canthus), chinmuscle EMG, left and right anterior tibial mus-cle EMG, lead II ECG, instantaneous heart rate(R-R interval by flicker fusion threshold analysis),nasal/oral airflow by capnography and pressuretransduction, end-tidal CO2 trend by capnome-try, chest and abdominal wall respiratory move-ment by piezocrystal belts, body position, upper

respiratory tract sound by sonography, and oxy-gen saturation by pulse oximetry. Continuouspulse volume is measured by using fingerplethysmography. The record is scored and eval-uated according to the accepted criteria ofRechtschaffen and Kales for children over 6months of age42 or Anders, Parmalee, and Emdeefor children younger than 6 months of age.43

Analysis is conducted in 30-second epochs; allepochs of the recording are analyzed.

ARCHITECTURE

The analysis of sleep structure throughout thenight is known as its architecture. Sleep latencyis the time from lights out to sleep onset. REMlatency is the time from sleep onset to the firstREM sleep period. WASO means “wake aftersleep onset.” Sleep efficiency is calculated bydividing the WASO by the total sleep time. Thesleep state percentages for the entire recordingare also reported. Laboratory effects may resultin a somewhat longer sleep latency period thanat home and a slight decrease in REM sleep.

ELECTROENCEPHALOGRAM

There is a comprehensive screening EEG elec-trode array. It is not a diagnostic EEG recording.

CHIN ELECTROMYOGRAM

The submental muscles are continuously evalu-ated for a snore artifact and/or bruxism. Thegenioglossus muscle is the principal muscle beingrecorded.

ANTERIOR TIBIAL MUSCLE ELECTROMYOGRAM

The continuous recording of the EMG of the leftand right anterior tibial muscle is analyzed forperiodic and/or episodic limb movements andrestless legs syndrome as well as phasic muscleactivity during REM sleep.

ELECTROCARDIOGRAM

The continuous recording of lead II is assessedfor rhythm, rate, and abnormalities that may beassociated with central or obstructive respiratoryevents.

RESPIRATORY ANALYSIS

The comprehensive evaluation of sleep-relatedrespiratory status is conducted. The numberof occlusive (apneas) and partially occlusive(hypopneas) respiratory events is provided.


Ch06.qxd 1/25/05 3:14 AM Page 68

Significant central apneas, hypopneas, and/orperiodic breathing are reported in the Impression.The average length of each is calculated. Theapnea index represents the total number of apneasdivided by the total sleep time. The A + H indexrepresents the total number of apneas plus hypo-pneas per hour of sleep. The REM RDI representsthe respiratory disturbance index (total number ofapneas plus hypopneas) during REM sleep only.Baseline and nadir SaO2 are reported, as are base-line and maximum end-tidal CO2.

OTHER CONSIDERATIONS

Several other issues are important in under-standing the processes of polysomnography.These issues include patient preparation by thepractitioner for the sleep study, patient safety,and the physical layout of the laboratory.

Patient PreparationThe proper preparation of the patient for thestudy will alleviate anxiety about the procedureand increase patient compliance. Before thestudy, the practitioner should discuss with thepatient the reason for the testing, the proceduresthat will be used, and what the patient canexpect from the laboratory staff. Directly address-ing these issues will help alleviate the anxietyassociated with sleeping in a strange laboratoryenvironment and will result in a more represen-tative night’s sleep. The extra effort and additionaltime with younger patients and their familiescan significantly ease tension. Most laboratorieswill permit a tour of the facilities before thestudy to familiarize the child with the equipmentthat will be used. Parents are also encouragedto spend the night in the laboratory with theirchild.

Patients and their families often feel that theplacement of multiple electrodes, sensors, andmonitoring equipment will interfere with sleep.Although the potential exists, experience showsthat children (and adults) sleep quite well in thelaboratory, and the sleep recorded is generallyrepresentative of the patient’s typical sleep phys-iology. However, there may be a discrepancy inthe measurement of sleep between the first nightin the laboratory and consecutive nights. Onoccasion, two or more nights of polysomnogra-phy may be required for an accurate assessmentof the presenting problem. This can frequently

be anticipated from historical information beforethe study night, and many sleep laboratories willrequire a clinical evaluation in the sleep centerbefore the investigation. The staff of the sleepcenter can determine the appropriate polysomno-graphic montage and the number of nightsrequired for an accurate diagnosis to be made.

Patient Safety

The safety of the patient is of primary impor-tance within the sleep laboratory. Because thepatient is connected to complex electrical equip-ment, the potential for exposure to incomingelectrical currents and shock hazards is mini-mized by compulsive attention to the inspectionand maintenance of the recording apparatus.A single electrical ground is affixed to the patient.If more than one is used, the creation of a“ground loop” is greatly enhanced.

It is necessary for the technician to continu-ously observe the patient throughout the entirerecording period. Patients may require promptmedical attention during the course of the sleepstudy, and the laboratory staff is prepared forthese situations. Therefore, the techniciansremain vigilant with the patient and recordingthroughout the entire study period.

Sleep Laboratory Environment

Although sleep laboratories differ markedly inlayout, design, and decoration, they generallyconsist of one or more bedrooms for monitoringpatients, an adjacent control room, a restroom,and a storage area. To reduce the probability ofone patient disturbing the sleep of another, eachpatient is provided with a private room. Ideally,the sleep rooms simulate a home environment,are painted a neutral color, and are light andsound attenuated. The rooms are always easilyaccessible to the technician. Some laboratorieshave separate patient preparation rooms. Thesupplies needed for patient setup can be keptin this area and patients will be able to relax andattempt to sleep in the bedroom. Furthermore,efficiency can be maximized because additionaltime-consuming “cleanup” procedures can beperformed in areas other than the sleep room.

The control room is large enough to accom-modate the equipment required for all record-ings and should be safe and comfortable for the


Ch06.qxd 1/25/05 3:14 AM Page 69

technician. In addition, the room must be con-structed in a manner that minimizes electricalinterference and artifact intrusion. Since poly-somnographic records consist of voluminouspages of data, an adjacent reading and storageroom is helpful.

The design of sleep laboratories is gearedtoward both patient comfort and safety and thetechnician’s needs. With the appropriate man-agement strategies, the physical laboratory envi-ronment can ensure efficient patient care andaccurate, reproducible recordings.

References

1. Bowersox SS, Kaitin KI, Dement WC: EEG spin-dle activity as a function of age: Relationship tosleep continuity. Brain Res 1985;334:303-308.

2. Spehlmann R: EEG Primer. Amsterdam, Elsevier,1981.

3. Willis J, Schiffman R, Rosman NP, et al:Asymmetries of sleep spindles and beta activityin pediatric EEG. Clin Electroencephalogr 1990;21:48-50.

4. Drake ME Jr, Pakalnis A, Padamadan H, et al:Sleep spindles in epilepsy. Clin Electroenceph-alogr 1991;22:144-149.

5. Orem J, Barnes CD: Physiology in Sleep. New York,Academic Press, 1980.

6. Louis J, Zhang JX, Revol M, et al: Ontogenesis ofnocturnal organization of sleep spindles: A longi-tudinal study during the first 6 months of life.Electroencephalogr Clin Neurophysiol 1992;83:289-296.

7. Soh K, Morita Y, Sei H: Relationship between eyemovements and oneiric behavior in cats. PhysiolBehav 1992;52:553-558.

8. Becker PT, Thoman EB: Rapid eye movementstorms in infants: Rate of occurrence at 6 monthspredicts mental development at 1 year. Science1981;212:1415-1416.

9. Feinberg I, Baker T, Leder R, March JD: Responseof delta (0-3 Hz) EEG and eye movement densityto a night with 100 minutes of sleep. Sleep1988;11:473-487.

10. Coakley D, Williams R, Morris J: Minute eyemovement during sleep. Electroencephalogr ClinNeurophysiol 1979;47:126-131.

11. Askenasy JJ, Yahr MD: Different laws governmotor activity in sleep than in wakefulness. JNeural Transm Gen Sect 1990;79:103-111.

12. DeKoninck J, Lorrain D, Gagnon P: Sleep posi-tions and position shifts in five age groups: Anontogenetic picture. Sleep 1992;15:143-149.

13. Hakamada S, Watanabe K, Hara K, Miyazaki S:Development of the motor behavior during sleepin newborn infants. Brain Dev 1981;3:345-350.

14. Kohyama J, Watanabe S, Iwakawa Y: Phasic sleepcomponents in infants with cyanosis during feed-ing. Pediatr Neurol 1991;7:200-204.

15. DeKoninck J, Lorrain D, Gagnon P: Sleep posi-tions and position shifts in five age groups: Anontogenetic picture. Sleep 1992;15:143-149.

16. Kohyama J, Watanabe S, Iwakawa Y: Phasic sleepcomponents in infants with cyanosis during feed-ing. Pediatr Neurol 1991;7:200-204.

17. Fukumoto M, Mochizuki N, Takeishi M, et al:Studies of body movements during night sleep ininfancy. Brain Dev 1981;3:37-43.

18. Myers A: Organization of spontaneous behaviorsof sleeping neonates. Percept Mot Skills1977;45:791-794.

19. Wilde-Frenz J, Schulz H: Rate and distribution ofbody movements during sleep in humans. PerceptMot Skills 1983;56:275-283.

20. Hashimoto T, Hiura K, Endo S, et al: Postural effectson behavioral states of newborn infants—a sleeppolygraphic study. Brain Dev 1983;5:286-291.

21. Hakamada S, Watanabe K, Hara K, et al: Bodymovements during sleep in full-term newborninfants. Brain Dev 1982;4:51-55.

22. Meier-Koll A, Fels T, Kofler B, et al: Basic restactivity cycle and stereotyped behavior of a men-tally defective child. Neuropadiatrie 1977;8:172-180.


24. Sussman P, Moffitt A, Hoffmann R, et al: Thedescription of structural and temporal character-istics of tonic electrophysiological activity duringsleep. Waking-Sleeping 1979;3:279-290.

25. Sadeh A, Lavie P, Scher A, et al: Actigraphichome-monitoring sleep-disturbed and controlinfants and young children: A new method forpediatric assessment of sleep-wake patterns.Pediatrics 1991;87:494-499.

26. Gadoth N, Costeff H, Harel S, Lavie P: Motorabnormalities during sleep in patients with child-hood hereditary progressive dystonia, and theirunaffected family members. Sleep 1989;12:233-238.

27. Mancia G, Baccelli G, Adams DB, Zanchetti A:Vasomotor regulation during sleep in the cat. AmJ Physiol 1971;220:1086-1093.

28. Sheldon SH, Spire JP, Levy HB: Pediatric SleepMedicine. Philadelphia, WB Saunders, 1992.

29. Guazzi M, Zanchetti A: Blood pressure and heartrate during natural sleep of the cat and their reg-ulation by carotid sinus and aortic reflexes. ArchItal Biol 1965;103:789-817.

30. Gassel MM, Ghelarducci B, Marchiafava PL,Pompeiano O: Phasic changes in blood pressureand heart rate during the rapid eye movementepisodes of desynchronized sleep in unrestrainedcats. Arch Ital Biol 1964;102:530-544.


Ch06.qxd 1/25/05 3:14 AM Page 70

31. Radvanyi MF; Morel-Kahn F: Sleep and heart ratevariations in premature and full term babies.Neuropadiatrie 1976;7:302-312.

32. Baust W, Gagel J: The development of periodicityof heart rate and respiration in newborn babies.Neuropadiatrie 1977;8:387-396.

33. Litscher G, Pfurtscheller G, Bes F, Poiseau E:Respiration and heart rate variation in normalinfants during quiet sleep in the first year of life.Klin Padiatr 1993;205:170-175.

34. VanGeijn HP, Jongsma HW, deHaan J, et al: Heartrate as an indicator of the behavioral state: Studiesin the newborn infant and prospects for fetalheart rate monitors. Am J Obstet Gynecol 1980;136:1061-1066.

35. Gabriel M, Albani M: Rapid eye movement sleep,apnea, and cardiac slowing influenced by pheno-barbital administration in the neonate. Pediatrics1977;60:426-430.

36. Mazza NM, Epstein MA, Haddad GG, et al:Relation of beat-to-beat variability to heart rate innormal sleeping infants. Pediatr Res 1980;14:232-235.

37. Arendt RE, Halpern LF, MacLean WE Jr,Youngquist GA: The properties of V in newborns

across repeated measures. Dev Psychobiol 1991;24:91-101.

38. Schechtman VL; Harper RH: Minute-by-minuteassociation of heart rate variation with basal heartrate in developing infants. Sleep 1993;16:23-30.

39. Haddad GG, Krongrad E, Epstein RA, et.al.: Effectof sleep state on the QT interval in normalinfants. Pediatr Res 1979;13:139-141.

40. Kumazawa T, Baccelli G, Guazzi M, et al: Hemo-dynamic patterns during desynchronized sleep inintact cats and cats with sinoaortic deafferenta-tion. Circ Res 1969;24:923-927.

41. Fagioli I, Salzarulo P, Salomon F, Ricour C: Sinuspauses in early human malnutrition during wak-ing and sleeping. Neuropediatrics 1983;14:43-46.

42. Rechtschaffen A, Kales A (eds): A Manual ofStandardized Terminology, Techniques andScoring System for Sleep Stages of HumanSubjects. Los Angeles, BIS/BRI, UCLA, 1968.

43. Anders T, Emde R, Parmelee A (eds): A Manualof Standardized Terminology, Techniqued andCriteria for Scoring of States of Sleep andWakefulness in Newborn Infants. Los Angeles,UCLA Brain Information Service, NINDSNeurological Information Network, 1971.


Ch06.qxd 1/25/05 3:14 AM Page 71

The physiology of the human organism has beenextensively studied, and many functions areclearly understood. Responses of organ systemsin various states of health and disease are wellknown during the waking state. For many years,function during the sleeping state had beenassumed to parallel that of waking. In 1963,Nathaniel Kleitman published his historicvolume, Sleep and Wakefulness.1 Kleitman pro-posed that physiologic processes vary accordingto state; that organ systems function and interre-late differently asleep and awake. Since then, sev-eral major texts describing human physiologyduring sleep have been published.2 Over the past40 years, it has become clear that focusing on thefunctioning of the human organism in health anddisease states only during periods of wakefulnessprovides limited insight on which to base manytherapeutic regimens. If organ systems functiondifferently during the sleeping state, the responseto disease processes will also vary between states.

This chapter focuses on specific variations inthe function of organ systems during the sleepingstate that may affect the response to diseaseprocesses and the efficacy of therapeutic regimens.Specific changes in the central nervous system,temperature regulation, and the endocrine, cardio-vascular, and respiratory systems are discussed.

CEREBRAL BLOOD FLOWDURING SLEEP

Consistency of cerebral blood flow (CBF) duringsleep depends on alterations in cerebral vascularresistance. Upper and lower limits of autoregula-tion are not fixed: they vary with the chemicalenvironment, metabolic needs, and neurogenicinput.3 Data regarding CBF in humans are lim-ited because of difficulties in measurement.However, in 1981, Meyer and coworkers4 devel-oped a noninvasive method for estimating local

and regional blood flow in the brain and openedthe door for other investigators to study theeffect of blood flow variations in the intacthuman central nervous system.

Animal experiments have revealed significantvariations in CBF during sleep. There appears tobe a significant decrease in CBF duringnon–rapid eye movement (NREM) sleep and aprofound increase during rapid eye movement(REM) sleep. In general, cerebral vasodilatationoccurs during sleep.5 Response to state changeis, however, heterogeneous, with differentregions of the brain exhibiting different magni-tudes of alteration. Additionally, differences existin blood flow between slow-wave sleep (SWS)and paradoxical sleep (REM).6 Townsend7 founda consistent decrease in CBF during SWS, withan overall average decrease of 10%. During REMsleep, there was a significant increase in bloodflow, with an overall increase of approximately8% over the baseline state.

The exact mechanisms responsible for varia-tions in CBF during sleep are unknown. It hasbeen suggested, however, that changes may be inresponse to variations in metabolic rates of cere-bral tissue during various stages of sleep. Braintemperature decreases in NREM sleep andincreases in REM sleep.8 This REM sleep–relatedincrease is attributed to increased blood flow andincreased metabolic rate during this stage.Neuronal activity in many central nervoussystem regions is higher during REM sleep thanduring NREM sleep, and this increased activitymay be responsible for the increased metabolicrate.9 In addition, there appears to be a neuro-genic component to the control of CBF duringsleep.10 Meyer and Toyoda have suggested thatthe variation in CBF during different stages ofsleep is a function of neurogenic control.11

During REM sleep, local blood flow has beenshown to increase in the rhombencephalon,

Physiologic Variations during Sleepin ChildrenStephen H. Sheldon

7

73

Ch07.qxd 1/25/05 3:15 AM Page 73

whereas blood flow in the mesencephalondecreases.12 Cortical blood flow has been notedto increase by 30% to 50% in REM sleep whencompared with SWS.8 Larger differences are seenin the brainstem. White matter and cortex showthe least alteration. During REM sleep, phasicoscillations in blood flow have been observedand appear to correlate with bursts of phasicactivity (e.g., twitches, eye movements). DuringSWS, oscillations are not as dramatic or as con-sistent.12

Intrinsic regulation of CBF during wakeful-ness and sleep is significantly affected by thechemical environment (Table 7–1). AlthoughCBF is dramatically altered by changes in arterialcarbon dioxide tension (PaCO2), moderatechanges do not appear to be associated with sig-nificant variation in cerebral circulation duringsleep.13 When PaCO2 disturbances occur,extravascular pH appears to be an importantvariable.14

Cerebral metabolic rate and CBF seem to beinsensitive to alterations in the arterial oxygentension (PaO2) within the physiologic range.8 Ifthe arterial PaO2 is lowered to below about 50mm Hg, CBF begins to increase precipitously inresponse to the hypoxia.15 Hypoventilation seenduring sleep in normal subjects results in a pro-gressive decrease in oxygen saturation (SaO2). Ina study reported by Doust and Schneider,16 adecrease in SaO2 from 96% during wakefulnessto 87% during SWS was seen in normal subjects.

Intracranial pressure also reveals sleep-relatedchanges. There is little variation of intracranialpressure from the waking state as the subjectenters NREM sleep.17 During REM sleep, how-ever, large intracranial pressure (ICP) wavesoccur that are almost double the steady statepressure. In most individuals, these ICP wavesare of little clinical significance. However, inpatients with little ICP reserve, small increases

during sleep may result in depressed neuronalfunction.

BODY TEMPERATURE REGULATIONDURING SLEEP

Core body temperature exhibits a highly stablecircadian rhythmicity and has been used as amajor marker for other endogenous rhythms. Asthe night progresses, core body temperature fallsand reaches its nadir during the early morninghours. During REM sleep, however, core tem-perature increases approximately 0.2° C.

Homeothermic temperature regulation ischaracteristic of mammals. Interestingly, animalexperiments have shown that during REM sleep,body temperature follows environmental tem-perature, increasing as the ambient temperatureincreases, decreasing as the ambient tempera-ture decreases. Return to NREM sleep is accompa-nied by a rise in core temperature to homeostaticlevels. This positive correlation of variation ofbody temperature with environmental tempera-ture during REM sleep suggests a shift towardpoikilothermy (i.e., thermoregulatory mecha-nisms are depressed). On the other hand, a neg-ative correlation exists during NREM sleep,indicating that thermoregulation remains intactduring this state.18

Sweating and shivering are major tempera-ture-regulating mechanisms during wakefulness(Table 7–2). Significant variations are seen inboth functions during sleep. Sweating remainsintact during NREM sleep in neutral or warmenvironments,19 but it is notably absent duringREM sleep.20 Similarly, shivering and thermoreg-ulatory vasomotor activity, which occur duringwakefulness and NREM sleep, are not present(or are significantly depressed) during REMsleep. It has been shown that absence of shiver-ing during REM sleep results from mechanisms

74 Chapter 7 Physiologic Variations during Sleep in Children

Parameter Wake Slow-Wave Sleep REM SleepBlood flow Neurogenic and Heterogeneous Increased 30%-50%

chemical control changes Phasic oscillationsBrain temperature Related to metabolic rate Decreased IncreasedIntracranial pressure Relatively stable Relatively stable Increased

Phasic oscillations

Table 7–1. Physiologic Variations in the Central Nervous System during Wake and Sleep

Ch07.qxd 1/25/05 3:15 AM Page 74

other than muscle atonia associated with thisstate. Animals lesioned in the pontine tegmen-tum (which results in abolition of REM atonia)still do not shiver during REM sleep.21

The diaphragm is normally unaffected bygeneralized skeletal muscle inhibition duringREM sleep, but it does not show increased activ-ity (i.e., tachypnea) to warming or cooling.Cooling of the hypothalamus during NREMsleep in animals results in increased oxygenconsumption and metabolic heat production,but neither cooling nor heating of the hypothal-amus during REM sleep results in a thermoreg-ulatory response.22 These data support theassumption that hypothalamic thermoregula-tion is significantly decreased or absent duringREM sleep.

ENDOCRINE VARIATIONSDURING SLEEP

The effects of sleep on hormone secretion areoutlined in Table 7–3.

Growth HormoneIn prepubertal children, secretion of growth hor-mone (GH) is clearly coupled with sleep onset. Itpeaks early in the first third of the night duringSWS and is secreted exclusively during sleep.23

During puberty and throughout adolescence, thepattern of secretion of GH is modified from theprepubertal paradigm. Several minor peaks occurthroughout the day, although GH still reaches itsmaximal concentration during sleep. Shifting of

Physiologic Variations during Sleep in Children Chapter 7 75

Parameter Wake Slow-Wave Sleep REM SleepRegulation Homeothermic Homeothermic Variable

Varies positively with environmental temperature

Sweating Present Present Relatively absentVasomotor activity Present Present Relatively absentShivering Present Present Absent

Table 7–2. Physiologic Variations in Temperature Regulation during Wake and Sleep

Normal Sleep Phase SleepHormone Peak and/or Trough Effect of Shifted Phase DependencyGrowth Secreted at sleep onset Secretion time follows Yes

hormone Peaks early in sleep period the shift in sleep phase.Prolactin Secreted 30-90 min Secretion time follows the Yes

after sleep onset shift in sleep phase.Peaks in the early

morning hoursThyroid-stimulating Peaks in the early evening There is no significant No

hormone Declines more gradually change in secretion. across the sleep period Sleep appears to modulate

(inhibit) secretion.Luteinizing Rises during sleep in There is a shift in secretion Yes

hormone prepubertal children that follows the shift in Secondary peaks occur sleep phase.

during wake during puberty.Follicle-stimulating Sleep-related rise in There is a shift in secretion Yes

hormone secretion occurs. that follows the shift insleep phase.

Cortisol Peaks at the end of the There is no significant Nosleep period change in secretion.

Nadir occurs early in the Sleep appears to modulate Nosleep period. (inhibit) secretion.

Table 7–3. Effects of Sleep on Hormone Secretion

Ch07.qxd 1/25/05 3:15 AM Page 75

the sleep phase to times of day other than thenormal sleep period is accompanied by shifts inthe timing of secretion of GH.24,25 A 180-degreereversal of sleep phase results in a 180-degreeshift in peak secretion of GH. This sleep-associ-ated release of GH has been related to Jouvet’stheory of monoaminergic sleep systems. Agree-ment rests in observations that NREM sleep andhypothalamic releasing factors are both triggeredby the same serotonergic neurons in the raphenuclei of the brainstem.26

ProlactinUnder basal conditions, prolactin rhythmicallypeaks each night. As with GH, summits arealmost entirely restricted to the sleep intervaland are clearly coupled to the sleeping state.27

Prolactin levels generally increase 30 to 90 min-utes after sleep onset and reach maximal levelsin the early morning hours.20 Prolactin secretionalso occurs during daytime naps and remainscoupled to sleep after acute sleep–wake phasereversals. This sleep-related pattern of secretionis present from late puberty to old age and per-sists during pregnancy, when higher levels ofprolactin are secreted. Although prolactin andGH secretion are both connected to the sleepingstate, peaks of each hormone are not coincident.In contrast to the adolescent and adult GHrhythm, the prolactin peak during sleep appearsto be the only major aligned episode of releaseacross 24 hours.28

CortisolLike core body temperature, cortisol follows aclear, well-established, consistent circadianrhythm; it has also been used as a phase refer-ence point for other endogenous rhythms.29

Maximal cortisol secretion generally occurs atthe end of the sleep period, and its nadir islocated early in the sleep period. Manipulationof the timing of sleep does not dramaticallychange adrenocorticotropic hormone–cortisolsecretion as it does GH and prolactin, and itsrhythm is most likely independent of sleep itself,controlled by a different endogenous oscillator.A reduction in plasma cortisol concentrationseems to occur regardless of the timing of thesleep period.26 This suggests that cortisol secre-tion and the rhythm of the sleep–wake cycle areindependent. Sleep appears to only modulate the

release of cortisol, inhibiting rather than con-trolling secretion.

Thyroid-Stimulating HormoneDaily maximums of thyroid-stimulating hor-mone (TSH) secretion are seen to rhythmicallyrecur each evening and precede the onset ofsleep.30 The TSH peak begins in the earlyevening and then declines across the sleep phase.This sleep-associated decline in TSH suggeststhat sleep is almost as important a determinantof the locus of the maximum TSH rhythm as it isfor GH and prolactin. However, with shifts in thesleep phase, the rise in TSH levels appears to becoupled to clock time rather than sleep period.Changing the timing of the sleep period revealsthat inhibition of TSH secretion occurs withsleep, although the rise to peak of this hormoneis still synchronized to clock time. Therefore,sleep appears to truncate the circadian secretoryepisode of TSH.26

Luteinizing HormoneLuteinizing hormone (LH) has been shown todrive male testosterone production by the Leydigcells of the testes, and it exhibits sleep-relatedrhythmicity during puberty.31-33 However, thesesleep-associated changes in LH secretion andtestosterone production have been shown tooccur prior to Tanner stage 2.34 LH increases dur-ing sleep seems to be responsible for initiation ofpubertal changes. LH surges are followedmomentarily by elevations in plasma testosteronelevels, which, in turn, induce the changes in sec-ondary sexual characteristics seen during thisstage of development.32,35 As boys traversepuberty, enhanced episodic release of LH andtestosterone occur during wakefulness untilTanner stage 5 is reached. At this time, an adultpattern of equivalent pulse height of LH secretionacross the sleep cycle occurs.32 Change of sleepphase results in a corresponding change in LHand testosterone activity. Testosterone levels peakduring sleep in the adult in spite of a weakenedor absent nocturnal rise in LH activity.20

Follicle-Stimulating HormoneFollicle-stimulating hormone (FSH) follows apattern similar to that of LH. Increases in LHoccur at sleep onset, and wakefulness appears toinhibit LH and FSH secretion. The delay of sleep


Ch07.qxd 1/25/05 3:15 AM Page 76

or the acute reversal of the sleep–wake cycleresults in corresponding changes in FSH activitypatterns. Pubertal girls show sleep-coupledrelease of LH and FSH during sleep.20 Estradiollevels, however, do not increase until 10 to 14hours later. Cycling women do not show sleep-related release, but there may be inhibition of LHsecretion 2 to 3 hours after sleep onset.

CARDIOVASCULAR VARIATIONSDURING SLEEP

The physiologic variations in the cardiovascularsystem during wake and sleep are outlined inTable 7–4.

Heart RateHeart rate is generally reduced during naturalsleep in most animal species. When comparedwith quiet wake (QW), the effect of synchronized(NREM) sleep on heart rate is a slight decrease inheart rate of 4 to 8 beats per minute.36 DuringREM sleep, a decrease in rate of 17 beats perminute (approximately 8% less than the mean forQW) has been noted. Reduction in heart rateduring sleep appears to parallel sleep-relatedalterations in blood pressure, in that the variationis more pronounced during REM sleep, and heartrate variability is clearly greater in REM sleepthan in NREM SWS.37

Heart rate is also significantly influenced byphasic phenomena of REM.38 During bursts ofrapid eye movements and body movements,

there are clear episodes of short-lasting tachycar-dia, followed occasionally by a rebound briefbradycardia before return to baseline levels.

Cardiac output is generally coupled to heartrate during wakefulness. During sleep, however,cardiac output has been shown to be onlyslightly reduced during SWS when comparedwith the waking state.36,39 Reduction in cardiacoutput becomes more pronounced, however,during the tonic REM state, its average thenbeing about 9% less than during QW. Changes incardiac output are not accompanied by changesin stroke volume, which tends to remain con-stant during QW, SWS, and REM cycles.

Blood Pressure

Extended studies of laboratory animals haverevealed a moderate, but significant decrease inblood pressure (approximately 14 mm Hg) dur-ing NREM sleep when compared with QW.37 Amore significant fall has been shown to occurduring REM sleep (approximately 25 mm Hg).Changes in blood pressure are more complexand variable during REM sleep than duringNREM sleep. Brief sharp increases in mean arte-rial pressure occur, which are superimposed onrelatively hypotensive values. These oscillationsin blood pressure appear to be related to briefexcitatory somatomotor phenomena that aretypical of REM sleep. Blood pressure rise beginswith, or occurs shortly following, the onset ofbursts of rapid eye movements and muscletwitches.38


Parameter Waking Slow-Wave Sleep REM SleepHeart rate Variable Mildly decreased Variable

Dependent on activity compared to wake Decrease with phasicincreases

Cardiac output Variable Mildly decreased Mildly decreased Dependent on activity compared to wake compared to wake

and SWSBlood pressure Variable Mildly decreased Decreased compared

Dependent on activity compared to wake to SWS and wakePhasic elevation

Peripheral Variable Mild vasodilation Vasodilation withvasomotor activity Dependent on activity periods of phasic

vasoconstriction

REM, rapid eye movement; SWS, slow-wave sleep.

Table 7–4. Physiologic Variations in the Cardiovascular System during Sleep

Ch07.qxd 1/25/05 3:15 AM Page 77

Two types of blood pressure changes indesynchronized sleep occur in the cat. First, atonic blood pressure alteration takes place, con-sisting of hypotension lasting throughout thedesynchronized sleep episode. Second, there arefrequent phasic pressure changes, which consistof brief blood pressure rises that occur simulta-neously with other phasic phenomena.40

Marked, brief cutaneous vasoconstrictionoccurs concomitantly with phasic events duringREM sleep.20 Other regional vasoconstrictiontakes place as well. For example, there is a sig-nificantly reduced urine volume production dur-ing REM sleep, which is considered to be theresult of a reduced renal blood flow during thisstate. A central inhibitory influence on vasocon-striction, which is present during REM sleep, hasbeen shown to be potentiated by sinoaortic dener-vation.20 Vasoconstriction normally seen in theiliac artery during REM sleep is converted tovasodilatation. Surprisingly, the buffering actionof the sinoaortic nerve depends more onchemoreceptive reflexes than on baroreceptorreflexes. The baroreflexes are diminished duringREM sleep in the cat, and this variation may benecessary to maintain homeostasis.

Cardiovascular variations during sleep holdspecific and significant clinical relevance. Phasicexcitation of the heart and coronary circulationmay precipitate decreased coronary artery bloodflow and precipitate myocardial infarction (orother myocardial dysfunction) during REMsleep. Nowlin and coworkers reported that noc-turnal exacerbations of angina generally occur

during REM sleep.41 As early as 1923,MacWilliams reported that deaths from cardiacdisorders occur more frequently during sleepthan wakefulness.42 Most deaths were found tooccur around 05:00 to 06:00, a time when sleepconsists primarily of the REM state.

RESPIRATORY VARIATIONSDURING SLEEP

Well-defined changes in respiratory patternsoccur during sleep (Table 7–5). These alter-ations result in modification of ventilatory con-trol, blood-gas and respiratory patterns, andregulatory responses to changes that are signifi-cantly different than in the waking state.

In general, hypoventilation occurs duringsleep in the normal individual. During SWS,there is a slight decrease in minute ventilation. Achange in metabolic rate is demonstrated by adecrease in oxygen uptake and an increase incarbon dioxide production by approximately10% to 20%.43 Subsequently, a change in ventila-tory control also occurs, as alveolar ventilationfalls more than expected in response to thesemetabolic changes.44 Consequently, the partialpressure of CO2 increases during SWS.

One of the most striking features of breathingduring SWS is its monotonous regularity andlack of breath-to-breath variability. Respiratoryrate is slightly lower and the tidal volume isslightly greater in SWS than in wakefulness.45,46

All major indices of respiratory function appearto be in a stable state during this time. In con-


Parameter Waking Slow-Wave Sleep REM SleepMinute ventilation Varies with activity Mildly decreased VariableRespiratory Rate Varies with activity Mildly decreased VariablePaCO2 Stable Mildly increased IncreasedPaO2 Stable Mildly decreased DecreasedHypercapnic Sensitive to elevation Mildly decreased Decreased to a greater

ventilatory response in carbon dioxide response extent when compared to wake and SWS

Hypoxic ventilatory Sensitive to decreases Mildly decreased Decreased to a greaterresponse in oxygen tension response extent when compared

to wake and SWS

PaCO2, arterial carbon dioxide tension; PaO2, arterial oxygen tension; REM, rapid eye movement; SWS, slow-wavesleep.

Table 7–5. Physiologic Variations in the Respiratory System during Sleep

Ch07.qxd 1/25/05 3:15 AM Page 78

trast, during REM sleep, significant irregularityis noted. Irregularities in respiratory patternswere part of the original description of REMsleep by Aserinsky and Kleitman in 1953.47

These irregularities of respiration and rapidbreathing patterns associated with REM sleephave been subsequently confirmed in studies inhuman infants, human adults, and otherspecies,48-50 and as seen with cardiovascularchanges, there is a significant temporal relation-ship to the phasic components of REMsleep.48,51,52

Respiratory changes seen in REM sleepappear to be secondary to internally activatedneural events.53 This is consistent with the pow-erful REM sleep–coupled neural influences onother systems, which apparently arise from thebrainstem.

The neonate (in particular the prematurenewborn) exhibits irregular breathing patterns.Breath-to-breath variability and long episodes ofperiodic breathing occur.54,55 Periodic breathingpatterns in premature infants occur duringwake, quiet sleep, and active sleep. Althoughpresent during quiet sleep, periodic breathing issignificantly increased during REM sleep.56

During quiet sleep, periodic breathing appears tobe quite regular: breathing and apneic intervalsare of similar durations. These intervals are, onthe other hand, quite irregular during activesleep.57 The cause of periodic breathing isunknown, but many investigators believe that itdepends on oscillations in blood gases duringsleep.58

REM atonia, which affects almost all skeletalmuscles, also affects muscles involved in respira-tion. The diaphragm, however, maintains itsactivity during REM sleep, but intercostal mus-cles and muscles of the upper airway are signifi-cantly hypotonic or atonic during this stage.20

Decrease in muscle tone of the accessory mus-cles of respiration results in hypoventilationand/or upper airway occlusion. Neonates andinfants appear to be particularly susceptible tointercostal muscle hypotonia resulting in ribcage collapse and “see-saw” respiratory efforts(paradoxical respiration).

Mucociliary clearance of pulmonary secre-tions is reduced during sleep. The cough reflex issignificantly suppressed during NREM and REMsleep, and the presence of coughing appears todepend on arousal from sleep.59 Decreased clear-

ance of secretions becomes clinically significantin patients with pathologic states that involveincreased mucous production (e.g., asthma).Sleep architecture is frequently disrupted bynumerous arousals, perhaps to assist in clear-ance of these pulmonary secretions. In addition,respiratory tract smooth muscle tone is alsoaffected by sleep. There is a decrease in this tonein NREM sleep when compared with wakeful-ness. Tone continues to decrease in tonic REMsleep; however, there is a phasic increase super-imposed on this decrease during bursts of rapideye movements.20

During wakefulness, the newborn infantresponds to hypoxia in a manner different fromthat of older children and adults.60 At loweredPaO2, a biphasic response is seen, with an initialperiod of hyperventilation followed by a fall inventilation to a level below the baseline. A simi-lar response curve is noted during REM sleep,but the initial hyperventilatory response is lessdramatic. During NREM sleep, hyperventilationis noted and sustained, without the fall seen inthe other two states.

Ventilatory response to hypoxemia isdecreased during NREM sleep (compared withwakefulness) in adult men. In contrast, theresponses during NREM and during wakefulnessare similar in women. The reason for this sex dif-ferentiation is unknown.61-63 During REM sleep,the ventilatory response to hypoxia is below thatof NREM sleep in both males and females.

Ventilatory response to hypercapnia isdepressed during sleep in the adult human.64

The slope of the ventilatory–CO2 response curvefalls during NREM sleep when compared withwakefulness.65-67 This decrease in response fromwakefulness to NREM sleep has been reported tobe approximately 50%.68,69 This change does notappear to occur in females.70 Hypercapnicresponse appears to be lowest in REM sleepwhen compared with wakefulness and NREM.The mean hypercapnic response during REMsleep seems to be approximately 28% lower thanin the other two states.68,70,71

In transient or semi–steady state experimentsin human neonates, hypercapnic ventilatoryresponse differs from in the adult. Studies indi-cate that there is no difference in the ventilatoryresponse to CO2 between active and quietsleep.72-75 Using rebreathing techniques, how-ever, a lower response in REM sleep is


Ch07.qxd 1/25/05 3:15 AM Page 79

suggested,60,76 and the hypercapnic response islower in preterm neonates than in term infants.77

SKELETAL MUSCLE ACTIVITYVARIATIONS DURING SLEEP

As early as 1894, a reduction in spinal reflexesand loss of patellar reflex activity during sleepwas noted.78 These observations have been sys-tematically confirmed. Hodes and Dement79

documented a decrease in spinal reflexes duringsleep, and a decrease in body motility associatedwith the progression into SWS was corroboratedby Rohmer and coworkers.80

Skeletal muscle activity and spinal cord reflexactivity are diminished to their greatest extentduring REM sleep.79,81 Periodically occurringtwitches and muscular tremors of the face andlimbs frequently occur during this stage ofsleep.82 Progressive relaxation of the skeletalmusculature and attenuation of reflexes are mostprominently seen during REM sleep. DuringREM sleep, periodic bursts of excitatory activitybreak through the generalized tonic inhibition.Hyperpolarization of the intracellular membraneis correlated with active inhibition of the motorneuron at the spinal level.83 Stimulation of neu-rons in the pons (specifically in the nucleus pon-tis oralis) results in different peripheral effectsdepending on whether the subject is awake, inNREM sleep, or in the REM sleep state. Duringwakefulness, stimulation of these neuronsresults in excitatory postsynaptic potential gen-eration. During REM sleep, however, inhibitorypostsynaptic potentials are recognized. Thus,membrane potentials are reversed depending onthe state of arousal. There appears to be a neu-ronal gate that opens specifically during REMsleep. In addition, coactivation of facilitative andinhibitory neuronal activity explains thetwitches and jerks accompanying REM sleep.

Major body movements are strongly related tothe stage of sleep, with the rate of body move-ments progressively decreasing from waking,through stage 1, REM sleep, and stage 2, toSWS.84 Although body movements occur duringall stages of sleep, they are most frequent duringREM sleep and least frequent during SWS.85

Within a single night’s sleep, the number of bodymovements can vary considerably (from 70 to200 movements).86

VARIATIONS IN GENITOURINARYFUNCTION DURING SLEEP

Changes in renal function occur during sleepand are characterized by a decrease in urine vol-ume and an increase in urine osmolality.87 Thesechanges generally occur in conjunction withREM sleep, but they appear to be more related todecrease in regional blood flow and not pri-marily mediated by REM-related antidiuretichormone release (though a nocturnal peak invasopressin secretion can be demonstrated).88 Inaddition to renal conservation of water duringsleep, excretion of sodium chloride is reducedduring this state.

Penile tumescence occurs during REM sleepin children and adult males.89-92 The most strik-ing amount of nocturnal penile tumescenceepisodes occur in the prepubertal and pubertalyears, gradually declining in frequency through-out the later years of life. A similar phenomenonappears to occur in females during REM sleep.Most evidence for this phenomenon stems fromsimilarities between phasic shifts of vaginal vas-cular blood flow and penile blood flow duringREM sleep. Karacan and coworkers93 reportedclitoral erections in three women during REMsleep. Abel and coworkers94 documentedincreases in relative pulse pressure within thevagina during REM sleep.

VARIATIONS INGASTROINTESTINAL FUNCTIONDURING SLEEP

It has been shown that physiologic activity dur-ing sleep is not homogeneous. It varies as a func-tion of the stage of sleep as well as the time ofnight. Parasympathetic activity seems to domi-nate NREM sleep, and sympathetic activity dom-inates REM sleep.95

Gastrointestinal function is extrinsically andintrinsically modified during sleep. Finch andcoworkers96 demonstrated that gastroduodenalmotility during sleep shows a strong relationshipwith body movements and with sleep stagechanges. Although there appear to be no statisti-cally significant relationships between gastricacid secretion and stages of sleep,97 a circadianvariation of secretion is seen, with peaks occur-ring about 02:00. There appears to be a lack of


Ch07.qxd 1/25/05 3:15 AM Page 80

inhibition of gastric acid secretion during sleepin patients with duodenal ulcers. It is speculatedthat epigastric pain during sleep in duodenalulcer patients may be more related to the pro-longed fast associated with sleep than to factorsassociated with gastric acid secretion.98

Swallowing and esophageal function areimportant determinants of gastroesophagealreflux during sleep. Swallowing is significantlysuppressed during sleep.99 This suppression ofswallowing is most noticeable during SWS, andrefluxed gastric contents maintain mucosal con-tact for longer periods of time. Acid contact andprolonged acid clearance is considered to be theprime factor in the development of esophagitisin patients with gastroesophageal reflux.100

Salivary flow, another important determinantof acid clearance, nearly ceases during sleep.101

Body position also appears to play an importantrole in acid clearance, with supine positionsassociated with markedly prolonged esophagealacid clearance.100

Orr and coworkers99 documented that sleepimpairs esophageal acid clearance in both nor-mal subjects and patients with esophagitis.However, patients with esophagitis took a signif-icantly longer time for acid clearance even dur-ing waking hours. This is especially noteworthy,as most gastroesophageal reflux occurred duringarousals from sleep in this study, as well as in astudy of spontaneous gastroesophageal reflux innormal subjects.102 These data clearly implicatedefective acid clearance as a major determinantin the pathogenesis of esophageal inflammationin patients with gastroesophageal reflux. Fewdata are available regarding esophageal motilityand acid clearance during sleep in the infantwith significant gastroesophageal reflux.

References


2. Orem J, Barnes CD: Physiology in Sleep. NewYork, Academic Press, 1980.

3. Ekstrom-Jodal, Haggendal E, Linder LE, NilssonNJ: Cerebral blood flow autoregulation at higharterial pressures and different levels of carbondioxide tension in dogs. Eur Neurol 1971-72;6:6-10.

4. Meyer JS, Hayman LA, Amano T, et al: Mappinglocal blood flow of human brain by CT scanningduring stable xenon inhalation. Stroke 1981;12:426-436.

5. Bridges TH, Clark K, Yahr MD, et al:Plethysmographic studies of the cerebral circula-tion: Evidence for cranial nerve vasomotor activ-ity. J Clin Invest 1958;37: 763-772.

6. Reivich M, Isaacs G, Evarts E, Kety S: The effectsof slow-wave sleep and REM sleep on regionalblood flow in cats. J Neurochem 1968;15:301-306.

7. Townsend RE, Prinz PN, Obrist WD: Humancerebral blood flow during sleep and waking.J Appl Physiol 1973;35:620-625.

8. Greenberg JH: Sleep and the cerebral circulation.In Orem J, Barnes CD (eds): Physiology in Sleep.New York, Academic Press, 1980, pp 57-95.

9. Reivich M: Blood flow and metabolism couple inthe brain. In Plum F (ed): Brain Dysfunction inMetabolic Disorders. New York, Raven Press,1974, pp 125-140.

10. Perves MJ: The Physiology of the Cerebral Cir-culation. London, Cambridge University Press,1972.

11. Meyer JS, Toyoda M: Studies of rapid changes incerebral circulation and metabolism duringarousal and rapid eye movement sleep in humansubjects with cerebrovascular disease. In ZulchKJ (ed): Cerebral Circulation and Stroke. NewYork, Springer-Verlag, 1971, pp 156-163.

12. Baust W: Local blood flow in different regions ofthe brain-stem during natural sleep and arousal.Electroencephalogr Clin Neurophysiol 1967;22:365-372.

13. Alexander SC, Cohen PJ, Wollman H, et al:Cerebral carbohydrate metabolism duringhypocarbia in man: Studies during nitrous oxideanesthesia. Anesthesiology 1965;26:624-632.

14. Wahl M, Deetjen P, Thurau K, et al: Micro-puncture evaluation of importance of perivascu-lar pH for the arteriolar diameter on the brainsurface. Pflugers Arch 1970;316:152-163.

15. Kogure K, Scheinberg P, Reinmuth OM, et al:Mechanism of cerebral vasodilatation in hypoxia.J Appl Physiol 1970;29:223-229.

16. Doust LJW, Schneider RA: Studies on the physi-ology of awareness: Anoxia and the levels ofsleep. Br Med J 1952;1:449-455.

17. Gucer G, Viernstein LJ: Intracranial pressure inthe normal monkey while awake and asleep.J Neurosurg 1979;51:206-210.

18. Parmeggiani PL, Franzini C, Lenzi P, Cianci T:Temperature changes in cats sleeping at differentenvironmental temperatures. Brain Res 1971;33:397-404.

19. Henane R, Buguet A, Roussel B, Bittel J: Variationsin evaporation and body temperatures duringsleep in man. J Appl Physiol 1977;42:50-55.

20. Orem J, Keeling J: A compendium of physiology insleep. In Orem J, Barnes CD: Physiology in Sleep.New York, Academic Press, 1980, pp 315-335.


Ch07.qxd 1/25/05 3:15 AM Page 81

21. Hendricks JC, Bowker RM, Morrison AR:Functional characteristics of cats with pontinelesions during sleep and wakefulness and theirusefulness for sleep research. In Koella WP,Levin P (eds): Sleep 1976. Basel, Karger, 1977,pp 207-210.

22. Parmeggiani PL, Agnati LF, Zamboni G, Cianci T:Hypothalamic temperature during sleep cycles atdifferent ambient temperatures. Electroen-cephalogr Clin Neurophysiol 1975;38:589-596.

23. Parker DC, et al: Rhythmicities in humangrowth hormone concentration in plasma. InKrieger DT (ed): Endocrine Rhythms. New York,Raven Press, 1979, pp 143-173.

24. Parker DC, Rossman LG: Physiology of humangrowth hormone release in sleep. In Scow RO(ed): Endocrinology. Amsterdam, ExerptaMedica, 1973, pp 655-660.

25. Parker DC, Rossman LG: Sleep-wake cycle andhuman growth hormone, prolactin and luteiniz-ing hormone. In Raiti S (ed): Advances inHuman Growth Hormone Research. Washing-ton, DC, U.S. Government Printing Office, 1974,pp 294-312.

26. Parker DC, et al: Endocrine rhythms acrosssleep-wake cycles in normal young men underbasal state conditions. In Orem J, Barnes CD(eds): Physiology in Sleep. New York, AcademicPress, 1980, pp 145-179.

27. Parker DC, Rossman LG, Vander Laan EF: Sleep-related, nychthermeral and briefly episodic vari-ation in human plasma prolactin concentrations.J Clin Endocrinol Metab 1973;36:1119-1124.

28. Sassin JF, Frantz AG, Weitzman ED, Kapen S:Human prolactin: 24-hour pattern with increasedrelease during sleep. Science 1972;177:1205-1207.

29. Aschoff J: Circadian rhythms: General featuresand endocrinological aspects. In Krieger D (ed):Endocrine Rhythms. New York, Raven Press,1979, pp 1-61.

30. Parker DC, Pekary AE, Hershman JM: Effect ofnormal and reversed sleep-wake cycles uponnyctohemeral rhythmicity of plasma thy-rotropin: Evidence suggestive of an inhibitoryinfluence in sleep. J Clin Endocrinol Metab1976;43:318-329.

31. Boyar R, Finkelstein J, Roffwarg H, et al:Synchronization of augmented luteinizing hor-mone with sleep during puberty. N Engl J Med1972;287:582-586.

32. Parker DC, Judd HL, Rossman LG, Yen SS:Pubertal sleep-wake patterns of episodic LH,FSH, and testosterone release in twin boys.J Clin Endocrinol Metab 1975;40:1099-1109.

33. Kapen S, Boyar RM, Finkelstein JW, et al: Effectof sleep-wake cycle reversal on luteinizing hor-

mone secretory pattern in puberty. J ClinEndocrinol Metab 1974;39:293-299.

34. Judd HL, Parker DC, Yen SSC: Sleep-wake pat-terns of LH and testosterone release in prepuber-tal boys. J Clin Endocrinol Metab 1977;44:865-869.

35. Judd HL, Parker DC, Siler TM, Yen SS: The noc-turnal rise of plasma testosterone in pubertalboys. J Clin Endocrinol Metab 1974;38:710-713.

36. Mancia G, Baccelli G, Adams DB, Zanchetti A:Vasomotor regulation during sleep in the cat.Am J Physiol 1971;220:1086-1093.

37. Guazzi M, Zanchetti A: Blood pressure and heartrate during natural sleep of the cat and their reg-ulation by carotid sinus and aortic reflexes. ArchItal Biol 1965;103:789-817.

38. Gassel MM, Ghelarducci B, Marchiafava PL,Pompeiano O: Phasic changes in blood pressureand heart rate during the rapid eye movementepisodes of desynchronized sleep in unre-strained cats. Arch Ital Biol 1964;102:530-544.

39. Kumazawa T, Baccelli G, Guazzi M, et al:Hemodynamic patterns during desynchronizedsleep in intact cats and cats with sino-atrial deaf-ferentation. Circ Res 1969;24:923-937.

40. Mancia, G, Zanchetti, A: Cardiovascular regula-tion during sleep. In Orem J, Barnes CD:Physiology in Sleep. New York, Academic Press,1980, pp 1-55.

41. Nowlin JB, Troyer WG Jr, Collins WS, et al: Theassociation of nocturnal angina pectoris withdreaming. Ann Intern Med 1965;63:1040-1046.

42. MacWilliam JA: Blood pressure and heart actionin sleep and dreams: Their relation to haemor-rhages, angina and sudden death. Br Med J1923;22:1196-1200.

43. Phillipson EA, Murphy E, Kozar LF: Regulationof respiration in sleeping dogs. J Appl Physiol1976;40:688-693.

44. Sullivan CE: Breathing in sleep. In Orem J,Barnes CD (eds): Physiology in Sleep. New York,Academic Press, 1980, pp 213-272.

45. Orem J, Netick A, Dement WC: Breathing dur-ing sleep and wakefulness in the cat. RespPhysiol 1977;30:265-269.

46. Lugaresi E, et al: Breathing during sleep in manin normal and pathological conditions. InFitzgerald H, Gautier H, Lahiri S (eds): The reg-ulation of Respiration during Sleep andAnesthesia. New York, Plenum, 1978, pp 35-45.

47. Aserinsky E, Kleitman N: Regularly occurringperiods of eye motility, and concomitant phe-nomena, during sleep. Science 1953;118:273-274.

48. Phillipson EA, Kozar LF, Murphy E: Respiratoryload compensation in awake and sleeping dogs.J Appl Physiol 1976;40:895-902.


Ch07.qxd 1/25/05 3:15 AM Page 82

49. Snyder F, Hobson JA, Morrison DF andGoldfrank F: Changes in respiration, heart rate,and systolic blood pressure in human sleep.J Appl Physiol 1964;19:417-422.

50. Snyder F: Autonomic nervous system manifesta-tions during sleep and dreaming. In Kety SS,Evarts EV, Williams HL: Sleep and Altered Statesof Consciousness. Baltimore, Williams &Wilkins, 1967, pp 469-487.

51. Aserinsky E: Periodic respiratory pattern occur-ring in conjunction with eye movements duringsleep. Science 1965;150:763-766.

52. Baust W, Holzbach E, Zechlin O: Phasic changesin heart rate and respiration correlated withPGO-spike activity during REM sleep. PflugersArch 1972;331:113-123.

53. Pompeiano O: The neurophysiological mecha-nisms of the postural and motor events duringdesynchronized sleep. In Ketty SS, Evarts EV,Williams HL (eds): Sleep and Altered States ofConsciousness. Baltimore, Williams & Wilkins,1967, pp 351-423.

54. Rigatto H, Brady JP: Periodic breathing andapnea in preterm infants: I. Evidence forhypoventilation possibly due to central respira-tory depression. Pediatrics 1972;50:202.

55. Rigatto H, Brady JP: Periodic breathing andapnea in preterm infants: II. Hypoxia as a pri-mary event. Pediatrics 1975;55:604.

56. Rigatto H, Kalapesi Z, Leahy FN, et al:Ventilatory response to 100% and 15% O2 dur-ing wakefulness and sleep in preterm infants.Early Hum Dev 1982;7:1.

57. Rigatto H: Control of breathing during sleep inthe fetus and neonate. In Kryger MH, Roth T,Dement WC (eds): Principles and Practice ofSleep Medicine. Philadelphia, WB Saunders,1989, pp 237-248.

58. Waggener TB, Stark AR, Cohlan BA, Frantz ID3rd: Apnea duration is related to ventilatoryoscillation characteristics in newborn infants.J Appl Physiol 1984;57:536.

59. Bateman JRM, Clarke SW, Pavia D, Sheahan NF:Reduction in clearance of secretions from thehuman lung during sleep. J Physiol 1978;284:55P.

60. Rigatto H: Control of ventilation in the new-born. Annu Rev Physiol 1984;46:661-674.

61. White DP, Douglas NJ, Pickett CK, et al:Hypoxic ventilatory response during sleep innormal premenopausal women. Am Rev RespirDis 1982;126:530-533.

62. Douglas NJ, White DP, Weil JV, et al: Hypoxic ven-tilatory response decreases during sleep in normalmen. Am Rev Resp Dis 1982;125:286-289.

63. Hedemark LL, Kronenberg RS: Ventilatory andheart rate response to hypoxia and hypercapniaduring sleep in adults. J Appl Physiol 1982;53:307-312.

64. Douglas NJ: Control of ventilation during sleep.Clin Chest Med 1985;6:563.

65. Birchfield RI, Sieker HO, Heyman A: Alterationsin respiratory functions during natural sleep.J Lab Clin Med 1959;54:216-222.

66. Douglas NJ, White DP, Pickett CK, et al:Respiration during sleep in normal man. Thorax1982;37:840-844.

67. Gothe B, Altose MD, Goldman MD, CherniackNS: Effects of quiet sleep on resting and CO2-stimulated breathing in humans. J Appl Physiol1981;50:724-730.

68. Bulow K: Respiration and wakefulness in man.Acta Physiol Scand 1963;59(Suppl 209):1-110.

69. Douglas NJ, White DP, Weil JV, et al:Hypercapnic ventilatory response in sleepingadults. Am Rev Resp Dis 1982;126:758-762.

70. Berthon-Jones M, Sullivan CE: Ventilation andarousal response to hypercapnia in normal sleep-ing adults. J Appl Physiol 1984;57:59-67.

71. White DP: Occlusion pressure and ventilationduring sleep in normal humans. J Appl Physiol1986;61:1279-1287.

72. Anderson JV Jr, Martin RJ, Abboud EF, et al:Transient ventilatory response to CO2 as a func-tion of sleep state in full-term infants. J ApplPhysiol 1983;54:1482-1488.

73. Davi M, Sankaran K, Maccallum M, et al: Theeffect of sleep state on chest distortion and onthe ventilatory response to CO2 in neonates.Pediatr Res 1979;13:982-986.

74. Haddad GG, Leistner HL, Epstein RA, et al:CO2-induced changes in ventilation and ventila-tory pattern in normal sleeping infants. J ApplPhysiol 1980;48:684-688.

75. Rigatto H, Kalapesi Z, Leahy FN: Chemical con-trol of respiratory frequency and tidal volumeduring sleep in preterm infants. Respir Physiol1980;41:117-125.

76. Honma Y, Wilkes D, Bryan MH, Bryan AC: Ribcage and abdominal contributions to ventilatoryresponse to CO2 in infants. J Appl Physiol 1984;56:1211-1216.

77. Moriette G, Van Reempts P, Moore M, et al: Theeffect of rebreathing CO2 on ventilation anddiaphragmatic electromyography in newborninfants. Respir Physiol 1985;62:387-397.

78. Tarchanoff J: Quelques observations sur le som-meil. Arch Ital Biol 1894;21:318-321.

79. Hodes R, Dement WC: Depression of electricallyinduced reflexes (“H” reflexes) in man duringlow voltage EEG sleep. Electroencephalogr ClinNeurophysiol 1964;17:617-629.

80. Rohmer F, et al: La motilité spontanée, lafréquence cardiaque et la fréquence respiratoireau cours du sommeil chez l’homme normal:Leurs relations avec les manifestations electroen-cephalographiques et la profondeur du sommeil.


Ch07.qxd 1/25/05 3:15 AM Page 83

In La Société d’Electroencephalographie et deNeurophysiologie Clinique de Langue Française(eds): Le Sommeil de Nuit Normal etPathologique, Études Electroencephalogra-phiques. Paris, La Société d’Electroencephalo-graphie et de Neurophysiologie Clinique deLangue Française, 1965, pp 192-207.

81. Jacobson A, Kales A, Lehmann D, HoedemakerFS: Muscle tonus in human subjects during sleepand dreaming. Exp Neurol 1964;10:418-424.

82. Baldridge BJ, Whitman RM, Kramer M: The con-currence of fine muscle activity and rapid eyemovements during sleep. Psychosom Med1965;27:19-26.

83. Chandler SH, Nakamura Y, Chase MH:Intracellular analysis of synaptic potentialsinduced in trigeminal jaw-closer motoneuronsby pontomesencephalic reticular stimulationduring sleep and wakefulness. J Neurophysiol1980;44:372-382.

84. Wilde-Frenz J, Schulz H: Rate and distributionof body movements during sleep in humans.Percept Mot Skills 1983;56:275-283.

85. Oswald I, Berger RJ, Jaramillo RA, et al:Melancholia and barbiturates: A controlled EEGand eye-movement study of sleep. Br JPsychiatry 1963;109:66-78.

86. Gardner R, Grossman WL: Normal motor pat-terns in sleep in man. In Weitzman ED (ed):Advances in Sleep Research, vol 2. New York,Spectrum, 1975, pp 67-107.

87. Mandell A, et al: Dreaming sleep in man:Changes in urine volume and osmolality.Science 1966;151:1558-1560.

88. Rubin RT, Poland RE, Gouin PR, Tower BB:Secretion of hormones influencing water andelectrolyte balance (antidiuretic hormone, aldos-terone, prolactin) during sleep in normal adultmen. Psychosom Med 1978;40:44-59.

89. Halverson H: Genital and sphincter behavior onthe male infant. J Gen Psychol 1940;56:95-136.

90. Ohlmeyer P, Brilmayer H, Hullstrung H:Periodische vorgange im Schlaf. Pflugers Arch1944;248:559-560.

91. Fisher C, Gross J, Zuch J: Cycle of penile erec-tions synchronous with dreaming (REM) sleep:Preliminary report. Arch Gen Psychiatry1965;12:29-45.

92. Karacan I, Williams RL, Thornby JI, Salis PJ:Sleep-related penile tumescence as a function ofage. Am J Psychiatry 1975;132:932-937.

93. Karacan I, Rosenblum A, Williams R: The cli-toral erection cycle during sleep. Psycho-physiology 1970;7:338.

94. Abel GG, Murphy WD, Becker JV, Bitar A:Women’s vaginal responses during REM sleep.J Sex Marital Ther 1979;5:5-14.

95. Snyder F, Hobson JA, Morrison DF, Goldfrank F:Changes in respiration, heart rate, and systolicpressure in human sleep. J Appl Physiol1964;19:417.

96. Finch PM, Ingram DM, Henstridge JD,Catchpole BN: Relationship of fasting gastro-duodenal motility to the sleep cycle.Gastroenterology 1982;83: 605-612.

97. Orr W, Hall WH, Stahl ML, et al: Sleep patternsand gastric acid secretion in duodenal ulcer dis-ease. Arch Intern Med 1976;136:655-660.

98. Orr W, Robinson M: The sleeping gut. Med ClinNorth Am 1981;65:1359-1376.

99. Orr W, Robinson M, Johnson L: Acid clearanceduring sleep in the pathogenesis of refluxesophagitis. Dig Dis Sci 1981;26:423-427.

100. Johnson LF, DeMeester TR, Haggitt RC:Esophageal epithelial response to gastroe-sophageal reflux: A quantitative study. Am J DigDis 1978;23:498-509.

101. Helm JF, Dodds WJ, Riedel DF, et al: Determinantsof esophageal acid clearance in normal subjects.Gastroenterology 1980;78:1181.

102. Dent J, Dodds WJ, Friedman RH, et al:Mechanism of gastroesophageal reflux in recum-bent asymptomatic human subjects. J ClinInvest 1980;65:256-257.


Ch07.qxd 1/25/05 3:15 AM Page 84

A circadian rhythm is defined as an intrinsic bio-logic rhythm with a periodicity (peak-to-peakduration) of approximately 24 hours. Circadianrhythms are a basic property of higher life forms,including plants and animals. In mammals, cir-cadian rhythms are coupled to the timing ofsleep and wakefulness. Circadian rhythms undergoa developmental course, from their beginningsin utero to adult expression. This chapter describesbasic properties of circadian rhythms and thenreviews some of their fundamental properties asthey develop during childhood. Animal studiesare employed to illustrate findings that are appli-cable to infants and children.

Entrainment of the fetus to day and nightbegins in utero as a passive response to maternalmelatonin secretion. In the perinatal period, vari-ous expressions of an endogenous circadian pace-maker neuroanatomic structure (including thesuprachiasmatic nucleus, the pineal body, and theirconnections to the retina) develop sequentially,resulting in the successive appearance of circadianrhythms of melatonin secretion, temperature,wakefulness, and sleep. Circadian rhythms arefully expressed within the first 6 months of life.

In the infant and young child, the timing oflight exposure controls the timing of melatoninsecretion and consequently the timing of sleep.In early childhood and in preadolescent children,the phase relationship of the child’s circadianrhythm to environmental light and darkness isvariable. The hour at which the child is awak-ened, the timing and brightness of artificial light-ing after sunset and before sunrise, the timing ofexercise, the child’s feeding schedule, and thetiming of social interactions might each influencethe child’s timing of sleep propensity. The num-ber of hours of darkness (less than 3 lux) towhich the child is exposed may influence theduration of the child’s melatonin secretion andthe number of hours that the child sleeps.

It is becoming increasingly clear that sleepquality impacts human development. Both thehours at which a child sleeps and the duration ofsleep appear to be modifiable. Beginning inutero, and continuing through childhood, envi-ronmental and social activities are capable ofinfluencing the timing and duration of sleep.

THE CIRCADIAN RHYTHM SYSTEM:BASIC PROPERTIES

The duration of daylight changes continually.The magnitude of change from the shortest to thelongest day of the year is directly dependent onlatitude. The farther one is from the equator, thegreater the change from shortest day to longestday. Without artificial light, human and mam-malian ancestors were in synchrony with changesin photoperiod (day length) that varied fromgreater than to less than 12 hours. The circadianrhythm system must be viewed as having thecapacity to track continually changing times oflight exposure, including changing times of sun-rise and of sunset.

In contrast to sleep homeostasis, in whichsleep drive increases with increased awake time,the body’s internal clock mechanism, or circadianrhythm, exhibits a cyclical tendency that is mostlyindependent of prior wakefulness—this is thecircadian rhythm of sleep propensity.1 Thus,a child or adolescent may be unable to beginnocturnal sleep at a desired hour despite sleep-ing little on previous weekday nights andexhibiting daytime sleepiness during schoolhours. For example, a child with a 9 PM bedtimemight remain awake until midnight. Biologicmarkers of circadian rhythms, such as the dailyrise and fall in core body temperature, continuein the absence of normal social, lighting, orother time cues. Each individual possesses aprecise internal clock, the periodicity of which

Chronobiology of Sleep in ChildrenJohn H. Herman

8

85

Ch08.qxd 01/28/05 8:25 PM Page 85

can be “unmasked” in various types of labora-tory protocols. This clock typically has a peak-to-peak cycle duration of approximately 24.1 to24.3 hours.2,3

Period LengthThe duration of the circadian rhythm (Fig. 8–1),sometimes referred to as period length, or tau,appears to be a genetically heritable property ofthe suprachiasmatic nucleus (SCN). Periodlength (see Fig. 8–1A) is measured from onecycle’s temperature minimum to the next, or fromthe onset of melatonin secretion on consecutivedays, in experimental conditions that eliminateenvironmental influences.4 The SCN contains anautonomous circadian pacemaker. It is the site ofgeneration of circadian rhythmicity. Metabolicand electrical activity rhythms in the SCN havebeen observed in vivo, and the SCN maintainsrhythmicity in vitro. A transplanted SCN from adonor can restore circadian function after thedestruction of a host’s SCN. Single SCN “clockcells” exhibit independent firing-rate rhythms.The proteins responsible for rhythmicity withinthe SCN have been identified in circadian mutants(tau mutant hamsters and Clock mutant mice).These mutants have enabled the isolation of so-called clock genes.5

AmplitudeThe amplitude of the circadian rhythm—that is,the change in a measured parameter from nadirto peak (acrophase)—is an estimate of the bio-logic capacity to oscillate every day from deepsleep to intense alertness. The amplitude of thecircadian rhythm appears to be minimal at birthand to attain adult levels within the first year oflife, as measured by melatonin or temperature.Figure 8–1B shows the effects of constant rou-tine (such as staying in bed under conditions ofcontinuous dim illumination for 24 hours) ondamping circadian rhythm amplitude. Factorssuch as exercise and a constant bed and wake-uptime do not affect the amplitude of the circadianrhythm.6

PhaseThe phase of a circadian rhythm refers to therelationship between the internal timekeeping

mechanism and the environmental time. Ideally,core temperature drops and sleepiness occursat the desired hour of sleep, and body tempera-ture is rising and an individual feels rested at thedesired hour of awakening. This is referred toas being “in phase.” Phase advance refers to theinternal clock’s being set early with respectto environmental time, and phase delay refers tothe internal clock’s being set late with respect toenvironmental time (Fig. 8–1C). The tendency for

86 Chapter 8 Chronobiology of Sleep

Normal activityConstant routine

Amplitude (nadir to peak)

37.5

37.0

36.5

36.0

Tem

pera

ture

(�C

)

0

A

B

12 24

Hour

Tau

Nadir

Am

plitu

de

12 24

Dee

p bo

dy te

mpe

ratu

re (

�C)

–12 –9 0–6 –3

Time from sleep onset (hours)

Legend

3 6 9 12

38.0

37.5

37.0

36.5

36.0

Figure 8–1. Properties of a single cycle of the circadianrhythm. A, One cycle of the circadian rhythm of tempera-ture drawn twice to illustrate the duration of the circadianrhythm (tau, or τ), here measured from succeeding tem-perature nadirs, and the amplitude of the circadian rhythm,in this case, temperature difference from apogee to nadir.B, The amplitude of the circadian rhythm is reduced by thesubject remaining in bed for 24 hours (constant routine,CR) compared to the amplitude of temperature from peakto nadir when the subject is engaged in normal wakingactivity.

Ch08.qxd 01/28/05 8:25 PM Page 86

Chronobiology of Sleep Chapter 8 87

1200 1600 2000 2400 0400 0800

Advancedsleep phase

syndrome

Typical sleep phase

1200

Delayed sleep phasesyndrome

C

20

15

10

5

0

Pla

sma

cort

isol

(μg/

dL–1

)

50

40

30

20

10

0

Pla

sma

mel

aton

in(p

g/m

L–1 )

4

3

2

Sle

epin

ess

(Sta

nfor

d sc

ale)

2 4 6 8 10 12

Sleep

Darkness

Sleep

2

Period ofmelatoninsecretion

LD 10:14

n = 10

LD 16:8

4 6 8 10 12 2 2 4 6 8 10 12 2 4 6PM AMPM AM

8 10 12 2

D

Period ofmelatoninsecretion

Figure 8–1. cont’d—C, Phase of circadian rhythm refers to the relationship of the internal clock to envi-ronmental timing of light and dark, the top bar illustrates advanced sleep-phase syndrome (sleep from 7:00PM to 5:00 AM), the middle bar illustrates normal phase timing (sleep from 11:00 PM to 8:00 AM), and thebottom bar illustrates delayed sleep phase syndrome (sleep from 2:30 AM to 11:00 AM). D, The relativeduration of biological day to biological night changes as the light-to-dark ratio is changed from 16hours of light and 8 hours of darkness (left, black bar) to 10 hours of light and 14 hours of darkness(right, black bar), resulting in an increased duration of melatonin secretion (top), delayed cortisol secre-tion (middle), and increased duration of temperature suppression (bottom). (A and C, redrawn fromBaker SK, Zee PC: Circadian disorders of the sleep wake cycle. In Kryger M, Roth T, Dement W (eds):Principles and Practice of Sleep Medicine, 3rd ed. Philadelphia, WB Saunders, 2000, pp 607, 610; Dredrawn from Wehr TA, Aeschbach D, Duncan WC: Evidence for a biological dawn and dusk in thehuman circadian timing system. J Physiol 2001;535:937-951.)

Ch08.qxd 01/28/05 8:25 PM Page 87

phase to be delayed appears to increase fromchildhood to adolescence.7

Morningness–EveningnessPreference and PhaseThe duration of an individual’s circadian rhythmis correlated with “owl” versus “lark” behavior(referred to as morningness versus eveningness):the longer the period length, the greater the owltendencies. Evening types (owls) have a lateronset of melatonin secretion, their morningtemperature rise is later, and they wake up laterthan do morning types. Dropping temperature isrelated to sleep onset and rising temperatureis correlated with preferred wake-up time.8

Individuals who are maximally alert in themorning have an earlier peak in circadian tem-perature than do individuals who are most alert

in the evening. Individuals with morning prefer-ence have a greater change in temperature fromnadir to peak. These lark types awaken andbegin sleep earlier than individuals with eveningpreference.9 The phase of temperature and mela-tonin secretion can be predicted fairly accuratelyby having an individual maintain a sleep log forseveral weeks, even with irregular sleep times.10

With the approach of adolescence, circadianphase shifts to a later hour, as does an increase inevening preference for activities. This occurs atabout age 12 to 13 years.11

Biological Day and Biological NightA property of any waveform is the proportionof the wave cycle during which values areabove the mean as opposed to below the mean.From the perspective of circadian rhythms, this


Figure 8–2. A, The phase-response curve to light (in Syrian hamsters) showing the change in thephase of the circadian rhythm in response to exposure to light at different times. The phase-responsecurve shows no phase-shifting effect of light during the circadian hours of normal light exposure(typically, the hours of daylight). Exposure to light before the moment of singularity results in a maxi-mal phase-delaying effect during the hours of normal darkness (negative numbers). Exposure to lightafter the moment of singularity results in a maximal phase-advancing effect during the hours of nor-mal darkness (positive numbers). A person’s circadian time must be known to predict whether theresponse to bright light will have a phase-advancing or a phase-delaying effect. Point of singularity:the time at which light exposure (or another zeitgeber) switches from having a phase-advancing to aphase-delaying effect on the circadian rhythm. B, Circadian sleep propensity double-plotted for onecircadian cycle, showing the forbidden sleep zone (a brief time period typically preceding sleep onset),during which sleep is least likely to occur during the circadian day. (A, redrawn from Takahashi JS, ZatzM: Regulation of circadian rhythmicity. Science 1998;2178:1104-1111; B, redrawn from Dijk D-J,Czeisler CA: Contribution of the circadian pacemaker and the sleep homeostat to sleep propensity, sleepstructure, electroencephalographic slow waves and sleep spindle activity in humans. J Neurosci1995;15:3526-3538.)

2

1

0

–1

–2

Pha

se s

hift

(hou

rs)

0 6 12

Circadian time (hours)

Phase delay portion

Phase advance portion

18 24

Point ofsingularity

PHASE-RESPONSE CURVE

A

Forbidden sleep zone

30

20

10

0

Wak

eful

ness

(% o

f rec

ordi

ng ti

me)

6 14 22

Time of day during entrainment

6 14 22 6

CIRCADIAN SLEEP PROPENSITY

B

Ch08.qxd 01/28/05 8:25 PM Page 88


refers to the ratio of biological day (above themean) to biological night (below the mean).Figure 8–1D displays two different lengths ofenvironmental light and dark: a 16-hour lightperiod followed by an 8-hour dark period (on theleft) and a 10-hour light period followed by 14hours of darkness (on the right). At latitudesboth north and south of the equator, half the yearhas day length greater than night length. Thismeans that environmental night is greater than12 hours for half of the year in almost everycountry, a situation that was far more behav-iorally significant before the invention of indoorlight, beginning with candles. Figure 8–1Dshows that the duration of biological night canchange in response to changing durations ofenvironmental night length, including a length-ening in the duration of melatonin secretion.12

The capacity to increase the duration of bio-logical night is a critical factor in appreciatingthe flexibility of the human sleep–wake cycle.

Contemporary humans, in essence, attempt toexist year round on a schedule similar to a shortsummer night.13

Implicit in the description of long versusshort biological nights is the hypothesis that thereis one switch (entrained to dusk) for the begin-ning of biological night, and a second switch(entrained to dawn) for the beginning of biolog-ical day (Fig. 8–3).13 The end of biological nightdoes not occur at a set time after its beginning,and the time of awakening is not a constantinterval after dim-light melatonin onset(DLMO). Instead, the switches for biologicalnight and day are capable of tracking changingenvironmental day lengths.9 This is referred to asa two-oscillator model.14 The duration of sleep,temperature suppression, and cortisol suppres-sion each increases as biological nights lengthen.The concept of a flexible (within limits) biologi-cal night has significant implications for humandevelopment.

Figure 8–3. Biological day coincides with the environmental time from dawn to dusk, and in diurnalspecies, including humans, it exhibits qualities compatible with alertness and activity. In diurnalspecies, biological night, which coincides with darkness, is compatible with rest and immobility. Thedurations of biological day and biological night are responsive to changes in the duration of daylightas it varies from less than 12 to greater than 12 hours, depending on latitude and season. (Redrawnfrom Wehr TA, Aeschbach D, Duncan WC: Evidence for a biological dawn and dusk in the human cir-cadian timing system. J Physiol 2001;535:937-951.)

Dawn

Legendtypical daily rhythmaveraged daily rhythm

Biological day

No melatonin secretionIncreasing core body temperatureDecreasing sleepinessDecreasing sleep propensityWakefulness

Biological night

Melatonin secretionDecreasing core body temperatureIncreasing sleepinessIncreasing sleep propensitySleep

Dusk

Time

Tem

pera

ture

Temperaturerhythm

Ch08.qxd 01/28/05 8:25 PM Page 89


Consolidated Sleep PeriodOptimal sleep is one manifestation of biologicalnight. The timing of biological night and anincreased sleep propensity (sleepiness) are initi-ated by the beginning of darkness, or dusk, whichis signaled from the retina to the SCN via the reti-nal hypothalamic pathway (Fig. 8–4). Dusk per-mits the expression of melatonin (i.e., DLMO) bythe pineal body, and it begins the nocturnal phaseof the circadian rhythm, or biological night. Thisincludes a drop in core body temperature, whichreaches a nadir approximately 2 hours prior tohabitual wake-up time (see Figs. 8–1A andFig. 8–3), and a concomitant decrease in glucoseutilization,15 urinary metabolism, and appetite.After the temperature begins its nightly drop andserum melatonin begins its rise, sleep propensityincreases rapidly and the consolidated period ofnocturnal sleep begins (see Fig 8–3).16

Rate of Change from BiologicalDay to Biological NightHuman circadian rhythms have traditionally beenrepresented graphically as a plot of core bodytemperature or melatonin secretory levels aver-aged from several subjects. The resulting plot ofmean values resembles a sine wave (Figs. 8–3and 8–5A) that is similar to a frequency his-togram recording electrophysiologically from

SCN neurons.17 However, if melatonin secretoryprofiles are measured consecutively, time-lockedto melatonin onset or offset, the resulting wave-form resembles more of a square wave, witha rapid transition from biological day to biologi-cal night, as shown in Figure 8–5B.8 This rapidtransition is consistent with experimental find-ings in constant-routine protocols of a brief sleep“gate,” before which sleep is unlikely to occurand after which sleep is likely to occur.18 As daychanges to night, the output of the SCN switchesan organism from one mode of functioning,in which it is interactive with its environment,to another, in which it is nonresponsive andimmobile.13

Eye~

SCN~

Outputs

TemperatureCortisolMelatoninSleep-wake cycleBehaviorCardiovascular function

RHTλ

70

60

50

40

30

20

10

0Pla

sma

mel

aton

in (

pg/m

L–1)

2 4 6

PMA

B

AM

n=55

Melatonin offset

Melatoninonset

PM AM

8 10 12 2 4 6 8 10 12 2

70

60

50

40

30

20

10

0Pla

sma

mel

aton

in (

pg/m

L–1)

2 4 6 8 10 12 2 4 6 8 10 12 2

Figure 8–4. The input to the suprachiasmatic nuclei(SCN) is via special retinal receptors that form a synapticconnection to the retinohypothalamic tract, which car-ries light information to the SCN. These light receptorsdo not form synaptic connections with the optic nerve,and they transmit no visual information to the CNS visualprocessing centers. Therefore, conscious light perceptionand SCN input are via distinct pathways; if either isdestroyed, the other can remain fully functional. (FromRivkees S: Mechanisms and clinical significance of circa-dian rhythms in children. Curr Opin Pediatr 2001;13:352-357.)

Figure 8–5. A, Melatonin secretory profile derived fromaveraging serum levels of melatonin. The result is a sinu-soidal waveform with a periodicity of approximately24 hours. B, Plotted values are derived from averagingforward and backward from each subject’s time-lockedmelatonin onset and offset. B reveals melatonin as anon-or-off switch, demarcating the onset of biologicalnight and the onset of biological day. (Redrawn fromWehr TA, Aeschbach D, Duncan WC. Evidence for a bio-logical dawn and dusk in the human circadian timingsystem. J Physiol 2001;535:937-951.)

Ch08.qxd 01/28/05 8:25 PM Page 90


Point of SingularityEach cycle of a circadian rhythm has an acrophaseand a nadir (see Fig. 8–1D). The circadian nadiris significant: it is near the minimum of body tem-perature. Before this time, light has a phase-delaying effect, and after this time, light hasa phase-advancing effect19 (see Fig. 8–2A). Some-times referred to as the point of singularity, thecircadian nadir is coincident with the beginningof the rises of core body temperature, cortisolsecretion, and the proportion of sleep occupiedby rapid eye movement sleep. Exposure to brightlight centered at the point of singularity signifi-cantly reduces the amplitude of core body tem-perature, plasma cortisol secretion, and melatoninsecretion.20

Forbidden Sleep ZoneThe “forbidden sleep zone” immediately precedesthe onset of biological night. This is the timeat which sleep is least likely to occur during a cir-cadian cycle (see Fig. 8–2B). This 1- to 3-hourperiod is well established in human adults andadolescents. Because the forbidden sleep zone pre-cedes the normal hour of sleep, many individualswith delayed sleep phase have difficulty advancingtheir hour of sleep despite sleep deprivation anddaytime sleepiness.21,22 Although forbidden sleepzones have not been studied in children, theyappear to be particularly robust, as even childrenwho are sleepy enough to fall asleep in school arenot able to begin nighttime sleep at an earlier hour.

Circadian OutputsAn extremely large number of biologic, behavioral,cognitive, and emotional factors demonstrateconsistent and predictable circadian rhythms.Core body temperature (see Fig. 8–1A), hormonesecretion, and cell metabolism are among the bet-ter-known biologic parameters that demonstrate acircadian rhythm.23 Human performance is welldocumented to demonstrate a circadian rhythm,and a variety of measures of cognitive skills showcircadian changes in performance. Mood and emo-tions also follow predictable circadian pat-terns.24,25 It is currently believed that most of ourorgans contain their own clock mechanisms, andthat each of these maintains a synchronous rela-tionship with the body’s master clock via outputsfrom the SCN.26

Phase ShiftsOnly disrupted phase, and no other aspect of thecircadian rhythm, is considered a circadian sleepdisorder. The phase of the circadian rhythm issynchronized with environmental time by a vari-ety of zeitgebers (or time givers, and thus clocksetters), principally bright light, but also includ-ing the timing of sleep and the timing of socialcontacts. Suitably timed exposures to brightlight result in large shifts in the circadian phase.27

In contrast to bright light (and melatonin), mostother factors have only weak and inconsistenteffects in shifting circadian phase.28

The phase-shifting effect of bright light (orany other zeitgeber) requires not only that it bedelivered at a suitable time (see Fig. 8–2A), forsufficient duration, and at necessary intensity,but also that the individual not be exposed tolight at times where the phase-shifting effectwould be in the opposite direction.29 For exam-ple, exposure to bright light during normalhours of morning sleep (after the point of singu-larity) will not result in a phase advance unlessdarkness is present during the earlier (evening)time interval to which sleep is to be phaseadvanced.

It appears that both light of ordinary roomintensity30,31 and daytime napping32 are capableof inducing a small phase shift. These so-calledweak zeitgebers33 have been demonstrated tochange circadian phase under conditions of con-stant routine, an experimental technique thateliminates all stimuli that would interfere with aphase shift. It is not clear how effective weakzeitgebers would be in shifting circadian rhythmsunder normal conditions.

MelatoninMelatonin is secreted by the pineal gland duringdarkness, and the timing of its secretion appearsto be under the control of the SCN.34 Melatoninis sometimes referred to as the hormone of dark-ness. The endogenous rhythm of melatoninsecretion is entrained by the light–dark cycle.35

Light is able to both suppress and entrain mela-tonin production.34 The normal timing of sleepis tightly coupled to the melatonin secretoryprofile. Melatonin is associated with sleep indiurnal mammals and with waking in nocturnalanimals.

Ch08.qxd 01/28/05 8:25 PM Page 91

Melatonin administration and light exposureshift circadian phase in opposite directions.Thus, melatonin administration before DLMOresults in a phase advance, and melatoninadministration after the morning melatoninoffset results in a phase delay. The phase-response curve to melatonin is similar to thephase-response curve to light but in the oppo-site direction.36,37 Melatonin is more effectivein altering the timing of circadian rhythms thanthat it is as a hypnotic.38 The administrationof exogenous melatonin alters the timing ofendogenous melatonin secretion on subsequentdays.39

In a constant-routine experimental paradigm,in which sleep is permitted for only 10 minutes,followed by 20 minutes of forced wakefulnessfor multiple 24-hour periods, sleep occurs onlyduring intervals that coincide with melatoninsecretion (greater than 10 pg/mL has been sug-gested as a minimal sleep-inducing melatoninlevel4). During intervals when melatonin is notsecreted, sleep does not occur.40

The nocturnal onset of melatonin secretionis time-locked in humans to the opening of thenocturnal sleep gate: as melatonin secretorylevels increase, body temperature falls. In avariety of free-running conditions, sleepoccurs only when body temperature is drop-ping or low. These findings suggest that mela-tonin participates in sleep–wake regulation inhumans.18

MaskingThe environment exerts influences that keepan individual entrained with local day and night.These are exogenous influences that “mask”the body’s endogenous circadian rhythm.Masking is so universally present that thebody’s circadian rhythm may be observed onlyunder extremely controlled experimental con-ditions that eliminate the multiple zeitgebersof everyday life that keep all of us entrained.The timing of bright and dim light exposure,social activities, exercise, and meals all inducea masking effect.41 Some degree of masking isoccurring continually in all mammals whoseendogenous circadian rhythm is not precisely24 hours, as environmental zeitgebers synchro-nize circadian rhythms with the local hours oflight and darkness.

EntrainmentEntrainment refers to an individual’s hours ofbiological day and biological night being in syn-chrony, or in phase, with environmental day andnight. Entrainment of circadian rhythms to anenvironmental schedule appears to be accom-plished by a variety of factors, many of whichhave already been discussed: bright or dim light,social schedule, melatonin,42,43 social activities,44

and exercise.45 Bright light induces greaterentrainment than dim light; specifically, expo-sure to bright light induces a greater eveningdrop and morning rise in core body temperature.Daytime exposure to bright light also results inan advance in the circadian rhythm of body tem-perature compared to dim light.46 These findingssuggest that both dim and bright light entrainan organism to environmental phase, or time, butthat bright light further enhances entrainmentby inducing a greater amplitude of the tempera-ture rhythm and an earlier rise in core body tem-perature. These findings suggest that children andadolescents with sleep initiation problems wouldbenefit from daytime bright light exposure.

Effect of Indoor Light on PhaseThe light of dawn (an average of 155 lux, simi-lar to indoor lighting) is sufficient to induce aphase advance in normal subjects. It is also suf-ficient to prevent phase delay in subjects livingin constant dim illumination.47 Subjects kept inconstant dim light (as opposed to timed indoorlight) for 21 days who had knowledge of clocktime and who had social cues in a group settingdemonstrated free-running circadian rhythms,with a periodicity of 24{1/4} hours in melatoninand temperature rhythms.48 This study showsthat normal individuals may not be able toremain entrained to environmental time in theabsence of timed light as a zeitgeber. Ordinaryroom light is capable of shifting the rhythms oftemperature, melatonin,49 and cortisol, indicat-ing that the master circadian pacemaker hasshifted in response to indoor light.15,31 Both tem-perature and hour of sleep shifted to the earlierphase, indicating a change in the output of themaster clock, similar to results observed in stud-ies using bright light. It has also been shown thatthe timing of exposure to normal room lightmodulates the effect of bright light on circadian


Ch08.qxd 01/28/05 8:25 PM Page 92

phase, such as room light exposure during bio-logical night, including television sets and com-puter monitors. This suggests that even in thepresence of normally timed bright light, indoorlight after or before normal daylight hours mayinduce a phase shift in the circadian rhythm.50

These studies collectively indicate that indoorroom light is sufficient to continually exert weakinfluences on circadian rhythms, but that it isinsufficient to abruptly switch rhythms to a vastlyaltered sleep–wake schedule.

Resetting Circadian Time in ChildrenThe phase-response curve (PRC), or the degreeto which the circadian rhythm will be advancedor delayed, is based on the time at which brightlight or melatonin is administered. Typically, theadvance or delay of the circadian system is meas-ured in the number of minutes that the circadiantemperature nadir is advanced or delayed.

Timing of Light

The PRC shows that there is virtually no changein circadian rhythms (for example, timing of thetemperature nadir) induced by exposure to lightduring the middle of the circadian “day”; onthe other hand, there is maximal responsivenesswhen exposure occurs immediately before orafter the temperature nadir.51 Other investiga-tors have found that under certain experimentalconditions, light delivered at any time during thecircadian day had a slight phase-shifting effect.50

At about the time of the temperature minimum(point of singularity), a switch is thrown, andat this point light changes from having a phase-delaying to a phase-advancing effect: immedi-ately before the temperature minimum, light hasits maximal phase-delaying effect, and immedi-ately after the temperature minimum, light hasits maximal phase-advancing effect.52

Intensity of Light

The intensity of the light greatly influences themagnitude of the response, with bright light(about 5000 to 10,000 lux) having a muchgreater effect than normal room light (100 to200 lux). Even dim light (less than 100 lux)might have a slight effect on shifting circadian

rhythms.53 The timing of sleep and darkness isas important as the timing of the bright light.54

A homologous phase-response relationshipprobably exists for melatonin and circadianrhythms, in which melatonin has the oppositeeffect of light: evening melatonin would advancethe circadian rhythm and morning melatoninwould exert a phase-delaying effect.22 A singledose of melatonin properly timed is capable ofadvancing the human circadian clock.55 In sub-jects living in constant dim light for a month,melatonin administered 2 hours before thetemperature acrophase had a maximal effecton advancing the circadian rhythm.34 Melatoninsynthesis normally begins about the hour of sun-set; in indoor-living humans, it typically beginsto be present in plasma shortly after sunset(about 8 PM to 9 PM) and gradually rises to a peakabout 3 AM to 4 AM, followed by a decline inplasma concentration to near zero by 9 AM. BothDLMO and melatonin offset have been foundto be as reliable as core body temperature inestimating circadian rhythms.3

DEVELOPMENT OF CIRCADIAN RHYTHMS

In Utero and in the Neonatal PeriodSome aspects of circadian rhythm appear tobegin in utero in mammalian species. The maturecircadian rhythm system requires connectionsfrom a fully functioning SCN to retinal inputsand to multiple targeted outputs. The circadiansystem is not fully functional at birth in mostmammalian species. However, circadian rhythmsbegin in utero as the fetus becomes entrained tothe mother’s rhythm. From the standpoint of thefetus, this is an exogenous rhythm, externallyimposed by factors such as maternal temperatureor melatonin, prior to the development of thefetus’s own fully developed SCN. In mammals,maternal melatonin and the maternal circadiantemperature cycle convey information aboutenvironmental time. In some species, the retino-hypothalamic tract transmits light informationto the fetal SCN in utero, which is capable ofbecoming entrained to environmental lightshortly prior to birth. However, light appears tobe a less potent entraining factor in utero thanmaternal melatonin secretion.56


Ch08.qxd 01/28/05 8:25 PM Page 93

The Suprachiasmatic Nucleus

Neurogenesis of the SCN occurs 3 to 5 daysbefore birth in rats and hamsters, respec-tively.57,58 If a pregnant hamster is injected withmelatonin the day before delivery, after weaning(on day of life 20) the midpoint of the pups’ sub-jective night will coincide with the time of theinjection. If the pregnant hamster is injectedwith a dopamine agonist, after weaning the mid-point of the pups’ subjective day will coincidewith the time of the injection. This demonstratesthat dopamine and melatonin either are, or mimic,maternal entraining signals that represent dayand night.59,60

The responsiveness of a pup’s circadian clockto melatonin and dopamine agonist injectionsis gone by day 4 of life, replaced by an equallyrobust response to light (subjective night inrodents) and darkness (subjective day in ro-dents).61 If two groups of pregnant hamsters areinjected with a dopamine agonist 12 hours apart,the circadian rhythms of the two groups will be12 hours out of synch on the day of weaning,demonstrating that the prenatal dopamine ago-nist set the phase of the offspring’s circadianrhythms.60 In rats, at about the same time—theend of the first week of life for the rat pup—the SCN develops the ability to be entrainedby light. At this time, the circadian rhythm ofN-acetyltransferase first appears.

The human SCN develops early in gestation,and circadian rhythms are present in the fetusand newborn. The circadian system seems to befunctional in human fetal life and can receivecircadian inputs through the mother.62

Uniqueness of SCN Tissue

For the lateral geniculate nucleus (LGN) todevelop normally, afferents from the retina mustinnervate LGN cells during development. Invisual deprivation studies, LGN hypertrophyresults from interruption of retinal afferentfibers. Visual input from the retina is crucial incell development and neural arborization in theLGN, as in most other CNS structures that arepart of a sensory system. In contrast, there is noevidence that afferents from the retinohypotha-lamic tract, the raphe, or the intrageniculatenucleus influence the development of the SCN.Lesioning afferent fibers does not result in SCNhypotrophy. No other tissue takes on pacemaker

properties if the SCN is enucleated early indevelopment.63,64

Pacemaker properties of SCN cells are intrin-sic: SCN fetal tissue develops its intrinsic rhyth-micity even if transplanted to a totally differentenvironment, such as the anterior chamber of anadult eye.65 Fetal SCN tissue may be mincedprior to transplantation, destroying specific spa-tial relationships, yet intrinsic rhythmicity devel-ops normally.66

Role of Melatonin in Development

When the fetus’s SCN begins to function, mela-tonin from the mother is probably the principalfactor entraining it to the prevailing light–darkcycle.67 Even when a pregnant mother is keptin constant light, the mother’s rhythm and therhythms of her offspring demonstrate a consis-tent phase relationship with each other.68 Thisindicates that entrainment of the newborn toits mother results from intrauterine factors, espe-cially when environmental zeitgebers are absent.

Nonphotic Zeitgebers in Development

In the absence of light cues for day and night,social contact may act as an entraining factor indeveloping mammals. Mice born and raised tothe age of weaning in constant lighting condi-tions remain entrained to their mother’s circadianrhythm. The mouse pup’s daily onset of wheelrunning (nighttime) is entrained to the mother’stermination of nursing, and therefore, the pupstime of wheel running coincides with themother’s absence. In constant lighting conditions,the mother’s return is taken by the pups as theirrest time and her absence as their wake time.67

The newborn’s temperature and melatoninsecretory rhythms are in phase with environ-mental night and day, probably resulting frominfant–mother synchronization. Feeding andrest–activity synchronization to mother and envi-ronment develop rapidly during the neonatalperiod.

In the First Year of LifeFrom birth to 6 months, the human infant devel-ops robust circadian rhythms. The retinohypo-thalamic tract is present in humans before birth,and plasma melatonin demonstrates a rhythm


Ch08.qxd 01/28/05 8:25 PM Page 94

influenced by light within the first 2 days afterbirth.69 Human rhythms are in phase with theirenvironment when they appear, suggestingentrainment that precedes the ability to measurethem. The SCN may be functioning before birth inhumans but may not yet have developed output-entraining mechanisms. Human fetal rhythms areentrained to those of the mother, and the probablesynchronizing substance is melatonin.70 Thehuman fetus shows a rhythm in heart rate andfetal movement activity that is also circadian.71

A weak circadian rhythm, showing newborn-mother entrainment, is observable in 24-hourconsecutive plots of the infant’s tympanic mem-brane temperature during days of life 1 to 14.72

This implies that some degree of entrainment ofthe circadian rhythm of temperature has occurredin utero in the human fetus. If a human infant isexposed to only sunlight (no incandescent orflorescent light) for the first 6 months of life, thesequential development of behavioral and phys-iologic properties of the infant’s circadianrhythm become evident.72 The circadian rhythmof temperature appears first, soon after birth,and becomes statistically significant within1 week. Most apparent in 24-hour temperatureplots of the first 2 weeks of life are a morningincrease and an evening fall in temperature. Thewake circadian rhythm appears soon after,attaining significance at day 45, approximatelythe same time that increased melatonin concen-tration begins to occur at sunset (DLMO,defined as salivary concentration greater than20 pg/mL, equivalent to adult levels).72

The circadian rhythm of sleep appears last,attaining significance after day 56. Ninety- to120-minute zones of sustained wakefulness firstappear in the second month of life, after awak-ening and prior to sleep onset (forbidden sleepzone). The infant’s nocturnal sleep-onset is cou-pled to sunset before day 60, and subsequentlyit is coupled to family bedtime, giving evidenceof initial photic entrainment followed by socialentrainment. Even in the absence of artificiallight after sunset, the infant’s sleep onset timetracks the family’s schedule, as the infant remainsawake for 1 to 3 hours every night in darkness.72

Actigraphy monitoring of rest–activity cyclesin the 3-week-old human infant also shows acircadian rhythm.73

Before sleep or wake demonstrates a circadianrhythm, the infant’s morning awakening appears

to become entrained to the rise in body temper-ature, and the two appear to remain coupled.Morning entrainment of awakening to risingbody temperature appears to precede the cou-pling of sleep onset to sunset; dawn is a morepowerful zeitgeber than dusk. A coupling ofsleep onset to evening darkness first appears ona consistent basis during week 7 of life. A pro-clivity for sleep to occur during the night andwake to occur during the day is present in thefirst weeks of life as well. By week 14, the infanthas developed a wake maintenance zone begin-ning at sunset and continuing for 1 to 3 hours.72

By 6 months of life, the human infant, whenexposed only to sunlight, is entrained to bothit and the social rhythm of the household.Maintenance of wakefulness zones are apparentas the infant remains awake after sunset.72 Justas the organization of sleep stages attains adultcriteria by age 6 months,74 the human circadianrhythm displays period, amplitude, and phaseactivity at 6 months of age that are similar to theseelements in adult human circadian rhythms.It appears that circadian rhythms are one of theearliest maturing physiologic-behavioral systems.

In the Preschool ChildLittle research is available on the preschoolchild’s circadian rhythms, probably because thisis a period when aberrations and idiosyncrasiesare more readily tolerated due to the absence ofsocial scheduling requirements. It appears thatpreschool children are behaviorally similar topreadolescents, who are relatively phase advancedcompared to adolescents.75 The preschool childfrequently becomes entrained to maternal orfamilial rhythms before becoming synchronizedwith a school schedule.

In the School-Aged ChildRemaining awake later on weekends and sleep-ing to a later hour has a phase-delaying effect,making it more difficult for the school-agedchild to initiate sleep on Sunday night and toawaken on Monday morning.76 This effect(remaining awake later and subsequent phasedelay) may be greater over extended schoolvacations. During preadolescence in most chil-dren, sleep gradually becomes more delayedwith respect to biological night, resulting in a


Ch08.qxd 01/28/05 8:25 PM Page 95

propensity to initiate sleep and awaken at a laterhour despite little change in tau, or periodlength. Adolescents are better able to sleep aftertheir body temperature rises and to remainawake after body temperature falls than youngerchildren.77

Carskadon and colleagues2 measured the cir-cadian rhythm duration of 12- to 15-year-olds byhaving them follow a 28-hour day for 12 con-secutive days. They were in dim light for twothirds of the 28 hours (waking period) then indarkness the remaining third of the time (sleepperiod). Because the human circadian clock isunable to synchronize to the 28-hour schedule,the timing of sleep and waking become “uncou-pled” from the body’s temperature and mela-tonin rhythm. In this protocol, measurement ofthe time from one temperature or melatoninpeak to the next indicates the unmasked cyclelength of the body’s clock. The circadian rhythmduration of both temperature and melatonin wasfound to be 24.3 hours, longer than 24 hours,but similar to that found in older individuals.2

This study demonstrates that the age-related,adolescent “night owl” propensity, or tendencytoward delayed sleep and awakening, does notresult from a longer adolescent clock cycle butmust be caused by other factors, such as the tim-ing of sleepiness with respect to the circadiantemperature cycle.7

Another study found that 10th grade and 3rdgrade children sleep the same number of hourson weekends, about 9{1/4} hours, but the bed-times and wake-up times of older children are2 hours later than those of younger children.This indicates that the biological clock of thelate adolescent is set 2 hours later than that ofthe younger child, and that sleep need has notchanged significantly from early childhood toadolescence.75

CONCLUSIONS

Circadian rhythms in humans develop in utero,under control of maternal melatonin early ingestation, and later under control of the fetus’developing SCN. Light suppresses the secretionof melatonin, a hormone that is secreted in dark-ness and is related to biological night in humans.Melatonin levels are higher at approximatelythe time when sleep occurs. The pacemaker sys-tem (anatomically centered in the SCN) receives

information about the presence of light, whichsuppresses melatonin secretion, from the retina.The timing of the pregnant mother’s exposure tolight affects her secretory melatonin levels andthose of the fetus. If the mother’s light exposureduring the later portion of gestation is restrictedto daytime hours, this could increase the likeli-hood that the newborn will first express a sleep-wake pattern in synchrony with environmentallight and dark.

Circadian rhythms of some aspects of humanphysiology are present at birth. A sleep–wakecircadian rhythm that is quite robust emergesin the first 6 months of life, including the hourat which sleep begins and ends. The history of anindividual’s recent light exposure determinesa specific time at which the circadian clockpermits sleep to begin (biological night) and aswell as a time at which an individual is likely toawaken (biological day).

The greater the proportion of the 24-hourperiod that is occupied by darkness (biologicalnight), the longer is the duration of melatoninsecretion, which is associated with an increasein the number of hours of sleep. Thus, boththe circadian rhythm (timing) of sleep and thehomeostasis (amount) of sleep are influencedby light exposure. Data from healthy humansubjects suggest that sleep time increases if theduration of environmental night (dark period)is increased. The same data suggest that sleeptime decreases if the duration of environmen-tal night is decreased. Humans are biologicallyand behaviorally equipped to accommodateto long nights in winter and short nights insummer.

The timing of biologic rhythms is also par-tially under the control of factors other than light,especially if exposure to bright light is curtailed.The timing of social and family activities mayinfluence the timing of sleep. Thus, regular tim-ing of family activities, including meals and bed-times, is likely to facilitate the development ofmore regular rhythms in a human infant, child,or adolescent.

Indoor light can be disruptive to sleep. Wepossess a circadian mechanism that has notadapted to the variety of light- and sound-emit-ting devices that may be present in many house-holds during the hours of night. The presenceof indoor light allows each family (or individual)to select its hours of darkness (and quiet), a phe-


Ch08.qxd 01/28/05 8:25 PM Page 96


nomenon that could be detrimental to theoptimal expression of sleep in children. In effect,artificial light enables the contraction of envi-ronmental night, possibly decreasing a child’stotal hours of sleep.

A family’s timing of activities and their hoursof light exposure influence when and how mucha child sleeps throughout development. Theencroachment of (artificial) daytime into hourshistorically reserved for sleep has unknowninfluences on human development, as this is arecent phenomenon from a biologic perspective.

References

1. Dijk DJ, Czeisler CA: Contribution of the circa-dian pacemaker and the sleep homeostat to sleeppropensity, sleep structure, electroencephalo-graphic slow waves and sleep spindle activity inhumans. J Neurosci 1995;15:3526-3538.

2. Carskadon MA, Labyak SE, Acebo C, Seifer R:Intrinsic circadian period of adolescent humansmeasured in conditions of forced desynchrony.Neurosci Lett 1999;260:129-132.

3. Czeisler CA, Duffy JF, Shanahan TL, et al: Stability,precision, and near-24-hour period of the humancircadian pacemaker. Science 1999;284:2177-2181.

4. Lewy AJ, Cutler NL, Sack RL: The endogenousmelatonin profile as a marker for circadian phaseposition. J Biol Rhythms 1999;14:227-236.

5. Weaver DR: The suprachiasmatic nucleus: A 25-year retrospective. J Biol Rhythms 1998;13:100-112.

6. Waterhouse J, Minors D, Folkard S, et al: Lack ofevidence that feedback from lifestyle alters theamplitude of the circadian pacemaker in humans.Chronobiol Int 1999;16:93-107.

7. Carskadon MA, Vieira C, Acebo C: Associationbetween puberty and delayed phase preference.Sleep 1993;16:258-262.

8. Duffy JF, Rimmer DW, Czeisler CA: Association ofintrinsic circadian period with morningness-eveningness, usual wake time, and circadianphase. Behav Neurosci 2001;115:895-899.

9. Horne JA, Ostberg O: A self-assessment question-naire to determine morningness-eveningness inhuman circadian rhythms. Int J Chronobiol 1976;4:97-110.

10. Martin SK, Eastman CI: Sleep logs of youngadults with self-selected sleep times predict thedim light melatonin onset. Chronobiol Int 2002;19:695-707.

11. Shinkoda H, Matsumoto K, Park YM, NagashimaH: Sleep-wake habits of schoolchildren accord-ing to grade. Psychiatry Clin Neurosci 2000;54:287-289.

12. Wehr TA, Aeschbach D, Duncan WC: Evidencefor a biological dawn and dusk in the humancircadian timing system. J Physiol 2001;535:937-951.

13. Wehr T: The impact of changes in nightlength(scotoperiod) on human sleep. In Turek FW,Zee PC (eds): Regulation of Sleep and CircadianRhythms. New York, Marcel Dekker, 1999, pp263-285.

14. Pittendrigh CS, Daan S: A functional analysisof circadian pacemakers in nocturnal rodents:V. Pacemaker structure: A clock for all seasons. JComp Physiol A 1976;106:333-355.

15. Braun AR, Balkin TJ, Wesenten NJ, et al: Regionalcerebral blood flow throughout the sleep-wakecycle: An H2(15)O PET study. Brain 1997;120:1173-1197.

16. Krauchi K, Wirz-Justice A: Circadian clues tosleep onset mechanisms. Neuropsychopharma-cology 2001;25:S92-96.

17. Zhang L: New electrophysiological approaches tothe suprachiasmatic circadian pacemaker. BolEstud Med Biol 1994;42:31-36.

18. Shochat T, Luboshitzky R, Lavie P: Nocturnalmelatonin onset is phase locked to the primarysleep gate. Am J Physiol 1997;273:R364-370.

19. Czeisler CA, Wright KP: Influence of light on circa-dian rhythmicity in humans. In Turek FW, Zee PC(eds): Regulation of Sleep and Circadian Rhythms.New York, Marcel Dekker, 1999, pp 149-180.

20. Jewett ME, Kronauer RE, Czeisler CA: Light-induced suppression of endogenous circadianamplitude in humans. Nature 1991;350:59-62.

21. Folkard S, Barton J: Does the “forbidden zone” forsleep onset influence morning shift sleep dura-tion? Ergonomics 1993;36:85-91.

22. Lavie P: Ultrashort sleep-waking schedule: III.“Gates” and “forbidden zones” for sleep. Electro-encephalogr Clin Neurophysiol 1986;63:414-425.

23. Porkka-Heiskanen T, Stenberg D: Cellular andmolecular mechanisms of sleep. In Turek FW, ZeePC (eds): Regulation of Sleep and CircadianRhythms. New York, Marcel Dekker, 1999,pp 287-307.

24. Wood C, Magnello ME: Diurnal changes in per-ceptions of energy and mood: J R Soc Med 1992;85:191-194.

25. Boivin DB, Czeisler CA, Dijk D, et al: Complexinteractions of the sleep-wake cycle and circadianphase modulates mood in healthy subjects. ArchGen Psychiatry 1997;54:145-152.

26. Yamazaki S, Numano R, Abe M, et al: Resettingcentral and peripheral circadian oscillators intransgenic rats. Science 2000;288:682-685.

27. Minors DS, Waterhouse JM, Wirz-Justice A: Ahuman phase-response curve to light. NeurosciLett 1991;133:36-40.

Ch08.qxd 01/28/05 8:25 PM Page 97


28. Duffy JF, Kronauer RE, Czeisler CA: Phase-shift-ing human circadian rhythms: Influence of sleeptiming, social contact and light exposure. JPhysiol 1996;495:289-297.

29. Mitchell PJ, Hoese EK, Liu L, et al: Conflictingbright light exposure during night shifts impedescircadian adaptation. J Biol Rhythms 1997;12:5-15.

30. Boivin DB, Czeisler CA: Resetting of circadianmelatonin and cortisol rhythms in humans byordinary room light. Neuroreport 1998;9:779-782.

31. Waterhouse J, Minors D, Folkard S, et al: Light ofdomestic intensity produces phase shifts of thecircadian oscillator in humans. Neurosci Lett1998;245:97-100.

32. Buxton OM, L’Hermite-Baleriaux M, Turek FW,van Cauter E: Daytime naps in darkness phaseshift the human circadian rhythms of melatoninand thyrotropin secretion. Am J Physiol RegulIntegr Comp Physiol 2000;278:R373-382.

33. Minors DS, Waterhouse JM: Deriving a “phaseresponse curve” from adjustment to simulated timezone transitions. J Biol Rhythms 1994;9:275-282.

34. Geoffriau M, Brun J, Chazot G, Claustrat B: Thephysiology and pharmacology of melatonin inhumans. Horm Res 1998;49:136-141.

35. Reiter RJ: Pineal melatonin: Cell biology of itssynthesis and of its physiological interactions.Endocr Rev 1991;12:151-179.

36. Lewy AJ, Ahmed S, Sack RL: Phase shifting thehuman circadian clock using melatonin. BehavBrain Res 1996;73:131-134.

37. Lewy AJ, Ahmed S, Jackson JM, Sack RL: Melatoninshifts human circadian rhythms according to aphase-response curve. Chronobiol Int 1992;9:380-392.

38. Turek WF, Czeisler CA: Role of melatonin in theregulation of sleep. In Turek FW, Zee PC (eds):Regulation of Sleep and Circadian Rhythms. NewYork, Marcel Dekker, 1999, pp 181-195.

39. Zaidan R, Geoffriau M, Brun J, et al: Melatoninis able to influence its secretion in humans:Description of a phase-response curve. Neuro-endocrinology 1994;60:105-112.

40. Lavie P: Melatonin: Role in gating nocturnal rise insleep propensity. J Biol Rhythms 1997;12:657-665.

41. Dijk D-J, Edgar DM: Circadian and homeostaticcontrol of wakefulness and sleep, in the regula-tion of sleep. In Turek FW, Zee PC (eds): Regu-lation of Sleep and Circadian Rhythms. New York,Marcel Dekker, 1999, pp 111-147.

42. Lockley SW, Skene DJ, James K, et al: Melatoninadministration can entrain the free-running circa-dian system of blind subjects. J Endocrinol 2000;164:1-6.

43. Arendt J: Complex effects of melatonin. Therapie1998;53:479-488.

44. Klerman EB, Rimmer DW, Dijk DJ, et al: Nonphoticentrainment of the human circadian pacemaker.Am J Physiol 1998;274:991-996.

45. Baehr EK, Fogg LF, Eastman CI: Intermittentbright light and exercise to entrain human circa-dian rhythms to night work. Am J Physiol 1999;277:1598-1604.

46. Park SJ, Tokura H: Effects of different light inten-sities during the daytime on circadian rhythm ofcore temperature in humans. Appl Human Sci1998;17:253-257.

47. Danilenko KV, Wirz-Justice A, Krauchi K, et al: Thehuman circadian pacemaker can see by the dawn’searly light. J Biol Rhythms 2000;15:437-446.

48. Middleton B, Arendt J, Stone BM: Human circadianrhythms in constant dim light (8 lux) with knowl-edge of clock time. J Sleep Res 1996;5:69-76.

49. Brainard GC, Rollag MD, Hanifin JP: Photic regula-tion of melatonin in humans: Ocular and neural sig-nal transduction. J Biol Rhythms 1997;12:537-546.

50. Jewett ME, Rimmer DW, Duffy JF, et al: Humancircadian pacemaker is sensitive to light through-out subjective day without evidence of transients.Am J Physiol 1997;273:R1800-1809.

51. Takahashi JS, Zatz M: Regulation of circadianrhythmicity. Science 1982;4565:1104-1111.

52. Jewett ME, Kronauer RE, Czeisler CA: Phase-amplitude resetting of the human circadian pace-maker via bright light: A further analysis. J BiolRhythms 1994;9:295-314.

53. Boivin DB, Duffy JF, Kronauer RE, et al. Dose-response relationships for resetting of humancircadian clock by light. Nature 1996;379:540-542.

54. Mitchell PJ, Hoese EK, Liu L, et al: Conflictingbright light exposure during night shifts impedescircadian adaptation. J Biol Rhythms 1997;12:5-15.

55. Lewy AJ, Sack RL: Exogenous melatonin’s phase-shifting effects on the endogenous melatoninprofile in sighted humans: A brief review andcritique of the literature. J Biol Rhythms 1997;12:588-594.

56. Weaver DR, Reppert SM: Direct in utero percep-tion of light by the mammalian fetus. Brain ResDev Brain Res 1989;47:151-155.

57. Altman J, Bayer SA: Development of the dien-cephalon in the rat: III. Ontogeny of the special-ized ventricular linings of the hypothalamic thirdventricle. J Comp Neurol 1978;182:995-1015.

58. Altman J, Bayer SA: Time of origin of neurons ofthe rat superior colliculus in relation to othercomponents of the visual and visuomotor path-ways. Exp Brain Res 1981;42:424-434.

59. Viswanathan N, Davis FC: Single prenatal injec-tions of melatonin or the D1-dopamine receptoragonist SKF 38393 to pregnant hamsters sets the

Ch08.qxd 01/28/05 8:25 PM Page 98


offspring’s circadian rhythms to phases 180 degreesapart. J Comp Physiol [A] 1997;180:339-346.

60. Viswanathan N, Weaver DR, Reppert SM, Davis FC:Entrainment of the fetal hamster circadian pace-maker by prenatal injections of the dopamine ago-nist SKF 38393. J Neurosci 1994;14:5393-5398.

61. Weaver DR, Reppert SM: Definition of the devel-opmental transition from dopaminergic to photicregulation of c-fos gene expression in the ratsuprachiasmatic nucleus. Brain Res Mol Brain Res1995;33:136-148.

62. Seron-Ferre M, Torres-Farfan C, Forcelledo ML,Valenzuela GJ: The development of circadianrhythms in the fetus and neonate. Semin Perinatol2001;25:363-370.

63. Mosko S, Moore RY: Retinohypothalamic tractdevelopment: Alteration of suprachiasmatic lesionsin the neonatal rat. Brain Res 1979;174:1-15.

64. Mosko S, Moore RY: Neonatal suprachiasmaticnucleus ablation: Absence of functional and mor-phological plasticity. Proc Natl Acad Sci 1978;75:6243-6246.

65. Roberts MH, Bernstein MF, Moore RY: Dif-ferentiation of the suprachiasmatic nucleus infetal rat anterior hypothalamic transplants. DevBrain Res 1987;32:59-66.

66. Wiegand SJ, Gash, DM: Organization and efferentconnections of transplanted suprachiasmaticnuclei. J Comp Neurol 1988;267:562-579.

67. Reppert SM: Interaction between the circadianclocks of mother and fetus. CIBA Found Symp1995;183:198-207.

68. Viswanathan N: Maternal entrainment in the cir-cadian activity rhythm of laboratory mouse(C57BL/6J). Physiol Behav 1999;68:157-162.

69. Rivkees SA, Hofman PL, Fortman J: Newborn pri-mate infants are entrained by low intensity light-ing. Proc Natl Acad Sci 1997;94:292-297.

70. Stark RI, Daniel SS: Circadian rhythm of vaso-pressin levels in cerebrospinal fluid of the fetus:Effect of continuous light. Endocrinology1989;124:3095-3101.

71. Patrick, J, Campbell K, Carmichael L, et al:Patterns of gross fetal body movement over 24-hour observation intervals during the last 10weeks of pregnancy. Am J Obstet Gynecol1982;142:363-371.

72. McGraw K, Hoffmann R, Harker C, Herman JH:The development of circadian rhythms in ahuman infant. Sleep 1999;22:303-310.

73. Nishihara K, Horiuchi S, Eto H, Uchida S: Thedevelopment of infants’ circadian rest-activityrhythm and mothers’ rhythm. Physiol Behav2002;77:91-98.

74. Rechtschaffen A, Kales A (eds): A Manual for SleepStages of Human Subjects: A Manual ofStandardized Terminology: Techniques andScoring System, Los Angeles, UCLA BrainInformation Service/Brain Research Institute, 1968.

75. Israel DN, Ancoli-Israel S: Sleep and rhythms intenth vs third graders. Sleep 2000;23(Suppl 2):A197.

76. Yang CM, Spielman AJ, Martinez E, et al: Theeffects of delayed weekend sleep schedule on sub-jective sleepiness and cognitive functioning. Sleep1998;21(Suppl):202.

77. Dijk DJ, Duffy JF: Circadian regulation of humansleep and age-related changes in its timing, con-solidation and EEG characteristics. Ann Med1999;31:130-140.

Ch08.qxd 01/28/05 8:25 PM Page 99

Circadian rhythm disorders are characterized bynormal sleep quantity and quality at the wrongtime, as dictated by societal or familial demands.The circadian pacemaker may be delayed or ad-vanced with respect to the desired hour ofsleep, or clock time and circadian time may beout of phase. This may result at least partly froma genetic propensity when a parent suffersfrom the same symptoms or propensity. Thephase delay or advance is typically aggravatedby the child being exposed to light at the wrongtime and not exposed to light at the right time.The wrong time to be exposed to light is shortlybefore or during the desired hours of sleep. Theright time to be exposed to light is duringthe desired hours of waking.

Some ConsequencesCircadian rhythm disorders are frequently asso-ciated with daytime sleepiness. The strugglesthat occur when a child is unable to fall asleepuntil an inappropriately late hour may have fea-tures in common with sleep onset associationdisorder or sleep phobia. Children who are sleepdeprived secondary to a circadian rhythm disorderfrequently have symptoms similar to those ofattention deficit hyperactivity disorder (ADHD).Circadian rhythm disorders commonly lead tosleep deprivation during the night and sleepinessduring the day at school.

Biologic Clocks and Circadian RhythmsAll individuals possess inherent circadianrhythms of greater or less than 24 hours, withthe vast majority longer than 24 hours (Fig. 9–1).Circadian rhythms are entrained to be in phasewith environmental time by “zeitgebers,” or timecues, mainly light but also including activity

schedules,1 exercise, and other environmentalfactors.2 Circadian rhythms develop duringinfancy and are robust by 1 year of age.2

Individuals with circadian rhythm disordershave failed to properly entrain to environmentalzeitgebers and are either phase delayed (delayedsleep phase syndrome [DSPS]) or phase advanced(advanced sleep phase syndrome [ASPS])(Fig. 9–2). Some individuals “free-run” withrespect to environmental time, and their hour ofsleep moves progressively later or earlier, goingin and out of synchrony with the expected hourof sleep (non–24-hour sleep–wake disorder). It issuspected that individuals with longer geneticcircadian rhythms might require stronger zeitge-bers to remain in phase with environmental time,and this might explain why both heredity andenvironment are components of circadianrhythm disorders. Also, motivation appears to bea factor, and certain individuals with psychiatricor behavioral symptomatology appear to be morelikely to become phase delayed or advanced. It isdifficult to parse the genetic, environmental,behavioral, and psychiatric contributions to achild’s circadian rhythm disorder.

Forbidden Sleep ZoneThe “forbidden sleep zone” immediately pre-cedes the onset of biologic night. This is thetime at which sleep is least likely to occurduring a circadian cycle.3 This 1- to 3-hourperiod is well established in human adults andadolescents. Since the forbidden sleep zone pre-cedes the normal hour of sleep, many ado-lescents with delayed sleep phase seem unableto advance their hour of sleep despite sleepdeprivation and daytime sleepiness. Althoughforbidden sleep zones have not been studied inchildren, they appear to be particularly robust,as many children who fall asleep in school

Circadian Rhythm Disorders:Diagnosis and Treatment

John H. Herman

9

101

Ch09.qxd 01/28/05 10:29 AM

are not able to begin night-time sleep at an ear-lier hour.

PrevalenceCircadian rhythm disorders appear in at least10% of children. DSPS alone affects 5 to 10% ofadolescents but may appear in children as youngas those who are just starting school.4 ASPS isseen less frequently but also appears at the age ofstarting school. Non–24-hour circadian rhythmis relatively rare and appears in adolescents.Some evidence suggests that more males thanfemales display DSPS.

Differential DiagnosisA careful history is essential to distinguish circa-dian rhythm disorders from other sleep disorders.The hallmark of circadian rhythm disorders isthat when the child is permitted to sleep on his orher desired schedule, sleep is normal and daytimesleepiness rapidly subsides. Circadian rhythm dis-orders should be distinguished from primaryinsomnia, settling disorders, generalized anxietydisorders, mood disorders, attention deficit disor-

der (ADD)/ADHD, restless legs/periodic limbmovements, and obstructive sleep apnea syn-drome (Table 9–1). School avoidance/refusal maybe confused with DSPS.

DELAYED SLEEP PHASE SYNDROME

DSPS is a disorder in which the major sleepepisode is delayed significantly and persistently,typically by 1 or more hours beyond the desiredbedtime, in children or adolescents. Bedtime and

102 Chapter 9 Circadian Rhythm Disorders: Diagnosis and Treatment

8

6

4

2

0

Num

ber

of s

ubje

cts

Intrinsic circadian period (h)

Older

Young

23.8 23.9 24.0 24.1 24.2 24.3 24.4 24.5 24.6

38.0

37.5

37.0

36.5

36.0

Tem

pera

ture

(�C

)

0 12 24

Time of day (h)

Time of day (h)

Nadir7:35 AM

36 48

38.0

37.5

37.0

36.5

36.0

Tem

pera

ture

(�C

)

0 12 24

DSPS

ASPS

Nadir1:20 AM

36 48

Figure 9–1. Duration of circadian rhythm period in 25adults who participated in a constant routine protocoldesigned to reveal the intrinsic period length. Half wereyounger and half older adults. Note that over 90% of thesubjects had period lengths greater than 24 hours. (FromDijk DJ, Duffy JF, Czeisler CA: Contribution of circadianphysiology and sleep homeostasis to age-related changesin human sleep. Chronobiol Int 2000;17:285-311.)

Figure 9–2. Illustrations of the temperature curves ofindividuals with delayed sleep phase syndrome (DSPS)(temperature nadir for DSPS at 7:35 AM and for advancedsleep phase syndrome [ASPS] at 1:20 AM). The tempera-ture minimum is approximately 5-6 hours after sleeponset in an 8-hour sleeper. From these temperature min-imums, a sleep period from 2:00 to 10:00 AM can be pre-dicted for the DSPS illustration and from 8:00 PM to 4:00AM for the ASPS illustration. The timing of the tempera-ture nadir must be estimated when timing the adminis-tration of bright light in treating DSPS or ASPS.

Ch09.qxd 01/28/05 10:29 AM

Circadian Rhythm Disorders: Diagnosis and Treatment Chapter 9 103

Syn

dro

me

Prim

ary

Sett

ling

M

oo

d

AD

D/

Sym

pto

mD

SPS

Inso

mn

iaD

iso

rder

An

xiet

yD

iso

rder

AD

HD

RLS

/PLM

DO

SAS

Bedt

ime

Usu

alU

nusu

al b

ut

Usu

al u

nles

s U

sual

, be

dtim

e U

sual

Usu

alRa

re,

child

Ra

re,

child

st

rugg

les

may

be

seen

par

ent

may

be

typ

ical

ly

typ

ical

ly

pre

sent

pho

bia

slee

py

slee

py

Diff

icul

ty

Alw

ays,

U

sual

or

Usu

al u

nles

s U

sual

Usu

alRa

re,

slee

p

Usu

alRa

rein

itiat

ing

esse

ntia

l for

di

srup

ted

par

ent

is n

orm

alsl

eep

diag

nosi

ssl

eep

pre

sent

Dis

rup

ted

slee

p

Nev

erU

sual

Usu

alU

sual

Usu

alRa

re,

slee

p

Usu

alU

sual

due

to

at r

ando

m

is n

orm

alre

spira

tory

tim

esar

ousa

lsD

isru

pte

d sl

eep

N

ever

Rare

Usu

alN

ever

Nev

erN

ever

Nev

er,

Nev

er

at f

ixed

tim

esaw

aken

ings

in

firs

t ha

lf of

nig

htSl

eep

ing

late

A

lway

sRa

reRa

re,

Rare

, ov

erac

tive

Usu

alU

sual

Usu

alU

sual

, sl

eep

on

wee

kend

sfr

eque

nt

wak

ing

rebo

und

nap

ssy

stem

Awak

enin

g U

sual

Rare

, m

ay f

ollo

w

Rare

Rare

, ty

pic

ally

U

sual

, w

ish

Usu

alRa

reU

sual

, sl

eep

st

rugg

les

poo

r sl

eep

com

plia

nt

to is

olat

eno

n-re

fres

hing

child

Exce

ssiv

e da

ytim

e U

sual

Rare

Rare

Rare

Usu

alRa

reU

sual

Usu

alsl

eep

ines

sC

omp

lain

ts o

f U

sual

Usu

alRa

reRa

reU

sual

Rare

Usu

alU

sual

fatig

ueA

tten

tiona

l/U

sual

due

to

Usu

al d

ue t

o U

sual

due

to

Rare

Usu

al d

ue

Alw

ays

Usu

al d

ue t

o U

sual

due

to

hyp

erac

tivity

sl

eep

loss

poo

r sl

eep

p

oor

slee

p

to m

ood

poo

r sl

eep

p

oor

slee

p

sym

pto

ms

qua

lity

qua

lity

diso

rder

qua

lity

qua

lity

AD

D/A

DH

D a

tten

tion

defic

it di

sord

er/a

tten

tion

defic

it hy

per

activ

ity d

isor

der;

OSA

S, o

bstr

uctiv

e sl

eep

ap

nea

synd

rom

e; R

LS/P

LMD

, re

stle

ss l

egs

synd

rom

e/p

erio

dic

limb

mov

emen

t di

sord

er.

Tab

le 9

–1.

Dif

fere

nti

al D

iag

no

sis

of

Del

ayed

Sle

ep P

has

e Sy

nd

rom

e (D

SPS)

Ch09.qxd 01/28/05 10:29 AM

awakening become issues for them. Childrenappear to be less tolerant of sleep deprivationthan adults, and a 1-hour delay in sleep onsetmay result in pathologic symptoms.

Presenting ComplaintsThe presenting complaints are listed below.

● Bedtime struggles or difficulty awakening atthe desired time

● Complaints of insomnia at bedtime and/orexcessive sleepiness in the morning

● Inability to wake up at the desired time● Falling asleep at school or being too sleepy in

the morning to participate in normal activities● Symptoms consistent with behavioral dysregu-

lation, ADHD, or depression● Preferring to not eat breakfast and being hun-

gry at or close to bedtime

Diagnostic CriteriaThe criteria below are similar to those describedin the International Classification of SleepDisorders, Revised: Diagnostic and Coding Manualand are required for diagnosis.

● The sleep pattern is significantly delayed, typ-ically by more than 1 hour.

● There is an inability to fall asleep at the desiredclock time and an inability to awaken sponta-neously at the desired time of awakening.

● This sleep pattern has been present for at leastone month.

● Normal quality and quantity of sleep areobserved when the child can sleep on his orher desired schedule.

● Children with DSPS sleep later on weekendsand holidays.

● They report less daytime sleepiness on week-ends, when they awaken spontaneously at alater hour.

● No other sleep or psychiatric disorder is pres-ent that could explain the patient’s symptoms.

● The delayed phase is not the result of socialpreference or an overloaded school, social activ-ity, or work schedule.

Family DynamicsIn DSPS, both the child and the parents are usu-ally perplexed that he or she cannot fall asleepmore quickly. Frequently, family disputes eruptaround bedtime and awakening time. Control bat-

tles ensue. Parents’ efforts to advance the timingof their child’s sleep onset by earlier bedtime,extra alarm clocks, relaxation techniques, andsleeping pills are not successful. The hour of sleepand waking become the center of family conflicts.When other behavioral or psychiatric syndromesare present in the child or family members, thechild’s sleep habits become flashpoints for dis-putes. Sometimes the family dynamics must beaddressed before DSPS in a child can be treated. Ifthe other family members are sleeping to a latehour on non-workdays, the child may have noresources for dealing with DSPS.

Interpersonal Dynamics

Patients with DSPS, especially adolescents, typi-cally are “night owls” or “night people” and saythey feel and function best and are more alert dur-ing the late evening and night hours. Children withDSPS nap more frequently during school days,complain of daytime sleepiness, have more atten-tion problems, poorer school achievement, andmore injuries and are more emotionally upset thanthe other children.5 DSPS may become a “lifestyle”in some children whose identities becomeenmeshed with late-night activities, such as phonecalls, Internet interaction, and chat rooms. Late-night children often become close friends.Children who awaken and retire at an earlier hourare viewed with condescension by cliques of nightowls. In some instances DSPS becomes ingrained,and the child may pursue a career that involves orpermits late-night activity, such as writing, soft-ware, or the service or entertainment sector. Theearlier in life DSPS is treated, the less resistance istypically encountered.

Coexistence withPsychiatric/Behavioral SymptomsADHD, oppositional symptoms, conduct disorder,aggressive symptoms, and symptoms of depressionappear frequently in many but not all children withDSPS (see Table 9–1).6 In some instances DSPSleads to chronic sleep deprivation, which mayaggravate underlying psychopathologic tendenciesin a child.7 When psychiatric symptoms and DSPSare present conjointly, it is important to establishwhether the psychiatric symptoms are presentindependent of the DSPS or only occur with it.8

For example, on weekends or holidays, when thechild sleeps until spontaneous awakening, do psy-


Ch09.qxd 01/28/05 10:29 AM

chiatric symptoms lessen? If the psychiatric symp-toms and DSPS coexist temporarily, it is best totreat the DSPS before addressing the psychiatricsymptoms. If the psychiatric symptoms are presentregardless of prior sleep, then treating both the psy-chiatric symptoms and the DSPS concurrentlymight be warranted.

Age of OnsetDSPS may be encountered in young children.Frequently, a child’s late hours of sleep becomeproblematic only with the start of school. ManyDSPS patients report that their difficulties beganafter a summer vacation. After a summer ofsleeping late, these children find it impossible tosleep on a normal schedule when schoolresumes. Adolescence appears to be the mostcommon period of life for the onset of DSPS.

EvaluationThe following factors are considered in the eval-uation of children for DSPS.

● Usually have one parent with DSPS symptomsor night owl preference

● Typically accomplish tasks such as homeworkbest at a later hour

● Feel best at or near bedtime● Experience academic, emotional, and behav-

ioral problems during morning hours● Must be evaluated for substance abuse● Should be evaluated for depression, ADHD,

school refusal, conduct disorder, oppositionaldefiant disorder, and anxiety disorders

● In some cases it is necessary for the family tomaintain a diary of bedtime, time of sleep onset,and wake-up time for a minimum of two weeksand as long as 1 month to establish a pattern.

● May describe their sleep habits significantly dif-ferently from the way their parents portray them

● Tend to cover windows to block out sunlight● Watch TV or computer monitors near or after

the desired bedtime

Singular Significance of Sunday NightSunday night bedtime is frequently the majorflashpoint for school aged children. Many chil-dren with DSPS awaken and retire at a consider-ably later hour on weekends, effectively“moving” to a more westward time zone. Sunday

night the child is expected to return to a week-day schedule (or move east) and finds it impos-sible. Late-night sleep onset is followed by abattle when the child must awaken for schoolMonday morning, and the week is off to a badstart, frequently with behavioral, emotional, andfamilial conflicts followed by disruptive behav-ior, hyperactivity, distractibility, failure to attend,and learning problems in school.

In adolescents, failure to cooperate with a planto reschedule their sleep may be a sign of clini-cal depression or oppositional defiant symp-toms. Adolescence is a particularly vulnerablelife stage for the development of this syndrome:For some children, adolescence, with increasedautonomy and noncompliance, becomes syn-onymous with DSPS.

TREATMENT OPTIONS FORDELAYED SLEEP PHASE SYNDROME

The various options for the treatment of DSPSare discussed below. They are also summarizedin Table 9–2 along with a summary of additionaltherapeutic interventions useful in the treat-ment of other causes of sleep onset insomnia.

Establishing the Patient’sTemperature NadirThe nadir of core body temperature, or the min-imum temperature in the circadian cycle, typi-cally occurs about two thirds of the way throughthe sleep cycle. A child who begins sleep at 9:00PM and spontaneously awakens at 7:00 AM onweekends typically has a temperature minimumat approximately 3:00 to 4:00 AM. It is crucial toestablish the timing of the temperature mini-mum, also known as the point of singularity, aslight administered before the minimum causesa phase delay, or results in sleep occurring later,and light administered after the minimumcauses a phase advance, or results in sleep occur-ring earlier (see Fig. 9–2). The timing of thetemperature nadir may best be established by acareful history of when the child or adolescentprefers to begin sleep and to awaken, which mayrequire keeping a diary. Sometimes the child’spreferred hour of sleep onset is as late as 4:00AM and their preferred hour of awakening isremarkably late, such as 2:00 PM. Such childrenmight have a core body temperature minimum


Ch09.qxd 01/28/05 10:29 AM

at as late as 11:00 AM. Therefore, exposure tolight before 10:00 AM will cause a further phasedelay and not the expected phase advance.

Phase Shifting: A Family System ApproachMoving a child’s propensity to sleep is calledphase advance when shifting to an earlier hourand phase delay when shifting to a later hour. Thechild and the family first must understandthe basic principles behind the child’s inability tosleep at a desired time. This may be more difficultthan expected when one or more family membersbelieve the problem to be essentially behavioral.

Nevertheless, behavioral issues frequently over-lay circadian rhythm disorders. The entire familyand the patient must be motivated to undergo thephase shift. Resistance may arise from a variety ofareas, as one or more family members may beinvested in the child’s propensity to sleep late,such as a parent who sleeps until noon on week-ends. Expectations must be explored. The familyand child must comprehend that this may be a life-long issue that will affect the child long after he orshe leaves the home of origin. This may be the firsttime that someone has said to the child, “After youare grown up and no longer live with your parents,this may continue to be an issue that you will haveto cope with on your own.”

The child must be encouraged to “own” theproblem, or take responsibility for it. The parentsmust be convinced that the root of the problemis not intentional defiance. Children should beencouraged to set their own alarms and wake upon their own. Parents must recognize that the lessinvolved they are with their child’s awakening, themore the child may assume responsibility.

Sleep HygieneThe treatment of DSPS must begin with ensuringgood sleep hygiene if manipulations with lightand melatonin are to be effective. Instructions forgood sleep hygiene include the following:

● Sleep only when sleepy. If a child is not sleepy,reading in bed with a dim light might be anoption for some families. This reduces the timethe child is struggling with initiating sleep.

● If the child can’t fall asleep within 20 minutes,allow him or her to do something boring untilthe child feels sleepy. The child should not beexposed to bright light after bedtime.

● Minimize the child’s naps. This will optimizethe probability that the child will be tired atbedtime. If the child requires a nap, allow lessthan 1 hour of sleep, before 3:00 PM if possible.Lights should remain on during naps.

● The child must adhere to approximately thesame wake-up time and bedtime 7 days a week.Other children might be able to sleep late onweekends and still fall asleep earlier on Sundaynight, but the hallmark of children with DSPS isthat they can’t. Children with the same wake-uptime every morning will be more stable emo-tionally and less oppositional in the morningthan children who sleep late on weekends.

● A quiet time with minimal light and less activ-ity should precede bedtime. Television andcomputer monitors should be avoided. Maxi-mizing active playing and exercise is recom-mended to help you sleep well, but the timingof the workout is important. Exercising in themorning or early afternoon will not interferewith sleep.

● Every child deserves a consistent bedtime. Asmuch as a child might protest, a set bedtime,similar to a home-cooked meal, is appreciatedby the child as a basic component of goodparenting.

● Develop sleep rituals for your child. Children’ssleep is enhanced by rituals, such as readingstories, quiet games, a drink, or just chatting.Having a set time for bathing and getting intopajamas helps a child prepare for sleep. It isimportant to give your child cues that it is timeto slow down and prepare for sleep.

● Allow your child to use his or her bed only forsleeping. Refrain from permitting your childto use the bed to watch TV, do homework, oruse the computer. The child then knows thatwhen it is time to go to bed, it is time to sleep.

● Avoid all foods or beverages with caffeine afterlunch. This includes chocolate and many bev-erages with caffeine as an additive. Coffee, tea,cola, cocoa, chocolate, and some prescriptionand nonprescription drugs contain caffeine.

● Some children benefit from a light snackbefore bed. For these children, if their stom-achs are too empty, it can interfere with sleep.However, if your child eats a heavy meal beforebedtime, this can interfere as well. Dairy prod-ucts and turkey contain tryptophan, whichacts as a natural sleep inducer. A warm glass ofmilk is sometimes recommended probablybecause of its tryptophan content.


Ch09.qxd 01/28/05 10:29 AM

● Give your child a hot bath 90 minutes beforebedtime. A hot bath will raise your bodytemperature, but it is the drop in body tem-perature that may leave you feeling sleepy.

● Make sure your child’s bed and bedroom arequiet and comfortable. A hot room can be un-comfortable. A cooler room along with enoughblankets to stay warm is recommended. Toomany blankets can overheat a child even in acold room.

● Use sunlight to set your child’s biologic clock. Assoon as your child gets up in the morning, havehim or her play in sunlight, preferably outside.

Bright-Light Therapy

Bright-light therapy should include control oflight and dark exposures across the whole day.The following list of therapeutic componentsillustrates this point.

● The patient with DSPS should use bright-lightexposure early in the morning and avoid lightafter the hour of sunset as much as possible.9

The light of dawn is extremely important inentraining circadian rhythms10 in children aswell as adults.11 Such control of light exposureshould produce a phase advance. One to 2hours outdoors or indoors in early morningdirect sunlight or (if rising before sunrise inthe early morning) in front of a light box thatemits 2500 lux will usually produce a phaseadvance in a few days.

● Parents and patient are encouraged to removecoverings from windows and to place thechild’s bed where the most direct sunlight willbe present.

● Children are encouraged not to cover theirheads with blankets and to face the light whensleeping.

● If possible, parents are encouraged to movetheir child to a bedroom facing east or south,with morning-light exposure.

● Avoid light after sunset as much as possible, aseven ordinary indoor light may affect a child’scircadian rhythm.12

● Bright light is more effective when it is expe-rienced in conjunction with active play orexercise.13

Bright-Light BoxIf a family is considering purchasing a bright-light box, encourage them to go to one of several

web sites that show illustrations of products.Frequently, a portable unit that can be placed nextto the child as he or she goes through morningroutines, does homework, reads, or is next to avideo monitor is preferable. Instructions for eachunit must be followed to allow the correct amountof light exposure, as light intensity decreases as afunction of the square of the distance. The childshould not look directly at the light. If he or shehas an adverse reaction to the light box, increasethe distance. Some children become irritable orexhibit other symptoms of hypomania whenexposed to bright morning light. In such cases,light exposure should be stopped or the durationdecreased until hypomanic symptoms subside.

MelatoninMelatonin has been shown to be effective as a hyp-notic in children,14,15 but there is more evidencethat it is effective in altering sleep phase.16 Variousstudies show it to be effective in the treatment ofDSPS.17 It is a hormone secreted by the pinealgland associated with body temperature dropand lowered metabolism. Melatonin administered1 hour before the desired time of sleep onset willadvance the timing of sleep if given in conjunctionwith the blocking of light exposure.18 Many ado-lescents may be insufficiently motivated to exposethemselves to early morning sunlight or a bright-light box but may be willing to take melatonin.Ideally, melatonin before the hour of sleep isadministered in conjunction with early morningbright light. A physiologic level of melatonin isachieved with a dose of less than 1 mg. In children,it is best to begin with 1 mg and proceed to 3 mg ifthis is ineffective. Melatonin may be administeredfor a few weeks during phase shifting and reintro-duced if phase delay symptoms recur.

Advancing BedtimeAlong with the administration of morning lightand/or evening melatonin, bedtime is typicallyadvanced 15 minutes per day, as is the time ofmelatonin administration. If the child has diffi-culty falling asleep on a given night, bedtimeshould not be further advanced until sleep onsetis less than 30 minutes after bedtime.

ChronotherapyChronotherapy is a behavioral technique thatsystematically delays bedtime, following the


Ch09.qxd 01/28/05 10:29 AM


Syn

dro

me

Prim

ary

Sett

ling

M

oo

d

Trea

tmen

tD

SPS

Inso

mn

iaD

iso

rder

An

xiet

yD

iso

rder

AD

D/A

DH

DR

LS/P

LMD

OSA

SBe

havi

oral

A

lway

s M

ay b

e A

lway

s H

elp

ful

Hel

pfu

lH

elp

ful

Rare

ly h

elp

ful

Nev

er

inte

rven

tions

emp

loye

dhe

lpfu

lem

plo

yed

Rela

xatio

n M

ay b

e he

lpfu

l H

elp

ful

Rare

ly

Hel

pfu

lM

ay b

e M

ay b

e he

lpfu

l Ra

rely

Ra

rely

hel

pfu

lte

chni

que

sas

sec

onda

ry

help

ful

help

ful

as s

econ

dary

he

lpfu

ltr

eatm

ent

trea

tmen

tM

orni

ng b

right

-Fi

rst-

line

Rare

ly

Rare

ly

Rare

ly

May

be

If

DSP

S, m

ay

Rare

ly

Nev

erlig

ht t

reat

men

ttr

eatm

ent

help

ful

help

ful

help

ful

help

ful,

be h

elp

ful

help

ful

alon

g w

ithes

pec

ially

PM

dar

knes

sSA

DS

Mor

e st

ruct

ured

A

lway

sH

elp

ful

Alw

ays

May

be

M

ay b

e M

ay b

e he

lpfu

l, N

ever

, Ra

rely

hel

pfu

lbe

dtim

e an

d em

plo

yed

help

ful,

help

ful,

esp

ecia

lly

awak

enin

gs

awak

enin

g es

pec

ially

es

pec

ially

ro

utin

esty

pic

ally

tim

ero

utin

esro

utin

esoc

cur

in

first

hal

f of

ni

ght

Mel

aton

inH

elp

ful

May

be

Rare

ly

Rare

ly

Rare

ly

Rare

ly h

elp

ful

Rare

ly h

elp

ful

Rare

ly h

elp

ful,

help

ful

help

ful

help

ful

help

ful

alw

ays

trea

t O

SA f

irst

Hyp

notic

sN

ever

May

be

Nev

erRa

rely

Ra

rely

N

ever

May

be

Nev

erhe

lpfu

lhe

lpfu

lhe

lpfu

lhe

lpfu

lSS

RIs

Nev

erN

ever

Nev

erM

ay b

e H

elp

ful

Rare

ly h

elp

ful

Nev

erN

ever

help

ful

5-H

T-2A

N

ever

Rare

ly

May

be

May

be

Rare

ly h

elp

ful

May

be

Rare

ly h

elp

ful,

anta

goni

sts

help

ful

Nev

erhe

lpfu

lhe

lpfu

lhe

lpfu

lal

way

s tr

eat

OSA

firs

tSt

imul

ants

Nev

erN

ever

Nev

erN

ever

Nev

erH

elp

ful o

nly

if N

ever

Rare

ly h

elp

ful,

true

AD

HD

alw

ays

trea

t O

SA f

irst

Psyc

hoth

erap

yRa

rely

hel

pfu

lH

elp

ful

May

be

Hel

pfu

lH

elp

ful

May

be

help

ful

Nev

erN

ever

help

ful

AD

D/A

DH

D a

tten

tion

defic

it di

sord

er/a

tten

tion

defic

it hy

per

activ

ity d

isor

der;

DSP

S, d

elay

ed s

leep

pha

se s

yndr

ome;

5-H

T, s

erot

onin

; O

SAS,

obs

truc

tive

slee

p a

pne

a sy

n-dr

ome;

RLS

/PLM

S, r

estle

ss le

gs s

yndr

ome/

per

iodi

c lim

b m

ovem

ent

diso

rder

; SA

DS,

sea

sona

l affe

ctiv

e di

sord

er s

yndr

ome;

SSR

Is,

sele

ctiv

e se

roto

nin

reup

take

inhi

bito

rs.

Tab

le 9

–2.

Var

iou

s T

her

apeu

tic

Inte

rven

tio

ns

for

Syn

dro

mes

Pre

sen

tin

g w

ith

Sle

ep O

nse

t In

som

nia

Ch09.qxd 01/28/05 10:29 AM

natural tendency of human biology. Bedtime isdelayed by 3-hour increments each day, estab-lishing a 27-hour day. The procedure is main-tained until the desired bedtime is reached (e.g.,9:00 PM), at which point the child is properlyentrained. This approach is favored by adoles-cents who are extreme night owls and whotherefore are delighted at the concept of a laterbedtime and wake-up time. It is difficult toadminister, as a parent typically must be presentto oversee the temporarily unusual sleep andwake-up times, such as 12:00 noon to 8:00 PM.However, schedule normalization takes no morethan a week.

Successful Phase Advance TherapyThe first sign that the light and/or melatonin arehaving their desired effect is an earlier hour ofsleep onset and fewer struggles with morningawakening. This is often followed by a reportfrom the parents that the child’s morning appetitehas increased.

Maintenance PhaseOnce the desired sleep schedule is attained, thenext problem is its maintenance. When adoles-cents with DSPS stay up late studying, socializing,or engaging in other activities, it is probable thatthey will be drawn back to sleeping late. A strat-egy must be determined in advance for counter-acting this tendency. Frequently, the adolescentagrees that he or she may remain awake late attimes but must arise and be in sunlight relativelyearly. Parents need to be vigilant that weekends,school holidays, or summer vacations do notresult in a return of phase delay.

ADVANCED SLEEP PHASE SYNDROME

ASPS is a disorder in which the major sleepepisode is advanced in relation to the desiredclock time, which results in symptoms of com-pelling evening sleepiness, an early sleep onset,and an awakening that is earlier than that desired.

The following indications show the sympto-matology of ASPS.

● Inability to stay awake until the desired bed-time or inability to remain asleep until thedesired time of awakening. Children and ado-

lescents with ASPS fall asleep doing home-work, at social events, immediately after din-ner, or before dinner if it is at a late hour.

● Symptoms are present for at least 3 months.● When not required to remain awake until the

later bedtime, patients will have a habitualsleep period that is of normal quality andduration; with a sleep onset earlier than thatdesired, they will awaken spontaneouslyearlier than desired, and they will maintainstable entrainment to a 24-hour sleep–wakepattern.

The major complaint may concern the inabil-ity to stay awake in the evening, early morninginsomnia, or both.

Unlike other sleep maintenance disorders, theearly morning awakening occurs after undis-turbed sleep. Unlike other causes of excessivesleepiness, daytime school or activities early inthe day are not affected by sleepiness. This prob-lem becomes apparent in children and adoles-cents when social and academic activities stretchinto the evening hours. The evening activitiesare cut short by the need to retire much earlierthan the social norm.

Typical sleep onset times are as early as 5:00to 6:00 PM but no later than 8:00 PM, and wake-up times can be as early as 1:00 to 3:00 AM butno later than 5:00 AM. These sleep onset andwake-up times occur despite the family’s bestefforts to delay sleep to later hours.

There can be negative personal or social con-sequences to children and adolescents whentheir peers are engaged in evening activities andthe patient avoids them due to early sleep.Frequently, this interferes with family functionssuch as going out for dinner.

Attempts to delay sleep onset to a later timemay result in embarrassment due to fallingasleep during social gatherings.

If a child or adolescent with ASPS is continu-ally forced to stay up later for social or voca-tional reasons, the early awakening aspect of thesyndrome could lead to chronic sleep depriva-tion, daytime sleepiness, or napping.

EpidemiologyASPS is more likely to appear in children whohave one parent or grandparent with ASPS ten-dencies, which may not be sufficiently severe tobe syndromic. Similar to narcolepsy, the gene


Ch09.qxd 01/28/05 10:29 AM

and gene product responsible for ASPS havebeen identified.19

Differential DiagnosisASPS should be differentiated from other disor-ders producing early evening sleepiness, such asobstructive sleep apnea, disrupted nocturnalsleep, chronic sleep deprivation, and narcolepsy.Sometimes the early morning awakening is mis-takenly thought to be indicative of depressive ill-ness. Like DSPS patients, ASPS patients have nointrinsic sleep abnormalities; they report thatsleep itself is restful and restorative.

Treatment

ASPS is treated with chronotherapy or bright-light therapy. Chronotherapy would involve asystematic delay of bedtime until the desiredtime is achieved. Bright-light therapy wouldinvolve inducing a phase delay, with the lightexposure in the early evening. There are limiteddata about the effectiveness of bright-light ther-apy for ASPS.

NON–24-HOUR SLEEP–WAKE DISORDER

This condition can occur when the intrinsicperiod of the circadian pacemaker free-runs withrespect to the 24-hour day. Alternatively, chil-dren’s and adolescents’ self-selected exposure toartificial light may drive the circadian pacemakerto a schedule longer than 24 hours. Affectedchildren are not able to maintain a stable phaserelationship with the 24-hour day. Parents ini-tially report that their child’s sleep behavior iscompletely random and has no pattern. Only bymaintaining a sleep log for 2 to 6 weeks is it pos-sible to “smooth out” aberrations and estimatethe circadian periodicity. These children are thentypically revealed to have an incremental patternof successive delays in sleep onsets and waketimes (free-running), progressing in and out ofphase with local time. When the child’s endoge-nous rhythm is out of phase with the family andschool, both insomnia and excessive daytimesleepiness are present. Conversely, when the child’sendogenous rhythm is in phase with home andschool, symptoms remit. The intervals betweensymptomatic periods may last several weeks toseveral months. Blind individuals unable to per-

ceive any light are particularly susceptible to thisdisorder.

TreatmentMelatonin administration has been reported toimprove the timing of sleep in non–24-hoursleep–wake disorder, especially in the blind.Begin at 3 mg 1 hour before the desired hour ofsleep onset. Once a patient is entrained, mela-tonin may be lowered to 1 mg. The patient shouldbe in as much light and as active as possible dur-ing the intended wake period and in conditionsof dark, quiet, and bedrest during the intendedsleep period. In nonblind individuals, exposureto the light of dawn or a bright-light box atthe time of dawn may be essential to maintainentrainment.

CONCLUSIONS

Circadian rhythm disorders are fairly common inchildren and frequently masquerade as othersleep disorders, most notably insomnia and day-time sleepiness. Many children have been evalu-ated extensively from a medical perspective be-fore they are referred to a sleep specialist. For themost part, an in-depth interview includingthe child and caretakers is sufficient to estab-lish the diagnosis of circadian rhythm disorder.In many cases a 1-week to 1-month diary isessential to make the diagnosis, especially whenthe family describes the child’s sleep as “chaotic”and not following any pattern. In almost everycase, a sleep diary that is collected over a suffi-cient period of time will reveal an underlyingpattern of when sleep occurs.

Frequently, a second disorder coexists alongwith the circadian rhythm disorder, such asdepression, an anxiety disorder, obstructivesleep apnea, or restless legs syndrome/periodiclimb movement disorders. In some instancespolysomnography is required to establish theseverity of a coexisting disorder, and it should betreated before the circadian rhythm disorder canbe addressed. Treating obstructive sleep apneafirst, for example, will make it much easier for achild to comply with changing his or her sleepschedule. In some instances, for example in achild with a psychiatric disorder, correcting thecircadian rhythm disorder can alleviate symp-toms of depression or anxiety. To cite another


Ch09.qxd 01/28/05 10:29 AM

example, it is well known that early morningbright light has an antidepressant effect in someindividuals. Children with phase delay syn-drome and anxiety symptoms frequently becomevery anxious about their inability to initiate sleep.The correction of DSPS may exert an anxiolyticeffect in such children.

Discussions with children and their familiesabout circadian rhythm disorders are frequentlyprofound, as the sleep disorder specialist informsthe child about symptoms that may continuethroughout his or her life and offers suggestionsfor treatment strategies to address them. It isextremely rewarding when diagnosing and treat-ing a circadian rhythm disorder in a child withonly bright-light therapy and sometimes mela-tonin results in an enormous improvement inthe child’s quality of life.

References

1. Klerman EB, Rimmer DW, Dijk DJ, et al:Nonphotic entrainment of the human circadianpacemaker. Am J Physiol 1998;274:R991-R996.

2. Rivkees SA: Mechanisms and clinical significanceof circadian rhythms in children. Curr OpinPediatr 2001;13:352-357.

3. Lavie P: Ultrashort sleep-waking schedule: III.‘Gates’ and ‘forbidden zones’ for sleep. Electro-encephalogr Clin Neurophysiol 1986; 63: 414-425.

4. Mindell JA, Owens JA: A Clinical Guide toPediatric Sleep: Diagnosis and Management ofSleep Problems. Philadelphia, Lippincott, Williams& Williams, 2003.

5. Giannotti F, Cortesi F, Sebastiani T, Ottaviano S:Circadian preference, sleep and daytime behav-iour in adolescence. J Sleep Res 2002;11:191-199.

6. Cardinali DP: The human body circadian: Howthe biologic clock influences sleep and emotion.Neuroendocrinol Lett 2000;21:9-15.

7. Dahl RE, Lewin DS: Pathways to adolescenthealth sleep regulation and behavior. J AdolescHealth 2002;31(Suppl 6):175-184.

8. Kayumov L, Zhdanova IV, Shapiro CM: Melatonin,sleep, and circadian rhythm disorders. Semin ClinNeuropsychiatry 2000;5:44-55.

9. Lafrance C, Dumont M, Lesperance P, Lambert C:Daytime vigilance after morning bright light expo-sure in volunteers subjected to sleep restriction.Physiol Behav 1998;63:803-810.

10. Danilenko KV, Wirz-Justice A, Krauchi K, et al:The human circadian pacemaker can see by thedawn’s early light. J Biol Rhythms 2000;15:437-446.

11. Clodore M, Foret J, Benoit O, et al: Psycho-physiological effects of early morning bright lightexposure in young adults. Psychoneuroendo-crinology 1990;15:193-205.

12. Boivin DB, Czeisler CA: Resetting of circadianmelatonin and cortisol rhythms in humans byordinary room light. Neuroreport 1998;9:779-782.

13. Baehr EK, Fogg LF, Eastman CI: Intermittentbright light and exercise to entrain human circa-dian rhythms to night work. Am J Physiol1999;277:R1598-R1604.

14. Ivanenko A, Crabtree VM, Tauman R, Gozal D:Melatonin in children and adolescents withinsomnia: A retrospective study. Clin Pediatr(Phila) 2003;42:51-58.

15. Pires M, Benedito-Silva A, Pinto L, et al: Acuteeffects of low doses of melatonin on the sleepof young healthy subjects. J Pineal Res 2001;31:326-332.

16. Sack RL, Lewy AJ, Hughes RJ: Use of melatoninfor sleep and circadian rhythm disorders. AnnMed 1998;30:115-121.

17. Dagan Y, Yovel I, Hallis D, et al: Evaluating therole of melatonin in the long-term treatment ofdelayed sleep phase syndrome (DSPS).Chronobiol Int 1998;15:181-190.

18. Garcia J, Rosen G, Mahowald M: Circadianrhythms and circadian rhythm disorders in chil-dren and adolescents. Semin Pediatr Neurol2001;8:229-240.

19. Reid KJ, Chang AM, Dubocovich ML, et al:Familial advanced sleep phase syndrome. ArchNeurol 2001;58:1089-1094.


Ch09.qxd 01/28/05 10:29 AM

The hypothesis that colic might be a sleep disor-der caused by developmental or biologic factorswithin the child1 was based on the natural his-tory of colic and links between colic, sleepingproblems, and difficult temperament.2,3 Onlyretrospectively obtained data obtained from par-ents’ reports supported the hypothesis.4 It wasalso suggested that “postcolic” sleep problemsoccur after 3 or 4 months of age because someparents experienced difficulty in establishingage-appropriate sleep routines.1 Recent researchutilizing more objective measures has providedprospectively obtained data associating colicwith sleep problems.

NATURAL HISTORY OF COLIC

Wessel and coworkers5 diagnosed as colicky aninfant “who, otherwise healthy and well-fed, hadparoxysms of irritability, fussing or crying lastingfor a total of more than 3 hours a day and occur-ring on more than 3 days in any 1 week . . . and. . . the paroxysms continued to recur for morethan 3 weeks.” The criterion “more than 3weeks” was added because nannies left familiesafter about 3 weeks of crying. He thought thatnannies knew that if babies cried for more than3 weeks, then the crying would continue. Themothers, now alone caring for their babies atnight, came to his pediatric office after 3 weekscomplaining that their children were always cry-ing (M. A. Wessel, personal communication,2002). About 26% of infants in their study hadcolic. Illingworth6 defined colic as “violentrhythmical, screaming attacks which did notstop when the infants were picked up, and forwhich no cause, such as underfeeding, could befound.” Together, they studied about 150infants.

The age of onset of these behaviors is charac-teristic. Wessel and colleagues and Illingworth

found that the attacks were absent during thefirst few days but were present in 80% of affectedinfants by 2 weeks and about 100% by 3 weeks.The attacks begin in premature babies shortlyafter the expected due date, regardless of theirgestational age at birth. The time of day whenthese behaviors occur is another characteristic.During the first month, crying appears to occurrandomly, but later it occurs predominantly inthe evening hours. In 80% of infants, the attacksstart between 5 and 8 PM and end by 12 PM. For12% of infants, the attacks start between 7 and10 PM and end by 2 AM. In only 8% are theattacks randomly distributed throughout the dayand night. The age of termination of these spellsis also characteristic. They disappear by 2months of age in 50% of infants, by 2-3 monthsof age in 30%, and by 3-4 months of age in 10%.Using Wessel’s criteria, the infant’s behavioralstate was observed to be associated with colickybehavior.2 In 84% of colicky infants, the cryingspells begin when they are awake, in 8% theystart when asleep, and in 8% they occur undervariable conditions. When the crying spell ends,83% of the infants fall asleep.

As there are other definitions of colic, datafrom different studies may not be comparable.Some studies claim to use Wessel’s criteria butexclude “fussing” or exclude the criterion of“more than 3 weeks.” The problem with exclud-ing the 3-week criterion is that an extremelyhigh prevalence rate may result. For example, inWessel’s study, 49% of infants would have beendiagnosed with colic without that criterion.Reijneveld and coworkers7 showed that evensmall differences between definitions of colicbased on the duration of crying result in largechanges in prevalence rates, and they concludedthat different definitions of colic could not becompared because they mostly identified differ-ent infants. Their study excluded “fussing” from

Sleep and ColicMarc Weissbluth 10

113

Ch10.qxd 1/25/05 3:18 AM Page 113

the definition of colic, but they recommendedthat future studies should define colic using theduration of both crying and fussing, plus theamount of parental distress. This is an importantpoint, because it is now known that fussing, asopposed to crying, is a major part of colickybehavior,8 and parental distress over colic maybe a major factor in producing postcolic sleepproblems.9 Because of this concern of compara-bility, this chapter includes the definitions ofcolic utilized by different researchers.

Studies of colic are also difficult to comparebecause they have been performed at differentages. Finally, small sample sizes and modestgroup differences have led researchers to differ-ent interpretations of the clinical significance ofdata from a single study.

ETIOLOGY

A recent study showed that colicky infants,defined as crying or fussing for at least 3 hoursper day for a minimum of 3 days per week, hadhigher urinary levels of 5-hydroxy indoleaceticacid, a metabolite of serotonin.10 This supportedthe hypothesis that some temporal and behav-ioral features of colic might be caused by aserotonin–melatonin counterbalancing systeminvolving the gastrointestinal smooth muscles.11

Concentrations of serotonin are high in infantsduring the first month of life, and after 3 monthsthey decline. Immediately after delivery, concen-trations of serotonin show a circadian rhythmwith highest levels at night. Melatonin is derivedtransplacentally from the mother; concentra-tions are high in infants immediately after birth,they fall rapidly to extremely low levels withinseveral days, they increase slightly between1 and 3 months, and only after 3 months is therean abrupt increase in melatonin concentrations,with a circadian rhythm that demonstrateshigher concentrations at night.

Serotonin and melatonin have oppositeeffects on intestinal smooth muscle: serotonincauses contraction and melatonin causes relax-ation. Animal studies show that serotonin po-tency causing intestinal smooth musclecontraction is much greater in the neonatalperiod than at older ages. It was hypothesizedthat in some infants, the balance between circu-lating serotonin concentrations and intestinalsmooth muscle sensitivity to serotonin might

lead to painful gastrointestinal cramps in theevening when serotonin concentrations arehighest. The high nocturnal melatonin concen-tration antagonizes the contraction induced byserotonin on intestinal smooth muscle.

Alternatively, independent of the action ofmelatonin and serotonin on intestinal smoothmuscle, their effects might be mediated by thedeveloping central nervous system. For exam-ple, the occurrence of pineal melatonin rhyth-micity is coincident with consolidation of nightsleep.

In a study that defined colic as fussing or cry-ing for 3 or more hours on each of 3 successivedays, it was observed that colicky infants had ablunted rhythm in cortisol production comparedwith control infants.12 The control infants exhib-ited a clear and marked daily rhythm in cortisolthat was not observed in the colicky infants.However, the level of cortisol production aver-aged over the day was similar for both groups.Additionally, researchers in this study (who wereblind to the colic status) coded behavioral meas-ures from videotapes and arrived at the sameconclusion as many other studies: the crying ofthese infants was not caused by differences inhandling by the mother, and the colic was notsimply a maternal perception.

Maternal smoking was evaluated as a cause inone study, where colic was defined as more than3 hours of crying per day for more than 3 days aweek or more than 1.5 hours of crying per day in6 of 7 days. Maternal smoking occurred twice asoften among infants with colic.13 In contrast,food hypersensitivity has not been linked toinfantile colic.14

CRYING

Some degree of irritability, fussing, or crying isuniversal; crying for “unknown reasons” occursin all babies.15 Brazelton16 reported that themedian duration of crying was 1.75 hours dur-ing the second week, with a gradual increase to2.75 hours at 6 weeks, followed by a decrease incrying thereafter to 1 hour or less by 12 weeks ofage. The fussiest infants, whom he called “col-icky,” cried 2 to 4 hours per day every day, andtheir crying also increased between 6 and 8weeks of age. The distress experienced by par-ents over their inability to deal with this cryingcannot be overstated. Recent data have shown

114 Chapter 10 Sleep and Colic

Ch10.qxd 1/25/05 3:18 AM Page 114

infant homicides to increase after the secondweek and to peak at the eighth week, andPaulozzi and Sells17 concluded that the “peak inrisk in week 8 might reflect the peak in the dailyduration of crying among normal infantsbetween weeks 6 and 8.”

In studies of crying in infants, there are noclear discontinuities in measurements of irri-tability, fussing, or crying, whether by directobservation in hospital nurseries, by voice-activated tape recordings in homes, or by parentdiaries. Thus, colic appears to represent anextreme amount of normally occurring andunexplained crying or fussing present in allhealthy babies.

Because the spells of irritability, fussing, orcrying are universal, differing only in degreebetween infants, because the occurrence ofspells is related to postconceptual age and inde-pendent of parenting practices, and because thebehaviors exhibit state specificity and a circadianrhythm, it is reasonable to think that thesebehaviors reflect normal physiologic processes.An example is the normal physiologic processesinvolving the development of arousal–inhibitoryor wake–sleep control mechanisms. In all babies,the consolidation of night sleep develops duringthe second month (when the peak of cryingoccurs), and this periodic alternation of sleepand wake states is well developed by 3 to 4months of age (when colic ends).

FUSSING

Recent research has shown that persistent low-intensity fussing, rather than intense crying,characterizes infants diagnosed as having colic.8

In fact, the title of Wessel’s paper5 was“Paroxysmal fussing in infants, sometimes called‘colic.’ ” Fussing is not a well-defined behavior,and although not defined in Wessel’s paper, it isusually described as an unsettled, agitated,wakeful state that would lead to crying if ignoredby parents. Because sucking is soothing toinfants, some parents misattribute the fussingstate to hunger and vigorously attempt to feedtheir baby. These parents may misinterpret theirinfants as having a growth spurt at 6 weeksbecause they appeared to be hungry all the time,especially in the evening. They do not view theirchild as fussy. Even if the time spent feedingtheir infants at night to prevent crying is more

than 3 additional hours a day, on more than 3days a week, for more than 3 weeks, parents donot view them as colicky because there is apaucity of crying. Over a 34-month period, atnewborn visits, I routinely questioned every newparent who joined my general pediatric practicewhether their child fulfilled Wessel’s exact diag-nostic criteria for colic. All families had been fol-lowed since the child’s birth and receivedcounseling regarding the normal development ofcrying or fussing. There were 118 colicky infantsout of 747 (16%). However, the vast majority ofinfants had little or no crying. Instead, theyfulfilled Wessel’s criteria because they had longand frequent bouts of fussing, which did notlead to crying because of intensive parentalintervention.

In one study, fussing was defined by “Yourbaby is unsettled, irritable, or fractious and maybe vocalizing but not continuously crying” andcrying was defined as “periods of prolonged dis-tressed vocalizations” (see p. 2530 of St. James-Roberts and Plewis8). This study is noteworthybecause it supports previous research thatshowed that between 2 to 6 weeks, there is pre-dominately an increase in fussing, not crying.Furthermore, fussing and sleeping, but notably,not crying, were found to be stable individualcharacteristics from 6 weeks to 9 months of age.The amount of crying during the first 3 monthsdid not predict crying behavior at 9 months.More pointedly, and in agreement with anotherstudy,9 “high amounts of early crying do notmake it highly probable that an infant will . . .have sleeping problems at 9 months of age” (seep. 2538 of St. James-Roberts and Plewis8).

COLIC AND SLEEP

Kirjavainen and associates18 made the diagnosisof colic if infants exhibited “intensive crying(despite consoling efforts) for 3 hours or more aday, on 3 or more days a week” and were other-wise healthy and thriving. Parents kept a dailydiary, and night-time sleep onset was defined asthe time when the first 15-minute episodeoccurred after 9 PM. Sleep recordings were per-formed only at night between 9 PM and 7 AM.At about 4.5 weeks, the total sleep time from thediary was significantly shorter in the colic group(12.7 versus 14.5 hours/day, P < .001). Althoughcolicky infants slept less than the control group

Sleep and Colic Chapter 10 115

Ch10.qxd 1/25/05 3:18 AM Page 115

throughout the 24-hour record, the most dra-matic decrease in sleep between the groupsoccurred at night between 6 PM and 6 AM. Thediary data showed that by 6 months of age, thecolicky infants slept slightly less than the non-colicky infants, but the group differences werenot significant. The first polygraph recordingwas performed when the infants were about 9weeks old. There were no differences betweenthe groups in the night recordings in sleep char-acteristics. A subgroup analysis was performedon 11 colicky infants whose diary data wereclosest to the polygraph data, to take intoaccount the delay between the diaries and therecordings. In this older subgroup of colickyinfants, the total sleep time was increased (from12.7 to 13.2 hours/day). The second polygraphicrecordings were performed at about 30 weeks ofage, and again there were no differences in sleepcharacteristics between the infants formerly withand without colic. Also, at this age, the formerlycolicky infants had fewer nocturnal arousals.

This study by Kirjavainen and associates18 of15 colicky infants showed clearly that amonginfants with colic, parent diary data showedshorter total sleep times compared with the age-matched control group at 4.5 weeks, but that by9 weeks, there were no group differences in poly-graphic data obtained during the night. Also,this report suggests that over time, between theages of 5 and 9 weeks, sleep duration in colickyinfants increases. On the basis of only the poly-graphic data, the authors concluded that infan-tile colic was not associated with a sleepdisorder.

The enormous day-to-day variability ininfant colicky behavior noted in this study( J. Kirjavainen, personal communication, 2002)has been noted in another study.2 This variabil-ity makes it difficult to interpret the resultsderived from a single night-time polygraphicstudy. In addition, every child, includingKirjavainen’s son, had poorer sleep quality(more awakenings, shorter sleep duration) inthe laboratory than at home. The day-to-dayvariability and a laboratory effect may havemasked group differences.

St. James-Roberts and colleagues used theterm persistent criers to describe colicky infants,because their amount of fussing or cryingexceeded 3 hours per day most days in a week,and the term moderate criers for those who usu-

ally fussed or cried for 30 minutes or less per6 hour period of the day.19 Utilizing diaries thathad been validated with concurrent 24-houraudio recordings, the colicky infants, at 6 weeksof age, slept significantly less than noncolickyinfants (12.5 versus 13.8 hours/day). There wereno group differences regarding time spent awakeor time spent feeding. Colicky infants slept lessthroughout the 24-hour diary record. Unlikeother studies,12,18 the clearest group differencesfor sleep were in the daytime. In fact, there wereno significant group differences regarding sleepat night (between midnight and 6:00 AM).Additionally, at night, there were no statisticallysignificant group differences for crying or fuss-ing behavior. The clearest group differences forcrying or fussing behavior were in the daytime.The groups were similar in the timing and dura-tion of the infant’s longest sleep period. Thisanalysis of sleep cycle maturation led to the con-clusion that the “chief difference between themlies in amounts of daytime fuss/crying and sleep-ing, rather than in the diurnal organization ofsleep and waking behavior” (from p. 148 ofSt. James-Roberts et al.19). Additionally, at6 weeks of age, substantial negative correlationsbetween amounts of fuss/crying and sleepingwere observed in both groups. Because theauthors observed no deficit in wakefulness, onlysleeping, they felt that there was a specific trade-off between fussing or crying and sleep, “ratherthan a more global disturbance of infant behav-ioral states” (from p. 150 of St. James-Robertset al.19). The researchers concluded that persist-ent crying is associated with a sleep deficit.

Utilizing sensors embedded within a mattressto continuously monitor body movements andrespiratory patterns, a prospective study of 20colicky infants, who cried an average of at least2.5 hours per day, showed that at 7 and 13 weeksof age, they slept less than control infants. Thecolicky infants had more difficulty falling asleep,were more easily disturbed, and had less quiet,deep sleep.20

At about 8 weeks of age, utilizing diary datapreviously validated by audio taping, infantswere diagnosed with colic if they had fussing orcrying for 3 hours or more on each of 3 days dur-ing a 3-day study.12 Per 24 hours, the colickyinfants slept significantly less (11.8 versus 14.0hours per day). The colicky infants slept lessduring the day, evening, and night; however, the


Ch10.qxd 1/25/05 3:18 AM Page 116

difference in sleeping between the colickyinfants was significant only during the night-time. Crying and sleeping showed a significantnegative correlation in the colicky infants butnot in the control infants. White and colleagues12

concluded that colic might be associated with adisruption or delay in the establishment of thecircadian rhythm of sleep–wake activity. At4 months of age, a prospective study showedthat the mean total sleep duration, based onparental reports, of 48 infants who had hadcolic, based on Wessel’s exact definition, was13.9 ± 2.2 hours.2

In my general pediatric practice, where atevery visit all parents receive anticipatory adviceregarding sleep hygiene, the parents of colickyinfants describe the development of early bed-times, self-soothing at night-time sleep onset,longer night sleep periods, fewer night awaken-ings, and regular, longer naps later than the par-ents of noncolicky infants describe them. Thissuggests that colic may be associated with adelay in maturation of sleep–wake control mech-anisms. However, the data show that by 6, 8, and12 months, there are no significant differencesbetween colicky and control groups regardingduration of night sleep,18,21 although night wak-ing has been reported to be more common at 4,8, and 12 months in infants who previously hadcolic.4,21 This might be interpreted as a slightlydelayed maturation of normal nocturnal sleepduration coupled with a more persistent impair-ment of the learned ability to return to sleepunassisted during a naturally occurring noctur-nal arousal from sleep.

COLIC AND WAKEFULNESS

Parents of colicky infants often report that day-time sleep periods are extremely irregular andbrief. Also, some parents of colicky infantsdescribe a dramatic increase in daytime wakeful-ness and sometimes a temporary but completecessation of napping when their infantsapproached their peak fussiness at age 6 weeks.It has been suggested that, before 3 to 4 monthsof age, the period of inconsolability in theevening hours (when the infant cries and cannotsleep) might reflect periods of high arousal sim-ilar to the circadian forbidden zone (from p. 77of Weissbluth22). In adults, the forbidden zone isa period during which sleep onset and prolonged

consolidated and restorative sleep states do noteasily occur. In this context, it might be moreappropriate to describe colic not as a disorder ofimpaired sleep but as a disorder of excessivewakefulness in the evening. This view is sup-ported by recent polygraph investigations show-ing that in infants, a circadian forbidden zonedoes exist between 5 PM and 8 PM.23

COLIC AND TEMPERAMENT

Studies evaluating colic and temperament aredifficult to compare because of different ages ofinfants studied, small sample sizes (often withnull findings), small but statistically significantgroup differences, different temperament-assess-ment instruments, comparisons between tem-perament ratings and global impressions oftemperament, and possible measurement biasresulting from reliance on parental reports forboth crying and temperament.

Temperament characteristics of mood, inten-sity, adaptability, and approach/withdrawal ten-dency are statistically related, and infants whowere described as negative in mood, intense,slowly adaptable, and withdrawing were diag-nosed as having difficult temperaments becausethey were difficult for parents to manage. Theseinfants were also observed to be irregular in allbodily functions. When parents performed atemperament assessment at 2 weeks of age and a24-hour behavior diary at 6 weeks of age, it wasobserved that more difficult temperaments at 2weeks predicted an increased frequency andduration of crying and fussing at 6 weeks.24

Although statistically significant, the authorthought it was a weak association and concludedthat “early infant temperament predisposes toearly crying and fussing but is of limited use as aclinical predictor”(from p. 514 of Barr24).However, recent research utilizing a more com-prehensive assessment of infant temperamentfound the association between early infant tem-perament and later infant crying to be morerobust. At 4 weeks of age, infants who were moredifficult in general, more intense, and less dis-tractible (consolable) cried more during theirsecond month of life than other infants.25

Another prospective study of infants definedcolic as paroxysmal crying that is difficult toconsole and lasts for 3 or more hours a day dur-ing 3 or more days a week.21 Temperament


Ch10.qxd 1/25/05 3:18 AM Page 117

assessments were done at the ages of 3 and 12months. At 3 months, the colicky infants weremore intense, less persistent, less distractible,and more negative in their mood. However, at 12months, ratings on the temperament question-naire showed no group differences between thecolicky infants and the control group, but thegeneral impression of the mothers of the colickygroup was that they were more difficult (23%versus 5%).

Infants who had colic, using Wessel’s criteria,are significantly more likely to have a difficulttemperament than noncolicky babies when thetemperament assessment is performed at 4months of age.2 Furthermore, this progressionoccurs even when colic is successfully treatedwith dicyclomine hydrochloride,2 an atropinicdrug that may act centrally in the nervous systemto relax gastrointestinal smooth muscle. Similarresults were observed in babies who cried morethan 3 hours per day: when behavioral manage-ment significantly reduced evening fussing andcrying, this successful treatment had no effect ontemperament ratings, and the infants were stilldescribed by their parents as difficult.26 Theseresults suggest that congenital factors, not colic-induced parental distress or fatigue, causeincreased crying or fussing behavior during thefirst 3 to 4 months and subsequently lead to anassessment of difficult temperament at age 3 to 4months.

Colic does not appear to be an expression of apermanently difficult temperament. In one studyof only 10 colicky infants, in which colic wasdefined as cry bouts occurring more than 3 hoursper day more than 3 times per week, subsequentmeasurements of temperament at 5 and 10months did not show group differences betweenformerly colicky and noncolicky infants.27

SLEEP AND TEMPERAMENT

Continuous recordings of sleep patterns in thesecond day of life were correlated with tempera-ment assessments at 8 months. It was observedthat infants with the most extreme values on allsleep variables were more likely to have difficulttemperaments.28

Temperament assessments performed at amean age of 3.6 months showed an associationbetween problems of sleep–wake organization,difficult temperament, and extreme crying.29

These babies were selected for persistent cryingand were compared with babies with no currentcrying problem. Mothers of crying infants scoredhigh on depression, anxiety, exhaustion, anger,adverse childhood memories, and marital dis-tress. The authors concluded that factors relatedto parental care, although they did not cause per-sistent crying, did function to maintain or exac-erbate the behavior. The persistence of parentalfactors may explain why at 1 year there arereported to be more difficulty in communication,more unresolved conflicts, more dissatisfaction,and greater lack of empathy in families with acolicky infant30 and after 4 years, formerly col-icky children have been reported to be more neg-ative in mood on temperament assessments.31

In a study of 60 5-month-old infants, theinfants rated as difficult had mean sleep timessubstantially less than the infants rated as easy(12.3 ± 1.5 versus 15.6 ± 1.6 hours; P < .001).3

Although nine infant temperament characteris-tics were measured, only five were used to estab-lish the temperament diagnosis of difficult. Fourof these (mood, adaptability, rhythmicity, andapproach/withdrawal tendency) were highly sig-nificantly correlated negatively with total sleepduration. No significant correlations occurredwith the remaining temperament characteristics.Because the temperament characteristic “mood”relates to crying or fussing behavior, someauthors describe the association between colicand difficult temperament as an artifact (fromp. 43 of Barr and Gunnar32). However, mood isonly one of the five temperament characteristicscomprising the difficult temperament.

When the original study of 60 5-month-oldinfants was extended to include 105 infants,those with difficult temperaments slept 12.8hours and those with easy temperaments slept14.9 hours.33 This observation was subsequentlyconfirmed in another ethnic group with differentparenting practices.34 It thus appears that infantswho have a difficult temperament have briefertotal sleep durations when assessed at 4 to 5months of age.

Support for an association between sleep andtemperament is also based on a study in whichobjective measures of sleep–wake organization,derived from time-lapse video recordings, werecompared with parental perceptions of infanttemperament at 6 months of age.35 The authorsstate, “Infants considered (temperamentally)


Ch10.qxd 1/25/05 3:18 AM Page 118

easy have longer sleep periods and spend lesstime out of the crib for caretaking interventionsduring the night” (from p. 769 of Keeneret al.35). However, the authors’ analysis also ledthem to the conclusion that night waking iscaused by environmental rather that biologicfactors. Thus, this study has been cited as notsupporting a general association between tem-perament and sleep–wake patterns.36 It shouldbe noted that the increased time out of the cribfor temperamentally more difficult children at 6months corresponds to the observation thatincreased night waking occurs in formerly col-icky infants at 4, 8, and 12 months.9,21

Utilizing a computerized movement detector,it was observed that for 12-month-old children,those with the temperament trait of increasedrhythmicity went to sleep earlier and had longersleep durations,37 and by 18 months of age, therewas again the observation that both subjectiveand objective measures of improved sleep wereassociated with easier temperament assess-ments.38 The authors correctly state that it isunclear whether the association is a biologiccontinuity between daytime temperament fea-tures and night-time sleeping, or whether theinfluence of the child’s sleep behavior affects themother’s perceptions or ratings on temperamentassessment instruments.38

The same 60 infants who were evaluated at 4months of age3 were restudied at 3 years.39 Again,temperamentally easy children had longer sleepdurations compared with children with more dif-ficult temperaments. However, there was no indi-vidual stability of temperament or sleep durationbetween the ages of 4 months and 3 years. Thus,temperament ratings and associated sleep pat-terns at age 4 months do not predict tempera-ment or sleep patterns at 3 years.

POSTCOLIC SLEEP

A retrospective study of 141 infants between 4and 8 months of age from middle-class familiesshowed that the history of colic, using Wessel’sexact definition, significantly covaried with theparent’s judgment that night waking was a cur-rent problem.4 The frequency of awakening wasa problem in 76% of infants, the duration ofawakenings was a problem in 8%, and both fre-quency and duration were a problem in 16%.The frequency of night awakenings significantly

correlated positively with the duration of nightawakenings. Other studies also reported morenight waking at 8 and 12 months21 and ages 14 to18 months40 in postcolic children. Among infantswho had had colic, the total sleep duration wasless than among those who had not had colic(13.5 ± 0.3 versus 14.3 ± 0.2 hours, respectively;P < .05).4

Group differences in sleep duration betweencolicky and noncolicky infants, and betweeneasy and difficult infants were observed to gen-erally decrease over time.41 Data obtained evenduring the first 4 months support a trend towardnormalization.20 For example, in one study,infants were diagnosed as irritable if cryingoccurred an average of at least 2.5 hours per dayat 1 month of age. The irritable infants slept less,but over the 3-month study period, differencesbetween the two groups diminished as theamount of crying subsided. Also, the irritableinfants “spent less time in the content-awake orquiet-alert state that is optimal for parent-infantinteraction” (from p. 8 of Keefe et al.20).

Some studies suggest that both infant irri-tability and sleep deficits are moderately stableindividual characteristics during the first year oflife or beyond.8,9,40 One study showed that chil-dren with colic had more sleeping problems andthe families exhibited more distress than a con-trol group at age 3 years.41 The trend of decreas-ing group differences with increasing agebetween colicky and noncolicky infants withregard to sleep, taken together with the normalpolygraph recordings of colicky infants at 9weeks of age, suggests that it is parenting prac-tices, not biologic factors, that contribute toenduring sleep problems beyond 9 weeks of age.

As previously reported,42 it may be difficultfor parents of formerly colicky infants to elimi-nate frequent night awakenings and to lengthensleep durations after those infants are 4 monthsold. Parents, because they are fatigued, mayunintentionally become inconsistent and irregu-lar in their responses to their infant. It cannot beoveremphasized that “parents are never trulyprepared for the degree to which the babies’sleep–wake patterns will dominate and com-pletely disrupt their daily activities” (from p. 390of Parmelee43). They may become overindulgentand oversolicitous regarding night awakeningsand fail to appreciate that they are inadvertentlydepriving their child of the opportunity to learn


Ch10.qxd 1/25/05 3:18 AM Page 119

how to fall asleep unassisted. Some mothershave difficulty separating from their child, espe-cially at night, and other mothers have a ten-dency toward depression that might be aggra-vated by the fatigue from struggling to cope witha colicky infant. In either case, simplistic sug-gestions to help the child sleep better often failto motivate a change in parental behavior. If thechild fails to learn to fall asleep unassisted, theresult is sleep fragmentation or sleep deprivationdriven by intermittent positive parental rein-forcement. This causes fatigue-driven fussinesslong after the colic has resolved, which ulti-mately creates a sleep-deprived family.

Support for this view has come from researchon infants at 5 months of age who were followedto 56 months of age. Wolke and Meyer9 showedthat “long crying duration and [the parents] hav-ing felt distressed about crying during the first 5months were significant predictors of night wak-ing problems at 20 months” (p. 198 [italicsmine]) but not at 56 months. In particular, latersleep problems are mainly related to the comor-bidity of crying with sleep problems at 5 monthsrather than to crying problems alone. Sleep prob-lems at 5 months remained the best predictor ofsleep problems (especially night waking) at 20months. These authors concluded that postcolic“sleep problems are likely to be due to a failure ofthe parents to establish and maintain regularsleep schedules. . . . This conclusion does notblame parents for sleep difficulties: rather, it rec-ognizes why many parents adopt strategies todeal with night waking in the least conflictualmanner by night feeding or co-sleeping. Thismay be especially true of parents who are dealingwith a temperamentally more difficult infant”(from p. 203 of Wolke and Meyer9). The authorsalso concluded that postcolic sleeping problemsare not caused simply by increased crying per sebut appear to be the consequence of associatedinfant sleeping problems and altered care-takingpatterns for dealing with night waking in infancy.

Bates and associates recently evaluated theinteraction between family stress, family manage-ment, disrupted child sleep patterns (variabilityin amounts of sleep, variability in bedtime, andlateness of bedtime), and preschool adjustment inchildren about 5 years old.44 Children with dis-rupted sleep did not adjust well in preschool, and,in this analysis, disrupted sleep directly causedthe behavior problems. These authors did not find

any evidence that family stress or family manage-ment problems caused both disrupted sleep andbehavior problems. They concluded that “sleepirregularity accounted for variation in (behav-ioral) adjustment independently of variation infamily stress and family management” (from p. 70of Bates et al.44). Interestingly, in support of theirconclusion that disrupted sleep and not familystress was the main cause of behavioral problems,they observed that “in clinical treatment of youngoppositional children, we have seen some spec-tacular improvements in manageability associatedwith the parents instituting a more adequateschedule of sleep for their children. Our clinicalimpression in these cases was that the changeswere too rapid to be accounted for by otherchanges, such as in parental discipline tactics”(from p. 71 of Bates et al.44). The senior authoragrees with my hypothesis that sleep modulatestemperament and that “parenting responses to(sleep) issues would be involved in the continu-ity/discontinuity of temperament. . . . If parentsmake the effort to manage their kids’ sleep sched-ules consistently, I would think that over yearsthey are going to see less difficult and unmanage-able behavior” ( J. E. Bates, personal communica-tion, 2002).

In a recent report, 64 children who had had,as infants, “persistent crying” (defined as fussingor crying more than 3 hours on 3 days in theweek) were studied at 8 to 10 years of age.45 Theauthors concluded that they were at risk forhyperactivity problems and academic difficul-ties. In addition, at 8 to 10 years of age, the pre-vious persistent criers took longer to fall asleep,suggesting to the senior author that “they wereless effective in controlling their own behav-ioural state to fall asleep” (D. Wolke, personalcommunication, 2002).

Therefore, it appears that increased crying orfussing behavior alone in infancy does notdirectly cause later sleep problems. Although thepostcolicky child’s family may be stressed, itappears that it is the failure to establish age-appropriate sleep hygiene that specifically leadsto later disrupted sleep and behavioral problems.

SUMMARY AND DISCUSSION

Different definitions of colic and different ages ofinfants at the time of the study make it difficultto come to firm conclusions. During the first 4


Ch10.qxd 1/25/05 3:18 AM Page 120

months, colicky infants, by definition, exhibitmore crying or fussing behavior. Data from par-ent diaries obtained at 4.5, 6, 7, and 8 weeks ofage show that colicky infants sleep less thannoncolicky infants (about 12 to 12.5 hours ver-sus about 14 to 14.5 hours),8,12,18,20 but there isdisagreement over whether the decreased sleepoccurs predominately during the day or nighthours. By 9 weeks of age, polygraphic data donot show group differences in sleeping betweencolicky and noncolicky infants. Although groupdifferences in sleeping duration between infantswho had had colic and infants who had not beencolicky were present even at 4 months of age,4

these differences disappear by 6 to 8 months.18,21

This raises the suggestion that parenting prac-tices might be especially important in affectingsleep patterns after 9 weeks, especially regardingthe development of a night waking habit. Also,by 6 months of age, researchers are more apt toview parents as contributing to sleeping prob-lems,35 especially night waking.

There appears to be agreement that infantcrying alone does not predict the development of sleep problems.8,9 Rather, the comorbidity ofcrying plus parental distress at 5 months, orof crying plus sleep problems at 5 months, pre-dicts night waking at 20 months but not at 56months.

Temperament assessments rely on parentalreports, which often raises questions about howmuch the assessments reflect parental percep-tions as opposed to objective descriptions ofbehavioral styles in children. At 2 and 4 weeks ofage, infant difficultness predicted increased cry-ing or fussing at about 6 weeks of age.24,25

Infants with colic are more likely to have a diffi-cult temperament when assessed at 4 months ofage2 but not at 12 months.21 A difficult tempera-ment is associated, at many ages, with problemsin sleeping, such as shorter sleep durations andnight waking, but this association is not predic-tive of later sleep problems. At 4 to 5 months ofage, infants with a difficult temperament havetotal sleep durations of about 13 hours asopposed to about 15 hours for infants with aneasy temperament.

The association of difficult temperament andsleeping problems during and shortly after age 4months occurs despite successful treatment ofcolic. This suggests that crying or fussing behav-ior does not simply cause a measurement bias

that makes parents imagine their children bothto be difficult and to not sleep well. My impres-sion is that it is exactly those parents with thewillingness and resources to invest heavily insoothing during periods of fussing who are ableto prevent some of the fussing that would other-wise escalate into crying, thus preventing somepostcolic sleep problems. On the other hand,some parents are unable to manage severe infantfussing and become overwhelmed by the crying.Or perhaps they feel that because they cannotinfluence their child’s crying behavior during thefirst few months, they also cannot influencetheir child’s sleeping habits later.

It is important to help parents of postcolickyinfants establish healthy sleep habits. Some ofthese children have difficulty falling asleep andstaying asleep. At about 4 months, they have notdeveloped self-soothing skills, perhaps becauseparents had invested constant soothing to pre-vent fussiness from developing into crying, orperhaps because inability to self-soothe is in factan integral component of colic. A successfulintervention effort to help families cope withinfant crying during colic reduces parental dis-tress. Continued education about age-appropri-ate sleep hygiene after colic ends is likely toprevent sleep problems from persisting beyond4 months. Unsuccessful intervention and educa-tion increases the likelihood that temperamentissues, family stress, and sleeping problems willpersist beyond 4 months.

LINKS BETWEEN COLIC, SLEEP,AND TEMPERAMENT: AN ALTERNATIVE VIEW

The previous discussion assumes that individ-ual differences in the neurologic maturation ofsleep–wake control mechanisms, and the abilityto regulate state (or self-soothe) are the primarydeterminants of how much fussiness, crying,and sleeping occur in babies during thefirst 4 months. Parenting practices duringthe first 4 months were ascribed a secondaryrole that develops in reaction to infant fussinessand crying. After 4 months, parenting practiceswere considered to be more important in main-taining sleep problems. An alternative view isthat parenting practices do indeed make a majorcontribution to the outcome during the first4 months of life.


Ch10.qxd 1/25/05 3:18 AM Page 121

This analysis is based on data from an out-come study using Wessel’s exact definition ofcolic and data on temperament at 4 months ofage,2 an analysis of sleep and temperament atabout 4 months of age,3 and an assumption thatthe difficult temperament at about 4 months rep-resents an overtired behavioral style. Certainly,the natural history of crying and fussy behavior(age of onset, age of cessation, evening time ofoccurrence, and occurrence mainly during thefirst 3 or 4 months with a peak at about 6 weeks)and its occurrence in all cultures strongly arguefor biologic factors regarding etiology. However,the possibility remains that the severity (orintensity) and the duration of these behaviorsare influenced by parenting practices. For exam-ple, infants in traditional or tribal cultures thatpractice constant holding and frequent breast-feeding may have less fussiness or crying. Also,the videotape studies that show no differencesbetween parenting behaviors for colicky andnoncolicky infants might reflect either an obser-vation period that is too short, a too-small sam-ple size, or extra effort on the part of the parent,motivated by the presence of video cameras inthe home during the study.

Consider a balance between the baby’s dispo-sition to express distress and the parent’s capabil-ity to soothe the baby. Not only do babies vary intheir expression of fussing or crying behavior butalso parents vary in their capability to soothe.Factors that affect parents’ ability to soothe fussi-ness and crying and to promote their baby’s sleep,and that might influence the outcome over thefirst few months, include the following:

1. Father involvement versus absent father2. Agreements or disagreements between

mother and father regarding child rearing,such as breast-feeding versus bottle-feedingor crib versus family bed

3. Absence or presence of marital discord orintimacy issues between wife and husband

4. Absence or presence of baby blues or post-partum depression

5. Absence or presence of other children requir-ing attention

6. Ease or difficulty of breast-feeding7. Absence or presence of medical problems in

child, mother, father, or other children8. Number of bedrooms in the home9. Presence or absence of relatives, friends, or

neighbors to help

10. Help or interference from grandparents11. Ability or inability to afford housekeeping

help12. Ability or inability to afford childcare help13. Absence or presence of financial pressures,

such as for mother to return to work soon

For example, assuming that 20% of infantshave colic, consider three groups of childrenwho developed along different paths:

A. In a cohort of 20 infants with colic, about 75%,or 15 infants, will not have a difficult tempera-ment at 4 months of age.2 At 4 months of age,these children will have an easier temperamentand sleep longer than other children. The colicwill have resolved. These children may havehad a milder form of colic, or their parents mayhave vast resources for soothing.

B. In the same cohort of 20 infants with colic,about 25%, or five infants, will have a difficulttemperament at 4 months of age and sleepless than other children.2 At 4 months of age,the discrete spells of evening fussing or cryingwill be less well defined, but the dispositionto express distress at any time will persistbecause these children remain overtired.Therefore, they will tend to be at their worstat the end of the day when they are most over-tired. I call this fatigue-driven fussiness.These children may have had a more severeform of colic, or their parents may have lim-ited resources for soothing.

C. Among 80 noncolicky infants, about 5%, orfour infants, will have a difficult temperamentat 4 months of age and resemble those infantsin group B: they sleep less than other chil-dren, they are overtired, and they fuss and cry,especially at the end of the day. The parents ofthese children may have limited resources forsoothing, or they may allow their child tobecome overtired by skipping naps, allowingtoo late a bedtime, or contributing to frag-mented night sleep by giving too much atten-tion at night. Alternatively, circumstancesbeyond the parents’ control, such as extensiveand necessary traveling, interfere with estab-lishing sleep routines.

Therefore, at 4 months of age, there are nineinfants who are overtired and who fuss or crymore than other children. The five (56%) fromgroup B had colic and the four (44%) from groupC did not have colic. These percentages are in


Ch10.qxd 1/25/05 3:18 AM Page 122

close agreement with a recent study of colickyinfants, in which colic was defined as “parox-ysms of irritability, fussing, or crying lasting for3 or more hours in any 1 day and occurring on 3or more days in any 1 week” (from Clifford et al.,p. 118446). Mothers who participated, comparedwith mothers who declined or dropped out ofthe study, were more likely to be married, tohave completed more formal education, to havea higher family income, to have higher levels ofsocial support and lower levels of anxiety anddepression, to breast-feed, and to be nonsmok-ers. The researchers described three groups ofinfants. In their first group, colic resolved by 3months of age in 86% of infants, similar to groupA (75%). In their second group, at 3 months ofage, there were 35 children with colic, of whom18 (51%) had colic at 6 weeks of age, similar togroup B (56%). The authors referred to thesechildren as having “persistent colic.” In theirthird group, at 3 months of age, there were also17 children (49%) with colic but they did nothave colic at 6 weeks of age, similar to group C(44%). The authors referred to this group ofinfants as having “latent distress.” They did notpublish data comparing the mothers of infantswith “persistent colic” versus “latent distress.”

Although the authors46 suspected that thepersistent colic occurred in the 18 infantsbecause of “a persistent mother-infant distresssyndrome,” they noted that colic “does notresult in lasting effects to maternal mentalhealth.” However, they acknowledged that theirresults regarding the effects of colic on maternalhealth might be different in other populations“with different access to resources that may easethe burden of caring for a fussy infant . . . (whomight not) be able to seek respite when they feltoverwhelmed . . . (or did not have) adequateresources to buffer the effects of their infants’colic (p. 1186-1187).” The reason that I object tothe term mother-infant distress syndrome is that itputs undue blame on mothers, when fathers,marriage, family, and financial concerns mayplay an equal if not greater role in the family’sability to soothe a baby. In fact, one recent studyshowed dysfunctional interactions between thecolicky baby and the father to be much greaterthan between the colicky baby and the mother.47

After 4 months of age, it is my impression thatthe persistence of limited resources for soothingusually causes a continuation of sleeping prob-

lems and parental distress. However, interventioncan correct the sleep problems and alleviate someof the parental distress. An improvement in par-enting practices regarding sleep produces rapidand dramatic improvement in sleeping andbehavior in children from group C, but theimprovement in children from group B is muchslower and more incremental. This is most likelybecause the infants from group B and their par-ents are more overtired, the children have beenmore consistently parent-soothed, and there maybe lingering biologic delays in the developmentof sleep consolidation and nap rhythms.

The clinician should be aware of the socialresources available to a family, because lack ofthese resources interferes with the parents’ abil-ity to soothe their child and help their childsleep. Clear identification of the social strengthsand weaknesses in the family’s support systempermits a more individualized intervention pro-gram.

References

1. Weissbluth M: Crybabies: Coping with Colic,What To Do When Baby Won’t Stop Crying. NewYork, Arbor House, 1984, pp 66-79, 131-146.

2. Weissbluth M, Christoffel KK, Davis AT: Treat-ment of infantile colic with dicyclomine hydro-chloride. J Pediatr 1984;6:951-955.

3. Weissbluth M: Sleep duration and infant tempera-ment. J Pediatr 1981;99:817-819.

4. Weissbluth M, Davis AT, Poncher J: Night wakingin 4- to 8-month-old infants. J Pediatr 1984;104:477-480.

5. Wessel MA, Cobb JC, Jackson EB, et al: Parox-ysmal fussing in infancy, sometimes called“colic.” Pediatrics 1954;14:421-435.

6. Illingworth RS: Three-months’ colic. Arch DisChild 1954;29:165-172.

7. Reijneveld SA, Brugman E, Hirasing RA: Excessiveinfant crying: The impact of varying definitions.Pediatrics 2001;108:893-897.

8. St. James-Roberts I, Plewis I: Individual differences,daily fluctuations, and developmental changes inamounts of infant waking, fussing, crying, feeding,and sleeping. Child Dev 1996;67:2527-2540.

9. Wolke D, Meyer R: Co-morbidity of crying andfeeding problems with sleeping problems ininfancy: Concurrent and predictive associations.Early Dev Parenting 1995;4:191-207.

10. Kurtoglu S, Uzum K, Hallac IK, Coskun A:5 Hydroxy-3-indole acetic acid levels in infantilecolic: Is serotoninergic tonus responsible for thisproblem? Acta Paediatr 1997;86:764-765.


Ch10.qxd 1/25/05 3:18 AM Page 123

11. Weissbluth L, Weissbluth M: Infant colic: Theeffect of serotonin and melatonin circadianrhythms on intestinal smooth muscle. Med Hy-potheses 1992;39;164-167.

12. White BP, Gunnar MR, Larson MC, et al:Behavioral and physiological responsivity, sleep,and patterns of daily cortisol production ininfants with and without colic. Child Dev 2000;71:862-877.

13. Sondergaard C, Henriksen TB, Orbel C, WisborgK: Smoking during pregnancy and infantile colic.Pediatrics 2001;108:342-346.

14. Hill DJ, Hosking CS: Infantile colic and foodhypersensitivity. J Pediatr Gastroenterol Nutr2000;30:S67-S76.

15. Aldrich CA, Sung C, Knop C: The crying of newlyborn babies: 1. The community phase. J Pediatr1945;26:313-326.

16. Brazelton TG: Crying in infancy. Pediatrics 1962;29:40-47.

17. Paulozzi L, Sells M: Variation in homicide riskduring infancy—United States, 1989-1998.MMWR Morb Mortal Rep 2002;51:187-189.

18. Kirjavainen J, Kirjavainen T, Huhtala V, et al:Infants with colic have a normal sleep structure at2 and 7 months of age. J Pediatr 2001;138:218-223.

19. St. James-Roberts I, Conroy S, Hurry J: Linksbetween infant crying and sleep-waking at sixweeks of age. Early Child Dev 1997;48:143-152.

20. Keefe MR, Kotzer AM, Froese-Fretz A, Curtin M:A longitudinal comparison of irritable and nonir-ritable infants. Nurs Res 1996;45:4-9.

21. Lehtonen L, Korhonen T, Korvenranta H:Temperament and sleeping patterns in colickyinfants during the first year of life. J Dev BehavPediatr 1994;15:416-420.

22. Weissbluth M: Colic. In Ferber R, Kryger MH(eds): Principles and Practice of Sleep Medicinein the Child. Philadelphia, WB Saunders, 1995,pp 75-78.

23. Giganti F, Fagioli I, Ficca G, Salzarulo P:Polygraphic investigation of 24-h waking distribu-tion in infants. Physiol Behav 2001;73:621-624.

24. Barr RG: Feeding and temperament as determi-nants of early infant crying/fussing behavior.Pediatrics 1989;84:514-521.

25. Blum NJ, Taubman B, Tretina L, Heyward RY:Maternal ratings of infant intensity and dis-tractibility: Relationship with crying during thesecond month of life. Arch Pediatr Adolesc Med2002;156:286-290.

26. Wolke D, Gray P, Meyer R: Excessive infant cry-ing: A controlled study of mothers helping moth-ers. Pediatrics 1994;94:322-332.

27. Stifter CA, Braungart J: Infant colic: A transientcondition with no apparent effects. J Appl DevPsychol 1992;13:447-462.

28. Novosad C, Freudigman K, Thoman EB: Sleeppatterns in newborns and temperament at eightmonths: a preliminary study. J Dev Behav Pediatr1999;20:99-105.

29. Papousek M, von Hofacker N: Persistent crying inearly infancy: A non-trivial condition of risk forthe developing mother-infant relationship. ChildCare Health Dev 1998;24:395-424.

30. Raiha H, Lehtonen L, Korhonen T, KorvenrantaH: Family life 1 year after infantile colic. ArchPediatr Adolesc Med 1996;150:1032-1036.

31. Canivet C, Jakobsson I, Hagander B: Infantilecolic: Follow-up at four years of age—Still more“emotional.” Acta Paediatr 2000;89:1-2.

32. Barr RG, Gunnar M: Colic: The ‘TransientResponsivity’ Hypothesis. In Barr RG, Hopkins B,Green JA (eds): Crying as a Sign, a Symptom, anda Signal. Series: Clinics in DevelopmentalMedicine, vol 152, Cambridge: Cambridge Uni-versity Press, 2000, pp 42-44.

33. Weissbluth M, Liu K: Sleep patterns, attentionspan, and infant temperament. J Dev BehavPediatr 1983;4:34-36.

34. Weissbluth M: Chinese-American infant tempera-ment and sleep duration: An ethnic comparison.Dev Behav Pediatr 1982;3:99-102.

35. Keener MA, Zeanah CH, Anders TF: Infant tem-perament, sleep organization, and nighttimeparental interventions. Pediatrics 1988;81:762-771.

36. Anders TF, Sadeh A, Appareddy V: Normal sleepin neonates and children. In Ferber R, Kryger M(eds): Principles and Practice of Sleep Medicinein the Child. Philadelphia, WB Saunders, 1995,p 15.

37. Scher A, Tirosh E, Lavie P: The relationshipbetween sleep and temperament revisited:Evidence for 12-month-olds—A research note.J Child Psychol Psychiatry 1998;39:785-788.

38. Scher A, Epstein R, Sadeh A, et al: Toddlers’ sleepand temperament: Reporting bias or a validlink—A research note. J Child Psychol Psychiatry1992;33:1249-1254.

39. Weissbluth M: Sleep duration, temperament, andConners’ ratings of three-year-old children. J DevBehav Pediatr 1984;5:120-123.

40. Stahlberg M-R: Infantile colic: Occurrence andrisk factors. Eur J Pediatr 1984;143:108-111.

41. Rautava P, Lehtonen L, Helenius H, Sillanpaa M:Infantile colic: Child and family three years later.Pediatrics 1995;96:43-47.

42. Weissbluth M: Sleep and the colicky infant. InGuilleminault C (ed): Sleep and Its Disordersin Children. New York, Raven Press, 1987, pp129-140.

43. Parmelee AH Jr: Remarks on receiving theC. Anderson Aldrich Award. Pediatrics 1977;59:389-395.


Ch10.qxd 1/25/05 3:18 AM Page 124

44. Bates JE, Viken RJ, Alexander DB, et al: Sleep andadjustment in preschool children: Sleep diaryreports by mothers relate to behavior reports byteachers. Child Dev 2002;73:62-74.

45. Wolke D, Rizzo P, Woods S: Persistent infant cry-ing and hyperactive problems in middle child-hood. Pediatrics 2002;109:1054-1060.

46. Clifford TJ, Campbell MD, Speechley KN,Gorodzinsky F: Sequelae of infant colic: Evidence

of transient distress and absence of lasting effectson maternal mental health. Arch Pediatr AdolescMed 2002;156:1183-1188.

47. Raiha H, Lehtonen L, Huhtala V, et al: Excessivecrying infant in the family: Mother-infant, father-infant, and mother-father interaction. Child CareHealth Dev 2002;28:419-429.


Ch10.qxd 1/25/05 3:18 AM Page 125

Sleeplessness is one of the most frequent com-plaints presented to the sleep practitioner, sincethe entire family is affected by a child who can-not sleep at night. Symptoms of sleep depriva-tion may be identified in one or all members ofthe family. It is not unusual for a worn, haggard,frustrated parent to seek help for a child whoappears happy, active, alert, and well rested.

Sleeplessness appears to be a final commonpathway for a heterogeneous group of disordersthat may have many etiologies. Underlying causesmay be chronophysiologic, medical, psychologi-cal, social/environmental, or a combination offactors. Indeed, considerable overlap exists amongunderlying conditions. Some children may haveno difficulty falling asleep but are unable to main-tain continuity of sleep throughout the night,resulting in multiple nocturnal awakenings.Other children may have significant difficultyfalling asleep at a prescribed time but, onceasleep, will sleep throughout the night. Still oth-ers may suffer from both difficulty settling andnight-time awakenings.

The typical connotation of the term “insom-nia” does not appear to be applicable to sleeplesschildren. The underlying causes of disordersof initiating and maintaining sleep (DIMS) dur-ing childhood appear to be considerably differ-ent from those of insomnia in adults, anddifferent approaches to diagnosis and therapy arerequired.

Sleep onset and maintenance require appro-priate interaction between biologic and psy-chological determinants. When one or more ofthese factors are disrupted or dysfunctional, sleep-lessness results. Disorders of partial arousal fromsleep stages and sleep–wake transition disorders(such as sleepwalking and night terrors) are dis-cussed in Chapter 9.

To appropriately manage these problems, accu-rate diagnosis is essential. This chapter focuses on

the epidemiology, clinical presentation, diagnosticfeatures, differential diagnosis, and managementparadigms for DIMS in childhood.

GENERAL CONSIDERATIONS

The problem of sleeplessness during childhooddoes not appear to be solely neurodevelopmen-tal. Instead, sleeplessness reflects a pattern ofinteraction between physiologic and behavioralvariables.1 Interactions between the parent andthe child at times of sleep transition are impor-tant. Parents display a wide variety of normalresponses to their sleeping infant, and theseresponses change as the infant ages. Under nor-mal circumstances, parents frequently respondto a signaling (crying) infant during the firstfew months of life. With older infants, parentalintervention usually involves occasional checksduring sleep. It is interesting to note that duringthe first two thirds of the first year of life, wheninfants are undergoing rapid neurodevelop-mental changes and entrainment to appropriatesleep–wake cycles, almost half of all infants arealready asleep when placed in the crib.2

As the infant ages, sleeplessness may con-tribute to significant parental anxiety, anger, fear,feelings of helplessness, and feelings of hope-lessness. Parental concerns focus not only on thechild’s difficulties sleeping but also on their ownsleeplessness. Prolonged night-time awakeningsshould be thought of as being derived from aninteraction between the child’s temperamentalpredisposition for awakening at night and otherfactors of health, development, and parentalmanagement.3 The concerns of parents may bejustified. Recent studies in adults have suggestedthat short sleep duration is related to increasedmorbidity.4 During childhood, a lack of sleepseems to significantly affect school performance.A study of 9000 children with superior school

Disorders of Initiating and Maintaining Sleep

Stephen H. Sheldon

11

127

Ch11.qxd 01/28/05 10:30 AM Page 127

performance5 and higher test scores on intelli-gence tests6 were associated with a longer dura-tion of night-time sleep.

Despite nocturnal awakenings, it is rare forchildren less than 5-years of age to suffer from sig-nificant sleep deprivation. Young infants and tod-dlers sleep when necessary at any place or time.Although the parents are often concerned thata child awakening at night might be sleep dep-rived, it is not usually the case. Problems arisewhen social and educational demands are juxta-posed with sleep demands.7

The parents of older children often considerthe total sleep time that is affected by televisionviewing in the evening. Weissbluth and colleagueshave shown that total sleep time is the sametoday as it was 60 years ago.8 Television viewinghas had minimal or no effect on daytimeor evening sleep in children 5 years of age orolder. Although girls sleep slightly longer thanboys at night, there appears to be no sex differ-ence for daytime sleep or television viewing.The sex difference in total nocturnal sleep timeis also seen in adults. Though there was no changein total sleep time, increased television viewingcorrelated with consistently decreased scoreson standardized tests for reading, written expres-sion, and mathematics.8

INCIDENCE AND PREVALENCE

Brief awakenings at night are normal in childrenand adults. Most often the individual turns over,immediately falls back asleep, and does not recallthe arousal.7 Studies of incidence and prevalenceof night-time awakening in the pediatric popula-tion have been limited because of their sub-jective nature, relying on parental reports andquestionnaires. Despite the limitations of thismethod of investigation, the frequency of sleep-lessness during childhood seems high. Bax hasshown that nocturnal awakenings during thefirst 5 years of life approximate 20% in the 0- to2-year-old age group. There seems to be a slightdecline in 3- to 4-year-old children, but the prob-lem is still common in 10% of 4- to 5-year-oldchildren.7 Of the children exhibiting nocturnalawakenings at 12 months of age, 40% still expe-rienced them at 18 months. Forty percent ofchildren who were still awakening at 18 monthscontinued to awaken at 2 years (but not neces-sarily the same 40%). When children awakening

at 18 months were studied, 54% were still awak-ening at 2 years and 23% at 3 years of age. At 5years nocturnal awakening still occurred, thoughless significantly and less frequently after schoolentry.

In another study by Jenkins and associates,night-time awakening was reported in 23% ofchildren at 1 year, 24% at 18 months, and 14% at3 years of age.9 Richman and coworkers showedthat at 3 years of age, 13% of children have prob-lems settling and 14 percent awaken at night.10

Inherent difficulties are present in the evalua-tion of data collected by parental reports andquestionnaires, since night-time awakeningsare noted only if the child is fussy and requiresattention. Periods of quiet wakefulness duringnormal sleep hours are missed. To overcomethis problem, Anders longitudinally videotapedparent and baby interactions during normalsleep periods.11 Settling was defined as sleepingwithout removal from the crib between midnightand 5:00 AM for at least 4 weeks and night-timeawakening as an arousal during that time at leastonce weekly for 4 weeks. A simple awakeningoccurred when the baby awoke after beingasleep and then returned to sleep without fuss-ing or being removed from the crib. A complexawakening occurred when the baby awoke afterbeing asleep and then fussed and was removedfrom the crib. Sixty percent of 2-month-oldinfants and 62% of 9-month-old infants wereinitially placed in their cribs awake. Seventy per-cent of the males were placed in their cribsawake, compared with 50% of the females. Therewas also a difference between sexes in the timethe infants were placed in their cribs. Femalesat 2 and 9 months of age were placed in theircribs between 8:00 and 9:00 PM and removedat 7:00 and 8:00 AM. Males at 2 months wereplaced in their cribs between 7:00 and 10:00 PM

and removed at 6:00 and 8:00 AM. Males at9 months were placed in their cribs between7:00 and 9:00 PM and removed at 6:00 and 8:00AM. Average sleep onset latency for the groupof infants placed in their cribs awake was 27.5minutes for 2-month-olds and 16.4 minutes for9-month-olds. Two thirds of the infants whowere placed in their cribs awake entered activesleep first. This correlated well with polysomno-graphic studies that showed active sleep-onsetREM (rapid eye movement) during infancy.12

Half of the awakenings for those at 2 and 9 months

128 Chapter 11 Disorders of Initiating and Maintaining Sleep

Ch11.qxd 01/28/05 10:30 AM Page 128

of age were simple and half were complex. Thelength of simple awakenings was 9 minutes forthose at 2 and 9 months and tended to be moreprominent as morning approached. Complex awak-enings lasted longer and the duration changedwith age, being 75 minutes for 2-month-olds and24 minutes for 9-month-olds.11 At 2 months ofage, complex awakenings occurred predomi-nantly between midnight and 6:00 AM. Femalesseemed to awaken more regularly between 3:00and 6:00 AM. At 9 months, males demonstrated2 peaks of complex awakening. The first peakoccurred between 9:00 PM and midnight and thesecond between 3:00 and 6:00 PM. Femalesremained more regular, with a single peakbetween 3:00 and 6:00 AM.

Moore and Ucko reported that 50% of theinfants who had settled successfully awakenedagain at night when they were 7 to 12 monthsof age.13 Most parents and child health care prac-titioners will attest to the presence of this secondpeak of nocturnal awakenings. The etiology maybe developmentally or physiologically based,however, night-time awakenings during this timemay also be secondary to shifts and stressesin the environment.13

BEHAVIORAL CHARACTERISTICS,TEMPERAMENT, AND NIGHT-TIMEAWAKENING

Zuckerman and colleagues conducted a longitu-dinal study of night-time awakening based oninterviews with 308 parents.14 For infants at8 months of age, 10% of the mothers reportedthat their babies awoke three or more times pernight, 8% reported that their babies took an houror longer to settle after awakening, and 5% com-plained that their own sleep was severely dis-rupted by the child; 18% reported at least one ofthese problems. By 3 years of age, 29% of thechildren enrolled in the study had difficulty set-tling and/or staying asleep. Of the children withsleep problems at 8 months of age, 41% still hada problem sleeping at 3 years of age, comparedwith only 26% of the children without a sleepingproblem at 8 months of age who developeda problem by 3 years of age. In this study, childrenwith persistent sleep problems were more likelyto have behavior problems, especially tantrumsand behavior management problems, than thosewithout persistent sleep problems. On the other

hand, children with sleep problems were notmore likely to have fears, anxiety, or other behav-ior problems as measured by the BehaviorScreening Questionnaire.

Some evidence exists that there is a tempera-mental predisposition to night-time awakening.If this is true, the evaluation of the sleeplesschild requires a broader perspective. In a privatepediatric practice, 25% of 60 randomly selectedpatients between 1 and 12 months of age sufferedfrom night-time awakenings, which correlatedsignificantly with temperamental characteristicsof low sensory threshold .3

To determine if common sleep disturbancesin young children, such as night-time awakeningand bedtime struggles, tend to persist; if they arerelated to environmental stress factors andare accompanied by other behavioral problems;and if their persistence is related to other factors,Kataria and associates studied 60 childrenbetween the ages of 15 and 48 months.15 Themothers were interviewed at the beginningof the investigation and after a period of 3 years.The children identified with sleep disorders werecompared with those without sleep disorders.Forty-two percent of the children studied haddisturbed sleep at the initial interview, and 84%of these children had persistence of the sleepdisturbance after 3 years. The persistence of thesleep disorder had a significant relationship withthe increased frequency of stress factors in theenvironment. Other generalized behavior diffi-culties were present in 30% of sleep-disturbedchildren and 19% of non–sleep-disturbed children.The latter finding is not statistically significant.Cosleeping (sleeping with a parent or sibling)was noted more frequently in sleep-disturbedchildren (34%) than in non–sleep-disturbed chil-dren (16%). Twenty percent of the mothers atthe initial interview and 30% of the mothersat the 3-year followup perceived their children’ssleep disturbances as stressful to them andto their family life. They concluded that earlyidentification of the child with a sleep distur-bance and timely intervention would help boththe child and the family.

It is apparent that settling problems, night-time awakenings, or a combination of both seri-ously disrupt family life in many instances andlead to fatigue, irritability, the limitation of theparents’ activities, and, on occasion, maritalstrain.16 Night-time awakening is undoubtably

Disorders of Initiating and Maintaining Sleep Chapter 11 129

Ch11.qxd 01/28/05 10:30 AM Page 129

a trigger for child abuse on some occasions, andthe symptoms of difficulty settling and night-timeawakening should be taken seriously by the prac-titioner.7 Stress and depression, which are com-mon in parents with young children, are oftenmade worse by a lack of sleep. Parents who them-selves have not slept cannot think rationallyabout ways of coping with their child’s night-time awakenings.

Stress in the home is frequently associatedwith poor sleeping at night. Marital discord, sep-aration, divorce, financial or professional difficul-ties, parental affective illness, medical disorders,death, family, move, school entry or change, toi-let training, and the birth of a new sibling may allcontribute.1 A cry certainly is a powerful biologicsignal. It is indeed difficult to ignore, whether theresponse is positive or negative.

Although night-time awakening and settlingproblems have medical, biologic, and psychoso-cial underpinnings, some consider these symp-toms to be a conditioned reflex. Wilkes has statedthe following:

“The child wakes, cries, the mother intervenes withloving attention and a drink; the child repeats his cry-ing nightly and each time is quickly rewarded. Withsuch regular reinforcement, the reflex is soon estab-lished.”17

However, this proposition is too simplistic whenconsidering DIMS, since a combination of inter-actions among many variables contribute to thefinal common pathway of sleeplessness.

In contrast to sleep duration in older chil-dren, adolescents, and adults (which appearsto be related to mental and physical health statusand environmental and social factors), sleepduration in infancy seems determined primarilyby neurologic maturation and to some degreeby temperament. Weissbluth compared infanttemperament characteristics with total sleepduration and found that 4 of the 5 infant tem-perament characteristics (mood, adaptability,rhythmicity, and approach/withdrawal) that areused to establish temperament diagnoses of easyor difficult were highly negatively correlated withtotal sleep duration.18

There appear to be consistent patterns ofbehavior in babies. Some children are definitelymore quiet and less active than others. Therefore,a consideration of the baby’s personality may beimportant in the assessment of the sleep disor-der.3,19 Persistent sleep problems have been shown

to be associated with increased rates of tem-per tantrums and generally greater difficulty inmanaging the child’s behavior at 3 years of age.This suggests that persistent sleep problems area part of more pervasive behavioral difficultiesbetween the parent and child involving limitsand boundaries.14

Sensory thresholds, defined as the levels ofextrinsic stimulation required to evoke a dis-cernible response,3 may also be important inevaluating the child’s behavior and sleep prob-lems. A low sensory threshold may partiallyexplain problems such as teething and sleepless-ness. In Carey’s study, the distribution of thetemperament syndrome revealed that 76% wereeasy or intermediate low and 24% were interme-diate high or difficult. Of the 15 infants whoawoke at night, 13 had low sensory thresholds.3

There are two theoretical mechanisms of howlow sensory threshold contributes to night-timeawakening. First, a greater response to stimuliduring the day makes infants more arousableat night. Second, the infant may generally bemore responsive to internal and external stimuliat night.

However, great care must be taken in evaluat-ing infant temperament and sleep problems.Oberklaid and coworkers have shown thattoddler temperament ratings differ according tosex, age, social class, and cultural contexts. Futuretemperament norms may need to specify thecharacteristics of the group of children fromwhich they were derived to allow for more validcomparisons.20

After 3 to 4 months of age, a major featureof a child’s difficult temperament is the procliv-ity to cry and fuss in many situations.21 Infantsbeyond this age range who cry excessively alsotend to have irregular sleep or sleep patternscharacterized by frequent awakenings.22 Also,it appears that in the age range of 3 to 18 months,infants who cry more at any time during theday and night are also more likely to cry duringsleep–wake transitions.

COSLEEPING

Cosleeping (sharing a bed with another individ-ual) is common in many cultures. Concerns aboutthe suffocation of children and night-time awak-enings appear to be unfounded.7 It is often rec-ommended for the infant to sleep in a separateroom soon after birth to avoid contamination


Ch11.qxd 01/28/05 10:30 AM Page 130

of the sleep–wake transition with the need fora parent to be present in order to fall asleep.Parents are often concerned that putting thechild in a separate room at night may resultin feelings of inattentiveness to the child’s needs.However, Bax reports that parents often becomemore attentive to their child when the sleepingenvironments are separated. A cry from a distantbedroom easily wakes up the parent to respondto the child’s needs.7 Loudspeaker systemsand intercom units have been marketed to ensurethis arousal response in the parent. These tech-niques appear to conflict with the contentionthat cosleeping is not detrimental. However, it isinteresting to note that Bax reported 68% of thechildren with sleep problems to have experi-enced cosleeping, compared with 22% of thosewithout sleep problems.

The prevalence and correlations of sleeping inthe parental bed among healthy children between6 months and 4 years of age were studiedby Lozoff and colleagues.23 In this cross sectionof families in a large U.S. city, cosleeping wasa routine practice in 35% of white and 70% ofblack families. Cosleeping in both racial groupswas associated with approaches to sleep man-agement at bedtime that emphasized parentalinvolvement and body contact. Cosleeping chil-dren were significantly more likely to fall asleepout of bed and to have adult company and bodycontact at bedtime. Among white families only,cosleeping was associated with the older child,lower level of parental education, less profes-sional training, increased family stress, a moreambivalent maternal attitude toward the child,and disruptive sleep problems for the child.

Objective evidence of cosleeping resulting insleep disruption has been reported by Monroe.24

Couples who were good sleepers and habituallyslept together were studied polysomnographi-cally when they slept together and apart. Whenthe couple slept apart, more slow-wave sleep andless REM sleep were identified. This seems toindicate a rebound of slow-wave sleep similarto that seen in sleep-deprived subjects. It wasconcluded that sleeping together may be goodfor marital bliss, but sleeping apart eliminateddisturbances of sleep when partners change posi-tions. Similar studies have not been performedin children.

Breast-feeding may be associated with a con-current mother-infant adaptation that involves

frequent night-time awakenings. However, whena mother stops breast-feeding, the rate of sleepproblems among breast-fed children at 8 monthsof age is the same as that for children whosemothers never breast-fed.14 Breast-feeding andethnic differences may not be linked to persist-ent sleep problems because these characteristicsmay be associated with different child-rearingexpectations and behaviors.25 Night-time awak-ening during infancy is considered to be a nor-mal occurrence in some cultures, especially whenchildren sleep with their parents.

EVALUATION OF THESLEEPLESS CHILD

The evaluation of the sleepless child should beapproached in the same manner as that for anyother clinical problem. Evaluation begins with acomprehensive medical, developmental, andbehavioral history. The physical exam should becomprehensive, paying special attention to devel-opmental landmarks and the central nervoussystem (CNS).7

Parents will usually seek help for the problemonly if the child’s sleeplessness has resultedin parental sleeplessness or fear that the child isbecoming significantly sleep deprived. It is clearthat unless the clinician takes the initiative toobtain a comprehensive sleep history during thenormal course of well-child care and health main-tenance, problems will often be missed. Persistentnight-time awakenings have been frequentlyascribed to parental mismanagement of noctur-nal arousals. Appropriate attention should bepaid to the methods with which the parents haveresponded to their sleepless child in the past.The following questions may elucidate signifi-cant aspects of the patient’s history and provideinsight into sleep patterns and habits.

1. What is the normal evening bedtime?2. What specific activities are performed during

a period of 2 hours immediately before bed-time?

3. How long does it normally take the child tofall asleep?

4. If it takes the child longer than 30 minutesto fall asleep, what does the child do duringthe period of time from lights out to sleeponset? What are the parents’ responses to thesebehaviors?


Ch11.qxd 01/28/05 10:30 AM Page 131

5. What steps have the parents/caretakerstaken to assist the child in falling asleep?

6. Is the child permitted to fall asleep somewhereother than in his/her own bed and bedroom?

7. Exactly what associations does the childrequire in order to fall asleep (e.g., stuffedanimal, special toy, pacifier, bottle, beingheld, watching television, night-light)?

8. After sleep onset, does the child awaken?For how long? At what times? How manytimes per night?

9. If the child awakens at night, what interven-tions are required for the child to fall backasleep?

10. Does the child exhibit any unusual behaviorsor movements during sleep? Does the childwalk in his/her sleep? Bang his/her head?Wake suddenly with a piercing scream? Jerkhis/her legs? Grind his/her teeth?

11. Does the child snore? Is the snoring mild,moderate, or severe? Does the child snoreevery night?

12. What time does the child awaken in themorning? Does the child awaken sponta-neously, or is the child awakened by a clock?Parent? Sibling? Is it difficult to wake up thechild? How does the child feel on awaken-ing? Grumpy? Happy? Tired?

13. At what time during the day is the child themost active and alert? The most tired andcranky?

14. Does the child nap during the day? How manytimes each day? At what times?

15. Does the child exhibit any unusual behav-iors during daytime naps?

16. Where does the child nap? Are the same sleepassociations present for the child during day-time naps?

A sleep diary or sleep log maintained fora period of 2 weeks is of great assistance in ana-lyzing children’s sleep problems. The diary pro-vides more objective longitudinal evidenceof sleep patterns and a baseline from which tocompare the progress and success of treatmentregimens. By keeping the sleep diary duringthe course of treatment, clear evidence of changeis provided for the parents and practitioner.

Despite parental concern that their child isbeing sleep deprived because of difficulty set-tling and/or night-time awakening, this is usu-ally not the case. If the child remains active, alert,and playful during normal waking hours, the

volume of sleep during the previous 24 hoursis most likely appropriate.7 On the other hand,parental sleep deprivation is most often real.The treatment of the child’s sleep problem willmost likely result in resolution of the parents’sleep problems.

GENERAL CONCEPTS ON THEMANAGEMENT OF CHILDHOODDISORDERS OF INITIATINGAND MAINTAINING SLEEP

A variety of solutions to sleep onset and main-tenance problems have been suggested, andthere have been significant contradictory recom-mendations regarding optimal management.No single approach appears to be universallyeffective. Recommendations have included atten-tion to night-time rituals, firm handling, lettingthe child “cry it out,” the use of hypnotics andsedatives, general support and sympathy, and dif-ferent combinations of these.16 Many are appro-priate, but some are not. For example, Wilkes17

has described a method for deconditioningnight-time awakenings in one night.

“The child cries. Father stamps in, flings down theside of the cot, and slaps him once. But firmly, he flingsup the cot side, and stamps out, slamming the doorbehind him (locking it with older children). The childscreams and shouts for hour after hour until, depend-ing on his strength and determination, eventually hesubsides in a lather of sweat and exhaustion.”

Obviously, this method of treatment goesagainst most acceptable methods and may resultin numerous difficulties in addition to problemswith sleep.

The behavioral methods of treatment havebeen used successfully in the management ofchildhood psychological and behavioral prob-lems. However, a behavioral approach may notbe appropriate for all problems related to sleep-lessness. This approach requires careful analysisof the individual problem and the establish-ment and resulting gradual attainment of thespecific goals of treatment. Emphasis should beplaced on removing the factors that reinforcethe problem and planning a replacement orsubstitution with appropriate behaviors. Manysleep problems will resolve spontaneously,and time is always on one’s side in the manage-ment of many sleep disorders.7 Simple behavioral


Ch11.qxd 01/28/05 10:30 AM Page 132

procedures may occupy the parents’ attentionlong enough that a spontaneous remission mayoccur in the interim.

Night-time awakenings and settling prob-lems are very often interactional in nature,affecting the entire family and not merely thechild. Therefore, the management of the prob-lem should be directed at the family and not justthe child. In many circumstances, the bestmethod of treatment is habit training (behaviormodification) and attention to the principlesof sleep hygiene (Table 11–1). If there is failureto respond, some recommend the short-term useof sedation.3 It is thought that a short-termcourse of a hypnotic or sedative may alter theinfant’s arousal threshold enough to facilitatethe learning of a more acceptable sleep pattern.However, there are many conflicting opinions,and it is thought by many that medication rarelyhas a place in the management of most sleepdisorders.

Changes in the physical sleeping environmentmay cause dramatic changes in sleep patterns.Children who experience a persistent problemwith sleeplessness may sleep well at a grandpar-ent’s home only to have the symptoms return intheir own home environment. This is frequently

irritating to young parents, who may feel inse-cure about their role.7 Appropriate support andmanagement of the entire family are essentialduring these times.

According to Bax,7 the short-term use of chlo-ral hydrate or Phenergan may be necessary. Inthese situations, it is important to start withan adequate dose and be weaned quickly fromthe medication. This method seems more effica-cious than starting with small doses and increas-ing them until the desired effect is achieved. Thelatter method may be accompanied by increasingtolerance to the medication, and larger dosesmay be required.

During any behavioral management regimen,parents must be warned that the child’s sleepproblem may appear to worsen for a few daysat the beginning of treatment. This characteristicfrequently results in the abandonment of the treat-ment protocol if there is no adequate warning,support, and reassurance. When there is improve-ment, it tends to happen quickly.

Jones and Verduyn16 reported that 53% of chil-dren’s sleep problems resolved completely bymeans of a behavioral approach. Thirty-sevenpercent more showed partial resolution, and10% were unchanged at the end of the treatment


The child’s bedroom should be dark and quiet.Bedtime routines should be strictly enforced.The time of morning awakening should be firmly and consistently structured.Bedroom temperatures should be kept comfortably cool (<75° F).Environmental noise should be minimized as much as possible. Background music and/or white

noise may inhibit extraneous noise.Children should not go to bed hungry and may have a snack before bedtime.Excessive fluids before bedtime may result in bladder distention and arousal and disrupt sleep.Children should learn to fall asleep alone (i.e., without their parents’ presence in the room).Vigorous activity should be avoided before bedtime.A bath is often a stimulating activity for children. If bedtime struggles are present after the child’s

bath, it may be moved to the morning or separated from the child’s bedtime by at least 2 hours.Methylxanthine-containing beverages and food (e.g., caffeinated beverages/colas, tea, chocolate)

should be avoided for several hours before bedtime.Parents should read the labels on all over-the-counter and prescription medications. Some

medications contain alcohol or caffeine and may disrupt sleep.Naps should be developmentally appropriate. Brief naps may be refreshing, but prolonged naps

or napping too frequently may result in significant accumulation of sleep during the day andmake it more difficult to fall asleep at night.

Modified with permission from Sheldon SH, Spire JP, Levy HB: Pediatric Sleep Medicine. Philadelphia, WB Saunders,1992, p 76.

Table 11–1. Principles of Sleep Hygiene in Childhood

Ch11.qxd 01/28/05 10:30 AM Page 133

regimen. Maternal psychiatric history was notrelated to the outcome. The success of sleepingthrough the night remained unchanged at6-month followup. It is important to note thatthe problem was less likely to resolve when mar-ital discord was present. Response was bestwhen both parents were persuaded to attendsessions with the physicians. Consistency and per-sistency are imperative in management, andcooperation between the parents during treat-ment regimens will result in the more rapid andcomplete resolution of symptoms.

BEHAVIORAL AND/ORCONDITIONED FACTORS CAUSINGSLEEPLESSNESS

Disorder of Sleep OnsetAssociationsIt has been pointed out in Chapter 1 that briefnocturnal awakenings are normal, especiallyduring transitions between REM and non-REMsleep. These brief arousals may serve an impor-tant survival function. The sleeping environ-ment may be checked and the body repositioned.Bedclothes may be adjusted to the sleeper’s com-fort. The return to sleep is most often rapid, andthe short arousal is not remembered. However,if the sleeping environment has changed, return-ing to sleep may be difficult and the individualmay become fully alert.

Sleep onset occurs through complex behav-ioral and physiologic interactions. First, thebody must be physiologically ready for sleepat its appropriate position in the circadian cycle.A series of bedtime rituals follow. On retiring,a pillow might be fluffed and covers or blanketspositioned. Other factors may also be required,such as a night-light, soft music, television, and/orthe presence of a bed partner. Without thesebehavioral rituals, sleep onset may be difficult.In other words, we learn to fall asleep in a par-ticular manner, and this learned behavior isrepeated night after night. When these sleep onsetassociations are absent, falling asleep may betroublesome.

Children learn to fall asleep in the same man-ner. Certain conditions must be present to easethe transition. The child may require a particularbedroom, lying in a certain crib or bed, and hold-ing a favorite stuffed animal or special blanket.1,26

These conditions are usually present all night.Being held and rocked, the presence of televi-sion, being read a story, and pacifier sucklingmay not be present all night. After a normalarousal, it may be hard for the child to returnto sleep, since the child has not learned to fallasleep on his/her own without these associa-tions. Ferber persuasively states the importanceof these associations as follows:

“Thus the problem is not one of abnormal wakingsbut one of difficulty in falling back to sleep. And thedifficulty arises because of the child’s particular asso-ciations with falling asleep.”26

A disorder of sleep onset associations occursmost commonly in infants and toddlers. Theolder the child, the more control s/he has overthe environment and the less likely s/he is torequire associations that are not present withinthe sleep environment during nocturnal arousals.Association problems become less significantafter the age of 4 years, although they may occa-sionally persist into adolescence.

Clinical Presentation and Diagnosis

The usual complaint is one of prolonged night-time awakenings. Settling may not be of concernto the parents, since associations may be presentearly in the evening. During the night, the childmay fuss significantly until s/he is removed fromthe crib or bed, permitted to sleep in the parents’bed, held, or rocked. Daytime sleepiness is mostoften not present, although the parents may besignificantly sleep deprived from chronicallyresponding to their child’s nocturnal crying. Thetime of settling is usually normal. The reviewof a sleep diary usually reveals normal sleeponset latency time and frequent, prolonged noc-turnal arousals. Total sleep is usually normal,the timing of the major sleep period is normal,and the morning wake-up time is usually nor-mal. Parents often report that the child rapidlyreturns to sleep if given a bottle. The volumeof fluid consumed by the youngster during thesenocturnal feedings is usually small. For the breast-fed infant, sleep returns quickly once nursingbegins and the nocturnal feedings are perceived bythe mother to be different from the daytime feed-ings. These characteristics help differentiateassociation problems from disorders of excessivenocturnal fluids. Physiologic processes controlling


Ch11.qxd 01/28/05 10:30 AM Page 134

the sleep–wake cycle can be assumed to be func-tioning normally if the child sleeps well whilebeing held or rocked. If there is significant dys-function of the central mechanisms controllingsleep, the child will not sleep well under anycircumstances. The physical exam is usually nor-mal. A rapid return to sleep when habitual asso-ciations are reestablished helps differentiatenormal from abnormal arousals and determinethe diagnosis of disordered sleep onset associa-tions as the problem.

Treatment

Treatment is straightforward. The child mustlearn to make the transition from wakefulnessto sleep without expecting participation of theparent. Success rates are high if parents are givensufficient support. Relearning usually takes lessthan 1 week.26

The child must learn to fall asleep alone andunder conditions that can be easily reestablishedafter normal nocturnal arousals. Bedtime ritualsshould not be stimulating or require ongoingactivity or parental participation. Holding, rock-ing, nursing, and pacifier suckling should be dis-continued. The child should be placed in thecrib or bed alone and should learn to fall asleepby himself/herself.

Two basic principles must be followed. Theparents must be consistent in reestablishing appro-priate sleep onset associations and persistentwith the regimen. The following paradigm isusually rapidly successful if the parents are con-sistent and persistent. For some families, sleep-less nights may seem worse for the first dayor two of treatment. Therefore, the treatmentregimen should begin on days when parentalsleeplessness will not affect their performancethe next day. Although some degree of cryingshould be expected during the beginning of thetreatment phase, it is usually kept to a minimumby gradually establishing appropriate sleep onsetassociations. Letting the child “cry it out” willusually maintain crying at its maximum andoften results in parental frustration and treat-ment failure.

1. The first night

● The child should be placed in the crib orbed with only those association objects thatwill be present during normal nocturnalawakenings.

● The room should be dark (although a smallnight-light may be used), quiet, and at a com-fortable temperature. A room that is too hotor cold will tend to inhibit sleep. When theroom temperature cannot be controlled,the child should be dressed appropriatelyfor the ambient temperature.

● The parent may soothe and comfort the childuntil s/he is lying down quietly. Allowingsome crying will rarely result in psycholog-ical trauma to the child and is often moredifficult for the parent.

● Once the child is quiet in the bed or crib,the parent should leave the room.

● If the child begins to cry, the parents shouldallow it for a short period of time. The exacttime should be developmentally appropriate;however, a period of at least 1 to 2 minutes(according to a clock or a stopwatch) shouldelapse before the parent first responds.

● The parent may then return to the roomto comfort the child; however, the childshould not be removed from the crib or bed.The parent may remain in the room quietlysoothing the child until s/he lies back downin the crib or bed and is quiet, at which timethe parent should leave the room.

● If and when the child begins to cry again,the parent should wait slightly longerbefore responding (a period of 2 to 3 min-utes is usually appropriate) and repeat theimmediately preceding step.

● The above process should be repeated (keep-ing the wait–response period at 2 to 3 minuteson the first night) until the child is sleeping.

● On the first night, this process of cry–wait–respond may last for several hours before thechild falls asleep. Parents must be warnedthat this might occur. It is important that theybe supported and reassured that the processwill be successful if they are persistent andconsistent with the regimen. If they give upand remove the child from the crib or bed onthe first or second night of management, atreatment failure is likely. It is also importantto involve all caretakers in the treatment pro-gram so that sharing responsibilities mayoccur and consistency be assured.

2. The second night

● Treatment on the second night is similar tothat on the first; however, the wait–response


Ch11.qxd 01/28/05 10:30 AM Page 135

period may be lengthened to 2 to 4 minutesfor the first and 5 minutes for subsequentwaiting times. If the child is beginning tocalm down at the end of the wait–responsetime, it is usually better to wait slightly longerto see if the child settles on his/her own.

● Parental interventions should be supportive.The child should know that the parent isunderstanding and is nearby .

● Parents must not exhibit anger or frustra-tion or escalate these feelings as the nightprogresses.

3. Subsequent nights

● Treatment on subsequent nights shouldparallel that on the first two nights.

● However, by the third night improvement isusually seen. Night-time awakenings areusually shorter, loud crying may be replacedwith mild whimpering, and a return to sleepwithout parental intervention occurs.

● Nocturnal arousals may still occur, but cry-ing does not. The child has learned to fallback asleep on his/her own without parentalassistance.

During the course of management, parentsshould record progress in a sleep diary. Thisis important for the parents and practitionerso that improvement can be documented. Duringthe course of treatment, follow-up visits shouldoccur at weekly intervals. The sleep diariesshould be evaluated at each visit. Once the childis falling asleep on his/her own, s/he will mostoften continue to sleep well. Occasional disrup-tions might occur, especially during times whenthe usual regimen is altered (e.g., vacations,birthdays, holidays). The parental responseto and management of these intermittent sleepdisruptions will determine whether or not theywill persist. If the disruption is managed ina manner consistent with that for the originalregimen, rapid resolution will occur and thechild will continue to sleep well.

If the mother is breast-feeding the child, it isnot necessary to discontinue nursing to correctthe sleep onset association. It is necessary onlyto dissociate the act of nursing from fallingasleep. Association problems are most likelypresent when the infant falls asleep during noc-turnal breast-feedings after nursing for onlya few moments. Instead of nursing at bedtime,it is often better to recommend moving the last

feeding of the evening to a time several hoursearlier. If the child tends to fall asleep duringbreast-feeding, the mother may discontinue thefeeding before the child falls asleep and place thechild in the crib or bed. The same techniquesmay be used during daytime naps.

Excessive Nocturnal FluidsAlthough this condition is considered under therubric of behavioral etiologies, excessive noctur-nal fluids may have a physiologic basis for caus-ing sleeplessness. In older children and adults,bladder distention typically results in an arousalresponse (see Chapter 27). In children with pri-mary functional nocturnal enuresis, this arousalresponse appears to be developmentally delayed.Other evidence that bladder distention has a cen-tral arousal effect stems from observations thatparasomnias (disorders of sleep stages and par-tial arousal from sleep) appear to be exacerbatedby bladder distention.


Infants consuming excessively large quantities offluid during the night typically awaken continu-ously and frequently. This problem occurs in bothbreast- and bottle-fed infants. The volume of fluidconsumed during the night may range from 8to 32 ounces, and these children awaken from3 to 8 times per night.27 The diapers are usuallyheavily soaked by morning.

From 7 to 12 months of age, infants no longerhave a physiologic need for feeding during thenight. Awakening to eat may be secondary to oneof three major factors: there may be an associa-tion of feeding with sleep onset and returningto sleep; there may be learned hunger during thenight; or bladder distention may be present andcausing the arousal. Children awakening and feed-ing due to inappropriate sleep onset associationstypically awaken less often than those awaken-ing due to excessive nocturnal fluids. Fluidintake during arousals is also less. The presenceof only one or two arousals per night, witha rapid return to sleep after only a brief periodat the bottle, breast, or pacifier, suggests that thenipple or parent is more important than the fluidintake. If sleep onset association problems arepresent, the fluid intake is usually less than6 ounces per night. More frequent awakeningsand the consumption of large volumes of fluid


Ch11.qxd 01/28/05 10:30 AM Page 136

(8 to 32 ounces) suggest excessive fluid intake asthe problem. Breast-feeding more than two timesper night and nursing sessions of longer than2 to 3 minutes suggest excessive fluid intake.Finally, if the child has learned to eat during thenight, frequent nocturnal awakening will occur.Habitual nocturnal feedings may distort the cir-cadian hunger rhythm and disrupt the sleep–wakecycle as much as the fluid/food intake.

Treatment

Although nocturnal awakenings are typically fre-quent and severe in patients ingesting excessivenocturnal fluids, treatment is straightforwardand not difficult. The aim of therapy is to gradu-ally discontinue the nocturnal fluids. For thebottle-fed infant, gradually decreasing the vol-ume of fluid in each nocturnal bottle over thecourse of 1 to 2 weeks will usually result in rapidimprovement. Often infants will discontinueawakening for feeding when only 1 or 2 ouncesare provided. One alternative is to graduallydilute the formula with water before weaningfrom the bottle. However, it is important in eithercase that during the treatment phase, the associ-ation of sleep onset with suckling not be estab-lished. If this is the case, the regimen should bemodified to include the principles discussedin the section on the treatment of sleep onsetassociation disorder. Treatment of the breast-fedinfant ingesting excessive nocturnal fluids isslightly more problematic. Nursing mothers mayexperience a letdown on hearing their cryinginfant. It may be necessary for the nursing motherto manually express her milk (or use a breastpump) and have the father or other caretakerrespond. Diluting the mother’s milk with waterand gradually weaning the child from the bottleare then conducted in the same manner as thatfor formula-fed infants. Once the infant is sleep-ing through the night, maternal letdown duringsleep will usually resolve.

ColicColic is one of the more common disordersaffecting sleep in young infants, affecting aboutone in five.28 The patient is usually less than4 months of age, and the typical complaint is oneof inconsolable fussiness during late afternoonor evening hours. Symptoms of colic will often

resolve, but disordered sleep remains. The prob-lem does not appear to be biologically based,although it may reflect altered chronophysio-logic factors during development. The sleep prob-lem more likely occurs in response to alteredsleep–wake schedules and habitual patternsof parental responsiveness persisting into thepostcolic period.1

Weissbluth has stated that colicky behaviorappears to be a reflection of the developmentof the CNS during the first few months of life.28

Subsequently, some colicky infants developirregularity of behavior, heightened activityor arousal, and sensitivity to stimuli that cause,directly or indirectly, difficulty in maintainingregular, prolonged, and consolidated sleep pat-terns. Mismanagement of the infant’s sleepschedule and habits after the colicky period haspassed, a result of early difficulties in handlingthe infant, is thought to be the most commoncause of persistence of sleep problems after4 months of age.


Patients appear to suffer violent, rhythmic, scream-ing attacks that are unresponsive to parentalintervention and for which no other cause couldbe found.29 Wessel and colleagues described col-icky infants as those who experience paroxysmsof irritability, fussing, or crying lasting for a totalof more than 3 hours a day, occurring more than3 days per week, and continuing or recurring formore than 3 weeks.30

The age of onset, time of occurrence, and ageat the resolution of symptoms are characteristic.“Attacks” generally do not occur during the firstfew days of life. Symptoms typically begin dur-ing the third (80%) and fourth (100%) weeks oflife.30 An important exception is the prematureinfant. Three studies have documented the onsetof colic in the premature infant within 2 weeksafter the infant’s expected date of birth (70%between the 39th and 44th gestational weeks)regardless of the gestational age at birth.31-33

Once the symptoms begin in the preterm infant,the duration of colic appears to be similar to thatof term infants. Thus, the ages of the onset anddisappearance of colic appear to be locked intime to conceptional age. Episodes of colic tendto start between 5:00 and 8:00 PM and endat about midnight. By 2 months of age almost


Ch11.qxd 01/28/05 10:30 AM Page 137

half of the infants suffering from colic will havea resolution of symptoms, which will subside in90% of all infants with colic who have no moreattacks by 4 months of age.

Another characteristic feature of colic is rela-tively uninhibited motor activity. During the cry-ing spells, affected infants have been describedas hypertonic or neurolabile. This does not occurat other times during the day or night. Physicalmovements during the colicky spell includetonic stiffening of the entire body with fiststightly clenched; legs flexed rigidly over theabdomen; writhing, twisting, and turning motions;and jerky, uncoordinated movements such asbatting or flapping of the arms and kicking ofthe legs. Facial grimacing suggests severe pain,and affected infants seem unusually sensitiveto light.34

Daytime sleep periods for colicky infantsappear to be extremely irregular and brief. Someinfants will temporarily discontinue daytime napsduring the period of maximum fussiness (approx-imately 6 weeks of age). Colicky infants miss peri-ods of quiet sleep during their evening attacks.28

It is interesting to note that after the cessationof daytime and night-time spells, the infantinvariably falls asleep. This may be secondary toexhaustion or the time of day or could representa period of transient but excessive neurologicarousal, excitation, or lack of inhibition that isterminated by naturally occurring periods of quietsleep or increased inhibition.28

A number of studies have evaluated the sleeppatterns of infants in the postcolic period.35-37

The parents of infants who have had colic reportshorter total sleep duration at 4 to 5 months ofage. A history of colic appears to be significantlyassociated with frequent night-time awakeningswhen compared with infants with no history ofcolic (a ratio of almost 2.5 to 1). The durationof night-time awakenings is a problem for someinfants, as is frequency and duration in others.

Some postcolic infants are exquisitely sensi-tive to irregularities in their sleep–wake sched-ules. Disruptions of sleep–wake routines becauseof illness, holidays, or vacations cause extremedisruptions of settling and night-time awak-enings that may last several days. According toWeissbluth, these prolonged recovery periodsmight reflect easily disorganized, endogenousbiologic rhythms caused by enduring congenitalimbalances in arousal/inhibition of sleep–wake

control mechanisms.28 This suggests that disor-dered chronophysiologic mechanisms may besignificant in postcolic sleep–wake disturbances,and treatment should focus on mechanisms thatmay functionally reset and entrain the endoge-nous circadian pacemaker.

Treatment

The clinical observations of infants who sufferfrom frequent night-time awakenings and shortsleep period times in the postcolic period may besuccessfully treated only if the parents establishand maintain a regular sleep schedule. It seemsthat most postcolic sleep problems are the resultof the parents’ failure to establish regular sleeppatterns when the colic dissipates at approxi-mately 4 months of age. Therefore, treatmentshould focus on the reestablishment of a normalsleep–wake routine. If the parents strictly pro-gram their child’s sleep schedule, regular day-time and night-time sleep patterns will re-emerge. The most powerful point of entrainmentis the morning wake-up time, which must be fixedand consistent. Nocturnal bedtime and naptimes must also be strictly adhered to. Consistencyand persistency are essential in establishing asuccessful regimen. The maintenance of a sleepdiary for 1 to 2 weeks before instituting therapyand during the course of treatment will greatlyassist parents in maintaining an appropriateroutine and provide accurate feedback for thepractitioner regarding the success of the regimenduring follow-up visits.

MEDICAL DISORDERS RESULTINGIN SLEEPLESSNESS

Cow’s Milk AllergyAn allergy to cow’s milk protein may result ina severe disturbance of sleep during earlyinfancy. It is often difficult to clinically differen-tiate cow’s milk allergy from colic, since both maybegin at similar ages and are associated withsleeplessness, fussiness, intermittent cryingepisodes, and short sleep period times.


Frequent night-time awakenings (five or moretimes per night) and short total sleep times (often


Ch11.qxd 01/28/05 10:30 AM Page 138

as low as 4.5 hours) are the typical sleep com-plaints.1,38 Infants with cow’s milk allergy oftencry frequently during daytime hours and aredescribed by their parents as fussy. The physicalexam may be unremarkable; however, anemiaand hematochezia may be present in someinfants. Polysomnographic analysis reveals a sig-nificantly disturbed sleep pattern and is usefulin determining the absence of other causes ofarousal and short total sleep times. The diagno-sis is based on a high clinical index of suspicionof atopy. Allergy testing usually reveals elevatedimmunoglobulin E levels, and radioallergosor-bent testing is often positive for cow’s milk pro-tein. Discontinuing a cow’s milk–based formulaand substituting a hydrolyzed milk protein for-mula result in the resolution of symptoms.

Treatment

Once the diagnosis is established, treatmentconsists of the removal of cow’s milk proteinand the beginning of hydrolyzed milk-proteinformula. Within 2 weeks of dietary modifica-tions, sleep patterns will generally normalize.A decrease or complete resolution of daytimesymptoms occurs, total sleep time increases toa developmentally normal level, and nocturnalawakenings resolve. Reintroduction of cow’smilk protein formula will result in the exacerba-tion of symptoms. At times, exacerbation is severeand clinical judgment should determine whetherreintroduction of the responsible protein is pru-dent. In general, it should be done under onlyrare circumstances and only under highly con-trolled conditions.

Otitis MediaAny condition that results in pain, discomfort,or fever may be associated with sleep disruptionand sleeplessness. Otitis media is one of themost common childhood illnesses presentingto the practitioner, and acute middle ear diseaserarely goes unrecognized. The cause of thechild’s sleep complaint is often readily apparent.In contrast, chronic middle ear disease is oftenpresent despite a paucity of clinical symptoms.Serous or secretory otitis media associated withpersistent middle ear effusions may be clinicallyasymptomatic except for the presence of disruptedsleep.


Children with acute suppurative otitis media oftenpresent with fever, otalgia, changes in appetite,and vomiting. Frequent and prolonged nocturnalarousals, associated daytime sleepiness, and a sig-nificant decrease in daytime activity often occur.The physical exam reveals a bulging tympanicmembrane, with the loss of normal landmarks andimmobility of the drum on pneumo-otoscopyand/or tympanography.

Chronic serous or secretory otitis media isoften associated with only a few symptoms.Hearing may be decreased; however, this maynot be clinically apparent. Complaints of otalgiaare often absent. The physical exam may reveala retracted, immobile tympanic membrane.Evidence of eustachian tube dysfunction is oftenpresent, and air and/or fluid levels may be visu-alized. Regardless of the absence of clear clinicalcomplaints during the day, sleep may be signifi-cantly disrupted.

Treatment

Treatment of the sleep disturbance associatedwith acute or chronic middle ear disease firstfocuses on the adequate management of theunderlying pathology. Appropriate courses ofantibiotic therapy for acute suppurative otitismedia should result in the resolution of sleepcomplaints. Treatment of chronic serous orsecretory otitis media and persistent middle eareffusion has included tympanocentesis, tympan-otomy tube placement, chronic antibiotic therapy,and/or a combination of antibiotics and a shortcourse of steroid treatment. Resolution of theeffusion will often result in the concomitant res-olution of nocturnal symptoms. During thecourse of medical and/or surgical therapy, appro-priate attention must be paid to the principlesof sleep hygiene. Sleep schedules must be regu-larized and parents should adhere to strict pat-terns of age-appropriate sleep–wake timetables.

Neurologic DisordersNeurologically impaired children frequentlyawaken at night and exhibit disordered sleep–wakeschedules. It has been thought that this sleepdisruption may be due to chronic cerebral irrita-tion; however, sleeplessness and night-time


Ch11.qxd 01/28/05 10:30 AM Page 139

awakening are also reported by the parents ofchildren with significant handicaps that arenot associated with cortical irritability. Childrenwith static encephalopathies, noncortical blind-ness, and deafness may exhibit significant disor-ders of sleep–wake cycles.


Sleep problems in children with neurologic dis-orders may be due to various underlying etiolo-gies. Indeed, a child with obvious neurologicimpairment may also exhibit symptoms of sleep-lessness secondary to any of the other non-neurologic conditions discussed in this chapter.Therefore, the assessment of the child’s sleepproblem must include a comprehensive evalua-tion without prematurely attributing the disorderof sleep to the prominent neurologic condition.All factors that may result in sleeplessnessshould be considered.

The primary neurologic diagnosis is oftenclinically obvious. However, the characteristicsof the sleep problem may be more obscure, andmaintenance of a sleep diary is extremely impor-tant in evaluating the sleep–wake cycle of theneurologically impaired child. Reasonable datamay be collected to differentiate a primarilybehavioral etiology from one secondary toimpairment of the neurologic mechanismsresponsible for controlling sleep–wake cycles,sleep onset, and/or sleep maintenance. The med-ications used for the management of the princi-pal neurologic condition may themselves beresponsible for disordered sleep. Nocturnalpolysomnography is often helpful in determin-ing a central origin of the sleep disorder and willassist in detecting nocturnal seizure activity thatmay not be apparent during the day. The diagno-sis of an organic basis for disordered sleep isexclusionary. Only after the comprehensive eval-uation of other possible mechanisms has failedto reveal the etiology should neurologic mecha-nisms be considered causative.

Treatment

Treatment of sleeplessness in the neurologicallyimpaired child is dependent on the identifiedunderlying cause. Attention should be paid firstto all the factors that may influence sleep and thesleep–wake cycle, other than the neurologic

disorder. If a behavioral or parent managementproblem is suspected, it may be handled in themanner previously described. However, it mightbe necessary to conduct the therapeutic regimenmore slowly than it would be in the non-neurologically impaired child. Appropriateattention should be paid to the child’s sleep–wakeschedule, especially in children with special sen-sory disorders such as blindness. The correctionof these schedule abnormalities is usually notdifficult and will result in the rapid improvementof sleep patterns and daytime function.

If the medication used in the treatment of theprincipal neurologic disorder is considered to beresponsible for the sleep problem, attentionshould be paid to the dosage and timing ofadministration of the drug. Windows of timeexist in which a particular medication may sig-nificantly affect circadian timekeeping mecha-nisms. Administration within the window willresult in significant shifts of circadian rhythm,and administration of the same medication anddosage outside of the window will result in littlechange in pacemaker timing. Therefore, a shiftin the time of administration of prescribed med-ications may greatly change the child’s sleep–wakepatterns. Under certain circumstances, thedosage of the medication may be modified toimprove the child’s sleep. In other situations,though less commonly, specific medications mayrequire discontinuation with substitution ofan alternative drug for improvement in sleepto occur.

In situations in which the underlying neuro-logic abnormality is considered the cause ofsleeplessness, attention should still be paid toappropriate sleep hygiene. If an improvementin reduced sleeplessness occurs with mild med-ications, such as diphenhydramine, the childwill most often show similar improvementwithout the use of drugs. Hypnotic or sedativemedication is occasionally required. Chloralhydrate is often effective in the treatment ofsleeplessness in children with neurologicor special sensory impairment. If medication isnecessary, it is usually best to use adequatedosages. It is more appropriate to begin witha dosage schedule that is significant enoughto control the symptoms rather than beginningwith a small dose and slowly increasing it untila response occurs. However, in the neuro-logically impaired child, the development of


Ch11.qxd 01/28/05 10:30 AM Page 140

medication resistance occurs less commonlythan in children without neurologic disorders.Once medication is instituted, it is important tomonitor the child’s daytime function to ensureimprovement. Periodic withdrawal of medica-tion is also recommended. Once the symptomsof sleeplessness are controlled, it may be possi-ble to discontinue the medication withoutexacerbation of symptoms.39

Attention Deficit HyperactivityDisorderThe parents of hyperactive children considertheir children to have many more sleep problemsthan the parents of children without hyperac-tivity do.40 Most commonly, night-time awaken-ings and restless sleep are reported. In a studyby Salzarulo and Chevalier, the parents of chil-dren with attention deficit hyperactivity disor-der (ADHD) reported sleep onset problems16.5% of the time and night-time awakenings39% of the time.41 The Diagnostic and StatisticalManual of Mental Disorders (Third Edition)lists motor restlessness during sleep as one ofthe defining characteristics of hyperactivity,42

even though support from sleep studies isambiguous and controversial results do exist.43,44

It is most likely that hyperactivity and attentiondeficits are symptoms of a heterogeneous groupof disorders with varied etiologies, manifestingincreased daytime activity and disordered sleep.For example, children with obstructive sleepapnea syndrome manifest daytime symptomsthat are virtually indistinguishable from thosewith ADHD on clinical grounds alone. Attentionspan problems may reflect the daytime resultsof sleep deprivation with the appearance ofmicrosleep episodes. It is obvious that furtherinvestigation is necessary to determine whetherdysfunctional sleep is a cause or effect of ADHDin some children.


The diagnosis of ADHD may be difficult. Thesechildren tend to be fidgety, have difficulty stay-ing on and completing tasks, often disturb otherchildren in school, are easily distracted, oftencry easily, and have rapid mood swings. They mayexhibit restlessness and increased activity. Theyare often easily frustrated in their efforts and

may become destructive. The physical examis often normal, and the child may not exhibitincreased activity during the clinical evaluation.Soft neurologic signs may be present in somechildren; however, the significance of these find-ings is questionable. Visual tracking may be poorand speech dysfluency may be present. Letterreversals on writing from dictation, dysdiado-chokinesia with significant overflow-associatedmovements, difficulty hopping and skipping,and right/left confusion have also been reportedwith increased frequency in children withADHD.45,46

Night-time awakenings and restless sleep arecharacteristic sleep complaints. However, theredoes not appear to be a difference in sleep-onsetlatency or total sleep time between children withADHD and normal children. Enuresis and nightsweats have also been reported more frequentlyin hyperactive children than in control subjects.40

Differentiating the sleep disruption of ADHDfrom that of obstructive sleep apnea may be dif-ficult. The parents of virtually all children withobstructive sleep apnea syndrome will admit thatsignificant snoring is present. Daytime symptomsof hyperactivity may also alternate with periodsof somnolence in children with obstructive sleepapnea syndrome. Traditional nocturnal polysom-nography may or may not be helpful. If signifi-cant sleep-related airway obstruction is present,the polysomnogram may reveal apneas, hypo-pneas, oxygen desaturation, and arousals.However, in many instances the results of non-invasive sleep studies are inconclusive. Despiteincreased respiratory resistive load during sleep,clear-cut apneas and hypopneas may be absent.Diagnosis would then require more invasivetechniques, such as continuous monitoring ofintrathoracic pressure through an indwellingesophageal balloon manometer. Noninvasivetechniques for continuous monitoring of intra-thoracic pressure during sleep are currently beinginvestigated.

Treatment

The appropriate treatment for ADHD is contro-versial and beyond the scope of this text.Stimulant medication, counseling, and behaviormodification are the most widely acceptedmethods of therapy. Sleep problems tend toimprove along with daytime symptoms despite


Ch11.qxd 01/28/05 10:30 AM Page 141

the use of stimulants that often result in sleep-lessness in children without ADHD. At thepresent time, how the medication functions toimprove the daytime and night-time symptomsof ADHD is obscure. It is interesting to specu-late on the mechanism of this paradoxical effect.First, stimulant medication may provide enoughcortical stimulation to result in an alertingresponse. Children may then be more capableof focusing attention on the task at hand. On theother hand, if microsleep episodes are presentand responsible for attention span deficits andconcomitant symptoms, stimulant medicationmay decrease the frequency or eliminate thesemicrosleep episodes, resulting in better atten-tion and focus. If microsleep episodes are fre-quent and sleep begins to accumulate duringdaytime hours, nocturnal sleep may be disrupted.Administration of stimulant medication maythen result in the reduction in frequencyor elimination of microsleep episodes. Sincedaytime sleep accumulation does not take place,nocturnal sleep improves. Though engaging,these explanations are highly speculative andawait documentation.

MedicationsVirtually any medication may cause sleep dis-ruption. Hypnotics are the most commonly pre-scribed drugs in the United States. It has becomeapparent that this classification of medicationhas generally resulted in more significant exac-erbation of sleep problems than cures. Hypnoticsand sedatives are often inappropriately used inchildren. Rarely will the use of hypnotics resultin long-term improvement of sleep in childrenwho are otherwise normal. If there is improve-ment initially, a behavioral approach and atten-tion to appropriate sleep hygiene most likelywould have resulted in the resolution of symp-toms as well without the risk of side effects. Themost commonly prescribed medications in child-hood are antihistamines (which may themselvescause sleeplessness), major sedatives (such aschloral hydrate and phenobarbital; which maythemselves cause paradoxical hyperactivity),and short-acting benzodiazepines. These classesof medication may also adversely affect the child’sperformance the next day.

Relatively innocuous medications prescribedfor acute or chronic illness may also be responsi-

ble for sleeplessness. Antibiotics, especially liquidpreparations, have been associated with DIMS.39

It is thought that the vehicle and not the antibi-otic itself is responsible for the sleep disruption.Over-the-counter medications, especially com-bination drugs, may also be implicated. Oralbronchial smooth muscle relaxant medicationsoften cause sleep disruption, although the exactmechanism is unclear.


A comprehensive medication history shouldbe taken from the parents. It is important toinclude over-the-counter medication, since thevehicle may be responsible for sleep disruption.If the child is taking medications, it is importantto determine their dosage and timing of admin-istration. The history may suggest the onset ofthe sleep problem commensurate with the insti-tution of medication. Most medications do notresult in any typical patterns of sleep disruption.Nocturnal settling and awakening problemsmay be present. Hypnotics, sedatives, and neu-roleptics may cause significant changes in sleeparchitecture, and nocturnal polysomnographymay be useful in assisting in the diagnosis.Depending on the medication, total sleep timemay be decreased or shifts in sleep–wake sched-uling problems may occur. Maintenance ofa sleep diary for a period of two weeks may beof great value in accurately documenting thechild’s typical circadian sleep–wake rhythm.

Treatment

If possible, the suspected offending medicationmight be discontinued. If this is not possible,modification of the timing of administrationand/or dosage may be attempted. Switching toa similar medication or the same medicationprepared differently may be successful (e.g.,changing from oral bronchodilators to inhaledpreparations). At times, merely changing brandsof medication will result in the resolution ofthe sleep problem.

Chronic IllnessAny chronic condition may contribute to per-sistent sleep problems. Pain or discomfort fromthe illness or from treatment regimens may be


Ch11.qxd 01/28/05 10:30 AM Page 142

contributory. Disorders such as migraine cephalgia,asthma, diabetes mellitus, gastroesophagealreflux, and seizures have all been associated withsleep disturbances. The problem of sleeplessnessmay be directly caused by the underlying disor-der or may be an indirect consequence of ther-apy, medication, or anxiety.


The underlying disturbance may be readily appar-ent or may be obscure. For example, a childwhose sleep is disrupted because of pruritisassociated with chronic eczematous dermatitiswill most frequently have easily diagnosablesigns. On the other hand, a sleepless child awak-ening from pain secondary to esophagitis andchronic gastroesophageal reflux may exhibitonly a few findings. Likewise, chronic serousor secretory otitis media may present with only afew subjective symptoms.

However, it is often difficult to sort out whichfactors related to the illness are precipitatingthe sleep problems. Confusion regardingwhether sleep disruption is caused by the pri-mary illness, associated symptoms, the sideeffects of medication or therapy, or the family’sand/or the child’s response to the illness is fre-quently present. Diagnosis rests on a compre-hensive history and physical exam.

Treatment

Treatment is based on management of the under-lying chronic disorder, control of associated symp-toms, and appropriate attention to sleep hygiene.If parental or patient anxiety about the chronicillness is suspected, appropriate supportiveinterventions should be recommended.

PSYCHOSOCIAL FACTORSRESULTING IN SLEEPLESSNESS

Childhood Affective DisordersIt is extremely difficult to diagnose depressionin children because the clinical picture may varydepending on the child’s age and developmentallevel.47-49 Disrupted sleep has long been noted aspart of the symptom complex. The polysomno-graphic characteristics seen in adult patientswith affective disorders have been well defined

and reproducible. Kane and associates47 havedescribed a patient with childhood depressionwho manifested polysomnographic variablesdifferent from the adult, but similar in charac-teristics to adult studies when compared withnormative data for children of the same age.A significant disturbance in sleep continuity wasdescribed. Early morning awakening, problemssettling, and intermittent nocturnal arousals wereall present. Sleep efficiency was decreased, anda stable but persistently shortened REM latencywas also noted. Although the REM latency wasat the lower end of normal for the patient’s age,it was significantly shorter than the mean valuefor the patient’s age group.

Stress may be associated with sleeplessnessin children. Transient DIMS may be precipitatedby acute life events, such as the death ofa grandparent or parental marital discord.Persistent disorders of sleeplessness in the childis more likely due to parental mismanagementof transient sleeplessness caused by acute stressreactions. Another possibility is that stressfullife events do not affect children directly butrather are mediated by parental affect and changesin their responsiveness and caretaking, espe-cially during infancy, when an infant’s sleep–wakeorganization may be affected by responsive care-taking.50

Depression is not just another problem buta central link between many kinds of problems,among them those that lead to depression andthose that may follow.51 Psychosocial factors iden-tified as being associated with sleep problems mayrepresent a parent’s withdrawal of psychologicalattention.52 There is evidence that an associationexists between parental depressed feelings anda child’s sleep problems. Depressed maternal feel-ings, rather than other measured psychosocialstresses, such as separation experiences, are asso-ciated with the development of sleep problems.Long-term sleep continuity problems and theassociation of persistent sleep problems withparental difficulty in behavior management havebeen reported.14

To determine whether the sleep problems com-monly seen in pediatric practice are associatedwith more pervasive disturbances in the childor family, Lozoff and coworkers studied twogroups of healthy children.52 Five experiencesdistinguished children with sleep problems fromthose without them: an accident or illness in the


Ch11.qxd 01/28/05 10:30 AM Page 143

family, an unaccustomed absence of the motherduring the day, maternal depressed mood(s),sleeping in the parental bed, and the maternalattitude of ambivalence toward the child. Thesefindings attest to the importance of sleep prob-lems as an early childhood symptom. Bedtimeconflicts and night-time awakening seem to bequantifiable, easily ascertainable behavior pat-terns that could alert pediatric health profession-als to the existence of more pervasive disturbancesin the child and family.

Night-time FearsChildhood fears are normal. Most often they aredevelopmentally related and are manifested ina variety of ways. Children 2 to 3 years old mayexhibit aggression directed toward siblings (sib-ling rivalry), may fear death or the loss of a par-ent, or may be troubled by separation fromparents and emerging socialization. The clinicalpresentation may vary significantly, and devel-opmentally appropriate fears may be generallyobscure. Children may complain of a fear ofrobbers or monsters. These fears may be mani-fested as frightening dreams. In general, the eti-ologies of nocturnal fears and nightmares aresimilar.1 If these fears are extreme, excessivelystrict limit setting by parents will require thechild to deal with these fears alone, are usuallyunsuccessful in resolving the sleep disturbance,and may be detrimental to the emotional healthof the child.

To determine the origin of the fears that maybe manifested at night and by sleep disturbances,evaluation should occur during waking hours.Evaluation and intervention by a psychologistor psychiatrist should be considered, especiallyif the problem has persisted for a significantperiod of time. If fears (and symptoms) havebeen present for only a short period of time,parental understanding and firm support maybe all that is necessary. Positive reinforcement,progressive relaxation techniques, biofeed-back, and techniques of self-control may beattempted. A change in the bedtime ritual, withthe parent remaining close to the child for sup-port and with gradual withdrawal, may be suc-cessful in some children. Others may respondwell to modifications in the physical sleepingenvironment.

Inadequate Limit SettingDisordered sleep secondary to inadequate limit-setting is most often seen in children duringmiddle childhood and early adolescence. Theusual complaint is one of difficulty settling atnight and bedtime struggles. The child will oftenrefuse to remain in bed and frequently evenin the bedroom. Although the child is physio-logically prepared for sleep, the parents give into the child’s protestations easily and are unwill-ing or unable to enforce night-time rituals androutines consistently. Therefore, the child doesnot remain in bed long enough for sleep onsetto occur. The child repetitively gets out of bedand protests, the parents become disturbed, ten-sion escalates, and the parents ultimately givein and the child’s behavior is reinforced. Parentsmay lack knowledge regarding limit setting or beunaware that they are not appropriately settinglimits for the child. Environmental factors, suchas sharing a bedroom with a sibling or with par-ents, and/or marital discord, may contribute tosleep disruption. Parental lack of recognitionof the intensity of the child’s fears and the inabil-ity to provide supportive firmness in the faceof the child’s protestations may increase anxietyand exacerbate the problem.

Bedtime struggles in children who are notphysiologically ready for sleep due to circadianrhythm disturbances or who suffer from inordi-nate night-time fears may present in a mannersimilar to that in children who do not haveappropriate limits set. Clinical differentiation maybe difficult, and evaluation of the family, intra-familial relationships, and adequacy of parentingis necessary.

If inappropriate limit setting is suggested clin-ically, treatment should be designed to addressthe underlying etiology. Psychological causesand parental dysfunction should be addressedfirst. Parents should be counseled on mecha-nisms to enforce consistency and to provide sup-portive firmness. A regular bedtime ritual shouldbe implemented and parents must be unwillingto modify the regimen, regardless of protesta-tions by the child. Closing the door or a gatemight be used to keep the child in his/her room.However, door closing must be associated witha supportive parental response, and this may bedifficult if significant parental dysfunction is


Ch11.qxd 01/28/05 10:30 AM Page 144

concomitantly present. Anger, punishment, andescalation of tension must be eliminated. Childrenover 3 years of age may respond to specificbehavior modification techniques and negotia-tions using positive reinforcement for appropri-ate behavior and compliance with objectivesof the regimen. The concepts of persistency, con-sistency, and winning should be stressed. Theparents must decide who is going to win, theyor the child. Parental patience and persistencyare frequently less than those of the child, believ-ing it is easier to give in to the child than tostruggle. By letting the child win, the parentsreinforce similar behaviors that are often mani-fested during waking hours as well.

CHRONOPHYSIOLOGIC FACTORSRESULTING IN SLEEPLESSNESS

Advanced Sleep Phase SyndromeAdvanced sleep phase syndrome is a less com-mon chronophysiologic disorder resulting insleeplessness. Total sleep time is normal; how-ever, sleep occurs at an inappropriate time. Whensleep phase is advanced, children retire and set-tle early and easily. Bedtime struggles are notcommon, and children may become cranky anddisagreeable if kept awake when they are physi-ologically ready for sleep. Nocturnal awakeningsand protestations are infrequent or absent.Parental complaints most often center aroundearly morning awakening. The child is awake,alert, happy, and playful during early morninghours. Children with an advanced sleep phasewill often go unrecognized because early bed-time and awakening time may fit well withparental schedules and cause little social inter-ruption. School attendance and performance arerarely affected, and parents may appreciate thefree time afforded to them in the evening.Maintenance of a sleep diary will greatly assistin the diagnosis.

Treatment is straightforward and a rapidresponse should be expected. Delaying the sleepphase is significantly easier than advancingit, since the master circadian pacemaker, underfree-running, non-entrained conditions, is spon-taneously delayed (usually by 1 hour for each24-hour period). Bedtime should be graduallydelayed by 30 to 60 minutes each night, depend-

ing on the developmental level of the child.Morning awakenings should remain sponta-neous and will be delayed reflexively once set-tling is delayed. When the desired bedtime andawakening time have been achieved, they shouldbe fixed and enforced every night (includingweekends and holidays). Again, persistency andconsistency are imperative for the success of theregimen. A sleep diary should be maintainedduring the course of treatment to objectively mon-itor progress.

Delayed Sleep Phase SyndromeDelayed sleep phase syndrome is the result ofa shift from the normal period of sleep to a timelater than that expected or desired. The charac-teristics of sleep are normal, but the childrensleep at the wrong times. Children are physio-logically ready for sleep late in the evening, andthe morning spontaneous awakening time mayoccur during the late morning or early afternoonhours, depending on the degree of the shift).Nocturnal awakenings do not appear to bea problem. Once the child is asleep, s/he tends tostay asleep. Parental expectations for bedtimemay be quite earlier than the time the child isbiologically ready for sleep, resulting in the typ-ical complaint of sleep onset difficulties and sig-nificant bedtime struggles. Because the normalsleep period has shifted to a time later in the day,parents will usually report profound difficultiesin waking up the child in the morning. Thechild’s behavior and activity in the morninghours are quite sluggish. School performance inthe morning classes may be poorer than thatin the afternoon classes, and the child may fallasleep in school, especially at the beginning ofthe week. On weekends and holidays, when bed-time tends to be later in the evening, there arefewer struggles and the latency from “lights out”to sleep is shorter. Children suffering fromdelayed sleep phase syndrome tend to be sleepdeprived during the week because of the earlymorning awakenings but catch up on the week-ends. It may be difficult to differentiate delayedsleep phase syndrome from other disorders pre-senting with sleep onset difficulties, such as dis-orders of inappropriate and inconsistent limitsetting. A comprehensive evaluation and mainte-nance of a sleep diary for 2 to 3 weeks are of


Ch11.qxd 01/28/05 10:30 AM Page 145

great assistance in evaluating a child for sleepphase delays.

Small phase shifts in younger children are usu-ally best treated with a controlled phase advance.The regimen should begin when the child doesnot have important daytime social and develop-mental responsibilities (i.e., school), since somedegree of sleep deprivation occurs during thefirst few days of management. The regimenbegins by initially delaying the hour of bedtimefor 2 or 3 days until the child is physiologicallyready for sleep and sleep onset latency is rela-tively normal; bedtime struggles will disappearquickly. Also at the beginning of the regimen,the morning wake-up time should be strictlyfixed and maintained at an appropriate time,even on weekends and holidays. If it is develop-mentally appropriate, naps should be continuedbut prolonged naps avoided. If the child doesnot normally nap, daytime sleep should not bepermitted, since the child may accumulate sleepduring these hours at the expense of nocturnalsleep. Once the child is settling easily, the bed-time should be slowly advanced until the desiredtime is reached. The degree of advance should bedetermined by the child’s tolerance and responseto treatment. However, it must be rememberedthat the total sleep time should be commensu-rate with the child’s age and development.Parents should be counseled regarding appropri-ate sleep period times at various ages and shouldexpect realistic compliance from their child. Again,maintenance of a sleep diary will greatly assistin the assessment and monitoring of the thera-peutic regimen.

Treatment of extensive sleep phase delays inadolescent patients may involve further delayof the sleep phase around the clock (movingin the direction of a free-running pacemaker)until the desired sleep and wake-up times arereached. This should be accomplished only whenthe youngster has no other social responsibilities,since s/he will be sleeping all day and remainingawake all night for a time. In addition, a parentmay have to stay up with the youngster andphase delay with him/her in order for the regi-men to succeed. This form of treatment may beextremely difficult and inappropriate for youngeradolescents and prelatency children.

Exposure to intense light, especially whenit occurs during windows of entrainment, may

quickly shift the sleep phase to an appropriatetime. This form of therapy appears promisingfor the adult patient, although it is still experi-mental. Similar trials have not been performed inchildren with phase shifts.

Regular but InappropriateSleep–Wake ScheduleChildren may present with a complaint of nap-ping too late or too early in the day. Naps may betoo frequent or prolonged. Daytime sleep mayalso be too infrequent for the child’s develop-mental level. Occasionally, the child complainsof an inappropriately early bedtime or earlymorning awakenings (which appear similar toa phase advance), but significant daytime sleepalso occurs, resulting in appropriate total sleeptimes. Although any of these symptoms may bepresent, they tend to be consistent and recurdaily. A sleep diary for a period of 2 to 3 weekswill reveal these regularly recurring but inappro-priate times of sleep.

Treatment is directed at normalization of thesleep–wake schedule. Once a regular and appro-priate sleep–wake schedule is attained, sleepnormalizes.

Irregular Sleep ScheduleEntrainment of circadian rhythms occurs only ifappropriate zeitgebers are present and sleep–wakeschedules remain constant. Inappropriate napsand schedules may result in internal desyn-chronization of other systems as well. In thesepatients, not only bedtime is disrupted but alsodaily living in general is chaotic and lacks formalstructure. Social instability is often present. Mealtimes tend to be irregular from day to day, andmeals may be taken by different family membersat different times.

Family dysfunction must be addressed first.With little structure in the lives of all membersof the household, there is little chance that struc-turing sleep for the child will be successful unlessthe entire ecology of the family is treated. Sleepshould be charted regularly and appropriateschedules for meals and sleep established.Appropriate time cues must be provided, andstructured wake-up times and bedtimes must bemaintained in a consistent and persistent manner.


Ch11.qxd 01/28/05 10:30 AM Page 146

OTHER FACTORS RESULTINGIN SLEEPLESSNESS

Childhood Onset of Disorders ofInitiating and Maintaining SleepOne form of primary insomnia in adults has theonset of symptoms before puberty. Some evi-dence may place the presence of this disorderearly in infancy.53 Sleep onset and/or sleep main-tenance complaints may be present. Daytimesymptoms of inadequate sleep may occur. It isdifficult to differentiate this form of DIMS fromothers, since it may be finally diagnosed onlyif the sleep complaint persists through pubertyand into adulthood. There seems to be no iden-tifiable precipitating cause. Childhood-onsetDIMS may also constitute a shift in the balanceof the circadian pacemaker responsible for gen-eration of the sleep–wake rhythm. The patternsof poor sleep are often more difficult to treatthan the other forms of insomnia during child-hood, and poor sleep tends to continue through-out both the negative and positive periodsof emotional and developmental adaptation.This group is also differentiated from shortsleepers by the presence of daytime symptomsof fatigue, irritability, tenseness, mild depression,difficulties in waking attention, and at times day-time sleepiness.

Short SleeperAlthough it is not typically discussed in the pedi-atric literature, the short sleeper is presentedhere because it may be genetically or congeni-tally determined. Short sleeper is the designationfor an individual who consistently sleeps sub-stantially less in a single day than the customaryamount of sleep for the patient’s age and devel-opmental level.53 Sleep, though short, is normal.There are typically no specific complaints aboutthe quality of sleep, although the symptoms ofbedtime struggles and/or early morning awaken-ings may occur during childhood. Daytimesleepiness, difficulty with daytime behavior, andpoor performance are notably absent. True DIMSdoes not exist despite the occasional desire andunsuccessful attempts to sleep longer. Short sleepappears to be at one end of the normal individ-ual sleep requirement continuum.

As previously mentioned, short sleep has beenlinked to reduced life expectancy. This relation-ship probably has its source mainly in short totalsleep time patterns resulting from medicaland/or other sleep pathologies and not in shortsleep itself, as represented by the short sleeper.53

Insomnia is a symptom and not a diagnosis.The causes of insomnia are varied and rangefrom the medical (i.e., drug-related, pain-induced,associated with primary sleep disorders suchas obstructive sleep apnea) to the behavioral(i.e., associated with poor sleep hygiene or sleeponset association disorder) and are often a com-bination of these factors. In adults insomnia isgenerally defined as difficulty initiating and/ormaintaining sleep and/or early morning awaken-ing and/or nonrestorative sleep. However, thedefinition of insomnia or problematic sleep inchildren is much more challenging for a numberof reasons. Clinically significant sleep problems,like many behavioral problems in childhood,may best be viewed as more loosely occurringalong a continuum of severity and chronicitythat ranges from a transient and self-limitingdisturbance to a disorder that meets specificdiagnostic criteria. Unlike strict research defini-tions of sleep problems, the validity of parentalconcerns and opinions regarding their child’ssleep patterns and behaviors and the resultingstress on the family must be considered in defin-ing sleep disturbances in a clinical context. Therelative prevalence and the various types of sleepproblems that occur throughout childhood mustalso be understood in the context of normalphysical, cognitive, and emotional phenomenathat are occurring at different developmentalstages. Parental recognition and reportingof sleep problems in children also vary acrosschildhood, with the parents of infants and tod-dlers more likely to be aware of sleep concernsthan those of school-aged children and adoles-cents. Thus, the range of sleep behaviors thatmay be considered “normal” or “pathologic”is broad and the definitions are often highlysubjective.

To more clearly define the clinical situationsin which the use of pharmacotherapy for pedi-atric sleep problems might be appropriate, a con-sensus panel of pediatric sleep medicinespecialists agreed that the development of a con-sensus definition of pediatric insomnia was


Ch11.qxd 01/28/05 10:30 AM Page 147

a necessary and important first step. The follow-ing key components of the consensus definitionwere developed by the panel.

1. Pediatric insomnia may be defined as diffi-culty initiating or maintaining sleep that isviewed as a problem by the child or caregiver.

2. The significance of the sleep problem may becharacterized by

● its severity, chronicity, and frequency and● associated impairment in daytime function

for the child or family.

3. The sleep problem may be due to a primarysleep disorder or occur in association withother sleep, medical, or psychiatric disorders.

Although prevalence rates are only approx-imations because the definition of difficultyinitiating or maintaining sleep in children variesacross studies, approximately 25% of all childrenare reported to experience some type of sleepproblem at some point during childhood. Specificstudies have reported an overall prevalence ofa variety of parent-reported sleep problems rang-ing from 25% to 50% in pre–school-aged sam-ples54 to 37% in a community sample of 4to 10 year olds55 to up to 40% in adolescents.56

Furthermore, sleep concerns are one of the mostfrequent parental complaints in pediatric prac-tices, reported as the fifth leading concern of par-ents after illness, feeding, behavior problems, andphysical abnormalities.57

Although these sleep disturbances are tran-sient in many children, there is considerableevidence that sleep problems may persist orrecur in a substantial percentage of them.7,8

In addition to their high prevalence and chronic-ity, recent evidence also suggests that sleep dis-orders may have significant short- and long-termconsequences on children’s academic and socialfunctioning and on their health.9,10 A wealthof empirical evidence from several lines of researchclearly indicates that children and adolescentsexperience significant daytime sleepiness asa result of inadequate or disturbed sleep and thatsignificant performance impairment and mooddysfunction are associated with daytime sleepi-ness. Higher-level cognitive functions, such ascognitive flexibility and abstract thinking/rea-soning, seem to be particularly sensitive to theeffects of disturbed or insufficient sleep. Finally,the health outcomes of inadequate sleep include

an increase in accidental injuries (ranging fromminor injuries to drowsy driving–related motorvehicle fatalities) and the potentially deleteriouseffects on the cardiovascular, immune, and vari-ous metabolic systems, such as glucose metabo-lism and endocrine function. Sleep problems arealso a significant source of distress for familiesand may be one of the primary reasons for care-giver stress in those families with children whohave chronic medical illnesses or severe neu-rodevelopmental delays.

SCREENING FOR SLEEP PROBLEMS

A number of studies have suggested thatscreening for sleep problems in pediatric prac-tice is inadequate and may result in significantunderdiagnosis of sleep disorders. For example,in a recent survey of over 600 community-based pediatricians, over 20% of the respon-dents did not routinely screen for sleepproblems in school-aged children in the con-text of the well-child visit, and less than 40%questioned adolescents directly about theirown sleep habits. Recognizing this gap, the taskforce recommended that all children be regu-larly screened for sleep problems in pediatricclinical practice. One simple sleep-screeningalgorithm is BEARS (Bedtime problems,Excessive daytime sleepiness, Awakenings atnight, Regularity and duration of sleep, Snoring.The key areas of inquiry that are included in theBEARS screening for sleep problems in childrenand adolescents are (1) bedtime resistance anddelayed sleep onset; (2) frequent and/or pro-longed night-time awakenings; (3) regularity,pattern, and duration of sleep; (4) snoring andother symptoms of sleep-disordered breathing;(5) sleep-related anxiety behaviors (e.g., night-mares, night-time fears); and (6) excessive day-time sleepiness (e.g., difficulty awakening in themorning, naps). A number of other brief parentand self-reporting sleep survey tools have alsobeen developed that can facilitate the screeningprocess and yield important information aboutthe nature and severity of any coexisting sleepcomplaints.

EVALUATION OF SLEEP COMPLAINTS

The clinical evaluation of a child presenting witha sleep problem involves a careful medical his-tory to assess the potential medical causes ofsleep disturbances, such as allergies, concomitant


Ch11.qxd 01/28/05 10:30 AM Page 148

medications, and acute or chronic pain condi-tions. A developmental history is importantbecause of the aforementioned frequent associa-tion of sleep problems with severe developmen-tal delay. Assessment of the child’s current levelof functioning (e.g., school, home) is key in eval-uating the possible mood, behavioral, and neu-rocognitive sequelae of sleep problems. Currentsleep patterns, including the usual sleep dura-tion and sleep–wake schedule, are often bestassessed with a sleep diary, in which parentsrecord daily sleep behaviors for an extendedperiod. A review of sleep habits, such as bedtimeroutines, daily caffeine intake, and the sleepingenvironment (e.g., temperature, noise level) mayreveal environmental factors that contributeto the sleep problems. The use of additionaldiagnostic tools such as polysomnographic eval-uation are seldom warranted for the routineevaluation of pediatric insomnia but may beappropriate if organic sleep disorders such asobstructive sleep apnea or periodic limb move-ments are suspected.

Finally, referral to a sleep specialist for diag-nosis and/or treatment should be consideredunder the following circumstances: childrenor adolescents with persistent or severe bedtimeissues that are not responsive to simple behav-ioral measures or are extremely disruptive; chil-dren or adolescents with parasomnias who alsopresent with symptoms of another underlyingsleep disrupter (e.g., sleep disordered breath-ing) or for whom pharmacologic treatment isbeing considered; children with associated med-ical, psychiatric, and/or developmental conditionsthat create additional management challenges;and children and adolescents with circadianrhythm disorders.

Currently, there are no medications approvedby the Food and Drug Administration for thetreatment of difficulty initiating and/or main-taining sleep in the pediatric population.Although it is clear that the ideal pediatric hyp-notic medication does not exist, the task forcemembers agreed that it was important to con-sider the characteristics that would be presentin the optimum clinical situation to allow thepractitioner to compare the pharmacologicoptions that are currently available. The phar-macokinetic properties of the ideal pediatrichypnotic would include the high oral bioavail-ability, a property that encompasses solubility,

rapid absorption, stability, and invulnerabilityto extensive first-pass metabolism by the liverand/or the gut to ensure rapid, consistent, andreliable clinical responses and allow appropriatedosages to be accurately predicted. Metabolismshould be to inactive products or active metab-olites with short half-lives that are no longerthan that of the parent compound to minimizethe metabolite-associated side effects. Eliminationhalf-life should be short (2 to 3 hours).

Pharmacodynamic properties would includea rapid onset of effect, preferably within 30minutes, so that it could be administeredshortly before bedtime and a duration of actionthat would be sufficiently long so that onlyonce nightly dosing is necessary but not exces-sively long so as to avoid residual daytime seda-tion. There should also be no associatedrebound, tolerance, or withdrawal and only afew side effects—preferably the tolerabilityprofile of placebo—with little or no potentialfor drug–drug interactions. The ideal hyp-notic should also not affect sleep architec-ture because changes in slow-wave sleep, forexample, might lead to alterations in hormonelevels such as human growth hormone.35

Finally, the medication should be available ina palatable oral liquid as well as a tablet/capsuleformulation.

Disorders of Sleep OnsetAssociationsTreatment

The discussion of a variety of treatment modali-ties along with their essential components isprovided in the following sections.

DIET AND LIFESTYLE

● Treatment for this disorder and most otherdisorders associated with the problem ofsleeplessness in childhood involves imple-mentation of the principles of sleep hygiene.

● The sleeping environment should be quiet anddark.

● Morning awakening time should be firmlyfixed and consistently enforced. It is the mostpowerful time for the entrainment of thesleep–wake cycle.

● Bedtime should be firmly enforced. However,some flexibility must be allowed for a normalfamily lifestyle.


Ch11.qxd 01/28/05 10:30 AM Page 149

● Environmental temperatures should remain ata comfortable level, generally less than 75° F.Excessively hot temperatures can disturb sleepcontinuity.

● Hunger at bedtime should be avoided by theprovision of a small bedtime snack.

● Excessive bedtime and nocturnal fluids shouldbe avoided.

● Children must learn to fall asleep on theirown, without parental intervention. Infantsand children should be placed in their cribs orbeds in a drowsy but awake state at the begin-ning of their nocturnal sleep period.

● Vigorous activity should be avoided for 1 to2 hours before bedtime. Baths are stimulatingfor children and should be moved to themorning hours if there are problems settlingat the beginning of the nocturnal sleep period.

● Food or beverages containing caffeine (orother methylxanthines) should be avoidedfor several hours before bedtime. Commonfoods and beverages containing caffeine includebut are not limited to colas, chocolate, tea, andcoffee.

● A variety of medications contain alcohol(elixirs) or caffeine and should be avoided forseveral hours before bedtime.

● Naps should be developmentally appropriate.Prolonged naps close to the desired bedtime andnaps taken too frequently should be avoided.

● A faded-response program is typically themost reliable method to change sleep onsetassociations and assist the child in learning tofall asleep without the need for parental inter-vention.16 Faded response requires work onthe parents’ part. A resolution tends to occurrapidly if caretakers are persistent and consis-tent with the intervention.

● Since the sleeping environment must be safeand secure, the youngster needs to be assuredthat s/he can get his/her parents if needed. Buts/he should not need his/her parents in orderto fall asleep.

● The youngster should be allowed to protestfor a developmentally appropriate period oftime before the parents respond. Once theprotestation reaches this developmentallydetermined time limit, the parents respondand calm down the child. Once s/he is calmbut awake, they should again leave the roomand allow the child to protest slightly longerbefore they respond. Each intervention should

result in the child being calm, in bed, andawake when the parent leaves the sleepingenvironment.

● This process should be continued until thechild is asleep. Several hours of repeated inter-ventions may be required during the initialfew nights of treatment. Parents should be pre-pared for this and understand that if they areconsistent and persistent, the problem will beresolved quickly.

PHARMACOLOGIC TREATMENT

● Pharmacologic treatment is not indicated.

INTERVENTIONAL PROCEDURES

● Other interventions, such as environmentalmanipulation (rearranging the bedroom),night-lights, white noise, and other safe tran-sitional objects, may be used in combinationwith sleep hygiene and faded response.

ASSISTANCE DEVICES

● White-noise generators, sound machines, andcontinuous-playing compact disc players mayassist in masking environmental noise thatmay result in nocturnal awakening.

OTHER TREATMENTS

● Other behavior modification techniques havebeen used. Permitting the child to “cry it out”has been recommended; this can be successfulfor some children but is often very difficultfor the child and the family. As previouslynoted, security is quite important and permis-sive for sleep onset to occur. Anxiety relatedto an inability to alert parents to illness, pain,or other discomfort may create other sleep-related problems and settling difficulties.

● The treatment goal for excessive and repetitivenocturnal fluid intake is to gradually wean theyoungster from the fluids at night. Abruptly dis-continuing the fluids may be difficult and cancreate ongoing sleep maintenance problems.

● Weaning from a bottle is often quick and sim-ple. The volume of fluid is decreased by 1⁄4 to1⁄2 oz every night until there is a minimumamount in the bottle. If the child has diffi-culty with this level of reduction in fluidintake, the decrease may occur more slowly(e.g., every other night, every third night).

● The child should not be permitted to fall asleepwith the bottle.


Ch11.qxd 01/28/05 10:30 AM Page 150

● Once a minimum amount of fluid has beeningested, switching to a pacifier may beattempted.

● Weaning from breast-feeding, when associ-ated with excessive nocturnal fluids, may besomewhat more difficult, but attention to thetime spent nursing or switching to nocturnalfeeding with a bottle and then weaning(as mentioned above) might be considered.Nurturing may still be provided, but caremust be taken not to establish the nursingactivity as a sleep onset association. The nurs-ing mother is the best suited to assess thisneed. If the nursing period at night is consid-erably different, less vigorous, and shorterthan diurnal feedings, the nursing may havedeveloped into a sleep onset association.The child should not be permitted to fallasleep at the breast and should be placed inthe desired sleeping environment in a drowsybut awake state.

Neurologic DisordersTreatment

Treatment of problem sleeplessness in the neu-rologically challenged child depends on theidentified underlying etiology. All of the factorsthat influence sleeplessness and affect thesleep–wake cycle require attention. If behavioralor parental management of the sleeplessnessis suspected, it may be managed by attentionto appropriate sleep hygiene. However, it mightbe necessary to conduct the therapeutic regimenmore slowly than it would be in a non-neurolog-ically impaired child. Appropriate attentionshould be paid to the child’s sleep–wake schedule,circadian rhythm, and time cues, which are par-ticularly important for children with impairmentsof special senses (e.g., blindness). Correction ofthese schedule abnormalities is usually not diffi-cult and rapidly improves sleep patterns anddaytime function.


Pharmacologic treatment is often required inchildren with neurologic challenges. Treatmentmay be for the short or long term and may beinfluenced by significant comorbidity. The mostcommon pharmacologic agents used for problemsleeplessness in children with neurologic chal-lenges are listed below.17–19

● Diphenhydramine● Standard dosage: 2 to 6 years of age, 12.5 to

25 mg at bedtime; 6 to 12 years of age, 25 mgat bedtime; over 12 years of age, 25 to 50 mgat bedtime.

● Contraindications: Narrow-angle glaucoma,gastrointestinal or genitourinary obstruction.

● Main drug interactions: May interact withantihypertensive or antidepressant medica-tions; increased sedation when combinedwith alcohol.

● Main side effects: Nervousness, dizziness,hypertension, excitability, and drowsiness.

● Special points: Over-the-counter preparationsare often combined with other medications(e.g., pseudoephedrine) as decongestants.Elixirs contain 5% to 6% alcohol. Multiplepreparations are available. Patients’ parentsshould be instructed to read the medicationlabel before purchase and administration.

● Chloral hydrate● Standard dosage: Sedation and anxiety, 25 to

50 mg/kg (oral or rectal) every 8 hours, witha maximum of 500 mg; Hypnotic, one dose50 mg/kg (oral or rectal), with a maximumof 2 g.

● Contraindications: Hypersensitivity to chlo-ral hydrate; hepatic or renal impairment;gastritis or ulcers; severe cardiac disease.

● Main drug interactions: May potentiatethe effects of warfarin, CNS depressants,and alcohol; concomitant intravenousadministration of furosemide may resultin flushing, diaphoresis, and blood pressurechanges.

● Main side effects: Disorientation, excitement,dizziness, fever, headache, ataxia, rash,urticaria, gastric irritation with nausea, vom-iting, or diarrhea; respiratory depressionwhen combined with other sedatives or nar-cotics.

● Special points: Patients develop toleranceto the hypnotic effects, so chloral hydrateis not recommended for use for longer than2 weeks; taper dosage to avoid withdrawalwith prolonged use.

● Imipramine● Standard dosage: Imipramine is typically

used for the treatment of various forms ofdepression, often in conjunction with psy-chotherapy, and enuresis in children and hasbeen used as an analgesic for certain chronic


Ch11.qxd 01/28/05 10:30 AM Page 151

and neuropathic pain. Hypnotic doses havenot been clearly delineated. Depression,1.5 mg/kg/day, with dosage incrementsof 1 mg/kg every 3 to 4 days; maximum doseis 5 mg/kg/day in 1 to 4 divided doses.Enuresis, 10 to 25 mg at bedtime.

● Contraindications: Hypersensitivity toimipramine (cross-sensitivity to other tri-cyclic antidepressants may occur); patientsreceiving monoamine oxidase inhibitorswithin the past 14 days; narrow-angle glau-coma.

● Main drug interactions: May decrease orreverse the effects of guanethidine and cloni-dine; may increase the effects of CNS depres-sants, adrenergic agents, anticholinergicagents, and alcohol; hyperpyrexia, tachycar-dia, hypertension, seizures, and death mayoccur with monoamine oxidase inhibitors;similar interactions with other tricyclic anti-depressants may occur.

● Main side effects: Drowsiness, confusion,dizziness, fatigue, anxiety, nervousness,seizures, rash, photosensitivity, nausea, vom-iting, constipation, dry mouth, decreasedappetite; urinary retention, blood dyscrasia,hepatitis, blurred vision, increased intraocu-lar pressure, weakness, and hypersensitivityreactions.

● Special points: Do not discontinue abruptly inpatients receiving long-term, high-dose ther-apy; use with caution in patients with cardio-vascular disease, conduction disturbances,seizure disorders, urinary retention, anorexia,or hyperthyroidism or those receiving thyroidreplacement.

● Hydroxyzine● Standard dosage: 0.5 mg/kg orally 30 min-

utes before bedtime.● Contraindications: Hypersensitivity to

hydroxyzine or any of its components.● Main drug interactions: May potentiate

other CNS depressants or anticholinergicsand can antagonize the vasopressor effects ofepinephrine.

● Main side effects: Hypotension, dizziness,headache, ataxia, dry mouth, urinary reten-tion, and weakness.

● Special points: Competes with histamine forH1-receptor sites on effector cells in the gas-trointestinal tract, blood vessels, and respi-ratory tract.

● Diazepam● Standard dosage: 0.04 to 0.25 mg/kg at bed-

time.● Contraindications: Hypersensitivity to ben-

zodiazepines; should not be used in patientswith preexisting CNS depression, respira-tory depression, narrow-angle glaucoma,or severe and uncontrolled pain.

● Main drug interactions: Inducers ofcytochrome P450 2c may increase the metab-olism of diazepam. There is increased toxic-ity in the presence of CNS depressants.Cimetidine may decrease the metabolismof diazepam. Cisapride can significantlyincrease diazepam levels. Valproic acid maydisplace diazepam from binding sites, whichmay result in an increase in sedative effects.Selective serotonin reuptake inhibitors (e.g.,fluoxetine, sertraline, paroxetine) havegreatly increased diazepam levels by alteringclearance.

● Main side effects: Drowsiness, confusion,dizziness, amnesia, slurred speech, ataxia,paradoxical excitement, blurred vision,decreased respiratory rate, and apnea.

● Special points: Depresses all levels of theCNS, including the limbic system and retic-ular formation. Benzodiazepines must beused with caution in the presence of otherCNS depressants and also in patients withCNS lesions that affect the brainstem andrespiratory centers.

● Clonazepam● Standard dosage: Start at 0.01 mg/kg at bed-

time. Maximum dose should not exceed0.025 mg/kg.

● Contraindications: Hypersensitivity to clo-nazepam or other benzodiazepines; severeliver disease and acute narrow-angle glau-coma.

● Main drug interactions: Concomitant usewith other CNS depressants will increasesedation; phenytoin and barbiturates increaseclearance of clonazepam.

● Main side effects: Changes in behavior andpersonality, dizziness, headache, memoryimpairment, decreased concentration, hang-over effect, and paradoxical insomnia.

● Special points: Clonazepam-induced behav-ioral problems may be more frequent inchildren with mental handicaps; when dis-continuing treatment, taper slowly.


Ch11.qxd 01/28/05 10:30 AM Page 152

● Lorazepam● Standard dosage: 0.05 mg/kg at bedtime.● Contraindications: Hypersensitivity to ben-

zodiazepines; should not be used in patientswith preexisting CNS depression, respiratorydepression, narrow-angle glaucoma, orsevere and uncontrolled pain.

● Main drug interactions: Other CNS or respi-ratory center depressants may increase theadverse effects of lorazepam.

● Main side effects: Drowsiness, lethargy,hangover effect, dizziness, transitory hallu-cinations, and ataxia.

● Special points: Use with caution in patientswith hepatic or renal impairment. Careshould be taken when used in patients withcompromised pulmonary function.

● Clonidine● Standard dosage: Initial dosage is 2.5 to

5 μg/kg at bedtime.● Contraindications: Hypersensitivity to cloni-

dine hydrochloride or any component.● Main drug interactions: Tricyclic antidepres-

sants antagonize the hypotensive effects ofclonidine. Beta-blocking agents may potenti-ate bradycardia, and withdrawal may result inrebound hypertension. CNS depressants andalcohol may increase the sedative effects.

● Main side effects: Palpitations, drowsiness,sedation, headache, insomnia, anxiety,depression, rash, constipation, and fatigue.

● Special points: Clonidine is a very potentREM sleep suppressant. On abrupt with-drawal, there is considerable REM sleeprebound.

● Melatonin● Standard dosage: There are only a few well

controlled studies of the use of melatonin inchildren. Typical doses have ranged from 0.5mg to 10 mg taken 1 to 5 hours before bed-time.20

● Contraindications: None.● Main drug interactions: None.● Main side effects: One report has shown

an increase in seizure activity in childrenwith significant neurologic conditions andseizures.21 No other significant side effectshave been reported.

● Special points: Many over-the-counter prepa-rations may contain other ingredients. Sinceformulations are proprietary, other ingre-dients may not be known. Pure synthetic

melatonin is not soluble in water, and liq-uid preparations may contain 7% to 13%alcohol.

INTERVENTIONAL PROCEDURES

In some patients with hypertonicity, decreasingmuscle tone may result in improved sleep.Medication used to decrease tone and spasticitymay result in more consolidated sleep.

Surgical interventions are not specific to thechild’s sleep-related disorder; they are aimed atimprovement of the underlying neurologic con-dition (e.g., tendon release in spastic quadriple-gia, posterior fossa decompression in patientswith Chiari II malformations). A beneficial effecton sleep is secondary to improvement in theyoungster’s underlying condition.

Approaches to the child with underlying neu-rologic conditions are similar to those usedin other youngsters. White-noise devices orsound generators, compact disc players, night-lights, and other assistance devices may be usedfor problem sleeplessness in children with neu-rologic disorders.

PHYSICAL/SPEECH THERAPY AND EXERCISE

Physical therapy and functional improvementmay also ameliorate problem sleeplessness.Brushing or combing for 15 to 20 minutes beforebedtime may be helpful for youngsters with sen-sory integration problems. A soft brush is rhyth-mically stroked directly on the skin of the arms,legs, chest, abdomen, and back. Brushing may bedone 2 to 3 times per day and should cease beforesleep onset so that it will not become a sleeponset association.

Advanced Sleep Phase SyndromeTreatment

DIET AND LIFESTYLE

● Increase early evening activity and bathe thechild before the desired bedtime.

● Increase environmental light exposure duringthe late afternoon and early evening hours.Decrease environmental light exposure duringthe early morning hours.

● Bedtime may be delayed from 30 minutes to2 hours, depending on the child’s age anddevelopmental level.

● If it is developmentally appropriate, providea brief mid-afternoon nap.


Ch11.qxd 01/28/05 10:30 AM Page 153


In adults, 0.5 to 10 mg of melatonin taken after4:00 AM or at the time of morning sleep offset hasbeen recommended. Similar data in children arelacking, and it is unclear whether melatonin hasa place in the treatment of advanced sleep phasesyndrome in children.

Delayed Sleep Phase SyndromeTreatment

The therapeutic components constituting dietand lifestyle as well as pharmacology in the treat-ment and management of delayed sleep phasesyndrome are discussed below.

Diet and Lifestyle● Small phase shifts are typically best treated

with a controlled phase advance. The regimenshould begin when the child does not haveimportant daytime responsibilities, since somedegree of sleep deprivation will occur.

● The regimen begins by delaying the hour ofbedtime for 2 to 3 days to a point at which thechild is physiologically ready for sleep. If thereis significant stimulation during the eveningbefore bedtime, this should be minimized andswitched to the morning hours.

● Watching television before bedtime should beavoided. Television programming requires thechild to delay sleep onset in order to finishwatching the show. Videotapes that may bestopped and restarted may help.

● Exposure to bright light in the evening andat around bedtime should be avoided. Brightlight and stimulation from computer gamesand activities may both result in delayed sleeponset.

● Morning awakening time should be firmlyfixed. It should not vary by more than 1 to 1.5hours on school days, weekends, and holi-days.

● Exposure to bright light in the morning mayassist in the slow advance of the sleep phase.Bright light may be provided through a varietyof phototherapy devices or exposure to sun-light. Bright computer screens surrounded byother lamps may also provide the intensity oflight required for assisting in the advanceof the sleep phase.

● If it is developmentally appropriate, napsshould continue but prolonged naps should beavoided.

● If the child does not normally nap, daytimesleep should be avoided.

● Once the youngster is settling easily, the bedtimemay be slowly advanced until its desired timeis reached. The child’s tolerance and responseto treatment should determine the degree ofadvance.

● Total sleep time should remain appropriateto the child’s age and development.

● Parental expectations should be realistic.● Occasionally, large phase delays in adolescents

may be managed by a continuous (around-the-clock) phase delay.


● Melatonin has been used for the short term inadults with delayed sleep phase syndrome andhas been reported to aid in the achievementof and adherence to a new sleep schedule. Inaddition, chronic use may assist in maintain-ing the schedule. Nonetheless, no data are sim-ilarly available in children.

● Over-the-counter hypnotics and sedatives aswell as prescription medication have not beenshown to be effective in the long-term resolu-tion of delayed sleep phase syndrome.

Non–24-Hour Sleep–WakeScheduleTreatment

This sleep-related disorder is more common insightless children. It is also common in childrenwith middle fossa tumors both before and aftersurgical and/or radiologic intervention.

● Begin and maintain a firm sleep–wake schedule.● Morning awakening time should be firmly fixed.● Providing sufficient and powerful time cues is

essential in assisting entrainment to a 24-hoursleep–wake cycle.

● Caffeine-containing food and beveragesshould be avoided after 2:00 PM.


● Melatonin 0.5 to 10 mg may be taken 1 to4 hours before the desired bedtime.

● Vitamin B12 has been reported to be helpfulin some adult patients. Doses range from 0.5 to3 mg per day.


Ch11.qxd 01/28/05 10:30 AM Page 154

Primary Insomnia in ChildhoodTreatment● Treatment is centered on the institution of a

comprehensive program of sleep hygiene andappropriate attention to sleep–wake patternsacross the 24-hour continuum.

● The sleeping environment should be quiet anddark.

● Morning awakening time should be firmlyfixed and consistently enforced.

● Bedtime should be firmly enforced.● The environmental temperature should remain

at a comfortable level.● Hunger at bedtime should be avoided by the

provision of a small bedtime snack.● Excessive fluids before bedtime should be

avoided.● Children must learn to fall asleep on their

own, without parental intervention.● Vigorous activity should be avoided for 1 to

2 hours before bedtime.● Food or beverages containing caffeine (or other

methylxanthines) should be avoided for severalhours before bedtime.

● A variety of medications contain alcohol (elixirs)and should be avoided for several hours beforebedtime.

● Naps should be developmentally appropriate.● Appropriate parental expectations should be

established. These expectations should be basedon principles related to the ontogeny of thesleep–wake cycle.

● Some children require less sleep than others.The timing of the nocturnal sleep period shouldbe coincident with the youngster’s normal sleeprequirements.

● Behavioral approaches should be attemptedbefore other interventions.

● When behavioral interventions fail to resolveproblem sleeplessness, assessment of the par-ents’ or caretakers’ ability to comply with thetherapeutic protocol is warranted.

● When compliance fails, more intensive behav-ioral and family counseling may be indicated.


General recommendations for the use of medica-tions in pediatric insomnia:

● Since the ideal pediatric hypnotic does notcurrently exist, rational treatment selectionshould be based on the clinician’s judgment

of the best possible match between the clinicalcircumstances (e.g., type of sleep problem,patient characteristics) and the individualproperties of currently available drugs (e.g.,onset and duration of action, safety, tolerabil-ity). This principle presumes some degreeof clinician familiarity with the pharmacologicprofile of the available sedatives/hypnoticscurrently available for use in the pediatricpopulation.

● Treatment must be diagnostically driven andbased on a careful clinical evaluation of thesymptoms and consideration of all possible dif-ferential diagnoses. It is incumbent upon theclinician to choose the most appropriate phar-macologic and/or behavioral therapies basedon the actual diagnosis rather than the symp-tom complex. For example, sleep initiationinsomnia may be due to a primarily behavioralsleep disorder (e.g., limit-setting sleep disor-der), a physiologically based sleep disorder(e.g., restless legs syndrome, delayed sleepphase syndrome), or a combination of thesetwo disorders (e.g., psychophysiologic insom-nia), each necessitating a different treatmentapproach.

● Sleep problems in infants and very young chil-dren are almost always related to developmentalasynchrony between the child’s sleep develop-ment and parental expectations (e.g., the develop-ment of nocturnal-diurnal sleep–wake rhythms,“sleeping through the night”). Therefore, med-ication is rarely if ever indicated in this agegroup.

● In almost all cases, medication is not the firsttreatment choice nor the sole treatment strat-egy. Medication use, except for very self-limit-ing circumstances such as travel, should beviewed only within the context of a more com-prehensive treatment plan.

● Medication should always be used in combina-tion with nonpharmacologic strategies (e.g.,behavioral interventions, parent education).This is analogous to a number of other condi-tions in children, such as attention deficithyperactivity disorder (ADHD), in whicha combination of pharmacologic and behav-ioral strategies are often superior to drug treat-ment alone.

● Before consideration of pharmacologic treat-ment, sleep hygiene should always be opti-mized. All sleep disorders in children may be


Ch11.qxd 01/28/05 10:30 AM Page 155

exacerbated by poor sleep habits, such as exces-sive caffeine use and irregular sleep-wakeschedules, which must be addressed beforemedication is recommended.

● Treatment goals should be realistic, clearlydefined, and discussed with and agreed uponby the family before treatment is initiated.In addition, there must be a clear plan for thefollowup and reassessment of therapeuticgoals. In particular, the parental expectationsregarding the degree of amelioration of thesleep problem by medication and the antici-pated duration of drug treatment should beclearly established.

● Medication should be used only for the shortterm; no prescription refills should be givenwithout reassessment of the target symptomsand assessment of patient compliance with bothpharmacologic and behavioral management.

● Adolescents should be screened for alcohol/drug use and pregnancy before the initiation oftherapy. Many recreational substances may havesynergistic clinical effects when combined withsedatives/hypnotics. In addition, hypnotics withhigh toxicity levels in overdose should be usedwith extreme caution in situations in whichthere is any risk of nonaccidental overdose.

● Patients should also be screened for the con-current use of self-initiated nonprescriptionsleep medications (e.g., Excedrin PM, mela-tonin, herbals). Some of these medications havesimilar ingredients (e.g., diphenhydramine isthe soporific ingredient in both Benadryl andTylenol PM); while generally viewed by par-ents as safe, the potential drug–drug interac-tions between most herbal preparations andsedatives/hypnotics are largely unknown.

● Medication selection, particularly in terms ofduration of action, should be appropriate to thepresenting complaint—that is, for problemswith sleep onset, a shorter-acting medicationis generally desirable. For problems with sleepmaintenance, longer-acting medications maybe considered but are more likely to result in“hangover” effects the following morning.

● All medications should be used with cautionand monitored closely for efficacy and sideeffects. Since there are so few data on thesafety and efficacy of these medications in chil-dren, a conservative approach similar to thatwhich should be exercised with any pediatricoff-label drug use is warranted.

● There are potential indications for hypnoticuse in otherwise normal children (generallyshort-term).

● The safety or welfare of the child is threat-ened—for example, the parent is overwhelmedand/or unable to implement nonpharmaco-logic interventions. Because of their rapidonset of action, in this situation medicationsmay assist in “breaking the cycle” and allow forthe implementation of effective behavioralstrategies.

● There is a failure of, or an inability to com-ply with, an adequate trial of accepted non-pharmacologic and/or behavioral treatment,for example, the older child or adolescentwith psychophysiologic insomnia who failsto respond to standard behavioral manage-ment tools such as stimulus control andsleep restriction.

● Medication is used as an adjunct to sleephygiene and/or chronotherapy in circadianrhythm disturbances, for example, in an oth-erwise normal adolescent with delayed sleepphase syndrome.

● The insomnia occurs in the setting of medicalillness with associated issues, including paincontrol, concomitant medications, and hospi-talization, for example, a child on steroids forchronic illness.

● The insomnia occurs in the context of an acutestressor, for example, a death in the family.

● The insomnia occurs or is anticipated in thecontext of travel, for example, a prolongedplane ride with an accompanying time change.

Contraindications to hypnotic use in otherwisenormal children:

● The insomnia occurs in the presence ofuntreated sleep-disordered breathing, for exam-ple, obstructive sleep apnea. Not only is hyp-notic medication often inappropriate for treatingthe underlying condition but also sedatives withrespiratory depressant properties (e.g., chloralhydrate) may be dangerous in the situation of acomorbid sleep-related breathing disorder.

● The insomnia is due to developmentally basednormal sleep behavior; for example, there areinappropriate expectations from the parents orpractitioner regarding the child’s sleep behavior.

● The insomnia is due to a self- limiting conditionthat temporarily results in night-time awaken-ings, for example, teething.


Ch11.qxd 01/28/05 10:30 AM Page 156

● There are potential drug interactions withconcurrent medications (e.g., opiates) orunrecognized substance abuse or alcohol use.

● There is a limited ability to follow up with andmonitor the patient; for example, the parentfrequently misses scheduled appointments.

PHARMACOLOGIC TREATMENT OF PEDIATRICINSOMNIA IN CHILDREN WITH SPECIAL NEEDS

Sleep disturbances are a prominent part of themorbidity of neurobehavioral, psychiatric, andchronic medical conditions. Whether it is theprimary condition or is secondary to the chroniccondition, pediatric insomnia may contribute tothe exacerbation of these conditions and have anadverse impact on the quality of life for the childand family. Children with chronic medical anddevelopmental conditions are particularly proneto insomnia because of the unique combinationsof family stresses, caregiver interactions, social(peer) relationships, medical needs, sedatingor activating medications, physical challenges,and psychiatric morbidity. For example, the fac-tors common to many medical conditions thatcan also create or exacerbate insomnia includepain, pruritis, cough, abnormal movements, andother disturbances. The physician must considereach of these potential factors, realizing that sev-eral may coexist and frustrate therapeutic inter-ventions.

Because the neural and cognitive requisitesto develop and express most sleep symptoms arebasic, children with chronic medical and devel-opmental conditions are susceptible to the sameemotional, behavioral, medical, and circadiansleep disorders as normal children. Therefore,these common causes of insomnia should beconsidered first in children with chronic healthconditions and developmental disabilities andtreated appropriately. As is the case with normalchildren, it is preferable to identify and controlthe underlying process that creates a sleep distur-bance instead of simply prescribing symptomatictherapy for insomnia. However, sleeplessnessmay overshadow other indicators of primarymedical and psychiatric disorders and thusappear to arise de novo in an otherwise healthychild. For example, insomnia is often a present-ing symptom of many psychiatric disorders,such as depression, post-traumatic stress disor-der, and generalized anxiety disorder. If they arenot carefully sought, other symptoms of psychi-

atric and medical conditions may be missed andan opportunity to eliminate the basis of theinsomnia would be lost.

Children with neurologic injury and specificgenetic, psychiatric, and behavioral syndromesand conditions are particularly susceptibleto specific types of insomnia. Examples includeautism and pervasive developmental disorder,blindness (circadian rhythm disorders), Smith-Magenis syndrome (severe insomnia), Williamssyndrome (periodic limb movement disorder),Rett syndrome (prolonged sleep onset latency,sleep fragmentation), and Tourette’s syndrome(increased nocturnal movements and awaken-ings). Children with attention deficit hyperactiv-ity disorder are often reported by parents to havesleep onset difficulties and restless sleep andpresent one of the more common chronic condi-tions for which sedatives are recommendedby pediatric practitioners. Children with asthmaand atopy, renal failure, and epilepsy are alsohighly prone to sleep disruption related to theunderlying chronic condition and/or medica-tion. Thus, the physician may be guided by theliterature relevant to the specific diagnosis.

The approach to insomnia in children withunderlying medical or developmental conditionsmust take into account the expected efficacy inlight of a patient’s behavioral strengths and chal-lenges, the capabilities and needs of familiesand caregivers, health care priorities or urgen-cies that may temporarily supersede long-terminterests, and the impact of insomnia therapy onunderlying medical conditions and concurrentmedications. For example, a teenager with mildmental retardation might not be able to takea hypnotic in a reliable manner; a family in cri-sis may need the rapid effect of a hypnotic untilbehavioral interventions take effect; and mela-tonin has the potential to lower the seizurethreshold in a child with an underlying seizuredisorder. In addition, because of frequent con-current use of other medications and the idio-syncratic response that these children may haveto sedatives/hypnotics, extreme caution shouldbe exercised with regard to the choice of med-ication and monitoring of side effects.

Therefore, the task force recommends that thespecific history about the quality of sleep be anintegral part of the evaluation and maintenancecare of children with chronic medical and/ormental health conditions. Pharmacologic therapy


Ch11.qxd 01/28/05 10:30 AM Page 157

should be considered for sleep problems asa part of the overall management strategy in con-junction with behavioral therapy and sleephygiene. Sleep problems in these children shouldbe managed aggressively to avoid exacerbationof the underlying condition and improve theoverall quality of life; a longer duration of drugtherapy is often necessary. Consultation witha pediatric neurologist, developmental/behav-ioral pediatrician, and/or sleep specialist is oftenwarranted.

● Pharmacologic management of primary insom-nia in childhood is rarely indicated. Behavioralmanagement is indicated in almost all cases.Nevertheless, certain situations exist that jus-tify pharmacologic treatment in some casesof primary childhood insomnia. These situa-tions include those children with developmen-tal delays, neurologic problems, and sensoryintegration issues.

● Pharmacologic management may also be indi-cated in some situations in which childhoodproblem sleeplessness is the primary and under-lying cause of family dysfunction.

● Once the situation indicates pharmacologicmanagement, treatment should be for the shortterm (<3 weeks).

● Pharmacologic management must be used inconjunction with other behavioral/family inter-ventions that should continue beyond the lim-ited use of medication.

The most common medications used for thislimited purpose are listed below.

● Diphenhydramine● Standard dosage: 2 to 6 years of age, 12.5 to

25 mg at bedtime; 6 to 12 years of age, 25 mgat bedtime; over 12 years of age, 25 to 50 mgat bedtime.

● Contraindications: Narrow-angle glaucoma,gastrointestinal or genitourinary obstruction.

● Main drug interactions: May interact withantihypertensive or antidepressant medica-tions; increased sedation when combinedwith alcohol.

● Main side effects: Nervousness, dizziness,hypertension, excitability, and drowsiness.

● Special points: Over-the-counter preparationsare often combined with other medications(e.g., pseudoephedrine) as decongestants.Elixirs contain 5% to 6% alcohol. Multiplepreparations are available. The patients’

parents should be instructed to read themedication label before purchase andadministration.

● Cost-effectiveness: 25 mg capsules, $0.03;12.5 mg elixir, $0.13 per teaspoon.

● Hydroxyzine● Standard dosage: 0.5 mg/kg orally 30 minutes

before bedtime.● Contraindications: Hypersensitivity to

hydroxyzine or any of its components.● Main drug interactions: May potentiate

other CNS depressants or anticholinergics,and can antagonize the vasopressor effects ofepinephrine.

● Main side effects: Hypotension, dizziness,headache, ataxia, dry mouth, urinary reten-tion, and weakness.

● Special points: Competes with histamine forH1-receptor sites on effector cells in the gas-trointestinal tract, blood vessels, and respi-ratory tract.

● Lorazepam● Standard dosage: 0.05 mg/kg at bedtime.● Contraindications: Hypersensitivity to ben-

zodiazepines; should not be used in patientswith preexisting CNS depression, respiratorydepression, narrow-angle glaucoma, orsevere and uncontrolled pain.

● Main drug interactions: Other CNS or respi-ratory center depressants may increase theadverse effects of lorazepam.

● Main side effects: Drowsiness, lethargy, hang-over effect, dizziness, transitory hallucina-tions, and ataxia.

● Special points: Use with caution in patientswith hepatic or renal impairment. Careshould be taken when used in patients withcompromised pulmonary function.

References

1. Ferber R: Sleeplessness in the child. In Kryger MH,Roth T, Dement WC (eds): Principles and Practiceof Sleep Medicine. Philadelphia, WB Saunders,1989, pp 633-639.

2. Anders TF: Night-waking in infants during thefirst year of life. Pediatrics 1979;63:860-864.

3. Carey WB: Night waking and temperament ininfancy. J Pediatr 1974;84:756-758.

4. Palmer CD, Harrison GA, Hiorns RW: Sleep pat-terns and life style in Oxfordshire villages. J BiosocSci 1980;12:437-467.

5. California Assessment Program: Student Achieve-ment in California Schools: 1979-1980. Annual


Ch11.qxd 01/28/05 10:30 AM Page 158

report, prepared under the direction of AlexanderI. Law, Chief, Office of Program Evaluation andResearch, 1980.

6. Hayashi Y: On the sleep hours of school childrenof 6 to 20 years. Psychological Abstracts 1927;1:439.

7. Bax MCO: Sleep disturbance in the young child.Br Med J 1980;280:1177-1179.

8. Weissbluth M, Poncher J, Given G, et al: Sleepduration and television viewing. J Pediatr 1981;99:486-488.

9. Jenkins S, Bax MCO, Hart, H: Behaviour problemsin preschool children. J Child Psychol Psychiatry1980;21:5-17.

10. Richman N, Stevenson JE, Graham PJ: Prevalenceof behaviour problems in 3-year-old children:An epidemiological study in a London borough.J Child Psychol Psychiatry 1975;16:277-287.

11. Anders TF: Night-waking in infants during thefirst year of life. Pediatrics 1979;63:860-864.

12. Metcalf D: The ontogenesis of sleep-awake statesfrom birth to 3 months. Electroencephalogr ClinNeurophysiol 1970;28:421.

13. Moore T, Ucko LE: Night waking in early infancy.Arch Dis Child 1957;32:333-342.

14. Zuckerman B, Stevenson J, Bailey V: Sleep prob-lems in early childhood: Continuities, predictivefactors, and behavioral correlates. Pediatrics 1987;80:664-671.

15. Kataria S, Swanson MS, Trevathan GE: Persistenceof sleep disturbances in preschool children.J Pediatr 1987;110:642-646.

16. Jones DPH, Verduyn CM: Behavioural manage-ment of sleep problems. Arch Dis Child 1983;58:442-444.

17. Wilks JM: The sleepless child (letter). Br Med J1977;2:704-705.

18. Weissbluth M: Sleep duration and infant tempera-ment. J Pediatr 1981;99:817-819.

19. Thomas A, Chess S, Birch HG: Temperament andBehavior Disorders in Children. New York, NYUPress, 1968.

20. Oberklaid F, Prior M, Sanson A, et al: Assessmentof temperament in the toddler age group.Pediatrics 1990;85:559-566.

21. Lounsbury ML, Bates JE: The cries of infants ofdiffering levels of perceived temperamental diffi-culties: Acoustic properties and effects on listen-ers. Child Dev 1982;53:677-686.

22. Snow ME, Jacklin CN, Maccoby EE: Cryingepisodes and sleep-wakefulness transitions in thefirst 26 months of life. Infant Behav Dev 1980;3:387-394.

23. Lozoff B, Wolf AW, Davis NS: Cosleeping in urbanfamilies with young children in the United States.Pediatrics 1984;74:171-182.

24. Monroe LJ: Transient changes in EEG sleep pat-terns of married good sleepers: The effects of alter-ing sleeping arrangement. Psychophysiology 1969;6:330-337.

25. Caudill W, Ploth D: Who sleeps by who. Psychiatry1964;32:12-43.

26. Ferber R: Solve Your Child’s Sleep Problems. NewYork, Simon and Schuster, 1985.

27. Ferber R, Boyle MP: Nocturnal fluid intake: A causeof, not treatment for, sleep disruption in infants andtoddlers. Sleep Res 1983;12:243-249.

28. Weissbluth M: Sleep and the colicky infant.In Guilleminault C (ed): Sleep and Its Disordersin Children. New York, Raven Press, 1987, pp129-141.

29. Illingworth RS: “Three months” colic. Arch DisChild 1954;29:167-174.

30. Wessel MA, Cobb JC, Jackson EB, et al: Paroxysmalfussing in infancy, sometimes called colic.Pediatrics 1954;14:421-434.

31. Pierce P: Delayed onset of “three months” colicin premature infants. Am J Dis Child 1948;75:190-192.

32. Breslow L: A clinical approach to infantile colic:A review of 90 cases. J Pediatr 1957;50:196-206.

33. Meyer JE, Thaler MM: Colic in low birth weightinfants. Am J Dis Child 1971;122:25-27.

34. Jorup S: Colonic hyperperistalsis in neurolabileinfants: Studies in so-called dyspepsia in breastfed infants. Acta Paediatr 1952;85:1-92.

35. Weissbluth M, Christoffel KK, Davis AT: Treatmentof infantile colic with dicyclomine hydrochloride.J Pediatr 1984;104:951-955.

36. Weissbluth M, Liu K: Sleep patterns, attentionspans, and infant temperament. J Dev Behav Pediatr1983;4:34-36.

37. Weissbluth M, Davis AT, Poncher J: Night wakingin 4- to 8-month-old infants. J Pediatr 1984;104:477-480.

38. Kahn A, Mozin MJ, Casimir G, et al: Insomniaand cow’s milk allergy in infants. Pediatrics 1985;76:880-884.

39. Ferber R: The sleepless child. In Guilleminault C(ed): Sleep and Its Disorders in Children. NewYork, Raven Press, 1987, pp 141-163.

40. Kaplan, BJ, McNicol J, Conte RA, Maghadam HK:Sleep disturbance in preschool-aged hyperactiveand nonhyperactive children. Pediatrics 1987;80:839-844.

41. Salzarulo P, Chevalier A: Sleep problems in chil-dren and their relationship with early distur-bances of the waking-sleeping rhythms. Sleep 1983;6:47-51.

42. Diagnostic and Statistical Manual of MentalDisorders, 3rd ed. Washington, DC, AmericanPsychiatric Association, 1980.


Ch11.qxd 01/28/05 10:30 AM Page 159

43. Greenhill L, Puig-Antich J, Goetz R, et al: Sleeparchitecture and REM sleep measures in pre-pubertal children with attention deficit disorderwith hyperactivity. Sleep 1983;6:91-101.

44. Busby K, Firestone P, Pivik RT: Sleep patterns inhyperkinetic and normal children. Sleep 1981;4:366-383.

45. Peters JE, Romaine JS, Dyckman RA: A specialneurological examination of children with learn-ing disabilities. Dev Med Child Neurol 1975;17:63-78.

46. Touwen BCL, Prechtl HFR: NeurologicalExamination of the Child with Minor NervousDysfunction. London, Spastics InternationalMedical Publications, 1970.

47. Kane J, Coble P, Conners K, Kupfer DJ: EEG sleepin a child with severe depression. Am J Psychiatry1977;134:813-814.

48. Cytryn L, McKnew D: Proposed classificationof childhood depression. Am J Psychiatry 1972;129:149-155.

49. Poznanski E, Frull JP: Childhood depression:Clinical characteristics of overtly depressed chil-dren. Arch Gen Psychiatry 1970;23:8-15.

50. Sander LW, Stechler G, Burns P, et al: Earlymother-infant interaction and 24-hour patterns ofactivity and sleep. J Am Acad Child Psychiatry1970;9:103-123.

51. Brown GW, Harris T: Social Origins of Depression.London, Tavistock Publications, 1978.

52. Lozoff B, Wolf AW, Davis NS: Sleep problems seenin pediatric practice. Pediatrics 1985;75:477-483.

53. Association of Sleep Disorders Centers: DiagnosticClassification of Sleep and Arousal Disorders, 1sted. Prepared by the Sleep Disorders ClassificationCommittee, H.P. Roffwarg, Chairman. Sleep 1979;2:1-137.

54. Mindell JA, Carskadon MA, Owens JA: Develop-mental features of sleep. Child Adolesc PsychiatrClin N Am 1999;8:695-725.

55. Owens J, Spirito A, McGuinn M, Nobile C: Sleephabits and sleep disturbance in school-aged chil-dren. J Dev Behav Pediatr 2000;21:27-36.

56. Gianotti F, Cortesi F. Sleep patterns and daytimefunctions in adolescents: An epidemiological sur-vey of Italian high school student population. InCarskadon MA (ed): Adolescent Sleep Patterns:Biological, Social, and Psychological Influences.New York, Cambridge University Press, 2002.

57. Mindell J, Owens J: Sleep in the pediatric practice.In Mindell J, Owens J (eds): A Clinical Guide toPediatric Sleep: Diagnosis and Management ofSleep Problems in Children and Adolescents.Philadelphia, Lippincott-Williams-Wilkins, 2003,pp 1-10.


Ch11.qxd 01/28/05 10:30 AM Page 160

Pediatric sleep medicine once focused mainly onnocturnal behavior: parasomnias, bedtime ritu-als, and interactions between children and theirparents. More recently, clinicians have realizedthat intrinsic dyssomnias such as obstructivesleep apnea, upper airway resistance syndrome,restless legs syndrome (RLS), and periodic limbmovement disorder (PLMD) also occur in chil-dren and may have important effects on daytimebehavior. Furthermore, these effects differ fromthose seen in adults with the same disorders.Whereas adults often display overt, excessivedaytime sleepiness, children are more likely topresent with cognitive and behavioral morbidity.For example, some affected children show inat-tentive and hyperactive behaviors that lead to adiagnosis of attention deficit hyperactivity disor-der (ADHD) and treatment with stimulantsbefore the underlying sleep problem is recog-nized. This chapter will highlight evidence thatabnormal daytime behaviors, many reminiscentof ADHD, are among the most important out-comes of childhood sleep disorders. Data thathave challenged or qualified this hypothesis alsowill be discussed.

SLEEP, ADHD, AND POTENTIALPATHOPHYSIOLOGIC LINKS

The American Psychiatric Association’s Diagnosticand Statistical Manual of Mental Disorders (DSM)once listed disturbed sleep as one of the diagnosticfeatures of ADHD. Although the current editionhas dropped this criterion, the parents of childrenwith ADHD commonly note disturbed and rest-less sleep in their children.1-4 In practice, sleepdisruption is often attributed to stimulant med-ications or to persistence of daytime hyperactivityinto nocturnal hours. Monitoring of movement(actigraphy) during 5-day intervals in childrenwith ADHD did not demonstrate nocturnal over-

activity but did show relative instability of sleeponset and total sleep time in comparison withcontrols.5 Polysomnographic studies in ADHDchildren did not identify specific primary sleepdisorders.6-11 A review of polysomnographic stud-ies of ADHD,12 as well as a subsequent video-polysomnographic study,13 failed to identifyconsistent abnormalities except for increasedmovements during sleep. However, these studiesusually were performed with a focus on sleepstaging; they did not record breathing or legmovement data later found to be as important inchildren as they are in adults. Clinicians andresearchers without subspecialty interest in sleepmedicine often do not consider the possibilitythat primary sleep disorders could contribute toADHD or related symptoms.14-17 However, grow-ing evidence suggests that sleep disorders,because of associated daytime sleepiness or otherconsequences, could promote inattentive andhyperactive behaviors.

Several types of studies indicate that hyperac-tive children tend to have deficient levels ofarousal. More than 60 years ago, Bradley andBowen observed that stimulants rather thansedatives provide effective treatment for hyper-activity.18,19 In contrast, sedatives can worsen thebehavior. Early electrophysiologic studies—based on skin conductance, EEG, and auditoryand sensory evoked responses—also suggestedthat children with ADHD show hypoarousalrather than hyperarousal.20 Recent multiplesleep latency tests among 26 ADHD children and21 controls showed increased daytime sleepinessin the ADHD subjects,21 and a survey of the par-ents of more than 800 children at general pedi-atrics clinics suggested a substantial associationbetween sleepiness and behavior that character-izes ADHD.22

The behaviors that define ADHD may be aproduct of cognitive changes, particularly deficits

Attention Deficit, Hyperactivity, andSleep Disorders

Ronald D. Chervin

12

161

Ch12.qxd 1/25/05 3:21 AM Page 161

in executive functioning,23 that can result fromdisordered sleep or daytime sleepiness. In adults,cognitive abilities are sensitive to sleep disor-ders,24-26 and executive functioning, whichincludes working memory, decision making, andthe ability to apply old information to new chal-lenges, is especially vulnerable. Research in chil-dren has been limited but does include a study of16 children randomized to either 5 or 11 hours ofsleep for one night: On the following day, thesleep-deprived group showed impairment of ver-bal creativity and abstract thinking.27 Anotherstudy of 82 children randomized to either 4 or 10hours of sleep for one night showed increasedinattentive behaviors after sleep deprivation butnot hyperactive-impulsive behavior or impairedperformance on tests of response inhibition andsustained attention.28

Cognitive deficits after sleep disruption maylocalize to the prefrontal cortex.29 Adults with

obstructive sleep apnea show neuropsychologi-cal deficits, particularly in executive function-ing, that can be attributed to the prefrontalcortex and sometimes improved by treatment.30

Preliminary data from functional magnetic reso-nance imaging in adults with obstructive sleepapnea show decreased activity of the dorsolateralprefrontal cortex during performance of a work-ing memory task and restoration of normalactivity levels after treatment.31 Prefrontal corti-cal activity is also affected by experimental sleepdeprivation in adults.32-34 These data combine tosuggest a potential mechanism for cognitive andbehavioral changes in childhood dyssomnias:Insufficient or inadequate sleep may contributeto prefrontal cortical dysfunction and impairexecutive functions such as behavioral inhibi-tion, emotional regulation, and working mem-ory and thereby promote inattentive andhyperactive behaviors (Fig. 12–1).35

162 Chapter 12 Attention Deficit, Hyperactivity, and Sleep Disorders

Sleepdisruption

Intermittenthypoxia andhypercarbia

Disruption of restorativefeatures of sleep

Prefrontal cortical

Dysfunction of cognitive executive

Behavioralinhibition

Setshifting

Self-regulation ofaffect and arousal

Workingmemory

Adverse daytime effects

Problems mentally manipulating informationPoor planning and haphazard execution of plansDisorganizationPoor judgment/decision-makingRigid thinkingDifficulty maintaining attention and motivationEmotional lability (“mood swings”)Overactivity/impulsivity (especially in children)

Analysis/synthesis

Contextualmemory

Disruption of cellular orchemical homeostasis

Figure 12–1. A model by which sleep disruption or physiologic effects of sleep apnea could alter day-time behavior in children. (Reprinted with permission from Beebe DW, Gozal D: Obstructive sleepapnea and the prefrontal cortex: Towards a comprehensive model linking nocturnal upper airwayobstruction to daytime cognitive and behavioral deficits. J Sleep Res 2002;11:1-16.)

Ch12.qxd 1/25/05 3:21 AM Page 162

Other pathophysiologic mechanisms maypertain to specific sleep disorders. Obstructivesleep apnea is associated with hypoxemia, whichcould affect brain development. A model of sleepapnea, produced by subjecting young rats tointermittent hypoxia, has produced evidence ofhippocampal and cortical apoptosis.36 Althoughthe histologic changes are transient, associateddeficits in a spatial learning task do not com-pletely resolve 2 weeks after the insult.Furthermore, rats seem to have an age-specificwindow of vulnerability that may have impor-tant analogies in human development: Fifteen-to 20-day-old animals show prominent neuronalapoptosis after intermittent hypoxia, whereasthe effect is much less at younger and olderages.37 Although it is attractive as a potentialexplanation for the neurobehavioral problemsseen in obstructive sleep apnea, hypoxemia can-not explain similar problems seen in childrenwith upper airway resistance syndrome or othersleep disorders.38

COGNITIVE AND BEHAVIORALCHANGES IN CHILDREN WITHSPECIFIC SLEEP DISORDERS

Obstructive Sleep ApneaIn 1976, the first published series of childrenwith obstructive sleep apnea described learningdifficulties, inattention, and hyperactivity asprominent daytime symptoms.39 Subsequentseries in the early 1980s re-emphasized the fre-quency of these behaviors, noted that some ofthese children had been diagnosed with ADHDbefore presentation to a sleep center, and foundthat the symptoms of ADHD improved or disap-peared after treatment for sleep-disorderedbreathing.38,40 Although no subsequent random-ized, placebo-controlled, double-blind studiestested these observations, one study did use aquasiexperimental design: Twelve children withsleep-disordered breathing and scheduled toundergo tonsillectomy were compared with 11primary snorers also scheduled for tonsillectomyand with 10 control subjects, most of whom hadbeen scheduled for unrelated surgery.41

Postoperatively, validated parental scales foraggression, inattention, and hyperactivity and acontinuous performance test all showedimprovement in the group with sleep-disordered

breathing but no change in the control group.Children with primary snoring showed improve-ment in some areas.

Another study used a validated pediatric sleepquestionnaire to compare 27 ADHD patients at achild psychiatric clinic, 43 patients with otherdiagnoses at the same clinic, and 73 children ata general pediatric clinic.42 The ADHD childrenwere three times as likely as the non-ADHD psy-chiatric patients and the general pediatricpatients to have a history of habitual snoring.Habitual snoring explained 16% of the variancein a DSM-IV–based inattention and hyperactivityscale. In a subsequent study, the parents of 866children attending two general pediatric clinicscompleted a validated 22-item pediatric sleep-disordered breathing scale and the hyperactivityindex of the Conners’ Parent Rating Scale(CPRS).22 High CPRS scores, defined as greaterthan 1 standard deviation (SD) above normal,were found in 13% (95% confidence interval[CI], 11, 16) of all subjects, 22% (95% CI, 15,29) of habitual snorers, and 12% (95% CI, 9, 14)of nonsnorers. High CPRS scores were associ-ated with high overall sleep-disordered breath-ing scores (P < .01). Stratification by age and sexshowed that most of the association betweenhyperactivity and snoring derived from boys lessthan 8 years old (Table 12–1). These findingssuggest that if sleep-disordered breathing doescontribute to hyperactive behavior, young boysmay be the most vulnerable to this effect.

In an evidence-based review, an AmericanAcademy of Pediatrics subcommittee found 12publications that evaluated the associationbetween sleep-disordered breathing and a rangeof cognitive and behavioral problems.43 Noreport was a randomized trial, and each was across-sectional or cohort study or a case series.When the authors pooled the results of six cross-sectional studies, they concluded that the com-bined odds ratio (OR) for neurobehavioralproblems and snoring in children was 2.93(2.23, 3.83).

Published reports that hyperactivity is partic-ularly common in children with sleep-disorderedbreathing generally have not prospectivelyevaluated children using DSM-IV criteria. Thepreliminary results of one study did include psy-chiatric evaluations of 5- to 13-year-old childrenscheduled for adenotonsillectomy (AT) and acontrol group scheduled for general surgery.44

Attention Deficit, Hyperactivity, and Sleep Disorders Chapter 12 163

Ch12.qxd 1/25/05 3:21 AM Page 163

Nearly all of the AT children were suspectedby their otolaryngologists of having sleep-disordered breathing, although many also hadadditional surgical indications. Although thechild psychiatrists who evaluated the childrenwere not blinded to surgical group, they did usetwo well-validated structured interviews, theDiagnostic Interview Schedule for Children andthe Children’s Psychiatric Ratings Scale. A dis-ruptive behavior disorder—ADHD, oppositionaldefiant disorder, or conduct disorder—was diag-nosed in 20 (44%) of 45 AT children but in0 (0%) of 8 control subjects (Fisher’s exact test;P = .019). Among the AT children, 7 (16%)had the ADHD-inattentive subtype, 2 (4%) hadthe ADHD-hyperactive subtype, 6 (13%) had theADHD-combined type, and 9 (20%) had opposi-tional defiant disorder. The children were alsoevaluated for several other major DSM-IV diag-noses—social phobia, seasonal affective disorder,generalized anxiety disorder, obsessive-compulsivedisorder, post-traumatic stress disorder, enure-sis, encopresis, tic disorder, major depressivedisorder, dysthymia, mania, hypomania, andpervasive development disorder—but noneshowed increased frequency in the AT group.These data suggest that DSM-IV–based diag-noses of disruptive behavior disorders are partic-ularly common among children scheduled forAT, most of whom carry a clinical diagnosis ofsleep-disordered breathing.

In addition to these behavioral problems, cog-nitive changes in children with sleep-disorderedbreathing have been the subject of a small butgrowing body of literature. One study focusedon 16 obese children, among whom 5 hadobstructive sleep apnea on polysomnography.45

The children with sleep apnea demonstrateddeficits in learning, memory, and vocabulary, andin the entire sample, apnea severity correlatedinversely with memory and learning perform-ance. A second study compared neurocognitivefunctioning in 16 childhood snorers, who hadlittle or no sleep apnea, to that in 16 nonsnor-ers.46 The snorers showed lower attention, mem-ory, and intelligence scores but no differences insocial competency and problematic behavior.A third study of 28 children with obstructive sleepapnea and 10 control subjects found preliminaryevidence that the former group had significantlymore somatic complaints, mood problems, anddifficulty relating to peers.47 However, manymeasures showed no differences betweengroups, and children with mild sleep apneashowed more behavioral and emotional prob-lems overall than those with severe sleep apnea.Finally, data from the study of children sched-uled for AT or other procedures show thatalthough the two groups did not differ signifi-cantly with respect to age, grade, verbal IQ, andperformance IQ, the AT children performedmore poorly on academic achievement tests,


Regression Estimated OddsAge and Gender (N) Coefficient (SE) Ratio (95% CI) PAll subjects (866) 0.81 (0.25) 2.2 (1.4, 3.6) .0011

Boys (469) 1.05 (0.31) 2.8 (1.5, 5.2) .0008Girls (397) 0.39 (0.43) 1.5 (0.6, 3.3) .3633

Age < 8 yr (565) 1.02 (0.30) 2.8 (1.5, 4.9) .0007Age ≥ 8 yr (301) 0.32 (0.46) 1.4 (0.5, 3.3) .4831

Boys < 8 yr (295) 1.46 (0.38) 4.3 (2.0, 9.1) <.0001Remaining children (571) 0.33 (0.35) 1.4 (0.7, 2.7) .3500

*Results adjusted for age and sex. Hyperactive behavior defined as Conners’ Parent Rating Scale hyperactivity index;T score > 60.

CI, confidence interval; SE, standard error.Data from Chervin RD, Archbold KH, Dillon JE, et al: Inattention, hyperactivity, and symptoms of sleep-disordered

breathing. Pediatrics 2002;109:449-456.

Table 12–1. Conditional Logistic Regression Coefficients and Odds Ratios forHyperactive Behavior and Habitual Snoring (Present vs Absent) in aStudy of Children Aged 5 to 13 Years*

Ch12.qxd 1/25/05 3:21 AM Page 164

including mathematical reasoning, spelling, andnumeric operations.48 A similar trend emergedfor basic reading performance, though not for ameasure of reading comprehension. In aggre-gate, these studies suggest specific academicdeficits in children with sleep-disordered breath-ing that would predict an important practicalimpact, including poor school performance.

School performance was the main outcomevariable in one study that enrolled 297 poorlyperforming first-grade children and screenedthem for sleep-disordered breathing by parentalquestionnaire, home oximetry, and capnogra-phy: Fifty-four children (18%) were identified aslikely to have sleep-disordered breathing.49 Theparents of these children were advised to seekfurther evaluation and possible treatment. Oneyear later, 24 (44%) of the 54 children had beendiagnosed and treated for obstructive sleepapnea. These children showed significantimprovement in their school grades, whereas the30 children with untreated sleep-disorderedbreathing and the 243 children without evidenceof sleep-disordered breathing showed noimprovement. In a subsequent survey of 1,588seventh and eighth graders, students whose per-formances ranked them in the bottom quartile oftheir classes were two to three times as likely astop-quartile students to have had frequent andloud snoring between the ages of 2 and 6 years.50

Finally, a study performed among 201 medicalstudents found that the failure rate for a year-endexamination was 13% for 78 nonsnorers, 22%for 99 occasional snorers, and 42% for 24 fre-quent snorers (P < .01).51

Restless Legs Syndrome andPeriodic Limb Movement DisorderAlthough the prevalence of RLS in children isnot well defined, more than one third of 138adults with this condition reported that symp-toms had begun before the age of 10.52 Earlystudies of restless legs in children found thatmany patients had problems with inattentionand hyperactivity.53,54 One report found that 117(91%) of 129 children referred to a sleep labora-tory and confirmed to have periodic leg move-ments during sleep (>5 movements per hour)also met the criteria for ADHD.55 Additional datasuggested that children with ADHD may bemore likely to have restless legs and related com-

plaints than other children.56,57 These studiesfocused on children referred for sleep or behav-ioral concerns, but subsequent data collected atgeneral pediatric clinics also suggested promi-nent associations between hyperactive behaviorand symptoms that characterize RLS and peri-odic leg movements during sleep.58 Hyper-activity was assessed with the CPRS and periodicleg movements, restless legs, and growing painsby a validated subscale of the Pediatric SleepQuestionnaire (Fig. 12–2).59 High hyperactivityindices (>1 SD elevation) were found in 13% ofall 866 subjects, 18% of children who haddescribed “restlessness of the legs when in bed,”and 11% of those without restless legs (P < .05).The OR for a high hyperactivity score and a 1 SDincrease in the periodic leg movement subscalewas 1.6 [1.4, 1.9], and this association retainedsignificance after statistical adjustment forsleepiness, snoring, restless sleep in general, orstimulant use.


100

80

60

40

0.00–0.25(n = 670)

0.26–0.50(n = 147)

0.51–0.75(n = 41)

PLMS score

HI

0.76–1.00(n = 8)

Figure 12–2. Among 866 children aged 5 to 13 years,a validated questionnaire subscale for symptoms of rest-less legs syndrome and periodic leg movements duringsleep (PLMS Score) displayed a dose–response associa-tion with the hyperactivity index (HI). Box plots showmedians and 10th, 25th, 75th, and 90th percentiles. Thehorizontal black stripe represents the mean. (Reprintedwith permission from Chervin RD, Archbold KH, Dillon JE,et al: Associations between symptoms of inattention,hyperactivity, restless legs, and periodic leg movements.Sleep 2002;25:213-218.)

Ch12.qxd 1/25/05 3:21 AM Page 165

Further support for the possibility that RLS orperiodic leg movements might contribute toADHD was provided by a report that sevenaffected children experienced improvement inbehavior, as assessed by well validated measures,when their RLS was treated with levodopa or adopamine agonist.60 However, as the authorsnoted, these medications could have improvedbehavior by a direct effect on brain mechanismsthat control behavior rather than through aneffect on RLS.

Other Sleep DisordersChildren with sleep disorders other than sleep-disordered breathing or RLS also may suffer fromcognitive and behavioral consequences. A casereport of a child with delayed sleep phase syn-drome carefully documented attention deficitthat improved considerably when the circadianrhythm disorder was treated.61 Children withnarcolepsy often have problems with inattentionand hyperactivity that may improve upon treat-ment.62,63 In the United States, adolescents areoften chronically sleep deprived. Those who per-form poorly academically tend to obtain lesssleep at night than those who receive bettergrades.64 Recent efforts to augment high schoolstudents’ nocturnal sleep and address their ten-dency to have delayed sleep phases by delayingschool start times in some Minnesota school dis-tricts have been associated with longer times inbed and slightly higher grades.

DO SLEEP DISORDERS CAUSECOGNITIVE AND BEHAVIORALMORBIDITY IN CHILDREN?

The evidence presented above suggests thatsleep disorders may contribute to cognitiveimpairment and disruptive behavior in children.However, these studies do not constitute ran-domized, double-blind, placebo-controlled tri-als, and proof is still lacking for the hypothesisthat substantial minorities of the many childrenwith disruptive behavior disorders have undiag-nosed sleep disorders as a contributing factor. Inaddition, some studies have raised questions ormethodological concerns about demonstratedlinks among sleep, cognition, and behavior.

One study of 113 children, aged 2 to 18 yearsand referred for suspected sleep-disordered

breathing to a sleep laboratory, found that the59 children confirmed to have sleep-disorderedbreathing did show high hyperactivity scores,but these scores were no higher than those ofthe 54 children without sleep-disorderedbreathing.65 Furthermore, hyperactivity levelsshowed no significant association with the rateof hypopneas and apneas, minimum oxygen sat-uration, or most negative esophageal pressure(investigated in a subset of 19 children).Instead, the level of hyperactivity was associatedwith the presence or absence of five or moreperiodic leg movements per hour of sleep (P = .02).The rate of periodic leg movements duringsleep showed a linear association with hyperac-tivity among those subjects with sleep-disor-dered breathing (P = .002) but no associationamong those without sleep-disordered breath-ing (P = .64). The association between behaviorand periodic leg movements rather than apnea,even in a population referred for suspectedsleep-disordered breathing, raises some doubtabout whether sleep-disordered breathingcauses disruptive behavior. Corroborativethough preliminary evidence has emerged froma second study.66 Such data are similar to thoseobtained from adults with sleep apnea, in whomthe major behavioral outcome—sleepiness—shows little correlation with apnea severity.67,68

One possible explanation for a correlationbetween periodic leg movements and hyperac-tive behavior only among children with sleep-disordered breathing is that this disorder acts asan effect modifier. However, the lack of astronger, direct association between apneaseverity and hyperactivity in children raisessome doubt about whether sleep apnea directlycauses this behavior.

Furthermore, studies that have reported asso-ciations between sleep-disordered breathing andeither poor school performance or hyperactivebehavior have been correlational and could notaccount for many potential confounding fac-tors.22,42,49-51 For example, recent data from 146second and fifth grade students in Ypsilanti,Michigan, showed significant associationsbetween school performance and symptoms ofsleep disordered breathing before but not aftersocioeconomic status was taken into account.69

These results raise the possibility that previousstudies of cognitive and behavioral problems inchildhood sleep disorders also would have


Ch12.qxd 1/25/05 3:21 AM Page 166

reached different conclusions if they hadaccounted for socioeconomic status.

CONCLUSIONS

In short, increasing evidence suggests that a vari-ety of childhood sleep disorders are associatedwith inattention, hyperactivity, and underlyingcognitive impairment that could have signifi-cantly adverse effects on such important out-comes as development and school performance.However, proof that sleep disorders cause neu-robehavioral deficits in substantial numbers ofchildren is still lacking, and some studies havegenerated conflicting data. Mechanisms bywhich sleep disorders may cause cognitive andbehavioral changes have been proposed but notyet thoroughly investigated.

If sleep disorders do influence cognition andbehavior, then this morbidity may constitute animportant public health problem. Problemsrelated to sleep are highly prevalent among chil-dren70 but seldom discussed, recognized, ortreated at general pediatric clinics.71 Further-more, those children who are suspected of havingsleep-disordered breathing usually undergo ATwithout having any formal polysomnographic,psychiatric, or cognitive testing.72 Argumentsthat preoperative evaluation should be moreextensive, as now suggested by the AmericanAcademy of Pediatrics,73 will be strengthened iffuture research more effectively shows cause-and-effect relationships between sleep disordersand neurobehavioral consequences, and if futuredata more convincingly link specific polysomno-graphic findings with these important adversehealth outcomes.74

References

1. Ring A, Stein D, Barak Y, et al: Sleep disturbancesin children with attention-deficit /hyperactivitydisorder: A comparative study with healthy sib-lings. J Learn Disabil 1998;31:572-578.

2. Owens JA, Maxim R, Nobile C, et al: Parental andself-report of sleep in children with attention-deficit / hyperactivity disorder. Arch PediatrAdolesc Med 2000;154:549-555.

3. Marcotte AC, Thacher PV, Butters M, et al:Parental report of sleep problems in children withattentional and learning disorders. J Devel BehavPediatr 1998;19:178-186.

4. Corkum P, Tannock R, Moldofsky H, et al:Actigraphy and parental ratings of sleep in chil-

dren with attention-deficit /hyperactivity disorder(ADHD). Sleep 2001;24:303-312.

5. Gruber R, Sadeh A, Raviv A: Instability of sleeppatterns in children with attention-deficit /hyper-activity disorder. J Am Acad Child AdolescPsychiatry 2000;39:495-501.

6. Busby K, Firestone P, Pivik RT: Sleep patterns inhyperkinetic and normal children. Sleep1981;4:366-383.

7. Greenhill L, Puig-Antich J, Goetz R, et al: Sleeparchitecture and REM sleep measures in pre-pubertal children with attention deficit disorderwith hyperactivity. Sleep 1983;6:91-101.

8. Haig JR, Schroeder CS, Schroeder SR: Effectsof methylphenidate on hyperactive children’ssleep. Psychopharmacologia (Berl) 1974;37:185-188.

9. Small A, Hibi S, Feinberg I: Effects of dextroam-phetamine sulfate on EEG sleep patterns ofhyperactive children. Arch Gen Psychiatry1971;25:369-380.

10. Palm L, Persson E, Bjerre I, et al: Sleep and wake-fulness in preadolescent children with deficits inattention, motor control and perception. ActaPaediatr 1992;81:618-624.

11. Platon MJR, Bueno AV, Sierra JE, Kales S:Hypnopolygraphic alterations in attention deficitdisorder (ADD) children. Intern J Neurosci1990;53:87-101.

12. Corkum P, Tannock R, Moldofsky H: Sleep dis-turbances in children with attention-deficit/hyperactivity disorder. J Am Acad Child AdolescPsychiatry 1998;37:637-646.

13. Konofal E, Lecendreux M, Bouvard M, Mouren-Simeoni MC: High levels of nocturnal activity inchildren with attention-deficit hyperactivity dis-order: A video analysis. Psych Clin Neurosci2001;55:97-103.

14. Zametkin AJ, Ernst M: Problems in the manage-ment of attention-deficit-hyperactivity disorder.N Engl J Med 1999;340:40-46.

15. Chervin RD: Attention-deficit-hyperactivity disor-der (Letter). N Engl J Med 1999;340:1766-1767.

16. Biederman J: A 55-year-old man with attention-deficit /hyperactivity disorder. JAMA 1998;280:1086-1092.

17. Yuen KM, Pelayo R: Sleep disorders and atten-tion deficit /hyperactivity disorder. JAMA 1999;281:797.

18. Bradley C: The behavior of children receivingBenzedrine. Am J Psychiat 1937;94:577-585.

19. Bradley C, Bowen M: Amphetamine (Benzedrine)therapy of children’s behavior disorders. AmJ Orthopsychiat 1941;11:92-103.

20. Satterfield JH, Cantwell DP, Satterfield BT:Pathophysiology of the hyperactive child syn-drome. Arch Gen Psychiatry 1974;31:839-844.


Ch12.qxd 1/25/05 3:21 AM Page 167

21. Lecendreux M, Konofal E, Bouvard M, et al: Sleepand alertness in children with ADHD. J ChildPsychol Psychiatry 2000;41:803-812.

22. Chervin RD, Archbold KH, Dillon JE, et al:Inattention, hyperactivity, and symptoms of sleep-disordered breathing. Pediatrics 2002;109:449-456.

23. Barkley RA: Genetics of childhood disorders,XVII: ADHD, Part 1: The executive functions andADHD. J Am Acad Child Adolesc Psychiatry2000;39:1064-1068.

24. Naegele B, Thouvard V, Pepin JL, et al: Deficits ofcognitive executive functions in patients withsleep apnea syndrome. Sleep 1995;18:43-52.

25. Engleman HM, Kingshott RN, Martin SE, DouglasNJ: Cognitive function in the sleep apnea / hypop-nea syndrome (SAHS). Sleep 2000;23(Suppl4):S102-S108.

26. Kim HC, Young T, Matthews CG, et al: Sleep-disordered breathing and neuropsychologicaldeficits: A population-based study. Am J RespirCrit Care Med 1997;156:1813-1819.

27. Randazzo AC, Muehlbach MJ, Schweitzer PK,Walsh JK: Cognitive function following acutesleep restriction in children ages 10-14. Sleep1998;21:861-868.

28. Fallone G, Acebo C, Arnedt JT, et al: Effects ofacute sleep restriction on behavior, sustainedattention, and response inhibition in children.Percept Mot Skills 2001;93:213-229.

29. Dahl RE. The impact of inadequate sleep on chil-dren’s daytime cognitive function. Semin inPediatr Neurol 1996;3:44-50.

30. Feuerstein C, Naegele B, Pepin JL, Levy P: Frontallobe-related cognitive functions in patients withsleep apnea syndrome before and after treatment.Acta Neurologica Belgica 1997;97:96-107.

31. Thomas RJ, Rosen BR, Bush G, Kwong KK:Working memory in obstructive sleep apnea: Afunctional magnetic resonance imaging study. SocNeurosci Abstr 2001. 27:845-849.

32. Drummond SP, Brown GG, Gillin JC, et al: Alteredbrain response to verbal learning following sleepdeprivation. Nature 2000;403:655-657.

33. Drummond SP, Gillin JC, Brown GG: Increasedcerebral response during a divided attention taskfollowing sleep deprivation. J Sleep Res2001;10:85-92.

34. Drummond SP, Brown GG, Stricker JL, et al: Sleepdeprivation-induced reduction in cortical func-tional response to serial subtraction. Neuroreport1999;10:3745-3748.

35. Beebe DW, Gozal D: Obstructive sleep apnea andthe prefrontal cortex: Towards a comprehensivemodel linking nocturnal upper airway obstruc-

tion to daytime cognitive and behavioral deficits.J Sleep Res 2002;11:1-16.

36. Gozal D, Daniel JM, Dohanich GP: Behavioral andanatomical correlates of chronic episodic hypoxiaduring sleep in the rat. J Neurosci 2001;21:2442-2450.

37. Gozal E, Row BW, Schurr A, Gozal D:Developmental differences in cortical and hip-pocampal vulnerability to intermittent hypoxia inthe rat. Neurosci Lett 2001;305:197-201.

38. Guilleminault C, Winkle R, Korobkin R,Simmons B: Children and nocturnal snoring—evaluation of the effects of sleep related respira-tory resistive load and daytime functioning. Eur JPediatr 1982;139:165-171.

39. Guilleminault C, Eldridge F, Simmons FB: Sleepapnea in eight children. Pediatrics 1976;58:23-30.

40. Guilleminault C, Korobkin R, Winkle R: A reviewof 50 children with obstructive sleep apnea syn-drome. Lung 1981;159:275-287.

41. Ali NJ, Pitson D, Stradling JR: Sleep disorderedbreathing: Effects of adenotonsillectomy onbehaviour and psychological functioning. EurJ Pediatr 1996;155:56-62.

42. Chervin RD, Dillon JE, Bassetti C, et al:Symptoms of sleep disorders, inattention, andhyperactivity in children. Sleep 1997;20:1185-1192.

43. Schechter MS, Section on Pediatric Pulmonology,Subcommittee on Obstructive Sleep ApneaSyndrome. Technical report: Diagnosis and man-agement of childhood obstructive sleep apneasyndrome. Pediatrics 2002;109:e69.

44. Dillon JE, Ruzicka DL, Champine DJ, et al: DSM-IV diagnoses in children scheduled for adenoton-sillectomy or hernia repair (Abstract). Sleep 2002;25:A346-A341.

45. Rhodes SK, Shimoda KC, Wald LR, et al:Neurocognitive deficits in morbidly obese chil-dren with obstructive sleep apnea. J Pediatr1995;127:741-744.

46. Blunden S, Lushington K, Kennedy D, et al:Behavior and neurocognitive performance in chil-dren aged 5-10 years who snore compared to con-trols. J Clin Exp Neuropsychol 2000;22:554-568.

47. Lewin DS, Rosen RC, England SJ, Dahl RE:Preliminary evidence of behavioral and cognitivesequelae of obstructive sleep apnea in children.Sleep Med 2002;3:5-13.

48. Huffman JL, Giordani B, Layne JR, et al: Academicachievement and attention in children scheduledfor adenotonsillectomy in comparison to controls(Abstract). Sleep 2002;25:A197-A198.

49. Gozal D: Sleep-disordered breathing and school per-formance in children. Pediatrics 1998;102: 616-620.


Ch12.qxd 1/25/05 3:21 AM Page 168

50. Gozal D, Pope DW Jr: Snoring during early child-hood and academic performance at ages thirteento fourteen years. Pediatrics 2001;107:1394-1399.

51. Ficker JH, Wiest GH, Lehnert G, et al: Are snor-ing medical students at risk of failing theirexams? Sleep 1999;22:205-209.

52. Walters AS, Hickey K, Maltzman J, et al: A ques-tionnaire study of 138 patients with restless legssyndrome: The ‘Night-Walkers’ survey. Neurology1996;46:92-95.

53. Picchietti DL, Walters AS: Severe periodic limbmovement disorder in childhood and adoles-cence. Sleep Res 1996;25:333.

54. Picchietti DL, Walters AS: Restless legs syndromeand periodic limb movement disorder in childrenand adolescents: comorbidity with attention-deficit hyperactivity disorder. Child AdolescPsychiatr Clin North Am 1996;5:729-740.

55. Picchietti DL, Walters AS: Moderate to severeperiodic limb movement disorder in childhoodand adolescence. Sleep 1999;22:297-300.

56. Picchietti DL, England SJ, Walters AS, et al:Periodic limb movement disorder and restlesslegs syndrome in children with attention-deficit-hyperactivity disorder. J Child Neurol 1998;13:588-594.

57. Picchietti DL, Underwood DJ, Farris WA, et al:Further studies on periodic limb movement dis-order and restless legs syndrome in children withattention-deficit hyperactivity disorder. MoveDisord 1999;14:1000-1007.

58. Chervin RD, Archbold KH, Dillon JE, et al:Associations between symptoms of inattention,hyperactivity, restless legs, and periodic leg move-ments. Sleep 2002;25:213-218.

59. Chervin RD, Hedger KM: Clinical prediction ofperiodic leg movements during sleep in children.Sleep Med 2001;2:501-510.

60. Walters AS, Mandelbaum DE, Lewin DS, et al:Dopaminergic therapy in children with restlesslegs/periodic limb movements in sleep andADHD. Pediatr Neurol 2000;22:182-186.

61. Dahl RE, Pelham WE, Wierson M: The role ofsleep disturbances in attention deficit disordersymptoms: A case study. J Pediatr Psychol1991;16:229-239.

62. Dahl RE, Holttum J, Trubnick L: A clinical pictureof child and adolescent narcolepsy. J Am AcadChild Adolesc Psychiatry 1994;33:834-841.

63. Guilleminault C, Pelayo R: Narcolepsy in prepu-bertal children. Ann Neurol 1998;43:135-142.

64. Wolfson AR, Carskadon MA: Sleep schedules anddaytime functioning in adolescents. Child Dev1998;69:875-887.

65. Chervin RD, Archbold KH: Hyperactivity andpolysomnographic findings in children evaluatedfor sleep-disordered breathing. Sleep 2001;24:313-320.

66. Chervin RD, Giordani B, Ruzicka DL, et al:Polysomnographic findings and behavior in chil-dren scheduled for adenotonsillectomy or herniarepair (Abstract). Sleep 2002;25:A431.

67. Chervin RD, Aldrich MS: Characteristics ofapneas and hypopneas during sleep and relationto excessive daytime sleepiness. Sleep 1998;21:799-806.

68. Chervin RD, Aldrich MS: The Epworth SleepinessScale may not reflect objective measures of sleepi-ness or sleep apnea. Neurology 1999;52:125-131.

69. Clarke DF, Huffman JL, Szymanski E, et al: Schoolperformance, race, and symptoms of sleep-disor-dered breathing (Abstract). Sleep 2002;25:A83-A84.

70. Archbold KH, Pituch KJ, Panahi P, Chervin RD:Symptoms of sleep disturbances among childrenat two general pediatric clinics. J Pediatr 2002;140:97-102.

71. Chervin RD, Hedger KM, Panahi P, Pituch KJ:Sleep problems seldom addressed at two generalpediatric clinics. Pediatrics 2001;107:1375-1380.

72. Weatherly RA, Mai EF, Ruzicka DL, Chervin RD:Adenotonsillectomy in children: Indications,practices, and outcomes reported by otolaryn-gologists (Abstract). Sleep 2000;24(Suppl):A212-A213.

73. Section on Pediatric Pulmonology, Subcommitteeon Obstructive Sleep Apnea Syndrome: Clinicalpractice guideline: Diagnosis and management ofchildhood obstructive sleep apnea syndrome.Pediatrics 2002;109:704-712.

74. American Thoracic Society: Cardiorespiratorysleep studies in children: Establishment of nor-mative data and polysomnographic predictors ofmorbidity. Am J Resp Crit Care Med1999;160:1381-1387.


Ch12.qxd 1/25/05 3:21 AM Page 169

In 1880, Gelineau coined the term narcolepsie todescribe a pathologic condition that was charac-terized by recurrent, brief attacks of sleepiness.1

He recognized that the disorder was accompa-nied by falls, or astasias, that were subsequentlytermed cataplexy. Narcolepsy is a lifelong neuro-logic disorder of rapid eye movement (REM)sleep in which there are attacks of irresistibledaytime sleepiness, cataplexy (sudden loss ofmuscle control in the legs, trunk, or neck inresponse to emotional stimuli such as laughter,fright, or rage), hypnagogic hallucinations (vividand often terrifying dreams at sleep onset), sleepparalysis (momentary inability to move at thetime of sleep onset), and fragmented nightsleep.2

EPIDEMIOLOGY

In a community-based survey in OlmstedCounty, Minnesota, the incidence of narcolepsywas 1.37 per 100,000 persons per year—1.72 formen and 1.05 for women.3 It was highest in the2nd decade of life, followed by a gradual decline.The prevalence was approximately 56 personsper 100,000 persons. In Japan, the prevalencehas been estimated at 1 in 600,4 and in Israel at1 in 500,000.5 The exact prevalence of nar-colepsy in childhood has been difficult to estab-lish. Nevertheless, between 1957 and 1960, in aseries of 400 narcolepsy patients seen at theMayo Clinic by Yoss and Daly,6 15 (4%) wereyounger than 15 years. These data were, how-ever, gathered prior to the introduction ofpolysomnography and the establishment of poly-somnographic criteria for the diagnosis ofnarcolepsy. Although the disorder is most oftendiagnosed in the 3rd and 4th decades, a meta-analysis of 235 subjects derived from three stud-ies by Challamel and coworkers7 found that 34%of all subjects had onset of symptoms prior to

the age of 15 years, 16% prior to age 10 years,and 4.5 % prior to age 5 years (Fig. 13–1). A lagperiod of 5 to 10 years between the onset ofsymptoms and diagnosis has been observed8 inadult subjects. A similar but shorter lag periodperhaps also occurs in childhood narcolepsy.Cataplexy, the most specific clinical feature ofnarcolepsy, is present in only 50% to 70% of allsubjects. Some epidemiologic studies haverequired the presence of cataplexy as a prerequi-site for the diagnosis,9 whereas others10 have notmade this stipulation. This lack of uniformity inclinical diagnostic criteria may explain variabil-ity in estimations of the prevalence of nar-colepsy.11

CLINICAL FEATURES

Preschool-Aged ChildrenNarcolepsy is rare in preschool-aged children.Yoss and Daly observed that 11.7% of a group of85 subjects were below the age of 5 years.6 Intheir meta-analysis of 235 children, Challameland colleagues found that 4.6% were below theage of 5 years at the time of diagnosis.7 Sharpand D’Cruz have described a 12-month-old withhypersomnia who was subsequently confirmedto have narcolepsy.12 Nevsimalova and cowork-ers have described a 21⁄2-year-old who developedhypersomnolence at the age of 6 months, withthe presence of as many as 30 cataplectic attacksper day that mimicked atonic seizures, but thatsubsided after treatment with chlorim-ipramine.13 In general, it is difficult to diagnosenarcolepsy prior to the age of 4 to 5 years, aseven unaffected children of this age tend to takehabitual daytime naps and are not able to pro-vide an accurate history of cataplexy, hypnagogichallucinations, or sleep paralysis. The diagnosismay, however, be facilitated by the documentation

Narcolepsy in ChildhoodSuresh Kotagal

13

171

Ch13.qxd 01/28/05 10:31 AM Page 171

of cataplectic attacks on video-polysomnogra-phy, which shows skeletal muscle atonia andbursts of rapid eye movements coinciding withlow-voltage, mixed-frequency activity on theelectroencephalogram (EEG).

School-Aged ChildrenDaytime sleepiness is the most invariant and dis-abling feature of narcolepsy. It may develop asearly as 5 to 6 years of age. There is a back-ground of a constant, foggy feeling from drowsi-ness, and superimposed on this background areperiods of more dramatic sleep attacks. Habitualafternoon napping is uncommon in most 5- to6-year-olds and should raise suspicion. Lennreported a 6-year-old who would fall asleep 5 to10 times a day.14 Wittig and colleaguesdescribed a 7-year, 5-month-old boy with nar-colepsy who tended to fall asleep while watch-ing television for longer than a half hour, at thedinner table, and while seated in his mother’slap at a doctor’s office.15 The naps in childrenwith narcolepsy tend to be longer (30 to 90 min-utes) than those in adult patients, but they arenot consistently followed by a refreshed feel-

ing.16 These attacks of sleepiness are most likelyto occur when the patient is carrying out seden-tary activities such as sitting in a classroom orreading a book. The daytime sleepiness fre-quently leads to automatic behavior of whichthe subject is unaware, to impaired consolida-tion of memory, to decreased concentration, toexecutive dysfunction, and to emotional prob-lems. Mood swings are also common.17,18

Children with daytime sleepiness may be mis-takenly labeled “lazy” and frequently becomethe target of negative comments from theirpeers. Excessive sleepiness may also be over-looked by the parents until it starts adverselyimpacting mood, behavior, or academic per-formance. Adults who have been diagnosedwith narcolepsy frequently give a history of“attention deficit disorder” in childhood.19

Pollack studied the circadian sleep–wakerhythms in subjects with narcolepsy who wereisolated from their environmental cues.20 Hefound that the major sleep episode was stillabout 6 hours long, and it occurred about onceevery 24 hours, thus indicating that the circa-dian clock was functioning normally. Pollackconfirmed that patients with narcolepsy tendedto sleep more often, but not longer than peoplewithout narcolepsy.

Cataplexy, the second most common butmost specific feature of narcolepsy, consists of asudden loss of muscle tone in the extensormuscles of the thighs, back, or neck in responseto emotional triggers such as fright, rage,excitement, surprise, or laughter. It is causedby the intrusion of the skeletal muscle atonia ofREM sleep into wakefulness.21,22 It is accompa-nied by hyperpolarization of spinal alpha motorneurons, with resultant active inhibition ofskeletal muscle tone and suppression of themonosynaptic H-reflex and tendon reflexes.A history of cataplexy may be difficult to elicitin young children. I recall a 6-year-old girl withproven narcolepsy who denied any episodicmuscle weakness but would repeatedly falldown whenever she jumped on a trampoline.Consciousness remains fully intact during thecataplexy episodes, which can last 1 to 30 min-utes. Respiration and cardiovascular functionsremain unaffected. Challamel and coworkers7

found cataplexy in 80.5% of idiopathic nar-colepsy and in 95% of symptomatic narcolepsysubjects.

172 Chapter 13 Narcolepsy in Childhood

< 5

6–10

11–15

>15

Age

4.6%

68.1%

11.6%

15.7%

Figure 13–1. The age of onset of narcolepsy. (Adaptedfrom Challamel MJ, Mazzola ME, Nevsimalova S, et al:Narcolepsy in children. Sleep 1994;17S:17-20.)

Ch13.qxd 01/28/05 10:31 AM Page 172

Hallucinations at sleep onset (hypnagogic) orupon awakening from sleep (hypnopompic) areseen in 50% to 60% of narcolepsy patients andmay have an unpleasant, frightening quality tothem. They may be auditory or visual in nature.Sleep paralysis is a sudden momentary inability tomove as one is drifting off to sleep or awakeningfrom sleep. Like cataplexy, both sleep paralysis andhypnagogic hallucinations are caused by the intru-sion of fragments of REM sleep into wakefulness.

Night-time sleep is also disturbed, with fre-quent awakenings. Young and colleagues attrib-uted sleep fragmentation in part to periodic limbmovements, which were found in five (63%) ofeight children with narcolepsy in their series.23

Periodic limb movements are rhythmic limbmuscle contractions of 0.5 to 5 seconds’ dura-tion, with an intermovement interval of 5 to 120seconds, occurring in series of three or more,usually during stage 1 or 2 of non-REM (NREM)sleep. They may or may not be associated withEEG evidence of cortical arousal. Similarly, theymay or may not be accompanied by a feeling ofrestlessness in the limbs. Sleep fragmentation innarcolepsy may also occur, unrelated to periodiclimb movements or restless leg syndrome.

Neuropsychological and behavioral manifesta-tions are common in childhood narcolepsy, butthey have not been adequately studied, partlybecause of difficulties in developing valid andpractical batteries of neuropsychological tests forsleepy children. It is not uncommon for childrenwith narcolepsy to present with inattentivenessor mild depression. Adult patients with nar-colepsy have demonstrated selective cognitivedeficits in response latency, word recall, and esti-mation of frequency.24 Rogers and Rosenberg25

performed a battery of neuropsychological stud-ies on 30 adults with narcolepsy and age-matched controls. The subjects with narcolepsyexperienced more difficulty in maintainingattention than controls, as evidenced by moreperseveration errors on Strub and Black’s List ofLetters. It is unclear whether children with nar-colepsy exhibit similar deficits.

PATHOPHYSIOLOGY

Narcolepsy in AnimalsOver the past 2 decades, the study of narcolepsyin animals has advanced understanding of the

various phenomena that are also observed inhumans. Narcolepsy has been studied in cats,miniature horses, quarter horses, Brahman bulls,and about 15 breeds of dogs, and it shows amonogenic, autosomal recessive pattern ofinheritance.26-28 Cataplexy can be induced incats by the injection of carbachol (an acetyl-choline-like substance) into the pontine reticu-lar formation.29 Specifically, muscarinic type-2receptors of acetylcholine have been impli-cated.30,31 The food-elicited cataplexy test, usedto study cataplexy in dogs, uses the finding thatthe time taken for the consumption of food bynarcoleptic animals whose eating is interruptedby cataplexy attacks is much longer than in ani-mals without narcolepsy.

In 1999, Lin and coworkers demonstratedthat canine narcolepsy is caused by a mutation inthe hypocretin receptor-2 (orexin-2) gene.32

Around the same time, Chemelli and colleaguesestablished that a null mutation for the hypocre-tin-1 and hypocretin-2 peptides in mice pro-duces aspects reminiscent of human narcolepsy,including cataplexy.33 The hypocretin-containingneurons are located primarily in the dorsolateralhypothalamus, but they have widespread projec-tions to remote areas such as the basal forebrain,amygdala, medial nuclei of the thalamus, peri-aqueductal gray matter, reticular formation,pedunculo-pontine nucleus, locus coeruleus,raphe nucleus, pontine tegmentum, and dorsalspinal cord.32,34 Hypocretins 1 and 2 are peptidesthat are synthesized from preprohypocretin, andhave corresponding receptors. Although thehypocretin type 1 receptor binds only tohypocretin-1, the hypocretin type 2 receptor canbind to both type 1 and type 2 ligands.Hypocretins stimulate food intake, increase thebasal metabolic rate, and promote arousal.34

Decreased activation of the hypocretin system isthe underlying theme in canine and murinenarcolepsy. This receptor downregulation mayoccur as a result of either exon-skipping muta-tions in the hypocretin receptors (Labrador andDoberman models),33 or after point mutationsin the Hcrt2 receptor gene, with an amino acidchange from glutamic acid to lysine in the N-terminus portion of the receptor (Dachshundmodel).35 Also of significance is the finding thatthe intravenous administration of hypocretin-1(orexin A) in narcoleptic Doberman pinschersreduces cataplexy for up to 3 days, increases

Narcolepsy in Childhood Chapter 13 173

Ch13.qxd 01/28/05 10:31 AM Page 173

activity, promotes waking, and reduces sleepfragmentation in a dose-dependent manner.36

Histocompatibility Antigensand Human NarcolepsyIn 1984, an association between narcolepsy andhistocompatibility leukocyte antigen (HLA) DR2was reported in Japan by Juji and coworkers.37

This association was subsequently observed inother geographic regions of the world aswell.38,39 Consequently, an immunologic mecha-nism was suspected in the pathogenesis ofhuman narcolepsy, but this has not been estab-lished. It was then demonstrated that the associ-ation with DR2 is only secondary, and that thereis a stronger association of narcolepsy with theHLA DQ antigens, specifically DQB1*0602 andDQA1*0102, which are present in 95% to 100%of patients, as compared to a 12% to 38% preva-lence in the general population.40 In a study of525 healthy subjects, Mignot and colleaguesdemonstrated that DQB1*0602 positivity waslinked to shorter REM latency, increased sleepefficiency, and decreased time spent in stage 1NREM sleep.41 Pelin and coworkers havedemonstrated that homozygosity for these twohaplotypes is associated with a twofold to four-fold increase in the likelihood of developing nar-colepsy over heterozygotes, but that thepresence of these antigens does not influence theseverity of the disease.42

Hypocretins and HumanNarcolepsyUnlike narcolepsy in dogs and mice, human nar-colepsy is generally not associated with abnor-malities in hypocretin receptors but rather withlow to absent levels of cerebrospinal (CSF)hypocretin-1.43 In a postmortem study of humannarcolepsy, Thannickal and colleagues found an85% to 95% reduction in the number of hypocre-tin neurons in the hypothalamic region,44

whereas melanin-concentrating hormone neu-rons, which are intermingled with the hypocretinneurons, remained unaffected, thus suggesting atargeted neurodegenerative process. Using aradioimmunoassay, Nishino and coworkers45

found that the mean CSF level of hypocretin-1 inhealthy controls was 280.3 ± 33.0 pg/mL, and inneurologic controls it was 260.5 ± 37.1 pg/mL,

whereas in those with narcolepsy, hypocretin-1was either undetectable or below 100 pg/mL.The diagnostic sensitivity of low levels (less than100 pg/mL) was 84.2%. Low to absent levelswere found in 32 out of 38 patients, who were allHLA DQB1*0602 positive. HLA-negative nar-colepsy patients had normal to high CSFhypocretin-1 levels. In another recent study,92.3% of patients who were both DQB1*0602and cataplexy positive had undetectable CSFhypocretin-1 levels, whereas DQB1*0602-nega-tive patients with cataplexy and DQB1*0602-negative patients without cataplexy had normallevels.46 In a study of narcolepsy with cataplexy,narcolepsy without cataplexy, and idiopathichypersomnia, Kanbayashi and colleagues47 foundthat all nine CSF hypocretin-deficient patientswere HLA DR2 positive (including three preado-lescents). In contrast, narcolepsy without cata-plexy and idiopathic hypersomnia patients hadnormal CSF hypocretin levels (Fig. 13–2).


300

250

200

150

100

50

0

pg/m

L

Narcolepsy with cataplexy, HLA+ve

Narcolepsy with cataplexy, HLA–ve

Narcolepsy without cataplexy

Idiopathic hypersomnia

n = 9 n = 2 n = 5 n = 12

Figure 13–2. Mean cerebrospinal fluid hypocretin-1levels in various categories of hypersomnia. The nar-colepsy-cataplexy group included three prepubertal chil-dren of ages 6, 7, and 10. (Adapted from Kanbayashi T,Inoue Y, Chiba S, et al: CSF hypocretin-1 (orexin-A) con-centrations in narcolepsy with and without cataplexy andidiopathic hypersomnia. J Sleep Res 2002;11:91-93.)

Ch13.qxd 01/28/05 10:31 AM Page 174

The routine application of CSF hypocretin-1assays in the diagnosis of narcolepsy is aroundthe corner. The assay is most useful when anHLA DQB1*0602-positive patient with suspectednarcolepsy-cataplexy is receiving CNS stimulantson initial presentation to the sleep specialist, andwhen discontinuation of medications for the pur-pose of obtaining a multiple sleep latency test(MSLT) is inconvenient or impractical.

The Two-Hit HypothesisThe presence of histocompatibility antigenDQB1*0602 or DQA1*0102 is, per se, insuffi-cient to precipitate narcolepsy. This is substanti-ated by the fact that DQB1*0602-positivemonozygotic twins have been incompletely con-cordant for narcolepsy, with one of the pairdeveloping narcolepsy-cataplexy at age 12 yearsand the other not until after having sufferedemotional stress and sleep deprivation at the ageof 45 years.48 Human narcolepsy is thus bestexplained on the basis of a two-threshold hypoth-esis, with an interplay between genetic suscepti-bility and environmental factors—major lifeevents, such as a systemic illness, an injury, orbereavement, have been reported to be presentin 82% of narcolepsy patients, compared with a44% incidence in controls49 (P < .001). The com-bination of genetic susceptibility and anacquired stress seems to trigger most cases ofnarcolepsy.

Monoamine DisturbancesHypocretin deficiency in humans leads todown-regulation of arousal-mediating nora-drenergic and dopaminergic pathways inthe brainstem, and to upregulation of REMsleep–facilitating cholinergic pathways.50

Montplaisir and coworkers51 measured serumand CSF levels of several biogenic amines andtheir metabolites—dopamine (the metabolitehomovanillic acid), norepinephrine (themetabolite 3-methoxy-4-hydroxyphenylethyl-eneglycol), epinephrine, and serotonin (themetabolite 5-hydroxy indoleacetic acid)—inpatients with narcolepsy, in those with idio-pathic hypersomnia, and in normal controls.Both narcolepsy and idiopathic hypersomniapatients had significantly decreased concentra-tions of dopamine and indoleacetic acid, a

metabolite of tryptamine. Dopamine and trypt-amine are usually present in high concentra-tions in the basal ganglia. A relative deficiencyof these compounds, probably mediated bydown-regulation of hypocretin, is involved inthe evolution of sleepiness. Stimulants such asdextroamphetamine and methylphenidate,which are used in the treatment of hypersom-nolence, are known to enhance dopaminerelease from presynaptic terminals. Activationof selective dopamine D2 receptor agonists alsosuppresses cataplexy.52 This inhibition isbelieved to be indirectly mediated via activationof noradrenergic pathways. A reduction in cen-tral dopamine activity might also underlie theperiodic leg movements that are common in thenight sleep of patients with narcolepsy—theyare usually relieved by treatment with levodopaor dopamine agonists. A cholinergic distur-bance also coexists—physostigmine, a choliner-gic agent, leads to a transient increase ofcataplexy.

Secondary NarcolepsyAlthough the majority of narcolepsy is idio-pathic, structural lesions of the diencephalonor brainstem may on rare occasion precipitatesecondary narcolepsy in those who are biologi-cally predisposed, perhaps subsequent to dis-ruption of the hypocretin pathways. Cerebellarhemangioblastomas, temporal lobe B-cell lym-phomas, pituitary adenoma, third ventriculargliomas, craniopharyngiomas, head trauma,viral encephalitis, ischemic brainstem distur-bances, sarcoidosis, and multiple sclerosis havebeen associated with narcolepsy.53-60 In suchpatients, the finding of cataplexy increases thelikelihood that the hypersomnolence is nar-colepsy related. Isolated cataplexy has alsobeen observed in Niemann-Pick type C dis-ease,61 but these patients cannot be labeledas having secondary narcolepsy as they lackother prerequisite clinical and polysomno-graphic features. Arii and colleagues describeda hypothalamic tumor in a 16-year-old girl whomanifested hypersomnolence, obesity, and lowCSF hypocretin-1 levels, without cataplexy.62

Although this patient did not meet the criteriafor the diagnosis of narcolepsy, the report high-lights the importance of hypocretins in the reg-ulation of alertness.


Ch13.qxd 01/28/05 10:31 AM Page 175

DIAGNOSIS

The diagnosis of narcolepsy is established on thebasis of the history, combined with characteristicfindings on the nocturnal polysomnogram and amultiple sleep latency test.63 In preparation forthe sleep studies, the patient should be with-drawn from all central nervous stimulants,hypnotics, antidepressants, and any other psy-chotropic agents for 2 weeks prior to the sleepstudies, as these drugs may impact sleep archi-tecture. In the case of drugs with a long half-life,such as fluoxetine, the drug-free interval mayneed to be 3 to 4 weeks. During this time, thepatient should maintain a regular sleep–wakeschedule, which can be verified by wrist actigra-phy and sleep logs that are maintained by thepatient for 1 to 2 weeks prior to the sleep stud-ies.64 A general physical examination shouldalso be carried out to assess the Tanner stage ofsexual development, as normal values for noc-turnal total sleep time, daytime sleep latency,and daytime REM sleep latency are closelylinked to the Tanner stages.65

During the nocturnal polysomnogram, multi-ple parameters of physiologic activity, such asthe EEG (C3-A1, O2-A1 montage), eye move-ments, chin and leg electromyogram, nasal pres-sure, thoracic and abdominal respiratory effort,and oxygen saturation, are recorded simultane-ously on a computerized sleep monitoring andanalysis system.66 The test helps exclude sleeppathologies such as obstructive sleep apnea, theperiodic limb movement disorder,67,68 and idio-pathic hypersomnia69,70 that can also lead todaytime sleepiness and mimic narcolepsy.

Patients with narcolepsy may exhibit onset ofREM sleep within 15 minutes of sleep onset(sleep-onset REM period). Gross sleep efficiencyis generally high (greater than 90%), but theremay be fragmentation of sleep from an increasednumber of arousals and periodic limb move-ments. There is no significant disordered breath-ing or oxygen desaturation. A useful clue tonarcolepsy in adolescents is the presence ofdecreased nocturnal REM sleep latency (the timebetween sleep onset and the onset of the first 30-second epoch of REM sleep). For example, inone study,71 the nocturnal REM sleep latency innarcoleptic subjects was less than 67 minutes(mean, 24.5 minutes; standard deviation, 30;range, 0 to 66.5 minutes; n = 8), as compared to

a mean nocturnal REM sleep latency of 143.7minutes in age-matched controls (range, 82.5 to230.5; SD, 50.9 minutes; P < .001). A similarconclusion was reached by Challamel andcoworkers in their meta-analysis of 235 sub-jects,7 and in the pioneering work of Carskadonand colleagues.72

An MSLT should be started at 2 hours afterthe final morning awakening. It consists of fourto five nap opportunities of 20-minute lengths at2-hour intervals in a darkened, quiet room (e.g.,at 1000, 1200, 1400, and 1600 hours).63 Eyemovements, chin electromyogram and elec-troencephalogram (generally C3-A2, O2-A1) arerecorded. The MSLT provides quantitative infor-mation about the degree of sleepiness and quali-tative information about the nature of thetransition from wakefulness into sleep (i.e.,wakefulness → NREM sleep, or wakefulness →REM sleep). The time interval between “lightsout” and sleep onset is termed the sleep latency;a mean sleep latency is then derived from all thenaps. A urine drug screen is obtained betweenthe naps if the patient appears to be fallingasleep very quickly. A concerted effort should bemade to ensure that the patient does not acci-dentally fall asleep between the naps. Typically,the mean sleep latency is markedly shortened toless than 5 minutes in patients with narcolepsy,whereas in controls it is generally 15 to 18 min-utes66 (Table 13–1).

In unaffected children, the initial transition isfrom wakefulness into NREM sleep. Narcolepsyis, however, associated with a transition fromwakefulness directly into REM sleep (sleep-onsetREM period [SOREMP]). This diagnostic featureof two or more SOREMPs may not be consis-tently present in the early stages of the disorderin children and young adults, and sometimesserial sleep studies are needed to establish adefinitive diagnosis73 (Fig. 13–3). Opinions dif-fer about normative values for the MSLT-derivedmean sleep latency in preadolescents. Althoughthe normative data derived by Carskadon65 arewidely used, it is conceivable that mean sleeplatencies may be greater than 20 minutes inpreadolescents.74,75

Serologic testing for the HLA DQB1*0602haplotype is an adjunctive test, but it cannot beused alone to diagnose narcolepsy, as this haplo-type is also present in 12% to 38% of the generalpopulation.41 CSF hypocretin analysis is useful


Ch13.qxd 01/28/05 10:31 AM Page 176

in establishing the diagnosis when psychotropicmedications cannot be discontinued for safetyreasons. Unfortunately, the testing is not readilyavailable, so nocturnal polysomnography andthe MSLT remain the gold standard at this timefor making a definitive diagnosis of narcolepsy.

DIFFERENTIAL DIAGNOSIS

By far, the most common disorder leading toexcessive daytime sleepiness in an adolescent is

insufficient nocturnal sleep.76 To begin with, thereis a physiologic shift in the sleep-onset time ofmost teenagers to between 10:30 and 11:00 PM.77

Also, the physiologic delay in the onset of dim-light melatonin secretion to a later time at nightand the resultant postponement of sleep onset isassociated with a corresponding shift to a latermorning awakening time in most adolescents.Linked to this phase shift, Carskadon andcoworkers have documented SOREMPs duringthe MSLT of 12 of 25 healthy students.78

Superimposed on this may be elements of abnor-mal sleep hygiene (e.g., use of stimulants, late-night television viewing/phone calls/computerchats), or circadian rhythm disturbances such asthe delayed sleep phase syndrome.79,80 The use ofprescription drugs or over-the-counter sedativessuch as antihistamines should also be consideredas a possible cause of sleepiness. A high indexof suspicion should be maintained for illicitdrug use.

It is not uncommon for narcolepsy to presentwith obesity and nocturnal snoring, thus mim-icking obstructive sleep apnea. The most commoncauses of childhood obstructive sleep apnea areadenotonsillar hypertrophy, neuromuscular dis-orders, and craniofacial anomalies such asCrouzon syndrome and Down syndrome.81-83

The upper airway resistance syndrome is anotherunder-recognized disorder leading to chronicdaytime sleepiness. It is characterized by habit-ual snoring, an increased number of breathing-related microarousals, and increased negativeintrathoracic pressures on monitoring ofesophageal pressures during polysomnographyusing balloons.84

The Kleine-Levin syndrome, or periodic hyper-somnia, is characterized by recurrent periods of


Stage of Development Mean Sleep Latency (min) Standard DeviationTanner stage I 18.8 1.8Tanner stage II 18.3 2.1Tanner stage III 16.5 2.8Tanner stage IV 15.5 3.3Tanner stage V 16.2 1.5Older adolescents 15.8 3.5

Adapted from Carskadon MA: The second decade. In Guilleminault C (ed): Sleeping and Waking Disorders:Indications and Techniques. Menlo Park, Calif, Addison-Wesley, 1982, pp 99-125.

Table 13–1. Normal Values for the Multiple Sleep Latency Test

20

15

10

5

01000 1200

Time of day

ControlTrial 1Trial 2

1400 1600

Min

utes

Figure 13–3. Serial multiple sleep latency tests (MSLTs)in a 13-year-old child with evolving narcolepsy. MSLT tri-als 1 and 2 were carried out 9 months apart. They arecompared with the MSLT findings of a healthy control.There is progressive decrease over time in the sleep laten-cies of the patient at 1000, 1200, and 1400 hours, alongwith an increase in the number of sleep-onset rapid eyemovement periods. Sleep latencies in the final nap (1600hours) may be physiologically prolonged in children as aresult of anticipation about going home—the “last nap”effect. (Reproduced from Kotagal S, Swink TD: Excessivedaytime sleepiness in a 13-year old. Semin Pediatr Neurol1996;3:170-172.)

Ch13.qxd 01/28/05 10:31 AM Page 177

sleepiness.85 The disorder is common in adoles-cent males, who manifest 1- to 2-week periods ofexcessive sleep, in association with hyperphagiaand hypersexual behavior. There may be a 2- to10-pound weight gain during the sleepy periods,which then remit spontaneously, only to reap-pear a few weeks later. No specific etiology hasbeen found. The disorder subsides graduallyover time. The nocturnal polysomnogram showsa high sleep efficiency and a decreased percent-age of time spent in stages 3 and 4 of NREMsleep.86 The MSLT shows moderate daytimesleepiness with a shortened mean sleep latencyof 5 to 10 minutes, but fewer than twoSOREMPs.

Idiopathic hypersomnia should also be consid-ered in the differential diagnosis of narcolepsy. Itis defined as a disorder associated with non-imperative sleepiness, long unrefreshing naps,difficulty reaching full awakening after sleep,and sleep drunkenness, as well as absence ofSOREMPs during the MSLT.70,87 Night sleep isquantitatively and qualitatively normal. TheMSLT shows a short mean sleep latency, gener-ally in the range of 5 to 10 minutes, but two ormore SOREMPs, which would be suggestive ofnarcolepsy, are not seen. I have studied six chil-

dren with narcolepsy who underwent serial noc-turnal polysomnograms and MSLTs for theassessment of daytime sleepiness.71 Long, unre-freshing naps were common. Although the ini-tial battery of sleep studies showed shortenedmean sleep latencies (less than 10 minutes), sug-gesting moderate daytime sleepiness, theylacked the necessary (two or more) SOREMPs todiagnose narcolepsy. Initially, therefore, they metthe diagnostic criteria for idiopathic hypersom-nia. A repeat battery of sleep studies severalmonths later, however, showed the appearanceof two or more SOREMPs on the MSLT of eachsubject, characteristic of narcolepsy. In somechildren, therefore, idiopathic hypersomnia maybe a transitional phase that precedes the devel-opment of classic narcolepsy.

MANAGEMENT

Because narcolepsy requires lifelong treatment,the nocturnal polysomnogram and MSLT find-ings should be unequivocally positive beforemaking the diagnosis. Drugs commonly used fortreatment are listed in Table 13–2. Daytimesleepiness is countered with stimulants such asmethylphenidate (regular or extended-release


Symptom Drug (Trade Name) DosageDaytime Methylphenidate hydrochloride 5 mg bid to a maximum of 60 mg

sleepiness Ritalin 20 mg bid to a maximum of 60 mgRitalin SR 18 mg qid to a maximum of 54 mg/dayConcerta 10 mg qid to a maximum of 60 mg/dayMetadate 100-500 mg/day in 1-2 doses

Modafinil (Provigil) 10-40 mg/dayDextroamphetamine (Dexedrine) 20-25 mg/dayMethamphetamine (Desoxyn) 10-40 mg/dayAmphetamine/dextroamphetamine

mixture (Adderall)

Cataplexy Clomipramine (Anafranil) 25-75 mg/day in 1-2 divided dosesImipramine 25-100 mg/dayProtriptyline (Vivactil) 2.5-10 mg/day in 1-2 divided dosesSodium oxybate (Xyrem) 3-9 g in two divided doses at night

Emotional Fluoxetine (Prozac) 10-30 mg/day every morningproblems Sertraline (Zoloft) 25-100 mg/day every morning

Periodic leg Clonazepam (Klonopin) 0.5-1.0 mg at bedtimemovements Levodopa-carbidopa (Sinemet) 25/100 or 50/200 at bedtime

Pramipexole (Mirapex) 0.125-0.25 mg at bedtime

Table 13–2. Drugs Used Commonly in the Treatment of Narcolepsy

Ch13.qxd 01/28/05 10:31 AM Page 178

formulations) or various preparations ofamphetamine.88-91 The side effects of theseagents include loss of appetite, nervousness, tics,headache, and insomnia. Modafinil (Provigil), adrug with an unspecified mode of action, hasalso been reported to be effective in enhancingalertness and improving psychomotor perform-ance in doses of 100 to 400 mg/day in one or twodivided doses. The side effects include headache,nervousness, anxiety, and nausea.

Because cataplexy is mediated by cholinergicpathways in the brainstem, drugs with anti-cholinergic properties such as protriptyline andclomipramine have been used for its control.Sodium oxybate (gamma hydroxybutyrate), anendogenous neurotransmitter/neuromodulator,has recently been found to alleviate cataplexy in arandomized, double-blind, placebo-controlled,multicenter study.92 It leads to consolidation ofsleep and a marked increase in stage 3-4 NREMsleep at night. It is unclear whether the signifi-cant improvement in cataplexy and the modestimprovement in daytime alertness with sodiumoxybate are linked to consolidation in the dis-rupted night-sleep that is so characteristic of nar-colepsy, or to other unknown mechanisms.Because of its positive effects on mood and libido,it has the potential of being abused as a recre-ational drug.50 One or two planned naps per dayof 25 to 30 minutes also help enhance daytimealertness and improve psychomotor perform-ance.93 Supportive psychotherapy and fluoxetineor sertraline might also be indicated if patientsdevelop emotional and behavioral problems.

The Narcolepsy Network (www.narcolepsynet-work.org) is a private, nonprofit resource forpatients, families, and health professionals.Because of the increased risk of accidents fromsleepiness, patients should be cautioned againstdriving long distances and working near sharp,moving machinery. Future treatments mightconsist of hypocretin analogues.50

References

1. Gelineau J: De la narcolepsie. Gaz Hosp (Paris)1880;53:626-628.

2. American Sleep Disorders Association: TheInternational Classification of Sleep Disorders:Diagnostic and Coding Manual. Rochester, Minn,American Sleep Disorders Association, 1997.

3. Silber M, Krahn L, Olson E, Pankranz V: The epi-demiology of narcolepsy in Olmsted County,

Minnesota: A population-based study. Sleep2002;25:197-202.

4. Honda Y: Clinical features of narcolepsy: Japaneseexperiences. In Honda Y, Juji T (eds): HLA inNarcolepsy. Berlin, Springer-Verlag, 1988, pp 24-57.

5. Lavie P, Peled R: Narcolepsy is a rare disease inIsrael. Sleep 1987;10:608-609.

6. Yoss RE, Daly DD: Narcolepsy in children.Pediatrics 1960;25:1025-1033.

7. Challamel MJ, Mazzola ME, Nevsimalova S, et al:Narcolepsy in children. Sleep 1994;17S:17-20.

8. Zarcone V: Narcolepsy. N Engl J Med 1973;288:1156-1166.

9. Matsuki K, Honda Y, Juji T: Diagnostic criteria fornarcolepsy and HLA-DR2 frequencies. TissueAntigens 1987;30:155-160.

10. Roffwarg HP, Sleep Disorders ClassificationCommittee, Association of Sleep Disorders:Diagnostic classification of sleep and arousal dis-orders. Sleep 1979;2:1-137.

11. Hublin C, Partinen M, Kaprio J, Kosekenvuo M,et al: Epidemiology of narcolepsy. Sleep1994;17(Suppl 1):7-12.

12. Sharp SJ, D’Cruz OF: Narcolepsy in a 12-month-old boy. J Child Neurol 2001;16:145-146.

13. Nevsimalova S, Roth B, Zouhar A, Zemanova H:Narkolepsie-kataplexie a periodicka hypersomniese zacatkem v kojeneckem veku. Cs Pediatr1986;41:324-327.

14. Lenn NJ: HLA-DR2 in childhood narcolepsy.Pediatr Neurol 1986;2:314-315.

15. Wittig R, Zorick F, Roehrs T, et al: Narcolepsy ina 7-year-old child. J Pediatr 1983;102:725-727.

16. Kotagal S, Hartse KM, Walsh JK: Characteristicsof narcolepsy in pre-teen aged children. Pediatrics1990;85:205-209.

17. Pearl PL, Efron L, Stein MA: Children, sleep, andbehavior: A complex association. Minerva Pediatr2002;54:79-91.

18. Wise MS: Childhood narcolepsy. Neurology1998;50(Suppl 1):S37-S42.

19. Navalet Y, Anders TF, Guilleminault C:Narcolepsy in children. In Guilleminault C,Dement WC, Passouant P (eds): Narcolepsy. NewYork, Spectrum, 1976, pp 171-177.

20. Pollack CP: The rhythms of narcolepsy.Narcolepsy Network 1995;8:1-7.

21. Guilleminault C, Wilson RA, Dement WC: Astudy of cataplexy. Arch Neurol 1974;31:255-261.

22. Guilleminault C, Pelayo R: Narcolepsy in prepu-bertal children. Ann Neurol 1998;43:135-142.

23. Young D, Zorick F, Wittig R, et al: Narcolepsy in apediatric population. Am J Dis Child1988;142:210-214.

24. Henry GK, Satz P, Heilbronner RL: Evidence of aperceptual encoding deficit in narcolepsy. Sleep1993;16:123-127.


Ch13.qxd 01/28/05 10:31 AM Page 179

25. Rogers AE, Rosenberg RS: Tests of memory in nar-coleptics. Sleep 1990;13:42-52.

26. Foutz AS, Mitler MM, Cavalli-Sforza LL, et al:Genetic factors in canine narcolepsy. Sleep1979;1:413-421.

27. Knecht CD, Oliver JE, Redding R, et al:Narcolepsy in a dog and a cat. J Am Vet MedAssoc 1984;162:1052-1053.

28. Strain GM, Olcott BM, Archer RM, et al:Narcolepsy in a Brahmin bull. J Am Vet MedAssoc 1984;185:538-541.

29. Mitler MM, Dement WC: Cataplectic-like behav-ior in cats after microinjection of carbachol inthe pontine reticular formation. Brain Res1974;68:335-343.

30. Nishino S, Reid MS, Dement WC, Mignot E:Neuropharmacology and neurochemistry of nar-colepsy. Sleep 1994;17(Suppl 1):S84-92.

31. Reid MS, Tafti M, Geary JN, et al: Cholinergicmechanisms in cataplexy: I. Modulation of cata-plexy via local drug administration into the para-median pontine reticular formation. Neuroscience1994;59:511-522.

32. Lin L, Faraco J, Li R, et al: The sleep disorder,canine narcolepsy, is caused by a mutation in thehypocretin (orexin) receptor 2 gene. Cell1999;98:365-376.

33. Chemelli RM, Willie JT, Sinton CM, et al:Narcolepsy in orexin knock-out mice: Moleculargenetics of sleep regulation. Cell 1999;98:437-451.

34. Silber MH, Rye DB: Solving the mysteries of nar-colepsy: The hypocretin story. Neurology2001;56:1616-1618.

35. Hungs Fan J, Lin X, Maki RA, Mignot E:Identification and functional analysis of muta-tions in the hypocretin (orexin) genes of nar-coleptic canines. Genome Res 2001;11:531-539.

36. John J, Wu M-F, Siegel JM: Systemic administra-tion of hypocretin-1 reduces cataplexy and nor-malizes sleep and waking durations in narcolepticdogs. Sleep Res Online 2000;3:23-28.

37. Juji T, Satake M, Honda Y, Doi Y: HLA antigens inJapanese patients with narcolepsy. TissueAntigens 1984;24:316-319.

38. Billiard M, Seignalet J, Besset A, Cadilhac J: HLADR2 and narcolepsy. Sleep 1986;9:149-152.

39. Langdon N, Lock C, Welsh K, et al: Immune fac-tors in narcolepsy. Sleep 1986;9:143-148.

40. Mignot E, Lin L, Rogers W, et al: Complex HLA-DR and -DQ interactions confer risk of nar-colepsy-cataplexy in three ethnic groups. Am JHum Genet 2001;68:686-699.

41. Mignot E, Young T, Lin L, Fin L: Nocturnal sleepand daytime sleepiness in normal subjects withHLA-DQB1*0602. Sleep 1999;22:347-352.

42. Pelin Z, Guilleminault C, Risch N, et al: HLADQB1*0602 homozygosity increases relative riskfor narcolepsy but not disease severity in two eth-nic groups. Tissue Antigens 1998;51:96-100.

43. Nishino S, Ripley B, Overeem S, et al: Hypocretin(orexin) deficiency in human narcolepsy. Lancet2000;355:39-40.

44. Thannickal TC, Moore RY, Nienhuis R, et al:Reduced number of hypocretin neurons inhuman narcolepsy. Neuron 2000;27:469-474.

45. Nishino S, Ripley B, Overeem S, et al: Low cere-brospinal fluid hypocretin (orexin) and alteredenergy homeostasis in human narcolepsy. AnnNeurol 2001;50:381-388.

46. Krahn LE, Pankrantz VS, Oliver L, et al:Hypocretin (orexin) levels in cerebrospinal fluidof patients with narcolepsy: Relationship to cata-plexy and DQB1*0602. Sleep 2002;25:733-738.

47. Kanbayashi T, Inoue Y, Chiba S, et al: CSFhypocretin-1 (orexin-A) concentrations in nar-colepsy with and without cataplexy and idio-pathic hypersomnia. J Sleep Res 2002;11:91-93.

48. Honda M, Honda Y, Uchida S, et al: Monozygotictwins incompletely concordant for narcolepsy.Biol Psychiatr 2001;49:943-947.

49. Orellana C, Villemin E, Tafti M, et al: Life eventsin the year preceding the onset of narcolepsy.Sleep 1994;17(Suppl 1):50-53.

50. Mignot E, Taheri S, Nishino S: Sleeping with thehypothalamus: Emerging therapeutic targets forsleep disorders. Nat Neurosci 2002;5(Suppl):1071-1075.

51. Montplaisir J, de Champlain J, Young SN, et al:Narcolepsy and idiopathic hypersomnia: Biogenicand related compounds in CSF. Neurology1982;32:1299-1302.

52. Nishino S, Arrigoni J, Valtier D, et al: DopamineD2 mechanisms in canine narcolepsy. J Neurosci1991;11:2666-2671.

53. Autret A, Lucas F, Henry-Lebras F, et al:Symptomatic narcolepsies. Sleep 1994;17S(Suppl1):21-24.

54. Tridon P, Montaut J, Picard L, et al: Syndrome deGelineau et hemangioblastome kystique ucervelet. Rev Neurol 1969;121:186-189.

55. Onofrj M, Curatola L, Ferracci F, Fulgente T:Narcolepsy associated with primary temporallobe B-cell lymphoma in a HLA DR2 negativesubject. J Neurol Neurosurg Psychiatry 1992;55:852-853.

56. Schwartz WJ, Stakes JW, Hobson JA: Transientcataplexy after removal of craniopharyngoma.Neurology 1984;34:1372-1375.

57. Lankford DA, Wellman JJ, Ohara C: Post-trau-matic narcolepsy in mild to moderate closed headinjury. Sleep 1994;17S(Suppl 1):25-28.


Ch13.qxd 01/28/05 10:31 AM Page 180

58. Bonduelle C, Degos C: Symptomatic narcolepsies:A critical study. In Guilleminault C, Dement WC,Passouant P (eds): Narcolepsy. New York,Spectrum, 1976, pp 312-332.

59. Rivera VM, Meyer JS, Hata T, et al: Narcolepsy fol-lowing cerebral ischemia. Ann Neurol1986;19:505-508.

60. Schrader H, Gotlibsen OB: Multiple sclerosis andnarcolepsy-cataplexy in a monozygotic twin.Neurology 1980;30:105-108.

61. Kandt RS, Emerson RG, Singer HS, et al:Cataplexy in variant forms of Niemann-Pick dis-ease. Ann Neurol 1982;12:284-288.

62. Arii J, Kanbayashi T, Tanabe Y, et al: A hypersom-nolent girl with decreased CSF hypocretin levelafter removal of a hypothalamic tumor.Neurology 2001;56:1775-1776.

63. Carskadon MA, Dement WC, Mitler MM, et al:Guidelines for the multiple sleep latency test(MSLT): A standard measure of sleepiness. Sleep1986;9:519-524.

64. Sadeh A, Hauri PJ, Kripke DF, Lavie P: The role ofactigraphy in the evaluation of sleep disorders.Sleep 1995;18:288-302.

65. Carskadon MA: The second decade. InGuilleminault C (ed): Sleeping and WakingDisorders: Indications and Techniques. MenloPark, Calif, Addison-Wesley, 1982, pp 99-125.

66. Kotagal S, Goulding PM: The laboratory assess-ment of daytime sleepiness in childhood. J ClinNeurophysiol 1996;13:208-218.

67. American Sleep Disorders Association: PeriodicLimb Movement Disorder. In The InternationalClassification of Sleep Disorders: Diagnostic andCoding Manual. Rochester, Minn, American SleepDisorders Association, 1997, pp 65-68.

68. Chervin RD, Archbold KH, Dillon JE, et al:Association between symptoms of inattention,hyperactivity, restless legs syndrome, and periodicleg movements. Sleep 2002;25:213-218.

69. Roth B: Narcolepsy and hypersomnia: Review andclassification of 642 personally observed cases.Scweiz Arch Neurol Neurochir Psychiatr1976;119:31-41.

70. Basetti C, Aldrich MS: Idiopathic hypersomnia: Aseries of 42 patients. Brain 1997;120:1423-1435.

71. Kotagal S: A developmental perspective on nar-colepsy. In Loughlin GM, Carroll JL, Marcus CL(eds): Sleep and Breathing in Children. New York,Marcel Dekker, 2000, pp 347-362.

72. Carskadon MA, Harvey K, Dement WC: Multiplesleep latency tests during the development of nar-colepsy. West J Med 1981;135:414-418.

73. Kotagal S, Swink TD: Excessive daytime sleepi-ness in a 13-year-old. Semin Pediatr Neurol1996;3:170-172.

74. Gozal D, Wang M, Pope DW: Objective sleepinessmeasures in pediatric obstructive sleep apnea.Pediatrics 2001;108:693-697.

75. Palm L, Persson E, Elmquist D, Blennow G: Sleepand wakefulness in normal pre-adolescents. Sleep1989;12:299-308.

76. Brown LW, Billiard M: Narcolepsy, Kleine-Levinsyndrome, and other causes of sleepiness inchildren. In Ferber R, Kryger M (eds): Principlesand Practice of Sleep Medicine in the Child.Philadelphia, WB Saunders, 1995, pp 125-134.

77. Carskadon MA: Factors influencing sleep pat-terns of adolescents. In Carskadon MA (ed):Adolescent Sleep Patterns: Biological, Social, andPsychological Perspectives. Cambridge, Mass,Cambridge University Press, 2002, pp 18-19.

78. Carskadon MA, Wolfson AR, Acebo C, et al:Adolescent sleep patterns, circadian timing, andsleepiness at a transition to early school days.Sleep 1998;21:871-881.

79. Garcia J, Rosen G, Mahowald M: Circadian rhythmsand circadian rhythm disorders in children andadolescents. Semin Pediatr Neurol 2001;8:229-240.

80. Dijk DJ, Bolos Z, Eastman CI, et al: Light treat-ment of sleep disorders: Consensus report: II.Basic properties of circadian physiology. J BiolRhythms 1995;10:113-125.

81. Marcus CL: Sleep architecture and respiratory dis-turbance in children with obstructive sleep apnea.Am J Respir Crit Care Med 2000;162(Part 1):682-686.

82. Marcus CL: Obstructive sleep apnea symptoms:Difference between children and adults. Sleep2000;23(Suppl 4):S140-S141.

83. Brooks LJ: Genetic syndromes affecting breathingduring sleep. In Loughlin GM, Carroll JL, MarcusCL (eds): Sleep and Breathing in Children. NewYork, Marcel Dekker, 2000, pp 737-754.

84. Guilleminault C, Pelayo R, Leger D, et al:Recognition of sleep-disordered breathing in chil-dren. Pediatrics 1996;98:871-882.

85. Papacostas SS, Hadjivasilis V: Klein Levin syn-drome: Report of a case and review of literature.Eur Psychiatr 2000;15:231-235.

86. Gadoth N, Kesler A, Vainstein G, et al: Clinicaland polysomnographic characteristics of 34patients with Klein Levin syndrome. J Sleep Res2001;10:337-341.

87. Billiard M, Rondouin G, Espa F, et al:Pathophysiology of idiopathic hypersomnia:Current studies and new orientation. Rev Neurol(Paris) 2001;157(Part 2):S101-S106.

88. Stahl SM: Awakening to the psychopharmacologyof sleep and arousal: Novel neurotransmittersand wake-promoting drugs. J Clin Psychiatr2002;63:467-468.


Ch13.qxd 01/28/05 10:31 AM Page 181

89. Mitler MM: Evaluation of treatment with stimu-lants in narcolepsy. Sleep 1994;17(Suppl 1):103-106.

90. Billiard M, Besset A, Montplaisir F, et al:Modafinil: A double blind multicentric study.Sleep 1994;17S:107-112.

91. Littner M, Johnson SF, McCall WV, et al: Practiceparameters for treatment of narcolepsy: Anupdate for 2000. Sleep 2001;24:451-466.

92. The U.S. Xyrem Multicenter Study Group: A ran-domized, double blind, placebo-controlled multi-center trial comparing the effects of three doses oforally administered sodium oxybate with placebofor the treatment of narcolepsy. Sleep 2002;25:42-49.

93. Garma L, Marchand F: Non-pharmacologicalapproaches to treatment of narcolepsy. Sleep1994;17S:97-102.


Ch13.qxd 01/28/05 10:31 AM Page 182

Excessive daytime sleepiness in the adolescent isepidemic in our society today. In the adolescentpopulation, hypersomnolence is manifested pri-marily as excessive daytime sleepiness. Symptomsare partially the result of school and lifestyle pres-sures combined with changes in circadian phaseand total sleep time. In contrast, excessive day-time sleepiness being manifested as hypersom-nolence is uncommon in young children. Theprimary manifestation of sleepiness in youngchildren is hyperactivity. Young children respondto the symptoms of sleepiness by increasing theirmotor activity to maintain alertness. However,even in this group, careful questioning does revealsymptoms of daytime sleepiness. The clinical def-inition of idiopathic hypersomnolence in child-hood must take into account this difference inthe clinical response of children to the same “sleeppressure” manifested as excessive daytime sleepi-ness in adults.

Idiopathic hypersomnia (IH) is characterizedby sleepiness and sleep episodes that are usuallyresistible, long lasting, not refreshing, and notassociated with frequent periods of sleep onsetwith rapid eye movement (REM). In addition, IHis sometimes associated with increased totalsleep, difficult awakenings, and sleep drunken-ness.1 Guilleminault and Pelayo have divided IHinto three types based on etiology: those patientswith a positive family history and positivehuman leukocyte antigen–Cw2, those witha history of viral infection, and those notincluded in the first two groups.2 Bassetti andAldrich divided their patients into three groupsbased on clinical symptoms.1 The “classic IH”group tended to have sleepiness that was notoverwhelming, take un-refreshing naps of up to4 hours, and have prolonged night-time sleepand difficulty awakening. The second group theytermed a “narcoleptic” type that presented withoverwhelming excessive daytime sleepiness;

took short, refreshing naps; and awakened with-out sleepiness. The third group was a “mixed”type, with features of both syndromes.

Idiopathic hypersomnolence is by definitiona diagnosis of exclusion. A complete evaluationof other causes of hypersomnolence must beundertaken. The differential diagnosis of idio-pathic hypersomnolence is quite broad, rangingfrom mood disorders to narcolepsy to circadianrhythm disturbances. Table 14–1 provides a frame-work based on etiology for further discussionof the differential diagnosis below.

OBJECTIVE EVALUATIONOF SLEEPINESS IN CHILDREN

Objective evaluation of sleepiness begins witha complete sleep history and physical examina-tion to eliminate other causes of excessive day-time sleepiness (Table 14–2) (see Chapter 1).Sleep logs should be obtained to document sleeptimes, and actigraphy may be utilized to confirmthe sleep log data. On the night prior to labora-tory testing for daytime sleepiness, a polysomno-gram should be obtained to confirm the absenceof a primary sleep disorder. Hypersomnolencecan be characterized quantitatively via the mul-tiple sleep latency test (MSLT)3 and mainte-nance of wakefulness test (MWT).4 Both testsmeasure the time to fall asleep during the day-time. The MSLT instructs the patient to fallasleep, while the MWT instructs the patient tostay awake. The mean adult sleep latency is 18.7minutes. It is generally accepted in the adultpopulation that pathologic average sleep latencyis 5 minutes or less as the result of an MSLT withan average sleep latency of 13.4 minutes.5

The results of the MSLT and MWT often dif-fer. Bonnet and Arand suggest that the MWTmeasures sleep propensity and also the arousalsystem secondary to motivation and posture,

Idiopathic HypersomniaMichael H. Kohrman

14

183

Ch14.qxd 1/25/05 3:22 AM Page 183

while the MSLT is done supinely and measuresonly sleep propensity.6 Carskadon and Dementhave demonstrated that sleep latency in childrenis age dependent.7 Using Tanner stages to groupchildren during adolescence, they observeda drop in average sleep latency on the MSLT of20% between Tanner stages 1 and 3. Thisdecreased sleep latency persisted into later ado-lescence (Tanner stages 4 and 5). In the same

subjects, total sleep time and total REM timeremained constant, but total slow-wave sleep(SWS) time decreased by 70%. The mean sleeplatency of Tanner stages 1and 2 children (meanage 12 years, sleeping their habitual average of9.1 hours) was 18 minutes, compared with thecollege student (mean age 19 years, habitualsleep 7.1 hours) of less than 5.5 minutes. Thus,the definition of pathologic sleepiness in chil-dren must account for this age dependency.

Clinical PresentationBassetti and Aldrich reviewed 42 cases of IH.1

The onset of hypersomnolence began at a meanage of 19 ± 8 years (range 6-43 years). Onset wasassociated with insomnia in 5, weight gain in 2,viral illness in 4, and minor head trauma in 3.Forty-five percent of the patients snored, and60% took one or more involuntary naps duringthe day. These naps lasted 30 minutes or more inover half the subjects in their study, and 77%reported that the naps were un-refreshing. Over50% of the patients had psychiatric complaints.Polysomnographic recordings demonstratedshort sleep latencies of 6.6 ± 5.7 minutes. Meanlatency on the MSLT was 4.3 ± 2.1 minutes. In12 patients who underwent esophageal pressuremonitoring, 5 were found to have upper airwayresistance syndrome. None of these patientsreported improved sleep with continuous posi-tive airway pressure.

Typically, symptoms begin during childhoodand include prolonged night-time sleep andawakening difficulties that often proceed to theonset of daytime sleepiness. Roth has reportedcontinuous nonimperative sleepiness; pro-longed, un-refreshing naps without dreaming;and difficult arousal.8

Polysomnographic studies of the 42 patientsof Bassetti and Aldrich1 demonstrated a sleepefficiency of 93% from a mean of 20 total awak-enings of greater than 1 minute, 8% slow-wavesleep, and 18% REM sleep. They also foundautomatic behaviors in 61% and sleep paralysisin 40%. The amount of sleep per day was 8.4 ±1.9 hours and arousal time in the morning42 minutes. Forza and colleagues studied10 patients with IH.9 All had onset before the ageof 21. There was no statistical difference betweenIH patients and control subjects for the Tannerstage test or sleep latency. Mean sleep latency

184 Chapter 14 Idiopathic Hypersomnia

Neurologic Disorders

Epilepsy and anticonvulsant therapyStrokeTumor/increased intracranial pressureNarcolepsy

Primary Sleep Disorders

Insufficient sleep syndromeInsomniaUpper airway resistance syndromeObstructive sleep apnea syndromeRestless legs syndrome

Behavioral Disorders

Chronic fatigue syndromeKleine-Levin syndromeMood disorders

Circadian Disorders

Delayed sleep phase syndrome

Medical Disorders

Infection: acute and chronicMuscle diseaseMetabolic disordersPrader-Willi syndrome

Table 14–1. Differential Diagnosisof Hypersomnolence

Sleep logPolysomnographyMultiple sleep latency test/maintenance

of wakefulness testActigraphyMRIEEGPsychological testing

Table 14–2. Diagnostic Studies to DefineIdiopathic Hypersomnia

Ch14.qxd 1/25/05 3:22 AM Page 184

was 9.1 minutes. IH patients demonstrateddecreased SWS and increased REM sleep per-cent. Mean sleep latency was 5.7 ± 0.7 minutes.

DIFFERENTIAL DIAGNOSIS OFIDIOPATHIC HYPERSOMNOLENCE

The differential diagnosis of idiopathic hyper-somnolence consists of those syndromes thatcan produce excessive daytime sleepiness. A listbased on etiology is seen in Table 14–1. History,physical examination, polysomnography, andassessment of daytime sleep latency can usuallyexclude most of these diagnoses. Other chaptersin this text cover most of these diagnoses, andthis chapter will be concerned only with thosethat are outside the scope of the primary sleepdisorders.

Primary Sleep DisordersWhile diagnostic criteria are well established forthe primary sleep disorders in adults, normativedata on which to base clinical decisions havebeen only recently available for children. Thus,the reader must apply pediatric normative valueswhen diagnosing sleep-disordered breathing10 orrestless legs syndrome/periodic leg movementdisorder.11 Insufficient sleep syndrome is still theprimary cause of daytime somnolence in mostchildren. A number of studies have led to con-flicting results with regard to school start timesand daytime somnolence. Epstein and associatesin Israel found decreased total sleep time andincreased somnolence in fifth grade pupilswho started school at 7:10 AM compared with8:00 AM.12 A study in Maryland demonstratedno correlation of total sleep time with academicperformance.13

Upper airway resistance syndrome remains adiagnosis that must be excluded in this patientpopulation. No consensus has yet been reachedon how to best evaluate this syndrome in chil-dren. The use of esophageal pressure manometryis controversial, and a consensus on the useful-ness of noninvasive measures, nasal pressure,and arterial tonometry has not been established.

Neurologic DisordersNarcolepsy remains the most difficult clinicalsyndrome to exclude in this group of disorders.

It can present at as young as 4 years of age, andoften cataplexy may not appear until later in life.However, the recent understanding of theabsence of Orexin/hypocretin as the pathophysi-ologic cause of narcolepsy will allow for accuratediagnosis of this patient group as cerebrospinalfluid assay for the absence of Orexin becomes aroutine clinical test.14

Those central nervous system lesions thatproduce increased intracranial pressure canbe associated with hypersomnia. Subduralhematomas can produce an indolent syndromeof decreasing mental status and lethargy.Associated with increased intracranial pressureare often headache and vomiting. Epilepsy canproduce a state in which the patient becomeslethargic and less responsive, the so-called petitmal status or spike-and-wave stupor, because ofthe atypical continuous spike-and-wave patternnoted on the EEG. This condition is oftenresponsive to valproic acid, with rapid return toa normal state. Anticonvulsants can also producesedation and excessive daytime sleepiness orparadoxical hyperactivity in children. Sedationappears to be common with carbamazepine,while hyperactivity is more common with phe-nobarbital.

Behavioral DisordersAttention deficit hyperactivity disorder is anactive area of investigation in terms of its rela-tionship to sleep disruption. Whereas bothobstructive sleep apnea syndrome and restlessleg syndrome/periodic leg movement disorderare associated with hyperactivity in children,attention deficit hyperactivity disorder is multi-factorial in origin and sleep disorders constitutean etiology in only a fraction of children withADHD.

Chronic fatigue/fibromyalgia syndrome isalso a diagnosis of exclusion. While these chil-dren often complain of diffuse muscle aches,trigger points may be seen in others. Thepatients often complain of tiredness rather thantrue hypersomnia. Polysomnography may demon-strate an alpha–delta pattern on the EEG.15

Depression and upper airway resistance syndromemust be excluded in chronic fatigue patientsas well.

Mood disorders are commonly associatedwith sleep difficulties in children. They may be

Idiopathic Hypersomnia Chapter 14 185

Ch14.qxd 1/25/05 3:22 AM Page 185

manifested as insomnia, early morning awaken-ings, daytime sleepiness, or hypersomnia.Vgontzas and coworkers found increased sleeplatency (54 vs 15 minutes), increased wake timeafter sleep onset (79 vs 56.8 minutes), andincreased total wake time (134 vs 72 minutes) inpatients with psychiatric etiologies for theirhypersomnia compared with those with IH. Inpatients with psychiatric disorders REM sleepwas decreased, compared with patients withIH.16 Bipolar disorder may also present withhypersomnia during the depression phase of theillness. At times these children may overlap withKleine-Levin Syndrome (see Chapter 16).

Circadian Disorders

Delayed sleep phase syndrome may also producehypersomnolence. However, these childrenreport normal total sleep times in the presence ofdelayed sleep onset and morning hypersomnia.(see Chapter 9). Non–24-hour sleep–wakecycles as well as advanced sleep phase syndromemay also produce hypersomnolence. Sleep logsand/or actigraphy are often helpful in recogniz-ing these problems.

Medical DisordersAcute illness often results in hypersomnolenceas one attempts to recover from bacterial or viralinfections. Hypersomnolence associated withfever and a stiff neck should alert the examinerto possible meningitis. However, a stiff neck andother meningeal signs may be absent in childrenunder 1 year of age. Lumbar puncture shouldalways be performed if any question of meningi-tis exists. Kubota and colleagues report a case ofacute disseminated encephalomyelitis associ-ated with hypersomnolence as the major pre-senting feature as well as low hypocretin levelsin the cerebrospinal fluid.17 Arii and associatesdescribe hypersomnolence and low hypocretinlevels in a child after removal of a hypothalamictumor.18

Metabolic disorders may produce episodicmental status changes which may appear to pro-duce sleepiness and altered mental status.Disorders of ammonia metabolism, lactate andpyruvate metabolism, and mitochondrial metab-olism can be associated with episodic hyper-somnia. Acute infection or other metabolicstress often precipitates these episodes. Muscle

weakness can also be a source of hypersom-nolence. Primary muscle disease, comprisingthe dystrophies, are often associated withdecreased respiratory effort secondary to fatigue.Polysomnography demonstrates increasingREM-related central apnea as the first sign ofweakness and the need for nocturnal respiratorysupport.19

Prader-Willi syndrome is associated withobesity, craniofacial dysmorphism, hypotonia,hyperphasia, hypersomnia, and hypothalamicdysfunction. Manni et al found sleep-onset REM(SOREM) periods in 5 of 10 children with Prader-Willi syndrome on MSLT.20 None of the patientswere Dr-15 or Dq-6 positive. Hypersomnia andSOREMs could not be accounted for by sleep-disordered breathing alone; however, upper air-way resistance syndrome was not excluded inthese patients.

TREATMENT

The treatment of IH is focused on symptom con-trol, as the primary etiology remains unknown.The goal of therapy is to allow the patient toenjoy normal, alert functioning and restful noc-turnal sleep.

The approach to therapy is sleep hygiene,appropriate use of stimulant medication, andsafety for the patient. Sleep hygiene measuresshould include naps in the afternoon and regularsleep periods of 8 to 10 hours nightly. Drugs,alcohol, or medications that promote sleepinessshould be avoided. Stimulant medication shouldbe used at the lowest effective doses toleratedand growth problems secondary to anorexia canbe a limiting factor of use in some children. Inaddition, a balanced plan of stimulant use mustbe maintained to not interfere with night-timesleep. The newer, longer-acting stimulants(Table 14–3) have been beneficial in manage-ment, keeping medicine out of the school settingand decreasing the need for late afternoon dos-ing. The American Academy of Child andAdolescent Psychiatry has recently produced apractice parameter for stimulant use in children,adolescents, and adults.21

SUMMARY

IH remains a diagnosis of exclusion, and at timesit can be difficult to differentiate it from the


Ch14.qxd 1/25/05 3:22 AM Page 186

other disorders in Table 14–1. Continuedresearch is necessary to better characterize thisdisorder in children. Treatment should be symp-tomatic; naps improve sleep hygiene and stimu-lants should be used in concert to provideimproved quality of life in these children. Therole of hypocretin in the etiology of IH in chil-dren requires further exploration.

References

1. Bassetti C, Aldrich MS: Idiopathic hypersomnia:A series of 42 patients. Brain 1997;120:1423-1435.

2. Guilleminault C, Pelayo R: Idiopathic CentralNervous System Hypersomnia. In Kryger MH,Roth T, Dement WC: Principles and Practiceof Sleep Medicine, 3rd ed. Philadelphia, WBSaunders, 2000, pp 687-692.

3. Carskadon MA, Dement WC, Mitler MM, et al:Guidelines for the multiple sleep latency test(MSLT): A standard measure of sleepiness. Sleep1986;9:519-524.

4. Mitler MM, Gujavarti KS, Browman CP:Maintenance of wakefulness test: A polysomno-graphic technique for evaluation of treatment effi-cacy in patients with excessive somnolence.Electroencephalogr Clin Neurophysiol. 1982;53:658-661.

5. Mitler MM, Carskadon MA, Hirshkowitz M:Evaluating Sleepiness. In Kryger MH, Roth T,Dement WC (eds): Principles and Practice ofSleep Medicine, 3rd ed. Philadelphia, WBSaunders, 2000.

6. Bonnet MH, Arand DL: Arousal componentswhich differentiate the MWT from the MSLT.Sleep 2001;24:441-447.

7. Carskadon MA, Dement WC: Sleepiness in theNormal Adolescent. In Guilleminault C (ed):Sleep and Its Disorders in Children. New York,Raven Press, 1987, pp 53-66.

8. Roth B: Narcolepsy and Hypersomnia. Basel,Karger, 1980.

9. Forza E, Gaudreau H, Petit D, Montplaisir J:Homeostatic sleep regulation in patients withidiopathic hypersomnia. Clin Neurophysiol2000;111:277-282.

10. Goodwin JL, Enright PL, Morgan WJ, et al:Correlates of obstructive sleep apnea in 6-12 yearold children: The Tucson Children Assessment ofSleep Apnea Study (TUCASA) (Abstract). Sleep2002;25(Suppl):A80.

11. Kohrman MH, Kerr SL, Schumacher S: Effect ofsleep disordered breathing on periodic leg move-ments of sleep in children (Abstract). Sleep1997;20(Suppl):635.

12. Epstein R, Chillag N, Lavie P: Starting times ofschool: Effects on daytime functioning of fifthgrade children in Israel. Sleep 1998;21:250-256.

13. King J, Gould B, Eliasson A: Association of sleep andacademic performance. Sleep Breath 2002;6:45-48.

14. Nishino S, Ripley B, Overeem S, et al: Hypocretin(Orexin) deficiency in human narcolepsy. Lancet2000;355:39-40.

15. Whelton CL, Salit I, Moldofsky H: Sleep, Epstein-Barr virus infection, musculoskeletal pain anddepressive symptoms in chronic fatigue syn-drome. J Rheumatol 1992;19:939-943.

16. Vgontzas AN, Bixler EO, Kales A, Criley C, Vela-Bueno A: Differences in nocturnal and daytimesleep between primary and psychiatric hypersom-nia: Diagnostic and treatment implications.Psychosomat Med 2000;62:220-226.

Idiopathic Hypersomnia Chapter 14 187

Duration Special Name of Action Contraindications ConsiderationsMethylphenidate 4-6 hr Increased risk of

Time-release 8-12 hr cardiac arrhythmias Longer actionpreparations

Dexmethylphenidate 4-6 hr Isomer of parent compound

Dextroamphetamine 4-6hrTime spansMixed salts 4-6 hrExtended-release 8-12 hr Longer duration

preparations of actionMonafinil 6-8 hr Non-amphetamine

drugPemoline 12 hr High risk of liver disease

Table 14–3. Stimulants Used for Treatment of Idiopathic Hypersomnia

Ch14.qxd 1/25/05 3:22 AM Page 187

17. Kubota H, Kanbayashi T, Tanabe Y, et al: A case ofacute disseminated encephalomyelitis presentinghypersomnia with decreased hypocretin levelin cerebrospinal fluid. J Child Neurol 2002;17:537-539.

18. Arii J, Kanbayashi T, Tanabe Y, et al: A hypersom-nolent girl with decreased CSF hypocretin levelafter removal of a hypothalamic tumor. Neurology2001;56:1775-1776.

19. Kerr SL, Kohrman MH: Polysomnogram inDuchenne muscular dystrophy. J Child Neurol1994;9:332-334.

20. Manni R, Politini L, Nobili L, Ferrillo F, et al:Hypersomnia in the Prader-Willi syndrome:Clinical- electrophysiological features and under-lying factors. Clin Neurophysiol 2001;112: 800-805.

21. Greenhill LL, Plixzka S, Dulcan MK: Work Groupon Quality Issues Practice Parameter for the useof stimulant medications in the treatment of chil-dren, adolescents, and adults. J Am Acad ChildAdolesc Psychiatry 2002;41:26S-49S.


Ch14.qxd 1/25/05 3:22 AM Page 188

Children who suffer significant head trauma fre-quently experience significant sleep distur-bances after the injury, particularly when thetrauma is severe enough to result in major loss ofconsciousness. Nonetheless, sleep disturbancesmay also follow minor trauma in which a briefloss of consciousness occurs. In all instances,sleep patterns after the insult vary notably fromthe pretrauma sleep habits.

Closed-head trauma is the most commonevent resulting in post-traumatic hypersomnia.Head injury as a result of an automobile accidentis the most common cause. Even so, similarsymptoms have been reported after neurosurgi-cal procedures and other brain traumas. Itappears that the mode of injury is less importantthan the location. Symptomatically, there is avariable period of initial coma that evolves into apost-traumatic hypersomnolence and excessivedaytime sleepiness with or without sleep attacksor unintentional sleep episodes. Nocturnal sleepmay or may not be prolonged compared with thepreinjury period. When total sleep time withineach 24-hour day is increased, the term “post-traumatic hypersomnia” is fitting. Associatedsymptoms are typically due to daytime sleepi-ness (e.g., concentration difficulties, amnesia ofrecent events, fatigue, and occasional visualproblems). Chronic headache and minor neuro-logic signs of traumatic brain injury may also bepresent.

Less commonly, the initial head injury is fol-lowed by problems initiating and maintainingsleep either with or without some degree of sub-jective daytime sleepiness. Post-traumatic psy-chogenic insomnia is also seen. Reports ofpost-traumatic narcolepsy and cataplexy havebeen described, but as it has been pointed out insome case reports, post-traumatic narcolepsy/cataplexy occurred 5 years after the head injury.Other case reports do not appear to be consis-

tent with the narcolepsy/cataplexy syndromeand more likely represent post-traumatic hyper-somnia.1

Symptoms resulting from closed-head injurydepend on the location of injury within sleep-regulating brain areas. Although data are limited,there appear to be several correlations, withhypersomnolence being the most common fea-ture. Areas of the brain involved are most com-monly those related to maintaining wakefulness,including but not limited to the brainstem retic-ular formation, posterior hypothalamus, and theregion of the third ventricle. Shearing forcesalong the direction of main fiber pathways canlead to microhemorrhages in these areas. Highcervical cord trauma has also been known tocause sleepiness and unintentional sleepepisodes.2,3 Whiplash injury may result inhypersomnia, with consequent sleep-disorderedbreathing. In these cases, the hypersomnolenceappears to be secondary to the respiratory abnor-mality.4

Countercoup injuries commonly occur at thebase of the skull and may result in organic post-traumatic sleeplessness. These types of injuryoccur in areas of bony irregularities (especiallythe sphenoid ridges), with consequent damageto the inferior frontal and anterior temporalregions,5,6 including the basal forebrain.7

Rodrigues and Silva have reported a patientwith aggressive body movements during rapideye movement (REM) sleep and periodic limbmovements after a traumatic brain injury—symptoms suggesting REM-sleep behavior disor-der.8 Multiple sleep latency testing revealedextremely short latencies. It was suggested thatdopaminergic pathways potentiate hypothalamichypocretin abnormalities that are involved in thepathophysiology.

Closed-head injury may also involve the supra-chiasmatic nucleus within the hypothalamus,

Post-Traumatic HypersomniaStephen H. Sheldon

15

189

Ch15.qxd 1/25/05 3:23 AM Page 189

resulting in disturbances of circadian rhythmicity.These symptoms then produce a combination ofhypersomnia and insomnia. Complete sleep phasereversal has also been reported.9

The prevalence of post-traumatic sleep disor-ders has not been determined. Head injury dueto motor vehicle accidents is common and rep-resents over 50% of all reported head injuries inindustrialized countries.10 Unintentional falls,intentional violence, and increased risk-takingbehavior also contribute to the frequency ofhead trauma in children.

Evaluating the underlying cause of hyper-somnia in patients who have suffered headtrauma involves determining the degree ofinjury and potential locations within the brainthat the trauma may have affected. In patientsrecovering from coma, the relationship is clear.Nonetheless, a careful history must always betaken, and the presence or absence of prior sleepdisturbances of similar types should be deter-mined. Obscure causes of hypersomnolenceshould also be considered in the evaluation ofthe patient with suspected post-traumatic hyper-somnia. Hydrocephalus, subdural hematoma,meningitis, encephalitis, and seizure disordersshould be assessed. The possibility of drug use(pharmacologic and/or recreational) should alsobe considered. Psychogenic insomnia related tohead injury must be assessed. When compre-hensive nocturnal polysomnography reveals thepresence of sleep-disordered breathing afterhead trauma, the sleep-related breathing disor-ders must first be managed before determiningthe extent to which each comorbid conditioncontributes to the hypersomnolence. Patientswith anatomic risk factors for obstructive sleepdisordered breathing may abruptly developsymptoms of head trauma. Similarly, clear-cutnarcolepsy with cataplexy has been shown tooccur after head trauma in patients who exhibithuman leukocyte antigen–DQ B1-0602. Therecognition that head trauma can be a precipitat-ing factor to syndromes leading to excessivedaytime sleepiness is important, and a compre-hensive evaluation of all patients who exhibitexcessive sleepiness after head trauma is needed.

Historical information forms the basis for theevaluation of the patient with possible post-traumatic hypersomnia in order to temporally doc-ument the association of the hypersomnia withthe trauma, determine the presence of preexist-

ing sleep-related pathology, and assess the evolu-tion of symptoms after the head injury. Physicalexamination can reveal the possibility of minorfocal neurologic signs, especially those of brain-stem origin.

In the hypersomnolent patient, polysomnog-raphy and multiple sleep latency testing shouldbe considered in order to determine the nature ofthe post-traumatic nocturnal sleep disturbancesas well as rule out coexisting pathologies such asobstructive sleep apnea syndrome and periodiclimb movement disorder. Polysomnography inpatients with post-traumatic hypersomnia gener-ally reveals an increase in sleep duration with orwithout changes in other aspects of sleep archi-tecture.11 The findings in patients with post-traumatic insomnia include long sleep latencies,low sleep efficiency, and a decrease in nightlysleep duration, especially in stage 2 sleep.12

Continuous polysomnography has confirmedan increase in total sleep per 24 hours in somepatients with post-traumatic hypersomnia.12 It isof interest that polysomnography during thecomatose period has prognostic value for the devel-opment of full recovery without post-traumaticsleep disturbances. The normal frequency ofsleep spindles, K-complexes, and normal non-REM and REM sleep architecture are favorableprognostic signs.13-16

If daytime sleepiness is a major complaint, amultiple sleep latency test is of help in deter-mining severity and circadian variation. In post-traumatic hypersomnia with daytime sleepi-ness, sleep latency may be markedly short-ened.11 Patients with post-traumatic insomniaexhibit prolonged sleep latencies on nocturnalpolysomnography, and multiple sleep latencytesting may be normal.12 The presence ofobstructive sleep apnea syndrome, other formsof sleep-disordered breathing, and/or sleep onsetREM periods (on nocturnal polysomnographyand/or multiple sleep latency testing) does notrule out the possibility that head trauma is a pre-cipitating or contributing factor.

Medical evaluation should include assess-ment of levels of activity and the habitualsleep–wake patterns of the youngster before thehead trauma. Assessment should include but notbe limited to the following: interviews of parentsand teachers, school performance records, priormedical history, and school tardiness and absen-teeism.

190 Chapter 15 Post-Traumatic Hypersomnia

Ch15.qxd 1/25/05 3:23 AM Page 190

The exact prognosis of patients with post-traumatic hypersomnia has not yet been eluci-dated. There are few systematic follow-upstudies. Based mostly on anecdotal information,once the hypersomnia has stabilized, sleep dis-turbances related to the head trauma show littleadditional change. Patients with post-traumatichypersomnia generally respond fairly well totreatment, and there are no known specific com-plications.

Youngsters with post-traumatic hypersomniamay require stimulant medication such asmethylphenidate or amphetamines in doses sim-ilar to those required for narcolepsy/cataplexysyndrome. In older children the drug of choicemay be modafinil, which has fewer side-effectsoverall than phenylethylamine stimulants. Thedosage of modafinil in children has yet to bedetermined, but a starting dose of 200 mg withgradual titration to 400 mg (if needed) might beattempted. On the other hand, modafinil has nonoradrenergic properties, and patients withsevere head trauma who have more significantlesions and also complain of intellectual slow-ness might benefit more from phenylethyl-amines. These medications will have a general“activating” effect that is not devoted solely tosleepiness.

Any coexisting sleep pathology or neurologicdisease requires independent management.Nasal continuous positive airway pressure maybe helpful in secondary sleep-disordered breath-ing. The potential beneficial effects of naps havenot been studied; they should probably berestricted to less than 30 minutes to avoid sig-nificant sleep inertia after the nap. They shouldbe taken when the patient feels the sleepiest butnot within 4 or 5 hours of habitual nocturnalsleep time.

References

1. Bonduelle M, Bouygues P, Faveret C: Narcolepsiepost-traumatique. Lille Med 1959;4:719-721.

2. Adey W, Bors E, Porter RW: EEG patterns afterhigh cervical lesions in man. Arch Neurol1968;19:377-383.

3. Hall CS, Danoff D. Sleep attacks: Apparent rela-tionship to atlantoaxial dislocation. Arch Neurol1975;32:58-59.

4. Guilleminault C, Yuen KM, Gulevich MG, et al:Hypersomnia after head-neck trauma: Amedicolegal dilemma. Neurology 2000;54:653-659.

5. Courville CB: Coup-contrecoup mechanism ofcranio-cerebral injuries: Some observations.Arch Surg 1942;55:19-43.

6. Ommaya AK, Grubb RL, Naumann RA: Coupand contre-coup injury: Observations on themechanics of visible brain injuries in the rhesusmonkey. J Neurosurg 1971;35:503-516.

7. Sterman MB, Clemente CD: Forebrain inhibitorymechanisms: Cortical synchronization inducedby basal forebrain stimulation. Exp Neurol1962;6:91-102.

8. Rodrigues RN, Silva AA: Excessive daytimesleepiness after traumatic brain injury:Association with periodic limb movements andREM behavior disorder. Case report. Arquivosde Neuro-Psiquiatria 2002;60:656-660.

9. Billiard M, Negri C, Baldy-Moulignier M, et al:Organisation du sommeil chez les sujets atteintsd’inconscience post-traumatique chronique. RevElectroencephalogr Neurophysiol 1979;9:149-152.

10. Marshall LF, Becker DP, Bowers SA, et al: TheNational Traumatic Coma Data Bank, Part 1:Design, purpose, goals, and results. J Neurosurg1983;59:276-284.

11. Guilleminault C, Faull KF, Miles L, van denHoed J: Posttraumatic excessive daytime sleepi-ness: A review of 20 patients. Neurology1983;33:1584-1589.

12. Manseau C, Broughton RJ: Severe head injury:Long term effects on sleep, sleepiness and per-formance. Sleep Res 1990;19:335.

13. Passouant P, Cadilhac J, Delange M, et al:Different electrical stages and cyclic organizationof post-traumatic comas: Polygraphic recordingsof long duration. Electroencephalogr ClinNeurophysiol 1965;18:726P.

14. Bergamasco B, Bergamini L, Doriguzzi T:Clinical value of the sleep electroencephalo-graphic patterns in posttraumatic coma. ActaNeurol Scand 1968;44:495-511.

15. Bricolo A, Gentilomo A, Rosadini G, Rossi GF:Long-lasting post-traumatic unconsciousness: Astudy based on nocturnal EEG and polygraphicrecording. Acta Neurol Scand 1968;44:512-532.

16. Lessard CS, Sances A, Larson SJ: Period analysisof EEG signals during sleep and posttraumaticcoma. Aerospace Med 1974;45:664-668.

Post-Traumatic Hypersomnia Chapter 15 191

Ch15.qxd 1/25/05 3:23 AM Page 191

Kleine-Levin syndrome is an unusual disorderfeaturing recurrent episodes of excessive sleepi-ness and prolonged total sleep time. It is rare,with an onset typically in late adolescence, butcases of younger and older individuals have beenreported. It has been suggested that a predispos-ing event occurs before the onset of symptoms innearly half of the reported cases. Symptoms thatare flulike often precede the onset of episodes ofrecurrent hypersomnia.

CLINICAL CHARACTERISTICSOF KLEINE-LEVIN SYNDROME

Hypersomnia is the characteristic symptom ofKleine-Levin syndrome. Sleepiness is profound,and during spells patients may sleep continu-ously for more than 20 hours. This excessivesleepiness and true hypersomnia can developsuddenly. However, symptoms typically have amore gradual onset of 1 to 7 days. During spellsof hypersomnia, patients rarely leave their bedsand sleep continuously. Sleep may be calm, butat times agitated, restless sleep occurs and vividdreams are occasionally reported.

In addition to excessively long sleep episodes,abnormal behaviors, including but not limited tocompulsive and excessive overeating and sexu-ally acting out behaviors, occur. Other mentaldisturbances may be present. Compulsiveovereating and hypersexuality may not be pres-ent with each episode. Excessive caloric intakeduring spells frequently results in weight gain bythe end of the spell.

Overt expression of hypersexuality mayinclude indiscriminate sexual advances, mastur-bation, and/or public display of sexual fantasies.Sexually acting-out is reported in approximatelyone third of the males with Kleine-Levin syn-drome. Hypersexuality is less often reported infemales. The presence of excessive eating and/or

hypersexuality is not required for the diagnosisof Kleine-Levin syndrome. In many cases,hypersomnia and its recurrence may be the onlysymptom.1 Psychological symptoms vary con-siderably, and irritability is common. Confusionand visual and auditory hallucinations fre-quently occur.

Recurrent episodes of hypersomnia may bebrief and last less than 1 week or may be pro-longed and last up to 30 days. Characteristic ofKleine-Levin syndrome are the recurrence ofepisodes of hypersomnia and abnormal behav-iors. Patients are normal between episodes.Nevertheless, neuropsychological sequelae,changes in personality, and a decrease in schoolperformance have been reported after a secondhypersomnolent episode. Physical examinationis generally normal.

Although Kleine-Levin syndrome is charac-terized by the triad of recurrent spells of hyper-somnolence, hyperphagia, and hypersexuality, itis likely that incomplete manifestation is morecommon than the complete triad. Isolated,abnormal sleepiness and recurrent periods ofhypersomnia without associated symptoms maybe a more common manifestation.

Periods of hypersomnia gradually decrease infrequency as the youngster ages. Spells graduallybecome less severe and eventually resolve.2

Nonetheless, patients with recurrent episodes ofhypersomnia occurring 20 years after the initialonset of symptoms have been reported.2-4

The social and professional consequences arenot negligible: Students miss classes, and youngworkers may be fired because of repeatedabsences.

Menstruation-associated hypersomnia hasbeen reported in female patients. Periods ofhypersomnolence recur and are coupled withmenses. Occasionally, mental disturbances alsooccur, frequently during the menstrual cycle.5

Kleine-Levin Syndrome and RecurrentHypersomnias

Stephen H. Sheldon

16

193

Ch16.qxd 1/25/05 3:23 AM Page 193

Several female patients with recurrent hyper-somnia had a notable absence of symptoms.1

The prognosis for menstruation-related hyper-somnia is similar to that of Kleine-Levin syn-drome.

The cause of recurrent hypersomnia isunknown. A possible viral etiology may be pres-ent in some cases, and in others there may be asuggestion of local encephalitis in the region ofthe diencephalon.6-8 Recurrent transient episodesof unresponsiveness and clearly identified stage 2sleep patterns with the presence of sleep spindlesdue to bilateral paramedian thalamic infarctionshave also been reported.9 Although the exactcause of recurrent hypersomnia and other mani-festations is still unclear, the symptoms of hyper-somnolence, excessive and compulsive eating,hypersexuality, recurrence, and absence of anyidentifiable abnormalities between spells aresuggestive of a functional abnormality at thelevel of the diencephalon. The hypothalamusmay also be involved. Indeed, identical symp-toms have been reported in patients with tumorsof the hypothalamus or third ventricle and inpatients with epidemic encephalitis. Most of theliterature and case reports suggest that thisrecurrent hypersomnia is caused by dysfunctionof the hypothalamus and midbrain limbic sys-tem. There also is evidence that brainstem dys-function may be involved. Recurrent hypersomniaassociated with decreased blood flow in the thala-mus on single-photon emission computed tomog-raphy has also been reported.10 During theremission period, there were no abnormal dataresulting from this testing.

Neuroendocrine function in patients withKleine-Levin syndrome has been assessed.However, only a limited number of subjectshave been investigated, and most of the litera-ture consists of case reports. Investigation iscomplicated by the inability to predict the tim-ing of episodes, lack of a prodromal period,and relatively rapid recovery once symptomsbegin.

Nonetheless, some abnormal laboratory datareported to date include a paradoxical growthhormone response to thyroid-releasing hormonestimulation,11 a blunted cortisol response toinsulin-induced hypoglycemia,12,13 and anabsent thyroid-stimulating hormone response tothyroid-releasing hormone.13 These findingssuggest a possible dysfunction within the hypo-

thalamic–pituitary axis. However, other basaland post-stimulation values of hormones areusually normal, and laboratory values are nor-mal during the asymptomatic periods betweenspells.

Studies of nocturnal or 24-hour secretorypatterns of pituitary hormones have been con-ducted in a limited number of patients. A nor-mal secretory pattern of growth hormone wasreported in one patient, but the sampling at 4-hour intervals was sparse.14 On the other hand,an elevated growth hormone secretory patternwas reported in two patients.15 A normal secre-tory pattern of cortisol and somewhat abnormalpatterns of growth hormone have been identi-fied in four other patients.16 A normal 24-hourpattern of melatonin, prolactin, and cortisolsecretion has been reported.17 Gadoth and col-leagues found an increased nocturnal prolactinsecretory pattern and an abnormally flat noctur-nal luteinizing hormone secretory pattern,whereas follicle-stimulating hormone and thy-roid-stimulating hormone secretory patternswere normal.11 Chesson and Levine compared24-hour secretions in the symptomatic andasymptomatic periods and found that valuesobtained during nocturnal sleep showed a sig-nificantly decreased growth hormone butincreased prolactin and thyroid-stimulatinghormone in a direction that supports thehypothesis that dopaminergic tone is reducedduring the symptomatic period in Kleine-Levinsyndrome.18 However, Mayer and colleaguesfound only minor hormonal changes in fivepatients.19 Altogether, these data suggest somefunctional disturbance in the hypothalamic-pituitary axis in Kleine-Levin syndrome, but thedisturbance may be in response to the sleep-related and behavioral changes rather than acause. An observation supporting a dien-cephalic dysfunction is the occurrence of dysau-tonomic features in some patients.20

Differentiating Kleine-Levin syndrome fromorganic etiologies is often difficult. Diagnosis isoften based on the clinical presentation of recur-rence of symptoms with asymptomatic intervalsand progressive improvement to resolution. Inaddition, it is often a diagnosis of exclusion.Hypersomnia may recur with space-occupyinglesions of the central nervous system, idiopathicrecurring stupor, and certain psychological/psy-chiatric conditions.

194 Chapter 16 Kleine-Levin Syndrome and Recurrent Hypersomnias

Ch16.qxd 1/25/05 3:23 AM Page 194

Tumors in the region of the third ventricle,such as cysts, astrocytomas, and /or cranio-pharyngiomas, may be responsible for intermit-tent obstruction of the third ventricle, leading toheadache, vomiting, sensory disturbances, andintermittent impairment of alertness. Less fre-quently, tumors in other locations of the centralnervous system may result in hypersomnia.Tumors in the middle fossa may disrupt thesuprachiasmatic nucleus, and an irregular sleepwake pattern or a free-running state mightresult. If a free running state occurs, hypersom-nia may alternate with sleeplessness at a regularinterval as the pacemaker cycles at its inherentrhythm. Recurrent hypersomnia may alsodevelop after encephalitis or head trauma.Periodic hypersomnia has also been reported ina patient with a Rathke cleft cyst.21

Major recurrent depression and bipolar affec-tive disorder can be associated with excessivesleepiness.22 Patients with psychogenic recur-rent hypersomnia complain of extreme sleepi-ness and fatigue and may spend many hours inbed. Continuous night and day polygraphicrecording often fails to demonstrate increasedtotal sleep time despite the complaint of sleepi-ness.23

EVALUATION

The diagnosis of Kleine-Levin syndrome is typi-cally based on clinical features. Laboratory eval-uations are occasionally useful, since thediagnosis may exclude other similar pathologies.Complete blood cell count, platelet count, elec-trolytes, renal and liver function tests, calcium,phosphorus, serum protein electrophoresis,immunoglobulin, antinuclear antibodies, rheuma-toid factor, and serum titers for herpes simplex,Epstein-Barr virus, cytomegalovirus, varicellazoster, mumps, and measles have all been foundto be normal in patients with Kleine-Levin syn-drome both during and between episodes.Cerebrospinal fluid evaluation with cultures forbacteria, mycobacterium, virus, and fungi havealso been shown to be negative.

Routine EEG obtained during attacks mayshow generalized slowing of background activityor may be unremarkable. MRI is normal duringand between spells of hypersomnia.19 Prolongedpolysomnographic monitoring may reveal theincreased total sleep time. During symptomatic

periods, the multiple sleep latency test can revealabnormal sleep latencies and sleep-onset REMperiods. Therefore, it has been suggested that themultiple sleep latency test may be useful in diag-nosis, especially when polysomnography andmultiple sleep latency testing are performed noearlier than the second night after the onset ofhypersomnolence.24 During asymptomatic peri-ods polysomnography and multiple sleeplatency testing are both normal.

TREATMENT

Kleine-Levin syndrome is typically treatedsymptomatically. Interventions are focused onterminating the episode of hypersomnia. Ampheta-mines, methylphenidate, and modafinil havebeen used in both treatment and prevention, butno treatment has been consistently successful.Generally, the effectiveness of maintaining wake-fulness using this therapeutic approach does notexceed a few hours.

References

1. Kesler A, Gadoth N, Vainstein G, et al: KleineLevin syndrome (KLS) in young females. Sleep2000;23:563-567.

2. Critchley M: Periodic hypersomnia and megapha-gia in adolescent males. Brain 1962;85:627-656.

3. Gran D, Begemann H: Neue Beobachtungen beieinem Fall von Kleine-Levin Syndrom. MunchMed Wochenschr 1973;115:1098-1102.

4. Bucking PH, Palmer WR: New contribution to theclinical aspects and pathophysiology of theKleine-Levin syndrome. Munch Med Wochenschr1978;120:1571-1572.

5. Billiard M, Guilleminault C, Dement WC: A men-struation-linked periodic hypersomnia. Neurology1975;25:436-443.

6. Takrani LB, Cronin D: Kleine-Levin syndrome ina female patient. Can Psychiatr Assoc J 1976;21:315-318.

7. Carpenter S, Yassa R, Ochs R: A pathological basisfor Kleine-Levin syndrome. Arch Neurol1982;39:25-28.

8. Fenzi F, Simonati A, Crosato F, et al: Clinical fea-tures of Kleine-Levin syndrome with localizedencephalitis. Neuropediatrics 1993;24:292-295.

9. Bjornstad B, Goodman SH, Sirven JI, Dodick DW:Paroxysmal sleep as presenting symptom of bilat-eral paramedian thalamic infarction. Mayo ClinProc 2003;78:347-349.

10. Nose I, Ookawa T, Tanaka J, et al: Decreasedblood flow of the left thalamus during somnolent

Kleine-Levin Syndrome and Recurrent Hypersomnias Chapter 16 195

Ch16.qxd 1/25/05 3:23 AM Page 195

episodes in a case of recurrent hypersomnia.Psych Clin Neurosci 2002;56:277-278.

11. Gadoth N, Dickerman Z, Bechar M, et al: Episodichormone secretion during sleep in Kleine-Levinsyndrome: Evidence for hypothalamic dysfunc-tion. Brain Dev 1987;9:309-315.

12. Koerber RK, Torkelson R, Haven G, et al:Increased cerebrospinal fluid 5-hydroxytrypta-mine and 5-hydroxyindoleacetic acid in Kleine-Levin syndrome. Neurology 1984;34:1597-1600.

13. Fernandez JM, Lara I, Gila L, et al: Disturbedhypothalamic-pituitary axis in idiopathic recur-ring hypersomnia syndrome. Acta Neurol Scand1990;82:361-363.

14. Gilligan BS: Periodic megaphagia and hypersom-nia—an example of the Kleine-Levine syndromein an adolescent girl. Proc Aust Assoc Neurol1973;9:67-72.

15. Kaneda H, Sugita Y, Masoaka S, et al: Red bloodcell concentration and growth hormone releasein periodic hypersomnia. Wak Sleep 1977;1:369-374.

16. Hishikawa Y, Iijima S, Tashiro T, et al:Polysomnographic Findings and GrowthHormone Secretion in Patients with PeriodicHypersomnia. In Koella WP (ed): Sleep 1980.Basel, S Karger, 1981, pp 128-133.

17. Thompson C, Obrecht R, Franey C, et al:Neuroendocrine rhythms in a patient with the

Kleine-Levin syndrome. Br J Psychiatry 1985;147:440-443.

18. Chesson AL, Levine SN: Neuroendocrine evalua-tion in Kleine-Levin syndrome: Evidence ofreduced dopaminergic tone during periods ofhypersomnolence. Sleep 1991;14:226-232.

19. Mayer G, Leonhard E, Krieg J, Meier-Ewert K:Endocrinological and polysomnographic findingsin Kleine-Levin syndrome: No evidence for hypo-thalamic and circadian dysfunction. Sleep1998;21:278-284.

20. Hegarty A, Merriam AE: Autonomic events inKleine-Levin syndrome. Am J Psychiatry 1990;147:951-952.

21. Autret A, Lucas B, Mondon K, et al: Sleep andbrain lesions: A critical review of the literatureand additional new cases. Neurophysiol Clin2001;31:356-375.

22. Jeffries JJ, Lefebvre A: Depression and mania asso-ciated with Kleine-Levin-Critchley syndrome.Can Psychiatr Assoc J 1973;18:439-444.

23. Billiard M, Cadilhac J: Les hypersomnies recur-rentes. Rev Neurol (Paris) 1988;144:249-258.

24. Rosenow F, Kotagal P, Cohen BH, et al: Multiplesleep latency test and polysomnography in diag-nosing Kleine-Levin syndrome and periodichypersomnia. J Clin Neurophysiol 2000;17:519-522.

196 Chapter 16 Kleine-Levin Syndrome and Recurrent Hypersomnias

Ch16.qxd 1/25/05 3:23 AM Page 196

Sleep-disordered breathing (SDB) is a commonand serious cause of morbidity during child-hood. This chapter is concerned with diagnosingthe spectrum of obstructive SDB, ranging fromthe frank, intermittent occlusion seen inobstructive sleep apnea syndrome (OSAS), topersistent, primary snoring (Fig. 17–1). OSAS ischaracterized by recurrent episodes of partial orcomplete airway obstruction resulting in hypox-emia, hypercapnia, and/or respiratory arousal.The sleep fragmentation and gas exchangeabnormalities observed with OSAS may produceserious cardiovascular and neurobehavioralimpairment. The upper airway resistance syn-drome (UARS) is characterized by brief, repeti-tive respiratory effort–related arousals duringsleep in the absence of overt apnea, hypopnea, orgas exchange abnormalities1 (Fig. 17–2). It hasbeen linked to significant cognitive and behav-ioral sequelae in children, including learningdisabilities, attention deficit, hyperactivity, andaggressive behavior.2 Obstructive hypoventila-tion features prolonged increased upper airwayresistance accompanied by gas exchange abnor-malities, but not frank apnea or hypopnea.3

Children with primary snoring may haveincreased respiratory effort, but they lack identi-fiable arousals, including respiratory effort–related,electroencephalographic (EEG), and subcorticalarousals.4 Although primary snoring has beentraditionally defined as a benign condition with-out polysomnographic abnormalities,5 recentevidence suggests that the increased respiratoryeffort in primary snoring per se may be asso-ciated with untoward neurobehavioral conse-quences.6,7

Habitual snoring has been reported in 3% to12% of the general pediatric population,although only 1% to 3% have OSAS.8-10 It isthus important to distinguish primary snoringfrom more serious forms of obstruction. Early

recognition of SDB is important for effectivetreatment with adenotonsillectomy or continu-ous positive airway pressure. Establishing strictcriteria for OSAS diagnosis and severity is thebasis for optimizing the surgical and medicalmanagement of this condition.11 The diagnosisand management of pediatric OSAS continuesto evolve as more precise measures of flow lim-itation and sleep fragmentation are introduced.Both the intermittent hypoxemia and the sleepfragmentation characteristic of OSAS pose a riskto the vulnerable developing brain. Theincreased recognition of subtle neurocognitiveimpairments in children with SDB has forced cli-nicians to rethink the threshold of disease requir-ing intervention.

HISTORY AND PHYSICALEXAMINATION

The high incidence of OSAS in children mandatesthat screening inquiries about sleep disturbancesshould be a routine part of the primary care inter-view12 (Table 17–1). Particular attention shouldbe given to conditions known to exacerbate SDB,such as craniofacial abnormalities, neuromuscularweakness, and genetic conditions. Parents typi-cally provide the clinical history, as the patients areusually oblivious to their condition. Loud snoring,on a nightly basis, is nearly universal in pediatricOSAS. A parental report of a snoring child is anaccurate predictor of polysomnographic snoring,but not of OSAS.13 The snoring is often accompa-nied by labored breathing, hyperextension of theneck, and witnessed apneic pauses. Subjectivereports of excessive daytime sleepiness are infre-quently present in children with OSAS. A recentstudy reported that only 7.5% of the children withpolysomnographically proven OSAS had a historyof daytime sleepiness.14 Using an objective meas-ure of sleepiness, such as a multiple sleep latency

Diagnosis of Obstructive Sleep ApneaSyndrome in Infants and Children

Eliot S. Katz

Carole L. Marcus

17

197

Ch17.qxd 1/25/05 3:24 AM Page 197

test, reveals that only 13% of children with OSAShave a sleep latency of less than 10 minutes.14

Although not considered sleepy by adult stan-dards, children with SDB are sleepier than unaf-fected children.15 Increasing body mass index andapnea index (usually greater than 15 to 20 events

per hour) were independently correlated withshorter sleep latencies.14

The use of standardized screening question-naires for OSAS has been disappointing.Brouillette and coworkers presented a question-naire suitable for distinguishing children with

198 Chapter 17 Diagnosis of Obstructive Sleep Apnea Syndrome in Infants and Children

Airw

ay c

olla

psib

ility

and/

or r

esis

tanc

e

Normal Primarysnoring

(PS)

UpperAirway

resistancesyndrome(UARS)

Obstructivehypoventilation

(OH)

Obstructivesleepapnea

syndrome(OSAS)

Figure 17–1. Spectrum of obstructive sleep-disordered breathing in children. (Adapted with permis-sion from Loughlin GM: Obstructive sleep apnea syndrome in children: Diagnosis and management.In Loughlin GM, Carroll JL, Marcus CL (eds): Sleep and Breathing in Children: A DevelopmentalApproach. New York, Marcel Dekker, 2000, pp 625-650.)

BODY

C4A1

C3A2

O1C4

O2C3

ROC

LOC

CHIN

EKG

R-LEG

L-LEG

SNORE

FLOW

NAF

THO

ABD

Raw data1:03:04

SpO2

10" 20" 30" 40" 50" 60"

RS RS RS RS RS RS RS RS RS RS RS RS RS RS RS RS RS RS RS RS RS RS RS RS RS RS RS RS RS RS

0

98 98 98 98 98 98 98 98 98 98 98 98 98 98 98 98 98 98 97 97 97 97 97 97 97 98 98 98 98 98

0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

Figure 17–2. A 60-second epoch from a 16 year old with snoring and excessive daytime sleepiness.Note flow limitation in the nasal pressure tracing leading to an EEG arousal. Inspiration is upward. Abd,abdomen; Body, body position; Flow, thermistor; NAF, nasal pressure; RS, right side; SpO2, pulseoximetry; Tho, thorax.

Ch17.qxd 1/25/05 3:24 AM Page 198

OSAS from normal controls.16 However, subse-quent application of this questionnaire to a pop-ulation of snoring children demonstratedwide-ranging positive predictive values (48.3%to 76.9%) and negative predictive values (26.9%

to 83.3%).16-18 Other questionnaires were alsounsuccessful19,20 or were not validated withpolysomnography.21,22 Although children withOSAS are statistically more likely to have reportedsymptoms (e.g., witnessed apnea, cyanosis, orlabored breathing), the sensitivity and specificityof such observations are low. Thus, the clinicalhistory alone is insufficient to diagnose OSAS in apopulation of snoring children.12

Physical examination of children with OSASis most often normal, with the exception of ade-notonsillar hypertrophy or craniofacial abnor-malities (Table 17–2). The majority of OSASchildren are of normal height and weight,although both obesity and failure to thrive mayoccur. Cardiovascular sequelae of SDB such ascor pulmonale and congestive heart failure areinfrequently observed in current clinical prac-tice, as heightened awareness has facilitated ear-lier diagnosis. Although blood pressure isstatistically elevated in children with OSAS, thewide range of normal makes this measurementa poor screening tool.23 The neurocognitiveconsequences of OSAS, such as poor school per-formance,24 aggressive behavior, and hyperactiv-ity,2,25 are nonspecific.

IMAGING AND LABORATORYEVALUATION

The diagnosis of SDB is firmly established usingpolysomnography, and ancillary testing is rarelyindicated. Further screening may be useful tofacilitate perioperative care and to excludeunderlying conditions (Table 17–3). For exam-ple, concern regarding right ventricular dysfunc-

Diagnosis of Obstructive Sleep Apnea Syndrome in Infants and Children Chapter 17 199

Sleep WakefulnessSnoring Poor school performanceWitnessed apnea Aggressive behaviorChoking noises HyperactivityIncreased work of breathing Attention deficit disorderParadoxical breathing Excessive daytime sleepinessEnuresis Morning headachesRestless sleepDiaphoresisHyperextended neckFrequent awakeningsDry mouth

Table 17–1. Clinical History of Obstructive Sleep Apnea Syndrome

General

SleepinessObesityFailure to thrive

Head

Swollen mucous membranesDeviated septumAdenoidal facies

Infraorbital darkeningElongated faceMouth breathing

Tonsillar hypertrophyHigh arched palateOverbiteCrowded oropharynxMacroglossiaGlossoptosisMidfacial hypoplasiaMicrognathia/retrognathia

Cardiovascular

HypertensionLoud P2 (heart sound)

Extremities

EdemaClubbing (rare)

Table 17–2. Physical Examination inObstructive Sleep ApneaSyndrome

Ch17.qxd 1/25/05 3:24 AM Page 199

tion may necessitate an electrocardiogram or anechocardiogram. Occasionally, chronic intermit-tent hypoxemia may induce polycythemia, andpersistent hypercarbia can elevate serum bicar-bonate. Routine pulmonary function testing isnot indicated in children suspected of havingOSAS unless restrictive or obstructive lung dis-ease is suspected. A fluttering pattern in the flowvolume loop has been described in adults withOSAS.26 Magnetic resonance imaging (MRI) ofthe upper airway in children with OSAS com-pared to unaffected controls reveals a statisticallysmaller upper airway luminal volume andelongation of the soft palate, as well as enlargedtonsils and adenoids.27 However, there is consid-erable overlap between the groups, renderingMRI a poor screening tool. The measurement ofthe ratio of the width of the tonsil to the depthof the pharyngeal space on a lateral neck radi-ograph was reported to have good sensitivity andspecificity for distinguishing mild from moder-ate/severe OSAS in a small number of patients.28

Dynamic fluoroscopy under sedation maydemonstrate glossoptosis, particularly in chil-dren with macroglossia, micrognathia, or neuro-muscular weakness.29 Anatomic localization ofthe site of obstruction may alter the therapeuticapproach in some children with craniofacialanomalies. Neck radiographs may be suggestiveof adenoidal hypertrophy, but direct or indirectvisualization of the adenoids remains the diag-nostic standard.

Video/Audio RecordingsDiagnosing OSAS using home audio recordings,in addition to a standard clinical history andphysical examination, revealed a sensitivity of92% but a specificity of only 29%.30 Subtle formsof SDB are particularly difficult to evaluate usingthis technique. However, computer-aided pro-cessing of audio signals for regularity mayimprove predictive value.31 Frequency domainanalysis of the snoring signal has also shownpromise in distinguishing OSAS from primarysnoring.32 Video recordings yield a noninvasivemeasure of movement and therefore arousal.33-35

Video is also a useful adjunct to a comprehensivepolysomnogram to evaluate body and head posi-tioning, paradoxical breathing, snoring, andmouth breathing. Studies correlating video scor-ing systems to standard polysomnography havebeen encouraging.34 Future research will be nec-essary to validate the utility of a particular domi-ciliary video or audio study in a population witha well-characterized symptomatology.

Overnight PolysomnographyPolysomnography is the gold standard for estab-lishing the presence and severity of SDB in chil-dren, and it can be performed in children of allages. An expert consensus panel has recom-mended that overnight polysomnography is thediagnostic test of choice in evaluating children


Serum Markers

HematocritSerum bicarbonate

Imaging

Brain magnetic resonance imaging (MRI)Anteroposterior and lateral neck radiographUpper airway computer tomography/MRI (complex craniofacial issues)Dynamic fluoroscopy

Sleep Monitoring

Multiple sleep latency test

Miscellaneous

EchocardiogramNeurocognitive testingElectrocardiogramFlow-volume loop

Table 17–3. Ancillary Diagnostic Studies in Obstructive Sleep Apnea Syndrome

Ch17.qxd 1/25/05 3:24 AM Page 200

with suspected SBD.12 Guidelines for performinglaboratory-based polysomnography in childrenhave been established.36 The sleep laboratoryshould be a nonthreatening environment thatcomfortably accommodates a parent during thestudy. Personnel with pediatric training shouldrecord, score, and interpret the study. The use ofsedatives37 and sleep deprivation38 may worsenSBD and are therefore not recommended. To theextent possible, sleep studies should conform tothe child’s usual sleep period. Infants may reason-ably be studied during the day, whereas adolescentstudies should generally start later at night. Thepolysomnographic montage will vary with thepatient’s age and suspected disorder (Table 17–4).

ElectroencephalogramConsensus guidelines for analyzing sleep architec-ture have been established in infants39 andadults.40 Standard practice is to apply adult EEGcriteria to children older than 6 months. Sleepstaging establishes that an adequate amount oftotal sleep time and sufficient rapid eye movement(REM) sleep were obtained on the night of thestudy. In addition to sleep staging, the EEG trac-ing is useful for scoring cortical arousals anddetecting epileptiform discharges. By consensus,an electrocortical (EEG) arousal is defined inadults as an abrupt 3-second shift in EEG fre-quency.41 This criterion appears to be appropriatefor use in children as well.42 However, visible EEGarousals are present in only 51% of obstructiveevents in children,43 complicating the diagnosis ofUARS in children. Frequency domain analysis of

the EEG tracing may further enhance the sensitiv-ity of detecting respiratory events or arousal.44,45

Initial reports indicated that children with snor-ing46 or even severe OSAS may have normal sleepstate distribution.47 However, a large cohort (N =559) comparing unaffected children to those withOSAS revealed that the latter have increases inslow-wave sleep (23.5% versus 28.8% of totalsleep time) and decreases in REM sleep (22.3%versus 17.3% of total sleep time).48 Furthermore,these authors observed a decline in the sponta-neous arousal index in patients with OSAS com-pared with controls (8.4 versus 5.3), suggesting ahomeostatic elevation in the arousal threshold.48

ArousalArousal from sleep is a protective reflex mecha-nism that restores airway patency through dila-tor muscle activation. Both mechanoreceptorsand chemoreceptors have a role in initiatingthe arousal response. Although arousals reversethe airway obstruction, they result in the unto-ward consequences of sleep fragmentation andsympathetic activation.49 Polysomnographically,arousal may be associated with EEG changes,41,45

increased airflow, elimination of airflow limita-tion, cessation of paradoxical breathing, tachy-cardia, movement,35 blood pressure elevation,50

and autonomic activation. In children, however,approximately 50% of obstructive events do notresult in an EEG arousal.43 In infants, EEGarousals are even less common.43 Thus, the EEGarousal index is unreliable for the diagnosis ofUARS. Frequency domain analysis may revealevidence of EEG arousal not readily visible, andit may be a clinically useful tool in the future.44,45

Obstructive events that terminate with auto-nomic activation are termed subcorticalarousals. Autonomic measures include heart ratevariability,51 blood pressure elevations,50 pulsetransit time,4 and peripheral arterial tonome-try.52 The pulse transit time was reported to be amore sensitive measure of respiratory arousalthan the 3-second EEG arousal.4 Subcorticalarousals alone have been demonstrated to resultin neurocognitive impairment in adults.53

Measures of Respiratory EffortA variety of methods to categorize central andobstructive respiratory events are amenable to


Electroencephalography (C3/A2, O1/A2)Electrooculogram (right/left)ElectrocardiogramAbdominal and thoracic excursionOximetry (3-sec averaging value), waveformEnd-tidal PCO2 (peak value, waveform)Flow: nasal pressure, oronasal thermistorSnore volumeBody position sensorpH (in special circumstances)Esophageal pressure (in special

circumstances)Video/audio taping

Table 17–4. Example of aPolysomnogram Montage

Ch17.qxd 1/25/05 3:24 AM Page 201

overnight polysomnography. Measures of tho-racic and abdominal excursion include piezo-electric belts and respiratory inductiveplethysmography (RIP). The classification ofcentral and obstructive apneas is achieved bydetermining if respiratory efforts are presentduring intervals of reduced flow. UncalibratedRIP tracings may also be used as an index ofthoracoabdominal asynchrony. The highly com-pliant chest wall in children results in asynchro-nous motion between the thoracic andabdominal tracings, termed paradoxical breath-ing. This disparity may be quantified usingphase angle analysis that is independent of therelative contributions of the two compart-ments.54 Thoracoabdominal asynchrony hasbeen demonstrated in children with increasedrespiratory effort due to upper airway obstruc-tion and OSAS.54,55 Paradoxical breathing is nor-mally seen in infants because of the highcompliance of their chest wall, particularly dur-ing REM sleep, but it is rare after 3 years of age.56

Esophageal manometry (Pes), the gold stan-dard for quantifying respiratory effort, permitsthe detection of subtle, partially obstructiveevents that may produce sleep fragmentation.57

However, Pes monitoring is uncomfortable andmay itself alter the frequency of respiratoryevents.58 The introduction of noninvasive,nasal pressure measurements has largely sup-planted Pes for establishing the diagnosis ofUARS (see Fig. 17–2). Our current practice isto reserve Pes monitoring for children withdiagnostic uncertainty even after standardovernight polysomnography.

Measures of AirflowAirflow can be quantitatively measured using anoronasal mask and a pneumotachograph.However, mask breathing is uncomfortable andhas been shown to alter respiratory mechanics.Practically speaking, this methodology isrestricted to research settings. Thermistors pro-vide qualitative measures of oronasal airflow bymeasuring the temperature of expired air.Although thermistors accurately indicatecomplete cessation of flow, they are not an accu-rate measure of tidal volume or therefore ofhypopnea.59 As a result, many laboratories haveadopted a definition of hypopnea that requiresan arterial oxygen desaturation of 3% and/or an

arousal in addition to a reduction in the ther-mistor signal. Another drawback of the thermis-tor is its long time constant that obscures thenuance of the flow profile, making the evalua-tion of flow limitation impossible.59

Nasal cannula pressure recordings provide aminimally invasive, semiquantitative measure ofairflow.60,61 The resulting signal has been shownto be proportional to flow squared. Because nasalpressure measurements have a fast time constant,it is possible to detect flattening of the inspiratorynasal pressure signal, termed flow limitation,which occurs in a collapsible tube when flowbecomes independent of driving pressure(Fig. 17–3). Limitations of the nasal cannulapressure recordings include obstruction of thetubing with secretions, mouth breathing (espe-cially in children with adenoidal hypertrophy),and the possible increase in nasal resistancecaused by obstruction of the nares. In children,the reported signal quality has varied in labora-tory-based studies. Trang and coworkers62

reported an uninterpretable nasal cannula signalduring only 4% of total sleep time. However, 17%of subjects had uninterpretable signals for morethan 20% of total sleep time. In contrast,Serebrisky and colleagues63 reported adequatenasal cannula flow signals (in greater than 50% oftotal sleep time) during sleep in only 71.8% ofpatients. In a domiciliary study, the nasal cannulachannel was not available, overall, for more than50% of the night.64

Airflow may be approximated using RIP byconsidering that lung volume can be approxi-mated by a two-compartment model (thoracicand abdominal excursion). RIP may be cali-brated using an isovolume maneuver or a statis-tical technique and used to derive a sumchannel proportional to tidal volume.65,66 Thus,the time derivative of the sum channel is pro-portional to flow. The RIP signal may thereforebe analyzed for apnea, hypopnea, and flow lim-itation. Our current practice is to combine RIP,nasal cannula pressure, and oronasal thermistorin our laboratory-based studies to ensure that aflow signal is likely to be available throughoutthe night.

Gas ExchangePulse oximetry, which is based on the absorptionspectra of hemoglobin, is the standard polysomno-


Ch17.qxd 1/25/05 3:24 AM Page 202

graphic measure of hypoxemia. The relationshipbetween the arterial oxygen tension (PaO2) andthe oxygen saturation (SpO2), the oxyhemoglo-bin curve, is sigmoidal. Patients with normalpulmonary function, whose baseline SpO2 is onthe flat portion of the curve, require largechanges in PaO2 to affect their SpO2. In contrast,patients with parenchymal lung disease may beoperating on the steep portion of the oxyhemo-globin curve, thereby experiencing a profounddecline in SpO2 with small decrements in PaO2.In sickle cell disease, the oxyhemoglobin curveis variably shifted to the right, thus limiting theutility of SpO2 measurements.67

The adequacy of ventilation can be assessednoninvasively during sleep by sampling CO2tension in respired air, termed capnography.Expired gas initially consists of dead space ven-tilation in the measuring apparatus as well as thephysiologic dead space. The terminal expiredconcentration of CO2 reaches a plateau andreflects alveolar gas. The concentration of CO2in a given alveolus is predicated on the ventila-tion–perfusion relationship. In subjects withnormal lung mechanics, end-tidal PaCO2 (P

ET

CO2) is approximately 2 to 5 mm Hg below the

arterial PaCO2 level.68 However, in diseases withuneven ventilation–perfusion or altered alveolartime constants, such as cystic fibrosis, the P

ETCO2

is not an accurate measure of PaCO2. Also, therapid respiratory rates characteristic of infantsmay not permit the alveolar CO2 to plateau andtherefore may underestimate the true CO2value.69 In OSAS, the shape of the capnogramwaveform qualitatively reflects flow, whereas theplateau level indicates the overall adequacy ofventilation. Normative data for P

ETCO2 during

sleep are presented later in Table 17–5. The cap-nometry signal is prone to artifact because ofnasal secretions that obstruct the sampling can-nula. Morielli and colleagues69 reported that27% of polysomnographic epochs in their labo-ratory had a poor P

ETCO2 waveform, stressing the

importance of careful attention to technique.Transcutaneous CO2 (TcCO2) monitors may bevaluable in patients in whom P

ETCO2 is not accu-

rate, such as infants or patients with parenchy-mal lung disease.69 Although absolute values ofTcCO2 are variable, the trend is proportional tothe PaCO2, with a lag of several minutes. Sleep isnormally associated with a 4- to 6-mm Hgincrease in TcCO2.


Figure 17–3. Nasal cannula pressure tracing demonstrating that flow becomes independent of driv-ing pressure during flow-limited breaths. (Reproduced with permission from Hosselet J, Norman RG,Ayappa I, Rapoport DM: Detection of flow limitation with a nasal cannula/pressure transducer system.Am J Respir Crit Care Med 1998;157:1461- 1467.)

Ch17.qxd 1/25/05 3:24 AM Page 203

Normative Polysomnographic DataNormative data for the EEG and respiratoryparameters of sleeping children are shown inTable 17–5.47,48,70-72 In infants, longitudinal arte-rial oxygen saturation monitoring revealed amedian SpO2 baseline of 98%, but the SpO2 wasless than 90% during 0.51% of epochs.73 Marcusand coworkers72 presented the respiratory dataof 50 normal, nonobese children between 1 and18 years of age (mean, 9.7 ± 4.6), using a ther-mistor and P

ETCO2 monitoring. Nine subjects had

at least one obstructive apneic event. One “nor-mal” child had an apnea index of 3.1/hour,although he had a sibling with OSAS. Includingthis outlier, an apnea index of greater than1/hour was determined to be statistically abnor-mal. However, the threshold level at which theapnea index becomes clinically significant hasnot been established. Only 15% of unaffectedchildren had any hypopneas, with the meanhypopnea index being 0.1 ± 0.1 (range, 0.1 to0.7).74 Acebo and colleagues71 also reported thathypopneas were rare in older children and ado-lescents. Normative data using nasal pressureduring sleep have not been presented in chil-dren. Experience from our laboratory indicatesthat hypopneas and respiratory effort–relatedarousals do not occur in normal young children.

Normative data for esophageal manometry(ΔPes) from 10 normal subjects in our laboratory,

aged 2 to 11 years, showed that control subjectshad a mean ΔPes of −8 ± 2 cm H2O (range, − 6 to−12), a peak ΔPes of −12 ± 3 cm H2O (range − 9to −19 cm H2O), and a ΔPes of less than or equalto −10 cm H2O for 8% ± 12% of breaths (range,3% to 61%).4 Other investigators have consid-ered ΔPes swings of −8 to −14 cm H2O as nor-mal75 and have suggested that normal childrenspend 10% or less of the night with inspiratoryesophageal pressure swings of −10 cm H2O orless.57 However, our data showed that two con-trol patients spent 21% and 61% of the nightwith a ΔPes of −10 cm H2O or less. Thus, nor-mative data from nonsnoring controls revealconsiderable variability in respiratory effort. TheΔPes is lower in normal infants (5 to 6 cmH2O).76

Domiciliary and Nap StudiesAttended overnight polysomnography in a des-ignated pediatric sleep laboratory is the goldstandard for the diagnosis of SDB children.36

However, such comprehensive testing is expen-sive, labor intensive, and not widely available.Recognizing that 10% to 12% of children snorebut that only 1% to 3% have OSAS8,9 hasprompted considerable interest in limited, nap,or unattended domiciliary studies. Choosing themost appropriate evaluation is predicated onunderstanding the relative risks of OSAS and its


Sleep Parameters Respiratory ParametersEEG arousal index, n/hr 7 ± 2 Obstructive apnea

index, n/hr TST 0.0 ± 0.1Sleep efficiency, % 84 ± 13 Obstructive hypopnea

index, n/hr TST 0.1 ± 0.1Stage 1, % TST 5 ± 3 Central apneas with

desaturation, n/hr TST 0.0 ± 0.1Stage 2, % TST 51 ± 9 Duration of

hypoventilation (PETCO2 ≥ 45 mm Hg), % TST 1.6 ± 0.8

Slow-wave sleep, % TST 26 ± 8 Peak PETCO2, mm Hg 46 ± 3REM sleep, % TST 19 ± 6 SpO2 nadir, % 95 ± 1REM cycles, n 4 ± 1

Data presented as mean ± 2 SD.EEG, electroencephalography; PETCO2, end-tidal arterial carbon dioxide partial pressure; REM, rapid eye movement;

SpO2, oxygen saturation.See references 47, 48, 70, and 72.

Table 17–5. Polysomnographic Data in Unaffected Children

Ch17.qxd 1/25/05 3:24 AM Page 204

treatments. Testing is indicated to determine thepresence and severity of SDB, the type of treat-ment, the preoperative care, and necessary fol-low-up.

Polysomnography during a daytime nap mayunderestimate the degree of SDB.77,78 REM sleepoften does not occur during naps. Furthermore,obstructive respiratory events worsen as sleepprogresses.47 Thus, although the positive pre-dictive value of an abnormal nap study is 100%,the negative predictive value is only 20%.77,78

Home study devices range in complexity fromoximetry alone79,80 to multichannel record-ings.81 Ambulatory studies may be used to eval-uate oxygenation in infants with chronic lungdisease, cystic fibrosis, restrictive lung diseases,and neuromuscular conditions. Jacob and col-leagues82 compared laboratory and domiciliarystudies in a population of symptomatic childrenwith adenotonsillar hypertrophy. Their multi-channel monitor included oxygen saturation,waveform, electrocardiogram, RIP, and videorecording, and therefore it cannot be comparedto commercially available ambulatory systems.82

Similar indices of apnea/hypopnea, desaturation,and arousal were observed during attended labo-ratory-based polysomnograms and domiciliarystudies.82

Documenting oximeters should include analgorithm for artifact reduction such as a plethys-mograph waveform or heart rate detected by theoximeter. The oximetry channel alone has beencompared to full polysomnography in a popula-tion of snoring children suspected to have SDBwith adenotonsillar hypertrophy.83 Oximetrydemonstrated an excellent positive predictivevalue (97%) but a poor negative predictive ability(53%).83 Many children with SDB and clinicallysignificant respiratory–effort related arousals donot demonstrate oxygen desaturation.1,57 Similarobservations have been made regarding the limi-tations of using oximetry as a screening tool foradult OSAS.79,80 Oximetry is inadequate to diag-nose a considerable percentage of pediatric SDBin which arousal and hypoventilation, ratherthan hypoxemia, predominate.84

Night-to-Night PolysomnographicVariabilityA single overnight polysomnogram is a well-rec-ognized diagnostic tool for determining the pres-

ence and severity of OSAS in symptomatic chil-dren.36 However, few data exist regarding thevariability of polysomnographic findings in chil-dren. That is, it is unknown whether the fre-quency and severity of sleep apnea syndromeamong snoring children is sufficiently consistentfrom night to night to be accurately assessed byone polysomnogram. The misclassification ofOSAS as primary snoring after a single night ofpolysomnography has been observed inadults,85,86 and it may also occur in children. Anadaptation effect of the sleep laboratory environ-ment, termed the first-night effect, has beenreported to disrupt sleep architecture87-90 andperhaps to underestimate respiratory distur-bance.88,91

In children, the clinical diagnosis of eitherOSAS or primary snoring remained the same intwo polysomnographic studies 1 to 4 weeksapart.92 Although the overall classification ofsubjects was unchanged, there were minimalchanges in OSAS severity between nights. Thissupports the view that a single polysomno-graphic night is sufficient for the diagnosis ofOSAS in otherwise normal, snoring childrenwith adenotonsillar hypertrophy. No night-to-night systematic bias was observed in anyof the mean intersubject respiratory parame-ters.90,92 The intrasubject respiratory param-eters, however, demonstrated considerablevariability, particularly in children with severedisease. The variability in respiratory parame-ters could not be accounted for by changes inbody position or percent REM time. Among thesleep variables, a first-night effect, includingincreased wakefulness and a reduction of REMsleep, was observed as the result of adaptationto the sleep laboratory environment.90

Circumstances in which a single study nightmay not be sufficient include inadequate REMtime, technical limitations in acquiring keychannels, or when the parents report that theparticular study night did not reflect a typicalnight’s sleep.

DIAGNOSTIC AND EVENTCLASSIFICATION

The optimal definition for respiratory events andclinical classification has not been established inchildren. There are no clinical studies evaluatingthe relative merit of specific event definitions


Ch17.qxd 1/25/05 3:24 AM Page 205

in relation to clinical outcomes.93 The poly-somnographic criteria for event scoring andclinical diagnosis, based on our experience, aresummarized in Tables 17–6 and 17–7. However,well-designed outcome studies are desperatelyneeded to validate these criteria. Additionalabnormal breathing patterns, including tachyp-nea and increased respiratory effort, have beendescribed in children with OSAS.94 Normativedata for respiratory effort–related arousals and

flow limitation are scant in children, and theseevents appear to be rare. Nevertheless, the clini-cal classification of symptomatic children cannotbe exclusively based on the apnea index alone.Consideration of the apnea-hypopnea index,flow limitation, SpO2, PET

CO2, and arousal indicesalso contribute to the diagnosis of OSAS. Thus,diagnostic interpretation of pediatric polysomnog-raphy will continue to require consideration ofall respiratory parameters.


Obstructive

Apnea Absence of oronasal airflow for any duration, with persistent respiratory effort

Hypopnea Thermistor: Reduction of oronasal flow >50% for a 2-breath duration or longer, with persistent respiratory effort accompanied by a desaturation of 3% or greater, or by an arousal

Nasal pressure: Discernible change in flow shape or amplitude for a 2-breath duration or longer, with persistent respiratory effort accompanied by a desaturation of 3% or greater, or by an arousal

Respiratory effort–related arousal Evidence of increased respiratory effort or flow limitation leading to an arousal, followed by normalization of effort and flow

Flow limitation Flattening of the inspiratory limb of the nasal pressure channel

Snoring Coarse, low-pitched inspiratory soundHypoventilation PETCO2 > 50 mm Hg for >10% TST, or PETCO2

peak > 53 mm HgAccompanied by paradoxical breathing or

obstructive events

Central

Apnea Absence of oronasal airflow for 20 sec or longer without respiratory effort

Shorter events are scored if associated with arousal, desaturation, or bradycardia.

Hypoventilation PETCO2 > 50 mm Hg for >10% TST, or PETCO2peak > 53 mm Hg

Accompanied by decreased respiratory effortPeriodic breathing Succession of 3 or more central apneas of 3-sec duration,

separated by < 20 sec of normal breathing

Miscellaneous

Mixed apneas Cessation of flow with a central and obstructive component

Rebreathing Evidence of increased CO2 in inspired air

PETCO2, end-tidal arterial carbon dioxide partial pressure; TST, total sleep time.

Table 17–6. Respiratory Pattern Scoring

Ch17.qxd 1/25/05 3:24 AM Page 206


PETCO2 > 50Apnea Index SpO2 Nadir PETCO2 Peak torr Arousals

Diagnosis* (Events/hr) (%) (torr) (% TST) (Events/hr)Primary snoring ≤1 >92 ≤53 <10 EEG < 11Upper airway

resistance ≤1r >92 ≤53 <10 RERA > 1syndrome EEG > 11

Mild OSAS 1-4 86-91 >53 10- 24 EEG > 11Moderate OSAS 5-10 76-85 >60 25-49 EEG > 11Severe OSAS >10 ≤75 >65 ≥50 EEG > 11

*Each diagnosis requires one or more of the measures to its right.EEG, electroencephalographic arousal; OSAS, obstructive sleep apnea syndrome; PETCO2, end tidal PCO2; RERA, res-

piratory-related arousal; Spo2, arterial oxygen saturation; TST, total sleep time.

Table 17–7. Diagnostic Classification of Sleep-Disordered Breathing

References

1. Guilleminault C, Stoohs R, Clerk A, et al: A causeof excessive daytime sleepiness: The upper airwayresistance syndrome. Chest 1993;104:781-787.

2. Guilleminault C, Korobkin R, Winkel R: A reviewof 50 children with obstructive sleep apnea syn-drome. Lung 1981;159:275-287.

3. Rosen C, D’Andrea L, Haddad G: Adult criteriafor obstructive sleep apnea do not identify chil-dren with serious obstruction. Am Rev Respir Dis1992;146:1231-1234.

4. Katz ES, Lutz J, Black C, Marcus CL: Pulse transittime as a measure of arousal and respiratory effortin children with sleep-disordered breathing.Pediatr Res 2003;53:580-588.

5. American Sleep Disorders Association: The inter-national classification of sleep disorders.Diagnostic and coding manual (revised edition).Rochester, Minn, American Sleep DisordersAssociation, 1997.

6. Urschitz MS, Guenther A, Eggebrecht E, et al:Snoring, intermittent hypoxia and academic per-formance in primary school children. Am J RespirCrit Care Med 2003;168:464-468.

7. O’Brien LM, Mervis CB, Holbrook CR, et al:Neurobehavioral implications of habitual snoringin children. Pediatrics 2004;114:44-49.

8. Ali NJ, Pitson DJ, Stradling JR: Snoring, sleep dis-turbance and behaviour in 4-5 year olds. Arch DisChild 1993;68:360-366.

9. Gislason T, Benediktsdottir B: Snoring, apneicepisodes, and nocturnal hypoxemia amongchildren 6 months to 6 years old. Chest1995;107:963-966.

10. Redline S, Tishler PV, Schluchter M, et al: Riskfactors for sleep-disordered breathing in children:Associations with obesity, race, and respiratory

problems. Am J Respir Crit Care Med1999;159:1527-1532.

11. Wilson K, Lakheeram I, Morielli A, et al: Canassessment for obstructive sleep apnea help pre-dict postadenotonsillectomy respiratory compli-cations? Anesthesiology 2002;96:313-322.

12. American Academy of Pediatrics, Section onPediatric Pulmonology, Subcommittee onObstructive Sleep Apnea: Clinical practice guide-line: Diagnosis and management of childhoodobstructive sleep apnea syndrome. Pediatrics2002;109:704-712.

13. Preutthipan A, Chantarojanasiri T, Suwanjutha 5,Udomsubpayakul U: Can parents predict theseverity of childhood obstructive sleep apnoea?Acta Paediatr 2000;89:708-712.

14. Gozal D, Wang M, Pope DW: Objective sleepinessmeasures in pediatric obstructive sleep apnea.Pediatrics 2001;108:693-697.

15. Melendres MCS, Lutz J, Rubin ED, Marcus CL:Daytime sleepiness and hyperactivity in childrenwith suspected sleep-disordered breathing.Pediatrics 2004;114:768-775.

16. Brouillette R, Hanson D, David R, et al: A diag-nostic approach to suspected obstructive sleepapnea in children. J Pediatr 1984;105:10-14.

17. Carroll J, McColley S, Marcus C, et al: Inability ofclinical history to distinguish primary snoringfrom obstructive sleep apnea syndrome in chil-dren. Chest 1995;108:610-618.

18. Rosen CL: Clinical features of obstructive sleepapnea hypoventilation syndrome in otherwisehealthy children. Pediatr Pulmonol 1999;27:403-409.

19. Croft CB, Brockbank MJ, Wright A, Swanston AR:Obstructive sleep apnoea in children undergoingroutine tonsillectomy and adenoidectomy. ClinOtolaryngol 1990;15:307-314.

Ch17.qxd 1/25/05 3:24 AM Page 207


20. Suen JS, Arnold JE, Brooks U: Adenotonsillectomyfor treatment of obstructive sleep apnea in chil-dren. Arch Otolaryngol Head Neck Surg1995;121:525-530.

21. Van Someren VH, Hibbert J, Stothers JK, et al:Identification of hypoxaemia in children havingtonsillectomy and adenoidectomy. Clin Oto-laryngol 1990;15:263-271.

22. Van Someren VH, Burmester M, Alusi G, Lane R:Are sleep studies worth doing? Arch Dis Child2000;83:76-81.


24. Gozal D: Sleep-disordered breathing and school per-formance in children. Pediatrics 1998;102: 616-620.

25. Guillemmault C, Winkel R, Korobkin R,Simmons B: Children and nocturnal snoring:Evaluation of the effects of sleep related respira-tory resistive load and daytime functioning. EurRespir J 1982;139:165-171.

26. Haponik EF, Smith PL, Kaplan J, Bleeker ER:Flow-volume curves and sleep-disordered breath-ing: Therapeutic implications. Thorax 1983;38:609-615.

27. Arens R, McDonough JM, Costarino AT, et al:Magnetic resonance imaging of the upper airwaystructure of children with obstructive sleep apneasyndrome. Am J Respir Crit Care Med 2001;164:698-703.

28. Li AM, Wong E, Kew J, et al: Use of tonsil size inthe evaluation of obstructive sleep apnoea. ArchDis Child 2002;87:156-159.

29. Donnelly LF, Strife JL, Myer CM: Glossoptosis(posterior displacement of the tongue) duringsleep: A frequent cause of sleep apnea in pediatricpatients referred for dynamic sleep fluoroscopy.AJR Am J Roentgenol 2000;175:1557-1559.

30. Goldstein NA, Sculerati N, Waisleben JA, et al:Clinical diagnosis of pediatric obstructive sleepapnea validated by polysomnography. OtolaryngolHead Neck Surg 1994;111:611-617.

31. Potsic WP: Comparison of polysomnography andsonography for assessing regularity of respirationduring sleep in adenotonsillar hypertrophy.Laryngoscope 1987;97:1430-1437.

32. McCombe AW, Kwok VK, Hawke WM: Anacoustic screening test for obstructive sleepapnoea. Clin Otolaryngol 1995;20:348-351.

33. Morielli A, Laden S, Ducharme FM, BrouilletteRT: Can sleep and wakefulness be distinguishedin children by cardiorespiratory and videotaperecordings? Chest 1996;109:680-687.

34. Sivan Y, Kornecki A, Schonfeld T: Screeningobstructive sleep apnoea syndrome by homevideotape recording in children. Eur Respir J1996;9:2127-2131.

35. Mograss M, Ducharme F, Brouillette R:Movement/arousals. Description, classificationand relationship to sleep apnea in children. Am JRespir Crit Care Med 1994;150:1690-1696.

36. American Thoracic Society: Standards and indica-tions for cardiopulmonary sleep studies inchildren. Am J Respir Crit Care Med 1996;153:866-878.

37. Hershenson M, Brouillette RT, Olsen E, Hunt CE:The effect of chloral hydrate on genioglossus anddiaphragmatic activity. Pediatr Res 1984;18:516-519.

38. Canet E, Gaultier C, D’Allest AM, Dehan M:Effects of sleep deprivation on respiratory eventsduring sleep in healthy infants. J Appl Phys1989;66:1158-1163.

39. Anders T, Ende R, Parmelee A: A manual of stan-dardized terminology, technique, and criteria forscoring of states of sleep and wakefulness in new-born infants. Los Angeles, UCLA BrainInformation Service, 1971.

40. Rechtschaffen A, Kales A: A manual of standard-ized terminology: Techniques and scoring sys-tems for sleep stages of human subjects. LosAngeles, UCLA Brain Information Service/BrainResearch Institute, 1968.

41. Sleep Disorders Atlas Task Force of the AmericanSleep Disorders Association: EEG arousals:Scoring rules and examples—A preliminary report.Sleep 1992;15:173-184.

42. Wong TK, Gaister P, Lau TS, et al: Reliability ofscoring arousals in normal children and childrenwith obstructive sleep apnea syndrome. Sleep2004; 27:1139–1145.

43. McNamara F, Issa F, Sullivan C: Arousal patternfollowing central and obstructive breathingabnormalities in infants and children. J ApplPhysiol 1996;81:2651-2657.

44. Bandla HP, Gozal D: Dynamic changes in EEGspectra during obstructive apnea in children.Pediatr Pulmonol 2000;29:359-365.

45. Black JE, Guilleminault C, Colrain IM, Carrillo O:Upper airway resistance syndrome: Central elec-troencephalographic power and changes inbreathing effort. Am J Respir Crit Care Med2000;162:406-411.

46. Fuentes-Pradera MA, Botebol G, Sanchez-Armengol A, et al: Effect of snoring and obstruc-tive respiratory events on sleep architecture inadolescents. Arch Pediatr Adolesc Med 2003;157:649-654.

47. Goh DY, Galster P, Marcus CL: Sleep architectureand respiratory disturbances in children withobstructive sleep apnea. Am J Respir Crit CareMed 2000;162:682-686.

48. Tauman R, O’Brien LM, Holbrook CR, Gozal D:Sleep pressure score: A new index of sleep dis-

Ch17.qxd 1/25/05 3:24 AM Page 208

ruption in snoring children. Sleep 2004;27:274-278.

49. Somers VK, Dyken ME, Clary MP, Abboud FM:Sympathetic neural mechanisms in obstructivesleep apnea. J Clin Invest 1995;96:1897-1904.

50. Davies RJO, Vardi-Visi K, Clarke M, Stradling JR:Identification of sleep disruption and sleep disor-dered breathing from the systolic blood pressureprofile. Thorax 1993;48:1242-1247.

51. Aljadeff G, Gozal D, Schechtman VL, et al: Heartrate variability in children with obstructive sleepapnea. Sleep 1997;20:151-157.

52. Schnall RP, Shlitner A, Sheffy J, et al: Periodic,profound peripheral vasoconstriction: A newmarker of obstructive sleep apnea. Sleep1999;22:939-946.

53. Martin SE, Wraith PK, Deary IJ, Douglas NJ: Theeffect of nonvisible sleep fragmentation on day-time function. Am J Respir Crit Care Med1997;155:1596-1601.

54. Sivan Y, Ward SD, Deakers T, et al: Rib cage toabdominal asynchrony in children undergoingpolygraphic sleep studies. Pediatr Pulmonol1991;11:141-146.

55. Kohyama J, Shiiki T, Shimohira M, Hasegawa T:Asynchronous breathing during sleep. Arch DisChild 2001;84:174-177.

56. Gaultier C, Praud JP, Canet E, et al: Paradoxicalinward ribcage motion during rapid-eye move-ment sleep in infants and young children. J DevPhysiol 1987;9:391-397.


58. Groswasser J, Scaillon M, Rebuffat E, et al: Naso-oesophageal probes decrease the frequency ofsleep apnoeas in infants. J Sleep Res 2000;9:193-196.

59. Farre R, Montserrat JM, Rotger M, et al: Accuracy ofthermistors and thermocouples as flow-measuringdevices for detecting hypopneas. Eur Respir J1998;11:179-182.

60. Norman RG, Ahmed MM, Walsleben JA,Rapoport DM: Detection of respiratory eventsduring NPSG: Nasal cannula/pressure sensor ver-sus thermistor. Sleep 1997;20:1175-1184.

61. Hosselet J, Norman RG, Ayappa I, Rapoport DM:Detection of flow limitation with a nasal can-nula/pressure transducer system. Am J Respir CritCare Med 1998;157:1461-1467.

62. Trang H, Leske V, Gaultier C: Use of nasal cannulafor detecting sleep apneas and hypopneas ininfants and children. Am J Respir Crit Care Med2002;166:464-468.

63. Serebrisky D, Cordero R, Mandeli J, et al:Assessment of inspiratory flow limitation in chil-dren with sleep-disordered breathing by a nasal

cannula pressure transducer system. PediatrPulmonol 2002;33:380-387.

64. Poels PJP, Schilder AGM, van de Berg S, et al: Evalu-ation of a new device for home cardiorespiratoryrecording in children. Arch Otolaryngol HeadNeck Surg 2003;129:1281-1284.

65. Sackner MA, Watson H, Belsito AS, et al:Calibration of respiratory inductive plethysmo-graph during natural breathing. J Appl Phys1989;66:410-420.

66. Adams JA, Zabaleta IA, Stroh D, Sackner MA:Measurement of breath amplitudes: Comparisonof three noninvasive respiratory monitors to inte-grated pneumotachograph. Pediatr Pulmonol1993;16:254-258.

67. Seakins M, Gibbs WN, Milner PF, Bertles JF:Erythrocyte Hb-S concentration: An importantfactor in the oxygen affinity of blood in sickle cellanemia. J Clin Invest 1973;52:422-432.

68. Bhavani-Shankar K, Moseley H, Kumar AY, Deiph Y:Capnometry and anesthesia. Can J Anaesth 1992;39:617-632.

69. Morielli A, Desjardins D, Brouillette RT:Transcutaneous and end-tidal carbon dioxidepressures should be measured during pediatricpolysomnography. Am Rev Respir Dis 1993;148:1599-1604.

70. Uliel S, Tauman R, Greenfeld M, Sivan Y: Normalpolysomnographic respiratory values in childrenand adolescents. Chest 2004;125:872-878.

71. Acebo C, Millman RP, Rosenberg C, et al: Sleep,breathing, and cephalometrics in older childrenand young adults: Part 1. Normative values. Chest1996;109:664-672.

72. Marcus CL, Omlin KJ, Basinski DJ, et al: Normalpolysomnographic values for children and adoles-cents. Am Rev Respir Dis 1992;146:1235-1239.

73. Hunt CE, Corwin MJ, Lister G, et al: Longitudinalassessment of hemoglobin oxygen saturation inhealthy infants during the first 6 months of age.J Pediatr 1999;134:580-586.

74. Witmans MB, Keens TG, Ward SLD, Marcus CL:Obstructive hypopneas in children and adoles-cents: Normal values. Am J Respir Crit Care Med2003;168:1540.

75. Miyazaki S, Itasaka Y, Yamakawa K, et al:Respiratory disturbance during sleep due to ade-noid-tonsillar hypertrophy. Am J Otolaryngol1988;10:143-149.

76. Skatvedt O, Grogaard J: Infant sleeping positionand inspiratory pressures in the upper airways andoesophagus. Arch Dis Child 1994;71:138-140.

77. Marcus CL, Keens TG, Ward SL: Comparison ofnap and overnight polysomnography in children.Pediatr Pulmonol 1992;13:16-21.

78. Saeed MM, Keens TG, Stabile MW, et al: Shouldchildren with suspected obstructive sleep apnea


Ch17.qxd 1/25/05 3:24 AM Page 209

syndrome and normal nap studies have overnightsleep studies? Chest 2000;118:360-365.

79. Cooper BG, Veale D, Griffiths CJ, Gibson GJ:Value of nocturnal oxygen saturation as a screen-ing test for sleep apnoea. Thorax 1991;46:586-588.

80. Ryan PJ, Hilton MF, Boldy DAR, et al: Validationof British Thoracic Society guidelines for the diag-nosis of the sleep apnoea/hypopoea syndrome:Can polysomnography be avoided? Thorax1995;50:972-975.

81. Redline S, Tosteson T, Boucher MA, MillmanRP: Measurement of sleep-related breathingdisturbances in epidemiologic studies: Assess-ment of the validity and reproducibility of aportable monitoring device. Chest 1991;100:1281-1286.

82. Jacob SV, Morielli A, Mograss MA, et al: Hometesting for pediatric obstructive sleep apnea syn-drome secondary to adenotonsillar hypertrophy.Pediatr Pulmonol 1995;20:241-252.

83. Brouillette RT, Morielli A, Leimanis A, et al:Nocturnal pulse oximetry as an abbreviated test-ing modality for pediatric obstructive sleep apnea.Pediatrics 2000;105:405-412.

84. Kirk VG, Boim SG, Flemons WW, Remmers JE:Comparison of home oximetry monitoring withlaboratory polysomnography in children. Chest2003;124:1702-1708.

85. Mosko S, Dickel M, Ashurst J: Night-to-nightvariability in sleep apnea and sleep-related peri-odic leg movements in the elderly. Sleep1988;11:340-348.

86. Bliwise D, Benkert R, Ingham R: Factors associ-ated with nightly variability in sleep-disorderedbreathing in the elderly. Chest 1991;100:973-976.

87. Aber W, Block A, Hellard D, Webb W: Consistencyof respiratory measurements from night to nightduring the sleep of elderly men. Chest 1989;96:747-751.

88. Le Bon O, Hoffman G, Tecco J, et al: Mild to mod-erate sleep respiratory events: One negative nightmay not be enough. Chest 2000;118:353-359.

89. Agnew H, Webb W, Williams R: The first nighteffect: An EEG study of sleep. Psychophysiology1966;2:263-266.

90. Scholle S, Scholle HC, Kemper A, et al: First nighteffect in children and adolescents undergoingpolysomnography for sleep-disordered breathing.Clin Neurophys 2003;14:2138-2145.

91. Bliwise D, Carey B, Dement W: Nightly variationin sleep-related respiratory disturbance in olderadults. Exp Aging Res 1983;9:77-81.

92. Katz ES, Greene MG, Carson KA, et al: Night-to-night variability of polysomnography in childrenwith symptoms of sleep-disordered breathing.J Pediatr 2002;140:589-594.

93. Marcus CL, England S, Annett RD, et al:Cardiorespiratory sleep studies in children:Establishment of normative data and polysomno-graphic predictors of morbidity. Am J Respir CritCare Med 1999;160:1381-1387.

94. Guilleminault C, Li KK, Khramstov A, et al:Breathing patterns in prepubertal children withsleep-related breathing disorders. Arch PediatrAdolesc Med 2004;158:153-161.


Ch17.qxd 1/25/05 3:24 AM Page 210

Obstructive sleep apnea (OSA) is a frequent con-dition and is currently estimated to affectbetween 1% and 3% of 2- to 8-year-old chil-dren.1-3 OSA is characterized by repeated eventsof partial or complete upper airway obstructionduring sleep, resulting in disruption of normalventilation, hypoxemia, and fragmentation ofsleep (Fig. 18–1).4 However, habitual snoringduring sleep is a much more frequent occurrenceand affects up to 27% of children,2,3,5-9 with adecrease in frequency in 9 to 14 year olds.10

OSA was first described by McKenzie over acentury ago,11 yet it was not until the mid-1970sthat it was recognized in children.12 Usingpolysomnography in eight children aged 5 to 14years, Guilleminault and colleagues publishedthe first detailed report of children with adeno-tonsillar hypertrophy and OSA and suggestedthat surgery may eliminate their clinical symp-toms. Since this initial report there have beenmultiple publications on OSA in children, and itis clear that the classic OSA in children is a dis-order distinct from the OSA that occurs in adults,in particular with respect to gender distribution,clinical manifestations, polysomnographic find-ings, and treatment.13,14 OSA is frequently diag-nosed in association with adenotonsillarhypertrophy and is also common in children withcraniofacial abnormalities and neurologic disor-ders affecting upper airway patency.

The primary symptom of OSA is snoring.While snoring is not normal and essentially indi-cates the presence of heightened upper airwayresistance, many snoring children may have pri-mary snoring, that is, habitual snoring withoutalterations in sleep architecture, alveolar ventila-tion, and oxygenation. However, the acquisitionof definitive criteria allowing for a reliable dis-tinction between primary snoring and OSA willhave to wait several years as the consequences ofOSA in children become better recognized and

understood and the threshold at which mor-bidity occurs becomes known. Despite suchlimitations, it is clear that polysomnographicassessment is currently required for the defini-tive diagnosis of OSA in children, since a clinicalhistory and physical examination are insufficientto confirm its presence or severity.15 Never-theless, it should be emphasized that alternativescreening methods have been recently ex-plored16-19 and that novel technologic advancesmay allow for improved diagnostic accuracy inthe future.20,21

The implications of OSA in children are mul-tifaceted and potentially complex. If leftuntreated or, alternatively, if treated late, pedi-atric OSA may lead to substantial morbidity thataffects multiple target organs and systems andthat may not be completely reversed with appro-priate treatment. The potential consequences ofOSA in children include behavioral disturbancesand learning deficits,2,22-25 pulmonary hyperten-sion,26 and systemic hypertension27 as well ascompromised somatic growth.28

The pathophysiology of OSA in children hasbeen recently reviewed29 and is discussed indetail in the previous chapter. It is clear that chil-dren with OSA have increased upper airway col-lapsibility30 and that enlarged tonsils andadenoids are a major contributor to this disor-der.31 However, adenotonsillar hypertrophyalone is not sufficient to cause OSA becausesome children with “kissing tonsils” do not haveOSA, whereas others are not cured after adeno-tonsillectomy.32

As with any other disorder in medicine, mor-bidity and mortality are the major incentives tointervene and treat it. In subsequent sections ofthis chapter, we will critically review the variousconsequences of pediatric OSA. For the purposeof clarity, the spectrum of morbidities associatedwith OSA can be examined as representing four

Consequences of Obstructive SleepApnea Syndrome

Louise Margaret O’Brien

David Gozal

18

211

Ch18.qxd 1/25/05 3:25 AM Page 211

212 Chapter 18 Consequences of Obstructive Sleep Apnea Syndrome

EO

G

EE

G

TcC

O2

SA

O2

EC

G

EM

G

L EM

G

Sno

re

Che

st

Abd AF

OA

OS

A

Des

atur

atio

nD

esat

urat

ionAro

usal

Aro

usal

Des

atur

atio

nAro

usal

Figure 18–1. Graphic recording of the obstructive sleep apnea (OSA) triad, namely the cessation ofairflow in the presence of respiratory effort, oxyhemoglobin desaturation, and sleep arousal. AF, air-flow; Abd, abdominal excursion; Chest, chest wall excursion; ECG, electrocardiogram; EEG, electroen-cephalogram; EMG, chin electromyogram; EOG, electro-oculogram; LEMG, left anterior tibial EMG;OA, obstructive apnea; SaO2, oxyhemoglobin saturation; Snore, sound channel; TcCO2, transcutaneousPCO2.

Ch18.qxd 1/25/05 3:25 AM Page 212

immediate consequences of upper airwayobstruction during sleep, namely, increasedwork of breathing, intermittent hypoxemia,sleep fragmentation, and alveolar hypoventila-tion. Obviously, the separation of these elementsas individual contributors is artificial due totheir anticipated interaction in end-organ effects.Nevertheless, this categoric classification per-mits more accurate representation of the relativecontributions of each of these alterationsinduced by OSA and provides a useful method-ological approach to their investigation.

INCREASED WORK OF BREATHING

A logical and potential consequence of increasedwork of breathing during sleep in children withOSA is the development of failure to thrive (FTT).Although in early reports of children with severeOSA FTT was often present,28,33,34 it is currentlyestimated that only a minority of children withOSA will present with this problem, most probablybecause of earlier recognition and referral.Adenotonsillectomy and complete resolution ofOSA will induce significant growth improvementsin those children who present with FTT. In a smallstudy of children with OSA and FTT, Brouilletteand associates35 found that adenotonsillectomyresulted in catch-up growth in all six children. Infact, adenotonsillectomy has been shown toincrease the height and weight velocities of chil-dren with normal growth and OSA.36 Interestingly,even obese children with OSA will demonstrateweight gain after surgical removal of their enlargedtonsils and adenoids for the resolution of OSA.37

Mechanisms for growth deceleration in OSAmay be varied. These mechanisms may include,but are not limited to, decreased appetite that ispossibly associated with reduced olfaction inchildren with adenoidal hypertrophy, dysphagiafrom tonsillar hypertrophy,38 decreased levels ofinsulin-like growth factor-1 (IGF-1), IGF bind-ing protein and possibly growth hormonerelease,39-42 hypoxemia, acidosis, and the in-creased metabolic cost of breathing during sleep.While the mechanisms underlying decreased cir-culating levels of IGF-1 in children with OSAremain unclear, IGF-1 will recover after tonsil-lectomy and adenoidectomy and parallel therebound growth that characterizes the responsesof these patients.41 It is possible that disruptionof non–rapid eye movement (NREM) sleep that

is not immediately recognizable when usingvisual inspection of the sleep records may play arole, since growth hormone release and itsdownstream signaling through IGF-1 occur dur-ing NREM sleep in general and more specificallyduring delta sleep.43,44 This is particularly appli-cable to pediatric OSA because even in snorerswho do not fulfill the current criteria for OSA,there is evidence of stunted growth and disrup-tion of IGF binding protein–3 circulating levels,suggesting that some of these children may suf-fer from disrupted sleep that in turn would affectgrowth hormone release.42

Alternatively, changes in energy expenditureduring sleep could account for a reduction in thelinear growth of children with OSA. Indeed,Marcus and coworkers45 studied 14 childrenwith moderate OSA (mean age 4 ± 1 years) anddemonstrated that increased metabolic require-ments were present during sleep and normalizedafter adenotonsillectomy, with a concomitantgain in body weight. In this study no evidencefor decreased appetite was apparent, sincecaloric intake was similar both before and aftersurgery in all children. These findings suggestthat poor growth in some children with OSAmay be secondary to the increased energy expen-diture that directly results from the increase inwork of breathing during sleep rather than beassociated with decreased caloric intake.1

However, this a priori logical mechanism hasbeen recently challenged by Bland and col-leagues46 in a study of 11 children (mean age6 ± 2 years) with a polysomnographically con-firmed diagnosis of OSA. These investigatorsfound no evidence of increased energy require-ments either before or after adenotonsillectomyin these patients; in fact, their total daily energyexpenditures were similar to those of healthyage- and gender-matched control subjects.

Thus, the FTT in pediatric OSA does notappear to be associated with increased dailyenergy requirements but involves a combina-tion of increased night-time energy expenditurecaused by increased respiratory effort and,more importantly, disruption of the growthhormone-IGF-1 pathway. In general, completerecovery of growth and resumption of its nor-mal patterns will occur after adenotonsillec-tomy.28,34,37 Furthermore, we are unaware ofany persistent, residual somatic deficits afterthe resolution of OSA.

Consequences of Obstructive Sleep Apnea Syndrome Chapter 18 213

Ch18.qxd 1/25/05 3:25 AM Page 213

INTERMITTENT HYPOXEMIA

Frequent oxygen desaturations during sleep arecommon in children with OSA. In a study of 11children, Gislason and Benediktsdottir3 foundthat 73% had three or more respiratory eventsper hour that were associated with desaturationsof >4%. Similarly, Stradling and associates38

found that 61% of the 61 children selected foradenotonsillectomy because of recurrent tonsil-litis had hypoxemia during sleep. Elevation ofpulmonary artery pressure due to hypoxia-induced pulmonary vasoconstriction is a seriousconsequence of OSA in children and can lead topersistent pulmonary hypertension and cor pul-monale.26 The overall magnitude of pressureincreases in the pulmonary circulation is proba-bly smaller than that occurring during the sus-tained, chronic hypoxemia observed at highaltitudes. Indeed, direct measurements of pul-monary artery pressures in rats and miceexposed to hypoxia revealed that the increase inpulmonary artery pressure was greater whencontinuous hypoxia was applied than it was withepisodic hypoxia.47 In 27 pediatric patients withmoderate to severe OSA, radionuclide assess-ment of right ventricular function revealedreduced ejection fraction in 37% and wallmotion abnormality in 45% of children withOSA.48 These abnormalities were reversible aftersurgery. However, there is no prospective studyassessing the overall prevalence and severity ofpulmonary hypertension in a large cohort ofchildren with sleep-disordered breathing (SDB).It also remains unclear whether untreated orlong-lasting OSA will be associated with vascularremodeling of the pulmonary circulation inthese children. Furthermore, evidence from ani-mal models exposed to hypoxia for a shortperiod of time during early postnatal life revealsthat pulmonary hypertension is increased whenexposed to hypoxia later in infancy.49 Theseauthors speculated that the disrupted alveolar-ization and vascular growth after the briefhypoxic insult may increase the risk of severepulmonary hypertension later in life.

Normally, the cardiovascular effects of breath-ing are recognizable in the systemic circulationbut are of small magnitude, as revealed by respi-ratory sinus arrhythmia and minor fluctuationsin blood pressure that are synchronous with res-piratory periodicity. In the context of OSA, sys-

temic hypertension has emerged as a major car-diovascular consequence in adult patients.Indeed, elevated blood pressures are present dur-ing wakefulness in approximately 50% of adultpatients with OSA.50 Although the pathophysio-logic mechanisms of such elevation in arterialtension are still under intense investigation, itappears that intermittent hypoxemia is the majorcontributor to this serious consequence of OSA,with less significant roles being played by sleepfragmentation and episodic hypercapnia. Not-withstanding such considerations, enhancedsympathoadrenal discharge, altered homeostasisof the renin-angiotensin pathway, and heightenedsympathetic autonomic nervous system tone willdevelop along with increases in small arteriolarvasomotor contractility suggestive of endothelialdysfunction.51-62

The evidence for alterations in autonomicnervous system tone is clearly not as extensivewhen addressing the pediatric OSA population.In an earlier study, Aljadeff and coworkers exam-ined heart rate variability (a noninvasive probeof autonomic nervous system tone) in childrenwith moderate to severe OSA and matchedcontrol subjects.63 Inspection of moment-to-moment changes in RR intervals revealed sus-tained increases in sympathetic activity as wellas increased vagal discharge during respiratoryevents. Similarly, Baharav and colleagues foundevidence of increased sympathetic tone asderived from spectral analysis of cardiac period-icities in children with OSA.64 The marked sen-sitivity of the sympathetic system to the in-creased respiratory effort that occurs throughoutthe night in OSA patients, with the subsequenthypoxia and arousal, will be manifested aschanges in vascular hemodynamics that can bederived from the time differential between theR wave (electrical cardiac contraction) and theperipheral pulse waveform, (i.e., pulse transittime). This relatively easy-to-obtain measuremay in fact be used as a circulatory corollary ofrespiratory and arousal events in children65 andmay ultimately become one of the useful param-eters for the OSA screening of snoring children.

Because children have greater vascular com-pliance than adults, the magnitude and severityof blood pressure elevation in pediatric patientswith OSA are not as prominent as those de-scribed for older patients. Nevertheless, they dooccur and are manifested primarily as diurnal


Ch18.qxd 1/25/05 3:25 AM Page 214

elevations of diastolic blood pressure.27 Theseautonomic alterations are mediated by both theepisodic hypoxia that characteristically accompa-nies the OSA syndrome and by the repeatedarousals and carbon dioxide elevations of thiscondition. The long-term implications of sucheffects on cardiovascular morbidity during thelater stages of childhood or even adulthood haveyet to be explored. However, in preliminary ani-mal experiments, when rodents were exposed toa model of OSA66 at an age corresponding to thatfor the peak prevalence of pediatric OSA in child-hood, marked attenuation of their baroreceptorfunction lasting into late adulthood was found.67

As a result, early childhood perturbations maylead to lifelong consequences; in other words,certain types of adult cardiovascular disease mayrepresent, at least in part, sequelae from a priori“unrelated events” during childhood. Therefore,early identification of children with alterations inbaroreceptor and autonomic nervous systemfunctions in the context of pediatric OSA maylead to detection of a population potentially atrisk for ulterior development of hypertensionand its cardiovascular-associated morbidity.

Evidence for an Effect of OSA on Learning and BehaviorAnother potentially very serious consequence ofintermittent hypoxia may involve its long-termdeleterious effects on neuronal and intellectualfunctions. Reports of decreased intellectualfunction in children with tonsillar and adenoidal

hypertrophy date back to 1889, when Hillreported on “some causes of backwardness andstupidity in children.”68 School problems havebeen repeatedly reported in case-series of childrenwith OSA and in fact may underlie more extensivebehavioral disturbances such as restlessness,aggressive behavior, excessive daytime sleepiness.and poor test performance.2,22,24,31,69,70,74 Table18–1 summarizes the evidence for OSA and neu-rocognitive and behavioral functions.

The neurocognitive and behavioral conse-quences of disrupted sleep architecture andintermittent hypoxemia in children with SDB areonly now being defined by appropriate scientificmethodology. There is increasing evidence tosupport an association between OSA and atten-tion deficit hyperactivity disorder (ADHD) inchildren.77,80,85-88 Several studies have docu-mented that children with SDB often have prob-lems with attention and behavior similar tothose observed in children with ADHD.2,22-24,31

In addition, three separate survey studiesencompassing almost 3300 children have docu-mented daytime sleepiness, hyperactivity, andaggressive behavior in children who snored.2,8,81

It has been further suggested that up to one thirdof all children with frequent, loud snoring willdisplay significant hyperactivity and inatten-tion,2 while a reciprocal relation also appears tobe true, that is, children with ADHD exhibitmore sleep disturbances more frequently thannormal children.81,85

However, most of these studies of ADHD chil-dren have relied on parental reporting rather


Symptom Reference(s)

Based on Parental Report

Daytime sleepiness 12, 38, 75, 76Behavior (including hyperactivity, inattention, 2, 12, 22-24, 31, 74, 75, 77-81

and aggression)Depression 82Enuresis 12, 76Impaired school performance 12, 71

Objectively Measured

Behavior 9, 73, 82Depression 83Impaired school performance 25, 72, 84

Table 18–1. Brief Summary of Evidence for Obstructive Sleep Apnea and Neurocognitive/Behavioral Function

Au: Refto notcited intext ortablebelow

Ch18.qxd 1/25/05 3:25 AM Page 215

than on objectively collected data (see Table18–1). In the few instances in which sleep stud-ies were conducted in patients with ADHD,attention was given to sleep architecture aloneand did not include cardiorespiratory data.89,90

In a recent study from our laboratory, both sub-jective and objective sleep measures wereobtained and showed that objectively measuredsleep and respiratory disturbances are relativelyfrequent among children with ADHD, albeit notas frequent as those anticipated from parentalreports.9 In this study we also found that theprevalence of OSA in a cohort of children withADHD, verified by neuropsychological testing,does not appear to differ from that found in thegeneral population.2,3,91 However, an unusuallyhigh frequency of OSA was found among chil-dren with mild to moderate increases in hyper-activity as measured by the Conner ParentRating Scale,92 a well validated method for dis-criminating between children with ADHD andnormal control subjects (i.e., without trueADHD). This suggests that while OSA caninduce significant behavioral effects being mani-fested as increased hyperactivity and inattention,it will not overlap with true clinical ADHD whenthe latter is assessed by more objective tools thanjust parental perception. Therefore, in a childpresenting with parental complaints of hyperac-tivity and who does not meet the diagnostic cri-teria of ADHD after undergoing a thoroughevaluation, as recently recommended by theAmerican Academy of Pediatrics,93 a carefulsleep history should be taken, and if snoring ispresent, an overnight polysomnographic evalua-tion should be performed.

The mechanism or mechanisms by which OSAmay contribute to hyperactivity remain unknown.It is possible that both the sleep fragmentation(see below) and episodic hypoxia that character-ize OSA will lead to alterations within the neuro-chemical substrate of the prefrontal cortex, withresulting executive dysfunction.94 Notwithstand-ing these considerations, sleep disturbances arefrequently reported by the parents of ADHD chil-dren even when snoring is excluded.88 Based onthe available literature, the comorbidity of OSAand ADHD could be shared by a substantial num-ber of hyperactive children. In fact, it has beensuggested that up to 25% of children with a diag-nosis of ADHD may actually have OSA.24

However, such rather extensive overlap may be

less prominent than that previously estimated ifmedication status and psychiatric comorbidity areaccounted for in the analysis.95

Inverse relationships among memory, learn-ing, and OSA have been documented.72,73 Inaddition, improvements in learning and behav-ior have been reported after treatment for OSA inchildren,25,38,73 suggesting that the neurocogni-tive deficits are at least partially reversible. In alarge cohort of first graders whose academic per-formance was in the lowest 10th percentile oftheir class, we found not only a six- to ninefoldincrease in the expected incidence of OSA butalso a significant improvement in school gradesafter adenotonsillectomy and resolution of OSA.However, since we did not know what the opti-mal learning potential was for these children, itis possible that long-term residual deficits mayoccur even after treatment. To further explorethis possibility, we examined the history ofsnoring during early childhood in two groups of13- to 14-year-old children who were matchedfor age, gender, race, school attending, andsocioeconomic status but whose performanceswere either in the upper or lower quartile oftheir class. We found that children who snoredfrequently and loudly during their early child-hood were at greater risk for poor academic per-formance in later years, well after snoring hadresolved.84 Therefore, these findings suggest thateven if a component of the OSA-induced learn-ing deficits is reversible, there may be a long-lasting residual deficit in learning capability andthat the latter may represent a “learning debt”;that is, the decreased learning capacity duringOSA may have led to such a delay in learnedskills that recuperation is possible only withadditional teaching assistance. Alternatively,the processes underlying the learning deficitduring OSA may have irreversibly altered theperformance characteristics of the neuronalcircuitry responsible for learning particular skills(Fig. 18–2; see below). More recent work furthersuggests that snoring children perform poorlyon measures of “executive functioning,”96 thatis, the ability to develop and sustain an organ-ized, future-oriented, and flexible approach toproblem solving. Executive dysfunction appearsto be the most prominent area of cognitiveimpairment with untreated SDB in adults,97,98

and our work now extends such observations tothe pediatric age range.9,96


Ch18.qxd 1/25/05 3:25 AM Page 216

SLEEP FRAGMENTATION

The physiologic and behavioral effects of sleeploss have been extensively investigated in adults,although the effects of OSA on sleep in childrenremain poorly understood. In adults OSA isknown to cause sleep fragmentation due to mul-tiple arousals. Sleep fragmentation may beachieved by auditory stimuli inducing arousalsthroughout the night and has been shown toresult in performance decrements the followingday.99-101 Functions requiring concentration anddexterity are significantly affected by the exces-sive daytime sleepiness (EDS) resulting fromsleep fragmentation, and subjects are often con-fused and disoriented. Aggressive outbursts, irri-tability, anxiety, and depression are all knownmanifestations of EDS in adults and appear to befully reversible once sleep recovery is allowed.However, in contrast to adults with OSA, EDSdoes not appear to be a major feature of child-hood OSA when assessed by either parental

reports15,102 or more objective tools such as themultiple sleep latency test.103 In fact, EDSoccurred in only 13% of the children with OSAand was linearly correlated with the severity ofrespiratory disturbance such that a multiple sleeplatency test of <10 minutes, which is indicative ofEDS, was extremely unlikely when the apnea-hypopnea index was <15 events/hr. However, evi-dence of increased sleepiness, measured bymultiple sleep latency testing, was reported byLecendreux and colleagues in 30 children withADHD who otherwise had no significantpolysomnographic alterations,104 raising the pos-sibility that a subset of children with hyperactiv-ity may in fact have an arousal disorder .

However, there is no reason to believe thatsleep fragmentation affects children differentlyfrom adults. If this assumption is true, we wouldanticipate that sleep fragmentation–related mor-bidity would be reduced in children, since dis-ruption of sleep architecture appears to beunusual in this population.105 Indeed, whenarousals are scored using criteria developed foradults,106 children with OSA have fewer EEGarousals than adults with OSA and thus are bet-ter able to preserve sleep architecture.45,102 Thelack of arousals may account for the increasedfrequency of nocturnal enuresis seen in this pop-ulation. Indeed, in a report by Guilleminault andassociates,22 18% of children presented with sec-ondary enuresis, while Weider and coworkers107

reported a 76% improvement in nocturnalenuresis after tonsillectomy and adenoidectomy.Nonetheless, EDS-like morbidity is in fact pres-ent in children with OSA, suggesting either thatcurrent techniques to define arousal are not sen-sitive enough to detect sleep fragmentation inchildren108 or that functional susceptibility tosleep fragmentation is higher in children.Alternatively, intermittent hypoxia rather thansleep fragmentation may be the major determi-nant of neurocognitive morbidity in children. Asmentioned above, separation of the effects ofsleep fragmentation and intermittent hypoxia onbehavior and cognition is not possible in chil-dren. Thus, the individual role played by each ofthese factors in relation to neurobehavioral mor-bidity will be impossible to determine in thecontext of human studies.

To overcome such constraints, we recentlydeveloped a rodent model that allows for separa-tion of the roles played by intermittent hypoxia


2 GHz chip

Simple computation: 32 ms versus 44 msComplex computation: 2.23 ms versus “freeze”

A B

SDBH

DG

EF

C

A B

H

DM

EF

C

Second best800 MHz chip

Figure 18–2. Schematic model illustrating how the lossof a given neuron within a neural circuit to sleep-disor-dered breathing (SDB) (in this case the neuron labeled“G”) leads to its replacement by the next best (2nd Best)option available (in this case the neuron labeled “M”).Such cellular changes, which underlie the fundamentalmechanisms of brain plasticity, are accompanied bychanges in the intrinsic properties of the neural circuit.Analogous to a computer chip, the performance of thecircuit is reduced from 2 GHz to 800 MHz. As such, sim-ple computations such as those required from early-school children would not obligatorily translate intovisible learning/performance deficits. However, as theintellectual requirements increase over time, the decre-ments in performance associated with SDB could trans-late into “FREEZE,” that is, the inability to learn orperform the computational task being requested. Thus,the deficits associated with pediatric OSA may not mani-fest themselves until later in life.

Ch18.qxd 1/25/05 3:25 AM Page 217

and sleep fragmentation.66 We found that inter-mittent hypoxia during sleep is associated withsignificant increases in neuronal cell loss andadverse effects on spatial memory tasks in theabsence of significant sleep fragmentation ordeprivation. Furthermore, when this model wasapplied to developing rodents, a unique periodof neuronal susceptibility to episodic hypoxiaduring sleep emerged and coincides with ages atwhich OSA prevalence peaks in children.109

Since this age coincides with that of the criticalperiod for brain development, it is possible thatduring this period the delayed diagnosis andtreatment of OSA will impose a greater burdenon vulnerable brain structures and ultimatelyhamper the overall neurocognitive potential ofchildren with SDB. This issue has been recentlyinvestigated, and indeed, developing rats exhibitgreater decreases in the acquisition and retentionof spatial tasks.110 Interestingly, male ratsexposed to intermittent hypoxia during sleepalso exhibit increased locomotor activity, uni-quely reminiscent of the hyperactivity of chil-dren with SDB, in whom a male predominance iswell established.

ALVEOLAR HYPOVENTILATION

Snoring children with and without OSA repre-sent a classic example of intermittent alveolarhypoventilation resulting from increased upperairway resistance associated with insufficientcompensatory respiratory drive mechanismsduring sleep. Since children are less prone to col-lapse their upper airways during sleep,29 longperiods of increased upper airway resistance andhypercapnia with and without hypoxemia willdevelop during the night rather than more fre-quent, discrete, obstructive apneic events, lead-ing to the term “obstructive hypoventilation.”13

This feature of OSA in children has led to therecommendation to monitor carbon dioxide lev-els using both end-tidal and transcutaneous ap-proaches, if possible.111

In adults with OSA, a blunted ventilatory driveto hypercapnia during wakefulness may developand may contribute to the pathophysiology ofupper airway obstruction. In addition, daytimehypercapnia may occur in adult patients withOSA and is probably the end result of multifacto-rial interactions involving respiratory dynamics,obesity, and leptin and other endocrine abnormal-

ities.112 While intermittent hypercapnia is fre-quent in pediatric OSA during sleep, it will almostalways disappear during wakefulness, indicatingthat hypercapnic and hypoxic ventilatory drivesare intact in children.27,45 However, arousalresponses to hypercapnia are attenuated duringsleep, suggesting that interactions among sleep,ventilatory drive, and neural mechanisms under-lying upper airway patency affect the recruitmentof arousal centers in OSA.27

Intermittent elevations in PaCO2 not only canexacerbate the effects of intermittent hypoxemiaon neural tissue but also can affect cerebral cir-culation and vasomotor activity.113 Furthermore,hypercapnia can directly suppress neural func-tion,114 and it is well established that CO2 eleva-tions will increase pulmonary artery pressure,possibly through increases in sympathetictone.115 Therefore, the typical intermittenthypercapnia that occurs in children with OSAmay either directly or, through interactions withepisodic hypoxia and arousal, exacerbate end-organ injury in this condition.

CONCLUSIONS

It is becoming increasingly clear that OSA inchildren can have adverse effects on somaticgrowth, induce cardiovascular alterations suchas pulmonary and systemic hypertension, andlead to substantial neurobehavioral deficits,some of which may not be reversible if treatmentis delayed. Based on our current understandingof the morbidity affecting pediatric OSA, it isimperative to direct future efforts toward animproved definition of the spectrum of the OSA-induced syndrome, such as providing moreaccurate guidelines for treatment.

References


2. Ali NJ, Pitson D, Stradling JR: Snoring, sleep dis-turbance and behaviour in 4-5 year olds. Arch DisChild 1993;68:360-366.

3. Gislason, T, Benediktsdottir B: Snoring, apneic epi-sodes, and nocturnal hypoxemia among children 6months to 6-years-old. Chest 1995;107:963-966.

4. American Thoracic Society: Standards and indica-tions for cardiopulmonary sleep studies in children.Am J Respir Crit Care Med 1995;153:866-878.


Ch18.qxd 1/25/05 3:25 AM Page 218

5. Teculescu DB, Caillier I, Perrin P, et al: Snoring inFrench preschool children. Pediatr Pulmonol1992;13:239-244.

6. Hulcrantz E, Lofstarnd TB, Ahlquist RJ: The epi-demiology of sleep related breathing disorders inchildren. Int J Pediatr Otorhinolaryngol 1995;6(suppl):S63-S66.

7. Owen GO, Canter RJ, Robinson A: Snoring, apneaand ENT symptoms in the paediatric community.Clin Otolaryngol Allied Sci 1996;21:130-134.

8. Ferreira AM, Clemente V, Gozal D, et al: Snoringin Portuguese primary school children. Pediatrics2000;106:E64.

9. O’Brien LM, Holbrook CR, Mervis CB, et al: Sleepand neurobehavioral characteristics in 5-7 yearold children with parentally reported symptomsof ADHD. Pediatrics 2003;111:554-563.

10. Corbo GM, Forastiere F, Agabiti N, et al: Snoringin 9- to 15-year-old children: Risk factors andclinical relevance. Pediatrics 2001;180:1149-1154.

11. McKenzie M: A Manual of Diseases of the Throatand Nose, Including the Pharynx, Larynx,Trachea, Oesophagus, Nasal Cavities, and Neck.London, Churchill, 1880.

12. Guilleminault C, Eldridge F, Simmons FB,Dement WC: Sleep apnea in eight children.Pediatrics 1976;58:28-31.

13. Rosen CL, D’Andrea L, Haddad GG: Adult criteriafor obstructive sleep apnea do not identify chil-dren with serious obstruction. Am Rev Respir Dis1992;146:1231-1234.

14. Carroll JL, McLoughlin GM: Diagnostic criteriafor obstructive sleep apnea in children. PediatrPulmonol 1992;14:71-74.

15. Carroll JL, McColley SA, Marcus CL, Curtis S,Loughlin GM: Inability of clinical history to dis-tinguish primary snoring from obstructive sleepapnea syndrome in children. Chest 1995;108:610-618.

16. Sivan Y, Kornecki A, Schonfeld T: Screeningobstructive sleep apnoea syndrome by homevideotape recording in children. Eur Respir J1996;9:2127-2131.

17. Lamm C, Mandeli J, Kattan M: Evaluation ofhome audiotapes as an abbreviated test forobstructive sleep apnea syndrome (OSAS) in chil-dren. Pediatr Pulmonol 1999;27:267-272.

18. Brouillette RT, Morielli A, Leimanis A, et al:Nocturnal pulse oximetry as an abbreviated test-ing modality for pediatric obstructive sleep apnea.Pediatrics 2000;105:405-412.

19. Brouillette RT, Lavergne J, Leimanis A, et al:Differences in pulse oximetry technology canaffect detection of sleep-disordered breathing inchildren. Anesth Analg 2002;94:S47-S53.

20. Schnall RP, Shlitner A, Sheffy J, et al: Periodic,profound peripheral vasoconstriction—a newmarker of obstructive sleep apnea. Sleep 1999;22:939-946.

21. Lavie P, Shlitner A, Sheffy J, Schnall RP:Peripheral arterial tonometry: A novel and sensi-tive non-invasive monitor of brief arousals duringsleep. Isr Med Assoc J 2000;2:246-247.


23. Guilleminault C, Winkle R, Korobkin R,Simmons B: Children and nocturnal snoring—evaluation of the effects of sleep related respira-tory resistive load and daytime functioning. Eur JPediatr 1982;139:165-171.

24. Chervin R, Dillon J, Bassetti C, et al: Symptoms ofsleep disorders, inattention, and hyperactivity inchildren. Sleep 1997;20:1185-1192.

25. Gozal D: Sleep-disordered breathing and schoolperformance in children. Pediatrics 1998;102:616-620.

26. Shiomi T, Guilleminault C, Stoohs R, Schnittger I:Obstructed breathing in children during sleepmonitored by echocardiography. Acta Paediatr1993;82:863-871.


28. Everett AD, Koch WC, Saulsbury FT: Failure tothrive due to obstructive sleep apnea. Clin Pediatr1987;26:90-92.

29. Marcus CL: Pathophysiology of childhoodobstructive sleep apnea: Current concepts. RespirPhysiol 2000;119:143-154.

30. Isono S, Shimada A, Utsugi M, et al: Comparisonof static mechanical properties of the passivepharynx between normal children and childrenwith sleep-disordered breathing. Am J Respir CritCare Med 1998;157:1201-1212.

31. Owens J, Opipari L, Nobile C, Spirito A: Sleep anddaytime behavior in children with obstructivesleep apnea and behavioral sleep disorders.Pediatrics 1998;102:1178-1184.

32. Suen JS, Arnold JE, Brooks LJ: Adenotonsillec-tomy for treatment of obstructive sleep apnea inchildren. Arch Otolaryngol Head Neck Surg1995;121:525-530.

33. Ahlqvist-Rastad J, Hutcrantz E, Melander H: Bodygrowth in relation to tonsillar enlargement andtonsillectomy. Int J Pediatr Otorhinolaryngol1992;24:55-61.

34. Freezer NJ, Bucens IK, Robertson CF:Obstructive sleep apnoea presenting as failure tothrive in infancy. J Paediatr Child Health 1995;31:172-175.


Ch18.qxd 1/25/05 3:25 AM Page 219

35. Brouillette RT, Fernbach SK, Hunt CE:Obstructive sleep apnea in infants and children.J Pediatr 1982;100:31-40.

36. Lind MG, Lundell BP: Tonsillar hyperplasia inchildren: A cause of obstructive sleep apneas,CO2 retention, and retarded growth. ArchOtolaryngol 1982;108:650-654.

37. Soultan Z, Wadowski S, Rao M, Kravath RE: Effectof treating obstructive sleep apnea by tonsillec-tomy and/or adeniodectomy on obesity in chil-dren. Arch Pediatr Adolesc Med 1999;153:33-37.

38. Stradling JR, Thomas G, Warley ARH, et al: Effectof adenotonsillectomy on nocturnal hypoxaemia,sleep disturbance, and symptoms in snoring chil-dren. Lancet 1990;335:249-253.

39. Waters KA, Kirjavainen T, Jimenez M, et al:Overnight growth hormone secretion in achon-droplasia: Deconvolution analysis, correlationwith sleep state, and changes after treatment ofobstructive sleep apnea. Pediatr Res 1996;39:547-553.

40. Chiba S, Ashikawa T, Moriwaki H, et al: Theinfluence of sleep breathing disorder on growthhormone secretion in children with tonsil hyper-trophy (in Japanese). Nippon Jibiinkoka GakkaiKaiho 1998;101:873-878.

41. Bar A, Tarasiuk A, Segev Y, Phillip M, Tal A: Theeffect of adenotonsillectomy on serum insulin-like growth factor-I and growth in children withobstructive sleep apnea syndrome. J Pediatr1999;135:76-80.

42. Nieminen P, Lopponen T, Tolonen U, et al:Growth and biochemical markers of growth inchildren with snoring and obstructive sleepapnea. Pediatrics 2002;109:e55.

43. Van Cauter E, Copinschi G: Interrelationshipsbetween growth hormone and sleep. GrowthHorm IGF Res 2000;10(Suppl B):S57-S62.

44. Obal F Jr, Krueger JM: The somatotropic axis andsleep. Rev Neurol (Paris) 2001;157:S12-S15.

45. Marcus CL, Carroll JL, Koerner CB, et al:Determinants of growth in children with theobstructive sleep apnea syndrome. J Pediatr1994;125:556-562.

46. Bland RM, Bulgarelli S, Ventham JC, et al: Totalenergy expenditure in children with obstructivesleep apnoea syndrome. Eur Respir J 2001;18:164-169.

47. Fagan KA: Selected contribution: Pulmonaryhypertension in mice following intermittenthypoxia. J Appl Physiol 2001;90:2502-2507.

48. Tal A, Leiberman A, Margulis G, Sofer S:Ventricular dysfunction in children with obstruc-tive sleep apnea: Radionuclide assessment.Pediatr Pulmonol 1988;4:139-143.

49. Tang J-R, Le Cras TD, Morris KG Jr, Abman SH:Brief perinatal hypoxia increases severity of pul-

monary hypertension after reexposure to hypoxiain infant rats. Am J Physiol. Lung Cell MolPhysiol 2000;278:L356-L364.

50. Shepard JW Jr: Hypertension, cardiac arrhyth-mias, myocardial infarction, and stroke in relationto obstructive sleep apnea. Clin Chest Med 1992;13:437-458.

51. Fletcher EC, Bao G: Effect of episodic eucapnicand hypocapnic hypoxia on systemic blood pres-sure in hypertension-prone rats. J Appl Physiol1996;81:2088-2094.

52. Peled N, Greenberg A, Pillar G, et al: Contri-butions of hypoxia and respiratory disturbanceindex to sympathetic activation and blood pres-sure in obstructive sleep apnea syndrome. Am JHypertens 1998;11:1284-1289.

53. Brooks D, Horner RL, Floras JS, et al: Baroreflexcontrol of heart rate in a canine model of obstruc-tive sleep apnea. Am J Respir Crit Care Med1999;159:1293-1297.

54. Fletcher EC, Bao G, Li R: Renin activity and bloodpressure in response to chronic episodic hypoxia.Hypertension 1999;34:309-314.

55. Khoo MC, Kim TS, Berry RB: Spectral indices ofcardiac autonomic function in obstructive sleepapnea. Sleep 1999;22:443-451.

56. Loredo JS, Ziegler MG, Ancoli-Israel S, et al:Relationship of arousals from sleep to sympa-thetic nervous system activity and BP in obstruc-tive sleep apnea. Chest 1999;116:655-659.

57. Duchna HW, Guilleminault C, Stoohs RA, et al:Vascular reactivity in obstructive sleep apnea syn-drome. Am J Respir Crit Care Med 2000;161:187-191.

58. Fletcher EC: Effect of episodic hypoxia on sym-pathetic activity and blood pressure. RespirPhysiol 2000;119:189-197.

59. Kraiczi H, Hedner J, Peker Y, Carlson J: Increasedvasoconstrictor sensitivity in obstructive sleepapnea. J Appl Physiol 2000;89:493-498.

60. Phillips BG, Somers VK: Neural and humoralmechanisms mediating cardiovascular responsesto obstructive sleep apnea. Respir Physiol2000;119:181-187.

61. Fletcher EC: Invited review: Physiological conse-quences of intermittent hypoxia: Systemic bloodpressure. J Appl Physiol 2001;90:1600-1605.

62. Elmasry A, Lindberg E, Hedner J, et al:Obstructive sleep apnoea and urine cate-cholamines in hypertensive males: A population-based study. Eur Respir J 2002;19:511-517.

63. Aljadeff G, Gozal D, Schechtman VL, et al: Heartrate variability in children with obstructive sleepapnea. Sleep 1997;20:151-157.

64. Baharav A, Kotagal S, Rubin BK, et al: Autonomiccardiovascular control in children with obstruc-tive sleep apnea. Clin Auton Res 1999;9:345-351.


Ch18.qxd 1/25/05 3:25 AM Page 220

65. Katz ES, Marcus CL: The pulse transit time as ameasure of respiratory arousal in children withsleep-disordered breathing. Sleep 2001;24:A207.

66. Gozal D, Daniel JM, Dohanich GP: Behavioral andanatomical correlates of chronic episodic hypoxiaduring sleep in the rat. J Neurosci 2001;21:2442-2450.

67. Soukhova GK, Roberts AM, Lipton AJ, et al: Earlypostnatal exposure to intermittent hypoxia atten-uates baroreflex sensitivity in adult rats(abstract). Sleep 2002;25:A335-336.

68. Hill W: On some causes of backwardness and stu-pidity in children and the relief of the symptomsin some instances by nasopharyngeal scarifica-tions. BMJ 1889;II:711-712.

69. Weissbluth M, Davis A, Poncher J, Reiff J: Signs ofairway obstruction during sleep and behavioral,developmental and academic problems. DevBehav Pediatr 1983;4:119-121.

70. Singer LP, Saenger P: Complications of pediatricobstructive sleep apnea. Otolaryngol Clin NorthAm 1990;23:665-676.

71. Leach J, Olson J, Hermann J, Manning S:Polysomnographic and clinical findings in chil-dren with obstructive sleep apnea. ArchOtolaryngol Head Neck Surg 1992;118:741-744.

72. Rhodes SK, Shimoda KC, Wald LR, et al:Neurocognitive deficits in morbidly obese chil-dren with obstructive sleep apnea. J Pediatr1995;127:741-744.

73. Ali NJ, Pitson D, Stradling JR: Sleep disorderedbreathing: Effects of adenotonsillectomy onbehaviour and psychological functioning. Eur JPediatr 1996;155:56-62.

74. Chervin RD, Archbold KH: Hyperactivity andpolysomnographic findings in children evaluatedfor sleep-disordered breathing. Sleep 2001;24:313-320.

75. Ali NJ, Pitson D, Stradling JR: Natural history ofsnoring and related behaviour problems betweenthe ages of 4 and 7 years. Arch Dis Child1994;71:74-76.

76. Stein MA, Mendelsohn J, Obermeyer WH, et al:Sleep and behavior problems in school-aged chil-dren. Pediatrics 2001;107:E60.

77. Kaplan BJ, McNichol J, Conte RA, MoghadamHK: Sleep disturbance in preschool-aged andhyperactive and nonhyperactive children.Pediatrics 1987;80:839-844.

78. Hansen DE, Vandenberg B: Neuropsychologicalfeatures and differential diagnosis of sleep apneasyndrome in children. J Clin Child Psychol1997;26:304-310.

79. Corkum P, Tannock R, Moldofsky H: Sleep dis-turbance in children with attention-deficit hyper-activity disorder. J Am Acad Child AdolescPsychiatry 1998;37:637-646.

80. Corkum P, Tannock R, Moldofsky H, et al:Actigraphy and parental ratings of sleep in chil-dren with attention deficit/hyperactivity disorder(ADHD). Sleep 2001;24:303-312.

81. Chervin RD, Archbold KH, Dillon JE, et al:Inattention, hyperactivity, and symptoms of sleepdisordered breathing. Pediatrics 2002;109:449-456.

82. Sadeh A, McGuire JP, Sachs H, et al: Sleep andpsychological characteristics of children on a psy-chiatric inpatient unit. J Am Acad Child AdolescPsychiatry 1995;34:813-819.

83. Benca RM, Obermeyer WH, Thisted RA, GillinJC: Sleep and psychiatric disorders: A meta-analy-sis. Arch Gen Psychiatry 1992;49:651-668.

84. Gozal D, Pope DW Jr: Snoring during early child-hood and academic performance at ages 13-14years. Pediatrics 2001;107:1394-1399.

85. Ball JD, Tiernan M, Janusz J, Furr A: Sleep pat-terns among children with attention-deficithyperactivity disorder: A re-examination of pa-rental perceptions. J Pediatr Psychol. 1997;22:389-398.

86. Ring A, Stein D, Barak Y, et al: Sleep disturbancesin children with attention-deficit/hyperactivitydisorder: A comparative study with healthy sib-lings. J Learn Disabil 1998;31:572-578.

87. Stein MA: Unraveling sleep problems in treatedand untreated children with ADHD. J ChildAdolesc Psychopharmacol 1999;9:157-168.

88. Owens J, Maxim R, Nobile C, McGuinn M, MsallM: Parental and self-report of sleep in childrenwith attention deficit/hyperactivity disorder. ArchPediatr Adolesc Med 2000;154:549-555.

89. Busby K, Firestone P, Pivik RT: Sleep patterns inhyperkinetic and normal children. Sleep1981;4:366-383.

90. Greenhill L, Puig-Antich J, Goetz R, et al: Sleeparchitecture and REM sleep measures in prepu-bertal children with attention deficit disorderwith hyperactivity. Sleep 1983;6:91-101.

91. Redline S, Tishler PV, Schluchter M, et al: Riskfactors for sleep-disordered breathing in children:Associations with obesity, race, and respiratoryproblems. Am J Respir Crit Care Med1999;159:1527-1532.

92. Conners CK: Conners’ Parent Rating Scales. NewYork, MHS Publishing, 1989.

93. American Academy of Pediatrics: ClinicalPractice Guideline: Diagnosis and Evaluation ofthe Child with Attention-Deficit/HyperactivityDisorder. Pediatrics 2000;105:1158-1170.

94. Beebe DW, Gozal D: Obstructive sleep apnea andthe prefrontal cortex: Towards a comprehensivemodel linking nocturnal upper airway dysfunc-tion to daytime cognitive and behavioral deficits.J Sleep Res 2002;11:1-16.


Ch18.qxd 1/25/05 3:25 AM Page 221

95. Corkum P, Moldofsky H, Hogg-Johnson S, et al:Sleep problems in children with attention-deficit/hyperactivity disorder: Impact of sub-type, comorbidity, and stimulant medication.J Am Acad Child Adolesc Psychiatry 1999;38:1285-1293.

96. Gozal D, Holbrook CR, Mehl RC, et al:Correlation analysis between NEPSY batteryscores and respiratory disturbance index in snor-ing 6-year old children: A preliminary report.Sleep 2001;24:A126.

97. Greenberg GD, Watson RK, Deptula D: Neuro-psychological dysfunction in sleep apnea. Sleep1987;10:254-262.

98. Naegele B, Thouvard V, Pepin JL, et al: Deficitsof cognitive executive functions in patients withsleep apnea syndrome. Sleep 1995;18:43-52.

99. Stepansky EJ, Lamphere P, Badia P: Sleep frag-mentation and daytime sleepiness. Sleep 1984;7:18-26.

100. Stepansky EJ, Lamphere J, Roehrs T: Experi-mental sleep fragmentation in normal subjects.Int J Neurosci 1987;33:207-214.

101. Chugh DK, Weaver TE, Dinges DF: Neuro-behavioral consequences of arousals. Sleep1996;19:S198-S201.

102. Frank Y, Kravath RE, Pollak CP, Weitzman ED:Obstructive sleep apnea and its therapy: Clinicaland polysomnographic manifestations. Pedi-atrics 1983;71:737-742.

103. Gozal D, Wang M, Pope DW: Objective sleepi-ness measures in pediatric obstructive sleepapnea. Pediatrics 2001;108:693-697.

104. Lecendreux M, Konofal E, Bouvard M, et al:Sleep and alertness in children with ADHD. JChild Psychol Psychiatry 2000;41:803-812.

105. Goh DY, Galster P, Marcus CL: Sleep architectureand respiratory disturbances in children withobstructive sleep apnea. Am J Respir Crit CareMed 2000;162:682-686.

106. Sleep Disorders Atlas Task Force, GuilleminaultC (ed). EEG arousals: Scoring and rules andexamples. Sleep 1992;15:173-184.

107. Weider DJ, Sateia MJ, West RP: Nocturnalenuresis in children with upper airway obstruc-tion. Otolaryngol Head Neck Surg 1991;105:427-432.

108. Bandla HPR, Gozal D: Dynamic changes in EEGspectra during obstructive apnea in children.Pediatr Pulmonol 2000;29:359-365.

109. Gozal E, Row BW, Schurr A, Gozal D:Developmental differences in cortical and hip-pocampal vulnerability to intermittent hypoxiain the rat. Neurosci Lett 2001;305:197-201.

110. Row BW, Kheirandish L, Neville JJ, Gozal D:Impaired spatial learning and hyperactivity indeveloping rats exposed to intermittent hypoxia.Pediatr Res 2002; 52:449-453.

111. Morielli A, Desjardins D, Brouillette RT: Trans-cutaneous and end-tidal carbon dioxide pres-sures should be measured during pediatricpolysomnography. Am Rev Respir Dis 1993;148:1599-1604.

112. Gozal D: Determinants of daytime hypercapniain obstructive sleep apnea: Is obesity the onlyone to blame? Chest 2002;121:320-321.

113. Loeppky JA, Miranda FG, Eldridge MW:Abnormal cerebrovascular responses to CO2 insleep apnea patients. Sleep 1984;7:97-109.

114. Itoh Y, Yoshioka M, Kemmotsu O: Effects ofexperimental hypercapnia on hippocampallong-term potentiation in anesthetized rats.Neurosci Lett 1999;260:201-220.

115. Myers JL, Domkowski PW, Wang Y, Hopkins RA:Sympathetic blockade blunts hypercapnic pul-monary arterial vasoconstriction in newbornpiglets. Eur J Cardiothorac Surg 1998;3:298-305.


Ch18.qxd 1/25/05 3:25 AM Page 222

PATHOPHYSIOLOGY

The anatomy and neural control of the pharynxhave evolved to serve its multiple functions as aconduit for air, liquids, and solids. The pharynxmust be collapsible to facilitate swallowing andspeech. Coordinated neuromuscular action isrequired to propel food into the esophagus butnot the nose or trachea, as well as for speech.However, the collapsibility required for speechand swallowing can be an impediment to respi-ration, when the patency of the pharynx isrequired to conduct air from the nose to the lar-ynx. In conscious individuals, the pharynx mustremain open at all times except during momen-tary closure associated with swallowing, speech,regurgitation, and eructation.

The patency of the pharynx is determined byboth anatomic and physiologic factors. Duringinspiration, intraluminal pressure is negative(subatmospheric) and tends to produce pharyn-geal collapse. Assuming constant flow, Bernoulli’sprinciple dictates that a smaller airway wouldresult in even more negative intraluminal pres-sure, with a greater tendency to collapse. Thus,one would expect that individuals with a smallpharynx would be at a greater risk for airwaycollapse during sleep, but this is not necessarilythe case. The airway of a child is inherentlysmaller than that of an adult, yet the prevalenceof obstructive sleep apnea (OSA) is about 1% inschool-aged children,1-3 compared with a preva-lence of 4% in middle-aged men.4 Similarly, theprevalence of OSA is higher in men thanwomen,4 even though women have a smallerpharynx.5 Thus, the anatomic size of the phar-ynx cannot be the only factor predisposing anindividual to OSA.

Modeling the pharynx as a Starling resistorhas been useful in understanding its mechanicalproperties as a collapsible tube (Fig. 19–1). On

inspiration, pressure at the airway opening (nares)is atmospheric, and downstream pressure isequal to tracheal pressure. Therefore, pressurewithin the pharyngeal lumen is negative withrespect to the atmosphere. Collapse occurs whenthis negative pressure is lower than that sur-rounding the segment (Pcrit). In adults Pcrit is thelowest in nonsnorers, increasing progressively insubjects who snore and those with hypopneaand obstructive apnea. The results of one studyshowed that subjects with OSA had a positivePcrit, indicating that the airway would collapseduring sleep.6 Similar results have been found inchildren, in whom Pcrit correlated with the sever-ity of sleep-disordered breathing (Fig. 19–2).7

In awake adults, a small pharynx can be over-come by stiffening of the pharynx,8 probably byaugmenting upper airway tone.9 During sleepthis compensatory mechanism can be lost,resulting in airway closure despite (and partiallybecause of ) continuing respiratory effort.10

Hypoxemia and hypercapnia ensue, with theapnea terminating on arousal. Frequent arousalresults in sleep fragmentation and daytimesleepiness. Therefore, any factors that decreasepharyngeal size or increase pharyngeal compli-ance might be expected to predispose an indi-vidual to OSA syndrome (OSAS) (Fig. 19–3). Inone early description of OSAS in children, 32%had facial dysmorphism, presumably contribut-ing to an anatomically small pharynx, and anadditional 12% had underlying neuromusculardisorders, probably causing OSA from pharyn-geal hypotonia.11

Factors Affecting Pharyngeal SizeThe anatomic size of the pharynx stems from acombination of the underlying bony structureand soft tissue, including the tonsils and ade-noids. It may also be affected by dynamic factorssuch as obesity.

Obstructive Sleep Apnea Syndrome inInfants and Children: Clinical

Features and PathophysiologyLee J. Brooks

19

223

Ch19.qxd 1/25/05 3:25 AM Page 223

The underlying bony structure of the faceclearly contributes to the risk for OSAS. Childrenwho snore have a narrower anterior-posteriordimension of the pharynx.12 Several studieshave documented familial aggregation ofOSAS13 independent of weight, body massindex, or neck circumference.14,15 Family mem-bers tended to share a small pharyngeal volume,retroposed maxillae with shorter mandibles,16

and/or other craniofacial features, such as a

high, narrow hard palate.17 Thus, genetic fac-tors, particularly those controlling pharyngealstructure, can place even normal individuals ata risk for OSAS.

Children with craniofacial abnormalities maybe at a particularly high risk for OSAS.18 Micro-gnathia is a hallmark of Treacher Collins syn-drome and the Robin sequence. Mutations in thefibroblast growth factor receptor genes result inmidfacial hypoplasia of children with such dis-orders as Apert-Crouzon syndrome and achon-droplasia. These disorders are discussed in detailelsewhere.

Soft tissues can also narrow the pharynx, par-ticularly in cases of adenotonsillar hypertrophy.The tonsils and adenoids are at their largest,with respect to the underlying bony structures,between 3 and 6 years of age,19 coinciding withthe peak incidence of childhood OSAS. Theseverity of obstructive events is related to thesize of the adenoids20 and most otherwise nor-mal children with OSAS respond well to adeno-tonsillectomy.21 However, OSAS is not caused byadenotonsillar hypertrophy alone. Several stud-ies have been unable to show a correlationbetween upper airway or adenotonsillar size andthe number of respiratory events that occur dur-ing sleep.20,22-24 In addition, some children withadenotonsillar hypertrophy but no other knownrisk factors for OSAS are not cured by adenoton-sillectomy.21 Finally, Guilleminault and colleagues

224 Chapter 19 Obstructive Sleep Apnea Syndrome in Infants and Children

RUS

PcritRDS

PUS PDS

Upstreamsegment

Downstreamsegment

Collapsible segment

Figure 19–1. The Starling resistor model of the upperairway. The upper airway is represented as a tube with acollapsible segment. The upstream (nasal) and down-stream (tracheal) segments have fixed diameters andresistances (RUS, RDS) and pressures (PUS, PDS). Collapseoccurs when the pressure surrounding the airway (Pcrit) is greater than that within the airway. (Redrawn with permission from Marcus et al: Upper airway collapsibilityin children with obstructive sleep apnea syndrome. J ApplPhysiol 1994;77:918-924.)

40

30

20

10

0

Apn

ea in

dex

–30 –25

Pcrit (cm H2O)

–20 –15 –10 –5 0 5

OSAS

Abnormal UAneuromotor

tone

UA Narrowing– Adenotonsillar hypertrophy– Craniofacial structure– Obesity

Other factors– Genetic– Hormonal– ?

Figure 19–3. Obstructive sleep apnea syndrome(OSAS) in children may result from a combination of fac-tors. Structural anomalies may result in a narrow upperairway (UA), and decreased neuromuscular tone mayresult in a floppy UA. Genetic, hormonal, or other factorsmay also have an impact on these conditions.

Figure 19–2. Pcrit (the pressure surrounding the col-lapsible pharyngeal segment) vs. apnea index (numberof obstructive apnea events per hour of sleep). There is asignificant correlation between Pcrit and the severity ofobstructive sleep apnea (r = .80, P < .02). (Redrawn withpermission from Marcus et al: Upper airway collapsibilityin children with obstructive sleep apnea syndrome. J ApplPhysiol 1994;77:918-924.)

Ch19.qxd 1/25/05 3:25 AM Page 224

described a cohort of children who were cured ofOSAS by adenotonsillectomy but developed arecurrence during adolescence.25 Thus, it appearsthat childhood OSAS is a dynamic processresulting from a combination of structural andneuromotor abnormalities rather than from struc-tural abnormalities alone.

Obesity may narrow the pharynx due to thedeposition of adipose tissue within the musclesand soft tissue surrounding the airway.26 Althoughmost adults with OSAS are obese, children withOSAS are often of normal weight or even fail tothrive.27 However, obesity does increase the child’srisk for OSAS.20,28-31

Factors Affecting Pharyngeal ComplianceWomen and children have a smaller pharynxthan men yet are at a lower statistical risk forOSA. Acoustic studies of the upper airwaydemonstrate that women have a relatively stiffpharynx compared with men.5 Children alsohave different pharyngeal pressure–flow char-acteristics, suggesting a stiffer upper airwaythan that in adults.32 These findings suggestthat both genetic and hormonal factors likelyaffect pharyngeal compliance and hence therisk for OSA. This can be modified by many dis-ease states, particularly those affecting neuro-muscular tone. Children with cerebral palsyand other neurologic disorders are at anincreased risk for OSA. Children with Downsyndrome and Prader-Willi syndrome are at anincreased risk for OSA because of decreasedneuromuscular tone as well as anatomic fac-tors. Obesity may make the pharynx more com-pliant and may also increase the risk forsleep-disordered breathing by decreasing oxy-hemoglobin saturation because of reducedfunctional residual capacity and/or blunting ofcentral respiratory drive.

CLINICAL FEATURES

The clinical appearance of children with OSAS isoften different from that of adults. However, inboth groups the problem is usually first recog-nized by the female caregiver (wife or mother),and there is often some denial and reluctance onthe part of the patient to seek medical care.Unlike adults, gender distribution in children, at

least during the prepubertal years, is approxi-mately equal.31

SnoringClinical findings may occur during both sleepand wakefulness (Table 19–1). It is important toobtain the clinical history from a parent or evena sibling who shares a room and might haveobserved the patient while s/he was sleeping; fewof us are aware of (and admit to) snoring andother nocturnal symptoms. Snoring is usuallythe presenting complaint of patients with OSAS.Up to 20% of children snore intermittently, and7% to 10% of them are habitual or regular snor-ers.1,2 From 1% to 3% of school-aged childrenhave significant disease.1,2 Respiratory pauses,often ending with a snort or gasp, are also oftendescribed. The snoring may be cyclical, worsen-ing with a decrease in motor tone associatedwith rapid eye movement sleep. These symp-toms are often the worst when the patient is inthe supine position, where gravity pulls thegenioglossus to occlude the airway. However,non-obese prepubertal children may breathe bet-ter in the supine position,33 suggesting a differ-ent pathophysiology from that in older childrenand adults.

Not all children who snore have OSAS. Nocombination of historical or physical factors hasbeen useful in distinguishing OSAS from pri-mary snoring. Several studies have attemptedto predict the presence of OSAS in children witha suggestive clinical presentation, including

Obstructive Sleep Apnea Syndrome in Infants and Children Chapter 19 225

Nocturnal

SnoringRespiratory pausesRestless sleepDiaphoresisParadoxical respiratory movementsEnuresis

Diurnal

SleepinessBehavior problemsPoor school performanceMorning headaches

Table 19–1. Symptoms of ObstructiveSleep Apnea in Children

Ch19.qxd 1/25/05 3:25 AM Page 225

snoring, witnessed apneas, and/or daytime som-nolence. However, only one third to one half ofthe children suspected of having OSAS satisfiedpolysomnographic criteria.21,34-36 Therefore, lab-oratory confirmation is essential to confirm thediagnosis and establish the severity of the dis-order so that the appropriate therapy may beinstituted. These studies are best performed in alaboratory where the staff members are experi-enced with children; preparation of the child forstudy and scoring and interpretation of the studymay be different from that for adults. In general,children have fewer clear obstructive events thanadults, and close attention must be paid to theidentification of partial obstructive events, orobstructive hypopneas, in children.37 An apnea/hypopnea index of 5 or greater is consideredabnormal in young adults, whereas normal chil-dren usually do not have more than one event ofobstructive apnea per hour of sleep.38 However,the clinical significance of a small number of res-piratory events, even if statistically abnormal, isnot known. Children are also more likely thanadults to have significant oxyhemoglobin desat-uration with shorter obstructive events, proba-bly due to lower functional residual capacity andsmaller oxygen stores.

Restless SleepRestless sleep may be a sign of frequent arousalsecondary to respiratory pauses. It is often help-ful to inquire about the condition of the bed-clothes in the morning; the only clue to theextent of restless sleep may be disheveled sheetsand blankets. The child may prefer to sleep inthe upright position or with his or her neck hyper-extended to maximize the patency of the upperairway.

EnuresisNocturnal enuresis can occur during any stage ofsleep.39 Several anecdotal reports describe a rela-tionship between OSAS and enuresis in chil-dren.40,41 In one study, 47% of children withmore than one respiratory event per hour ofsleep had nocturnal enuresis, compared with17% of children with no or one respiratory eventper hour (P < .05).42 Enuresis was more preva-lent among boys than girls but was not related toobesity. This may be a function of an insufficient

arousal response,43 impaired urodynamics,44

and/or insufficient vasopressin production dur-ing sleep.45

Obesity or Failure to ThriveAs in adults, obesity may predispose the child toOSAS.20,28-31 However, in contrast to adults, chil-dren with OSAS may fail to thrive. In one report,27% of the children with OSAS were failing tothrive,46 perhaps due to the increased work ofbreathing during sleep.27 After treatment fortheir OSAS, a group of 14 children demonstratedan increase in weight percentile, accompanied bya decrease in sleeping energy expenditure.27

Endocrine factors such as insulin-like growthfactor–1 (IGF-1) may also play a role. Thirteenchildren with OSA showed an improvement inweight and IGF-1 after adenotonsillectomy.47

Conversely, OSA may itself be a risk factor forthe metabolic syndrome, a complex of hyperin-sulinemia/fasting hyperglycemia, abdominalobesity, and dyslipidemia. The severity of OSAcorrelated with fasting insulin levels in a groupof obese children independent of the degree ofobesity.48 Treatment of OSA by continuous posi-tive airway pressure in adults with type 2 dia-betes resulted in a significant improvement intheir insulin responsiveness.49 This may be theresult of catecholamine and/or cortisol releaseassociated with arousal and/or hypoxemia.

Daytime FunctioningFrequent arousal and poor quality of sleep mayresult in daytime sleepiness. Sometimes this isobvious, as in the child who falls asleep in schoolor at the dinner table. It is frequently more sub-tle, being manifested as a short temper, behaviorproblems, or academic difficulties. Childrenwhose parents report snoring and sleep distur-bances were believed by their teachers to be morehyperactive and less attentive than control sub-jects.1 These children may be misdiagnosed asthose with attention deficit hyperactivity disorder.These episodes of daydreaming or inattentive-ness may represent brief “microsleeps” as thechild attempts to repay his or her sleep debt.Children with a lower academic performance inmiddle school are more likely to have snoredduring early childhood than their better perform-ing schoolmates.50 Children with sleep-associated


Ch19.qxd 1/25/05 3:25 AM Page 226

gas exchange abnormalities who were perform-ing poorly in school showed an increase in theirgrades after adenotonsillectomy.51 Deficienciesin cognitive function, when they occur, may bethe result of intermittent hypoxemia52 and/orfrequent arousal, preventing good-quality noc-turnal sleep. Sleepiness in the absence of snoringmay suggest narcolepsy or simply inadequatesleep due to insomnia or schedule problems.Nocturnal asthma and gastroesophageal refluxmay also disturb nocturnal sleep and result inexcessive daytime somnolence.

MortalityThere appears to be an excess mortality in adultswith OSAS. Adults with more than twentyapneic events per hour had a significant increasein mortality over those with fewer than twentysuch events per hour.53 The increased mortalitymay result indirectly from medical risk factorssuch as hypertension54 or an increased risk ofmotor vehicle accidents.55 Comparable data arenot available for children, but pediatric OSAScan be associated with systemic hypertension,56

ventricular dysfunction, and/or cor pulmonale.57,58

In one early report 12 of 22 children had cor pul-monale,46 but this is probably less common now,with greater awareness of the problem resultingin earlier diagnosis and treatment.

SUMMARY

OSA is common in children, resulting from acombination of factors influencing the size andcollapsibility of the pharynx. Children with OSASmay present with nocturnal complaints includ-ing snoring, restless sleep, and enuresis. In thedaytime they may exhibit sleepiness, behaviorproblems, and/or poor school performance.They may be obese or fail to thrive. OSAS mayresult in considerable morbidity and perhapsmortality if left untreated.

References

1. Ali NJ, Pitston DJ, Stradling JR: Snoring, sleepdisturbance, and behaviour in 4-5 year olds. ArchDis Child 1993;68:360-366.

2. Gisalson T, Benediktsdottir B: Snoring, apneicepisodes, and nocturnal hypoxemia among chil-dren 6 months to 6 years old. Chest 1995;107:963-966.

3. Anuntaseree W, Rookkapan K, Kuasirikul S,Thongsuksai P: Snoring and obstructive sleep apneain Thai school-age children. Pediatr Pulmonol2001;32:222-227.

4. Young T, Palta M, Dempsey J, et al: The occur-rence of sleep-disordered breathing amongmiddle-aged adults. N Engl J Med 1993;328:1230-1235.

5. Brooks LJ, Strohl KP: Size and mechanical prop-erties of the pharynx in healthy men and women.Am Rev Respir Dis 1992;146:1394-1397.

6. Gleadhill IC, Schwartz AR, Schubert N, et al:Upper airway collapsibility in snorers and patientswith obstructive hypopnea and apnea. Am RevRespir Dis 1991;144:1300-1303.

7. Marcus CL, McColley SA, Carroll JL, et al: Upperairway collapsibility in children with obstructivesleep apnea syndrome. J Appl Physiol 1994;77:918-924.

8. Hudgel DW, Brooks LJ, Harasick TM: Measure-ments of awake upper airway caliber do not pre-dict upper airway resistance during sleep. Am RevRespir Dis 1989;139:A374.

9. Mezzanotte WS, Tangel DJ, White DP: Wakinggenioglossal electromyogram in sleep apneapatients versus normal controls (a neuromuscularcompensatory mechanism). J Clin Invest 1992;89:1571-1579.

10. Mezzanotte WS, Tangel DJ, White DP: Influenceof sleep onset on upper airway muscle activity inapnea patients versus normal controls. Am JRespir Crit Care Med 1996;153:1880-1887.


12. Kulnis R, Nelson S, Strohl K, Hans M: Cephalo-metric assessment of snoring and nonsnoringchildren. Chest 2000;118:596-603.

13. Pillar G, Lavie P: Assessment of the role of inher-itance in sleep apnea syndrome. Am J Respir CritCare Med 1995;151:688-691.

14. Redline S, Tosteson T, Tishler PV, et al: Studies inthe genetics of obstructive sleep apnea. Am RevRespir Dis 1992;145:440-444.

15. Redline S, Tishler PV, Tosteson TD, et al: Thefamilial aggregation of obstructive sleep apnea.Am J Respir Crit Care Med 1995;151:682-687.

16. Mathur R, Douglas NJ: Family studies in patientswith the sleep apnea-hypopnea syndrome. AnnIntern Med 1995;122:174-178.

17. Guilleminault C, Partinen M, Hollman K, et al:Familial aggregates in obstructive sleep apneasyndrome. Chest 1995;107:1545-1551.

18. Brooks LJ: Genetic syndromes affecting breathingduring sleep in children. In Lee-Chiong TL, SateiaMJ, Carskadon MA (eds): Sleep Medicine.Philadelphia, Hanley & Belfus, 2002, pp 305-314.


Ch19.qxd 1/25/05 3:25 AM Page 227

19. Jeans WD, Fernando DCJ, Maw AR, Leighton BC:A longitudinal study of the growth of thenasopharynx and its contents in normal children.Br J Radiol 1981;54:117-121.

20. Brooks LJ, Stephens B, Bacevice AM: Adenoid sizeis related to severity but not the number ofepisodes of obstructive apnea in children.J Pediatr 1998;132:682-686.

21. Suen JS, Arnold JE, Brooks LJ: Adenotonsil-lectomy for treatment of obstructive sleep apneain children. Arch Otolaryngol Head Neck Surg1995;121:523-530.

22. Fernbach SK, Brouillette RT, Riggs TW, Hunt CE:Radiologic evaluation of adenoids and tonsils inchildren with obstructive apnea: Plain films andfluoroscopy. Pediatr Radiol 1983;13:258-265.

23. Mahboubi S, Marsh RR, Potsic WP, PasquarielloPS: The lateral neck radiograph in adenotonsillarhyperplasia. Int J Pediatr Otorhinolaryngol1985;10:67-73.

24. Laurikainen E, Erkinjuntti M, Alihanka J, et al:Radiological parameters of the bony nasopharynxand the adenotonsillar size compared with sleepapnea episodes in children. Int J Pediatr Otorhino-laryngol 1987;12:303-310.

25. Gulleminault C, Partinen M, Praud JP, et al:Morphometric facial changes and obstructivesleep apnea in adolescents. J Pediatr 1989;114:997-999.

26. Horner RL, Mohiaddin RH, Lowell DG, et al: Sitesand sizes of fat deposits around the pharynx inobese patients with obstructive sleep apnoeaand weight matched controls. Eur Respir J1989;2:613-622.


28. Mallory GB, Fiser DH, Jackson R: Sleep-associ-ated breathing disorders in morbidly obesechildren and adolescents. J Pediatr 1989;115:892-897.

29. Silvestri JM, Weese-Mayer DE, Bass MT, et al:Polysomnography in obese children with a his-tory of sleep-associated breathing disorders.Pediatr Pulmonol 1993;16:124-129.

30. Marcus CL, Curtis S, Koerner CB, et al:Evaluation of pulmonary function and polysom-nography in obese children and adolescents.Pediatr Pulmonol 1996;21:176-183.

31. Redline S, Tishler PV, Schluchter M, et al: Riskfactors for sleep disordered breathing in children.Am J Respir Crit Care Med 1999;159:1527-1532.

32. Marcus CL, Lutz J, Hamer A, et al: Developmentalchanges in response to subatmospheric pressureloading of the upper airway. J Appl Physiol1999;87:626-633.

33. Fernandes do Prado LB, Li X, Thompson R,Marcus CL: Body position and obstructive sleepapnea in children. Sleep. 2002;25:66-71.

34. Wang RC, Elkins TP, Keech D, et al: Accuracy ofclinical evaluation in pediatric obstructive sleepapnea. Otolaryngol Head Neck Surg 1998;118:69-73.

35. Carroll JL, McColley SA, Marcus CL, et al:Inability of clinical history to distinguish primarysnoring from obstructive sleep apnea syndrome inchildren. Chest 1995;108:610-618.

36. Leach J, Olson J, Herman J, Manning S: Poly-somnographic and clinical findings in childrenwith obstructive sleep apnea. Arch OtolaryngolHead Neck Surg 1992;118:741-744.

37. Rosen CL, D’Andrea L, Haddad GG: Adult criteriafor obstructive sleep apnea do not identify chil-dren with serious obstruction. Am Rev Respir Dis1992;146:1231-1234.

38. Marcus CL, Omlin KJ, Basinki DJ, et al: Normalpolysomnographic values for children and adoles-cents. Am Rev Respir Dis 1992;146:1235-1239.

39. Bader G, Neveus T, Kruse S, Sillen U: Sleep of pri-mary enuretic children and controls. Sleep2002;25:579-583.

40. Sakai J, Herbert F: Secondary enuresis associatedwith obstructive sleep apnea (letter). J Am AcadChild Adolesc Psychiatry 2000;39:140-141.

41. Timms DJ: Rapid maxillary expansion in thetreatment of nocturnal enuresis. Angle Orthod1990;60:229-233.

42. Brooks LJ, Topol HI: Enuresis in children withsleep apnea. J Pediatr 2003;142:515-518.

43. Berry RB, Kouchi KG, Der DE, et al: Sleep apneaimpairs the arousal response to airway occlusion.Chest 1996;109:1490-1496.

44. Yokoyama O, Amano T, Lee S, et al: Enuresis in anadult female with obstructive sleep apnea.Urology 1995;45:150-154.

45. Lin C, Tsan K, Lin C: Plasma levels of atrial natri-uretic factor in moderate to severe obstructivesleep apnea syndrome. Sleep 1993;16:37-39.

46. Brouillette RT, Fernbach SK, Hunt CE: Obstructivesleep apnea in infants and children. J Pediatr1982;100:31-40.

47. Bar A, Tarasiuk A, Segev Y, et al: The effect of ade-notonsillectomy on serum insulin-like growthfactor-1 and growth in children with obstructivesleep apnea syndrome. J Pediatr 1999;135:76-80.

48. de la Eva RC, Baur LA, Donaghue KC, Waters KA:Metabolic correlates with obstructive sleep apneain obese subjects. J Pediatr 2002;140:654-659.

49. Brooks B, Cistulli PA, Borkman M, et al:Obstructive sleep apnea in obese noninsulin-dependent diabetic patients: Effect of continuouspositive airway pressure on insulin responsive-ness. J Clin Endocrinol Metab 1994;79:1681-1685.


Ch19.qxd 1/25/05 3:25 AM Page 228

50. Gozal D, Pope DW: Snoring during early child-hood and academic performance at ages thirteento fourteen years. Pediatrics 2001;107:1394-1399.


52. Findley LJ, Barth JT, Powers DC, et al: Cognitiveimpairment in patients with obstructive sleep apneaand associated hypoxemia. Chest 1986;90:686-690.

53. He J, Kryger MH, Zorick FJ, et al: Mortality andapnea index in obstructive sleep apnea:Experience in 385 male patients. Chest 1988;94:9-14.

54. Lavie P, Herer P, Peled R, et al: Mortality in sleepapnea patients: A multivariate analysis of risk fac-tors. Sleep 1995;18:149-157.

55. Teran-Santos J, Jimenez-Gomez A, Cordera-Guevara J: The association between sleep apneaand the risk of traffic accidents. N Engl J Med1999;340:847-851.


57. Tal A, Lieberman A, Margulis G, Sofer S:Ventricular dysfunction in children with obstruc-tive sleep apnea: Radionuclide assessment. PediatrPulmonol 1988;4:139-143.

58. Sofer S, Weinhouse E, Tal A, et al: Cor pulmonaledue to adenoidal or tonsillar hypertrophy or bothin children. Chest 1988;93:119-122.


Ch19.qxd 1/25/05 3:25 AM Page 229

Several anecdotal reports suggest the likelihoodof a relationship between nocturnal enuresis andobstructive sleep apnea (OSA) in both adults1-6

and children.7-10 This is further elucidated in caseseries. A questionnaire study of children inIstanbul showed an increased prevalence of noc-turnal enuresis in habitual snorers comparedwith occasional snorers and nonsnorers.11 Cinarand colleagues12 noted that 35% of childrenoperated on for upper airway obstructionreported nocturnal enuresis, and two thirds ofthose who responded reported complete or par-tial relief of enuresis 3 months after surgery.Weider and coworkers13 described a series of115 children with upper airway obstruction andenuresis. Six months after “some type of surgeryto eliminate upper airway obstruction,” thegroup reported a 77% reduction in the numberof enuretic nights. However, neither of thesereports included a control group and very few ofthe patients had polysomnography to define andquantify their obstructive sleep apnea. A largepopulation study of home polysomnogramsin Hispanic and Caucasian children aged 6 to 11years found a prevalence of enuresis of 11.3% inchildren with a respiratory disturbance index (RDI;apneas plus hypopneas per hour of sleep) ofgreater than 1 compared with a prevalence of 6.3%in children with an RDI of less than 1 (P < .08).Enuresis was strongly associated with loudsnoring (P < .007), a known correlate of sleep-disordered breathing (SDB) in children.14

We described 160 pediatric patients (90 boysand 70 girls) who were referred to our sleep cen-ter with signs and symptoms suggestive of OSA15

(Table 20–1). The mean age of the children was9.6 ± 3.58 years (range, 4.2 to 17.9). There wasno relationship between RDI and age (rS = −.15,P > .20). A total of 66 children (41%) currentlydescribed enuresis, a percentage similar to thatfound in the data of Cinar et al.12 Forty-nine per-

cent of the boys were enuretic compared with31% of the girls (P < .05). There was no differ-ence in RDI between the boys and girls (P = .91).The mean body mass index (BMI; kg/m2) was22.6 ± 8.71 (range, 12.9 to 64.2). There was norelationship between BMI and RDI (rS = .09, P > .5)or the presence of enuresis (X2 = 3.08, P = 0.55).

Enuresis was more prevalent in the youngerchildren, but at all ages enuresis was more preva-lent than in literature controls.16,17 Of the childrenwho wet the bed, 64% had primary enuresis and36% had secondary enuresis. Children with anRDI of less than 1 had a significantly lowerprevalence of nocturnal enuresis (27%) than didchildren with an RDI of greater than 1 (47%) (X2

= 4.13, P < .05). Of children with an RDI of lessthan 1, 14% had enuresis frequently, comparedwith 32% of children with an RDI of greater than1 (X2 = 4.52, P < .05). There was no significantdifference in the prevalence of enuresis betweenchildren with an RDI between 1 and 5, those withan RDI between 5 and 15, and those with an RDIgreater than 15 (X2 = .18, P = .92) (Figure 20–1).

These data demonstrate a high prevalence ofnocturnal enuresis in children with suspectedSDB. Children with more than one respiratoryevent per hour of sleep (RDI > 1) were at signif-icantly greater risk of enuresis than were chil-dren with an RDI of less than 1. In other studies,more than one obstructive apnea per hour ofsleep has been suggested to indicate OSA inchildren.18

The predisposition of boys to nocturnalenuresis, and the 3-to-1 ratio of primary to sec-ondary enuresis are similar to descriptions ofotherwise-normal children.19 Wang and col-leagues found nocturnal enuresis in 24 of 82children (29%) referred for suspected OSA.20

Eleven of those 24 had OSA confirmed withpolysomnography, but no data were reported onthe severity of the OSA. Brouillette and coworkers

Enuresis in Children with Sleep Apnea

Lee J. Brooks

20

231

Ch20.qxd 1/25/05 3:26 AM Page 231

found nocturnal enuresis in 8% of children withOSA compared to 4% of controls.21 However,this was not a statistically significant difference,probably because of the small number of patients.

Finally, treatment of OSA with continuouspositive airway pressure, dental devices, or ade-notonsillectomy has been reported to ameliorateenuresis in adults2-6 and children.7-10,13,22

Why might children with SDB be at risk fornocturnal enuresis? Proposed etiologies for noc-turnal enuresis include insufficient arousalresponse, impaired urodynamics, and insuffi-cient vasopressin production during sleep.

Children who wet the bed are reported bytheir parents to be more difficult to arouse fromsleep than children without enuresis.23 Weiderand colleagues reported that “a number ofpatients [with upper airway obstruction] whocould not be aroused before surgery were able towake up and go to the bathroom by themselvesafter surgery.”13 SDB may therefore promoteenuresis by decreasing the arousal response,24

perhaps because of sleep fragmentation.24-26

Daytime urodynamics are similar in childrenwith and those without enuresis.27 However, theincreased intra-abdominal pressure caused byrespiratory efforts against an obstructed airwaycan be transmitted to the bladder. Cystometrog-raphy has demonstrated that bladder pressureduring sleep may rise as high as 60 cm H20in conjunction with respiratory efforts against anobstructed upper airway, in contrast to pressuresof about 5 cm H20 during quiet wakefulness.1

Thus, OSA may promote enuresis by increasingbladder pressure.

Finally, SDB can affect the secretion of urinaryhormones such as atrial natriuretic peptide

(ANP) and antidiuretic hormone (ADH). Adultswith OSA have elevated ANP and decreasedADH,28-30 and the normal nocturnal decreasein urine output does not occur.29-31 In patientswith OSA, plasma ANP correlates negatively withcumulative apnea duration and lowest arterialoxygen saturation, and positively with the high-est change in intrathoracic pressure.28 This mayresult from the more negative intrathoracic pres-sures increasing atrial volume, thus stimulatingANP production. Acute hypoxemia may resultfrom central or obstructive respiratory events,and it may stimulate ANP secretion.32 Urine out-put, sodium excretion, and secretion of ANP alldecrease, whereas renin and aldosterone increase,with successful treatment of OSA.28,30,31,33,34

Many of these studies describe patientsreferred for suspected SDB, and it is possible thatnocturnal enuresis was a factor in the primaryphysician’s decision to refer, thus inflating thereported prevalence of enuresis. Despite this, theevidence suggests that the association betweennocturnal enuresis and OSA is real, with a rea-sonable physiologic basis.

Obstructive sleep apnea, which can usually beeasily treated, should be considered in the differ-ential diagnosis of nocturnal enuresis in children.

232 Chapter 20 Enuresis in Children with Sleep Apnea

Boys GirlsN 90 70With enuresis (%) 49 31*

Age (yr) 9.7 ± 3.6† 9.4 ± 3.6BMI (kg/m2) 22.2 ± 7.9 24.0 ± 9.8RDI 11.3 ± 18.8 8.3 ± 13.4

*P < .05†Mean ± SDBMI, body mass index; RDI, respiratory disturbance

index.

Table 20–1. Patient Demographics

% o

f sub

ject

s

60

50

40

30

20

10

0�1 �151.1–5 5.1–15

RDI

44 46 42 28

*P � .05 compared to children with RDI�1

Any enuresisFrequent enuresis

*

*

Figure 20–1. Prevalence of nocturnal enuresis in 160children referred for suspected sleep-disordered breath-ing. Frequent enuresis was defined as wetting the bed3 nights a week or more. (From Brooks LJ, Topol HI:Enuresis in children with sleep apnea. J Pediatr 2003;142:515-518.)

Ch20.qxd 1/25/05 3:27 AM Page 232

References

1. Yokoyama O, Amano T, Lee S, et al: Enuresis in anadult female with obstructive sleep apnea. Urology1995;45:150-154.

2. Arai H, Furuta H, Kosaka K, et al: Polysomno-graphic and urodynamic changes in a case ofobstructive sleep apnea with enuresis. PsychiatryClin Neurosci 1999;55:319-320.

3. Everaert K, Pevernagie D, Oosterlinck W: Noctur-nal enuresis provoked by an obstructive sleep apneasyndrome. J Urol 1995;153:1236.

4. Ulfberg J, Thuman R: A non-urologic causeof nocturia and enuresis: Obstructive sleep apneasyndrome. Scand J Urol Nephrol 1996;30:135-137.

5. Steers WD, Suratt P: Sleep apnoea as a cause ofdaytime and nocturnal enuresis. Lancet 1997;349:1604.

6. Kramer NR, Bonitati AE, Millman RF: Enuresisand obstructive sleep apnea in adults. Chest 1998;114:634-637.

7. Wengraf C: Management of enuresis (letter). Lancet1997;350:221-222.

8. Simmons FB, Guilleminault C, Dement WC, et al:Surgical management of airway obstructions dur-ing sleep. Laryngoscope 1977;87:326-338.

9. Timms DJ: Rapid maxillary expansion in thetreatment of nocturnal enuresis. Angle Orthod1990;60:229-233.

10. Sakai J, Herbert F: Secondary enuresis associatedwith obstructive sleep apnea (letter). J Am AcadChild Adolesc Psychiatry 2000;39:140-141.

11. Ersu R, Arman AR, Save D, et al: Prevalence ofsnoring and symptoms of sleep-disordered breath-ing in primary school children in Istanbul. Chest2004;126:19-24.

12. Cinar U, Vural C, Cakir B, et al: Nocturnal enure-sis and upper airway obstruction. Int J PediatrOtorhinolaryngol 2001;59:115-118.

13. Weider DJ, Sateia MJ, West RP: Nocturnal enure-sis in children with upper airway obstruction.Otolaryngol Head Neck Surg 1991;105:427-432.

14. Goodwin JL, Kaemingk KL, Fregosi RF, et al:Parasomnias and sleep disordered breathing inCaucasian and Hispanic children: The Tucsonchildren’s assessment of sleep apnea study. BMCMed 2004;2:14.

15. Brooks LJ, Topol HI: Enuresis in children withsleep apnea. J Pediatr 2003;142:515-518.

16. Oppel WC, Harper PA, Rider RV: The age of attain-ing bladder control. Pediatrics 1968;42:614-626.

17. Sheldon SH, Spire JP, Levy HB: Sleep-related enure-sis. In Sheldon SH, Spire JP, Levy HB (eds): PediatricSleep Medicine. Philadelphia, WB Saunders, 1992,pp 151-166.


19. Rushton HG: Wetting and functional voiding dis-orders. Urol Clin North Am 1995;22:75-93.

20. Wang RC, Elkins TP, Keech D, et al: Accuracyof clinical evaluation in pediatric obstructivesleep apnea. Otolaryngol Head Neck Surg 1998;118:69-73.


22. Wang RC, Vordemark J: Effects of sleep apneatreatment upon enuresis in children. Surg Forum1994;45:643-644.

23. Neveus T, Hetta J, Cnattingius S, et al: Depth ofsleep and sleep habits among enuretic and incon-tinent children. Acta Paediatr 1999;88:748-752.

24. Berry RB, Kouchi KG, Der DE, et al: Sleep apneaimpairs the arousal response to airway occlusion.Chest 1996;109:1490-1496.

25. Phillipson EA, Bowes G, Sullivan CE, Woolf, GM:The influence of sleep fragmentation on arousaland ventilatory responses to respiratory stimuli.Sleep 1980;3:281-288.

26. Downey B, Bonnet MH: Performance during fre-quent sleep disruption. Sleep 1987;10:354-363.

27. Norgaard JP, Rittig S, Djurhuus JC: Nocturnalenuresis: An approach to treatment based onpathogenesis. J Pediatr 1989;114:705-710.

28. Krieger J, Follenius M, Sforza E, et al: Effects oftreatment with nasal continuous positive airwaypressure on atrial natriuretic peptide and argininevasopressin release during sleep in patients withobstructive sleep apnea. Clin Sci 1991;80:443-449.

29. Ichioka M, Hirata Y, Inase N, et al: Changes of cir-culating atrial natriuretic peptide and antidiuretichormone in obstructive sleep apnea syndrome.Respiration 1992;59:164-168.

30. Lin C, Tsan K, Lin C: Plasma levels of atrial natri-uretic factor in moderate to severe obstructivesleep apnea syndrome. Sleep 1993;16:37-39.

31. Rodenstein DO, D’Odemont JP, Pieters T, Aubert-Tulkens G: Diurnal and nocturnal diuresis andnatriuresis in obstructive sleep apnea. Am RevRespir Dis 1992;145:1367-1371.

32. Baertschi AJ, Adams JM, Sullivan MP: Acute hypox-emia stimulates atrial natriuretic factor secretionin vivo. Am J Physiol 1988;255:H295-H300.

33. Baruzzi A, Riva R, Cirignotta F, et al: Atrial natri-uretic peptide and catecholamines in obstructivesleep apnea syndrome. Sleep 1991;14:83-86.

34. Follenius M, Krieger J, Krauth MO, et al:Obstructive sleep apnea treatment: Peripheral andcentral effects on plasma renin activity and aldos-terone. Sleep 1991:14:211-217.

Enuresis in Children with Sleep Apnea Chapter 20 233

Ch20.qxd 1/25/05 3:27 AM Page 233

Childhood obstructive sleep apnea syndrome(OSAS) is a common condition that can result insevere complications if not diagnosed andtreated in a timely fashion. Nevertheless, mostpatients respond readily to treatment. The main-stay of treatment is tonsillectomy and ade-noidectomy (T&A). Nasal continuous positiveairway pressure (CPAP) is usually the second-line treatment for children who do not respondto T&A, or for those few children for whomT&A is not indicated. In selected cases, othertreatment options may be useful.

WHEN TO TREAT

One of the biggest dilemmas in pediatric sleepmedicine is the question of who warrants treat-ment. Very few outcome studies have been con-ducted in children. Some normative pediatricdata are available.1-4 However, these data tell usonly what is statistically abnormal. We do notyet know which clinical or polysomnographicparameters predict a poor outcome—that is,which parameters are clinically important. It isgenerally accepted that children with severeOSAS should always be treated, as these childrenare at risk for serious complications such as corpulmonale and failure to thrive. Many practi-tioners agree that children with primary snoringshould not be subjected to surgery, although thisgroup has not been rigorously studied. However,treatment of the child with mild abnormalitieson polysomnography is controversial. In suchcases, as with most medical conditions, treat-ment decisions should be based on the constel-lation of symptoms, physical examination, andlaboratory results (polysomnography), ratherthan on polysomnography results alone. If thedecision is made not to treat the child, thenfollow-up is important.

EMERGENCY TREATMENT

Most patients with diagnosed OSAS can awaitelective treatment. On a practical level, thesechildren have often waited weeks or monthsbefore seeing a physician or undergoing a sleepstudy, and the decision is therefore made thatthey can wait for surgery. Although these patientsdo not seem to have acute problems, it should berealized that the short-term effects of untreatedhypoxemia or obstructive apnea are unknown.

Occasionally, emergency treatment is war-ranted in children presenting with cardiorespira-tory failure or severe hypoxemia. Many childrencan be successfully managed with nasopharyn-geal tubes, pending definitive surgery,5 or withCPAP.6 Intubation is rarely needed. Sedationshould be avoided, and supplemental oxygenshould never be administered without carefulmonitoring for hypoventilation (see later).

TREATMENT MODALITIES

Tonsillectomy and AdenoidectomyT&A is the first-line treatment for childhoodOSAS. It is a relatively simple procedure with alow complication rate, and it usually results in acure. Therefore, it is preferable to long-term useof CPAP, where compliance is often a problem.Nevertheless, T&A should not be undertakenlightly. As with any surgery, there is associatedmorbidity and even mortality. For these reasons,it is recommended that surgery be performedonly after polysomnographic proof of OSAS hasbeen obtained.7,8

Efficacy of T&A

OSAS results from the relative size and structure ofthe upper airway components, and the relationship

Treatment of Obstructive SleepApnea Syndrome in Children

Carole L. Marcus

21

235

Ch21.qxd 1/25/05 3:28 AM Page 235

between structural narrowing and neuromotortone of the upper airway, rather than from theabsolute size of the tonsils and adenoid. Thus, evenin children with relatively small tonsils, or withadditional risk factors for OSAS such as cranio-facial syndromes, T&A is often successful in treat-ing OSAS. This has been demonstrated in childrenwith morbid obesity,9,10 Down syndrome,11 andcerebral palsy.12 However, these patients should bereevaluated postoperatively to ensure that theOSAS has resolved and that further treatment isnot necessary.

Several studies have evaluated the success ratefor T&A in children with OSAS secondary to ade-notonsillar hypertrophy. These studies are hard tocompare directly, as they utilized different patientpopulations, recording techniques, definitions ofrespiratory events, time periods for postoperativeevaluations, and types of surgery (e.g., some ofthe studies included patients who underwent ade-noidectomy with monotonsillectomy). Never-theless, the different studies all show that theoverwhelming majority of patients have animprovement in both symptoms and polysomno-graphic findings postoperatively.10,13-21 The mostcomplete data are given by Suen and colleagues.19

They studied 26 children with OSAS but withoutneurologic or craniofacial anomalies. Patientsunderwent follow-up polysomnography 6 to50 weeks after T&A (personal communicationfrom the authors), and showed a decrease in res-piratory distress index (RDI) from 16 ± 10/hr to 2± 1/hr. Four patients continued to have an RDI ofgreater than 5 postoperatively, with one still hav-ing an RDI of 45 per hour. The preoperative RDIwas a predictor of the response to surgery. Frankand coworkers18 evaluated seven children 4 to6 weeks after T&A and found that the averagetotal number of apneas fell from 194 to 7. Bar andcolleagues22 studied 10 normal children an aver-age of 3 months after T&A and found a decreasein the RDI from 8 ± 9 to 1 ± 2. Zucconi andcoworkers21 evaluated 14 patients with nap orovernight polysomnography 3 to 18 months post-operatively (surgery being adenoidectomy, ade-noidectomy with monotonsillectomy, or T&A)and showed a decrease in the apnea-hypopneaindex from 11 ± 10 to 3 ± 2/hr. Agren and col-leagues16 reported normalization of the oxygendesaturation index 1 year postoperatively.

Studies have shown that growth improvesafter T&A, even in children who were in the

normal range beforehand.20,22-25 In fact, evenobese children gain weight postoperatively.26

Enuresis has been reported to improve aftersurgery.27 Studies of behavioral and cognitivefunction have, in general, shown an improve-ment after T&A. Ali and coworkers28 found animprovement in the Conners Behaviour Scale,Matching Familiar Figures Test, and a continu-ous performance test postoperatively; childrenwith snoring showed a lesser response afterT&A, and controls (studied after a 6-monthtime period) did not change. Goldstein and col-leagues29 showed an improvement measured onthe Child Behavior Checklist (controls were notevaluated). Gozal showed an improvement inacademic scores after T&A in children withsleep-disordered breathing but no change ineither normal controls or untreated childrenwith sleep-disordered breathing.30 In contrastto these studies, a study of behavioral symp-toms after treatment, compared with an un-treated control group, showed an improve-ment in daytime symptoms but no significantchange in temperament or intelligence aftersurgery.31

Tonsillectomy or Adenoidectomy, or Both?

Many sleep specialists believe that both the ton-sils and the adenoid should be removed, even ifone or the other appears to be the more enlarged.This results from the belief that OSAS is causedby both structural and neuromotor abnormali-ties, and that therefore widening the airway asmuch as possible is desirable. There is alsoanecdotal evidence of a high persistence or re-currence rate in children treated with adenoid-ectomy or tonsillectomy alone. However, thereare no randomized, controlled trials evaluatingthe different types of surgery. Nieminen andcoworkers17 noted that 73% of children withOSAS had a history of prior adenoidectomy; allimproved after tonsillectomy. This suggestseither that adenoidectomy alone was insufficientor that the adenoids grew back. Zucconi and col-leagues21 reported that three children whounderwent adenoidectomy alone had only a tran-sient improvement in OSAS, whereas 16 childrenwho underwent T&A or adenoidectomy withmonotonsillectomy did better. However, furtherdetails were not provided. In children with a sub-mucous cleft palate, adenoidectomy may result in

236 Chapter 21 Treatment of Obstructive Sleep Apnea Syndrome in Children

Ch21.qxd 1/25/05 3:28 AM Page 236

velopharyngeal incompetence; in this group, ton-sillectomy alone is a consideration.

Adenoidal recurrence may occur, especially invery young children. One study evaluated ade-noidal regrowth in a sample of children who hadsymptoms of nasal obstruction 2 to 5 years afteradenoidectomy.32 One patient had alreadyundergone revision adenoidectomy. In theremaining patients, the authors reported lym-phoid regrowth occupying 0% to 40% of thenasopharynx. Thus, if children develop a recur-rence of OSAS after T&A, they should be evalu-ated to determine whether the adenoids haveregrown and revision adenoidectomy is required.

Morbidity and Mortality

Acute general complications of T&A includehemorrhage (in approximately 3% of cases),33 res-piratory decompensation, anesthetic complica-tions, pain, and poor oral intake. Death has beenreported.34 Preoperative sedation may precipitateacute upper airway obstruction.35 Postoperativerespiratory decompensation may result from tran-sient worsening of OSAS secondary to postopera-tive edema and increased secretions; respiratorydepression from anesthetic agents, narcotics, andthe use of oxygen in children with a bluntedhypoxic ventilatory drive; and the occurrence ofpostobstructive pulmonary edema.36,37 The latteris a poorly understood phenomenon occurringwith the relief of upper airway obstruction fromany cause (such as croup, epiglottitis, or OSAS).Long-term complications of T&A, such asnasopharyngeal stenosis and velopharyngealincompetence, occasionally occur.

A number of studies have evaluated theprevalence of postoperative respiratory compli-cations. In patients with polysomnographicallyproven OSAS, prevalence rates of 16% to 27%have been reported.6,38,39 Prevalence rates inthose studies in which polysomnography wasnot performed are lower.40,41 These studies prob-ably included children with primary snoring aswell as OSAS. Studies of postoperative complica-tions have been remarkably consistent in theirfindings about which patients are at greatest riskfor postoperative respiratory compromise. High-risk groups include children less than 3 years ofage, children with underlying disease (craniofa-cial anomalies, obesity, cerebral palsy, or a his-tory of prematurity), those with severe OSAS,and those with failure to thrive or cor pulmonale(both of which are associated with severeOSAS).6,38,39,42-44 High-risk patients can never-theless be treated safely if appropriate precau-tions are taken.45 Recommendations includeperforming surgery in tertiary care centers wherepediatric specialists and a pediatric intensivecare unit are available, avoiding preoperativesedation, ensuring intravenous access prior toinducing anesthesia, and monitoring patientsclosely postoperatively, so that appropriate air-way intervention can be performed as needed.Currently in the United States, T&A is usuallyperformed as outpatient surgery. The AmericanAcademy of Pediatrics Clinical PracticeGuideline7 recommends that high-risk patients(Table 21–1) be hospitalized overnight after sur-gery and monitored continuously with pulseoximetry. CPAP can be used in the perioperative

Treatment of Obstructive Sleep Apnea Syndrome in Children Chapter 21 237

Age younger than 3 yearsSevere obstructive sleep apnea syndrome (OSAS) on polysomnographyCardiac complications of OSAS (e.g., right ventricular hypertrophy)Failure to thriveObesityPrematurityRecent respiratory infectionCraniofacial anomaliesNeuromuscular disorders

Reproduced with permission from American Academy of Pediatrics, Section on Pediatric Pulmonology, Sub-committee on Obstructive Sleep Apnea: Clinical practice guideline: Diagnosis and management of childhoodobstructive sleep apnea syndrome. Pediatrics 2002;109:704-712.

Table 21–1. Risk Factors for Postoperative Respiratory Complications in Childrenwith OSAS Undergoing Adenotonsillectomy

Ch21.qxd 1/25/05 3:28 AM Page 237

period to stabilize patients prior to T&A and totreat postoperative complications.6,46 It mayreduce the need for postoperative intubation.

Follow-Up

All children should be reevaluated clinicallyafter surgery. Those with severe OSAS preopera-tively, continued risk factors for OSAS, or per-sistent symptoms should be considered forrepeat polysomnography. Traditionally, it isstated that OSAS improves 6 to 8 weeks afterT&A. However, this has not been studied sys-tematically. Significant improvement may occurmuch earlier. Helfaer and colleagues performedrepeat polysomnography the night after surgeryon a group of children with OSAS.47 This studywas limited to children with mild to moderatedisease and no underlying medical conditions.Postoperatively, patients showed an immediatereduction in the apnea-hypopnea index (from5 ± 1 to 2 ± 1) and an improvement in arterialoxygen saturation during rapid eye movementsleep (from 78 ± 5% to 92 ± 1%).

Continuous Positive Airway PressureThe vast majority of children with OSAS will becured by T&A. However, a small number ofpatients will require further treatment. Typically,these patients have underlying medical condi-tions, such as obesity, craniofacial anomalies, orneuromuscular disease. However, occasionalpatients have idiopathic OSAS that persists afterT&A despite the absence of any obvious risk fac-tors.48 In addition, some patients are not candi-dates for T&A, such as adolescents with minimaladenotonsillar tissue, or patients with severebleeding disorders. In these patients, nasal CPAPcan be a highly effective means of treatment.

Method of Action

During CPAP use, positive pressure is generatedby a blower attached to a mask. For bilevel pres-sure, the inspiratory (IPAP) and expiratory(EPAP) pressures can be adjusted independently,so that the expiratory pressure is lower than theinspiratory pressure. Unlike conventional venti-lators, the respiratory circuit has a single limb forboth inspiration and expiration. A fixed leak near

the patient’s mask prevents CO2 rebreathing.CPAP increases the intraluminal upper airwaypressure and elevates it above the upper airwaycritical closing pressure,49 thereby stenting theairway open. It reduces upper airway50 and dia-phragmatic51 inspiratory muscle activity. Becausethe upper airway is rich in sensory nerve end-ings, increased nasal airflow from CPAP can stim-ulate ventilation.52 The use of nasal CPAP inadults with OSAS has been shown to result inimproved cognitive and psychiatric function,decreased daytime somnolence, decreased wak-ing hypercapnia, improved cardiovascular func-tion, and possibly decreased mortality.53

Efficacy of CPAP in Children

In the United States, CPAP use is not approved forchildren weighing less than 30 kg. Nevertheless,there is widespread experience in the use of CPAPin children of all ages, including infants,54,55

showing that it is safe and effective. Several largeseries have been published. One study reportedCPAP use in 80 patients at the same institution,46

and a multicenter study reported on CPAP use in94 patients.48 Both studies showed that CPAP waseffective and tolerated in more than 80% ofpatients with OSAS, including children with cran-iofacial anomalies, obesity, or neurologic disor-ders. The required pressure varied from 4 to 20cm H2O, and this range was independent of ageand underlying diagnosis. Presumably, the pres-sure level was proportional to the severity ofOSAS, although different protocols at the varyinginstitutions did not allow this to be evaluated sys-tematically. Excluding patients in whom CPAPwas only used perioperatively, or in whom addi-tional treatment modalities were used, only 4% to16% of patients improved sufficiently to discon-tinue CPAP over the several years of follow-up.46,48 Long-term studies are needed.

Nasal CPAP has long been known to be effec-tive in treating central apnea in infants; recentstudies show that it is also effective in treatingobstructive apnea in this age group.54-56 Infantsare more likely to be weaned off CPAP than olderchildren.54,55 Unfortunately, nasal masks are notcommercially available in the United States forthis age group, although an infant mask(ResMed, Sydney, Australia) is available in othercountries.


Ch21.qxd 1/25/05 3:28 AM Page 238

Institution of CPAP Therapy

Although CPAP is effective in children, itrequires time and patience on the part of boththe family and the medical practitioner to get itto work. Several centers46,57 begin CPAP at homeon low pressures, to enable the child to graduallyhabituate to the system prior to bringing patientsinto the lab for a formal titration. It is importantto make the CPAP use part of a child’s bedtimeroutine. The alternative, placing the mask on ayoung child’s face while the child is asleep, canresult in frightened arousals during the nightand leads to poor tolerance. Behavioral condi-tioning may be useful.58,59 In children, pressurerequirements change over time.48,54 Therefore, itis necessary to frequently reevaluate pressurelevels in response to growth. In addition, it isimportant to pay attention to mask and headgearfit as time passes.

Patients with OSAS whose CPAP was discon-tinued for 1 night may have a temporaryimprovement in breathing compared to theirbaseline, pre-CPAP treatment state. Polysom-nography on the first night off CPAP demon-strates fewer and shorter obstructive apneas,decreased esophageal pressure swings, and lessarterial oxygen desaturation.60-62 This may resultfrom a number of factors, including decreasedsleep fragmentation, decreased upper airwayedema, and resetting of the ventilatory drive.Because of this transient improvement, CPAPshould be discontinued for at least several daysprior to polysomnography in patients being eval-uated for discontinuation of therapy.

Compliance

The major problem with CPAP use is poor com-pliance. A number of studies in adults haveshown that approximately 15% to 20% ofpatients drop out immediately.63,64 The remain-ing patients use CPAP for an average of 5 hoursa night.63-65 Comparable studies have not beenperformed in children, although a preliminaryreport suggests that their compliance may besimilar.66 It is recommended that compliancebe assessed objectively by regular readings of theequipment hour meter. Compliance can beimproved by increasing the comfort of the equip-ment and paying close attention to side effects.A ramp function that allows the machine to

gradually increase pressure as the patient fallsasleep may increase comfort. Bilevel pressuremay be more comfortable than CPAP, althoughthis requires further study in children. Only onestudy (in adults) has directly compared CPAP tobilevel ventilation for OSAS.63 This study did notdemonstrate a significant difference in effectiveuse between the two methods. However, theCPAP group had a significantly larger dropoutrate than the bilevel group. Furthermore, com-pliance in the bilevel group was affected by thedifference between inspiratory and expiratorypressures, suggesting that different ventilationprotocols may affect compliance. In practice,bilevel ventilation is often used for patients whofind CPAP uncomfortable.

Side Effects

The most common side effects of CPAP use arenasal symptoms, including nasal congestion, rhi-norrhea, dryness, and epistaxis (Table 21–2).These are caused by the cooling and dryingeffects of CPAP, which alters the nasal mucosaand impedes mucociliary clearance.67 In onestudy, more than two thirds of adults using CPAPcomplained of nasal symptoms.68 The most effec-tive treatment is use of a heated humidifier; coolhumidification is little better than placebo.69

Attention to hygiene (cleaning the equipmentand changing filters regularly) is important.Nasal steroids or saline drops may be efficaciousin some patients. Other side effects include dryor irritated eyes resulting from mask leak.Dermatitis and pressure lesions from the maskmay occur. This can be severe and can lead todecubitus ulcers, particularly in patients who areunable to adjust the mask themselves (e.g., thosewith cerebral palsy). Pressure lesions can be min-imized by ensuring good mask fit, avoiding over-tightening of the mask straps, using gel masks,placing spacers on the forehead rather than onthe bridge of the nose, using skin hydrocolloiddressings, or alternating different types of masks.Gastric distension occurs occasionally but is usu-ally self-limiting. More severe side effects, such aspneumothorax70 or barotrauma affecting theeyes,71 ears, and central nervous system,72 havebeen reported very rarely.

Some considerations are specific to the use ofCPAP in children. Central apneas can occur at


Ch21.qxd 1/25/05 3:28 AM Page 239

higher pressure levels46 because of an activeHering-Breuer reflex. This can be remedied byplacing the patient on bilevel ventilation with abackup rate. Children may have problems trigger-ing bilevel machines, as the triggering algorithmsused are not suitable for the lower flow rate andfaster respiratory rate seen in young or weak chil-dren. This must be taken into account whenadjusting settings. Certain brands of equipmenttrigger better in young children than others.Mouth breathing during CPAP use is not a prob-lem unless it interferes with ventilation, in whichcase a chin strap or full face mask can be used.However, there may be a risk of aspiration withthe use of a full face mask, particularly in veryyoung or developmentally impaired children.Occasionally, nasal damage (with nasal prongs)73

or midfacial depression (with nasal masks)74

occur in children who begin using CPAP at ayoung age.

UvulopalatopharyngoplastyUvulopalatopharyngoplasty (UVPP) is a surgicalprocedure involving excision of the uvula andthe posterior portion of the palate and tonsils,and trimming and reorientation of the tonsillarpillars. In adults, it has a success rate of approx-imately 50%.75 The procedure is not commonlyperformed in children and has not been system-

atically evaluated except in specific subpopula-tions. It is useful in patients in whom abnormalupper airway neuromuscular tone contributes toOSAS (e.g., patients with cerebral palsy andDown syndrome).76-78

Craniofacial and Other SurgeriesIn complex patients with OSAS, particularlythose with craniofacial anomalies, craniofacialsurgery may be effective. This is especially indi-cated in patients who do not tolerate CPAP orwho would also benefit from the cosmetic effectof surgery. Few large-scale studies have beenpublished; most of them were by one group ofsurgeons.79-82 This group has reported good suc-cess in treating patients with craniofacial anom-alies, Down syndrome, or cerebral palsy usingcomplex craniofacial surgery. They use an indi-vidualized surgical approach based on the site ofobstruction. Patients are evaluated by physicalexamination, radiologic studies, and endoscopy.Those with abnormalities in the region of thenares to the velum are treated with combinationsof adenoidectomy, septoplasty, and inferior tur-binectomy. Those with obstruction in the regionextending from the lips to the hypopharynx aretreated with procedures such as tonsillectomy,UVPP, tongue reduction, mandibular osteotomy,and tongue–hyoid suspension. These procedures


Side Effect TreatmentNasal symptoms (dryness, congestion, Heated humidification

rhinorrhea, epistaxis) Nasal steroidsSaline nose drops

Skin ulceration Correct mask fit.Avoid overtightening of head gear.Protect bridge of nose with hydrocolloid dressings.Gel masksSpacer on forehead instead of bridge of noseAlternate different interfaces (e.g., nasal cannula,

nasal mask).Eye irritation Correct mask fit.Inability to fall asleep while wearing CPAP Pressure rampCentral apnea Bilevel ventilation with backup rateHypercapnia (rarely seen in children) Nonrebreathing valveFailure to trigger bilevel ventilation Use ventilator with low triggering threshold

or adjustable sensitivity levels.If suboptimal ventilation, consider increasing

expiratory positive airway pressure and adding O2.Midfacial depression Custom mask

Table 21–2. Treatment of Side Effects of Continuous Positive Airway Pressure (CPAP)

Ch21.qxd 1/25/05 3:28 AM Page 240

are best performed in experienced, multidiscipli-nary centers, as they are fraught with potentialcomplications. Surgery may require prolongedintubation or temporary tracheostomy.

Select craniofacial procedures may be appro-priate for patients with isolated anomalies.Mandibular distraction procedures can be per-formed in children with micrognathia, includinginfants with Pierre Robin syndrome or hemifa-cial microsomia81 (Fig. 21–1). In this procedure,the mandible is divided and an external distrac-tion device applied. Distraction is graduallyapplied via the device, typically by 1 mm/day.However, long-term complications, such asasymmetrical jaw growth, may occur. Midfacialadvancement can improve both upper airwayfunction and cosmetic appearance in childrenwith craniofacial anomalies such as Crouzonsyndrome. Tongue reduction has been used totreat patients with Down syndrome or othercauses of macroglossia.77,83 However, the effectsof this surgery on swallowing and speech havenot been formally investigated. Epiglottoplasty isuseful in treating selected children with upperairway obstruction secondary to laryngomala-cia,84 although the overwhelming majority ofchildren with laryngomalacia do not requiretreatment, as they improve with age.

Weight LossWeight loss is recommended for all obesepatients. However, this is notoriously difficult toachieve. Even if weight loss does occur, it hap-pens slowly. Therefore, interim treatment isrequired. T&A is often effective even in mor-bidly obese children,9,85 and it is usually thefirst-line treatment unless there is minimal ade-notonsillar tissue. CPAP can be used pendingsurgery. Patients with significant obesity shouldundergo postoperative polysomnography todetermine whether further treatment, such asCPAP, is necessary.

Most of the data regarding weight loss andOSAS are from studies of adults. Major weightloss in adults, such as that seen with bariatricsurgery, has been reported to result in resolutionof apnea. Moderate weight loss (10% to 20% ofbody weight) is associated with a more modestimprovement in OSAS.86 One study evaluatedthe effects of bariatric surgery on OSAS in chil-dren.87 Eleven children underwent surgery, with

a resultant weight loss of 45% of body weight.Two children underwent simultaneous UVPP,and one child required a tracheostomy becauseof prolonged postoperative intubation. Ninepatients received polysomnographic evaluationpostoperatively and showed an improvement(from 73% to 93%) in mean arterial oxygen sat-uration. Additional details were not provided.Further study of bariatric surgery in morbidlyobese children with severe apnea, who have


Figure 21–1. Top, An 8-year-old girl with obstructivesleep apnea syndrome secondary to micrognathia isshown immediately after a mandibular distraction proce-dure with the placement of external fixators. Bottom,The patient is shown several years after the procedure.

Ch21.qxd 1/25/05 3:28 AM Page 241

failed other treatments, is warranted. Of con-cern, however, are several articles describinglong-term recurrence of OSAS in obese patientswho had lost weight, even though the weightloss had been maintained.88,89 This may haveresulted partly from the fact that the patientsnever attained a normal weight, even thoughthey weighed less than they had originally.

Oral Appliances and Orthodontic TreatmentOral appliances are recommended for adultpatients with mild OSAS or those who are intol-erant of CPAP,90 but few studies have evaluatedoral appliances in children. In one study,91 chil-dren with mild to moderate OSAS and dysgnathiawere fitted with an appliance. Twenty-six percentdiscontinued therapy. In the remainder, the apnea-hypopnea index decreased significantly but didnot normalize. A second study showed a modestimprovement in a group of children with mildobstructive apnea (apnea-hypopnea indexdecreased from 8 to 4 per hour).92 Further studyon oral and orthodontic appliances in children iswarranted. At present, these devices should beused only by specialists with expertise in pediatricorthodontics, as they have the potential to alterthe shape of the craniofacial skeleton in growingchildren. Conceivably, however, the early use oforal appliances could improve craniofacial char-acteristics that put children at risk for OSAS.

Rapid maxillary expansion using orthodonticappliances has been used successfully in onestudy of adults with obstructive apnea and max-illary constriction,93 and in one study of childrenwith obstructive apnea and maxillary constric-tion.94 In the latter study, children with adeno-tonsillar hypertrophy were excluded (the criteriafor diagnosing adenotonsillar hypertrophy werenot stated). Thus, this appears to be a promisingtechnique for a select population group.

Supplemental OxygenTwo studies have evaluated the effect of supple-mental oxygen on children with OSAS.95,96 Inboth studies, supplemental oxygen resulted inimproved arterial oxygen saturation, as expected.There were no clinically important differences inthe number or duration of obstructive apneaswhen subjects breathed supplemental oxy-gen.95,96 PCO2 levels did not change for the group

as a whole. However, a few individuals showed amarked increase in PCO2 when breathing supple-mental oxygen; this could not be predicted byage, weight, or severity of OSAS (Fig. 21–2).Supplemental oxygen use will not alleviate theincreased work of breathing or sleep fragmenta-tion associated with OSAS, and this is thereforenot an appropriate first-line treatment. However,these studies suggest that it may be useful as atemporizing measure, or when other treatmentsfail, in a few select individuals (e.g., in an infantwith mild OSAS resulting from craniofacialanomalies, who is expected to improve withgrowth). Supplemental oxygen should never beadministered to patients with OSAS without firstmeasuring their change in PCO2 in response tothe oxygen.

DrugsIn general, pharmacologic agents are not usefulin the management of OSAS. Systemic glucocor-ticosteroids have been used to shrink adenoidaltissue. However, only one study has evaluatedthis treatment. In an open-label, uncontrolledstudy, a 5-day course of prednisone did not


Room air CO2

90

80

70

60

50

40

Pea

k en

d-tid

al P

CO

2(m

m H

g)

Figure 21–2. The peak end-tidal PCO2 is shown for indi-vidual patients with obstructive sleep apnea syndromewhile breathing room air compared with supplementaloxygen during sleep. Group means and standard devia-tions are shown in the side bars. Although there was nodifference in the mean PCO2 between the two conditions,several individuals had an increase in PCO2 while breath-ing supplemental oxygen. (Adapted from Marcus CL,Carroll JL, Bamford O, et al: Supplemental oxygen duringsleep in children with sleep-disordered breathing. Am JRespir Crit Care Med 1995;152(Pt 1):1297-1301.)

Ch21.qxd 1/25/05 3:28 AM Page 242

significantly improve OSAS.97 Topical nasalsteroids have been shown to have a modest effectin children with OSAS.98 Following a 6-weekcourse of treatment, patients had a modestimprovement in apnea index from 11 ± 9 to 6 ±8/hr. However, the individual response to treat-ment was varied, with a wide standard deviation,and many subjects went on to have T&A. Theeffect of withdrawing topical steroid treatment isnot known. Thus, a trial of nasal steroids may bewarranted in some patients with mild OSAS, butfollow-up is essential. Further study is needed.

Environmental ConditionsThere is a high prevalence of allergy in childrenwith OSAS.99 Theoretically, avoiding allergensmay improve OSAS. Although passive smokingis associated with snoring,100 one study failed toshow a relationship between maternal smokingand OSAS.101 Nevertheless, it seems prudent tominimize exposure to cigarette smoke.

TracheostomyTracheostomy is the ultimate treatment forOSAS, as it bypasses the site of obstruction.However, it is associated with many complica-tions, including speech problems, chronic tra-cheitis, and interference with the activities ofdaily life. Fortunately, with the increased use ofCPAP and other treatment modalities, tra-cheostomy is now rarely required. It may beneeded in children with upper airway obstruc-tion during both wakefulness and sleep, particu-larly in children with cerebral palsy or severecraniofacial malformations.

CONCLUSION

A variety of effective treatment modalities areavailable for children with OSAS. However, manyunanswered questions remain. What degree ofOSAS warrants treatment? What is the correla-tion between polysomnographic abnormalitiesand clinical outcome? What is the long-term out-come of CPAP use in children? Which equipmentis best suited for them? What are the success andcomplication rates of complex surgical proce-dures such as UVPP and tongue reduction inchildren? Are successfully treated patients at riskfor recurrence of their disease as adults? Furtherlong-term studies are desperately needed.

References

1. Marcus CL, Omlin KJ, Basinki DJ, et al: Normalpolysomnographic values for children and ado-lescents. Am Rev Respir Dis 1992;146(Pt1):1235-1239.

2. Acebo C, Millman RP, Rosenberg C, et al: Sleep,breathing, and cephalometrics in older childrenand young adults. Chest 1996;109:664-672.

3. Poets CF, Stebbens VA, Samuels MP, Southall DP:Oxygen saturation and breathing patterns inchildren. Pediatrics 1993;92:686-690.

4. Hunt CE, Corwin MJ, Lister G, et al: Longi-tudinal assessment of hemoglobin oxygen satura-tion in healthy infants during the first 6 monthsof age. Collaborative Home Infant MonitoringEvaluation (CHIME) Study Group [see com-ments]. J Pediatr 1999;135:580-586.

5. Kravath RE, Pollak CP, Borowiecki B: Hypoven-tilation during sleep in children who havelymphoid airway obstruction treated bynasopharyngeal tube and T and A. Pediatrics1977;59:865-871.

6. Rosen GM, Muckle RP, Mahowald MW, et al:Postoperative respiratory compromise in childrenwith obstructive sleep apnea syndrome: Can it beanticipated? Pediatrics 1994;93:784-788.

7. American Academy of Pediatrics, Section onPediatric Pulmonology, Subcommittee onObstructive Sleep Apnea: Clinical practiceguideline: Diagnosis and management of child-hood obstructive sleep apnea syndrome.Pediatrics 2002;109:704-712.

8. American Thoracic Society: Standards and indi-cations for cardiopulmonary sleep studies inchildren. Am J Respir Crit Care Med 1996;153:866-878.

9. Kudoh F, Sanai A: Effect of tonsillectomy andadenoidectomy on obese children with sleep-associated breathing disorders. Acta OtolaryngolSuppl 1996;523:216-218.

10. Wiet GJ, Bower C, Seibert R, Griebel M: Surgicalcorrection of obstructive sleep apnea in the com-plicated pediatric patient documented by poly-somnography. Int J Pediatr Otorhinolaryngol1997;41:133-143.

11. Marcus CL, Keens TG, Bautista DB, et al:Obstructive sleep apnea in children with Downsyndrome. Pediatrics 1991;88:132-139.

12. Grundfast K, Berkowitz R, Fox L: Outcome andcomplications following surgery for obstructiveadenotonsillar hypertrophy in children withneuromuscular disorders. Ear Nose Throat J1990;69:756, 759-760.

13. Schechter MS: Technical report: Diagnosis andmanagement of childhood obstructive sleepapnea syndrome. Pediatrics 2002;109:E69.


Ch21.qxd 1/25/05 3:28 AM Page 243

14. Shintani T, Asakura K, Kataura A: The effect ofadenotonsillectomy in children with OSA. Int JPediatr Otorhinolaryngol 1998;44:51-58.

15. Nishimura T, Morishima N, Hasegawa S, et al:Effect of surgery on obstructive sleep apnea.Acta Otolaryngol Suppl 1996;523:231-233.

16. Agren K, Nordlander B, Linder-Aronsson S, et al:Children with nocturnal upper airway obstruc-tion: Postoperative orthodontic and respiratoryimprovement. Acta Otolaryngol 1998;118:581-587.

17. Nieminen P, Tolonen U, Lopponen H: Snoringand obstructive sleep apnea in children: A6 month follow-up study. Arch OtolaryngolHead Neck Surg 2000;126:481-486.

18. Frank Y, Kravath RE, Pollak CP, Weitzman ED:Obstructive sleep apnea and its therapy: Clinicaland polysomnographic manifestations. Pediatrics1983;71:737-742.

19. Suen JS, Arnold JE, Brooks LJ: Adenoton-sillectomy for treatment of obstructive sleepapnea in children. Arch Otolaryngol Head NeckSurg 1995;121:525-530.

20. Stradling JR, Thomas G, Warley AR, et al: Effectof adenotonsillectomy on nocturnal hypox-aemia, sleep disturbance, and symptoms in snor-ing children. Lancet 1990;335:249-253.

21. Zucconi M, Strambi LF, Pestalozza G, et al:Habitual snoring and obstructive sleep apneasyndrome in children: Effects of early tonsil sur-gery. Int J Pediatr Otorhinolaryngol 1993;26:235-243.

22. Bar A, Tarasiuk A, Segev Y, et al: The effect of ade-notonsillectomy on serum insulin-like growthfactor-I and growth in children with obstructivesleep apnea syndrome. J Pediatr 1999;135:76-80.


24. Williams EF III, Woo P, Miller R, Kellman RM:The effects of adenotonsillectomy on growth inyoung children. Otolaryngol Head Neck Surg1991;104:509-516.

25. Ahlqvist-Rastad J, Hultcrantz E, Melander H,Svanholm H: Body growth in relation to tonsil-lar enlargement and tonsillectomy. Int J PediatrOtorhinolaryngol 1992;24:55-61.

26. Soultan Z, Wadowski S, Rao M, Kravath RE:Effect of treating obstructive sleep apnea by ton-sillectomy and/or adenoidectomy on obesity inchildren. Arch Pediatr Adolesc Med 1999;153:33-37.

27. Weider DJ, Sateia MJ, West RP: Nocturnalenuresis in children with upper airway obstruc-tion. Otolaryngol Head Neck Surg 1991;105:427-432.

28. Ali NJ, Pitson D, Stradling JR: Sleep disorderedbreathing: effects of adenotonsillectomy onbehaviour and psychological functioning. Eur JPediatr 1996;155:56-62.

29. Goldstein NA, Post C, Rosenfeld RM, CampbellTF: Impact of tonsillectomy and adenoidectomyon child behavior. Arch Otolaryngol Head NeckSurg 2000;126:494-498.


31. Harvey JM, O’Callaghan MJ, Wales PD, et al:Six-month follow-up of children with obstruc-tive sleep apnoea. J Paediatr Child Health1999;35:136-139.

32. Buchinsky FJ, Lowry MA, Isaacson G: Do ade-noids regrow after excision? Otolaryngol HeadNeck Surg 2000;123:576-581.

33. Tami TA, Parker GS, Taylor RE: Post-tonsillec-tomy bleeding: An evaluation of risk factors.Laryngoscope 1987;97:1307-1311.

34. Wilkinson AR, McCormick MS, Freeland AP,Pickering D: Electrocardiographic signs of pul-monary hypertension in children who snore.BMJ 1981;282:1579-1581.

35. Yates DW: Adenotonsillar hypertrophy and corpulmonale. Br J Anaesth 1988;61:355-359.

36. Galvis AJ: Pulmonary edema complicating reliefof upper airway obstruction. Am J Emerg Med1987;5:294-297.

37. Feinberg AN, Shabino CL: Acute pulmonaryedema complicating tonsillectomy and ade-noidectomy. Pediatrics 1985;75:112-114.

38. Ruboyianes JM, Cruz RM: Pediatric adenotonsil-lectomy for obstructive sleep apnea. Ear NoseThroat J 1996;75:430-433.

39. McColley SA, April MM, Carroll JL, LoughlinGM: Respiratory compromise after adenotonsil-lectomy in children with obstructive sleepapnea. Arch Otolaryngol Head Neck Surg1992;118:940-943.

40. Rothschild MA, Catalano P, Biller HF: Ambula-tory pediatric tonsillectomy and the identifica-tion of high-risk subgroups. Otolaryngol HeadNeck Surg 1994;110:203-210.

41. Wiatrak BJ, Myer CM, Andrews TM: Compli-cations of adenotonsillectomy in children under3 years of age. Am J Otolaryngol 1991;12:170-172.

42. McGowan FX, Kenna MA, Fleming JA,O’Connor T: Adenotonsillectomy for upper air-way obstruction carries increased risk in chil-dren with a history of prematurity. PediatrPulmonol 1992;13:222-226.

43. Biavati MJ, Manning SC, Phillips DC: Predictivefactors for respiratory complications after tonsil-lectomy and adenoidectomy in children with


Ch21.qxd 1/25/05 3:28 AM Page 244

OSA. Arch Otolaryngol Head Neck Surg1997;123:517-521.

44. Gerber ME, O’Connor DM, Adler E, Myer CM:Selected risk factors in pediatric adenotonsillec-tomy [see comments]. Arch Otolaryngol HeadNeck Surg 1996;122:811-814.

45. Leiberman A, Tal A, Brama I, Sofer S: Obstruc-tive sleep apnea in young infants. Int J PediatrOtorhinolaryngol 1988;16:39-44.

46. Waters KA, Everett FM, Bruderer JW, SullivanCE: Obstructive sleep apnea: The use of nasalCPAP in 80 children. Am J Respir Crit Care Med1995;152:780-785.

47. Helfaer MA, McColley SA, Pyzik PL, et al:Polysomnography after adenotonsillectomy inmild pediatric obstructive sleep apnea. Crit CareMed 1996;24:1323-1327.

48. Marcus CL, Ward SL, Mallory GB, et al: Use ofnasal continuous positive airway pressure astreatment of childhood obstructive sleep apnea.J Pediatr 1995;127:88-94.

49. Smith PL, Wise RA, Gold AR, et al: Upper air-way pressure-flow relationships in obstructivesleep apnea. J Appl Physiol 1988;64:789-795.

50. Strohl KP, Redline S: Nasal CPAP therapy, upperairway muscle activation, and obstructive sleepapnea. Am Rev Respir Dis 1986;134:555-558.

51. Brochard L, Isabey D, Piquet J, et al: Reversal ofacute exacerbations of chronic obstructive lungdisease by inspiratory assistance with a facemask. New Engl J Med 1990;323:1523-1530.

52. McNicholas WT, Coffey M, Boyle T: Effects ofnasal airflow on breathing during sleep in normalhumans. Am Rev Respir Dis 1993;147:620-623.

53. American Thoracic Society: Indications andstandards for use of nasal continuous positiveairway pressure (CPAP) in sleep apnea syn-dromes. Am J Respir Crit Care Med 1994;150:1738-1745.

54. McNamara F, Sullivan CE: Obstructive sleepapnea in infants and its management with nasalcontinuous positive airway pressure. Chest1999;116:10-16.

55. Guilleminault C, Pelayo R, Clerk A, et al: Homenasal continuous positive airway pressure ininfants with sleep-disordered breathing. J Pediatr1995;127:905-912.

56. Downey R III, Perkin RM, MacQuarrie J: Nasalcontinuous positive airway pressure use in chil-dren with obstructive sleep apnea younger than2 years of age. Chest 2000;117:1608-1612.

57. Marcus CL: Ventilator management of abnormalbreathing during sleep: Continuous positive air-way pressure (CPAP) and nocturnal noninvasiveintermittent positive pressure ventilation. InLoughlin GM, Carroll JL, Marcus CL (eds): Sleepand Breathing in Children: A Developmental

Approach. New York, Marcel Dekker, 2000, pp797-812.

58. Koontz KL, Slifer KJ, Cataldo MD, Marcus CL:Improving pediatric compliance with positiveairway pressure therapy: The impact of behav-ioral intervention. Sleep 2003;26:1010-1015.

59. Rains JC: Treatment of obstructive sleep apneain pediatric patients. Clin Pediatr 1995;34:535-541.

60. Sforza E, Lugaresi E: Daytime sleepiness andnasal continuous positive airway pressure ther-apy in obstructive sleep apnea syndromepatients: Effects of chronic treatment and 1-nighttherapy withdrawal. Sleep 1995;18:195-201.

61. Kribbs NB, Pack AI, Kline LR, et al: Effects of onenight without nasal CPAP treatment on sleep andsleepiness in patients with obstructive sleepapnea. Am Rev Respir Dis 1993;147:1162-1168.

62. Boudewyns A, Sforza E, Zamagni M, Krieger J:Respiratory effort during sleep apneas afterinterruption of long-term CPAP treatment inpatients with obstructive sleep apnea. Chest1996;110:120-127.

63. Reeves-Hoche MK, Meck R, Zwillich CW: NasalCPAP: An objective evaluation of patient com-pliance. Am J Respir Crit Care Med 1994;149:149-154.

64. Krieger J, Kurtz D, Petiau C, et al: Long-termcompliance with CPAP therapy in obstructivesleep apnea patients and in snorers. Sleep1996;19(Suppl):S136-S143.

65. Kribbs NB, Pack AI, Kline LR, et al: Objectivemeasurement of patterns of nasal CPAP use bypatients with obstructive sleep apnea [see com-ments]. Am Rev Respir Dis 1993;147:887-895.

66. Marcus CL, Davidson Ward SL, Lutz JM, et al:Compliance with CPAP vs bilevel pressure inchildren. Am J Respir Crit Care Med 2004;169:A732.

67. Constantinidis J, Knobber D, Steinhart H, et al:Fine-structural investigations of the effect ofnCPAP-mask application on the nasal mucosa.Acta Otolaryngol 2000;120:432-437.

68. Pepin JL, Leger P, Veale D, et al: Side effects ofnasal continuous positive airway pressure in sleepapnea syndrome: Study of 193 patients in twoFrench sleep centers. Chest 1995;107:375-381.

69. Massie CA, Hart RW, Peralez C, Richards GN:Effects of humidification on nasal symptomsand compliance in sleep apnea patients usingcontinuous positive airway pressure. Chest1999;116:403-408.

70. Choo-Kang LR, Ogunlesi FO, McGrath-MorrowSA, Marcus CL: Recurrent pneumothoracesassociated with nocturnal noninvasive ventila-tion in a patient with muscular dystrophy.Pediatr Pulmonol 2002;34:73-78.


Ch21.qxd 1/25/05 3:28 AM Page 245

71. Alvarez-Sala R, Garcia IT, Garcia F, et al: NasalCPAP during wakefulness increases intraocularpressure in glaucoma. Monaldi Arch Chest Dis1994;49:394-395.

72. Bamford CR, Quan SF: Bacterial meningitis:A possible complication of nasal continuouspositive airway pressure therapy in a patientwith obstructive sleep apnea syndrome and amucocele. Sleep 1993;16:31-32.

73. Robertson NJ, McCarthy LS, Hamilton PA, MossALH: Nasal deformities resulting from flowdriver continuous positive airway pressure. ArchDis Child 1996;75:F209-F212.

74. Li KK, Riley RW, Guilleminault C: An unre-ported risk in the use of home nasal continuouspositive airway pressure and home nasal ventila-tion in children: Mid-face hypoplasia. Chest2000;117:916-918.

75. American Sleep Disorders Association: Practiceparameters for the treatment of obstructive sleepapnea in adults: The efficacy of surgical modifi-cations of the upper airway. Sleep 1995;19:152-155.

76. Kosko JR, Derkay CS: Uvulopalatopharyngo-plasty: Treatment of obstructive sleep apnea inneurologically impaired pediatric patients. Int JPediatr Otorhinolaryngol 1995;32:241-246.

77. Donaldson JD, Redmond WM: Surgical manage-ment of obstructive sleep apnea in children withDown syndrome. J Otolaryngol 1988;17:398-403.

78. Seid AB, Martin PJ, Pransky SM, Kearns DB:Surgical therapy of obstructive sleep apnea inchildren with severe mental insufficiency.Laryngoscope 1990;100:507-510.

79. Burstein FD, Cohen SR, Scott PH, et al: Surgicaltherapy for severe refractory sleep apnea ininfants and children: Application of the airwayzone concept. Plast Reconstr Surg 1995;96:34-41.

80. Cohen SR, Simms C, Burstein FD, Thomsen J:Alternatives to tracheostomy in infants and chil-dren with obstructive sleep apnea. J Pediatr Surg1999;34:182-187.

81. Cohen SR, Lefaivre JF, Burstein FD, et al:Surgical treatment of obstructive sleep apnea inneurologically compromised patients. PlastReconstr Surg 1997;99:638-646.

82. Lefaivre JF, Cohen SR, Burstein FD, et al: Downsyndrome: Identification and surgical manage-ment of obstructive sleep apnea. Plast ReconstrSurg 1997;99:629-637.

83. Morgan WE, Friedman EM, Duncan NO, SulekM: Surgical management of macroglossia in chil-dren. Arch Otolaryngol Head Neck Surg 1996;122:326-329.

84. Marcus CL, Crockett DM, Ward SL: Evaluationof epiglottoplasty as treatment for severe laryn-

gomalacia [published erratum appears inJ Pediatr 1991;118:168]. J Pediatr 1990;117:706-710.

85. Marcus CL, Curtis S, Koerner CB, et al: Eval-uation of pulmonary function and polysomnog-raphy in obese children and adolescents. PediatrPulmonol 1996;21:176-183.

86. Strobel RJ, Rosen RC: Obesity and weight loss inobstructive sleep apnea: A critical review. Sleep1996;19:104-115.

87. Breaux CW: Obesity surgery in children. ObesSurg 1995;5:279-284.

88. Sampol G, Munoz X, Sagales MT, et al: Long-term efficacy of dietary weight loss in sleepapnoea/hypopnea syndrome. Eur Respir J 1998;12:1156-1159.

89. Pillar G, Peled R, Lavie P: Recurrence of sleepapnea without concomitant weight increase 7.5years after weight reduction surgery. Chest1994;106:1702-1704.

90. American Sleep Disorders Association: Practiceparameters for the treatment of snoring andobstructive sleep apnea with oral appliances.Sleep 1995;18:511-513.

91. Villa MP, Bernkopf E, Pagani J, et al: Randomizedcontrolled study of an oral jaw-positioning appli-ance for the treatment of obstructive sleep apneain children with malocclusion. Am J Respir CritCare Med 2002;165:123-127.

92. Cozza P, Gatto R, Ballanti F, Prete L: Manage-ment of obstructive sleep apnoea in childrenwith modified monobloc appliances. Eur JPaediatr Dent 2004;5:24-29.

93. Cistulli PA, Palmisano RG, Poole MD: Treatmentof obstructive sleep apnea syndrome by rapidmaxillary expansion. Sleep 1998;21:831-835.

94. Pirelli P, Saponara M, Guilleminault C: Rapidmaxillary expansion in children with obstructivesleep apnea syndrome. Sleep 2004;27:761-766.

95. Marcus CL, Carroll JL, Bamford O, et al:Supplemental oxygen during sleep in childrenwith sleep-disordered breathing. Am J RespirCrit Care Med 1995;152(Pt 1):1297-1301.

96. Aljadeff G, Gozal D, Bailey-Wahl SL: Effects ofovernight supplemental oxygen in obstructivesleep apnea in children. Am J Respir Crit CareMed 1996;153:51-55.

97. Al-Ghamdi SA, Manoukian JJ, Morielli A, et al:Do systemic corticosteroids effectively treatobstructive sleep apnea secondary to adenoton-sillar hypertrophy? Laryngoscope 1997;107:1382-1387.

98. Brouillette RT, Manoukian JJ, Ducharme FM,et al: Efficacy of fluticasone nasal spray for pedi-atric obstructive sleep apnea. J Pediatr 2001;138:838-844.


Ch21.qxd 1/25/05 3:28 AM Page 246

99. McColley SA, Carroll JL, Curtis S, et al: Highprevalence of allergic sensitization in childrenwith habitual snoring and obstructive sleepapnea. Chest 1997;111:170-173.

100. Corbo GM, Fuciarelli F, Foresi A, De BenedettoF. Snoring in children: Association with respira-tory symptoms and passive smoking. BMJ1989;299:1491-1494.

101. Redline S, Tishler PV, Schluchter M, et al: Riskfactors for sleep-disordered breathing in chil-dren: Associations with obesity, race, and respi-ratory problems. Am J Respir Crit Care Med1999;159:1527-1532.


Ch21.qxd 1/25/05 3:28 AM Page 247

General and pediatric otolaryngologists are fre-quently asked to evaluate and consider surgicalmanagement in children with snoring, sleep-related breathing disorders, and possible obstruc-tive sleep apnea (OSA). OSA and upper airwayresistance syndrome (UARS) in the pediatric pop-ulation are most often caused by adenoid and ton-sillar (adenotonsillar) hypertrophy, which has apeak incidence from the ages of 2 to 5 years mir-roring the incidence of pediatric OSA. However,other causes of upper airway obstruction must beruled out and appropriately treated if present. Sincethe widespread availability and use of antibiotics,the morbidity and complications of tonsillitis havebeen reduced dramatically. Tonsillectomy and ade-noidectomy are the most commonly performedcurative procedures, and awareness of pediatricOSA syndrome (OSAS) and sleep apnea hasresulted in a changing trend in the indication forpediatric adenotonsillectomy over the traditionalinfectious indication.1 The definitions of OSA andUARS are described in detail in other chapters ofthis text; hence, this chapter will focus on the diag-nosis, surgical indications, and outcome of adeno-tonsillectomy specifically relevant to OSA.

PATHOPHYSIOLOGY OF OSA

The pathophysiology of sleep-related breathingdisorders, UARS, and OSA differs from patient topatient, depending on the presence of medicalcomorbidities, central versus obstructive apnea,and, if obstructive, the specific site of upper airwayobstruction. The upper airway includes the nasalpassages down to the glottic inlet, including poten-tially redundant soft tissue in the oropharynx andhypopharynx; therefore, evaluation must includeareas other than the adenoid and tonsillar tissue.

Nasal obstruction secondary to allergic rhinitisor polyps may respond to conservative manage-ment. Stridor may signify laryngeal or trachealpathology. Benign or malignant upper airwaymasses can present with snoring and sleep-disordered breathing. Comorbidities such as cran-iofacial abnormalities, neuromuscular deficienciescaused by cerebral palsy, and Down syndrome mayplay primary roles in the upper airway obstruc-tion.2 OSA specific to patients with craniofacialabnormalities and/or neurologically challengedchildren is carefully described in other chapters.Such patients require a more complex treatmentplan through a multidisciplinary approach.

Nonsurgical management of pediatric OSAincludes mechanical interventions such as contin-uous positive airway pressure (CPAP), bilevel pos-itive airway pressure (BiPAP), and medicaltherapy. The indications and technical detailsrelated to CPAP and BiPAP are covered elsewhere.Medical therapy is specifically relevant in the treat-ment of obesity, allergy, sinonasal pathology, andadenotonsillar hypertrophy due to acute tonsillitis.

ObesityObesity is becoming more prevalent in theUnited States and is recognized as a significanthealth problem in children. However, in contrastto adults, most preadolescent children with OSAare of normal weight or even underweight.Nonetheless, obesity inducing or contributing toOSA is growing more prevalent in the pediatricpopulation. Weight control alone has beenshown to result in the partial cure of sleep-related breathing disorders.3 For children withsevere or morbid obesity, referral to a weight lossspecialist or nutritionist and family education

Otolaryngologic Management ofSleep-Related Breathing Disorders

Julie L. Wei

Roberto L. Barretto

Mark E. Gerber

22

249

Ch22.qxd 1/25/05 3:28 AM Page 249

are critical in successfully reducing long-termsequelae of not only OSA but all obesity-relatedhealth problems.

Nasal ObstructionThe treatment of nasal pathology may improveobstructive symptoms. Many children presentwith a history of nasal congestion and nocturnalmouth breathing. This may be due to the nasalcycle, which is a natural phenomenon whenbreathing occurs via a single nasal passage suchas in turbinate mucosa alternate engorgementor due to vasomotor rhinitis or allergies. In chil-dren who present with nasal airway obstruction,significant adenoid hypertrophy causing nasopha-ryngeal obstruction needs to be ruled out. Over-the-counter topical nasal decongestants and oraldecongestants are not recommended for chronicuse in the pediatric population. Nasal salineirrigation may be effective in relieving nasal con-gestion, especially during upper respiratoryinfections. Consistent daily use of topical nasalsteroids may decrease the severity of snoring.Mometasone has been approved for children asyoung as 2 years of age, and fluticasone isapproved for children 4 years of age or older.Intranasal steroid sprays may relieve nasalobstruction secondary to adenoid hypertrophy,providing symptomatic improvement in thosewithout tonsillar hypertrophy. Aqueous intranasalbeclomethasone (Beconase AQ) used twice dailyfor 16 weeks was shown to decrease the ade-noid/choana ratio by 29% and reduce the nasalobstruction symptom score by 82%.10

Management of allergic rhinitis is essentialif this is the cause of nasal congestion and theresulting upper airway obstruction. Avoidance,antihistamines, mast cell stabilizers, intranasalsteroids, and immunotherapy, alone or in com-binations are effective measures in controllingatopic disease.4 Allergy testing may be helpful.Chronic sinusitis should also be considered inthe patient with persistent rhinitis and nasal air-way obstruction. In rare circumstances radio-graphic imaging studies such as coronal sinuscomputed tomographic screening may be help-ful in the child with a history that is suggestiveof but not conclusive enough to warrant medicalmanagement for sinusitis. Usually the clinicalhistory and physical findings are sufficient toinitiate a therapeutic trial regimen for sinusitis.

Appropriate sinusitis treatment should includebroad-spectrum antimicrobial therapy. Ancillarymeasures such as topical nasal steroids and salinenasal washes may be beneficial. Adenoidectomyalone or with maxillary sinus irrigation has beenshown to improve patients with purulent rhinor-rhea and chronic nasal congestion secondary tochronic sinusitis and adenoiditis.

Before medical therapy for nasal obstruction,the presence of sinonasal and pharyngeal massesmust be ruled out. The literature is replete withcases of benign and malignant neoplasms present-ing clinically with snoring and symptoms sug-gestive of OSA. A thorough history and completeexamination with heightened suspicions arenecessary to identify these rare conditions andprevent a physician from simply attributing thesymptoms of sleep-related breathing disorders toadenotonsillar hypertrophy.

GastroesophagealReflux/Gastrolaryngo-pharyngeal RefluxFunctional and pathologic gastroesophagealreflux (GER), and more specifically gastrolaryn-gopharyngeal reflux (GLPR), are widely recog-nized in the pediatric population and have beenshown to be associated with both chronic sinusi-tis and upper airway edema in children. Directvisualization of the upper airway by flexiblelaryngoscopy in a child with GLPR may revealedema of the pharyngeal and laryngeal mucosaas well as lingual tonsillar hypertrophy. Thesechanges in turn may result in sleep-relatedbreathing disorders and apnea in infants.5,6

When the history and/or physical findingssuggest that GLPR is involved, either a dual-channel, 24-hr pH probe study or a therapeutictrial of proton pump inhibitors or H2 blockers iswarranted.

DIAGNOSIS OF ADENOID ANDTONSILLAR HYPERTROPHY

Adenoid hypertrophy is usually diagnosed bylateral neck x-ray, whereas tonsillar hypertrophyis clinically diagnosed by direct visualization.The history of mouth breathing may suggestnasal obstruction and adenotonsillar hypertro-phy, but it is not specific to either adenotonsillarhypertrophy or OSA. Most children have promi-

250 Chapter 22 Otolaryngologic Management of Sleep-Related Breathing Disorders

Ch22.qxd 1/25/05 3:28 AM Page 250

nent adenoid and tonsillar tissue relative to theirsmall head size, but the majority of children donot have OSA. Children are susceptible to fre-quent viral infections involving the upper respi-ratory tract and will frequently have nasalcongestion and rhinorrhea associated with theupper respiratory illnesses even without adenoidhypertrophy. The lateral neck x-ray will demon-strate nasopharyngeal airway narrowing second-ary to prominent adenoid tissue, or alternatively,direct visualization using flexible fiber-opticnasopharyngoscopy will demonstrate adenoidhypertrophy if it is seen to occupy more thanapproximately 80% of the choana. Since bothmethods lack objectivity, the physician’s obser-vation, based on the clincal history and exam,should determine whether the patient undergoespolysomnography to assist in deciding if adeno-tonsillectomy is indicated.

When the oropharynx is inspected, the size,position, and characteristics of tonsils such asthe presence of exudate, tonsilloliths, or cryptsare noted. Tonsil size is determined clinically bythe space both tonsils occupy in the oropharynxin view. Tonsils occupying less than 25% of thepharynx are routinely described by otolaryngol-ogists as 1+, less than 50% 2+, less than 75% 3+,and finally, those occupying greater than 75%of the oropharyngeal space or tonsils making con-tact in the midline are 4+. This is a generalizedclassification for description. The decision toperform tonsillectomy should not be based onsize alone. The tonsils may be endophytic andappear as 1+ or 2+ in the office exam, and thetrue size is often not appreciated until the timeof surgery, when the patient is under anesthesiaand the oropharynx is completely exposed usinga mouth gag instrument. There is no strict cor-relation between the size of the tonsil and upperairway obstruction, and the oral exam may notrepresent the actual degree of obstruction com-pared with an exam of the posterior oropharynxwith nasopharyngoscopy.

ACUTE TONSILLITIS

Patients with acute tonsillitis may have suddenor new-onset symptoms of snoring and OSA.Tonsillar hypertrophy secondary to acute bacter-ial or viral infection may be unilateral or bilat-eral, both of which may cause new or worseningsymptoms of snoring and upper airway obstruc-

tion. Significant tonsillitis has been defined toinclude any of the following: temperature above38.5° C, cervical adenopathy of greater than 2 cm,the presence of tonsillar exudate, or positivegroup A b-hemolytic streptococcus (GABHS).Patients evaluated for acute tonsillitis with con-comitant symptoms of fever, odynophagia, dys-phagia, and possible airway obstruction shouldbe treated with a course of oral antibioticsif exudative tonsillitis is seen and/or a bacterialpathogen is confirmed by a rapid streptococcalswab test. Pain associated with acute tonsillitiscan be safely and adequately treated with a sin-gle dose of oral or intramuscular dexamethasonein patients older than 15 years of age.7 Acutemononucleosis may result in severe adenotonsil-lar hypertrophy with significant obstruction.A decision for adenotonsillectomy should takeinto consideration the patient’s infection historyand the frequency and severity of such airwayobstruction as well as the likelihood of recur-rence. Patients with acute tonsillar hypertrophysecondary to an infection should also be fol-lowed after infectious symptoms have resolvedto ensure that there is resolution of adenoidand/or tonsillar hypertrophy and symptoms ofupper airway obstruction.

CHRONIC ADENOTONSILLARHYPERTROPHY

In one study of cases of chronic adenotonsillarhypertrophy, broad-spectrum antibiotics suchas amoxicillin/clavulanate potassium given as a30-day course have been shown to decrease ade-notonsillar hypertrophy in the short term, but83% of the patients had a return of their sleep-disordered breathing–related symptoms andwent on to have surgical management for OSAat 24-month followup.8 There may be a role fora prolonged course of antibiotics in providingshort-term symptom relief or if the patient is nota surgical candidate due to medical comorbidities.

Dexamethasone 1 mg/kg/day over 3 to 5 dayscan result in a rapid reduction of tonsil and ade-noid size, with an improvement in the symptomsof upper airway obstruction. However, shortcourses of oral prednisone have been shown tobe ineffective in treating pediatric OSA causedby adenotonsillar hypertrophy.9 Long-term useof corticosteroids is generally not recommendedsecondary to systemic side effects.

Otolaryngologic Management of Sleep-Related Breathing Disorders Chapter 22 251

Ch22.qxd 1/25/05 3:28 AM Page 251

DIAGNOSIS OF OSA

While the clinical history and physical exam areimportant, they do not provide a reliable diag-nosis of OSA. Several recent studies have demon-strated that OSAS diagnosed by history alonewas confirmed in only 30% to 55.6% of patientswhen polysomnograms were perrformed.10-14

These studies suggest that there is a poor corre-lation between history and OSA with polysomnog-raphy (PSG). However, they were conductedbefore high UARS was recognized and thereforemost likely underestimated the incidence ofsleep-disordered breathing. In general, it doesappear that snoring as a symptom is not a reli-able indicator of OSA. Primary snoring withoutassociated abnormalities in sleep architecture,alveolar ventilation, or oxygenation is estimatedto occur in up to 10% of the pediatric popula-tion.15 PSG is considered the gold standard indiagnosing OSA. It is imperative to understandthat OSAS is differentially defined in adultsand children, and the criteria for an abnormalpediatric polysomnogram are different fromthose of adults.16 An apnea index of 1 and anapnea-hypopnea index (AHI) of greater than orequal to 5 are considered abnormal; however,the clinical significance of these numbers isunknown. Due to the difficulties in confirmingsleep-disordered breathing by history alone, PSGis a useful tool to assist in the confirmation ofthe diagnosis before considering surgical inter-vention.17 The severity of the OSA based on theapnea index, respiratory disturbance index, low-est oxygen saturation, and maximum end-tidalCO2 may direct the planning of postoperativecare, such as observation in the intensive careunit and parental counseling of possible pro-longed intubation during the immediate postop-erative period. However, not every child withsymptoms and an exam suggestive of upper air-way obstruction secondary to adenotonsillarhypertrophy needs to have PSG-proven OSAbefore consideration of surgical intervention.First of all, there are simply not enough availablesleep center resources to perform and interpretthe number of pediatric polysomnograms neces-sary to evaluate every child with the symptomsof sleep-related breathing disorders. Second,clinically significant sleep-disordered breathingcan be present even in the absence of an abnor-mal polysomnogram. UARS has recently been

described in which negative intrathoracic pres-sure swings during inspiration lead to EEGarousals and sleep fragmentation and, moreimportantly, to subsequent daytime symptoma-tology similar to that of OSAS. In UARS thepolysomnogram does not demonstrate signifi-cant apneas and hypopneas. It has been sug-gested to be more common than OSAS.18 UARSmay be formally diagnosed either by esophagealmanometry, which measures respiratory effortand intrathoracic pressure, or end-tidal CO2capnography in conjunction with a suprasternalnotch monitor, which detects increased intratho-racic pressure.

It is neither cost-effective nor possible to haveevery child who is a potential candidate for ade-notonsillectomy obtain a polysomnogram, regard-less of the severity of symptoms and the exam,or to undergo the aforementioned testing todiagnose UARS. The prevalence of UARS isunderestimated in children, since it is not rou-tinely assessed. However, since UARS leads tosymptomatology similar to that of OSAS, clinicalhistory alone that is suggestive of UARS can bean indication for surgical intervention. Thesepatients demonstrate improvement in the elimi-nation of snoring, sleep fragmentation, daytimesomnolence or hyperactivity, and other symp-toms associated with sleep-related breathingdisorders after adenotonsillectomy. PSG shouldalso be offered to families as an alternativeapproach to assist in providing additional objec-tive data to help their decision making. Forpatients who are not surgical candidates due tomedical comorbidity, PSG is necessary to deter-mine the parameters for nonsurgical treatmentoptions such as CPAP.

POLYSOMNOGRAPHY BEFORE SURGERY

There are currently no clear criteria according towhich patients should undergo polysomnogra-phy before adenotonsillectomy. This decision islargely dependent on the physician bias andtherefore should be based on the severity of theclinical history, the exam, and more specificallythe presence of medical comorbidity, includingobesity. For an otherwise healthy child withoutmedical comorbidity, including obesity, and whodoes not have any history of intubation, hospi-talization, or airway surgery, an oropharyngeal


Ch22.qxd 1/25/05 3:28 AM Page 252

exam consistent with significant tonsillar hyper-trophy may support the decision for adenotonsil-lectomy without first obtaining a polysomnogram.Complex patients who are at high risk for OSAmay benefit from a baseline polysomnogram. Inthis population, additional postoperative PSG isuseful to confirm the resolution of OSA, even inthe absence of residual symptoms. Table 22–1offers several predisposing factors for childhoodOSA but is by no means a complete list. In addi-tion, whenever the history does not match theexamination, PSG is indicated.

INDICATIONS FORADENOTONSILLECTOMY

Successful surgical management of pediatricsleep-disordered breathing depends on identifi-cation and intervention at every level of obstruc-tion, which may occur from the nasal vaultdown to the level of the glottis. A thorough his-tory of symptoms, the physical exam, and attimes flexible nasopharyngoscopy, laryngoscopy,and/or imaging studies will allow for the diagno-sis of a site-specific obstruction and the forma-tion of an appropriate surgical treatment plan.Most cases of sleep-disordered breathing in chil-dren are secondary to adenotonsillar hyper-trophy, and symptoms are readily alleviated afteradenotonsillectomy. Patients in whom thepolysomnogram demonstrates both central andobstructive hypoventilation may require a multi-disciplinary approach to achieve a satisfactoryresult. Decisions regarding when and how tointervene depend on several factors, amongthem the severity and site of obstruction; overallhealth of the child, including comorbid condi-tions; and family expectations. Treatment of pedi-atric sleep-disordered breathing varies dependingon the severity. For very mild cases, observationalone with monitoring for worsening signs andsymptoms may be sufficient. Surgical manage-ment has been shown to be beneficial not only topatients with mild cases secondary to adenoton-sillar hypertrophy alone but also to patientswith the aforementioned comorbid conditions.A combination of medical, surgical, and mechan-ical interventions may be necessary in treatingmoderate to severe cases.

Children who demonstrate mild or no evi-dence of OSA by polysomnography may stillhave UARS, causing sleep disturbances as well as

daytime chronic mouth breathing and behaviorproblems. A prospective study using parentalquestionnaires and sleep polysomnographydemonstrated that adenotonsillectomy provideslong-term benefits even in children in whom theobstruction did not meet the criteria for OSA(patients most likely have high upper airwayresistance).19 Many times the parents of childrenwith sleep-disordered breathing have a highlevel of anxiety and disrupted sleep patterns sec-ondary to their child’s sleep dysfunction. Surgicalintervention in these children can promptly


Neurologic AbnormalitiesSeizure disorder Head injuryCerebral palsy PrematurityHydrocephalus Central apneaArnold-Chiari Meningomyelocele

malformation Myotonic dystrophy

Craniofacial AbnormalitiesMicrognathia RetrognathiaMacroglossia Maxillary hypoplasia

DiseasesHypothyroidism Crouzon’s diseaseGoiter Gastroesophageal Morbid obesity reflux

Anatomic ObstructionsExcessive soft tissue Nasal stenosis

of the neck Choanal atresiaShort neck Septal deviationLingual tonsil Laryngeal papillomas/

hypertrophy tumorsRedundant Subglottic stenosis

oropharyngeal Subglottic mucosa/long uvula hemangioma

Adenoid hypertrophy Nasal polypsTonsillar hypertrophy Laryngeal web/Pharyngeal flap stenosis/mass

surgery

SyndromesAchondroplasia Beckwith-WiedemannApert’s Treacher CollinsDown Stickler’sPierre Robin Fetal alcoholPrader-Willi Marfan’sKlippel-Feil Hemifacial microsomia

Table 22–1. Factors Associated with HighRisk for Obstructive SleepApnea

Ch22.qxd 1/25/05 3:28 AM Page 253

improve the quality of the life for both the patientand the family.

TONSILLECTOMY AND ADENOIDECTOMY

Adenotonsillectomy is the most common majorpediatric surgical procedure performed in theUnited States. Sleep-related breathing disorderssecondary to adenotonsillar hypertrophy mayhave been more common than recurrent tonsilli-tis as an indication for surgery ever since theinception of the widespread use of antibiotic ther-apy. Daytime audible breathing and chronic mouthbreathing are also common parental concernsleading to referral to an otolaryngologist. Historytaking should focus on night-time symptomssuch as consistent nocturnal snoring, includingobservation of the degree of work of breathing; thepresence of retractions or paradoxical abdominalrespirations; obstructive pauses in airflow fre-quently followed by snorting or gasping for air;restless sleep; frequent arousals; and enuresis.Daytime symptoms include difficulty getting upin the morning, excessive irritability, hyperactiv-ity, a poor attention span, somnolence, and evenmorning headaches. Children with more severeadenotonsillar hypertrophy may have audiblebreathing while awake, which is easily notedduring an office visit.

PREOPERATIVE EVALUATION

Preoperative evaluation may vary depending onsurgeon preference and hospital guidelines.Children who are otherwise healthy may obtainmedical clearance for general anesthesia by athorough history and physical exam from theirprimary physicians. Any patient with a complexmedical history, including seizure, cardiac,endocrine, and pulmonary problems, shouldhave an appropriate preoperative evaluation bythe specialist treating the specific problem.Recommendations for perioperative, intraopera-tive, and/or postoperative management specificto each medical condition should be outlinedand communicated to the surgeon before theprocedure. Patients with complex medical comor-bidities or who may be at an increased risk withgeneral anesthesia should be evaluated by ananesthesiologist before the surgery. Many oto-laryngologists rely on a detailed bleeding ques-

tionnaire to help screen for a possible bleedingdisorder or the need for blood work and possiblehematology consultation. When there are posi-tive findings on the questionnaire, a screeningset of coagulation studies is obtained, includinga complete blood count, prothrombin time, partialthromboplastin time, and the recently availableplatelet function analysis, which is a replace-ment for bleeding time as a measure of plateletfunction. For African-American and other high-risk groups, a sickle cell screen should be con-sidered if the patient has not been previouslyevaluated for sickle cell disease. There remainscontroversy regarding the need for routine, pre-operative blood work before surgery, but dueto the lack of cost-effectiveness as well as theinability to predict which patient is at risk for thereported 1% to 4% of postoperative hemorrhage,routine blood work before adenotonsillectomy isnot usually performed.20 Laboratory evaluationsspecific to coexisting medical conditions may benecessary to determine clearance for surgery andgeneral anesthesia.

SURGICAL TECHNIQUES

Adenotonsillectomy is the most commonly per-formed procedure under general anesthesia bya variety of surgical techniques, including coldknife; electrosurgery, such as monopolar, bipolar,controlled radiofrequency, coablation, laser, andargon plasma coagulation; needlepoint or bladeelectrocautery; harmonic scalpel; and mostrecently the powered instrument microdébridertechnique, described as subtotal or intracapsulartonsillectomy.21-23 Electrosurgery and electro-cautery differ in that electrosurgery uses a high-frequency electrical current that passes throughtissue near the active electrode, causing the heat-ing of the tissue and resulting in cutting, co-agulation, and ablation of the tissue; in elec-trocautery the current passes through a heatingelement that conducts heat to the blade, creating a“hot blade” for cauterization or cutting of thetissue. Multiple studies have reviewed the differ-ences among the various techniques. Except forreported differences in operating time and intra-operative bleeding, there are minimal statisticallysignificant differences with respect to postopera-tive bleeding risk as well as outcome measuressuch as postoperative pain when a complete ton-sillectomy is performed.24-31


Ch22.qxd 1/25/05 3:28 AM Page 254

A subtotal/partial tonsillectomy is also de-scribed as intracapsular. This technique preservesthe tonsillar capsule and is believed to avoid directsurgical violation of the pharyngeal muscles, thusreducing postoperative pain and recovery timebecause there is less injury to and inflammationof the muscles and less disruption of nerve end-ings. A recent study concluded that when com-pared with the standard technique, intracapsulartonsillectomy is just as effective in relievingobstructive sleep-disordered breathing but pro-duces statistically significantly less postoperativepain and fewer episodes of delayed hemorrhageand dehydration.21 This study also reported thatthe numbers of days to return of normal activityand analgesic use were lower in the groups whounderwent an intracapsular tonsillectomy. Futuredata from current, prospective, randomizedstudies may confirm that statistically significantdifferences exist between this and all previoustechniques.

Adjunctive perioperative measures to improvepostoperative results include the use of dexa-methasone in a single perioperative dose. Meta-analysis of all blinded, randomized trials ofperioperative dexamethasone use has demon-strated decreased postoperative emesis and anearlier return to a soft or regular diet in thosethat receive it. A single dose of 0.5 mg/kg of dex-amethasone, up to 20 mg, given intraoperativelyhas been shown to be effective in reducing post-operative emesis and pain as well as improvingoral intake.32,33

POSTOPERATIVE MANAGEMENT

Candidacy for Ambulatory SurgerySeveral reviews of large series have demonstratedthat adenoidectomy and tonsillectomy are safelyperformed as outpatient procedures on thosechildren who meet the selection criteria.34-40

One prospective study showed the safety of asix-hour postoperative recovery room observa-tion period, and recommended a four-hourobservation period based on their finding.35

Children with OSAS have been shown to havean increased risk of respiratory compromiseimmediately after surgery, with the most signifi-cant risk factors being age less than 3 years andan AHI of greater than 10.41 Two other risk fac-tors include having an abnormal electro-

cardiogram and weight less than the 5th percentilefor age.41 An increased risk of respiratory com-promise applies not only to children who haveOSA secondary to adenotonsillar hypertrophybut also to those with OSA associated with cran-iofacial anomalies causing maxillary and/ormandibular hypoplasia and reduction in thepharyngeal airway. In addition to craniofacialanomalies, medical comorbidities involving neu-rologic problems such as seizure disorder, cere-bral palsy, Down syndrome, and others causesuch OSA to be more complex.42

One study identified postoperative respira-tory compromise after adenotonsillectomy in 10of 37 children with OSAS, and a review ofpatient characteristics resulted in a recommen-dation of overnight observation for patients whohave any of the following high-risk clinical crite-ria: less than 2 years of age, craniofacial anom-alies including midfacial hypoplasia ormicrognathia/retrognathia, failure to thrive,hypotonia, cor pulmonale, morbid obesity, or asignificant preoperative PSG abnormality.43

Several studies have demonstrated theincreased incidence of postoperative airwaycomplications in patients younger than 3 yearsof age; therefore, it is the common practice ofmost otolaryngologists to perform adenotonsil-lectomy as a 23-hour observation procedure inthose younger than 3 years of age.31,38,44,45 Othervariables found to be statistically significant inpredicting the risk of respiratory compromiseafter surgical intervention in pediatric patientsinclude neuromuscular disorders, chromosomeabnormalities, a history of restless sleep, diffi-culty in breathing while asleep, loud snoringwith apnea, and an upper respiratory tract infec-tion within 4 weeks of surgery.31 All of the afore-mentioned risk factors warrant consideration forovernight observation after surgery.

Recovery Room and Postoperative ObservationPostoperatively, patients are typically observedin the recovery room for 30 to 45 minutes oruntil they are sufficiently stable for transfer tothe ambulatory unit. Every ambulatory centerhas its own discharge criteria. The most commoncriteria include requiring demonstration ofcontrol of postoperative nausea and/or emesisand adequate pain control as well as an absence


Ch22.qxd 1/25/05 3:28 AM Page 255

of obstructive snoring, oxygen requirements,and other signs of respiratory compromise. Forpatients under 3 years of age as well as thosewith an increased risk of respiratory compro-mise, the signs of respiratory compromise mayoccur at any time within the first 12 to 24 hoursafter the procedure but commonly occur in therecovery room. Young children and those withsevere OSA are more likely to exhibit apneicepisodes, stridor, an inability to maintain oxygensaturation over 90% despite supplemental oxy-gen, and upper airway obstruction requiring anartificial airway such as a nasopharyngeal trum-pet or even endotracheal intubation. Inadequateair exchange can lead to CO2 retention and nar-cosis, which further depress respiratory drive.This must be promptly recognized during thepostoperative period and may be confirmed byusing transcutaneous CO2 monitoring. Suchpatients may require reintubation and respira-tory support by means of a ventilator for 24 to 48hours but are usually successfully extubatedwith an otherwise uneventful recovery period.

After fulfilling criteria in the recovery room,patients are typically transferred to an observa-tion room for up to 4 hours. Attention to ade-quate analgesia, stable vital signs, the ability tovoid, and adequate oral intake determine whethera patient may be discharged home.

Patients are usually discharged home with anarcotic or non-narcotic analgesic and instruc-tions for postoperative care. Those with severeOSA on preoperative PSG should be consideredfor postoperative PSG 6 weeks or later after sur-gical intervention to ensure that no further treat-ment is necessary. Postoperative PSG should alsobe considered in patients who have persistent orrecurrent symptoms of upper airway obstructionand sleep-related breathing disorders after surgi-cal intervention.

Postoperative use of antibiotics and analgesics,with or without narcotics, is based on the prefer-ences of the surgeon. Narcotics (e.g., codeine andoxycodone) should be used judiciously post-operatively, especially in children younger than3 years of age or those with neurologic disorders,medical comorbidities, or documented severeOSA. A recent study demonstrated equal efficacybetween the administrations of ibuprofen andacetaminophen with codeine for pain control.46

Another study demonstrated no difference inpain control and higher oral intake in the group

that received acetaminophen alone, most likelydue to the lack of potential nausea and vomitingand gastric irritation from the codeine.47 Inorder to maximize the benefits of acetaminophenand minimize the potential side effects of a nar-cotic, the senior author uses a scheduled dose ofacetaminophen every 4 hours (15 mg/kg) for thefirst few days after surgery. Hydrocodone is avail-able without acetaminophen in many coughsyrup formulations. A few of the formulations areliquid without an alcohol base. One of these isprescribed (dosing of hydrocodone at 0.15mg/kg) to be given every 4 hours but onlyif needed for breakthrough pain that is not ade-quately relieved by the acetaminophen.

POTENTIAL COMPLICATIONSAFTER ADENOTONSILLECTOMY

Complications after adenotonsillectomy may bedivided into immediate, short-term, and longterm categories.48 Primary and secondary hem-orrhages after surgery are reported to be smallbut potentially significant risks in mostseries.34,36,37,49 Primary hemorrhage occurswithin 24 hours after surgery, while secondaryhemorrhage usually occurs between postopera-tive days 5 and 10. There has been no identifi-able, statistically significant risk factor forpredicting post-tonsillectomy hemorrhage, withthe exception of one study in which age greaterthan 21 was found to be associated with anincrease in risk factor of 3%, almost twice that ofthose younger than 21 years of age.49 Post-obstructive pulmonary edema results from achange in pulmonary hydrostatic pressure afteran increase in intrathoracic pressure that devel-ops after the removal of long-term upper airwayobstruction. Other complications in the immedi-ately postoperative period include dehydrationsecondary to pain and decreased oral intake,nausea and vomiting, atelectasis, aspiration, andtemporary eustachian tube dysfunction or otalgiadue to referred pain. Parents are typicallyinstructed to aggressively encourage the oralintake of liquids, since it is speculated thathydration reduces hypertension as well as dryoropharyngeal mucosa and possibly traumaticeschar removal that may cause postoperativehemorrhage. Parents are also encouraged to keepthe patients from being too physically activeduring the first few days after surgery as a


Ch22.qxd 1/25/05 3:28 AM Page 256

measure to prevent hypertension, which mayalso cause postoperative hemorrhage.

Velopharyngeal insufficiency (VPI) mayoccur in the first few weeks to months afteradenotonsillectomy and is more likely to occurin patients with a history of cleft palate, submu-cosal cleft palate, or a neuromuscular disorderwith hypotonia. However, overaggressive resec-tion of superior tonsillar pillars may cause short-ening of the palate and may also produce thesymptoms of VPI. Patients and/or parents mayreport hypernasality or fluid regurgitation intothe nasopharynx. Most cases of post-adenoton-sillectomy VPI are resolved with observation.For those with symptoms persisting for longerthan 6 months, referral to speech therapy may beindicated. VPI as a complication after adenoton-sillectomy could theoretically require surgicalintervention but is rarely reported.

Nasopharyngeal stenosis is a rare but signifi-cant long-term complication secondary to scar-ring of raw mucosal surfaces during healing.Surgical repair may be necessary to re-establishan adequate nasal airway.50

OUTCOMES OFADENOTONSILLECTOMY IN THEMANAGEMENT OF OSA AND UARS

Various parameters have been used in the evalu-ation and reporting of outcomes after adenoton-sillectomy for the treatment of UARS and OSA.In one study, PSG performed before and afteradenotonsillectomy demonstrated that surgerysignificantly decreased the number of obstruc-tive apneas and the average number of totalapneas during all sleep stages but did not have asignificant effect on the duration and proportionof the various sleep stages.51 One prospectivestudy demonstrated that for mild OSA, definedas less than 15 OSA events per hour, 15 patientswithout underlying medical comorbiditiesdemonstrated a reduction in the number ofobstructive events and significant improvementin oxygen desaturation during rapid eye move-ment sleep on the operative night.52 A retrospec-tive case series evaluated 48 patients bycomparing AHI, percent of sleep time with oxy-gen saturation below 90%, and percent of sleeptime with end-tidal CO2 of greater than 50 onthe pre- and postoperative polysomnograms.This study included a diverse population of

patients such as those with cerebral palsy, Downsyndrome, morbid obesity, and other syndromeswho underwent adenotonsillectomy alone, ade-notonsillectomy with uvulopalatopharyngo-plasty (UPPP), or tonsillectomy as a part of UPPPonly. Statistically significant differences wereobserved in all three parameters after surgicalintervention.53

Before the availability and widespread use ofPSG for the evaluation of OSA, adenotonsillec-tomy was shown to have far-ranging benefits,with a reduction in mouth breathing as well asimprovement in behavioral problems evenwhen the upper airway obstruction did notdemonstrate severe OSA preoperatively.19 Thisprospective study used pre- and postoperativesleep sonography (recorded respiratory sounds)during sleep as well as parental questionnairesto compare 100 children with a diagnosisof adenotonsillar hypertrophy with 50 age-matched control subjects. Other improvementsreported include reduced dry mouth and halito-sis, decreased mouth breathing, and decreasedfussy morning behavior and daytime fatigue. Alldifferences in parameters were shown to be sta-tistically significant when comparing the out-comes of patients postoperatively with theirsymptoms preoperatively or with the controlsubjects.

In a review of 55 children with OSA treatedwith adenotonsillectomy, 86% of the patientsmarkedly improved, defined as a 75% reductionin AHI; an 8% demonstrated improvement,defined as a 50% to 74% reduction in AHI; and a3% demonstrated (slight) improvement, definedas a 25% to 49% reduction in AHI. Only onepatient had less than a 25% reduction in AHIpostoperatively.54 In a different review of 134children with OSA who underwent adenoidec-tomy and/or tonsillectomy, facial morphologywas correlated with adenotonsillar hypertrophy.Significant improvement in AHI and the lowestoxygen saturation (LSAT) was found in 77.6% ofpatients. Those who did not demonstrateimprovement tended to have smaller tonsils, nar-rower epipharyngeal spaces, and more poorlydeveloped maxillary and mandibular protru-sions.55 This study also stratified the patients intoage groups, and while the LSAT was improved inall age groups, significant improvements in AHIwere noted in patients 1 to 3 years of age and4 to 6 years of age but not in those 7 to 9 years


Ch22.qxd 1/25/05 3:28 AM Page 257

of age. There were no statistically significant dif-ferences in improvement among members withinan age group.

Adenotonsillectomy has been shown to effec-tively reverse cardiac changes in children withadenotonsillar hypertrophy by comparing thepre- and postoperative echocardiograms ofpatients with OSA as well as the postoperativeechocardiography parameters with a controlgroup.56 They used echocardiography to comparechildren with adenotonsillar hypertrophy andsleep-related breathing disorders with a controlgroup and found statistically significant differ-ences in right and/or left ventricular enlargementas well as decreased left ventricular compliance.

In a review of 31 children with obesity and anexpected body weight range of 130% to 260%,adenotonsillectomy was shown to decreaseirregular breathing periods to almost zero andresult in a mean of over 95% of the sleepingperiod during which oxygen saturation is greaterthan 90% in all patients.3 Weight loss in thesepatients demonstrated a partial cure of sleep-related breathing disorders, while adenotonsil-lectomy was effective even when severe obesityremained.

While an adenotonsillectomy with or withoutadditional surgery is expected to improve OSA inchildren with neuromuscular disorders, thesepatients are prone to increased postoperativeairway complications and prolonged postopera-tive hospitalization and recovery. Such patientsare likely to require insertion of a nasopharyngealairway to maintain a patent upper airway, meticu-lous secretion management, and nasogastric feed-ing temporarily due to delayed and inadequateoral intake.57 Avoidance of peri- and postopera-tive use of narcotics will decrease the depressionof hypoxic respiratory drive. An awareness ofincreased potential for postoperative complica-tions allows adequate discussion and preparationof the parents and family before surgery.

ADDITIONAL SURGICALPROCEDURES

Patients with residual symptoms suggestive ofupper airway obstruction after adenotonsillec-tomy, as well as those with a preoperativepolysomnogram and an AHI of greater than 20,should be considered for a second polysomno-gram typically 4 to 6 weeks after surgery. The

time period within which a second polysomno-gram is performed depends on the availability ofthe study center and time to complete recoveryfrom the adenotonsillectomy. Repeated PSG willagain determine the severity of the airway obstruc-tion and whether it is central or still obstructive,and the decision can then be made whether topursue additional surgical intervention or med-ical therapy such as CPAP or BiPAP. Outcomestudies have specifically focused on patientswith medical comorbidities that increase thelikelihood of having residual OSA despite stan-dard adenotonsillectomy. Such conditionsinclude Down syndrome, cerebral palsy, cranio-facial syndromes, and neuromuscular disordersother than cerebral palsy. While it is beyondthe scope of this chapter to discuss in detail thesurgical techniques that are used more fre-quently in adults for the treatment of OSA, theoutcomes of aggressive surgical intervention inthese patients are reviewed.

In patients with Down syndrome, adenoton-sillectomy as part of an aggressive managementprotocol including tongue reduction, tonguehyoid advancement, UPPP, and maxillary ormandibular advancement demonstrated improve-ment in the postoperative apnea index, respira-tory disturbance index, and LSAT.58 Oneprospective study evaluated the age-relatedoutcomes of soft tissue and skeletal sleep apneasurgery for infants and children with severeOSA refractory to conservative medical and sur-gical measures.59 The diagnoses among thesepatients included cerebral palsy, Down syn-drome, Pierre Robin syndrome, and other cran-iofacial anomalies. This study demonstratedthat children over 36 months of age demon-strated significant improvement in the respira-tory disturbance index, apnea index, and LSATpostoperatively, whereas only the respiratorydisturbance index improved significantly inthose between 12 and 36 months of age. Forchildren under 12 months of age, althoughthere was a trend toward improvement in respi-ratory indexes, these patients experienced longerhospital stays, a greater mean number of extuba-tion attempts, and a higher surgical failure ratethan the older patients. The importance of amultidisciplinary approach and soft tissue/skele-tal surgeries with adenotonsillectomy has beenshown in several studies of various patientgroups, specifically those who had residual OSA


Ch22.qxd 1/25/05 3:28 AM Page 258

after medical management and adenotonsillec-tomy. For 15 of the 18 patients with cerebralpalsy and OSA, aggressive surgical treatmentallowed sparing of tracheostomy as the defini-tive procedure for OSA.58 This study reported apostoperative reduction in the apnea index, res-piratory disturbance index, and LSAT, whichwere all statistically significant compared withthe same parameters on preoperative PSG.

UvulopalatopharyngoplastyUPPP may be necessary in children with longuvulae and redundant pharyngeal soft tissuewith or without adenotonsillar hypertrophy.This surgical procedure has been shown tobe effective in patients with OSA who also haveneurologic disorders.60,61 This is a much moreprevalent treatment option in adults, since long-term or severe sequelae such as a speech abnor-mality or VPI from uvulectomy and/or UPPPhave not been adequately studied in the pedi-atric population. In children whose anterior andposterior tonsillar pillars are prominent aftertonsillectomy, the surgeon may consider sutureapproximation of the two pillars to tightenredundant pharyngeal mucosa, which may con-tribute to the upper airway obstruction, at thetime of the tonsillectomy. Uvulectomy mayrarely be performed independent of UPPP inconjunction with the adenotonsillectomy ifit is found to be significantly elongated or ede-matous. As in adult patients, the main compli-cation to avoid in this procedure is the overlyaggressive resection of the soft palate, leadingto VPI.

Septoplasty, Turbinate Reduction,and/or Sinus SurgeryThe aforementioned procedures are all aimed atimproving only the nasal airway and are per-formed when appropriate and necessary. Thepediatric nasal septum is mostly cartilaginous,and injudicious surgical intervention may havea negative impact on long-term facial growth.Nasal septal dislocation secondary to traumaticbirth may occur in the neonate, and closedreduction may be effectively done at the bedsideafter prompt diagnosis. Otherwise, septoplastyfor a deviated nasal septum is rarely performedin the pediatric population, with the exception

of individual consideration given to those whohave significant internal nasal deformities caus-ing obstruction most commonly associated withnasal septal traumatic fractures. In summary,nasal surgery for the treatment of OSA in thepediatric population is uncommon.

Hypopharyngeal AirwayExpansion–Related ProceduresSuch procedures include geniohyoid advance-ment/expansion, tongue reduction, lingual ton-sillectomy, lingual suspension, sliding genioplasty,and maxillary/mandibular distraction. For thepediatric population, with their relatively smallercraniofacial dimensions, the most commonalternative surgical technique of relevance ismaxillary/mandibular advancement for thosewith maxillary/mandibular hypoplasia causingsignificant pharyngeal airway compromise. Withthe improvement of external and internal dis-traction devices, such procedures may be criticalin relieving or improving OSA and provide theopportunity for the patient to avoid a tra-cheostomy altogether or facilitate successfuldecannulation of patients who had previouslyrequired tracheostomy to relieve severe upperairway obstruction.62 Again, with the exceptionof maxillary and mandibular distraction tech-niques, the aforementioned procedures are morerelevant for treating adults with residual OSAafter standard procedures such as septoplastyand UPPP.

TracheostomyTracheostomy remains the most reliable andsignificant long-term surgical intervention forOSA refractory to all other interventions.Tracheostomy enables the complete bypass of alllevels of upper airway obstruction and allowsthe use of night-time supplemental oxygen orventilatory support for those patients requiringit. Tracheostomy is indicated if patients cannottolerate consistent and successful CPAP/BiPAPuse or if documented OSA persists on PSGdespite more conservative surgical proceduressuch as adenotonsillectomy and UPPP. Childrenwho have associated craniofacial anomalies,severe neurologic impairment such as cerebralpalsy, and/or chronic pulmonary disease maymeet multiple criteria for tracheostomy. This


Ch22.qxd 1/25/05 3:28 AM Page 259

procedure may be necessary at the time of orbefore pharyngeal or craniofacial surgery forairway protection, with subsequent decannula-tion after resolution of OSA as evaluated byPSG. The decision to perform tracheostomy isnot taken lightly. Despite the best of tra-cheostomy care, there are associated complica-tions including accidental decannulation,central apnea due to loss of hypoxic respiratorydrive, and mucous plugging resulting in venti-latory insufficiency.62 Tracheostomy involveslong-term equipment cost and significant levelsof education for all those involved in the care ofthe patient.

FOLLOW-UP

For patients whose symptoms of OSA do notresolve, those with an AHI of greater than 20preoperatively, and those with medical comor-bidity including persistent or worsening obesity,follow-up PSG is imperative. While residual,severe OSA may be found even in asymptomaticchildren who are otherwise healthy, it is muchmore likely in those with pre-existing neurologicimpairment. The decision to perform long-termmonitoring with intermittent PSG should bemade based on the clinical symptoms and his-tory of OSA severity and may be necessarybecause the recurrence of OSA is possible andrisk factors leading to recurrent OSA have notbeen well studied.

References

1. Rosenfeld R, Green R: Tonsillectomy and ade-noidectomy: Changing trends. Ann Otol RhinolLaryngol 1990;99:187-191.

2. Arnold J, Allphin A: Sleep apnea in the neurolog-ically-impaired child. Ear Nose Throat J 1993;72:80-81.

3. Kudoh F, Sanai A: Effect of tonsillectomy and ade-noidectomy on obese children with sleep-associ-ated breathing disorders. Acta Otolaryngol Suppl(Stockh) 1996;523:216-218.

4. Demain JG, Goetz DW: Pediatric adenoidal hyper-trophy and nasal airway obstruction: Reductionwith aqueous nasal beclomethasone. Pediatrics1995;95:355-364.

5. Phipps CD, Wood WE, Gibson WS, et al:Gastroesophageal reflux contributing tochronic sinus disease in children: A prospectiveanalysis. Arch Otolaryngol Head Neck Surg2000;126:831-836.

6. Plaxico DT, Loughlin GM: Nasopharyngeal refluxand neonatal apnea. Am J Dis Child 1981;135:793-794.

7. Wei JL, Kasperbauer JL, Weaver AL, et al: Efficacyof single-dose dexamethasone as adjuvant therapyfor acute pharyngitis. Laryngoscope 2002;112:87-93.

8. Sclafani AP, Ginsburg J, Shah MK, et al: Treatmentof symptomatic chronic adenotonsillar hypertro-phy with amoxicillin/clavulanate potassium:Short- and long-term results. Pediatrics 1998;101:675-681.

9. Al-Ghamdi SA, Manoukian JJ, Morielli A, et al: Dosystemic corticosteroids effectively treat obstruc-tive sleep apnea secondary to adenotonsillarhypertrophy? Laryngoscope 1997;107:1382-1387.

10. Goldstein NA, Sculerati N, Walsleben JA, et al:Clinical diagnosis of pediatric obstructive sleepapnea validated by polysomnography. OtolaryngolHead Neck Surg 1994;111:611-617.

11. Leach J, Olson J, Hermann J, et al: Polysomno-graphic and clinical findings in children withobstructive sleep apnea. Arch Otolaryngol HeadNeck Surg 1992;118:741-744.

12. Nieminen P, Tolonen U, Lopponen H, et al:Snoring children: Factors predicting sleep apnea.Arch Otolaryngol Suppl 1997;529:190-194.

13. Suen J, Arnold J, Brooks L: Adenotonsillectomyfor treatment of obstructive sleep apnea in chil-dren. Arch Otolaryngol Head Neck Surg1995;121:525-530.

14. Wang RC, Elkin TP, Keech D, et al: Accuracy ofclinical evaluation in pediatric obstructive sleepapnea. Otolaryngol Head Neck Surg 1998;118:69-73.

15. Ali N, Pitson D, Stradling J: Natural history of snor-ing and related behaviour problems between theages of 4 and 7 years. Arch Dis Child 1991;71:74-76.


17. Eliaschar I, Lavie P, Halperin E, et al: Sleep apneicepisodes as indications for adenotonsillectomy.Arch Otolaryngol 1980;106:492-496.


19. Potsic WP, Pasquariello PS, Baranak CC, et al:Relief of upper airway obstruction by adenoton-sillectomy. Otolaryngol Head Neck Surg1986;94:476-480.

20. Manning SC: Coagulation profile as a predictorfor post-tonsillectomy and adenoidectomy (T +A) hemorrhage. Int J Pediatr Otorhinolaryngol1995;32:261-263.

21. Koltai PJ, Solares CA, Mascha EJ, et al:Intracapsular partial tonsillectomy for tonsillar


Ch22.qxd 1/25/05 3:28 AM Page 260

hypertrophy in children. Laryngoscope 2002;112(8 pt 2):17-19.

22. Wiatrak BJ, Willging JP: Harmonic scalpel for ton-sillectomy. Laryngoscope 2002;112(8 pt 2):14-16.

23. Plant RL: Radiofrequency treatment of tonsillarhypertrophy. Laryngoscope 2002;112(8 Pt 2):20-22.

24. Hulcrantz E, Linder A, Markstrom A: Tonsillectomyor tonsillotomy? A randomized study comparingpostoperative pain and long-term effects. Int JPediatr Otorhinolaryngol 1999;51:171-176.

25. Lassaletta L, Martin G, Villafruela MA, et al:Pediatric tonsillectomy: Post-operative morbiditycomparing microsurgical bipolar dissection versuscold sharp dissection. Int J Pediatr Otorhino-laryngol 1997;41:307-317.

26. Linder A, Markstrom A, Hulcrantz E: Using thecarbon dioxide laser for tonsillectomy in children.Int J Pediatr Otorhinolaryngol 1999;50:31-36.

27. Pizzuto MP, Brodsky L, Duffy L, et al: A compari-son of microbipolar cautery dissection to hotknife and cold knife cautery tonsillectomy. Int JPediatr Otorhinolaryngol 2000;52:239-246.

28. Smith PS, Orchard PJ, Lekas MD: Predictingbleeding in common ear, nose, and throat proce-dures: A prospective study. R I Med J 1990;73:103-106.

29. Weimert TA, Babyak JW, Richter HJ: Electro-dissection tonsillectomy. Arch Otolaryngol HeadNeck Surg 1990;116:186-188.

30. Wexler DB: Recovery after tonsillectomy:Electrodissection vs. sharp dissection techniques.Otolaryngol Head Neck Surg 1996;114:576-581.

31. Wiatrak BJ, Myer CM, Andrews T: Complicationsof adenotonsillectomy in children under 3 yearsof age. Am J Otolaryngol 1991;12:170-172.

32. Steward DL, Welge JA, Myer CM: Do steroidsreduce morbidity of tonsillectomy? Meta-analysisof randomized trials. Laryngoscope 2001;111:1712-1718.

33. Tom LW, Templeton JJ, Thompson ME, et al:Dexamethasone in adenotonsillectomy. Int J PediatrOtorhinolaryngol 1996;37:115-120.

34. Colclasure JB, Graham SS: Complication of outpa-tient tonsillectomy and adenoidectomy: A reviewof 3,340 cases. Ear Nose Throat J 1990;69:155-60.

35. Gabalski EC, Mattucci KF, Setzen M, et al:Ambulatory tonsillectomy and adenoidectomy.Laryngoscope 1996;106(1 pt 1):77-80.

36. Lalakea ML, Marquez-Biggs I, Messner AH: Safetyof pediatric short-stay tonsillectomy. ArchOtolaryngol Head Neck Surg 1999;125:749-752.

37. Nicklaus PJ, Herzon FS, Steinle EW IV: Short-stayoutpatient tonsillectomy. Arch Otolaryngol HeadNeck Surg 1995;121:521-524.

38. Postma DS, Folsom F: The case for an outpatient“approach” for all pediatric tonsillectomies and/or

adenoidectomies: A 4-year review of 1419 cases ata community hospital. Otolaryngol Head NeckSurg. 2002;127:101-108.

39. Reiner SA, Sawyer WP, Clark KF, et al: Safety ofoutpatient tonsillectomy and adenoidectomy.Otolaryngol Head Neck Surg. 1990;102:161-168.

40. Rothschild MA, Catalano P, Biller HF: Ambulatorypediatric tonsillectomy and the identification ofhigh-risk subgroups. Otolaryngol Head NeckSurg 1994;110:203-210.

41. McColley S, April M, Carroll J, et al: Respiratorycompromise after adenotonsillectomy in childrenwith obstructive sleep apnea. Arch OtolaryngolHead Neck Surg 1992;118:940-943.

42. Bower C, Richmond D: Tonsillectomy and ade-noidectomy in patients with Down syndrome. IntJ Pediatr Otorhinolaryngol 1995;33:141-148.

43. Rosen GM, Muckle RP, Mahowald MW, et al:Postoperative respiratory compromise in childrenwith obstructive sleep apnea syndrome: Can it beanticipated? Pediatrics 1994;93:784-788.

44. Gerber ME, O’Connor DM, Adler E, et al: Selectedrisk factors in pediatric adenotonsillectomy. ArchOtolaryngol Head Neck Surg 1996;122:811-814.

45. Tom L, DeDio R, Cohen D, et al: Is outpatient ton-sillectomy appropriate for young children?Laryngoscope 1992;102:277-280.

46. St. Charles CS, Matt BH, Hamilton MM, et al: Acomparison of ibuprofen versus acetaminophenwith codeine in the young tonsillectomy patient.Otolaryngol Head Neck Surg 1997;117:76-82.

47. Moir MS, Bair E, Shinnick P, et al: Acetaminophenversus acetaminophen with codeine after pedi-atric tonsillectomy. Laryngoscope 2000;110:1824-1827.

48. Johnson LB, Elluru RG, Myer CM: Complicationsof adenotonsillectomy. Laryngoscope 2002;112(8 pt 2):35-36.

49. Wei JL, Beatty CW, Gustafson RO: Evaluation ofposttonsillectomy hemorrhage and risk factors.Otolaryngol Head Neck Surg 2000;123:229-235.

50. McDonald TJ, Devine KD, Hayles AB: Naso-pharyngeal stenosis following tonsillectomy andadenoidectomy: Report of six cases and theirrepair. Arch Otolaryngol 1973;98:39-41.

51. Frank Y, Kravath RE, Pollak CP, et al: Obstructivesleep apnea and its therapy: Clinical andpolysomnographic manifestations. Pediatrics1983;71:737-742.

52. Helfaer MA, McColley SA, Pyzik PL, et al:Polysomnography after adenotonsillectomy inmild pediatric obstructive sleep apnea. Crit CareMed 1996;24:1323-1327.

53. Wiet GJ, Bower C, Seibert R, et al: Surgical correc-tion of obstructive sleep apnea in the complicatedpediatric patient documented by polysomnography.Int J Pediatr Otorhinolaryngol 1997;41:133-143.


Ch22.qxd 1/25/05 3:28 AM Page 261

54. Nishimura T, Morishima N, Hasegawa S, et al:Effect of surgery on obstructive sleep apnea. ActaOtolaryngol Suppl (Stockh) 1996;523:231-233.

55. Shintani T, Asakura K, Kataura A: The effect ofadenotonsillectomy in children with OSA. Int JPediatr Otorhinolaryngol 1998;44:51-58.

56. Gorur K, Doven O, Unal M, et al: Preoperativeand postoperative cardiac and clinical findings ofpatients with adenotonsillar hypertrophy. Int JPediatr Otorhinolaryngol. 2001;59:41-46.

57. Grundfast K, Berkowitz R, Fox L: Outcome andcomplications following surgery for obstructiveadenotonsillar hypertrophy in children with neu-romuscular disorders. Ear Nose Throat J 1990;69:756, 759-760.

58. Cohen SR, Lefaivre JF, Burstein FD, et al: Surgicaltreatment of obstructive sleep apnea in neurolog-ically compromised patients. Plast Reconstr Surg1997;99:638-646.

59. Januszkiewicz JS, Cohen SR, Burstein FD, et al:Age-related outcomes of sleep apnea surgery ininfants and children. Ann Plast Surg 1997;38:465-477.

60. Derkay CS, Maddern BR: Innovative techniques foradenotonsillar surgery in children: Introduction andcommentary. Laryngoscope 2002;112(8 pt 2):2.

61. Kosko J, Derkay C: Uvulopalatopharyngoplasty:Treatment of obstructive sleep apnea in neurolog-ically impaired pediatric patients. Int J PediatrOtorhinolaryngol 1995;32:241-246.

62. Cohen SR, Suzman K, Simms C, et al: Sleep apneasurgery versus tracheostomy in children: Anexploratory study of the comparative effects onquality of life. Plast Reconstr Surg 1998;102:1855-1864.


Ch22.qxd 1/25/05 3:28 AM Page 262

PRIMARY SNORING

Primary snoring (PS) describes nightly or fre-quent snoring that is not associated with apnea,hypoventilation, or sleep fragmentation.1 By def-inition, the diagnosis is made by polysomnogra-phy, with objective measurement of sleep andrespiratory function. These criteria differentiatePS from obstructive sleep apnea syndrome(OSAS), which is associated with various degreesof hypoxemia, hypercapnia, and sleep fragmenta-tion related to complete or partial upper airwayobstructive events.

Clinical history alone cannot differentiate PSfrom OSAS.2 In a study of 83 snoring childrenreferred to a tertiary pediatric sleep clinic, par-ents completed a standardized, nurse-adminis-tered questionnaire that asked questions re-garding snoring frequency, observed apnea,struggling to breathe, and other daytime andnighttime symptoms. Children then underwentnocturnal polysomnography to assess for sleep-disordered breathing. Although there were sev-eral differences in symptom frequencies betweenthe children with PS and those with OSAS, bothsingle and multiple questions showed poor sen-sitivity and specificity in discriminating betweenPS and OSAS.

UPPER AIRWAY RESISTANCE SYNDROME

Although PS is a polysomnographic diagnosis, adiagnosis of PS suggests that no adverse sleep-related or daytime sequelae are associated withsnoring in the absence of gas exchange abnor-malities or abnormal sleep architecture. In otherwords, some children with snoring have no inter-ruption of normal sleep patterns, no cardiorespi-ratory compromise, and no neuropsychologicalmorbidity. It follows that no treatment is

required. Because snoring is common, affectingapproximately 10% of children and up to 40% ofadults, the presence of snoring without adverseconsequences is likely to occur. However, discus-sion of PS as a distinct diagnosis is complicatedby the emergence of the upper airway resistancesyndrome (UARS) as a distinct clinical entity.

UARS is defined as partial upper airwayobstruction that is not associated with gasexchange abnormalities but is accompanied byincreased respiratory effort (conventionally meas-ured by changes in intrathoracic pressure viaesophageal manometry) terminated by electroen-cephalographic arousal; the primary symptom inadults is daytime somnolence.3 Snoring occurs inmost affected individuals, but physiologic find-ings of UARS have been noted in patients withoutsnoring, especially in those who have had palatalsurgery for upper airway obstruction. UARS isassociated with increased upper airway collapsi-bility during sleep.4 Guilleminault and colleaguesinitially reported the clinical and polysomno-graphic characteristics of UARS in 25 childrenwho were referred for evaluation of snoring,excessive daytime somnolence, and behavioralproblems.5 They demonstrated marked differ-ences in the referred group compared to 25healthy control children. Subsequently, numerousstudies have demonstrated these polysomno-graphic findings in adults with daytime somno-lence. Daytime somnolence is significantlyimproved with nasal continuous positive airwaypressure therapy, as objectively measured by themultiple sleep latency test.3

In adults, excessive daytime somnolence hasa number of significant sequelae, including anincreased risk of motor vehicle and work-relatedaccidents and impaired mood. Hypertension isalso frequently seen in adults with UARS.Children identified as having UARS have a vari-ety of daytime symptoms, including symptoms

Primary Snoring in ChildrenSusanna A. McColley 23

263

Ch23.qxd 1/25/05 3:30 AM Page 263

of attention deficit hyperactivity disorder andacademic problems. Most studies of snoringchildren have not defined or described UARS asa clinical entity separate from OSAS or PS, andmany have not performed detailed physiologicrecording. However, many studies have sug-gested significant neurocognitive abnormalitiesin habitually snoring children,6-8 includingattention deficit hyperactivity disorder, academicproblems, and behavioral problems. Availabledata do not allow clear separation betweenpatients with UARS and those with OSAS.

Although sleep physiologists often viewsleep-disordered breathing as a spectrum, withprimary snoring being the mildest and OSAS themost severe form,1 no studies demonstrate aclear relationship between the degree of sleep-disordered breathing and symptoms or physio-logic sequelae. Furthermore, although thediagnosis of primary snoring has traditionallybeen made on the basis of polysomnographicfindings alone, the available evidence suggeststhat absence of daytime symptoms should be anadditional diagnostic criterion.

EPIDEMIOLOGY

A number of epidemiologic studies from diversegeographic locations have been publisheddescribing the prevalence of snoring in children.Most of these have utilized parental report viaeither questionnaire or interview format. Mosthave not included objective measurements ofrespiration during sleep, or they have includedsuch measures only in a subset of children felt tobe at high risk for OSAS.

A summary of epidemiologic studies of snor-ing in children is presented in Table 23–1. Theprevalence of habitual snoring ranges from aminimum of 3.2% in Iceland to a maximum of12.1% in England. This contrasts with the 40%frequency of regular snoring in adults. Studiesthat have attempted to identify a subgroup ofchildren with OSAS show a prevalence of thisdisorder of between 0.7% and 2.9%. None ofthese epidemiologic studies have attempted todistinguish UARS from PS. Rosen19 studied 326otherwise healthy children with snoring. Fifty-nine percent of children had OSAS, 25% had PS,6% had UARS, and 10% had no snoring. It isnotable that 28% of the children in the studywere obese.

NATURAL HISTORY

Limited data are available regarding the naturalhistory of primary snoring in children. Marcusand coworkers20 repeated polysomnography in20 children 1 to 3 years after the initial diagno-sis of PS; none had undergone airway surgery.All of these children had persistent snoring; in20%, snoring had increased, and in 70% therewas no change. Overall, there was no change inapnea index, oxyhemoglobin saturation, or peakend-tidal PCO2. However, two children had mildOSAS on repeat testing. The authors concludedthat most children with PS do not progress tohaving OSAS, and those who do progress haveonly mild OSAS. Daytime symptoms in persist-ently snoring children were not reported.

Topol and Brooks21 studied nine children withprimary snoring 3 years after initial polysomno-graphic diagnosis of PS; a control group of nineage-matched, nonsnoring subjects were also stud-ied for comparison. As in the study by Marcus andcolleagues, there was no overall change in respi-ratory parameters between the first and secondpolysomnograms; one snoring subject had signif-icant worsening of the respiratory disturbanceindex. Interestingly, the control group had signif-icantly better sleep efficiency and fewer briefarousals than the snoring group, which suggeststhat some of the children with PS may actuallyhave been affected by UARS.

DIAGNOSIS

History and Physical ExaminationA sleep history that includes a question regard-ing nocturnal snoring should be included inhealth maintenance or well child care visits.22

The hallmark of primary snoring is nightly ornear-nightly snoring without physiologic or neu-rocognitive consequences. Once a history ofhabitual snoring is elicited, further history tak-ing and a physical examination serve to assesssleep-related or daytime symptoms that may beassociated with significant upper airwayobstruction during sleep. Children who havenumerous episodes of observed obstructiveapnea, daytime somnolence, and problems withbehavior, attention, or school performancerequire prompt referral for diagnostic testing andtreatment for OSAS or UARS.

264 Chapter 23 Primary Snoring in Children

Ch23.qxd 1/25/05 3:30 AM Page 264

Primary Snoring in Children Chapter 23 265

Firs

t A

uth

or

Num

ber

H

abit

ual S

no

rin

g†

OSA

S (Y

ear)

(Ag

e R

ang

e)C

oun

try

Met

ho

ds*

Prev

alen

cePr

eval

ence

Ali

(199

3)9

782

(4-5

yr)

Engl

and

Post

al q

uest

ionn

aire

; ho

me

over

nigh

t vi

deo

12.1

%0.

7%an

d ox

imet

ry in

a s

ubse

t of

“hi

gh r

isk”

an

d co

ntro

lA

li (1

994)

1050

4 (6

-7 y

r)En

glan

dFo

llow

-up

stu

dy o

f ch

ildre

n in

pre

cedi

ng s

tudy

; 11

.4%

Not

exa

min

edre

pea

t p

osta

l que

stio

nnai

re o

nly

Gis

laso

n (1

995)

1145

4 (6

mo-

6 yr

)Ic

elan

dPo

stal

que

stio

nnai

re;

over

nigh

t re

spira

tory

3.

2%2.

9%m

onito

ring

in c

hild

ren

susp

ecte

d of

hav

ing

slee

p a

pne

aTe

cule

scu

(199

2)12

190

(5-6

.4 y

r)Fr

ance

Inte

rvie

w q

uest

ionn

aire

10%

Not

exa

min

edC

orbo

(19

89)13

1615

(6-

13 y

r)It

aly

Self-

adm

inis

tere

d q

uest

ionn

aire

7.3%

Not

exa

min

edC

orbo

(20

01)14

2209

(10

-15

yr)

Ital

ySe

lf-ad

min

iste

red

que

stio

nnai

re5.

6%N

ot e

xam

ined

Hul

tcra

ntz

(199

5)15

500

(4 y

r)Sw

eden

Inte

rvie

w;

PSG

in c

hild

ren

with

hab

itual

sno

ring

6.2%

Dat

a in

com

ple

teSm

edje

(19

99)16

1844

(5-

7 yr

)Sw

eden

Self-

adm

inis

tere

d q

uest

ionn

aire

7.7%

Not

exa

min

edA

nunt

aser

ee (

2001

)1710

08 (

6-13

yr)

Thai

land

Self-

adm

inis

tere

d q

uest

ionn

aire

; PS

G in

pat

ient

s 8.

5%0.

7%sn

orin

g “m

ost

nigh

ts”

Ferr

eira

(20

00)18

976

(6-1

1 yr

)Po

rtug

alSe

lf-ad

min

iste

red

que

stio

nnai

re8.

6%N

ot e

xam

ined

*All

que

stio

nnai

re a

nd in

terv

iew

dat

a w

ere

obta

ined

fro

m p

aren

ts.

†Hab

itual

sno

ring

is d

efin

ed a

s fr

eque

nt o

r ni

ghtly

sno

ring,

or

snor

ing

at le

ast

3 ni

ghts

per

wee

k, in

the

abs

ence

of

upp

er r

esp

irato

ry in

fect

ion.

OSA

S, o

bstr

uctiv

e sl

eep

ap

nea

synd

rom

e; P

SG,

pol

ysom

nogr

aphy

.

Tab

le 2

3–1.

Epid

emio

log

y o

f Sn

ori

ng

in

Ch

ild

ren

Ch23.qxd 1/25/05 3:30 AM Page 265

The physical examination may be normal, orit may reveal signs of upper airway obstructionsuch as mouth breathing and tonsillar hypertro-phy. Dolichocephaly and midface hypoplasia areassociated with snoring and OSAS. Obesityappears to be a predisposing factor in both snor-ing and OSAS, whereas growth failure may sig-nify severe sleep-disordered breathing.

Nocturnal PolysomnographyPrimary snoring is defined by polysomnographiccriteria. To differentiate PS from OSAS or UARS,it is important to carefully evaluate sleep stagingand the frequency of electroencephalographicarousals. Normal polysomnographic findings,including the absence of increased arousal fre-quency or abnormal tachypnea during sleep,suggest PS. Although the gold standard for diag-nosis of UARS is measurement of esophagealpressure, this technique is invasive. Other tech-niques have been evaluated but are not in wide-spread clinical use. Therefore, an otherwiseasymptomatic child with snoring and a normalpolysomnogram can be assumed to have PS. Onthe other hand, a child with normal polysomno-graphic findings and abnormal symptoms mayhave UARS, and treatment should be considered.

TREATMENT

By definition, primary snoring requires no spe-cific treatment. Because of the rare progressionfrom primary snoring to OSAS, children shouldbe monitored clinically and reevaluated if symp-toms increase over time. Because pediatric sleep-disordered breathing is a spectrum, individualtreatment decisions must be based on individualsymptoms and physical findings, not onpolysomnographic findings alone.

FUTURE DIRECTIONS

Further studies of snoring in children areneeded. The wide variability in published preva-lence may be secondary to the influence of eth-nicity and environment in different populations;further definition of prevalence differenceswould be useful in identifying patients at riskand in public health policy decisions. Better def-inition of the clinical consequences of snoring,and specific comparisons between patients withPS and those with UARS, are needed. Longer-

term natural history studies are essential, includ-ing health and functional outcomes in adoles-cents and adults who had childhood snoring.

References

1. Greene MG, Carroll JL: Consequences of sleep-disordered breathing in childhood. Curr OpinPulm Med 1997;3:456-463.

2. Carroll JL, McColley SA, Marcus CL, et al:Inability of clinical history to distinguish pri-mary snoring from obstructive sleep apnea syn-drome in children. Chest 1995;108:610-618.

3. Exar EN, Collop NA: The upper airway resis-tance syndrome. Chest 1999;115:1127-1139.

4. Gold AR, Marcus CL, Dipalo F, Gold MS: Upperairway collapsibility during sleep in upper airwayresistance syndrome. Chest 2002;121:1531-1540.

5. Guilleminault C, Winkle R, Korobkin R,Simmons B: Children and nocturnal snoring:Evaluation of the effects of sleep related respira-tory resistive load and daytime functioning. EurJ Pediatr 1982;139:165-171.

6. Chervin R, Dillon J, Bassetti C, et al: Symptomsof sleep disorders, inattention, and hyperactivityin children. Sleep 1997;20:1185-1192.


8. Gozal D, Pope DW: Snoring during early child-hood and academic performance at ages thirteento fourteen years. Pediatrics 2001;107:1394-1399.

9. Ali NJ, Pitson DJ, Stradling JR: Snoring, sleepdisturbance, and behaviour in 4-5 year olds.Arch Dis Child 1993;68:360-366.

10. Ali NJ, Pitson D, Stradling JR: Natural history ofsnoring and related behaviour problems betweenthe ages of 4 and 7 years. Arch Dis Child 1994;71:74-76.

11. Gislason T, Benediktsdottir B: Snoring, apneicepisodes and nocturnal hypoxemia among chil-dren 6 months to 6 years old. Chest 1995;107:963-966.

12. Teculescu D, Caillier I, Perrin P, et al: Snoring inFrench preschool children. Pediatr Pulmonol1992;13:239-244.

13. Corbo GM, Fuciarelli F, Foresi A, De BenedettoF: Snoring in children: association with respira-tory symptoms and passive smoke. BMJ1989;299:1491-1494.

14. Corbo G, Forastiere F, Agabiti N, et al: Snoring in9- to 15-year-old children: Risk factors and clin-ical relevance. Pediatrics 2001;108:1149-1154.

15. Hultcrantz E, Lofstrand-Tidestrom B, Ahlquist-Rastad J: The epidemiology of sleep relatedbreathing disorder in children. Int J PediatrOtorhinolaryngol 1995;32(Suppl):S63-S66.

266 Chapter 23 Primary Snoring in Children

Ch23.qxd 1/25/05 3:30 AM Page 266

16. Smedje H, Broman J-E, Hetta J: Parents’ reportsof disturbed sleep in 5-7-year-old Swedish chil-dren. Acta Paediatr 1999;88:858-865.

17. Anuntaseree W, Rookkapan K, Kuasirikul S,Thongsuksai P: Snoring and obstructive sleepapnea in Thai school-age children: Prevalenceand predisposing factors. Pediatr Pulmonol2001;32:222-227.

18. Ferreira AM, Clemente V, Gozal D, et al: Snoringin Portuguese primary school children. Pedia-trics 2000;106(5). Available at http://www.pedi-atrics.org/cgi/content/full/106/5/e64.

19. Rosen CL: Clinical features of obstructive sleepapnea in otherwise healthy children. PediatrPulmonol 1999;27:403-409.

20. Marcus CL, Hamer A, Loughlin GM: Naturalhistory of primary snoring in children. PediatrPulmonol 1998;26:6-11.

21. Topol HI, Brooks LJ: Follow-up of primary snor-ing in children. J Pediatr 2001;138:291-293.

22. Section on Pediatric Pulmonology, Subcom-mittee on Obstructive Sleep Apnea Syndrome:Clinical practice guideline: Diagnosis and man-agement of childhood obstructive sleep apneasyndrome. Pediatrics 2002;109:704-712.

Primary Snoring in Children Chapter 23 267

Ch23.qxd 1/25/05 3:30 AM Page 267

Disorders of the brain are often associated withsevere sleep disturbances. Frequently, childrenwith neurologic disabilities experience chronicsleep–wake problems related to circadian timingof sleep, sleep-related seizure disorders, sleep-related movement disorders, and sleep-relatedbreathing disorders. Many respond poorly to tra-ditional therapeutic intervention. Establishedtreatments include behavioral management,chronotherapy, phototherapy, faded response-cost programs, sedatives, hypnotics, and antide-pressants. These are often unsuccessful in theyoungster with a chronic disabling condition ofthe central nervous system (CNS). Some thera-peutic approaches may even exacerbate symp-toms or result in respite for only a few days.

Patients present with symptoms that mayinclude profound sleep onset difficulties atdesired bedtimes, inability to consolidate sleep,inability to maintain sleep, irregular sleep–wakeschedules, rapidly changing sleep–wake sched-ules, obstructive sleep apnea syndrome, prob-lems with central control of breathing, seizures,movement disorders, and arousal disorders.Presence of multiple symptoms is the rule ratherthan exception. Sleep deprivation and fragmen-tation of sleep continuity occur, and consider-able performance problems, as well as delay inthe response to rehabilitative efforts, can result.Interestingly, these disorders of sleep and thesleep–wake cycle not only deeply affect thepatient and his/her quality of life but commonlyresult in sleep disturbances and decreased qual-ity of life for the entire family.

CORRELATES OF CENTRALNERVOUS SYSTEM DEVELOPMENT

The CNS is the principal organ system governingsleep, sleep’s components, and the sleep–wakecycle. Major CNS alterations occur throughout

fetal life, neonatal life, infancy, and childhood.Understanding these changes is essential inassessing the patient with CNS dysfunction dur-ing wakefulness and during sleep. Indeed, acomprehensive awareness of maturationalchanges during sleep may provide insight intomanagement.

Genetic and environmental factors are impor-tant in determining morphologic and electrophys-iologic development of the CNS. Differentiationbegins very early in the evolution of the embryo,with a thickening of the dorsal ectoderm into theneural plate. The cells in this single layer rapidlyincrease in number and stratify, and two foldsand a neural groove develop. This central groovefuses to become the neural tube, giving rise tothe substance of all neural elements whose cellbodies and supporting elements lie within thebrain and spinal cord.1

During regional differentiation of the CNS,structural flexure begins. Three regions can beidentified: cephalic flexure (region of the mid-brain), cervical flexure (junction of the brain andspinal cord), and pontine flexure (junction of themetencephalon and myelencephalon). The lumenof the neural tube undergoes dramatic changesduring this period of development that corre-spond to regional specialization. The lumen inthe area of the telencephalon will ultimatelybecome the lateral ventricles. The lumen withinthe telencephalon and diencephalon will becomethe third ventricle. The cerebral aqueduct devel-ops from the lumen in the mesencephalon. Thelumen of the metencephalon and myelen-cephalon becomes the fourth ventricle.

Neuronal activity appears to be important inthe migration of neurons to appropriate posi-tions within the CNS, the degree of dendriticbranching, and the strength of synaptic inter-connections.2 Mitosis and migration continuethroughout development, and completion of the

Sleep in Neurologic DisordersStephen H. Sheldon

Daniel G. Glaze

24

269

Ch24.qxd 01/28/05 10:31 AM Page 269

location of individual neurons occurs about 1year after postconception term. Two internalprocesses result in a high degree of neuronalactivity: the waking state and active (rapid eyemovement [REM]) sleep. It is possible that thesetwo states are important during prenatal andearly postnatal life for appropriate ultrastruc-tural development of the CNS.

Centers responsible for control of sleep andthe sleep–wake cycles are contained in areas thatdevelop from the diencephalon. Appropriatediencephalic maturation is essential for normalsleep to occur. All neuronal activity that eventu-ally reaches the cortex passes through the dien-cephalon, with the sole exception of thoseoriginating from olfaction. The third ventricle iscontained within the diencephalon. During theseventh week of development, a small evagina-tion appears from the caudal wall of the thirdventricle. This eventually becomes glandularand forms the pineal body that is responsible forsecretion of melatonin.

Melatonin plays an important role in regulat-ing the sleep–wake cycle, presumably throughentrainment to light–dark cycling. Secretion ishighly responsive to afferent neural activity viathe retinohypothalamic tract. Secretion increasesin a dark environment and decreases when theretinas are exposed to light. Although dataregarding the function of melatonin are conflict-ing, evidence exists that it affects the timing ofsleep through its effect on circadian organiza-tion.3 Exogenous melatonin has been noted tobe useful in regulating sleep in children withsome sleep disorders4 and in improving sleepin some neurologically handicapped children.5 Itseems likely, therefore, that disorders of develop-ment of the diencephalon, as well as acquireddisorders that affect development or function ofcells in the caudal wall of the third ventricle, canresult in significant sleep–wake disorders.

After the seventh postconception week, tha-lamic regions undergo differentiation, and neu-ronal fibers separate the massive gray matter ofthe walls of the thalamus into numerous tha-lamic nuclei. Similarly, the wall of the hypothal-amus contains hypothalamic nuclei, the opticchiasm, the suprachiasmatic nucleus, and theneural lobe of the stalk of the body of the pitu-itary gland. The hypothalamus eventuallybecomes the executive region for regulation ofall autonomic activity including core body tem-

perature, temperature regulation, and sleep.Because the suprachiasmatic nucleus becomesthe governing region for the circadian timing ofmany major physiologic functions (the biologi-cal clock), it seems clear that dysfunctionaldevelopment of, or injury to, the ventral regionof the diencephalon can result in profoundsymptoms related to the sleep–wake cycle.

The cerebral hemispheres become prominentduring the sixth postconception week. Theyexpand rapidly until they cover the diencephalonand mesencephalon. The telencephalon becomesthe most specialized and complex portion of thebrain and can be quite sensitive to changes inintrauterine environment. In the presence ofdecreased neuronal electrical activity secondaryto hypoxemia from any cause, abnormal concen-trations of cellular elements, decreased dendriticbranching, and a lack of the synaptic strengthneeded to develop essential and mature neuralnetworks may result.

DISORDERS OF MATURATIONALDEVELOPMENT

Culebras6 has comprehensively described neu-roanatomic and neurologic correlates of a widevariety of sleep abnormalities. Lesions of themedial mesencephalon almost invariably cause areduction in the level of alertness. Symptomaticcataplexy, characterized by active inhibition ofskeletal muscle tone, has been described inpatients with rostral brainstem tumors thatinvade the floor of the third ventricle.7 Disordersof the lower mesencephalon and upper ponstegmentum involving the region around thelocus ceruleus are responsible for symptoms ofREM sleep without atonia.8 Extensive pontinetegmental lesions cause a reduction in total sleeptime, alterations in or abolition of non-REM(NREM) sleep states and REM sleep, and paraly-sis of lateral gaze.9

Disorders involving the medullary regions ofthe CNS commonly affect respiratory centers.A wide variety of sleep-related breathing problemsare seen in youngsters with Arnold-Chiari mal-formation.10,11 Central apnea, increased periodicbreathing during REM and NREM sleep, centralhypoventilation syndrome, and prolonged expi-ratory apneas can occur. If motor centers con-trolling pharyngeal musculature are involved,obstructive sleep apnea may also be present.

270 Chapter 24 Sleep in Neurologic Disorders

Ch24.qxd 01/28/05 10:31 AM Page 270

Many other correlations can be identified.Hypothalamic lesions have been associated withhypersomnia. Diffuse lesions of the thalamuslead to either ipsilateral decrease or completeabolition of sleep spindles and represent a usefulelectrographic sign of thalamic abnormalities.12

The cerebral hemispheres, although not primor-dial in the generation or maintenance of NREMand REM sleep, do have a modulating influence.Patients with extensive cortical laminar necrosisfail to exhibit slow waves or spindles duringNREM sleep but can express cortical desynchro-nization during REM sleep.13 Finally, space-occupying lesions of the CNS may causesleep–wake disturbances or specific sleep disor-ders by virtue of their location. They may alsocause symptoms indirectly through the develop-ment of increased intracranial pressure, hydro-cephalus, or both.

POLYSOMNOGRAPHICCORRELATES

The clinical discipline of pediatric sleep medi-cine and the study of sleep disorders in infantsand children are becoming increasingly focusedon brain dysfunction. Study of the sleeping brainhas been termed by Culebras as “neurosomnol-ogy.”14 The physiologic functions of most otherorgan systems differ significantly from the wak-ing state, and there are clear ontogenetic changesthat occur in sleep and its structure. Studyinglongitudinal changes of multiple physiologicvariables during sleep in the laboratory mightbe termed developmental polysomnography.15

Evaluation of maturation of sleep within the con-text of normal and abnormal human developmentmight provide a sensitive method of analysis.

It has been shown that electroencephalogra-phy (EEG) is an excellent tool for measuringbrain maturation.16 Each conceptional agereveals a characteristic pattern. The importantfeatures of normal EEG ontogeny, therefore, tendto reflect normal development. An apparentdelay of the appearance of these EEG patternsmight reflect an arrest or a delay in maturation ofthe CNS. It has been proposed that close atten-tion to stages of brain maturation in normal andabnormal EEGs, as well as the normal progres-sion of state development during sleep, mightallow more accurate timing of brain insult ininfants with neurologic sequelae.

Comprehensive polysomnography utilizingan EEG array that provides greater detail thanthe standard montage recommended for adultpolysomnography is recommended for the neo-natal and pediatric patient. However, diagnosingontogenetic EEG variations must be performedwith caution, as abnormalities in the EEG reflectgeneral pathophysiologic processes but show lit-tle specificity for any particular disease.17

Other polysomnographic variables can beimportant in the assessment of developmentalmaturation. Adding eye movement recordingsand electromyography to the EEG mightimprove specificity. Recording of eye movementsduring sleep helps to identify the sleep state. Eyemovement density and bursts of saccades mayhold special significance in prediction of mentaldevelopment and morbidity secondary to neona-tal illness. Becker and Thoman18 evaluated theoccurrence of REM storms in newborn infantsand again at 3, 6, and 12 months of chronologicage. The number of rapid eye movements withineach 10-second interval of active sleep was ratedon a scale based on the frequency and intensityof the eye movements. Bayley scales of mentaldevelopment were administered to the cohort ofinfants at 12 months of age. Interestingly, a sig-nificant negative correlation was found betweenthe frequency of REM storms and Bayley scores.By 6 months of age, REM storms seemed toexpress dysfunction or delay in the develop-ment of central inhibitory feedback control forsleep organization and phasic sleep-relatedactivities.

The degree of phasic electromyographic activ-ity during sleep may also reflect maturity of thedeveloping brainstem. Gross movements, local-ized body movements, and phasic muscle activ-ity are controlled by the CNS at differentorganizational levels. Phasic motor activity isontogenetically simpler and decreases early dur-ing development. Gross movements are quitecomplex and require a greater degree of centralintegration. The type and frequency of muscleactivity during sleep might, therefore, add toinformation about the integrity of the CNS.Hakamada and coworkers19 studied varioustypes of motor activity in term newborns withsignificant illnesses. Generalized body move-ments, localized tonic movements, and general-ized phasic movements were evaluated. Patientswith minimally depressed EEG background

Sleep in Neurologic Disorders Chapter 24 271

Ch24.qxd 01/28/05 10:31 AM Page 271

activity showed an increase in generalized move-ments and localized tonic movements duringquiet sleep. In contrast, patients with markedlysevere EEG abnormalities showed an increase inphasic movements. It was concluded that a sig-nificant decrease in generalized body move-ments, or an increase in generalized phasicmuscle activity, might indicate a poor progno-sis for particular infants. However, the pres-ence of even small amounts of localized tonicmovements suggest preservation of corticalfunction. Nonetheless, diagnostic use of poly-somnography and its components becomes mostcost effective when applied to specific problems.

SLEEP DISORDERS IN INFANTSAND CHILDREN

Sleep disorders that occur in adults also occur inchildren. Disorders of sleep and the sleep–wakecycle differ from adult disorders in etiology,pathophysiology, morbidity, and treatment.Indeed, symptomatology can be dramatically dif-ferent, and childhood sleep disorders are fre-quently overlooked or overshadowed by clinicalproblems that appear and are evaluated duringthe day. It must be remembered that disorderedsleep can underlie meaningful daytime symp-toms and can exacerbate other medical disorders.

There is often considerable delay in diagnos-ing disordered sleep in the neonate, infant, orchild. Brouillette and colleagues20 described sig-nificant delays in the diagnosis of sleep disor-dered breathing and demonstrated profoundmorbidity. In 22 patients with documentedobstructive sleep apnea, mean delay in referralfor 20 patients first evaluated after the neonatalperiod was 23 ± 15 months. Almost three quar-ters of the patients studied developed serioussequelae including cor pulmonale, failure tothrive, permanent neurologic deficits, behavioraldisturbances, hypersomnolence, and develop-mental abnormalities.

The following paragraphs discuss commonprimary sleep-related disorders in children.These include obstructive sleep apnea, centralsleep apnea, and other breathing disorders thatoccur during sleep, sleep-related seizures, partialarousal disorders, movement disorders associatedwith sleep, and sleep–wake schedule disorders.Focus is placed on clinical presentation, labora-tory diagnosis, and management considerations.

Sleep-Disordered Breathingin Infants and ChildrenEight types of respiratory pauses or respiratorydysfunction during sleep can be described. Theclinical significance of each varies according tothe conceptional age of the youngster, the devel-opmental status, and the presence of other medicalor congenital abnormalities. Respiratory pausesinclude obstructive apnea, central apnea, mixedapnea, expiratory apnea, post-sigh apnea, andperiodic breathing. Central hypoventilationand obstructive hypoventilation may not beassociated with pauses in breathing but may sig-nificantly affect ventilatory function.

An obstructive apnea is defined as anabsence of nasal and oral airflow from the noseand mouth despite continuing chest and/orabdominal effort (Fig. 24–1). Lack of airflowmay be brief, lasting 6 seconds or less, depend-ing on the rate of respiration. If there are atleast two obstructed respiratory efforts, theapnea may be clinically significant. Obstructiveapnea can be prolonged and associated withcardiac deceleration (generally occurring dur-ing the last third of the apnea), oxygen desatu-ration, and elevation of end-tidal CO2 (ETCO2)after airflow resumes.

Definition of central apnea varies according tothe age of the patient. In premature and veryyoung infants, central apnea is defined as absenceof nasal and oral airflow and respiratory effortslasting 20 seconds or longer. It may also be clin-ically significant if shorter than 20 seconds butassociated with heart rate changes or oxygendesaturation (Fig. 24–2). Prolonged centralapnea, related to an absence of both inspiratoryand expiratory neuronal activity, appears to bequite unusual in otherwise normal infants andchildren during NREM/quiet sleep. Brief centralapnea is normal during REM/active sleep. Whenprolonged central apneas occur during quietsleep, central hypoventilation syndrome or a pri-mary underlying central nervous system orbrainstem abnormality may be present.

Mixed apnea contains polysomnographiccomponents of both central and obstructive res-piratory pauses (Fig. 24–3). There is some evi-dence, however, that the “central” component ofa mixed apnea is either prolonged expiration oran expiratory apnea.21 It may also represent acombination of both.


Ch24.qxd 01/28/05 10:31 AM Page 272

Expiratory apnea has received little attentionin the medical literature. It is defined as anabsence of polysomnographically recorded nasaland oral airflow in the presence of continuedexpiratory effort against an occluded or partiallyobstructed upper airway. A prolonged expiratoryphase of breathing can be demonstrated(Fig. 24–4). Almost all expiratory apneas arepreceded by an augmented breath, or a sigh.22

Upper airway occlusion may occur at any level,from the nasopharynx to the larynx. Preliminaryevidence in some children indicates that a reflexglottic adduction may occur in the presence ofcontinued diaphragmatic contraction. Increasedparasympathetic tone may be present, affectingthe diaphragm via the vagus nerve and increas-

ing vocal cord adduction via the recurrent laryn-geal nerve. A significant fall in the heart rateseems to occur during the first third of the res-piratory pause in a manner similar to that seenduring the Valsalva maneuver. There is littlechange in the arterial oxygen saturation (SaO2)regardless of the length of the apnea. This mightbe explained by a temporary increase in lungvolume and positive expiratory pressure. Thepotential importance of recognizing expiratoryapnea is evident when assessing apnea or brady-cardia during home monitoring of the infant, aswell as in some patients with underlying CNSand autonomic nervous system abnormalities.

Although expiratory apneas are typically pre-ceded by a sigh, post-sigh apneas seem to be


Figure 24–1. This 60-second epoch demonstrates a prolonged obstructive apnea during REM sleep.Note the significant decrease in airflow (>50% reduction) in the end-tidal carbon dioxide (ETCO2) flowchannel (19). There is continued and increasing respiratory effort in the chest and abdominal channels(20 and 22) and continued intercostal electromyographic activity (21). After the event, there is anarousal and a return to normal respiration. Note the shift in the phase angle of chest and abdomenprior to the apnea, during the apnea, and after the apnea. During the apnea, effort shifts 180 degrees,and paradoxical breathing can be noted. There is mild oxygen desaturation with spontaneous andrapid return to oxygen saturation baseline. There is also a mild elevation of the ETCO2 during recoverybreathing after the event. Numbers in parentheses indicate channels in the figure.

Ch24.qxd 01/28/05 10:32 AM Page 273

associated with little or no upper airwayobstruction (Fig. 24–5). An augmented breath isfollowed by an apnea that appears to be central.There is little change in heart rate, almost nooxygen desaturation, and an absence of hyper-carbia during and after the event. In certaincases, continued rise in ETCO2 can be demon-strated. Identification of these respiratory pausesis also important in evaluating infants withapnea and in follow-up of infants being moni-tored at home.

Periodic breathing is a pattern of respirationcharacterized by three or more respiratorypauses of 3 seconds or longer, separated by nor-mal breathing for less than 20 seconds.23

Periodic breathing is a normal pattern of respira-tion in premature infants during active/REM and

quiet/NREM sleep (Fig. 24–6). The proportionof the total sleep time occupied by periodicbreathing decreases as the youngster matures.Periodic breathing is most often confined toactive/REM sleep in the term newborn.Persistence later in infancy and childhood maybe abnormal and reflect immaturity or an abnor-mality of brainstem respiratory control.Unfortunately, clear norms regarding appropri-ate percentages of periodic breathing across agegroups are not yet available.

Obstructive Sleep Apnea Syndrome

The hallmark of obstructive sleep apnea in chil-dren is snoring. Snoring, however, can be absenteven in the presence of severe obstructive sleep


Figure 24–2. Central apnea with a sinus pause noted in the electrocardiogram (EKG). In this 30-sec-ond epoch recorded during non-REM slow-wave sleep, a somewhat prolonged central apnea is noted.Air ceases, and there is also cessation of effort demonstrated in the chest, intercostal electromyogra-phy, and abdominal effort channels. Note that airflow on this tracing was measured using an end-tidalcarbon dioxide (ETCO2) waveform. Because a side-stream method of analysis is required (with the sam-pling chamber in the instrument rather than directly in the air stream as seen in main-stream analy-sis), there is about a 3-second sampling delay. Effort seems to stop before flow stops, and effort beginsbefore flow begins. This is artifact secondary to this sampling delay in ETCO2 waveform, and the apneais central rather than mixed. Note the 1.5-second sinus pause in the EKG at the end of the event.Oxygen desaturation did not occur during this respiratory pause.

Ch24.qxd 01/28/05 10:32 AM Page 274

apnea. Breathing is often punctuated withpauses and snorts. Difficulty breathing duringsleep, restless sleep, diaphoresis, morningheadaches, excessive morning thirst, night-mares, sleep terrors, and enuresis may be associ-ated symptoms. Daytime abnormalities includesleepiness, hyperactivity, poor school perform-ance, abnormal behavior, aggressiveness, patho-logic shyness, and social withdrawal. Learningproblems, frequent upper airway infections, fail-ure to thrive, or obesity may occur. In severecases, pulmonary hypertension and cor pul-monale can develop.

Obstructive sleep apnea during childhood isoften associated with an anatomic abnormalityof the upper airway. The most common cause ofupper airway obstruction in children is hyper-trophy of the tonsils and adenoids.24 Malforma-tions of the mandible and maxillae can alsocause upper airway obstruction during sleep.Central and peripheral neurologic abnormalities

can also result in obstructive sleep apnea as aresult of dysfunction of pharyngeal muscularmovements.

The upper airway serves a variety of functionsduring the respiratory cycle, including protec-tion of the lower airway and phonation. Little isknown about the physiologic function and reflexactivity of the upper airway and larynx in breath-ing during sleep in neonates, and there is apaucity of studies regarding normal and abnor-mal pharyngeal and laryngeal respiratory func-tion in prepubertal children. During the normalrespiratory cycle, the laryngeal airway is widelypatent during inspiration, and it narrows duringexpiration.25 Changes in the glottic apertureduring the respiratory cycle result from phasicactivity of the posterior cricoarytenoid musclesand the pharyngeal constrictor muscles.26 Theposterior cricoarytenoid muscles function towiden the glottis and are the principal abductorof the vocal cords. Along with the pharyngeal


Figure 24–3. This 30-second epoch demonstrates a mixed apnea. It reveals a pause in respiratory effortaccompanied by a pause in airflow. However, at least two respiratory efforts (resumption of chesteffort, abdominal effort, and intercostal muscle electromyographic activity) occur prior to resumptionof airflow. The return of effort is significantly greater than the 3-second sampling delay of end-tidalcarbon dioxide (ETCO2). The event is followed by an arousal, but no electrocardiographic (EKG)changes occurred. This event, however, was associated with a fall in oxygen saturation (SaO2) greaterthan 4% from the baseline, and the nadir SaO2 reached 86%.

Ch24.qxd 01/28/05 10:32 AM Page 275

constrictor muscles, the lateral cricoarytenoid,oblique and transverse arytenoid, and thyroary-tenoid muscles function to close the glottis byadduction of the vocal cords. Controlled declinein the activity of the posterior cricoarytenoidmuscles acts in consort with a decrease in activ-ity of thoracic inspiratory muscles to cause expi-ratory braking, a reflex-controlled regulation ofexpiratory airflow and end expiratory volume.27

Appropriate function of the upper airway in reg-ulating the respiratory cycle is critical. There isincreasing evidence that the upper airway has todilate before the diaphragm initiates inspiration.Children with neurologic deficits or an auto-

nomic nervous system abnormality may be athigh risk for abnormalities of respiration.Obstructive apnea predominates in childrenwith neurologic abnormalities, but other types ofapnea often occur.

Apnea of Prematurity

Apnea of prematurity (AOP) is defined as exces-sive periodic breathing with pathologic apnea ina premature infant. Almost half of all prematureinfants manifest periodic breathing during theneonatal period. Periodic breathing occurs withgreater frequency as gestational age decreases


Figure 24–4. Although they are not frequently discussed, expiratory apneas do occur. There is appar-ent resumption of expiratory neuronal activity and effort despite the occlusion of the upper airway.This is often preceded by an augmented breath (sigh) or an arousal, or both. The event shown illus-trates several characteristic features. There is an augmented breath during arousal (electromyographyshows chin muscle increases, and there is a brief increase in the heart rate). During the immediatepostinspiratory period, there is persistence of expiratory flow as demonstrated by the prolonged alve-olar plateau on the end-tidal carbon dioxide (ETCO2) waveform. There is also an initial fall in theR-R interval, with return to the baseline as the event continues and then resolves. This is quite differ-ent from a postsigh respiratory pause and a true central apnea, because during these events there is aprogressive deceleration in the heart rate, followed by an increase in the heart rate during the arousalafter the event has ceased.

Ch24.qxd 01/28/05 10:32 AM Page 276

and is present in almost all newborns less than28 weeks gestation.28 AOP occurs in about halfof all infants with periodic breathing. It isthought to be related to immaturity of the brain-stem respiratory centers, central and peripheralchemoreceptors, and pulmonary reflexes.

Even though many respiratory pauses in thepremature infant are central, there is evidencesuggesting that half of all apneas in prematureinfants may be obstructive or mixed in origin.29

In these immature and often ill newborns, com-plex neuromuscular events required to maintainpharyngeal patency during respiration are easilyoverwhelmed.30

AOP is often responsive to medical treatment.Treatment of recurrent, clinically significantapnea in a premature infant should ensure ade-quate oxygenation and ventilation. Physical stim-ulation may be all that is necessary. Continuouspositive airway pressure (CPAP), supplementaloxygen, or mechanical ventilation may also berequired.

Methylxanthines are currently the mostwidely used medications in the treatment ofAOP.31 Theophylline has been shown to reducethe number of apneic episodes and lessen devel-opment of respiratory failure.32 Theophylline,however, does not shorten the clinical course ofAOP. Caffeine has also been shown to produce asignificant increase in ventilation, tidal volume,and mean inspiratory flow. Caffeine appears toincrease ventilation mainly by increasing centralinspiratory drive. Both caffeine and theophyllinehave similar effects on episodes of apnea; how-ever, caffeine seems to have an earlier effect onrespiratory rate. Side effects of tachycardia, arousal,and gastrointestinal intolerance are more fre-quently observed with theophylline when com-pared to caffeine.33 Theophylline has also beenshown to be associated with a decrease in cere-bral blood flow.34 Caffeine provides stableplasma levels and has a significantly longer plasmahalf-life, allowing the prescription of only onedaily maintenance dose.


Figure 24–5. This 30-second epoch demonstrates a postsigh respiratory pause. There is an aug-mented breath (sigh), with rapid return of the flow to the zero baseline. There is little change in theelectrocardiogram, and there are no other physiologic effects of this event.

Ch24.qxd 01/28/05 10:32 AM Page 277

Nasal CPAP is quite effective in resolvingobstructive apneas in premature and very younginfants. Central apneas, on the other hand, areunaffected by nasal CPAP, although oxygenationhas been shown to increase, whether or notapnea is present.35 For central apneas in thepremature infant, methylxanthine treatmentappears to be significantly superior to CPAP.36

Doxapram, a potent respiratory stimulant,has been used experimentally to treat AOP. Itappears to affect both peripheral chemoreceptorsand central respiratory centers.37 Doxapraminfusions have been used in preterm infantswhen therapeutic concentrations of theophyllinehad failed to control episodes of central apnea.38

Obstructive Hypoventilation and HighUpper Airway Resistance

Obstructive hypoventilation occurs when thereis chronic upper airway obstruction associatedwith high upper airway resistance. Overt apneamay not be present (Fig. 24–7). If obstructiveapneas and hypopneas occur, they tend to bemost significant during REM sleep. Often, how-ever, the number of apneas and or hypopneasper hour of sleep is within age-appropriate nor-mal limits. Respiratory rate during sleep is oftenincreased and oxygen saturation may be normal.ETCO2 is, however, significantly elevated andremains high because of upper airway occlusion.


Figure 24–6. This 120-second epoch demonstrates periodic breathing, defined as three or more res-piratory pauses that are at least 3 seconds long and that are separated by less than 20 seconds of nor-mal breathing. This is normal in the premature infant in both quiet sleep and active sleep. As thebabies are born closer to full term, periodic breathing tends to decrease in prevalence throughout thesleep period and is typically specific to active sleep. Prolonged periods of periodic breathing duringquiet sleep in the term or near-term infant might suggest the presence of a central control of breath-ing problem, and further clinical investigation may be indicated. Care must be taken that the baby wasactually asleep at the time the spells of periodic breathing were noted in the respiratory channels.Crying can appear very much like periodic breathing; technologist notes and video evaluation areextremely helpful in differentiating between the two.

Ch24.qxd 01/28/05 10:32 AM Page 278


A

B

Figure 24–7. A, This 120-second epoch recorded during slow-wave sleep demonstrates obstructivehypoventilation. Airflow appears normal, but the end-tidal carbon dioxide (ETCO2) is elevated togreater than 50 mm Hg. This represents alveolar hypoventilation. However, considerable snoring isnoted in this patient, and considerable apneas and hypopneas were noted during REM sleep.Therefore, the elevation of the ETCO2 is most likely caused by obstructive hypoventilation. B, This 30-second epoch, recorded from a patient who had suffered from hypoxic-ischemic encephalopathy,demonstrates continuous epileptiform activity of sleep. The record could not be adequately scored forNREM sleep state because persistent and consistent spike and wave activity was noted in NREM sleep.

Ch24.qxd 01/28/05 10:32 AM Page 279

It has been recommended that obstructivehypoventilation be considered when ETCO2measurements remain above 46 mm Hg for morethan 60% of the total sleep time or when theETCO2 rises above 53 mm Hg.39

Central Hypoventilation

Central hypoventilation is an uncommon disor-der manifested by insufficient ventilatory effortdue to inadequate output from the brainstemcenters that control breathing—that is, there is afailure of the automatic control of breathing.Hypoventilation in the absence of lung diseaseor dysfunction of respiratory muscles exacer-bates during sleep. Respiratory rate may appearnormal, but tidal volume during sleep isextremely low. On the other hand, some childrenwith central hypoventilation have normal orincreased tidal volumes but a significantlyreduced respiratory rate. Hypoventilation mayalso occur during wakefulness.

Infants with central hypoventilation syn-drome may appear otherwise normal. Spontane-ous breathing occurs, especially during sleep, orthe infant may breathe erratically. Whenmechanical ventilation is instituted, weaning isdifficult. Blunted response to hypercapnia andhypoxemia is often demonstrable.

Clinically, infants may present early in lifewith apparent apneic episodes. However, symp-toms may be manifested later in infancy andinclude cyanosis and pulmonary hypertension.The clinical course is chronic, and life-long ven-tilatory support is often required, particularlyduring sleep. Some infants improve with timeand may be able to support ventilation duringwakefulness. Others deteriorate over time andrequire 24-hour mechanical ventilatory support.

Diagnosis of Sleep-Disordered Breathing

Accurate diagnosis of respiratory pauses duringsleep is based on continuous monitoring of res-piratory, cardiovascular, and electroencephalo-graphic parameters across the child’s habitualsleep period. Severity of breathing disorders dur-ing sleep often varies in a circadian manner. Thehighest incidence of pathologic apnea, even invery young infants, occurs during the earlymorning hours.40 Polysomnographic monitoringis most reliable when performed during themajor sleep period at night and attended by a

technician. Premature newborns and termneonates may be evaluated adequately duringdaytime hours if several continuous sleep–wakecycles are monitored.

The parameters that should be monitoredcontinuously during polysomnography includenasal and oral airflow by capnography, chest andabdominal respiratory effort, oxygen saturation,and measurement of ETCO2. Oxygen saturation ismonitored using pulse oximetry. Continuouselectrocardiography (lead II) is also recorded.This provides continuous feedback of the car-diovascular effects of respiratory pauses.Continuous monitoring of the EEG, with anexpanded electrode array, chin muscle and ante-rior tibialis electromyography, electro-oculo-gram, and objective recording of patientbehavior during the sleep study is essential.When indicated, esophageal pH may also becontinuously monitored. Recording multiplephysiologic variables provides an accurate iden-tification of sleep state–related respiratorychanges as well as identification of apneas asso-ciated with other significant sleep-related disorders(e.g., sleep-related seizure activity, gastroesoph-ageal reflux).

Pneumography (respiratory effort monitoringby thoracic impedance, and heart rate monitor-ing) should be avoided as a diagnostic method.23

Pneumograms have been widely used for screen-ing in asymptomatic premature and term infants.Unfortunately, integrity of respiration duringsleep can be significantly underestimated by thistype of recording. Sensitivity and specificity ofpneumography is poor. False-positive and false-negative results are common. Partial airwayobstruction may cause false breath detection,and breaths immediately after a sigh, or biphasicaugmented breaths are often missed.41 In addi-tion, unless nasal and oral airflows are recorded,only central apneas can be detected. Monitoringonly thoracic impedance may result in failure todetect obstructive apnea, can confuse cardiacpulse artifact with respiratory effort, and cannotdetect abdominal respiratory efforts.

Impedance pneumography done in the homeis almost as costly as comprehensive technician-attended polysomnography conducted in thelaboratory. Pneumograms should be reserved forevaluation of infants who trigger frequent homemonitor alarms, in which circumstances it mayhelp by differentiating true from false activation


Ch24.qxd 01/28/05 10:32 AM Page 280

of the monitor. Technician-attended polysomnog-raphy is still considered the most reliable and effec-tive method available for describing respirationduring sleep and for identification of normal ver-sus abnormal upper airway function duringsleep.

Paroxysmal DisordersSleep-related paroxysmal disorders may be dif-ferentiated into epileptiform and nonepilepti-form abnormalities. Interictal EEG evaluationsmay not be helpful in diagnosis. Often, spells donot occur in the laboratory and comprehensiveassessments and management must be based onclinical grounds. Continuous monitoring of EEGand other physiologic functions during poly-somnography in the sleep laboratory may bequite helpful in differentiating seizure disordersfrom nonepileptic paroxysmal disturbances.

Seizures

An abundant variety of paroxysmal motor disor-ders may occur during sleep. These recurringspells must be differentiated from sleep-relatednonepileptic motor activity. The sleep of patientswith true seizures is typically fragmented.42

Abnormal sleep patterns may, however, indicatea toxic effect of medication or CNS injury.

Interictal epileptiform activity tends toincrease during light stages of NREM sleep andis inclined to be suppressed during REM sleep.This is particularly true in patients sufferingfrom partial complex seizures.43 Sleep depriva-tion increases the rate of focal interictal epilepti-form discharges most markedly in stage 2 NREMsleep. Some epileptic seizures appear almostexclusively during sleep. For example, a syn-drome of continuous spike and wave activityduring sleep occurs in young children and isassociated with hyperkinesias, neuropsychologi-cal disturbances, and progressive aphasia—theLandau-Kleffner syndrome.44

Much literature confirms the observation thatepileptic seizures occur in relation to specificsleep stages and the sleep–wake cycle.45 Sleepdeprivation has been commonly used to pro-mote seizures in the laboratory, especially inpatients with temporal lobe seizure disorders.46

Although seizures are recognized clinically onlyduring the day, fluttering of the eyelids can beobserved during sleep in conjunction with

paroxysmal bursts of 3 cycles per second spike-and-wave activity.47 Partial complex seizuresoriginating in the frontal lobe occur most char-acteristically during NREM sleep. Amongpatients with sleep-related complex seizuresstudied by Cadilhac,48 almost two thirdsoccurred during NREM sleep. Approximately16% of seizures studied were isolated to REMsleep, and 20% occurred in both NREM andREM sleep states. There is also a strong correla-tion between seizures and sleep in patients withbenign partial epilepsy with centrotemporalspikes (Rolandic epilepsy).45

In addition to the facilitatory effect of sleepon seizure activity, seizure frequency may beaffected by the presence of other sleep-relateddisorders. For example, in a group of patientswith obstructive sleep apnea syndrome and par-tial epilepsy, six of seven patients studied byDevinsky and colleagues49 revealed a signifi-cant reduction in the frequency of seizureactivity and seizure severity after successfultreatment of the sleep-related breathing abnor-mality.

Clinical differentiation between epileptic andnonepileptic spells that occur during sleep canoften be difficult. Stores50 reviewed this issueand was able to divide patients fitting this diag-nostic dilemma into three categories. A firstgroup consisted of patients with nonepilepticprimary sleep disorders that are often associatedwith motor phenomena and similar presenta-tions, including some nightmares and sleep ter-rors, NREM sleep partial arousal disorders, andREM-sleep motor disorders. The second groupconsisted of patients with primary sleep disor-ders with motor components that can be incor-rectly diagnosed as epilepsy, including thepartial arousal disorders, REM-sleep motor dis-order, sleep-related rhythmic movement disor-ders (such as jactatio capitis nocturnes), somesymptoms associated with obstructive sleepapnea, automatic behaviors, idiopathic CNShypersomnia, and sleep-related enuresis. Finally,the third group of patients have some epilepticdisorders that occur during sleep and may bemistaken for sleep disorders. These consist ofnocturnal complex partial seizures of the tem-poral lobe and, particularly of frontal lobe ori-gin, nocturnal hypnogenic dystonia, episodicnocturnal wanderings, and nonconvulsive sta-tus epilepticus.


Ch24.qxd 01/28/05 10:32 AM Page 281

Paroxysmal Hypnogenic Dystonia

Paroxysmal hypnogenic dystonia was firstdescribed by Lugaresi and Cirignotta in 1981.51 Itis a rare disorder characterized by stereotypic,choreoathetotic movements and dystonic postur-ing during NREM sleep. Symptoms may begin inchildhood and be mistaken for normal (or abnor-mal) behavior patterns or other stereotypicmovement disorders. Episodes may be brief, last-ing less than a minute, or may be prolonged, per-sisting for hours. Eyes are often open andvocalizations may occur. If episodes occur fre-quently or are recurrent during a single sleepperiod, significant sleep disruption may occur.There are rhythmic, sometimes violent, stereo-typic movements (e.g., kicking, thrashing) of thelimbs and/or trunk, associated with dystonic pos-turing of the hands, feet, arms, legs, and/or face.At the termination of an episode, patients may becoherent, but they rapidly return to sleep.

Polysomnography usually reveals episodesarising out of stage 2 NREM sleep (although ithas been reported to occur in slow-wave sleep aswell). An EEG pattern of arousal may occur afew seconds before an episode. Significant move-ment artifact is seen on the EEG. Whether clearepileptiform activity occurs during a spell issomewhat controversial. Radiographic studiesand magnetic resonance imaging are notablynormal. It is unknown whether hypnogenicparoxysmal dystonia is associated with CNS (orother) pathology.

Symptoms generally run a chronic course andmay persist for many years. Carbamazepine, insmall doses, has ameliorated symptoms in somepatients.

ParasomniasParasomnias are classified as dysfunctions asso-ciated with sleep, sleep stages, or partial arousalsfrom sleep. They are a group of disorders withstrikingly dissimilar presentations, but theyshare many clinical and physiologic characteris-tics. Often, parasomnias present clear sympto-matology (e.g., sleepwalking, head-banging,bruxism). Manifestations appear early in child-hood and might be considered by parents andhealth care practitioners as normal, benign, orbehavioral in origin. As the child ages, benigncharacteristics can become exaggerated and dra-matic. However, few pathophysiologic abnor-

malities can be identified, despite occasionallysevere paroxysmal features.52

As with all other disorders of sleep and wake-fulness, evaluation begins with a comprehensivehistory and physical examination. Special atten-tion must be placed on a detailed description ofthe events. Neurodevelopmental landmarksmust be carefully assessed. Sleep–wake sched-ules, habits, and patterns require delineation.Morning wake time, evening bedtime, bedtimerituals, and nap time rituals should be described.The presence of excessive daytime sleepiness,snoring, or restlessness during sleep should beascertained. Whether there are or are not con-current medical illnesses and whether thepatient is taking any medications or drugsshould be obtained in the clinical interview.

In the complete physical examination,emphasis should be placed on a comprehensiveneurologic and developmental assessment. Theexistence of developmental delays or symptomssuggestive of neurologic disorders might indi-cate an organic basis for the patient’s presentingsymptoms. Evidence of other medical disordersshould be assessed as possible contributing orcoexistent factors.

Laboratory evaluations should be guided bythe presenting signs and symptoms. A urine drugscreen may be helpful if the symptoms might bea side effect of medication. Polysomnography isoften indicated. An expanded EEG electrodearray is recommended. A more extensive EEGmontage than typically recorded during poly-somnography is often helpful in differentiatinga nonepileptic parasomnia from sleep-relatedseizures. Concomitant video recording of thepatient while sleeping is indispensable and canclearly demonstrate motor manifestations andchronicle stereotypic movements. Attemptsshould be made to obtain at least 400 minutes ofnatural nocturnal sleep. It is often helpful to havethe patient drink fluids and avoid urination priorto settling, as bladder distention may precipitatesome parasomnias. The need for all-night EEGrecordings, routine EEG, and radiographic stud-ies depends on the presenting situation, night-time manifestations, and clinical symptoms.

Nonepileptic Stereotypic Parasomnias

Although the phenomenon of stereotypic move-ments during sleep has been recognized for


Ch24.qxd 01/28/05 10:32 AM Page 282

many years, little is known of the etiology.Stereotypic nonepileptic parasomnias are char-acterized by repetitive, meaningless movementsor behaviors. Large muscle groups are involvedand manifestations include rhythmic, repetitivemovements such as body rocking, head-banging,head-rolling, and body-shuttling. They are typi-cally associated with transition from wakeful-ness to sleep, may be sustained into light sleep,and/or occur after arousal from sleep. Childrenare usually developmentally, behaviorally, andmedically normal. Movements may be alarmingin appearance, and parents often become con-cerned for the child’s physical and mental well-being. Injury sometimes occurs.

Stereotypic movements occur in normalinfants and children. Lack of rhythmic activityduring infancy has occasionally been associatedwith developmental delays. When stereotypicmovements during wakefulness persist into olderchildhood and adolescence, a coexisting psy-chogenic component may be present. Stereotypicmovements may be a form of attention getting, ora mechanism of self-stimulation or self-soothing,in developmentally disabled children.

Rhythmic movements can be observed in twothirds of normal children by 9 months of age.The incidence of head-banging occurs in 3% to6.5% of the normal population; body rocking, in19.1% to 21%; and head-rolling, in 6.3%. By 18months of age, the prevalence of rhythmic move-ments decreases to less than 50%, and by 4 yearsof age to approximately 8%. This consistentdecrease and spontaneous resolution of symp-toms as the child grows and develops is consis-tent with a maturational origin of the disorder.

At times, motor activity and head-bangingcan be violent, and physical injury can occur,although it is uncommon. Cutaneous ecchymo-sis and callus formation can result. However,more serious injuries, including subduralhematoma and retinal petechiae, have beenreported. Rhythmic movements usually decreasein intensity and often resolve spontaneouslybetween 2 and 4 years of age. Rarely, symptomspersist into adolescence and adulthood.

Diagnosis is based on identification of charac-teristic symptoms in the absence of other med-ical or psychiatric disorders. Polysomnographydemonstrates typical rhythmic movements dur-ing the immediate presleep period and may per-sist into stage 1 NREM sleep. Occasionally,

activity is noted after spontaneous arousal. It canoccur during slow-wave sleep, but it is rare dur-ing REM sleep. Focal, paroxysmal, and epilepti-form EEG activity associated with the stereotypicactivity are absent; however, a full-montage EEGmay be necessary to rule out epilepsy. Sleep archi-tecture, stage progression, and stage volumes aretypically normal.

Partial Arousal Disorders

Arousal disorders are thought to be caused byimpaired or “partial” arousal from slow-wavesleep. A hierarchical model may exist, as itappears that a continuum of manifestations ofeach of these disorders of sleep is present.Symptoms most often begin in childhood andresolve spontaneously, although occasionallythey persist into adolescence and adulthood.Manifestations are quite alarming and injuryoften occurs.

Arousal disorders present with bizarre, dra-matic symptoms and share a number of commonfeatures. All seem to occur during stage 3/4sleep. Confusion, disorientation, and amnesiafor the events are present, and at times theepisodes are precipitated by external stimuli. Incontrast, forced arousal from REM sleep is moreoften followed by rapid awakening, clearthought processes, and vivid dream recall.

Partial arousal disorders occur more fre-quently during periods of stress, in the presenceof fever, after sleep deprivation, and in patientswith hypersomnolence syndromes. Partialarousals normally occur at the end of slow-wavesleep periods during ascent to lighter sleep stages.Because of the depth of slow-wave sleep and thehigh arousal threshold in children, these arousaldisorders may represent conflicting interactionbetween the mechanisms generating slow-wavesleep and arousal. Chronobiologic triggers thatcontrol sleep stage cycling may be more likelyto result in a partial arousal if the sleep scheduleis chaotic. There may be internal desynchroniza-tion, and the internal arousal stimulus may comeat the “wrong” time, resulting in incompletearousal and manifest characteristics of both states.As the child develops, these CNS mechanismsmature, synchronization occurs, and symptomsresolve spontaneously.

Confusional arousals consist of partialarousals from slow-wave sleep during the first


Ch24.qxd 01/28/05 10:32 AM Page 283

half of the sleep period. Episodes are sudden,startling, and may be precipitated by forcedawakenings. Children may appear to be awakeduring the episode, but they do not respondappropriately to commands and resist being con-soled. Confusion and disorientation are promi-nent. Attempts to abort the “attack” may, in fact,make the symptoms more severe and violent.

Factors that result in increased slow-wavesleep or those that impair arousal may precipitateor exacerbate confusional arousals. Hypersomniasecondary to rebound from sleep deprivation,narcolepsy syndrome, idiopathic hypersomnia, orobstructive sleep apnea may exacerbate symp-toms. Confusional arousals/sleep drunkenness isfrequently seen in patients with narcolepsy syn-drome after prolonged daytime naps (those thatare longer than 60 minutes and that contain slow-wave sleep). Stress, anxiety, fever, and excessiveexercise may precipitate attacks. Organic pathol-ogy is rarely noted, although CNS lesions of theperiventricular gray matter, reticular activatingsystem, and posterior hypothalamus have beenreported in some patients. Injuries during confu-sional arousals are common if there is displace-ment of the patient from the bed.

The onset of symptoms is usually prior to5 years of age. Children gradually arouse fromslow-wave sleep, and they may moan or mumbleunintelligibly. Symptoms then crescendo signifi-cantly. Patients may thrash about in bed or fallfrom the bed to the floor. During the episode, thechild appears profoundly confused and disori-ented. Combativeness and aggressiveness mayoccur and consolation or restraint may result inexacerbation of symptoms. Episodes may bebrief, lasting for only a few minutes, or they maybe prolonged and last for several hours. There isusually retrograde and/or anterograde amnesiafor the event. There may be associated night ter-rors or somnambulism. Enuresis may occur dur-ing or following the episode, resulting indifficulties in differentiating these spells frompartial complex seizures.

Diagnosis is based on identification of confu-sion, disorientation, agitation, or combativenesson arousal, most often during the first one thirdto one half of the night. Associated amnesia forthe event is present. There is rarely a clearlyidentifiable medical or psychiatric disorder pres-ent on clinical evaluation. Partial complexseizure disorders with confusional automatisms

need to be ruled out. Polysomnography revealssudden arousal from slow-wave sleep, brief peri-ods of delta activity, stage 1 theta patterns, recur-rent microsleeps, or a poorly reactive alphaactivity, or any combination of these. Focal,paroxysmal, and epileptiform activity are absentfrom the EEG. Symptoms may peak during mid-dle childhood and then undergo spontaneousremission. The clinical course is usually benign(although frightening). Physical injury canoccur and the child must be protected fromtrauma during the episode.

Somnambulism: Sleepwalking

Somnambulism may vary in presentation fromsimple sitting up in bed to agitated running andaggressive, violent behavior during sleep. Thecomplex series of automatic behaviors mani-fested may appear, on the surface, purposeful. Aswith other partial arousal disorders, somnambu-listic episodes occur out of slow-wave sleep, dur-ing the first third of the sleep period. Episodesmay be quite alarming. Patients are uncoordi-nated and clumsy during the walking episode,and injuries are common. Because of the highincidence of trauma during events, agitated som-nambulism should be considered a potentiallyfatal disorder and the major goal of managementis to protect the child from harm.

Somnambulism has been reported to occur in1% to 15% of the population. It occurs withgreatest frequency during childhood, decreasingsignificantly during adolescence, and it isuncommon in adults. Episodes vary in fre-quency, intensity, and length, making parentalreports quite inaccurate; the true incidence istherefore unknown. There appears to be anequal sex distribution. There also appears to be asignificant familial pattern, although cleargenetic transmission has not been identified.

Somnambulism usually begins in middlechildhood, between 4 and 8 years of age,although onset may occur at any time after thechild develops the ability to walk. Symptomsrange from simple sitting up in bed to extremelyagitated, semi-purposeful automatisms and fran-tic running. Most often the child will wanderaround the house and can perform complextasks, such as unlocking doors, taking food fromthe refrigerator, and eating. At times, childrenleave the house. Often, the behaviors are


Ch24.qxd 01/28/05 10:32 AM Page 284

meaningless and unusual. Verbalizations mayoccur but are usually garbled, confused, andmeaningless. Eyes are often open and the childmay appear awake, but behaviors are only semi-purposeful. Choreiform movements of the armsand head may occur. Often, enuretic episodesoccur and the child may urinate (or attempt tourinate) at unusual places around the house.During a somnambulistic spell, the child isextremely difficult to wake, although completearousal is possible. If awakened, confusion anddisorientation are usually present. Motor activitycan cease spontaneously, and the child may liedown and return to sleep at unusual placesaround the home, or the child may return to bedwithout ever becoming alert.

A number of factors may precipitate somnam-bulistic events. Fever and sleep deprivation arenotable. Any disorder that can produce signifi-cant disruption of slow-wave sleep, such asobstructive sleep apnea, may precipitate events.In addition, sleep walking can often be precipi-tated by urinary bladder distention in the suscep-tible patient. External noise may also trigger anevent. A number of medications can exacerbatethe disorder, including thioridazine, Prolixin,perphenazine, desipramine, and chloral hydrate.

Polysomnography typically reveals an arousalfrom stage 3 or stage 4 sleep during the firsthalf of the sleep period. Most of the backgroundEEG activity is obscured by muscle artifact. Seizureactivity is notably absent.

Although it is clinically difficult, somnambu-lism should be differentiated from other disor-ders of arousal, such as confusional arousals andnight terrors. Displacement from the bed andcalm nocturnal wanderings are less commonwith confusional arousals. Night terrors are moretypically associated with the appearance ofintense fear and panic and are less likely to beassociated with displacement from bed (althoughdisplacement from bed is more common withnight terrors than nightmares). Intense auto-nomic discharges and an initial scream herald asleep terror and are not present in somnambu-lism. Nocturnal seizure disorders typically revealepileptiform discharges during the events; how-ever, the interictal EEG may be normal. REMsleep behavior disorder has been described inchildren, characteristically occurs during REMsleep, and is associated with clear verbalizationsand seemingly purposeful movements.

Sleep Terrors

Sleep terrors are third in a continuum of partialarousals from slow-wave sleep. The onset of asleep terror (in contrast to the gradual onset ofconfusional arousals) is sudden, abrupt, striking,and frightening. These arousals are associatedwith profound autonomic discharges and behav-ioral manifestations of intense fear. The exactprevalence of sleep terrors, like that of other par-tial arousals, is unknown.

Onset of symptoms is usually between 2 and4 years of age. Although most frequent duringchildhood, sleep terrors can occur at any age. Aswith other nonepileptic parasomnias, precipitat-ing factors include fever, bladder distention,sleep deprivation, and CNS depressant medica-tion. Symptoms tend to significantly decreaseduring puberty and rarely persist into adoles-cence and adulthood. Psychopathology is rare inchildren.

A sleep terror begins suddenly. The child typ-ically sits upright in bed and emits a piercingscream. Severe autonomic discharge occurs.Eyes are usually widely open and pupils mayappear dilated. Tachycardia, tachypnea, diaphore-sis, and increased muscle tone are present.During the episode, the child is unresponsive toefforts to console, and parental efforts oftenexacerbate autonomic and motor activity. Duringa spell, the youngster may run hystericallyaround the house. The child may run wildlyinto walls, furniture, or windows. Episodes ofextreme agitation are commonly associated withinjury. Unintelligible vocalizations and enuresiscan occur. As with other partial arousal disor-ders, the child awakened from a spell may beconfused and disoriented, and there is amnesiafor the event. In contrast to confusional arousals,episodes of sleep terrors are usually brief, lastingonly a few minutes, and they subside sponta-neously.

Diagnosis is based on identification of thesesymptoms and exclusion of organic pathology.Polysomnography reveals sudden arousal fromslow-wave sleep during the first third of themajor sleep period. Sleep terrors, however, canoccur out of slow-wave sleep at any time duringthe night. Partial arousals without motor mani-festation occur more frequently in children withsleep terrors than in unaffected children.Autonomic discharges during these partial


Ch24.qxd 01/28/05 10:32 AM Page 285

arousals are identified by the presence of tachy-cardia without full-blown symptoms.

Sleep terrors require differentiation fromsleep-related epilepsy with automatisms. Inthese patients, EEG may show abnormal dis-charges from the temporal lobe, althoughnasopharyngeal leads may be required to iden-tify the focus of abnormal activity. Epilepticevents may also be distinguished from disordersof partial arousal by the presence of a combina-tion of clinical features, stereotypic behaviors,and the fact that they may occur during any partof the sleep period as well as during wakeful-ness. Identification of epileptiform activity, how-ever, does not completely rule out the presenceof a partial arousal, as they may occur concomi-tantly in the same patient.

Management of Parasomnias

There is no clear consensus regarding when apartial arousal parasomnia requires treatment.Symptoms are most often mild, occur less thanonce per month, and result in injury to neitherthe child nor the parents. In mild cases, expla-nation of partial arousal disorders and parentalreassurance may be all that is necessary. Sleephygiene also should be discussed. Parents shouldbe encouraged to let the event run its course andto intervene minimally. Interventions shouldfocus on preventing injury and simply guidingthe child back to bed. Too vigorous interventionmay prolong the episode.

Parents can be alerted of a quiet somnambu-listic episode by the use of an alarm system (e.g.,a bell placed on the doorknob of the child’sroom). Appropriate sleep hygiene is essential.Sleep deprivation should be avoided and regularsleep–wake schedules maintained. Brief daytimenaps might be attempted and a period of quietactivity or relaxation techniques instituted priorto bedtime. Fluids after the night-time mealshould be limited and the child encouraged toempty the bladder immediately prior to bedtime.Fevers, if present, should be appropriatelytreated.

Severity of partial arousals is considered mod-erate when symptoms occur less than once perweek and do not result in harm to the patient orto others. In these cases, reassurance and abehavioral approach (including behavior train-ing, sleep hygiene, psychotherapy, and/or hyp-nosis) have been successful.

In severe cases, when episodes occur almostnightly or are associated with injury, nondrugapproaches are considered first. Drug treatment,when used, should be prescribed for a shortperiod of time and should be used in conjunc-tion with sleep hygiene and behavioral manage-ment. The patient can be weaned from medicationwhen symptoms have been under good controlfor approximately 3 to 6 months.

The most commonly prescribed medication isdiazepam. However, lorazepam and clonazepamin small doses are also quite effective. Dosageshould be adjusted to the needs of the child.Prolonged use of medication increases thepotential for side effects and complications. Theyoung child generally responds well to bothbehavioral and medicinal approaches.

Parasomnias Associated with REM Sleep

Parasomnias previously discussed have beenrelated to dysfunctions associated with sleepstate transitions and partial arousal from NREMstage 3 and stage 4 sleep. Parasomnias have alsobeen reported to occur out of stage REM sleep.In many cases, manifestations are dissimilar andcan be differentiated on clinical grounds alone.Certain REM sleep parasomnias, however, sharethe symptoms of partial arousal disorders. Somefrequently occur in children (e.g., nightmares),whereas others are extremely rare and have onlyrecently been described in children (e.g., REMsleep behavior disorder). Disorders rarelyencountered during childhood are included(e.g., REM sleep behavior disorder) becausetheir importance to the practitioner may becomeclear when they are more completely understoodand dysfunction associated with the sleepingstate is further delineated in children.

Nightmares

A nightmare is a frightening dream that mayawaken the youngster from REM sleep. There isusually vivid, clear recall of disturbing dream con-tent. Anxiety and mild autonomic manifestationsoccur. Often, an anxiety dream contains elementsof danger to the individual; a sudden arousal fromREM sleep occurs; and after awakening, theyoungster is oriented to the environment withclear sensorium. Dream content usually involvesan experience of immediate and credible threat tosurvival, security, or self-esteem.


Ch24.qxd 01/28/05 10:32 AM Page 286

Dream anxiety attacks occur in REM sleep andare often associated with the longest, mostintense REM period, during the last third of thenight. Major body movements are rare (becauseof REM hypotonia), but REM sleep fragmenta-tion, increased phasic activity, and frequentmovement arousals and awakening from sleepwith clear mentation are typical. Manifestationsare generally mild and vocalizations are rare.Although autonomic activity increases duringnightmares, it is generally mild, differentiating itfrom a sleep terror. There is good recall for thedisturbing dream, and the child functions well onwaking. In contrast to sleep terrors, nightmaresare brief. A prolonged waking episode with diffi-culty returning to sleep is common after a night-mare. In addition, nightmares are generallyunassociated with violent outbursts, there is nodisplacement from the bed (until the child awak-ens), and injuries are quite rare. Return to sleepis generally delayed, and the child often respondswell to parental intervention.

Diagnosis of anxiety dreams is based on theidentification of the mild manifestations of disturb-ing dreams occurring during the early morninghours, absence of intense autonomic activation,clear recall of the dream, appropriate functioningand alertness on awakening, and a good responseto parental interventions. Polysomnography mayreveal an abrupt arousal from REM sleep. TheREM period from which the child awakens isusually the longest and most intense period ofthe night. It occurs later in the sleep period, dur-ing early morning hours, and is associated withmild tachycardia and tachypnea. Increased rapideye movement density may be noted. Focal,paroxysmal, and epileptiform EEG activitiesare absent.

Nightmares must be differentiated from sleepterrors, REM sleep behavior disorder, andepilepsy. Sleep terrors are usually more vivid,they are frightening to the observer, they occurduring the first third of the sleep period, andthey are associated with severe autonomic dis-charges. There is fragmented recall, the child isconfused on waking, somnambulism and agi-tated sleepwalking are common, and many chil-dren suffer injuries.

REM sleep behavior disorder has beenrecently described in childhood. Symptoms aresimilar to those seen in the adult patient. In theadult, the sudden arousal from REM sleep is

associated with significant purposeful motoractivity. Similar symptoms may be seen inpatients with post-traumatic stress disorder,where there is state dissociation including, butnot limited to, increased chin muscle tone,increased phasic activity, increased major bodymovements during REM sleep, and increasedperiodic limb movements. Partial complexseizure disorders may occur during any stage ofsleep and wake, and automatisms and stereotypyare common. Seizure episodes are associatedwith abnormal EEG activity.

Sleep Paralysis

Sleep paralysis is characterized by absence of vol-untary motor activity occurring at the beginningof a sleep period (hypnagogic) or immediatelyafter awakening from sleep (hypnopompic). Thepatient is conscious and aware of the environ-ment but feels paralyzed. All muscle groups areinvolved, but the diaphragm and extraocularmuscles are spared. Active inhibition of alphaand gamma motor neurons is present and is sim-ilar to that seen during REM sleep and cataplexy.Sleep paralysis typically lasts only several min-utes and subsides spontaneously. Occasionally,attacks can be aborted by rapid movements of theeyes or by being touched. Hypnagogic orhypnopompic hallucinations are unusual but canoccur and add to anxiety.

Isolated episodes of sleep paralysis can occurin unaffected individuals. Frequent spells arereported in patients with narcolepsy and in famil-ial sleep paralysis. Onset is usually during ado-lescence, but symptoms may begin duringchildhood. Children have difficulty describingthe events and may appear asleep during theepisode. Parents are unaware of the sleep paraly-sis spell, because the atonia can be aborted bytouching or shaking.

The clinical course varies significantly. Mostcases are isolated and may be exacerbated bysleep deprivation, excessive sleepiness, stress,irregular sleep–wake schedules, or acute changesin sleep phase. Sleep paralysis runs a morechronic course in patients with narcolepsy andin the familial form of the disorder.

Diagnosis of sleep paralysis is based on identi-fication of presenting symptoms. These may bequite difficult to interpret in children. Com-plaints of an inability to “get up” or inability to


Ch24.qxd 01/28/05 10:32 AM Page 287

“wake up” may be more common in children.The youngster complaining to the parent of aninability to move after sleep offset is rarelyencountered. Sleep paralysis associated with nar-colepsy can be differentiated from the isolatedform by the absence of chronic excessive daytimesleepiness, sleep attacks, hypnagogic hallucina-tions, and cataplexy. Atonic generalized seizuresoccur during wakefulness and may or may not beassociated with changes in level of conscious-ness. Syncope occurs during wakefulness as welland is most commonly associated with alteredlevels of consciousness.

Polysomnography usually reveals significantdecrease in skeletal muscle tone in the presenceof a normal waking EEG pattern and conjugateeye movements. Occasionally, patients entersleep during an episode of sleep paralysis andreveal an EEG pattern consistent with stage 1sleep. True sleep-onset REM periods may occur.

REM Sleep Behavior Disorder

REM sleep behavior disorder (RBD) or REMsleep motor anomaly (RMA) has been describedin adults.53 There is evidence that a similar syndrome also occurs during childhood.54

Nonetheless, RBD is an unusual disorder charac-terized by the appearance of elaborate, some-times purposeful movement during REM sleep.There is a paradoxical increase in muscle tone,and patients seem to be acting out their dreams.Violent behavior such as punching, kicking,leaping out of bed, and running are reported andoften correspond with dream mentation. Injuriesto the patient and to bed partners are common.

Cases of RBD/RMA have been reported inpediatric patients,54 and further understanding ofthis disorder may reveal the incidence and preva-lence to be higher than current descriptions sug-gest. The majority of cases are idiopathic, butneurologic disorders have been identified inapproximately 40% of affected adults.

Polysomnography reveals increased muscletone that persists throughout sleep. There isoften a paradoxical increase in muscle tone dur-ing REM sleep, increased phasic activity, andexcessive limb or body jerking. Complex behav-iors occur out of REM sleep, but no epileptiformactivity is noted on EEG during the complexmovements. Interestingly, REM sleep behaviordisorder in adults responds well to benzodi-azepines, especially clonazepam.

Imperative in management of youngsters withnonepileptic partial arousal disorders is a step-wise approach. Education of parents and reas-surance may be the only requirement. It isessential the child be protected from injury,especially if spells are frequent, there is displace-ment from the bed, or the child is significantlyagitated and violent. Behavioral managementincludes close attention to sleep hygiene, ade-quate total sleep time, and limited nocturnal flu-ids. Fever should be evaluated and treatedappropriately. Sleep deprivation should beavoided. Sources of stress and anxiety should beidentified and appropriately addressed. If motormanifestations are present, an alarm systemshould be established so that the parents/care-takers can be forewarned of episodes. A bell on adoorknob may be all that is needed. If medica-tions are indicated, benzodiazepines are typicallythe drugs of first choice. Clonazepam in smalldoses (e.g., 0.25 mg orally at bedtime) is quiteeffective for both NREM and REM disorders.Unfortunately, because of the long half-life ofclonazepam, a hangover effect can occur and theyoungster may do poorly and exhibit excessivesleepiness the following day. Lorazepam in simi-larly small doses has been quite successful. Smalldoses of diazepam at bedtime may be mostappropriate for partial arousal parasomnia,which occurs only during the first third of thesleep period time. If RBD/RMA is suspected,appropriate psychological, neurologic, and psy-chiatric evaluations should be considered. Aspreviously stated, RBD/RMA has been associatedwith post-traumatic stress disorder in adults,and preliminary data may support a similar phe-nomenon in children.

Cerebral PalsySleep problems commonly occur in patients withcerebral palsy (CP). Alterations of the sleep–wakecycle and specific primary sleep disorders mayincrease morbidity. Often, the quality of lifeof the youngster and the family is profoundlyaffected.

Physiologic changes occurring during sleepaffect specific reflexes in children with CP.Alterations in the H-reflex during sleep werestudied in 13 children with cerebral palsy (eightwith spastic tetraplegia, two with a mixed formof cerebral palsy without spasticity, three withhypotonic diplegia or tetraplegia).55 In unaffected


Ch24.qxd 01/28/05 10:32 AM Page 288

children, the maximum H-reflex progressivelydecreases in amplitude from wakefulness toREM sleep. In hypertonic patients, there is onlya slight decrease in the H-reflex in NREM sleepand no significant change in REM sleep; theamplitude of the H-reflex is always greater thanthat in the control group. In dystonic and hypo-tonic patients, the results obtained are similar tothose of the control group. In spastic patients (asopposed to the control and the dystonic andhypotonic groups), normal balance between thefunction of supraspinal systems regulating theamplitude of the spinal reflexes during wake andsleep is altered, probably through the scarcefunctionality of the supraspinal inhibitory struc-tures.

Activity in youngsters with CP is often lessthan physical activity in healthy peers. Childrenwith spastic diplegia were compared with nor-mal controls in an analysis of their physicalactivity during the day versus in sleep.56 It wasdetermined that children with spastic diplegiaare considerably less active than their healthypeers. This is often significant in stabilization ofthe sleep–wake cycle, as there is a correlation ofregularly scheduled physical activity during theday and improvement of sleep at night. Regularexercise as part of a comprehensive sleephygiene program is important in stabilization ofthe sleep–wake cycle and can improve sleep insome youngsters.

Disturbances in the sleep–wake cycle havealso been evaluated electroencephalographicallyin children with CP. Classification of the back-ground EEG was well correlated with distur-bances in the sleep cycle, which was also relatedto the interval after the hypoxic insult.57 Therelationship between sleep states and EEG pat-terns also had a close correlation with the classi-fication and became progressively disturbedwith increasing severity of the background EEG.Each grade of the background EEG abnormalityhad a different prognostic significance accordingto the time of the recording. The one in the firstweek offered the best prognostic value. A distur-bance of sleep—if it exists—always runs parallelwith the course of the disease.58

Sleep EEG patterns were compared in 23mentally retarded children (from 4 months to 5years old) with CP, and 39 reference mentallyretarded children with no abnormality exceptpsychomotor retardation.59 The children with

CP (unlike those without CP) exhibited absenceof EEG patterns characteristic of wakefulness,NREM sleep states that could not be differenti-ated, absence of NREM characteristics andNREM sleep without spindles, and absence ofREM sleep. In addition, short sleep times andlong waking times during the night were oftennoted.

Melatonin concentration in blood, urine, orsaliva may become a useful marker of the circa-dian rhythm in disorders of biologic rhythms.60

Of particular interest to pediatric clinicians is thepotential application of melatonin treatment inestablishing or reestablishing circadian rhythmsin infants and children maintained for long peri-ods under artificial light conditions, as encoun-tered in intensive care units, and in the treatmentof sleep and other rhythm disorders associatedwith developmental delay or blindness. Caremust be taken when using melatonin in young-sters with neurologic deficits, especially thosewith intractable seizure disorders. There may bea proconvulsant effect of exogenous melatoninin some children with cerebral palsy and seizuredisorder.61 Until there is a clear understandingof the efficacy and side effects of melatonin,it should be reserved for those youngsters withchronic neurologic disabilities not associatedwith epilepsy, and for sightless youngsters whoare unresponsive to light–dark cycling.

Infants and children with myelomeningo-cele, hydrocephalus, and Arnold-Chiari mal-formation often exhibit symptomatic apneaor hypoventilation. In a study of 18 asympto-matic infants, Ward and colleagues11 showedthat asymptomatic infants with myelomeningo-cele had longer total sleep time, longerepisodes of longest apnea, greater duration ofapnea greater than or equal to 6 seconds aspercent of total sleep time, and lower meanheart rate than did control infants. This sug-gested that asymptomatic infants withmyelomeningocele have a high incidence ofventilatory pattern abnormalities during sleep.This is currently being studied in a multicenterretrospective analysis of respiratory patterns inchildren with this CNS abnormality.

To assess hypoxic and hypercapnic arousalresponse in children with myelomeningoceleand apnea, Ward and colleagues62 evaluated 11infants in the presence of controlled hypoxemiaand 6 infants with controlled hypercarbia


Ch24.qxd 01/28/05 10:32 AM Page 289

challenges. During hypoxemia, only two infantswith myelomeningocele aroused in comparisonto eight of nine control infants. Similarly, duringhypercarbia, arousal occurred in only threeinfants with myelomeningocele compared to allseven control infants. Three infants withmyelomeningocele subsequently died. It wasconcluded that infants with myelomeningocele,Arnold-Chiari malformation, and apnea orhypoventilation have arousal deficits to normalrespiratory stimuli.

Circadian Rhythm DisordersChildren suffering from chronic debilitating neu-rologic disabilities frequently experience disor-ders of the sleep–wake cycle related to intrinsicbiologic (circadian) rhythms. In addition to disor-dered sleep architecture and composition, manyyoungsters (and their families) endure significantdisorganization of the timing of sleep. Disordersof timing of sleep may result in a significantlifestyle disarray, affect daytime functioning andperformance, and contribute to resistance to ther-apeutic interventions during daytime hours.

Delayed Sleep Phase SyndromeSleep phase delay is common during childhoodand is seen in many youngsters with and withoutabnormalities of neurologic status. Bedtimestruggles are common, and sleep onset difficul-ties can be profound. Once asleep, continuityand architecture may be normal, unless earlymorning waking is required or sleep mainte-nance difficulties are also present. The child whohas early morning responsibilities may experi-ence extreme difficulty arousing. Excessive day-time sleepiness most likely occurs, especiallyduring early morning hours. This may be mani-fested by actual sleep attacks (unintentionalsleep episodes), attention problems, behavioralabnormalities, or hyperactivity. Youngsters withdelayed sleep phase syndrome function best inthe afternoon or early evening hours. Spontaneoussleep offset (when no responsibilities are pres-ent) tends to be quite late in the morning, oftenextending into the afternoon. On weekends,recovery may occur because of the ability tosleep later in the day, only to have the problemrecur during weekdays. Treatment principallyfocuses on sleep hygiene. Firm, stable time ofmorning sleep offset is essential. Advancing bed-

time will follow. Faded bedtimes with responsecost is an appropriate behavioral intervention.If youngsters are not asleep within a reasonableperiod of time (e.g., 20 to 30 minutes after lightsout), they may be removed from bed to performquiet tasks until drowsiness is identified. Theyshould then be placed back into bed. Slowlyadvancing bedtime after sleep onset occurs rela-tively rapidly will assist in changing and ulti-mately fixing the timing of the major sleepperiod. Although there is typically a rapidresponse to this type of intervention, it may takesomewhat longer in the youngster with develop-mental or neurologic disability. Medication issometimes required in children with significantneurologic abnormalities, and it may delay sleeponset. Chloral hydrate and benzodiazepines aremost commonly prescribed. If chloral hydrate isused, the dosage should be adequate to assist insleep onset.

References

1. Martin JH, Jessell TM: Development as a guide tothe regional anatomy of the brain. In: Kandel ER,Schwartz JH, Jessell TM: Principles of NeuralScience (3rd ed). New York, Elsevier, 1991, pp296-308.

2. Oksenberg A, Marks G, Farber J, et al: Effect ofREM sleep deprivation during the critical periodof neuroanatomical development of the cat visualsystem. Sleep Res 1986;15:53.

3. Armstrong SM, Cassone VN, Chesworth MJ, et al:Synchronization of mammalian circadian rhy-thms by melatonin. J Neural Trans Suppl 1986;21:375-394.

4. Dahlitz M, Alvarez B, Vignau J, et al: Delayedsleep phase syndrome response to melatonin.Lancet 1991;333:1121-1124.

5. Jan JE, Espezel H, Appleton RE: The treatment ofsleep disorders with melatonin. Dev Med ChildNeurol 1994;36:97-107.

6. Culebras A: Neuroanatomic and neurologic corre-lates of sleep disturbances. Neurology 1992;42(Suppl 6):19-27.

7. Stahl SM, Layzer RB, Aminoff MJ, et al:Continuous cataplexy in a patient with a mid-brain tumor: The limp-man syndrome. Neurology1980;30:1115-1118.

8. Hendricks JC, Morrison AR, Mann GL: Differentbehaviors during paradoxical sleep without ato-nia depend on pontine lesion site. Brain Res1982;239:81-105.

9. Autret A, Laffont F, de Toffol B, et al: A syndromeof REM and non-REM sleep reduction and lateral


Ch24.qxd 01/28/05 10:32 AM Page 290

gaze paresis after medial tegmental pontinestroke: CT scans and anatomical correlation infour patients. Arch Neurol 1988;45:1236-1242.

10. Balk RA, Hiller FC, Lucas EA, et al: Sleep apneaand the Arnold-Chiari malformation. Am RevRespir Dis 1985;132:929-930.

11. Ward SL, Jacobs RA, Gates EP, et al: Abnormalventilatory patterns during sleep in infants withmeningomyelocele. J Pediatr 1986;109:631-634.

12. Jurko MF, Andy OJ, Webster CL: Disordered sleeppatterns following thalamotomy. Clin Electroen-cephalogr 1971;2:213-217.

13. Autret A, Carrier H, Thommasi M, et al: Etudephysiopathologique et neuro-pathologique d’unsyndrome decortication cerebrale. Rev Neurol(Paris) 1975;131:491-504.

14. Culebras A: The neurology of sleep. Neurology1992;42(Suppl 6):6-8.

15. Sheldon SH: Evaluating sleep in infants and chil-dren. New York, Lippincott-Raven, 1996, p 276.

16. Tharp BR: Electrophysiological brain maturationin premature infants: An historical perspective.J Clin Neurophysiol 1990;7:302-314.

17. Binnie CD, Prior PF: Electroencephalography.J Neurol Neurosurg Psychiatry 1994;57:1308-1319.

18. Becker PT, Thoman EB: Rapid eye movementstorms in infants: Rate of occurrence at 6 monthspredicts mental development at 1 year. Science1981;212:1415-1416.

19. Hakamada S, Watanabe K, Hara K, et al: Bodymovements during sleep in full-term newborninfants. Brain Dev 1982;4:51-55.

20. Brouillette RT, Fernbach SK, Hunt CE: Obstruc-tive sleep apnea in infants and children. J Pediatr1982;100:31-40.

21. Sanders MH, Rogers RM, Pennock BE: Prolongedexpiratory phase in sleep apnea: A unifyinghypothesis. Am Rev Respir Dis 1985;131:401-408.

22. Sheldon SH, Onal E, Lilie J, Spire JP: Sleep-relatedpost-inspiratory upper airway obstruction in chil-dren. Sleep Res 1993;22:270.

23. National Institutes of Health Consensus Develop-ment Conference: Infantile apnea and home mon-itoring. Bethesda, Md, US Department of Healthand Human Services, October 1, 1987, NIHPublication No 87-2950.

24. Sheldon SH, Spire JP, Levy HB: Pediatric SleepMedicine. Philadelphia, WB Saunders, 1992, p142.

25. Mathew OP, Remmers JE: Respiratory function ofthe upper airway. In Saunders NA, Sullivan CE(eds): Sleep and Breathing. New York, MarcelDekker, 1984, pp 163-200.

26. Bartlett D Jr: Effects of hypercapnia and hypoxiaon laryngeal resistance to airflow. Respir Physiol1979;37:293-302.

27. Remmers JE, Bartlett D Jr: Reflex control of expi-ratory airflow and duration. J Appl Physiol1977;42:80-87.

28. Bouterline-Young HJ, Smith CA: Respiration offull-term and premature infants. Am J Dis Child1953;80:753.

29. Dransfield DA, Spitzer AR, Fox WW: Episodicairway obstruction in premature infants. Am J DisChild 1983;137:441-443.

30. Milner AD, Boon AW, Saunders RA, Hopkin IE:Upper airway obstruction and apnoea in pretermbabies. Arch Dis Child 1980;55:22-25.

31. Kelly DH, Shannon DC: Treatment of apnea andexcessive periodic breathing in the full-terminfant. Pediatrics 1981;68:183-186.

32. Bairam A, Boutroy MJ, Badonnel Y, Vert P:Theophylline versus caffeine: Comparative effectsin treatment of idiopathic apnea in the preterminfant. J Pediatr 1987;110:636-639.

33. Sims ME, Yau G, Rambhatla S, et al: Limitationsof theophylline in the treatment of apnea of pre-maturity. Am J Dis Child 1985;139:567-570.

34. Rosenkrantz TS, Oh W: Aminophylline reducescerebral blood flow velocity in low birth weightinfants. Am J Dis Child 1984;138:489-491.

35. Miller MJ, Carlo WA, Martin RJ: Continuous pos-itive airway pressure selectively reduces obstruc-tive apnea in preterm infants. J Pediatr 1985;106:91-94.

36. Jones RAK: Apnoea of immaturity: I. A controlledtrial of theophylline and face mask continuouspositive airway pressure. Arch Dis Child 1982;57:761-765.

37. Hirsch K, Wang SC: Selective respiratory stimulat-ing action of doxapram compared to phenylene-tetrazol. J Pharmacol Exp Ther 1974;189:1-11.

38. Barrington KJ, Finer NN, Peters KL, Barton J:Physiologic effects of doxapram in idiopathicapnea of prematurity. J Pediatr 1986;108:124-129.


40. Guilleminault C: Sleep apnea in infancy. InGuilleminault C (ed): Sleep and Its Disorders inChildren. New York, Raven Press, 1987, pp 213-224.

41. Brouillette RT, Morrow AS, Weese-Mayer DE,Hunt CF: Comparison of respiratory inductiveplethysmography and thoracic impedance forapnea monitoring. J Pediatr 1987;111:377-383.

42. Baldy-Moulinier M, Touchon J, Besset A, et al:Sleep architecture and epileptic seizures. InDegen R, Neidermeyer E (eds): Epilepsy, Sleepand Sleep Deprivation. Amsterdam, Elsevier,1984, pp 109-118.

43. Rossi GF, Colicchio G, Pola P: Interictal epilepticactivity during sleep: A stereo-EEG study in


Ch24.qxd 01/28/05 10:32 AM Page 291

patients with partial epilepsy. ElectroencephalogrClin Neurophysiol 1984;58:97-106.

44. Hirsch E, Marescaux C, Maquet P, et al: Landau-Kleffner syndrome: A clinical and EEG study offive cases. Epilepsia 1990;31:756-767.

45. Bourgeois B: The relationship between sleep andepilepsy in children. Semin Pediatr Neurol1996;3:29-35.

46. Rajna P, Veres J: Correlation between night sleepduration and seizure frequency in temporal lobeepilepsy. Epilepsia 34;1993;574-579.

47. Niedermeyer E: Sleep electroencephalogram inpetit mal. Arch Neurol 1965;12:625-630.

48. Cadilhac J: Complex partial seizures and REMsleep. In Sterman MB, Shouse MN, Passouant P(eds): Sleep and Epilepsy. New York, AcademicPress, 1982, pp 315-324.

49. Devinsky O, Ehrenberg B, Barthlen GM, et al:Epilepsy and sleep apnea syndrome. Neurology1994;44:2062-2064.

50. Stores G: Confusions concerning sleep disordersand epilepsies in children and adolescents. Br JPsychiatry 1991;158:1-7.

51. Lugaresi E, Cirignotta F: Hypnogenic paroxysmaldystonia: Epileptic seizure or a new syndrome?Sleep 1981;4:129-138.

52. Sheldon SH: Pediatric Sleep Medicine. Philadelphia,WB Saunders, 1992, pp 119-135.

53. Schenck CH, Hurwitz TD, Mahowald MW: REMsleep behavior disorder. Am J Psychiatry 1988;145:652.

54. Sheldon SH, Jacobsen JJ: REM sleep motor disor-der in children. J Child Neurol 1998;13:257-260.

55. Cherubini E, Frascarelli M, Riccardi B, et al:[Changes in the monosynaptic reflex duringwakefulness and sleep of children with cerebralparalysis.] Riv Neurol 1978;48:228-241.

56. van den Berg-Emons HJ, Saris WH, de BarbansonDC, et al: Daily physical activity of school chil-dren with spastic diplegia and of healthy controlsubjects. J Pediatr 1995;127:578-584.

57. Watanabe K, Miyazaki S, Hara K, Hakamada S:Behavioral state cycles, background EEGs andprognosis of newborns with perinatal hypoxia.Electroencephalogr Clin Neurophysiol 1980;49:618-625.

58. Laffont F, Autret A, Minz M, et al: Polygraphicstudy of nocturnal sleep in three degenerative dis-eases: ALS, oligo-ponto-cerebellar atrophy, andprogressive supranuclear palsy. Waking Sleeping1979;3:17-30.

59. Shibagaki M, Kiyono S, Takeuchi T: Nocturnalsleep in mentally retarded infants with cerebralpalsy. Electroencephalogr Clin Neurophysiol1985;61:465-471.

60. Cavallo A: The pineal gland in human beings:Relevance to pediatrics. J Pediatr 1993;123:843-851.

61. Sheldon SH: Proconvulsant effects of melatoninin children with chronic neurological handicap-ping conditions. The Lancet 1998;351:1254.

62. Ward SL, Nickerson BG, van der Hal A, et al:Absent hypoxic and hypercapneic arousalresponses in children with myelomeningoceleand apnea. Pediatrics 1986;78:44-50.


Ch24.qxd 01/28/05 10:32 AM Page 292

PARASOMNIAS

Parasomnias represent a broad group of sleepdisorders that are defined as undesirable phe-nomena occurring predominantly during sleep.These sleep disorders are of great interest tosleep specialists, primary care providers, andpatients (and their parents) because this groupcomprises some of the most common and bizarresleep problems seen in children. The InternationalClassification of Sleep Disorders (ICSD),1 pub-lished in 1990, includes the following four sub-categories of parasomnia: arousal disorders,sleep–wake transition disorders, parasomniasusually associated with rapid eye movement(REM) sleep, and other parasomnias. The ICSDis currently under revision, and in the revisednosology the sleep–wake transition disorders,which include rhythmic movement disorders,will be reclassified in the sleep-related move-ment disorder category. This topic is covered inChapter 26. Parasomnias associated with REMsleep, which include REM sleep behavior disor-der and recurrent nightmares, is covered inChapter 26 and other parasomnias, which includeenuresis, in Chapter 27. This chapter addressesarousal disorders.

DISORDERS OF AROUSAL IN CHILDREN

Disorders of arousal constitute a common groupof sleep disorders seen in children2 that was firstdescribed as a distinct clinical entity by Broughtonin 1968.3 Disorders of arousal constitute a clini-cal spectrum that varies from a child who quietlysits up in bed, mumbles briefly, and then liesback down and returns to sleep to the adolescentwho has a sudden arousal that begins with abloodcurdling scream followed by headlongflight. All of the disorders of arousal share a

common pathophysiology and have many simi-larities in family history, genetic predisposition,timing during the sleep cycle, and clinicalfeatures.

CLINICAL DESCRIPTION

The clinical features common to most childrenexperiencing any of the disorders of arousalinclude the timing during the night-time sleepcycle, misperception of and unresponsivenessto the environment, automatic behavior, a higharousal threshold, varying levels of autonomicarousal, and variable retrograde amnesia. Thedisorders of arousal typically begin abruptly atthe transition from the first period of slow-wavesleep (non-REM [NREM] stage 4) (Figs. 25–1and 25–2) of the night, which accounts for thetypical timing 60 to 90 minutes after sleep onsetat the end of the first ultradian sleep cycle. Theduration of each event can vary from less than1 minute to over 90 minutes. In most cases, thearousal will terminate when the child returnsto sleep without ever fully awakening. Althoughonly a single event usually occurs on a givennight, some children may have multiple ones.When there are multiple events, they will typi-cally recur at 60- to 90-minute intervals duringthe first half of the night that correspond to sub-sequent transitions out of slow-wave sleep at theend of each subsequent ultradian sleep cycle.Successive events on the same night tend to beprogressively milder.

Although the clinical manifestations of thedisorders of arousal occur along a spectrum, forease of description and to establish a commonnomenclature, the ICSD has divided the spec-trum of arousal disorders into three distinct enti-ties: sleepwalking, confusional arousals, andsleep terrors. This nomenclature will be usedthroughout the present chapter. At the mildest

Disorders of Arousal in ChildrenGerald M. Rosen

Mark W. Mahowald

25

293

Ch25.qxd 1/25/05 3:32 AM Page 293

294 Chapter 25 Disorders of Arousal in Children

Figure 25–1. Polysomnogram of a disorder of arousal that occurred precipitously out of slow-wave sleep.

Figure 25–2. Idealized sleep hypnogram showing ultradian rhythm through the night. NREM–REMcycles are approximately 90 minutes; the majority of SWS occurs early in the sleep period, and themajority of REM sleep occurs late in the sleep period. Disorders of arousal generally occur during thetransition out of SWS (asterisks). NREM, non-REM; REM, rapid eye movement; SWS, short-wave sleep.(Rosen GM, Ferber R, Mahowald MW: Evaluation of parasomnias in children. Child Adolesc PsychiatrClin North Am 1996;5:601-616.)

Ch25.qxd 1/25/05 3:32 AM Page 294

end of the spectrum, a child will simply awakenfrom sleep, sit up in bed, look around briefly, lieback down, and return to sleep. These arousalsare rarely noticed unless the child sleeps with aparent. This type of arousal is usually not char-acterized as a problem by parents and is seldombrought to the attention of the child’s physician.These arousals may be noted as an incidentalfinding in children who are studied by overnightpolysomnography for other reasons.

SleepwalkingThe presentation of sleepwalking is similar at allages and includes, at a minimum, a partial arousalfrom sleep with some ambulation. The youngchild may simply awaken and crawl about in thecrib before returning to sleep. These events gen-erally go unnoticed unless the child sleeps withanother family member. An older child may getup and walk to the parents’ room or may simplybe found asleep at a location different fromwhere s/he went to bed, with no recollection ofhaving left his/her bed. Some unusual behavior,such as urinating in an inappropriate place suchas the closet or next to the toilet, is common.Such a child may be easily led back to bed, withlittle evidence of a complete awakening and norecall of the event the next day. Sleepwalking canbe triggered in most children by simply wakingand then standing them up within the first fewhours of sleep onset. Sleepwalking is not dan-gerous in and of itself, but children may putthemselves in danger during a sleepwalkingepisode by climbing out a window or leavingtheir homes. During presumed sleepwalking,children have drowned while camping near

water and have frozen to death after leaving theirhomes in the winter.

Sleepwalking is common in children, as doc-umented in two large-scale, population-basedstudies by Klackenberg4 in Sweden and Labergeand colleagues2 in Quebec. Klackenberg longitu-dinally studied a group of 212 randomly selectedchildren in Stockholm from ages 6 to 16 years.The prevalence of quiet sleepwalking occurringat least once during the 10-year data collectionperiod in this group was 40%. The yearly inci-dence varied from 6% to 17%, although only3% had more than one episode per month. InKlackenberg’s study, the sleepwalking persistedfor 5 years in 33% of children and for over10 years in 12%.

Laberge and associates studied 1353 randomlyselected children in Quebec whose parents com-pleted sleep questionnaires regarding the pres-ence of parasomnias in their children yearly fromthe time they were 10 to 13 years of age and ret-rospectively from when their children were ages 3to 10 years. In this group, occasional or frequentsleepwalking was present in 14% of all children atsome time between the ages of 3 and 13 years.The yearly incidence of sleepwalking fromLaberge’s study, as shown in Table 25–1, variedfrom 3% to 9%. In the majority of these children,the sleepwalking began and ended between theages of 3 and 10 years. At 13 years of age 3% ofthe children were still sleepwalking. The sleep-walking persisted beyond the age of 13 years in24% of these children. There was no gender dif-ference in sleepwalking prevalence. In the studiesof both Laberge and coworkers and Klackenburg,sleepwalking was frequently seen in those childrenwho had confusional arousals at a younger age.

Disorders of Arousal in Children Chapter 25 295

Age Sleepwalking (%) CA&NT (%) Sleeptalking (%)3-10 9.2 14.7 37.111 7 3.8 32.512 6.8 2.3 30.113 3.3 1.2 29.2Overall 13.8 17.3 55.5

CA&NT, confusional arousals and night terrors.Modified with permission from Laberge L, Tremblay RE, Vitaro F, Montplasir J: Development of parasomnias from

childhood to early adolescence. Pediatrics 2000;106:67-74.

Table 25–1. Prevalence of Parasomnias at Various Ages in the Same Children

Ch25.qxd 1/25/05 3:32 AM Page 295

Confusional ArousalsConfusional arousals may seem quite bizarre andfrightening to parents. The arousal usually startswith some movements and moaning, progress-ing to crying and often in association with intensethrashing about in the bed or crib. An infant maybe described simply as crying inconsolably.These arousals are common in infants and tod-dlers. The child is typically described as appear-ing confused, with the eyes open or closed. Theseevents can last anywhere from a few minutes toover 1 hour, with 5 to 15 minutes being typical.Even if the child calls for the parents, s/he oftendoes not recognize them. Even vigorous attemptsto wake the child are often unsuccessful.Holding and cuddling usually do not providereassurance; instead, the child often resists,twists, and pushes away and may become moreagitated. It is the parents’ inability to comforttheir child, who appears to be in great distress,that is often of the greatest concern to them.

In the study of Laberge and colleagues,2 theprevalence of occasional or frequent confusionalarousals between the ages of 3 and 13 yearswas 17%. The yearly incidence varied from 1%to 14%. In 85% of the children in their study,the confusional arousals first appeared betweenthe ages of 3 and 10 years, and in the majorityof these children, the confusional arousals disap-peared before the age of 10 years. The confu-sional arousals persisted beyond 13 years of agein 6.7% of the children who had confusionalarousals at a younger age. Sleepwalking and con-fusional arousals were often seen in the samechild at different ages in the studies of bothLaberge2 and Klackenberg.4 Thirty-six percent ofthe children with sleepwalking in the studyof Laberge and coworkers had confusionalarousals as preschoolers, and all of the childrenwith confusional arousals in Klackenburg’s studyhad at least one episode of sleepwalking. Labergeand colleagues used the term “night terrors”to describe the events that are called “confu-sional arousals” in the present chapter. Con-fusional arousals is used here to conform to thedefinitions described in the ICSD Classificationof Sleep Disorders.1

Sleep TerrorsSleep terrors are the most dramatic and leastcommon of the disorders of arousal. They are

seen more often in older children and youngadults. The events usually begin precipitouslywith the child bolting upright with a scream.There is generally a high level of autonomicarousal. The eyes are usually open, the heart isracing, and often there is diaphoresis and mydri-asis. The facial expression is one of intense fear.A youngster may jump out of bed and run blindlyas if to frantically avoid some unseen threat.These events are usually shorter than confu-sional arousals, generally terminating withina few minutes. The child may awaken at the con-clusion of the sleep terror or may simply returnto sleep without ever having completely awak-ened. The child may report some recollection,but it is most often fragmented and not charac-teristic of the imagery reported from dreams andnightmares. As measured by the Social BehaviorQuestionnaire,5 anxiety was associated withconfusional arousals and sleep terrors in thestudy of Laberge and associates2 and has beennoted by other authors in case reports of chil-dren and adolescents.6,7

PATHOPHYSIOLOGY OFDISORDERS OF AROUSAL

Disorders of arousal can best be understoodwithin the context of the basic neurophysiologyof sleep. This is described in detail in Chapter 4.However, those aspects of sleep–wake regu-lation that are particularly relevant to the un-derstanding of disorders of arousal will bediscussed here. Sleep is an active, complex,highly regulated neurologic process that is gen-erated by and primarily for the benefit of thebrain. Sleep is composed of two fundamentallydifferent sleep states: REM and NREM sleep.REM sleep, NREM sleep, and wakefulness eachhave very different patterns of cerebral bloodflow, glucose utilization, predominant neuro-transmitter systems, neuronal activation, andthalamic functioning. From the perspectiveof brain function, REM and NREM sleep are asdifferent from each other as each is from wake-fulness. From this perspective, there are threedifferent “states of being” in which humans canexist: REM sleep, NREM sleep, and wake-fulness.8-12 The determination of the state ofbeing may be made using a variety of criteria.For clinical evaluations, electrographic criteriaare generally used—the electroencephalogram,electro-oculogram, and chin electromyogram—


Ch25.qxd 1/25/05 3:32 AM Page 296

as described in the sleep stage–scoring manualof Rechtschaffen and Kales.13 However, behav-ioral criteria may also be used to differentiatewakefulness, NREM sleep, and REM sleep14 inhumans, and single-cell neuronal discharge pat-terns can be used to characterize states of beingin experimental animals.8

Each state has its own unique neuroanatomic,neurophysiologic, neurochemical, and neuro-pharmacologic correlates. There are a number ofphysiologic markers for each state that tendto occur in concert and cycle in a predictableand uniform manner, resulting in the behavioralappearance of a single prevailing state. Becausesleep is a fundamental property of numerousneuronal groups rather than a phenomenon thatrequires the whole brain, it is possible for differ-ent parts of the brain to be in the different statesof wakefulness, REM sleep, and NREM sleep atthe same time. This concept is fundamental tothe understanding of a number of sleep disor-ders, including the disorders of arousal. If differ-ent parts of the brain can be in different statessimultaneously, then the physical, clinical, andpolysomnographic manifestations of these statescan also occur simultaneously. When thisoccurs, the event can be described as a mixedor dissociated state.14

Sleep is a dynamic state. There are continuouscurrents driving and defining the propensitytoward NREM sleep, REM sleep, and wakeful-ness. The rhythmic alternation among thesethree states defines two important biologicrhythms, the circadian and ultradian rhythms.The alternation of wakefulness, REM sleep, andNREM sleep over the 24-hour light–dark cycledefines the circadian rhythm, whereas the inter-weaving of these states over the sleep perioddefines the ultradian rhythm (Fig. 25–2).Circadian cycling is controlled by a hypothala-mic pacemaker located in the suprachiasmaticnucleus.15 Sleep–wake cycling is the most obvi-ous manifestation of the circadian rhythm, butequally important are diurnal variations in bodytemperature, hormone secretion, drug metabo-lism, pulmonary function, immune response/reactivity, blood pressure, intestinal motility, gas-tric acid secretion, and the propensity to enterREM sleep. As the brain cycles among NREMsleep, REM sleep, and wakefulness, a dynamicreorganization occurs among multiple neuronalnetworks and neurotransmitters at many levels

of the neuraxis. A complex switch orchestratedin the midbrain and thalamus takes place.9,10

The transition among states usually occurssmoothly and completely and is not behaviorallyapparent, but this is not always the case. The tran-sition may be gradual and incomplete, resultingin the behavioral appearance of a mixed state.This is the neurophysiologic correlate of theclinical situation that is referred to as state disso-ciation. Narcolepsy, REM sleep behavior disorder,14

and disorders of arousal can all be best under-stood as clinical examples of mixed or dissoci-ated states of being.

During the disorders of arousal, some facetsof wakefulness appear during the transition outof slow-wave sleep. This usually occurs at theend of the first ultradian sleep cycle. As a conse-quence, the transition out of slow-wave sleep,which is usually quiet, becomes dramatic andperhaps violent. The child appears caughtbetween deep slow wave sleep and wakefulness.His/her behavior at this time has elements thatwe associate with wakefulness (walking, talking,complex motor behaviors) and sleeping (misper-ception of and unresponsiveness to the environ-ment, high arousal threshold, amnesia, automaticbehavior) occurring simultaneously. The EEGduring a partial arousal from sleep is typicallycharacterized by waking and sleeping rhythmswith the simultaneous occurrence of alpha,theta, and delta frequencies and likely showsthat different areas of the brain are in differentstates simultaneously. This dissociated state isinherently unstable, and eventually one stateis fully declared. In most cases, the child appearsto simply return to quiet sleep. Alternatively, thechild may awaken but will have no recall ofthe arousal and will usually rapidly return to sleep.The causes of the disorders of arousal are multi-factorial. The genetic predisposition, homeosta-tic drive, sleep–wake cycling/synchronization,and behavioral/emotional state all seem to playsome role in the clinical appearance of the disor-ders of arousal. Of these factors, genetic pre-disposition is probably the most important.A positive family history in a first-degree relativeis present in 60% of the children with these dis-orders16 compared with 30% in the general pop-ulation. Sleep–wake cycling and synchronizationare affected by age, homeostatic factors, circa-dian factors, hormones, and drugs. Affective dis-orders, anxiety, and environmental stress have


Ch25.qxd 1/25/05 3:32 AM Page 297

all been identified as important factors in theappearance of the disorders of arousal in clinicalstudies,2-4,6,7 although the mechanisms by whichthese factors lead to the arousal is not known.

GENETICS OF DISORDERS OF AROUSAL

A familial predisposition toward the disorders ofarousal has been recognized since these disor-ders were first described. The genetics of the dis-orders of arousal has been explored by Hublinand coworkers17,18 in population-based twinstudies and by Lecendreux and colleagues19 inthe human leukocyte antigen testing of patientswho sleepwalk and normal control subjects.Hublin and colleagues calculated the phenotypicvariance of sleepwalking attributable to geneticfactors at 65%, which he believed was the resultof many genes, each with minor effects. This isconsistent with the results of human leukocyteantigen testing for sleepwalking by Lecendreuxand associates.

The neurophysiologic mechanisms by whichgenetics influences the appearance of the disor-ders of arousal is not known. One possibleexplanation, for which there is good theoreticalsupport and some experimental evidence, is thatthe familial predisposition toward the disordersof arousal is mediated by the genetic controlof the kinetics of the sleep homeostatic process.This process has been shown to be under stronggenetic control in animal studies.20 Studies inmice have demonstrated that sleep loss leads toan increase in homeostatic drive, with a changein slow-wave sleep activity as measured by deltapower, in a dose–response fashion that varieswith the duration of prior wakefulness and isdifferent in different genotypes. A quantitativetrait-loci analysis revealed that this trait is theproduct of multiple genes. Human EEG studieshave also shown that slow-wave sleep and EEGslow-wave activity are markers for measuringhomeostatic drive.21-23 Increases in slow-wavesleep and slow-wave activity occur after sleepdeprivation and decline after sleep. In laboratoryanimals, the increase in the homeostatic drivecaused by sleep deprivation is not obliteratedafter the suprachiasmatic nucleus has beendestroyed; suggesting that the site of control forthe homeostatic system is different from thatfor the circadian system. Although the suprachi-

asmatic nucleus does not control the homeosta-tic process, the circadian system interacts withthe homeostatic system to regulate the timing ofsleep and wakefulness and hence the durationof each. The synchronization of the homeostaticand circadian systems has been shown to beessential for the attainment of optimum sleepand wakefulness. This interaction is describedin a comprehensive article by Dijk and Lockley.22

Adequate sleep duration occurs only when thecircadian and homeostatic systems are fully syn-chronized. The clinical implication of this obser-vation is that a child with an irregular and/orchaotic sleep–wake schedule will simply notbe able to have optimum synchronization ofthe homeostatic and circadian systems; thisinevitably leads to sleep disruption and sleepdeprivation, which leads to a clinical event inthe child predisposed to the disorders of arousal.

POLYSOMNOGRAPHY INDISORDERS OF AROUSAL

Polysomnographic studies have shown that indi-viduals with the disorders of arousal, when com-pared with normal control subjects, have noconsistent differences in sleep macroarchitec-ture23,24—that is, sleep efficiency, sleep stagepercentages, and sleep latencies, among its othercharacteristics. However, there are differencesin the sleep microarchitecture, including cyclicalternating patterns in arousal rates,25,26 arousalsfrom slow-wave sleep,23,26 and slow-wave sleepactivity delta counts between the sleep of patientswith the disorders of arousal and that of controlsubjects. These differences are the most promi-nent during the first ultradian sleep cycle, whichis when the disorders of arousal usually occur.There were (1) an increased number of briefEEG arousals from slow-wave sleep, (2) a decreasein slow-wave sleep activity delta counts at theend of the first ultradian cycle, and (3) an increasein arousals from the cyclic alternating patternscompared with control subjects. The subjects forthese research studies were carefully screenedfor the presence of other sleep disorders, specif-ically sleep-disordered breathing, sleep depriva-tion, and restless legs syndrome.

In one large case series of children referred toa sleep clinic for an evaluation of the disorders ofarousal, there was an association between sleep-disordered breathing and the disorders of


Ch25.qxd 1/25/05 3:32 AM Page 298

arousal or restless legs syndrome with periodicleg movements and these same disorders.27 Thedisorders of arousal disappeared after adenoton-sillectomy in the children with sleep-disorderedbreathing and after drug treatment in the chil-dren with restless legs syndrome, suggesting thatthese clinical causes of sleep disruption canunmask a disorder of arousal.27

All of these lines of evidence suggest that thefundamental abnormality that underpins the dis-orders of arousal is the instability of slow-wavesleep. This conclusion is supported by the clini-cal observation and experimental evidence thatin individuals predisposed to the disorders ofarousal, an event is much more likely to occurafter a night of sleep deprivation28 that increasesthe homeostatic drive and would be expected tomake an unstable homeostatic system moreunstable or after sleep fragmentation from sleepapnea or the periodic movements of sleep.

CLINICAL EVALUATION OFCHILDREN WITH DISORDERS OF AROUSAL

Children do not present to a clinician witha diagnosis but rather with a clinical problem.It is the responsibility of the clinician to arrive atthe diagnosis through a careful history and theappropriate use of diagnostic studies. The chil-dren described in this chapter will generallypresent with the complaint of unusual nocturnalawakenings. It is important to recognize thatthere are many causes of this disorder in chil-dren. The most important tool for evaluatingchildren with unusual nocturnal awakenings isa complete sleep and medical history. The sleephistory will usually allow the clinician to distin-guish between the different causes of unusualnocturnal arousals and to formulate an appro-priate evaluation and treatment plan. The pedi-atric sleep history is covered in Chapters 2 and5. Those facets of the sleep history that are themost important in the evaluation of the disor-ders of arousal are listed in Table 25–2. Anyproblem that causes sleep disruption, whichresults in awakenings, or affects sleep durationor synchronization can lead to the appearanceof the disorders of arousal in a child who is thuspredisposed.

There are a number of distinguishing clinicalcharacteristics in the histories of children with

the disorders of arousal. The salient features arelisted and discussed below.

Timing: Events generally though not alwaysoccur about 90 minutes after sleep onset. Ifa second event occurs during the night, it typ-ically occurs 90 minutes after the first, at thetime of the transition out of slow-wave sleep.Arousals occurring on or soon after awaken-ing are more likely to represent unusual sleep-related seizures.29

Description of the event: The onset can be grad-ual, with sleepwalking and confusionalarousals, or sudden, with sleep terrors. Thebehavior during an event is often bizarre andmay be complex but is not stereotypical. Thechild is not normally responsive to the envi-ronment, although s/he may be partiallyresponsive. A child will not typically recog-nize parents and often cannot be comfortedby them. The events generally terminate witha return to sleep and without a completeawakening.

Frequency: The frequency of occurrence ishighly variable, from several times a night toonce in a lifetime; multiple episodes in a sin-gle night may occur but are uncommon.

Level of consciousness: This child is generallynot arousable.

Memory of event: Most children will have norecall the day after the event.

Daytime sleepiness: Most children have no evi-dence of daytime sleepiness the next day.

Family history: The family history is often posi-tive in parents and siblings for any of thedisorders of arousal.

Table 25–3 lists those conditions that mimicthe disorders of arousal and those that may trig-ger them. The conditions that may mimic thedisorders of arousal include seizures; clusterheadaches; psychiatric disorders such as noctur-nal panic attacks, post-traumatic stress disorder,and nocturnal dissociative disorder; nightmares;REM sleep behavior disorder; and the rhythmicmovements of sleep. The conditions that maytrigger these disorders in a child who may bepredisposed include obstructive sleep apnea, theperiodic movements of sleep, gastroesophagealreflux, and behavioral/psychiatric disorders.Sleep deprivation and an irregular sleep–wakeschedule are the most common, easily corrected,nonspecific triggers for the disorders of arousal.


Ch25.qxd 1/25/05 3:32 AM Page 299

POLYSOMNOGRAPHY

The role of polysomnography in the evaluationof the disorders of arousal is limited because thesleep macroarchitecture in children with thesedisorders is generally normal. However, somedifferences in polysomnographic recorded sleephave been reported in children with these disor-ders, such as hypersynchronous delta,30 preced-

ing the arousal. This is a nonspecific finding thatis also seen in children undergoing polysomnog-raphy for other reasons. The primary role ofpolysomnography in the evaluation of childrenwith the disorders of arousal is to rule out othersleep disorders, such as obstructive sleep apnea,the periodic movements of sleep, gastroe-sophageal reflux, nocturnal psychogenic disso-ciative states, REM sleep behavior disorder, and


Circadian

Sleep log for 2 weeks (time in bed, sleep onset,awakening time) (weekdays/weekends)

Vacation sleep schedule24-hour daily schedule of activities (school,

work, meals, play)Amount of light in roomSeasonal variationsPreferred sleep time and duration

Sleep Environment

Describe bedroom (What is in it? Who isthere? How much natural light is there?Is there a television or radio?)

Sleep Onset

How does the child fall asleep?Who is present at sleep onset and what do

they do?Are there curtain calls, fears, hypnagogic

hallucinations, sleep-onset paralysis,restless legs, head banging, body rocking?

Arousals

Time of night FrequencyTriggers Association with injuryDescription of Way in which it

arousal terminatesLevel of agitation/ Manner in which the

ambulation child returns to sleep?Association with Recall the next day

eating/drinkingLevel of Age of onset

consciousnessDuration

Other Sleep Behavior

Seizures, enuresis, diaphoresis, restlessness,snoring, cough, choking, apnea, periodicmovements of sleep, vomiting, nightmares,bruxism

Waking Behavior

Hypnopompic hallucinations, paralysis,headaches

Daytime Sleep

Naps, cataplexy, excessive daytime sleepiness,settings where sleep occurs

Medical

Neurologic: Migraine headaches, attention-deficitdisorder, seizures, tics, mental retardation,narcolepsy, neuromuscular disease

Psychiatric: Depression, anxiety, dissociativedisorders, conduct disorder, panic disorder,physical/sexual abuse, post-traumatic stressdisorder

Ear, Nose, Throat: Ear infections, ear effusions,nasal airway obstruction, sinusitis,streptococci infections, swallowing problems

Cardiorespiratory: Asthma, cough, heartdisease, pneumonia

Gastrointestinal: Vomiting, diarrhea,constipation, swallowing problems

Growth: Failure to thriveAllergies: Milk, seasonal, asthma, eczemaDrug: Legal/illegalSchool/behavior: School/developmental

problems, behavioral problemsAcute medical illness

Family History

Sleep apnea/snoringArousals (sleepwalking, confusional arousals,

night terrors, restless legs/periodicmovements)

Psychiatric condition (depression)Social issues (stress at home, divorce, family

violence, drug/ethyl alcohol use)Narcolepsy, hypersomnolenceRestless legs syndromeDelayed/advanced sleep phase

Table 25–2. Sleep History

Ch25.qxd 1/25/05 3:32 AM Page 300

nocturnal seizures, which may either trigger ormimic the disorders of arousal.

If nocturnal seizures are suspected as thecause of the awakenings as a result of the historyof the stereotypical nature or the timing of thearousals throughout the sleep period or if thereis an increased likelihood of seizures because ofa concomitant neurologic problem, further diag-nostic studies should be obtained. In most clini-cal settings, sleep-related seizures can be bestevaluated with sleep-deprived EEG or an overnightstudy in an EEG telemetry unit. These childrenshould be evaluated in a sleep lab only if the labstaff are experienced in the diagnosis and treat-ment of sleep-related seizures. This complex topicof sleep-related epilepsy is discussed in Chapter24 and has recently been reviewed by Mallow.29

TREATMENT OF CHILDREN WITH DISORDERS OF AROUSAL

The most appropriate treatment for the childwith an unusual nocturnal arousal will dependon the diagnosis. It goes without saying thatbefore considering a treatment strategy, onemust have completed a comprehensive medical,neurologic, and sleep evaluation such as the onedescribed in this chapter. If the disorders ofarousal are thought to be the most likely causeof the awakenings, the treatment should beginwith the education of the child and the parents,reassurance, and safety. The following recom-mendations represent essential components inany effective treatment plan for children with thedisorders of arousal.

The education of the parents and their childrenabout the benign, self-limiting nature of thedisorders of arousal is always the startingpoint of treatment. It is often helpful to dis-cuss the pathophysiology of these disordersin a manner that is comprehensible to boththe child and parents.

The demystification of the symptoms is animportant part of the education of the parentsand child about the disorders of arousal. Theparents often misconstrue the problem andfear that the intensity of the arousal reflectssevere, unresolved conflicts and are symp-toms of psychological distress. Although it istrue that psychological stressors contribute tothe appearance of these disorders, they arerarely the only cause.

The child’s safety is of paramount concern inmanaging children with the disorders ofarousal. These disorders are not dangerous inand of themselves, but during such an eventchildren can put themselves in danger bywalking out of the house during winter or byrunning through a plate glass door. In mostcases these concerns can be addressed using asimple, commonsense approach. The childshould not sleep on the upper level of a bunkbed, obstructions should be removed fromthe room, double-cylinder locks may need tobe installed on the doors of the house, ora security system alerting the parents that adoor or window has been opened may need tobe installed.

Sleep extension and regularizing the child’ssleep schedule should always be considered

Neurologic*Seizures*Cluster headaches

Medical†Obstructive sleep apnea†Gastroesophageal reflux

Behavioral/Psychiatric*Conditioned arousals*Post-traumatic stress disorder*Nocturnal dissociative state*Nocturnal panic

Sleep*Nightmares*Rhythmic movements of sleep*Rapid eye movement sleep behavior

disorders†Periodic movements of sleep†Sleep deprivation†Irregular sleep–wake schedule

*Conditions that mimic disorders of arousal.†Conditions that trigger disorders of arousal.

Table 25–3. Conditions That Mimic* or Trigger† Disorders of Arousal


Ch25.qxd 1/25/05 3:32 AM Page 301

as part of the treatment of the disorders ofarousal. Sleep deprivation and an irregularsleep-wake schedule are very common prob-lems for children and are often causallyrelated to the appearance of these disorders.Correcting these problems often leads to theresolution of the arousal.

A history of caffeine use should be determined,and if found, its elimination is recommended.

The child’s bedtime routine should be pleasantfor both the parent and the child and ideallyends with his/her transition to sleep within15 minutes of bedtime. If there is conflictat bedtime, sleep onset latency is prolonged,or if the child is fearful or requires close con-tact with a parent, these issues will need to beaddressed.

Medication has also been described as an effec-tive treatment for the disorders of arousal inseveral series of case reports, although it hasnot been the subject of well controlled stud-ies. Clonazepam31,32 has been the mostwidely used medication for the treatment ofthese disorders. However, there is a generalhesitancy by many clinicians to begin phar-macotherapy for these disorders in childrenbecause in most cases the events are benignand self-limiting and have no direct, adverseimpact on the child. In a child with frequent,disruptive, or potentially dangerous arousals,medication may be a good short-term strategywhile other nonpharmacologic modalities areinitiated. If medication is used, it should begiven 1 to 11⁄2 hours before the anticipatedsleep onset to ensure that there are adequatedrug levels in the brain at the beginning ofthe night, when these disorders are mostlikely to occur. The beginning dose of clon-azepam is generally 0.25 mg. Tricyclic antide-pressants have also been anecdotally reportedto be effective in the case reports of childrenwith the disorders of arousal.33,34

Two behavioral treatments for confusionalarousals have been described anecdotally inthe literature. Scheduled awakening was firstsuggested by Lask35 as a treatment for the dis-orders of arousal; although it is not clear whythis intervention is helpful, several casereports have described its efficacy.36,37 Therecommended treatment is simple enough:The child is awakened 15 to 30 minutesbefore the time of the usual arousal and needs

to open his/her eyes and at least mumble aresponse before being allowed to return tosleep. The parents continue this interventionnightly for 1 month. In some cases this sim-ple intervention has been effective. However,it should be noted that in some cases thesescheduled awakenings actually trigger anarousal. Relaxation and/or mental imageryand biofeedback have both been described inseveral series of case reports as being usefultreatments for the disorders of arousal.38

Psychotherapy and counseling are importantinterventions for the child who has evidenceof significant psychological distress. This istrue for any child in whom these problems arediscovered when these psychological factorsare considered to have an adverse impact onthe child’s or family’s life. However, childrenare rarely able to compartmentalize their emo-tional lives as well as adults, so it is uncom-mon for the only manifestation of significantpsychological difficulties to be the nocturnalawakenings.

SUMMARY

To the casual observer, the disorders of arousalrepresent a paradox during which an individualappears to engage in waking behavior while stillasleep. With the understanding that sleep andwakefulness are not always mutually exclusivestates of being, the paradox disappears. The con-cept of a mixed state or state dissociation pro-vides an explanation for these events that isfounded in the current understanding of theneurophysiology of sleep. The disorders of arousalare common problems, especially in young chil-dren, and can usually be fully evaluated andtreated by a knowledgeable sleep clinician with-out the use of high technology.

References

1. Thorpy MJ, Chairman, Diagnostic ClassificationSteering Committee of the American Sleep DisorderAssociation: The International Classification ofSleep Disorders: Diagnostic and Coding Manual,1990.

2. Laberge L, Tremblay RE, Vitaro F, Montplasir J:Development of parasomnias from childhood toearly adolescence. Pediatrics 2000;106:67-74.

3. Broughton RJ: Sleep disorders: Disorders ofarousal? Science 1968;159:1070-1078.


Ch25.qxd 1/25/05 3:32 AM Page 302


4. Klackenberg G: Somnambulism in childhood—prevalence, course and behavioral correlates:A prospective longitudinal study (6-16 years). ActaPaediatr Scand 1982;71:495-499.

5. Tremblay RE, Loeber R, Gagnon C, et al: Disruptivesleep with stable and unstable high fightingbehavior patterns during junior elementaryschool. J Ab Child Psychol 1991;19:285-300.

6. Simonds JF. Parago H: Sleep behavior and disor-ders in children and adolescents evaluated at psy-chiatry clinics. Dev Behav Pediatr 1984;6:6-10.

7. Dahl RE, Puig-Antich J: Sleep disturbance in chil-dren and adolescent psychiatric disorders.Pediatrician 1990;167:32-37.

8. Jones B: Basic mechanisms of sleep-wake states.In Kryger MH, Roth T, Dement W (eds):Principles and Practice of Sleep Medicine.Philadelphia, WB Saunders, 1994, pp 145-162.

9. Basics of sleep behavior. In Chase M, McCarley R,Rechtschaffen A, Roth T (eds): Los Angeles, UCLASleep Research Society, 1993, pp 17-78.

10. Hobson J, Schiebel A: The brainstem care:Sensorimotor integration and behavioral statecontrol. Neurosci Res Prog Bull 1980;18.

11. Hobson J, Steriade M: Neuronal basis of behav-ioral state control. In Bloom F (ed): Handbookof Physiology, vol 4. Bethesda, Md, AmericanPhysiological Society, 1986, pp 701-823.

12. Hobson J, Lydic R, Baghdoyan H: Evolving con-cepts of sleep cycle generation: From brain cen-ters to neuronal populations. Behav Brain Sci1986;9:371-448.

13. Rechtschaffen A, Kales A: A Manual of Stan-dardized Terminology: Techniques and ScoringSystem for Sleep States of Human Subjects. LosAngeles, UCLA Brain Information Service/BrainResearch Institute, 1968.

14. Mahowald MW, Schenck CH: Dissociated states ofwakefulness and sleep. Neurology 1992;42:44-52.

15. Harrington M, Rusak B, Mistlberger R: Anatomyand physiology of the mammalian circadian sys-tem. In Kryger M, Roth T, Dement W (eds):Principles and Practice of Sleep Medicine, 3rd ed.Philadelphia, WB Saunders, 2000.

16. Richman N: Surveys of sleep disorders in childrenin a general population. In Guilleminault C (ed):Sleep and Its Disorders in Children. New York,Raven Press, 1987, pp 115-127.

17. Hublin C, Kaprio J, Partinen M, et al: Prevalenceand genetics of sleepwalking: A population-basedtwin study. Neurology 1997;48:177-181.

18. Hublin C, Kaprio J, Partinen M, et al: Parasomnias:Co-occurrence and genetics. Psychiatr Gen 2001;11:65-70.

19. Lecendreux M, Bassetti C, Dauvilliers Y, et al:HLA and genetic susceptibility to sleepwalking.Mol Psychiatry 2003;8:114-117.

20. Franken P, Chollet D, Tafti M: The homeostaticregulation of sleep need is under genetic control.J Neurosci 2001;21:2610-2621.

21. Dijk DJ, Czeisler CA: Contribution of the circa-dian pacemaker and the sleep homeostat to sleeppropensity, sleep structure, electroencephalo-graphic slow waves, and sleep spindle activity inhumans. J Neurosci 1995;15:3526-3538.

22. Dijk DJ, Lockley SW: Functional genomics ofsleep and circadian rhythm: Integration of humansleep-wake regulation and circadian rhythmicity(invited lecture). J Appl Physiol 2002;92:852-862.

23. Gaudreau H, Joncas S, Zadra A, Montplaisir J:Dynamics of slow-wave activity during the NREMsleep of sleepwalkers and control subjects. Sleep2000;23:755-760.

24. Guilleminault C, Poyares D, Aftab FA, et al: Sleepand wakefulness in somnambulism: A spectralanalysis study. J Psychosom Res 2001;51:411-416.

25. Zuconi M, Oldani A: Arousal fluctuation in non-rapid eye movement parasomnias: The role ofcyclic alternating pattern as a measure of sleepinstability. J Clin Neurophysiol 1995;12:147-154.

26. Smirne S, Ferini-Strambi L: Clinical applicationsof cyclic alternating pattern. In Comi G, LuckingC, Kimura J, Rossini PM (eds): Clinical Neuro-physiology: From Receptors to Perception.Philadelphia, Elsevier Science, 1999, pp 109-112.

27. Guilleminault C, Palombini L, Pelayo R, ChervinR: Sleepwalking and sleep terrors in prepubertalchildren: What triggers them? Pediatrics 2003;111:e17-e25.

28. Joncas S, Zadra A, Paquet J, Montplaisir J: Thevalue of sleep deprivation as a diagnostic tool inadult sleepwalkers. Neurology 2002;58:936-940.

29. Mallow BA: Paroxysmal events in sleep. J ClinNeurophysiol 2002;19:522-534.

30. Espa F, Ondze B: Sleep architecture, slow waveactivity and sleep spindles in adult patients withsleepwalking and sleep terrors. Clin Neuro-physiol 1999;111:929-939.

31. Mahowald M, Schenck C: NREM parasomnias.Neurol Clin North Am 1996;14:675-696.

32. Schenk C, Boyd J, Mahowald M: A parasomniaoverlap disorder involving sleepwalking, sleepterrors, and REM sleep behavior disorder in 33polysomnographically confirmed cases. Sleep1997;20:972-981.

33. Dahl R: The pharmacologic treatment of sleep dis-orders. Psych Clin North Am 1992;15:161-178.

34. Balon R: Sleep terror disorder and insomniatreated with trazadone: A case report. Ann ClinPsychiatry 1994;6:161-163.

35. Lask B: Novel and non-toxic treatment for nightterrors. Br Med J 1988;297:592.

Ch25.qxd 1/25/05 3:32 AM Page 303


36. Tobin J: Treatment of somnambulism with antici-patory awakening. J Pediatr 1993;122:426-427.

37. Frank C, Spirito A: The use of scheduled awaken-ings to eliminate childhood sleepwalking. J PediatrPsychol 1997;22:345-353.

38. Kohen DP, Mahowald MW, Rosen GM: Sleep-terrordisorder in children: The role of self hypnosis inmanagement. Am J Clin Hypnosis 1991;4:233-244.

Ch25.qxd 1/25/05 3:32 AM Page 304

Parasomnias are classified as dysfunctions asso-ciated with sleep, sleep stages, and partialarousal from sleep. The International Classifi-cation of Sleep Disorders (ICSD),1 published in1990, includes the following four subcategoriesof parasomnias: arousal disorders, sleep–waketransition disorders, parasomnias usually associ-ated with rapid eye movement (REM) sleep, andother parasomnias. The ICSD is currently underrevision, and in the revised nosology thesleep–wake transition disorders, which includerhythmic movement disorders, will be reclassi-fied in the category of sleep-related movementdisorders.

These disorders of sleep are strikingly dissim-ilar in presentation but share many clinical andbiologic characteristics. The symptoms appearearly in childhood. The steady, gradual transfor-mation and resolution of symptoms suggest thatthe etiology of parasomnias may be matura-tional. Few pathophysiologic abnormalities canbe identified despite the severe and often intensesymptoms. Spontaneous remission as the childages is common.

Longitudinal observations have shown thatmany parasomnias may appear gradually or havea sudden onset. The frequency of parasomniaspells can vary and can range from single rareepisodes to nightly events persisting for a pro-tracted period of time. During wakefulness, noobvious clinical abnormalities are present andthe patients appear medically and developmen-tally normal, only to express bizarre and some-times violent behaviors during sleep.

ETIOLOGY

The etiology of parasomnias is unknown. A mat-urational etiology has been hypothesized.However, any theoretical basis for the underly-ing cause of parasomnias must address the com-

mon features. The additional characteristicsaffecting classification in the pediatric patientinclude those parasomnias typically associatedwith non-REM (NREM) slow-wave sleep (SWS),those typically associated with REM sleep, thosetypically associated with the transition fromwakefulness to sleep, and other parasomnias.Classification is typically made with the empha-sis on observable behaviors.2

CLINICAL AND LABORATORYEVALUATION OF A CHILD WITHA PARASOMNIA

Evaluation begins with a comprehensive historyand physical exam. Special emphasis must beplaced on a detailed description of the nocturnalevents, including but not limited to the follow-ing variables.

1. Usual time of occurrence of the spell2. Description of behaviors, movements, or

symptoms manifested3. Whether intervention efforts by the caretaker

improve or exacerbate symptoms4. Whether the child leaves the bed5. Recall or amnesia of the event(s)6. Occurrence of symptoms during daytime

naps7. Presence or absence of symptoms during

wakefulness8. Presence of stereotypical movements or

rhythmic behaviors during the spell

Basic neurodevelopmental landmarks mustbe carefully assessed for the presence of daytimewaking behavioral or developmental abnormali-ties. Typical sleep–wake schedules, habits, andpatterns require delineation. Habitual morningwaking time, evening bedtime, and bedtime andnap time rituals should be described. The pres-ence of excessive daytime sleepiness, snoring, or

The ParasomniasStephen H. Sheldon

26

305

Ch26.qxd 1/25/05 3:33 AM Page 305

other respiratory symptoms3 may assist in deter-mining exacerbating factors. The presence ofconcurrent medical illnesses and medication his-tory should be obtained in the clinical interview.

A complete physical exam must be per-formed. Emphasis should be placed on a com-prehensive neurologic and developmentalevaluation. The existence of developmentaldelays or symptoms suggestive of neurologic dis-orders might indicate an organic basis for thepresenting symptoms. Other sleep disorders arecommon in children with parasomnias, and res-olution of the primary underlying sleep disorderoften results in resolution of the parasomniaspells.3 A urine drug screen might be helpful ifthe symptoms are considered to be due to a sideeffect of medication.

Video recording of the spell by the parentscan provide significant information. Videopolysomnography is often indicated.4,5 Anexpanded EEG electrode array assists in differ-

entiating a parasomnia from sleep-relatedseizures and might provide information localiz-ing focal pathology (Fig. 26–1). Concomitantvideo recording of the patient while s/he isasleep may demonstrate clinical manifestationsand chronicle movements.6 The study shouldbegin no later than 10:00 PM to avoid artificiallyshort sleep onset latency and end no earlier than6:00 AM to avoid missing the last REM episode.The patient should be allowed to awaken spon-taneously so that a realistic natural recordingmay be obtained. It is often helpful to have thepatient drink fluids and avoid urination beforesettling, since bladder distention may precipitatesome parasomnias.7 In analyzing the polysomno-gram, special emphasis is placed on the identifi-cation of other sleep-related pathologies thatmight precipitate the parasomnia. Attentionshould also be devoted to analyzing the ampli-tude of slow waves, synchronization of slow-wave activity,8 arousal rhythms occurring during

306 Chapter 26 The Parasomnias

Figure 26–1. Recording of centrotemporal spikes characteristic of benign epilepsy of childhood(benign rolandic epilepsy). In this 120-second epoch, spike activity can be seen in the left temporalregion; the series of spikes were followed by an arousal (not a seizure).

Ch26.qxd 1/25/05 3:33 AM Page 306

SWS (Fig. 26–2), and intrusion of 4 to 7 Hz EEGactivity into SWS. Partial arousals, frequentarousal rhythms without a state change, andhypersynchronous theta intrusion into SWS(theta–delta pattern) (Figs. 26–3 and 26–4) havebeen associated with but not diagnostic ofarousal disorders.9 The need for all-night EEGrecordings, routine sleep-deprived EEG, andradiographic studies depends on the presentingsituation, night-time manifestations, and clinicalsymptomatology.

SLEEP–WAKE TRANSITIONDISORDERS

Sleep–wake transition disorders (SWTDs) repre-sent a group of sleep disorders that are mani-fested by falling asleep, the transition betweensleep states, or the transition from sleep to wake-fulness. At times they may be thought of as statedissociations or overlapping states. The presen-tation varies considerably from rare and mildmovements during state transitions to frequentand sometimes injurious movements having thepotential to result in discomfort, pain, anxiety,embarrassment, and disturbance of the young-ster’s sleep as well as that of the entire family.Included in this classification of disorders arerhythmic movement disorders (head bangingand body rocking), sleep starts, sleep talking,and isolated sleep paralysis.

Rhythmic Movement DisordersRhythmic movement disorders involve stereo-typical body rocking or head banging that occursduring the transition from wakefulness tosleep.10-17 These movements may also occur dur-ing arousals from sleep and may persist intoNREM sleep. The movements vary in intensityand can sometimes be quite violent. The etiologyis unknown. Rhythmic movements surroundingthe sleep period are common and have beenreported in approximately two thirds of normalchildren. There appears to be a male-to-femaleratio of 4:1. They are typically self-limiting andare resolved spontaneously in the vast majorityof youngsters by 4 years of age.18

The diagnosis of rhythmic movement disor-ders is based on the identification of characteris-tic symptoms in the absence of other medicaland/or psychiatric disorders. Polysomnographyis rarely required for diagnosis, but when used itdemonstrates the typical rhythmic movementsduring the immediate presleep period that mayextend into early stage 1 sleep. They may also beseen during arousals from sleep and sleep cycletransitions. Occasional movements may be seenduring SWS, but it is rare during REM sleep.19

Focal, paroxysmal, and/or other epileptiformactivity associated with the stereotypical activityare absent. Video recordings can be quite helpfulin characterizing these rhythmic movements.

The Parasomnias Chapter 26 307

Figure 26–2. Recording of frequent arousal rhythms seen during short-wave sleep in children withnon–rapid eye movement parasomnia. In this epoch the rhythm is depicted as a significant theta burst.

Ch26.qxd 1/25/05 3:33 AM Page 307

Home video recordings can also be very helpfulin classification. An occasional sleep-deprived orprolonged EEG recording may be required torule out seizure disorders. Once asleep, sleeparchitecture, state progression, and stage vol-umes are typically normal (Fig. 26–5).

Although vigorous head banging may seemquite aggressive, injury is rare. Occasionalbruising and/or abrasions might occur. The spon-taneous degradation of symptoms and sponta-neous resolution occur. A variety of treatmentshave been suggested; however, most reports areanecdotal, and there have been few treatmentregimens that have been replicable. Treatmenthas included such varied modalities as inten-tional rhythmic movements before bedtime,rhythmic sounds in the sleeping environment,antihistamines, benzodiazepines, and carba-mazepine. The degree of success varies with eachtreatment regimen. Children should be protectedfrom injury, and other disorders must be ruledout, such as autism, pervasive developmental dis-order, and hypnogenic dystonia.

Sleep StartsSleep starts (hypnic myoclonia) has also beentermed hypnagogic jerks.1 This transition prob-lem from wakefulness to sleep is characterizedby a sudden, single, brief muscular contractionof the legs and occasionally the arms, head, andpostural muscles.20 Sensory hallucinations (hyp-nagogic hallucinations) often occur before thesleep start, and the subjective perception offalling may occur, ending with the myoclonicjerk. Sleep starts are common, occur in mostindividuals, and are not pathologic unless theyare frequent and result in sleep onset insomnia.21

They also must be differentiated from seizures,especially if they occur in patients with knownepilepsy.22

Hypnic myoclonia may occur at any age andmay be frightening when observed by a parent,especially if it is associated with a vocalization orcry. Injury from the massive movement is rare,but foot injuries secondary to kicking a bedpostor crib rail may occur.


Figure 26–3. Recording of a 30-second epoch demonstrating a theta-delta pattern during slow-wavesleep in a 13-year-old child. Note the intrusion of theta band activity during delta sleep. This theta-delta pattern has been observed in some patients with non–rapid eye movement (NREM) motor para-somnias. It may be consistent with but not diagnostic of NREM motor disorders.

Ch26.qxd 1/25/05 3:33 AM Page 308


Figure 26–4. Recording of arousal rhythms appearing as bursts of hypersynchronous delta activityduring short-wave sleep, which can be seen during the final one third of the 30-second epoch. Thisactivity has also been reported in some children before an episode of enuresis.

Figure 26–5. Rhythmic movement disorder demonstrated by a 60-second epoch on a recording of ayoungster who had been evaluated for sleep-related head-banging. This event typically occurs duringthe transition from wakefulness to sleep but may also be seen during sleep. In this epoch the patientsuddenly became aroused, with a rhythmic movement artifact noted in all channels. Video recording ofthis patient revealed the rhythmic shuttling of his whole body against the headboard of the bed.

Ch26.qxd 1/25/05 3:33 AM Page 309

Sleep TalkingSleep talking (somniloquy) is very common dur-ing childhood. It is typically of little concernto parents and caretakers. Significant outburstsand loud talking or utterances are rare but occa-sionally may be significant enough to disturbthe sleep of parents or other family members.23-25

Somniloquy is not associated with pathologicstates but may be related to other parasomnias,such as sleep terrors, confusional arousals,and sleepwalking.

The diagnosis is based on identification of thetypical manifestations of coherent speech, inco-herent mumbling, or utterances during the sleepperiod. Amnesia of the event is typical. The clin-ical course is usually self-limiting, but symptomsmay persist. Polysomnography reveals that som-niloquy can occur in any stage of sleep.

PARASOMNIAS USUALLYASSOCIATED WITH REM SLEEP

Parasomnias also occur during REM sleep. Inmost cases, the manifestations are markedly dis-similar to those of the NREM sleep parasomniasand sleep–wake transition disorders. Most canbe differentiated on clinical evaluation alone.Certain REM sleep parasomnias may occasion-ally share symptoms similar to partial arousaldisorders. The frequency varies considerablyfrom commonly seen REM sleep parasomnias(e.g., nightmares) to others that are rarelydescribed in childhood (e.g., REM sleep behav-ior disorders). Disorders rarely encountered dur-ing childhood are included because theirimportance to the practitioner might becomeclear when they are more completely understoodand the dysfunctions associated with the sleep-ing state are further delineated in children.

NightmaresA nightmare (anxiety dream) occurs duringREM sleep and is characterized by a frighteningdream that often results in a prolonged period ofwakefulness.26,27 There is typically clear recall ofthe dream, manifestations of anxiety may bepresent, and there may be some mild autonomicnervous system discharge. There is suddenarousal from REM sleep, arousal to a full wakingstate, full orientation to the environment, and aclear sensorium.

Nightmares most commonly occur during thelast one half to one third of the sleep period, dur-ing the longest and most intense REM episode.The “dream story” is often complex and mayinvolve a credible threat to survival, security, orself-esteem. The dreams are usually vivid, with aclear, action-packed story line. The dream recallis appropriate to the child’s developmental andmaturational level. The child is most often fullyawake and alert after the nightmare, and the reac-tion is emotional rather than associated with theintense autonomic nervous system dischargesseen in association with sleep terrors. Childrenare often easily comforted after an arousal from anightmare, and the return to sleep is delayed.

Nearly all children experience nightmares.Prevalence data are not clear. The age of onsetappears to parallel the development of dreams inchildhood. There seems to be equal sex distribu-tion and no clear familial pattern.

Movements are rare during nightmares due tonormal REM sleep hypotonia; however, arousalfrom sleep with clear mentation is typical.Manifestations are generally mild and vocaliza-tions rare.

The diagnosis of anxiety dreams is based onthe identification of the mild and characteristicmanifestations of disturbing dreams occurringduring the early morning hours, absence ofintense autonomic activation, clear recall of thedream, a story-like quality to the dream report,appropriate functioning and alertness on awak-ening, and a good response to parental interven-tions. Other diagnostic techniques are rarelyrequired, and nightmares are typically easily dif-ferentiated from sleep terrors on solely clinicalgrounds (Table 26–1).

If polysomnography is conducted, an abruptawakening from REM sleep is seen, followed bya prolonged period of wakefulness after sleeponset. There may be mild tachycardia. Increasedeye movement density during REM sleep mayaccompany the nightmare. Muscle tone is low,and there may be increased frequency of phasicmuscle twitches. Focal, paroxysmal, and epilep-tiform activities are absent.

Occasional nightmares are common duringchildhood. However, if they become frequent,persist for prolonged periods of time, or are asso-ciated with daytime behavioral or performancedysfunction, underlying medical or psychologi-cal causes should be considered.


Ch26.qxd 1/25/05 3:33 AM Page 310

Treatment is based on reassurance and educa-tion. The maintenance of appropriate sleephygiene is essential. Identifying and minimizingunderlying causes, especially those that are anx-iety or stress related, are essential to appropriatemanagement. Additional medical and/or psycho-logical evaluation and management may berequired.

Isolated Sleep Paralysis

Isolated sleep paralysis is characterized by aperiod of inability to voluntarily move occurringat the beginning of the sleep period (hypna-gogic) or immediately on awakening from sleep(hypnopompic).28,29 The child is conscious andawake and vigilant toward the environment. Allmuscle groups except the diaphragm and extra-ocular muscles are typically involved. Activeinhibition of alpha motor neurons is present,and this inhibition is similar (if not identical) tothat which is associated with normal REM sleep.Patients often have a sensation of difficultybreathing due to inhibition of the accessory

muscles of respiration, and the episodes arecharacteristically frightening.

Sleep paralysis spells are most often brief, lastonly a few minutes, and subside spontaneously.At times spells can be aborted by contact fromanother person or by volitional rapid move-ments of the eyes. Hypnagogic or hypnopompichallucinations are unusual during spells of sleepparalysis but can occur and increase the anxietyrelated to the episode.

Normal individuals experience isolatedoccurrences of sleep paralysis . More frequentevents are seen in patients with narcolepsy syn-drome (as part of the tetrad of excessive daytimesleepiness/sleep attacks, cataplexy, hypnagogichallucinations, and sleep paralysis) and in famil-ial sleep paralysis. There is equal sex distribu-tion in this isolated form and a femalepreponderance in the familial form.

The onset usually occurs during adolescence,but the symptoms may begin during childhood.Children have difficulty describing the event andmay appear asleep throughout its duration. Theparents may be unaware of its occurrence, since


Characteristic Sleep Terror NightmareTime of night First third to first half of the Last half to last third of the

sleep period sleep periodSleep stage Slow-wave sleep REM sleepAssociated activity Movement and significant motor Movements during nightmares

activity is common. are rare.Severity Severe outbursts Mild to moderate cryingVocalizations Common Mild and rareAutonomic Intense autonomic discharges Mild autonomic activity

nervous system including but not limited to tachycardia, tachypnea, diaphoresis,and pupillary dilation

Recall for the event None/amnesia for the event There is good recall for thenightmare. There is a story quality to the dream report (typically a good dream that turned bad).

State on awakening Confused/disoriented Fully awake and functioningInjuries/violence Common RareAssociated About 18% are associated with None

sleepwalking agitated sleepwalking.

REM, rapid eye movement

Table 26–1. Comparison of Sleep Terrors and Nightmares

Ch26.qxd 1/25/05 3:33 AM Page 311

touching or shaking might abort the hypotonia.The symptoms may be mistaken for resistance toawakening. Children who resist awakeningarouse cranky and obstinate. On the other hand,children awakening from an episode of sleepparalysis might be frightened and cry.

The clinical course varies significantly amongindividuals. Most cases are isolated and are pro-voked by sleep deprivation, excessive sleepiness,stress, and an irregular sleep–wake schedule orafter acute changes in the sleep phase. Sleepparalysis runs a more chronic course in patientswith narcolepsy syndrome and in the familialform of the disorder.

The diagnosis is based on the identification ofpresenting symptoms, which may not be obvi-ous. Sleep paralysis associated with narcolepsysyndrome can be differentiated from the isolatedform by the presence of other symptoms relatedto the clinical tetrad. Atonic generalized seizuresoccur during wakefulness and may or may notbe associated with changes in the level of con-sciousness. Syncope occurs during wakefulnessas well and is commonly associated with alteredlevels of consciousness.

Polysomnography might reveal a significantdecrease in chin muscle tone in the presence ofnormal waking EEG rhythm. Conjugate eyemovements may be present as well. Patients mayoccasionally enter sleep during an episode ofsleep paralysis and reveal an EEG pattern con-sistent with stage 1 sleep. True sleep-onset REMperiods may occur at night, and a multiple sleeplatency test may be required to differentiatethese episodes from narcolepsy syndrome.

REM Sleep Motor DisorderREM sleep motor disorder (RMD), originallydescribed in adults as REM sleep behavior disor-der, has also been described in childhood.30 RMDis an unusual abnormality seen in REM sleep thatis characterized by elaborate, sometimes purpose-ful movements accompanied by vocalizations.There is a paradoxical increase in muscle tonethat might be considered the absence of REMsleep atonia, resulting in patients “acting out theirdreams.” Violent behaviors occasionally occur,with patients punching, kicking, and/or leapingout of bed. These behaviors are associated withvivid dream recall. Injuries to the patient or to bedpartners are common. Episodes usually occur

during the first REM period of the night, approx-imately 90 minutes after sleep onset.

RMD usually begins during late adulthoodand progresses over a variable period of time.Children may also be affected, and a greaterunderstanding of this disorder may reveal theincidence and prevalence to be higher than cur-rent descriptions suggest. A majority of cases areidiopathic; however, neurologic disorders havebeen reported in approximately 40% of affectedadults. The signs and symptoms have also beenreported in post-traumatic stress disorder.

Polysomnography reveals increased muscletone that persists throughout sleep, especiallyREM sleep. There are increased phasic muscleactivity, excessive limb movements and bodyjerking, and periodic limb movements. Complexbehaviors occur during REM sleep. No epilepti-form activity is noted on EEG. Interestingly,RMD in both children and adults responds wellto benzodiazepines, especially clonazepam.

Sleep BruxismSleep bruxism is the forceful grinding or rhyth-mic clenching of the teeth or rhythmic move-ments of the mandible during sleep.31

These rhythmic movements are the result ofinvoluntary, repetitive contractions of the mas-seter, temporalis, and pterygoid muscles. Whenteeth grinding occurs, there is loud, unmistak-able noise produced. Predisposing factors for thedevelopment of bruxism have been reported toinclude minor abnormalities of the teeth, maloc-clusion, stress, and anxiety. Some anecdotal datahave shown that rhythmic movement and pro-trusion of the mandible also occur duringarousals associated with occlusive sleep-disor-dered breathing.

The prevalence of bruxism is unclear, but ithas been estimated that 5% to 20% of childrenhave manifested symptoms. Bruxism has beenreported in over 50% of children, with a meanage of onset of 10.5 years. Dental evidence ofbruxism can be identified in 10% to 20% of thegeneral population There appears to be an equalsex distribution, and the condition is most com-monly seen in children and young adults.Similar to many parasomnias, a familial patternwithout clear genetic transmission can beshown. There are no longitudinal studiesdemonstrating the natural course of bruxism.


Ch26.qxd 1/25/05 3:33 AM Page 312

Episodes of rhythmic jaw movements occureither periodically or paroxysmally in bursts of 5to 15 seconds or longer and are often repeatedmany times during the sleep period. The day-time symptoms are common and include jawpain, craniofacial pain, painful teeth, morningheadaches, chronic wear to the crowns of theteeth, periodontal tissue damage, and bleedingfrom the gums. Resorption of alveolar bone,hypertrophy of the masseter and temporalismuscles, and temporomandibular joint dysfunc-tion can occur. Sleep bruxism can also be mis-taken for atypical migraine cephalgia, especiallywhen the response to traditional treatment hasbeen poor.

The diagnosis is made by the identification ofthe loud, unmistakable sound of bruxism in theabsence of other medical or psychiatric disordersthat may produce abnormal movements during

sleep. Obstructive sleep-disordered breathingshould also be assessed, especially in the pres-ence of morning headaches, frequent nocturnalawakenings (with or without headaches), snor-ing, restless sleep, daytime sleepiness, hyperac-tivity, attention span problems, and performancedifficulties.

Polysomnography reveals paroxysmal, rhyth-mic muscle activity manifested by about 1 Hzmuscle artifact over the temporalis muscle(Fig. 26–6). This rhythmic activity may also beseen in the chin muscle electromyogram ormasseter muscle groups. If it is associatedwith occlusive sleep-disordered breathing, themuscle activity occurs during the arousal imme-diately after the obstructive respiratory event(Fig. 26–7).

A number of therapeutic approaches havebeen recommended, yet a most important factor


Figure 26–6. Recording showing a rhythmic temporalis muscle artifact associated with airwayobstruction. In this 60-second epoch the artifact is noted during an arousal after an episode of apnea,and this brief activity can again be seen during an arousal after hypopnea. The exact cause of this activ-ity is not known, but in patients with obstructive sleep-disordered breathing, it may be mandibularmovement in an attempt to assist in opening the airway.

Ch26.qxd 1/25/05 3:33 AM Page 313

is appropriate dental management. A mouthguard may be worn to prevent damage to theteeth. However, the mouth guard does not seemto prevent episodes of bruxism and is used pri-marily as a preventive dental intervention. Ifstress or anxiety is prominent, efforts to mini-mize the precipitating causes may be helpful.Treatment of dental and/or other anatomicabnormalities, if present, may not alter itscourse. If bruxism is associated with occlusivesleep-disordered breathing, this should beappropriately managed.

References

1. American Sleep Disorders Association: Inter-national Classification of Sleep Disorders: Diag-nostic and Coding Manual (rev). Rochester, Minn,American Sleep Disorders Association, 1997.

2. Brooks S, Kushida CA: Behavioral parasomnia.Curr Psychiatr Reports 2002;4:363.

3. Guilleminault C, Palombini L, Pelayo R, ChervinRD: Sleepwalking and sleep terrors in prepubertalchildren: What triggers them? Pediatrics 2003;11:e17.

4. Zucconi M, Oldani A, Ferini-Strambi L, et al:Nocturnal paroxysmal arousals with motorbehaviors during sleep: Frontal lobe epilepsy orparasomnia? J Clin Neurophysiol 1997;14:513.

5. Dyken ME, Yamada T, Lin-Dyken DC: Poly-somnographic assessment of spells in sleep:Nocturnal versus parasomnias. Semin Neurol2001;21:377.

6. Zucconi M, Ferini-Strambi L: NREM parasom-nias: Arousal disorders and differentiation fromnocturnal frontal lobe epilepsy. Clin Neuro-physiol 2000;41:1221.

7. Broughton R: Childhood sleep walking, sleep ter-rors and enuresis nocturna: Their pathophysiology


Figure 26–7. Transcoronal electrode array showing bruxism activity involving the rhythmic move-ment of the temporalis and masseter muscles. Bruxism is often associated with an obnoxious teeth-grinding noise. This disorder may occur during the day and/or during sleep and may be associatedwith significant wearing of the crowns and enamel of the teeth. There may or may not be similar activ-ity seen in an electromyographic recording of the chin muscle. In this epoch the muscle movement isseen over the temporal leads.

Ch26.qxd 1/25/05 3:33 AM Page 314

and differentiation from nocturnal epilepticseizures. Sleep 1978. Basel, S Karger, 1980, p 103.

8. Parrino L, Smerieri A, Terzano MG: Combinedinfluence of cyclic arousability and EEG syn-chrony on generalized interictal discharges withinthe sleep cycle. Epilepsy Res 2001;44:7.

9. Sheldon SH, Riter S, Detrojan M: Atlas of SleepMedicine in Infants and Children. Armonk, NY,Futura, p 208.

10. deLissovoy V: Head banging in early childhood: Astudy of incidence. J Pediatr 1961;58:803.

11. Kravitz H, Rosenthal V, Teplitz Z, et al: A study ofhead-banging in infants and children. Dis NervSys 1960;21:203.

12. Lewis MH, Baumeister AA, Mailman RB: A neuro-biological alternative to the perceptual reinforce-ment hypothesis of stereotyped behavior: Acommentary on “self-stimulatory behavior andperceptual reinforcement.” J Appl Behav Anal1987;20:253.

13. Lourie RS: The role of rhythmic patterns in child-hood. Am J Psychiatry 1949;105:653.

14. Lovaas I, Newsom C, Hickman C: Self-stimula-tory behavior and perceptual reinforcement.J Appl Behav Anal 1987;20:45.

15. Sallustro MA, Atwell CW: Body rocking, headbanging, and head rolling in normal children.J Pediatr 1978;93:704.

16. Schwartz SS, Gallagher RJ, Berkson G: Normalrepetitive and abnormal stereotyped behavior ofnonretarded infants and young mentally retardedchildren. Am J Ment Def 1986;90:625.

17. Klackenburg G: Rhythmic movements in infancyand early childhood. Acta Paediatr Scand1971;224(Suppl):74.

18. Kaneda R, Furuta H, Kazuto K, et al: An unusualcase of rhythmic movement disorder. PsychiatrClin Neurosci 2000;54:348.

19. Kohyama J, Matsukura F, Kimura K, Tachibana N:Rhythmic movement disorder: Polysomnographic

study and summary of reported cases. Brain Dev2002;24:33.

20. Oswald I: Sudden bodily jerks on falling asleep.Brain 1959;82:92.

21. Broughton R: Pathological fragmentary my-oclonus, intensified sleep starts and hypnagogicfoot tremor: Three unusual sleep-related disor-ders. In Koella WP (ed): Sleep 1986. New York,Fischer-Verlag, 1988, p 240.

22. Fusco L, Pachatz C, Cusmai R, Vigevano F:Repetitive sleep starts in neurologically impairedchildren: An unusual non-epileptic manifestationin otherwise epileptic subjects. Epileptic Dis1999;1:63.

23. Rechtschaffen A, Goodenough D, Shapiro A:Patterns of sleep talking. Arch Gen Psychiatry1962;7:418.

24. Saskin P, Whelton C, Moldofsky H, Akin F: Sleepand nocturnal leg cramps. Sleep 1988;11:307.

25. Laberge L, Tremblay RE, Vitaro F, Montplaisir J:Development of parasomnias from childhood toearly adolescence. Pediatrics 2000;106:67.

26. Fisher CJ, Byrne J, Edwards R, Kahn E: A psy-chophysiological study of nightmares. JAMA1970;18:747.

27. Ferber R: Sleeplessness in children. In Ferber R,Kryger M (eds): Principles and Practice of SleepMedicine in the Child. Philadelphia, WBSaunders, 1995, p 79.

28. Guilleminault C: Narcolepsy and its differentialdiagnosis. In Guilleminault C (ed): Sleep and ItsDisorders in Children. New York, Raven Press,1987, p 182.

29. Penn NE, Kripke DF, Scharff J: Sleep paralysisamong medical students. J Psychol 1981;107:247.

30. Sheldon SH, Jacobsen J: REM-sleep motor disor-der in children. J Child Neurol 1998;13:257.

31. Monaco A, Ciammella NM, Marci MC, et al: Theanxiety in bruxer child: A case-control study.Minerva Stomatol 2002;51:247.


Ch26.qxd 1/25/05 3:33 AM Page 315

Sleep-related enuresis (SRE) is characterized bythe involuntary and recurrent voiding of urineduring sleep. Biologic sequelae are uncommon,but psychological and emotional comorbidityare substantial. To minimize the consequences,an understanding of the etiology, diagnostic pro-cedures, and treatment is required.

Between 3 million and 7 million school-agedchildren suffer from SRE.1 Prevalence data aresimilar in all industrialized societies.2 Symptomsof SRE are often secreted, and generalizations tolarge populations should be accepted as approx-imations rather than as reflections of actual inci-dence and prevalence.

SRE may be primary or secondary dependingon the pattern of symptom presentation. PrimarySRE is the involuntary discharge of urine duringsleep, typically at night, that has been presentsince birth and has not been interrupted by sig-nificant asymptomatic periods. Secondaryenuresis refers to symptoms of involuntary void-ing, typically at night, when there has been anintervening period of at least 3 asymptomaticmonths followed by recurrence.3,4 SecondarySRE is more commonly associated with organicor psychological factors than is primary SRE.Approximately 80% of enuretic children wetonly at night, 5% wet only during wake duringthe day, and about 15% are enuretic both duringwake and sleep.5 Again, organic and psychologi-cal causes are more common when there is alsoa pattern of incontinence during wakefulness.

The age at which SRE becomes abnormal iscontroversial. The child’s coping ability andcommitment to solving the problem are impor-tant. The family’s overall cohesion and frustra-tion with the child who wets the bed at night caninfluence parental perceptions about the conse-quences of enuresis on the child and the family.6

Most agree that 5 years of age is the lower limitwhere concern over bedwetting should begin

and intervention should be considered.7 SRE ispresent in approximately 30% of 4 year olds butin only 10% of 6 year olds, 3% of 12 year olds,and 1% of 15-year-old adolescents. Spontaneousremission rates of about 14% to 19% per yearhave been reported.8 Boys tend to be affectedabout twice as frequently as girls when less than11 years of age. After age 11 years, the sex dis-tribution appears to be equal.9 Additionally,there appears to be a familial predisposition. Thehighest frequency of SRE occurs in childrenwhen both mother and father were enuretic aschildren.10

ETIOLOGY

The exact cause of primary SRE is unknown. Itis most likely that a variety of factors exist andcontribute to persistence of SRE after the age atwhich daytime continence occurs. Becausespontaneous resolution of SRE occurs at a rela-tively regular and steady rate thorough middlechildhood, the underlying cause is thought torelate to delayed maturation of bladder mecha-nisms, and perhaps to a delay in the develop-ment of portions of the central nervous systemrequired or maintenance of continence.

Although there is no conscious control overthe smooth muscle of the bladder, voluntarycontrol over voiding involves the maintenance ofcontrol over skeletal musculature that maydirectly influence smooth muscle. During wake-fulness, recognition of bladder distension mostlikely involves autonomic afferent stimuli origi-nating in the bladder wall.

Sleep-related continence appears to be relatedto several factors. First, functional bladdercapacity (i.e., the volume at which perception ofthe need for micturition occurs) must increaseto a level that can adequately maintain volumethroughout the entire sleep period without

Sleep-Related EnuresisStephen H. Sheldon

27

317

Ch27.qxd 1/25/05 3:34 AM Page 317

spontaneous contractions. An adequate func-tional volume is about 300 to 360 cc, and it isgenerally reached by 2 to 7 years of age.11

Children with primary SRE tend to have delayedfunctional bladder capacity when compared tononenuretic children. These children’s smallerfunctional bladder capacity is insufficient tomaintain sleep-related continence even thoughthey are 41⁄2 years of age or older.12

Spontaneous detrusor muscle contractionstend to be more frequent in enuretic children.13

These spontaneous bladder contractions areinhibited through complex interactions of vol-untary muscle groups. When contractions areuninhibited, involuntary detrusor muscle con-traction might occur during sleep.

Antidiuretic hormone secretion peaks duringnocturnal hours. This circadian variation resultsin decreased free-water clearance, a decrease inthe volume of urine production at night, and anincrease in urine osmolality during the sleepperiod. Night-time increase in antidiuretic hor-mone secretion has been shown to be signifi-cantly diminished in some enuretic childrenwith primary enuresis.14-16 Investigationsinvolving children with SRE compared withunaffected controls pointed to a central action ofdesmopressin, a defect at the central argininevasopressin receptor, or abnormality in the path-way of sensory signals.17

Nonenuretic children and adults arouse whenthe bladder reaches its functional capacity.Although the mechanism of this arousal is notclear, it is most likely viscerally mediated viaimpulses originating in the bladder muscle.18

This arousal response results in waking and con-scious perception of the need to void. Thisresults in activation of voluntary muscle groupsthat are involved in maintenance of urinary con-tinence.

None of these postulated mechanisms func-tion alone. The underlying cause of SRE mostlikely involves an interaction of factors related toongoing development of the urinary tract,endocrine system, autonomic nervous system,and areas of the central nervous system respon-sible for arousal and waking.

Underlying psychological factors are uncom-mon causes for primary SRE and seem to occurin less than 1% of prepubertal children.19,20

Symptoms related to emotional problems, how-ever, appear in about 10% to 15% of children and

appear in enuretic children more commonlythan in nonenuretic youngsters. Emotionalsymptoms might include, but are not limited to,thumb sucking, nail biting, tantrums, stuttering,eating disorders, and poor self-esteem.21,22 Withprimary SRE, these symptoms tend to resolvewith resolution of the SRE.23 On the other hand,psychological etiologies are more common inchildren with secondary SRE. Situational stress,separation from parents, abuse or neglect, anxi-ety, and birth of a new sibling have all been asso-ciated with secondary SRE.1,21,24-26

Achieving continence during sleep is part ofthe constitutional development of the child.Despite the lack of clear documentation for theconcept of maturational delay as an underlyingcause, it is based on longitudinal clinical obser-vations. First, in the absence of clinical interven-tion, complete sleep-related continence willoccur at about 15% per year prior to puberty.Functional bladder capacity increases as matura-tion progresses, despite a normal anatomic blad-der capacity.27 Uninhibited bladder contractionsthat occur in enuretic children improve as theyoungster grows and matures.13,28 Coordinatedcontractions of multiple skeletal muscles is nec-essary to initiate the urinary stream, as well as tomaintain continence.29 Control over these mus-cle groups follows a clear developmental patternsimilar to the development of other motor skills.Arousal from sleep is required when bladder dis-tension occurs. This arousal response is quitehigh in smaller children, and the threshold forarousal from internal and external stimulidecreases as maturation occurs.

Although straightforward genetic factors havenot been identified, a familial pattern of primarySRE has been demonstrated.23,30-32 Many parentsof children with primary SRE were also enureticas children. When both parents suffered fromSRE, almost three fourths of their children werealso enuretic. This is dramatically different fromchildren of parents who were not enuretic aschildren. When neither parent was enuretic as achild, only 15% of their offspring suffer fromSRE. Twin studies have also shown that whenone twin sibling suffers from SRE, there is agreater predisposition for younger siblings toalso suffer from similar symptoms.

Genetic factors may be pertinent in somecases. The pattern of involvement seems to beconsistent with an autosomal dominant charac-

318 Chapter 27 Sleep-Related Enuresis

Ch27.qxd 1/25/05 3:34 AM Page 318

teristic with incomplete penetrance. Geneticstudies suggest linkage of primary SRE to specificmarkers on chromosomes 12, 13, and 22.33-36

An association between genotype and pheno-type, and identification of intermediary pheno-types or traits of children with SRE need to bedocumented.37 Nonetheless, the expectations ofenuretic parents for their children may con-tribute to some degree of familial clustering ofsymptoms.

There has been some suggestion that earlyintroduction of toilet training is unsuccessful, asare rigid or punitive training practices.Development of diurnal and sleep-related uri-nary continence cannot be taught. Waking andsleep-related continence is acquired by a com-plex interaction between maturation of physio-logic systems and learning by trial and error (orsuccess). Too-rigid toilet training and introduc-tion of toilet training too early may be associatedwith SRE.38,39 However, other studies haveshown no correlation between the success orfailure of toilet training efforts and the develop-ment of SRE.40-42

Arousal from sleep seems to be difficult inchildren with SRE. Although the arousal thresh-old from slow-wave sleep may seem to underlieSRE in some children, youngsters with SRE tendto sleep no more deeply than nonenuretic chil-dren, and the structure of their sleep is not sub-stantially different.7 Enuresis occurs in all sleepstages, and spells appear to be random or relatedto time of night rather than sleep state.43,44

Although sleep of enuretic children does notseem to differ significantly from that ofnonenuretic youngsters, those who suffer fromprimary SRE spend a slightly longer time in bedand have an increased number of sleep cycles.Children who have enuretic spells during rapideye movement (REM) sleep are found to havemore REM sleep than comparison children.Tachycardia is often seen to precede the enureticspell. Although the sleep of children with SREseems to be polysomnographically normal, thesechildren can exhibit signs of autonomousarousal prior to voiding.45

ORGANIC FACTORS ASSOCIATEDWITH SLEEP-RELATED ENURESIS

Urinary tract infection (UTI) is 10 times morecommon in prepubertal girls suffering from

habitual SRE than in age-matched girls withoutSRE.20 Boys are much less likely to be prone toUTI than girls, even in the presence of uninhib-ited bladder contractions. Although most UTIsare symptomatic, occult infection can occur andtypical symptoms such as fever, dysuria, anddiurnal urinary frequency may not be present. Itis still unclear whether SRE in patients with aUTI is a symptom of a chronic underlying infec-tion or a result of the pathophysiologic mecha-nism responsible for SRE. After adequatetreatment of the UTI, SRE often remains presentand continues long after the infection hasresolved.8

Anatomic abnormalities of the urinary tracthave been associated with SRE. Any distalobstruction may be associated with the develop-ment of bedwetting.1,2,28 The presence of urinarytract abnormalities is very low in primary SRE,and expensive invasive urologic procedures aremost often not needed in the evaluation of chil-dren with SRE. However, consideration must begiven to the possibility of the presence of anorganic abnormality when diurnal symptomssuggest difficulty initiating or stopping the urinestream, dysuria, daytime enuresis, or excessiveurinary frequency.

Extrinsic pressure on the bladder from a vari-ety of causes may result in enuresis. Extrinsicpressure on the bladder decreases its functionaland anatomic volume, and a large fecal mass orextrinsic space-occupying lesion can result in along-term disorder of bladder control.1 In apatient with chronic constipation, the resolutionof the fecal mass pressing on the bladder oftenresults in resolution of the enuresis.

Patients with medical conditions that resultin polyuria often present with SRE as an initialproblem. Indeed, any condition that results inpolyuria can cause SRE. Enuresis as a primarycomplaint occurs commonly in patients withdiabetes mellitis or diabetes insipidus. There is ahigher than normal frequency of SRE in childrenwith sickle cell anemia, sickle cell trait, andother hemoglobinopathies.3,46 Enuresis is com-monly present in children with sleep-disorderedbreathing (see Chapter 20). In addition, enuresiscommonly occurs in children who suffer fromnon-REM sleep parasomnias, such as sleep ter-rors, sleepwalking, and confusional arousals.

Some neurologic disorders may result inenuresis. Enuresis may be associated with spinal

Sleep-Related Enuresis Chapter 27 319

Ch27.qxd 1/25/05 3:34 AM Page 319

cord lesions, including but not limited tomeningomyelocele, tumors, and agenesis.Enuresis may also occur as a manifestation of aseizure disorder.

Medications that result in a higher urine vol-ume, such as diuretics, may also result in SRE.Medications may be prescribed as short-term orlong-term medical regimens for other medicalconditions. Nonetheless, unintentional inges-tion of diuretics should also be considered in theinitial evaluation.

DIAGNOSIS OF SLEEP-RELATED ENURESIS

Successful treatment of SRE is based on an accu-rate diagnosis. The diagnostic workup beginswith a comprehensive history and physicalexamination. Secondary SRE (specifically ifassociated with symptoms suggesting a urinarytract abnormality), UTI, polyuria, polyphagia,polydipsia, seizure disorder, obstructive sleepapnea syndrome, history of prior surgery relatedto the genitourinary tract, allergies, or diurnalenuresis, should raise the suspicion of organicdisease. Initial workup should also include anassessment of possible psychopathology orabnormalities in family dynamics. The familyhistory should be evaluated for evidence ofenuresis in the patient’s mother or father. Sleephabits and patterns also require assessment.Maintenance of sleep logs or a sleep diary forabout 2 to 4 weeks may provide insight into theaccuracy of the sleep–wake history and will pro-vide a graphic longitudinal description of thechild’s sleep and wake behaviors as well as agraphic depiction of the pattern of bedwetting.

Measurement of the bladder’s functional urinevolume during the period of graphic measure-ment of the youngster’s sleep provides informa-tion regarding functional capacity. If thefunctional bladder capacity is normal, primarySRE is unlikely, and a search for other etiologiesmight be considered.

Laboratory evaluations begin with a urinaly-sis and urine culture. These tests may be all thatis required in the initial evaluation. The urinespecific gravity; the presence of glucose in theurine, pyuria, or proteinuria; and the presence ofcasts in the urinary sediment should be assessed.A urine culture as part of the initial workup torule out occult UTI is important. Other expen-

sive or invasive procedures are not typicallyrequired in the initial evaluation. Additionaltesting should be guided by the differential diag-nosis after the initial assessment.

Ultrasonography, vesicle sphincter elec-tromyography, and cystoscopy might be consid-ered in some children who continue to beenuretic after 3 months of treatment. If radi-ographic or cystoscopic examinations are nor-mal, continued enuresis may be associated withdetrusor or sphincter muscle problems.

If loud snoring, restless sleep, and sleep-related diaphoresis are present, comprehensivenocturnal polysomnography might be consid-ered as part of the patient’s comprehensive eval-uation. However, polysomnographic evaluationis rarely required in the diagnosis of primary SREand should be reserved for those situations inwhich an underlying sleep-related disorder isstrongly suspected. Indeed, diagnosis is mostoften based on assessment of the history andphysical examination alone.

TREATMENT

Treatment of SRE is based on a rational approachto diagnosis and management of any underlyingorganic or pathologic condition. Treatment ofunderlying sleep-disordered breathing oftenresults in resolution of the enuresis. Any psychi-atric or psychological cause must be addressedprior to or in conjunction with the institution ofany developmental, maturational, or behavioralmanagement program. Successful enuresis man-agement programs utilize combined methods ofintervention. Regardless of the treatment pro-gram of choice, behavior management and blad-der training are typically included and areimportant aspects of any intervention.

Treatment must begin with an assessment ofwhether the child has primary or secondaryenuresis. The majority of patients presentingwith this complaint suffer from primary SRE andrequire parental support, empathy, and patience.As the spontaneous cure rate for childrenbetween the ages of 5 and 16 years is about 15%per year, it is commonly recommended that notreatment be provided. Nonetheless, when thechild or parent desires a more rapid resolution orwhen secondary, emotional problems occur,treatment may also involve behavior modifica-tion, enuresis alarms, and retention control exer-


Ch27.qxd 1/25/05 3:34 AM Page 320

cises. Fluid restriction, pharmacotherapy, psy-chotherapy, hypnosis, or biofeedback, or a com-bination of these interventions, may beentertained.47 Behavioral intervention is morelikely to be successful in children over the age of6. Between the ages of 3 and 5 years, reassur-ance, motivational counseling, and simple blad-der training exercises may manage the problem.Continued input from the health care profes-sional and support during treatment contributeto compliance.

Persistence and consistency must be encour-aged and supported in any motivational andbehavioral counseling program. In the absenceof support and close followup, poor complianceis likely and persistence of the problem can beexpected to occur.

When a program involving any form ofbehavioral management begins, sufficient timemust be spent to reassure the youngster and theparents that the symptom of primary enuresis isnot their fault. Expectation of results of the man-agement program must be realistic, and theremust be clear communication regarding the timerequired for problem resolution. Support andpositive reinforcement by the health care profes-sional is essential. When expectations are clearand well understood, the likelihood of successincreases. The tone of the counseling shouldremain positive. Resolution may take months.Persistently working toward an agreed-on goalshould be the primary focus of managementrather than an instantaneous cure of the SRE.Parents and caretakers should avoid punishmentor exhibiting disappointment when the occa-sional wet night occurs, as is common duringmanagement. Interventions such as randomlywaking the child from sleep to void should bediscouraged; this generally results in prolonga-tion of the problem. Management must focus ontreatment of the child’s problem and not “treat-ment of the bed.”5

Respondent conditioning is very helpful.Parents and children should agree on a rewardsystem that is readily attainable at the beginningof management. Rewarding the child for a drybed in the morning can be more discouragingthan supportive because this is often beyond theyoungster’s capability at the outset of the pro-gram. The reward need not be large or expensive.Most importantly, the goal must be attainable bythe child, and the reward must be provided con-

sistently and immediately after the desiredbehavior is performed.

Initial management includes bladder-trainingexercises during the day. Exercises consist ofasking the child to terminate the urine streamabout halfway through the void, counting to 5,and then finishing the void. This final comple-tion is important because residual urine in thebladder may result in the development of a UTI.Rewards should be given for accomplishing eachof these behaviors.

Dry-bed exercises using guided imagery tech-niques also have been suggested. At bedtime, thechild is instructed to lie in bed for a fewmoments, visualizing being asleep and feelingbladder fullness. Assistance from the parents isrequired. Visualization of arousal and awakeningis then suggested and the child gets out of bed,goes to the toilet, and voids. This techniqueshould continue for several months, dependingon the child’s response. Counseling the young-ster and family that setbacks should be expected,at the beginning of treatment and throughout itscourse, improves compliance. Resolution ofsymptoms occurs at a rate of approximately 35%per year with the use of this regimen alone.7

Nevertheless, if there is no progress or change insymptoms after about 3 months, other interven-tions might be considered.

Enuresis alarms are also commonly used inbehavioral interventions. Although medicationswere once a common approach, health care pro-fessionals now appear to be more likely to rec-ommend behavioral techniques in managingSRE.7,48

A variety of instruments are available, somemore successful than others. Much of the suc-cess involves motivation of the child and family.Enuresis alarms involve a system that produces aloud sound when the sensor becomes wet. Allalarm systems function using the same principleof operant conditioning, with the physiologicfunction of bladder distension, detrusor musclecontraction, and voiding coupled to an externalstimulus. The sound (i.e., the stimulus) isintended to awaken (i.e., the response) the child,who will then stop voiding and will finish thevoid in the toilet. Eventually, by coupling thestimulus to the response of arousal, childrenawaken at the sensation of bladder distention.5

At first, the child may have difficult waking tothe sound of the buzzer or bell. Indeed,


Ch27.qxd 1/25/05 3:34 AM Page 321

confusional arousals might result. Parentsshould help the child awaken and guide thechild to the bathroom to attempt to void in thetoilet. If parents are consistent and persistent intheir response to the sound of the alarm, soonthe youngster will begin to spontaneously wakewith the alarm.

When treatment with an enuresis alarmbegins, the child typically completely emptiesthe bladder. Gradually, there will be a perceptibledecrease in the volume of urine voided prior toarousal and an increase in the volume voided inthe toilet. This progressive decrease in theamount of urine voided into the bed continuesuntil only a small spot of urine is voided, justenough to trigger the alarm. Then only theunderclothing may be wet but not the bed. Drynights will then begin to occur and increase infrequency until nocturnal continence isachieved. After about 1 month of consecutivedry nights, the alarm may be discontinued.

Behavioral modification, motivational coun-seling, and bladder training exercises may beused along with the enuresis alarm. Initialchange in the pattern of bedwetting typicallybegins after about 3 to 6 weeks of treatment.Youngsters’ responses can vary considerably,depending on the presence or absence of comor-bid conditions, motivation of the youngster, andconsistency of parental assistance. Some young-sters become dry quickly and others requiremuch longer time. In children where longertreatment regimens have been required, it isunclear whether the resolution was a result ofthe treatment or because of timing of maturationof nocturnal bladder continence.

As with all treatment regimens, exacerbationsmay occur. It is important for the child and theparents to understand that these are possible andcan be expected in many children, and theyshould not become discouraged. Reinstitution ofthe treatment protocol typically results in a veryrapid response.

Few contraindications exist. The child mustbe able to hear the alarm, and its volume must begreater than the arousal threshold. Hearing-impaired children or children of hearing-impairedparents may not be responsive to this method ofmanagement. Children and parents must be con-siderably motivated. Children who are fearful orresistant to the alarm are less likely to respondfavorably. Children must be developmentally

capable of responding appropriately to thealarm. The younger the child, the less likely it isthat the results will be positive. Finally, theenuresis alarm should not be used as a means oftoilet training. Diurnal continence results from acomplicated interaction of multiple reflexes,cognitive ability to inhibit bladder contraction,and conditioning. Nocturnal continence requirestime and patience. Parents’ expectations shouldbe appropriate to the intervention.

Resolution of SRE using enuresis alarms hasbeen reported at approximately 65% to 80%,with relapse rates of approximately 10% to15%.49 Comparative trials reveal that the enure-sis alarm may be superior to pharmacotherapyand may be more effective in treatment of pri-mary SRE when utilized with a behavior modifi-cation program. This technique may increasefunctional bladder capacity and result in sus-tained confidence.50,51

Most often, the enuresis alarm is combinedwith a positive reinforcement program, retentioncontrol training, and positive practice usingguided imagery. A combined approach to themanagement of enuresis is superior to the use ofan alarm alone.

Medications used in the treatment of SREhave included antidepressants, antidiuretics,antispasmodics, and prostaglandin synthesisinhibitors. Desmopressin, 10 to 40 microgramsintranasally or 200 to 600 micrograms orally, hasbeen successful in the treatment of primary SRE.Although desmopressin usually reduces the fre-quency of wetting, only a small minority ofpatients obtains complete dryness using thismedication.15,52,53 Treatment effects usually lastonly as long as the drug is taken. Recurrences arecommon. Side effects are uncommon. Significanthyponatremia associated with seizures has beenreported.54,55 There does not appear to be a rela-tionship between a family history of SRE andresponse to desmopressin.32 Utilization ofdesmopressin with an enuresis alarm or otherbehavioral intervention may be the best overalltherapy, and this combined approach appears tobe superior to either given alone.56,57

Oxybutynin chloride has a direct spasmolyticeffect on the bladder and is helpful in controllingsome children with SRE that is caused by unin-hibited bladder contractions. Oxybutynin has adirect effect on bladder smooth muscle, and con-tractions are directly inhibited. Traditionally, it


Ch27.qxd 1/25/05 3:34 AM Page 322

has been utilized in children with incontinencesecondary to neurogenic bladder. It has beenshown to be effective in controlling diurnalenuresis and uninhibited detrusor muscle con-tractions, and it has been used in some centers incombination with desmopressin and behavioralinterventions in comprehensive SRE manage-ment programs.

Tricyclic antidepressants, particularly imipra-mine, have been the most commonly prescribedmedication for the treatment of SRE. Their effec-tiveness appears to be derived from increasedalpha-adrenergic stimulation of proximaltubules of the kidneys.58,59 Unfortunately, theresponse rate is quite variable and the long-termcure rate is only about 25%. Relapses are com-mon. Tricyclic antidepressants significantlyaffect the sleep cycle and result in REM sleepsuppression. There is a narrow therapeuticwindow, and fatal overdose is a significantpossibility.60

In one randomized, double-blind, placebo-controlled trial, carbamazepine was shown to beuseful in the treatment of primary SRE.61

However, comprehensive polysomnography orEEG was not done prior to randomization, mak-ing the results questionable.

Other forms of interventions using alternativetechniques have been reported. Electroac-upuncture has been reported to result in 65%more dry nights than in control children.62

However, only 5 of 23 youngsters were consid-ered responders to treatment (respondersdefined as having a greater than 90% reductionin the number of wet nights at a 6-month evalu-ation). Interestingly, according to the youngsters’parents, the sleep arousal threshold had decreasedin about 50% of these children.

The success rate of any management programmay be attributed in part to the spontaneouscure rate or to the patient’s response to increasedattention to the problem. Emotional support ofthe child is probably one of the most critical ele-ments of any treatment program.

References

1. Schaefer CE: Childhood Encopresis andEnuresis. New York, Van Nostrand Reinhold,1979, p 89.

2. Schmidtt ED: Nocturnal enuresis: An update ontreatment. Pediatr Clin North Am 1982;29:21.

3. Nino-Murcia G, Keenan SA: Enuresis and sleep.In Guilliminault C (ed): Sleep and Its Disordersin Children. New York, Raven Press, p 253.

4. Wyker AW: Standard diagnostic considerations.In Gillenwater JY, Grayhack JT, Howard SS, et al(eds): Adult and Pediatric Urology. Chicago,Year Book Medical, 1987, p 62.

5. Sheldon SH: Sleep-related enuresis. ChildAdolesc Psychiatr Clin North Am 1996;5:661-672.

6. Landgraf JN, Abidari J, Cilento BG, et al:Coping, commitment, and attitude: Quantifyingthe everyday burden of enuresis on children andtheir families. Pediatrics 2004;113:334.

7. Sheldon SH, Spire JP, Levy HB: Pediatric SleepMedicine. Philadelphia, WB Saunders, 1992, pp151-166.

8. Forsythe WI, Redmond A: Enuresis and sponta-neous cure rate: Study of 1129 enuretics. ArchDis Child 1974;49:259.

9. Meadow SR: Enuresis. In Edelman CH (ed):Pediatric Kidney Disease. Boston, Little, Brown,1978, p 1176.

10. Cohen MW: Symposium on behavioral pedi-atrics. Pediatr Clin North Am 1975;22:3.

11. Esperanca N, Gerrard JW: Nocturnal enuresis:Studies in bladder function in normal childrenand enuretics. Can Med Assoc J 1969;101:721.

12. Muellner SR: Development of urinary control inchildren: Some aspects of the cause and treat-ment of primary enuresis. JAMA 1960;172:1256.

13. Koff SA, Murtagh DS. The uninhibited bladderin children: Effect of treatment on recurrence ofurinary tract infection and on vesicoureteralreflux resolution. J Urol 1983;130:1138-1141.

14. Norgaard JP, Rittig S, Djurhuus JC: Nocturnalenuresis: An approach to treatment based onpathogenesis. J Pediatr 1989;114(Pt 2):705.

15. Folwell AJ, Macdiarmid SA, Crowder HJ, et al:Desmopressin for nocturnal enuresis: Urinaryosmolality and response. Br J Urol 1997;80:480.

16. Aikawa T, Kasahara T, Uchiyama M: The argi-nine vasopressin secretion profile of childrenwith primary nocturnal enuresis. Eur Urol1998;33(Suppl 3):41.

17. Eggert P: What’s new in enuresis? Acta PaediatrTaiwan 2002;43:6-9.

18. Sheldon SH, Spire JP, Levy HB: Pediatric SleepMedicine. Philadelphia, WB Saunders, 1992, pp151-166.

19. Rutter M, Yule W, Grahm P: Enuresis and behav-ioral deviance. In Kolvin I, MacKeith RC, MeadowSR (eds): Bladder Control and Enuresis. London,Heinemann Medical Books, 1973, pp 137-147.

20. Taylor PD, Turner RK: A clinical trial of contin-uous, intermittent and overlearning “bell and


Ch27.qxd 1/25/05 3:34 AM Page 323

pad” treatment for nocturnal enuresis. BehavRes Ther 1975;13:281-293.

21. Benjamin LS, Stover DO, Geppert TV, et al: Therelative importance of psychopathology, trainingprocedure and urological pathology in nocturnalenuresis. Child Psychiatry Hum Dev 1971;1:215.

22. Faschingbauer TR: Enuresis: Its nature, etiology,and treatment—A review of the literature. JSASCatalog of Selected Documents for Psychology1975;5:194.

23. Hallgren B: Enuresis: A clinical and geneticstudy. Acta Psychiatr Scand Suppl 1957;114:1.

24. Douglas JWS: Early disturbing events and laterenuresis. In Klovin I, MacKeith RC, Meadows SR(eds): Bladder Control and Enuresis. London,Heinemann Medical Books, 1973.

25. Werry JS: Enuresis: A psychosomatic entity?Can Med Assoc J 1967;97:319.

26. Sheldon SH, Levy HB, Ahart S: Sleep disorders inabused and neglected children (Unpublisheddata.) 1990.

27. Troup CW, Hodgson NB: Nocturnal functionalbladder capacity in enuretic children. J Urol1971;129:132.

28. Lapides J, Diokno AC: Persistence of the infantbladder as a cause of urinary infection in girls.J Urol 1970;103:243.

29. Muellner SR: Physiology of micturition. J Urol1951;65:805.

30. White N: A thousand consecutive cases ofenuresis: Results of treatment. Child Fam 1971;10:198.

31. Young GC: The family history of enuresis. J RInst Public Health 1963;26:197.

32. Schaumberg HL, Rittig S, Djurhuus JC: No rela-tionship between family history of enuresis andresponse to desmopressin. J Urol 2001;166:2435-2437.

33. Eiberg H: Total genome scan analysis in a singleextended family for primary nocturnal enuresis:Evidence for a new locus (ENUR3) for primarynocturnal enuresis on chromosome 22q11. FurUrol 1998;33(Suppl 3):34.

34. Arnell H, Hjalmas K, Jagervall N, et al: The genet-ics of primary nocturnal enuresis: Inheritance andsuggestion of a second major gene on chromo-some 12q. J Med Genet 1997;34:360.

35. von Gontard A, Hollmann F, Eiberg H, et al:Clinical enuresis phenotypes in familial noctur-nal enuresis. Scand J Urol Nephrol 1997;183(Suppl):11.

36. Eiberg H: Total genome scan analysis in a singleextended family for primary nocturnal enuresis:Evidence for a new locus (ENUR3) for primarynocturnal enuresis on chromosome 22q11. EurUrol 1998;33(Suppl 3):34.

37. von Gontard A, Schaumburg H, Hollmann E,et al: The genetics of enuresis: A review.J Urology 2001;166:2438.

38. Benajmin LS, Serdahely W, Geppert TV: Nighttraining through parents’ implicit use of operantconditioning. Child Dev 1971;42:963.

39. Bindelgltas PM, Dee GH, Enos FA: Medical andpsychosocial factors in enuretic children treatedwith imipramine hydrochloride. Am J Psychiatry1968;124:125.

40. Broughton RJ: Sleep disorders: Disorders ofarousal? Science 1968;159:1070.

41. Despert J: Urinary control and enuresis.Psychosom Med 1944;6:294.

42. Klackenberg G: Primary enuresis: When is achild dry at night? Acta Paediatr Scand1955;44:513-518.

43. Kales A, Kales JD, Jacobson A, et al: Effect ofimipramine on enuretic frequency and sleepstages. Pediatrics 1977;60:431.

44. Mikkelsen EJ, Rapoport JL, Nee L, et al:Childhood enuresis: I. Sleep patterns and psy-chopathology. Arch Gen Psychiatry 1980;37:1139-1144.

45. Bader G, Neveus T, Kruse S, Sillen U: Sleep ofprimary enuretic children and controls. Sleep2002;25:579.

46. Readett DR, Morris JS, Serjeant GR: Nocturnalenuresis in sickle cell haemoglobinopathies.Arch Dis Child 1990;65:290.

47. Mattelaer P, Mersdorf A, Rohrmann D, et al:Biofeedback in the treatment of voiding disor-ders in childhood. Acta Urol Belg 1995;63:5-7.

48. Vogel W, Young M, Primack W: A survey ofphysician use of treatment methods for func-tional enuresis. J Dev Behav Pediatr 1996;17:90-93.

49. Moffatt ME: Nocturnal enuresis: A review of theefficacy of treatments and practical advice forclinicians. J Dev Behav Pediatr 1997;18:49-56.

50. Monda JM, Husmann DA: Primary nocturnalenuresis: A comparison among observation,imipramine, desmopressin acetate and bedwet-ting alarm systems. J Urol 1995;154:745-748.

51. Oredsson AF, Jorgensen TM: Changes in noctur-nal bladder capacity during treatment with thebell and pad for monosymptomatic nocturnalenuresis. J Urol 1998;160:166-169.

52. Moffatt ME, Harios S, Kirshen AJ, Burd L:Desmopressin acetate and nocturnal enuresis:How much do we know? Pediatrics 1993;92:420-425.

53. Skoog SJ, Stokes A, Turner KL: Oral desmo-pressin: A randomized double-blind placebocontrolled study of effectiveness in children withprimary nocturnal enuresis. J Urol 1997;158:1035-1040.


Ch27.qxd 1/25/05 3:34 AM Page 324

54. Thompson S, Rey JM: Functional enuresis: Isdesmopressin the answer? J Am Acad ChildAdolesc Psychiatry 1995;34:266-271.

55. Robson WL, Jackson HP, Blackhurst D, LeungAK: Enuresis in children with attention-deficithyperactivity disorder. South Med J 1997;90:503-505.

56. Bradbury MG, Meadow SR: Combined treatmentwith enuresis alarm and desmopressin for noc-turnal enuresis. Acta Paediatr 1995;84:1014-1018.

57. Bradbury M: Combination therapy for nocturnalenuresis with desmopressin and an alarm device.Scand J Urol Nephrol 1997;183(Suppl):61-63.

58. Fritz GK, Rockney RM, Yeung AS: Plasma levelsand efficacy of imipramine treatment for enure-sis. J Am Acad Child Adolesc Psychiatry1994;33:60-64.

59. Hunsballe JM, Rittig S, Pedersen ES, et al: Singledose imipramine reduces nocturnal urine outputin patients with nocturnal enuresis and noctur-nal polyuria. J Urol 1997;158:830-836.

60. Harari MD, Moulden A: Nocturnal enuresis:What is happening? J Paediatr Child Health2000;36:78-81.

61. Al-Waili NS: Carbamazepine to treat primarynocturnal enuresis: A double-blind study. Eur JMed Res 2000;5:40-44.

62. Bjorkstrom G, Hellstrom AL, Andersson S:Electro acupuncture in the treatment of childrenwith monosymptomatic nocturnal enuresis.Scand J Urol Nephrol 2000;34:21-26.


Ch27.qxd 1/25/05 3:34 AM Page 325

This chapter reviews the differential diagnosis ofsleep disorders in children and describes underwhat circumstances sedative/hypnotic medica-tions are useful for their treatment. It reviewsclasses of medications that affect sleep, with theimplications of their use in a pediatric popula-tion highlighted. It also reviews the mechanismof the sleep-promoting or -disrupting effects ofvarious classes of psychoactive compounds.

CRITICAL ASPECTS OF PEDIATRICSLEEP PHARMACOLOGY

Insomnia and hypersomnolence in children arerelated to the neurotransmitters γ-aminobutyricacid (GABA), serotonin, dopamine, norepineph-rine, histamine, and acetylcholine. The com-pounds that affect sleep share the commonproperty of being excitatory or inhibitory in thebrain in one or more of the above neurotrans-mitter systems.

The receptor mechanisms for the actions ofthe sleep-promoting medications include the lig-ands of the benzodiazepine receptor site, whichallosterically modulates the GABA receptor andits chloride ion channel; serotonin 5-hydrox-ytryptamine (5-HT)2A antagonists; anticholin-ergics/antimuscarinics; antihistamines; andα-adrenergic antagonists.

Examples of the classes of medication activeat these sites include benzodiazepines and ben-zodiazepine-like substances that bind to thebenzodiazepine recognition site; GABA agonistssuch as mood stabilizers; GABA reuptakeinhibitors, which are also mood stabilizers; andbenzodiazepine-mimicking compounds, includ-ing zaleplon and zopiclone. Serotonin 5-HT2Aantagonists include antidepressants such as tra-zodone and nefazodone. Anticholinergic med-ications include tricyclic antidepressants suchas amitriptyline and clomipramine. Antihista-

mines are present in a wide variety of com-pounds, including tricyclic antidepressants.Sedating antihistamines include hydroxyzineand diphenhydramine. α-Adrenergic antago-nists are present in antidepressants and bloodpressure medications.

The classes of medication that disrupt sleepinclude dopamine agonists, adenosine antago-nists, cholinergic agents, and serotonin reuptakeinhibitors. Norepinephrine and dopamine reup-take blockade also results in sleep disruption.Adenosine antagonists such as caffeine disruptsleep.

The selection of a sedating or stimulating psy-chopharmacologic agent for a child is always diag-nosis based.

In children, pharmacologic agents are typi-cally used only when behavioral interventionsare not effective.

The selection of a pharmacologic agent requiresa sleep diagnosis to be firmly established forwhich the agent is suitable. A thorough inter-view with the patient and parents or othersinvolved in managing the child’s sleep is required.A differential evaluation must include a completesleep history, establishing the severity of thesleep difficulty, its frequency, and its duration.The times of the night that the sleep difficultyand undisturbed sleep occur should be estab-lished. The extent to which the child is sleepyor is sleeping during the day should be reviewed.The impact of the sleepiness or tiredness onschool performance is important. Differencesbetween the number of hours of sleep on schoolday and non–school day nights is also importantbecause many children are partially sleepdeprived on school days and obtain reboundsleep on weekends.

Circadian rhythm disorders must be ruled outbefore considering difficulty initiating and main-taining sleep (DIMS). Sleep hygiene problems,

Pharmacology of Sleep Disorders in Children

John H. Herman

Stephen H. Sheldon

28

327

Ch28.qxd 1/25/05 3:35 AM Page 327

such as inappropriate noise or light in the sleep-ing environment, should be reviewed.

Listed below are some of the diagnoses thatmay lead to DIMS in a child, and subsequent sec-tions will review the treatment approachesfocusing on pharmacologic interventions.

Sleep onset association disorderSettling disorder of infancy/childhoodDIMS secondary to sleep phobiaSleep-related panic attacksChildhood-onset psychophysiologic insomniaPhase delay syndromePhase advance syndromeDIMS secondary to anxiety disorderDIMS secondary to mood disorderDIMS secondary to psychosis/schizophreniaNight terrorsNightmaresSleepwalkingHead banging, jactatio capitis nocturnaConfusional arousalDIMS in association with attention deficit hyper-

activity disorder (ADHD)

A comprehensive understanding of the child’semotional state at the time of the sleep difficulty,whether it is sleep onset insomnia or nocturnalawakening, is important. Chronic sleep-relatedanxiety, fear, a depressed mood, worryingthoughts, and apprehension each may be indica-tive of a psychiatric diagnosis. The behavioralrepertoire of the child on awakening should beestablished. Does the child remain in his or herown bed or enter the parent’s bedroom? The man-ner in which the parents respond to the child’ssleep difficulties is also important. Do they capit-ulate to the child’s wish to remain in their bed,or do they take the child to his or her bed?

The extension of any night-time symptomsinto daytime behavior should also be explored.If the symptoms of an anxiety or mood disorderappear to be related to the sleep difficulty, arethere comparable symptoms present during theday and in different situations? Many childrenexpress profound anxiety as bedtime approaches,have disturbed sleep for the first portion of thesleep period, and then settle into undisturbedsleep until morning awakening. Some of thesechildren may show no apprehensiveness withseparation or in other circumstances.

It is important to determine whether eitherbiologic parent or any first-degree relatives have

or have had sleep difficulties similar to thoseof the child. Detailed questioning will fre-quently reveal a first-degree relative with a sim-ilar sleep problem. Inquiry into its course andtreatment is helpful. Anxiety, depressive, andbipolar disorders each have a demonstrablefamily linkage. Adoption studies suggest astrong genetic component in the risk for thesedisorders.

If it is established that the patient (child) hascomorbid psychiatric symptoms and a familymember or members have similar symptoms, thesleep specialist is not diagnosing a psychiatricdisorder in the child but deducing the etiology ofthe sleep disorder and fashioning an approachthat considers psychiatric symptoms.

When both the psychiatric symptoms andDIMS are present, establishing the time course ofthe two is important in diagnosing the sleepdisorder. If the sleep disorder emerges concur-rent with the emotional symptoms and subsideswhen they subside, then treating the emotionaldisorder will most likely result in ameliorationof the sleep symptoms. If the sleep disorder pre-cedes the emotional disorder, then it may bean initial emerging symptom or an independentdisorder.

Before administering a sleeping agent toa child, the child’s family should complete asleep log for at least 2 weeks to provide a base-line measurement of the chief complaint. A tar-get symptom for which the drug is prescribed,such as sleep onset latency, sleep duration, orsleep fragmentation, should be selected. A realis-tic level of improvement, such as reducing sleeponset latency from greater than 1 hour to lessthan 30 minutes, should be discussed with thefamily. The family should maintain the sleep logafter the drug trial begins to enable the assess-ment of the effectiveness of the drug with regardto the specific selected complaint.

Caution in Children When UsingPsychoactive Compounds

All medications used in the treatment of insomniaor parasomnia require close monitoring. Sleep-inducing medications may precipitate nightterrors, sleepwalking, confusional arousal, anddaytime somnolence, which has the potential toexacerbate psychiatric and behavioral problemsand disrupt attention and learning.1

328 Chapter 28 Pharmacology of Sleep in Children

Ch28.qxd 1/25/05 3:35 AM Page 328

The recurrence of parasomnias on withdrawalof sedating medications is common, especially ifwithdrawal is rapid.1 Gradual withdrawal overseveral weeks may be required.

Sleep-Inducing AgentsThe term sedative-hypnotic refers to the ability ofa medication to sedate (or tranquilize) and inducesleep (hypno-). When sedative medications areused to decrease anxiety, they are referred to asanxiolytics. A number of agents are approvedas anxiolytics, among them various antidepres-sants such as doxepin and paroxetine, benzodi-azepines such as alprazolam and diazepam, andbuspirone, a 5-HT1A augmenting agent. Virtuallyevery hypnotic approved by the Food and DrugAdministration is a benzodiazepine or a com-pound that binds to the benzodiazepine receptorsite. Many agents approved as anxiolytics, anti-depressants, and antipsychotics have drowsinessas a side effect, and some physicians prefer theseagents over benzodiazepines despite the absenceof demonstrated efficacy.

Effects of GABA Allosteric Modulators on Sleep

Barbiturates, which are first-generation hyp-notics, increase stage 1 sleep and decrease para-doxical (rapid eye movement [REM]) sleep.Virtually all benzodiazepines, including triazo-lam, midazolam, and diazepam, also decreasesleep onset latency, increase non-REM (NREM)stages 1 and 2 sleep, and decrease REM sleepexcept for midazolam, which increases bothstage 1 and REM sleep at low doses. Zolpidemand zopiclone decrease REM sleep but do notincrease stage 1 sleep.2 Nearly all of these com-pounds have been shown to reduce the propor-tion of slow-wave activity. The activation ofGABAA receptors is involved in the initiationand maintenance of NREM sleep and the gener-ation of sleep spindles but disrupts the processesunderlying slow-wave sleep (SWS).

BENZODIAZEPINES AND SLEEP

All benzodiazepines share certain effects in com-mon—hypnotic, anxiolytic, muscle relaxant,and anticonvulsive—but some have greateremphasis on one or the other of these commoneffects. Therefore, certain benzodiazepines are

more effective as anxiolytics and others as hyp-notics. The speed at onset determines whethera benzodiazepine will be effective in treatingsleep onset insomnia, and the medication half-life determines whether it is effective in treatingdisorders of maintaining sleep or early morningawakening. Benzodiazepines with a short half-life tend to exert more amnestic effects, andthose with a long half-life result in more daytimesleepiness.

ZALEPLON

Zaleplon (Sonata), a nonbenzodiazepine that bindsto the benzodiazepine receptor, has an ultrashorthalf-life, making it useful in the treatment ofsleep onset insomnia. Recent publications sup-port its greater safety in children than zolpidemand benzodiazepines.3 Zaleplon is 14.3 timesmore potent in binding to membrane prepara-tions of the cerebellum than to those of thespinal cord. It produces significant increases inmuscimol binding similar to those of diazepam,and it is antagonized by flumazenil. Zaleplonshows little affinity for other receptors, and itproduces large increases in EEG power of thedelta frequency band without affecting the alphaor beta frequency band. In contrast, intravenousadministration of triazolam and zopicloneincreases the energy of the beta frequency band.The zaleplon-induced increase in the delta fre-quency band is antagonized by pretreatmentwith flumazenil, which does not affect the spon-taneous EEG alone. These results suggest thatzaleplon is a selective, full agonist of the ω1-receptor subtype; thus, zaleplon may induceSWS.4

ZOLPIDEM

Zolpidem (Ambien) has a relatively short half-life and no demonstrable morning hangover andmay be preferable to benzodiazepines withlonger half-lives, especially in children. It haspowerful hypnotic properties but little anxiolyticeffect. Zolpidem is a widely used hypnotic agentacting at the GABAA receptor benzodiazepinesite. It is especially useful in the treatment ofmiddle-of-the-night insomnia or early morningawakening because of its half-life of 2 to 3 hours.On recombinant receptors, zolpidem displaysa high affinity for only the α1-GABAA receptorsand an intermediate affinity for α2- and α3-GABAA receptors. It does not bind to α5-GABAA

Pharmacology of Sleep in Children Chapter 28 329

Ch28.qxd 1/25/05 3:35 AM Page 329

receptors. The sedative action of zolpidem isexclusively mediated by α1-GABA receptors.Similarly, the activity of zolpidem against GABAantagonist–induced tonic convulsions is alsocompletely mediated by α1-GABAA receptors.The sedative/hypnotic and anticonvulsant activ-ities of zolpidem are due to its action on α1-GABAA receptors and not on α2- and α3-GABA2Areceptors.5

TRIAZOLAM

Triazolam (Halcion) is useful for the treatmentof insomnia in children because of its short half-life but is to be avoided because of its amnesticeffects.

Tricyclic Antidepressants and Sleep

Tricyclic antidepressants should be selected withsedative properties taken into consideration. Thesedating or activating properties of antidepres-sants are side effects of the medication. Sedationwith certain compounds will occur on the firstnight of administration, which is beneficial inmany instances and in contrast to the therapeu-tic effect of the medication, which may require2 to 6 weeks to emerge.

Pediatric patients presenting with hypersom-nolence or insomnia whose onset coincided withthat of an affective disorder should be adminis-tered an antidepressant that addresses both theirsleep complaint and mood disorder. If sleep wasnormal until the advent of the mood disorder,then treating the mood disorder typically restoressleep to its levels before the mood disorder episode.That is, all antidepressants, whether they affectsleep or not, will probably result in improvementin the patient’s sleep as the mood disorder istreated. If a sleep complaint emerges or contin-ues independent of a mood complaint in thepast, it will be most efficacious to select an anti-depressant with an appropriate sleep- or wake-up–enhancing profile.

Some antidepressants powerfully promote orinterfere with sleep. Tricyclic antidepressantshave been used to treat children for many years.There is little evidence that they are effectiveantidepressants in children. However, those withsedating properties may be preferred in cases inwhich insomnia is present.

Most sedationAmitriptyline

LomipramineDoxepinTrimipramine

Moderate sedationImipramineNortriptyline

Least sedationAmoxapineDesipramineProtriptyline

For the hypersomnolent, anergic, depressedchild with psychomotor slowing, protriptylineis the most activating tricyclic antidepressant.Agitation and anxiety are its two most prevalentside effects. Doxepin, which is approved for bothdepression and anxiety, is effective in depressedindividuals with insomnia, especially with mixedanxiety/depression symptoms.

Selection of Serotonin-Specific ReuptakeInhibitors (Not So Specific)

Selective serotonin reuptake inhibitors (SSRIs)cause insomnia by stimulating 5-HT2A receptorsin brainstem and forebrain sleep-promoting cen-ters. 5-HT2A receptor stimulation results inincreased wakefulness and disruption of SWS.

More stimulationFluoxetineSertraline

Less stimulationFluvoxamineParoxetine

No stimulationCitalopram

Several members of this class of medicationhave demonstrated efficacy in children in con-trolled trials. Various studies have shown thatSSRIs disrupt sleep. Their administration gener-ally results in reduced SWS, greatly reduced REMsleep, and increased stage 1 sleep. The mecha-nism of action is believed to be the activationof 5-HT2A receptors.

Bupropion and Sleep

Bupropion’s mechanism of action is as a norep-inephrine and dopamine reuptake inhibitor.It inhibits sleep by increasing dopamine andnorepinephrine at the synaptic cleft and is the


Ch28.qxd 1/25/05 3:35 AM Page 330

least preferred for insomniac patients. It is themost preferred for retarded depression, whichpresents with cognitive slowing and hypersom-nia. It is also frequently prescribed for patientswho do not respond to serotonergic agents,including SSRIs and tricyclic antidepressants.Bupropion can be added to SSRIs to increasetheir clinical efficacy.

Serotonin and NorepinephrineReuptake Inhibitors

Venlafaxine (Effexor), a dual reuptake inhibitor,and serotonergic and noradrenergic reuptakeinhibitors may be more rapid in onset and moreeffective in severe depression. It is the most pre-ferred in depressed patients who present withhypersomnia, cognitive slowing, and anergia.Venlafaxine causes insomnia at medium dosesand severe insomnia at high doses due to block-ade at the NE, 5-HT and DA receptors.

Noradrenergic and SpecificSerotonergic Antidepressants

The mechanism of action of mirtazapine(Remeron) is as a noradrenergic and specificserotonergic 5-HT2A antagonist. It is sleep enhanc-ing due to its 5-HT2A antagonism, antihistamin-ergic actions, and alpha2 antagonist property,which disinhibits both the NE and 5-HT recep-tors. It is specifically indicated in depressedpatients with insomnia.

Serotonin Antagonists/ReuptakeInhibitors

Serotonin 5-HT2A antagonists/reuptake inhibitorsare sleep enhancing. Nefazodone, a member ofthis class, causes norepinephrine reuptake block-ade and acts as an antagonist at the serotonin5-HT2A receptor. A major side effect is sedation.Perhaps the most sedating of all antidepressantsis trazodone, which both acts as an antagonist atthe 5-HT2A receptor and blocks histamine recep-tors, leading to two sedating mechanisms.

Antiseizure Medications and Sleep

Valproic acid (Depakene) is used as an anti-seizure medication and in the treatment of bipo-lar disorder. It has sedating effects and may be

used effectively in the treatment of insomniaassociated with behavioral agitation. Carbama-zepine (Tegretol) is also used principally as ananti-seizure medication and in the treatmentof bipolar disorder. It also has sedating effects.Gabapentin (Neurontin) is sedating and appearsto increase SWS.

Atypical Antipsychotics and Sleep

Atypical antipsychotics that are sedating includeziprasidone, quetiapine, and olanzapine. Risper-idone, another atypical antipsychotic, has insom-nia as a side effect. Clozapine (Clozaril) is notdiscussed because of the risk of agranulocytosis.Atypical antipsychotics at low doses are extremelyuseful in treating severe agitation in combinationwith insomnia. Because they lack many of theside effects of typical neuroleptics, they are morefrequently administered to individuals who do notmeet the criteria of the Diagnostic and StatisticalManual–IV for psychosis or schizophrenia butare agitated due to disrupted sleep.

Combination Medications forDepression and Insomnia

Sedating agents such as trazodone, nefazodone,a sedating atypical antipsychotic, or a benzodi-azepine may be added to an SSRI at bedtime forthe treatment of insomnia with depression. Thedose of the SSRI should be considered, as theaddition of a second agent may result in increasedside effects.

Most benzodiazepines have half-lives that aretoo long to be effective in children.

Pharmacologic Treatmentof Specific Sleep Disordersin Children and AdolescentsSleep Onset Insomnia, Settling Disorders,and Sleep-Related Anxiety

Before prescribing medication, a differential diag-nosis should be performed differentiating a trueinsomnia from a variety of other sleep and psychi-atric disorders that result in difficulty in initiat-ing sleep. Sleep onset insomnia in a child consistsof minimum bedtime resistance and an elon-gated period of wakefulness without attempts atinteraction with caretakers. In contrast, settling


Ch28.qxd 1/25/05 3:35 AM Page 331

disorders are characterized by bedtime strugglesand/or repeated attempts at interaction withcaretakers. Settling disorders may be conceptual-ized as normal, early developmental variations ofoppositional defiant disorder. Sleep-related anx-iety is characterized by an increase in autonomicand central nervous system symptoms of anxietyand fear related to bedtime and sleep. Sleep-related anxiety may be a normal variation of theanxiety disorders as the child struggles to dealwith the early developmental task of individua-tion/separation.

It is important to differentiate among sleeponset insomnia, settling disorders, and sleep-related anxiety, as each is treated differently. Sleeponset insomnia is treated with behavioral ther-apy approaches such as stimulus control therapy,sleep restriction therapy, relaxation therapy, andsleep hygiene. It may also be treated with sedat-ing medications in certain circumstances. If sleepmaintenance is not an issue, a medication witha rapid onset and short half-life, such as zaleplon,is preferred.

Settling disorders are treated with a variety oftechniques. These behavioral techniques requireclose cooperation among all caretakers involvedin the child’s sleep initiation. Medications arenot indicated and have not been found to be effi-cacious in treating settling disorders. Prolongedsettling disorders may indicate other psychiatricdiagnoses.

Sleep-related anxiety is treated with approachesvirtually the opposite of settling disorders. A care-taker in the child’s presence, typically on a palatenext to the child’s bed, is required to minimize anx-iety. Antianxiety benzodiazepines with a rapidonset, such as diazepam, may be helpful for briefperiods. The dose should be low enough so asnot to induce daytime somnolence.

Most children with sleep onset insomnia, incontrast to those with a settling disorder, will notbe brought to the attention of a physician becausethey tend to remain in their own beds and notdisturb caretakers. Adults with childhood-onsetinsomnia come to the attention of physiciansonly after they reach their twenties or older, whenpoor sleep quality results in daytime fatigue. Incontrast to sleep onset insomnia, much more fre-quently observed is psychophysiologic insom-nia, which consists of DIMS.

Children with sleep difficulties had signifi-cantly increased odds of anxiety/depression based

on the mother’s but not the teacher’s reports. Theassociation increased with age and was independ-ent of the mother’s history of major depressivedisorder. The association between sleep difficul-ties and anxiety/depression was greater than thatfor other psychiatric problems.6

Circadian phase disorders and psychiatric dis-orders often have sleep onset insomnia as a pre-senting complaint, as do anxiety disorders anddepression. Although delayed sleep phase syn-drome may be treated with melatonin adminis-tered in the early evening, behavioral techniques,namely exposure to early morning bright light,are usually a first-line treatment in children (seeTable 9–2 in Chapter 9).

A 5-mg dose of melatonin at 6:00 PM is safefor short-term use in uncomplicated sleep onsetinsomnia and significantly more effective thanplacebo in advancing sleep onset and dim-lightmelatonin onset and increasing sleep duration inelementary school children. These results wereconfirmed in children with chronic sleep onsetinsomnia in a placebo-controlled trial. Sustainedattention was not affected.7 Other pharmaco-logic choices are described below.

Benzodiazepines are indicated mainly for tran-sient or short-term insomnia, for which prescrip-tions should be limited to a few days, occasionalor intermittent use, or courses not to exceed2 weeks, if possible. Temazepam, loprazolam, andlormetazepam, which have a medium durationof action, are suitable. Diazepam is also effectivein single or intermittent doses. Potent, short-acting benzodiazepines such as triazolam appearto carry a greater risk of adverse effects.8

Sleep Phobia

Children sometimes develop a specific phobicreaction to sleeping alone, with all the character-istics of a specific phobia, such as fear of flyingor fear of blood for adults. The phobia includesa panicky reaction, with autonomic activationand apparent terror, as if the child’s life is in dan-ger. This reaction is distinct from settling disor-ders, in which an overwhelming sense of palpableterror is not present, and anxiety disorders, inwhich the symptoms are present in many situa-tions throughout the day.

Specific phobias in adults are treated withbehavioral therapy, typically a deconditioningmodel. The same approach is the most effective


Ch28.qxd 1/25/05 3:35 AM Page 332

in children, which consists of introducing thechild to sleeping in his or her own room ina gradual and nonthreatening manner. Medicationsare a second-line treatment, most typically a ben-zodiazepine with anxiolytic properties.

The sleep specialist must carefully devisea plan, the first step of which is to achieve sepa-ration that is tolerated without overwhelminganxiety by the parents as well as the child. Thisoutcome may be realized with a parent on a cotin the child’s room. Considerable skill is requiredto develop a stepwise program that is neithertoo draconian nor overcapitulates to the child’sdemands.

Settling Disorders

Sleep difficulties characterized by the need forinteraction with caretakers, namely having a par-ent in the child’s bed or the child wishing to be inthe parents’ bed, are best treated behaviorally.These difficulties are settling disorders, in whichthe child repeatedly seeks the parents’ presence. Asettling disorder should be differentiated froma sleep phobia, as the latter is marked by over-whelming panic that would be exacerbated bya technique such as allowing a child to cry for aset period of time. Ferber describes techniques fordealing with settling disorders, which are varia-tions of behavioral therapy techniques. Flexibilityof approach is the hallmark of these interven-tions, the object of which is to find one that is tol-erable for both the parents and the child.

Children aged 6 to 12 months with severe,chronic sleep disorders such as DIMS were com-pared to peers without sleep problems. At theage of 5.5 years, 7 of the 12 children in the sleepdisorder group met the criteria for the diagnosisof ADHD. None of the control children metthese criteria.9 Sleep problems at 4 years of agepredict behavioral/emotional problems inmidadolescence after controlling for sex, adop-tion status, and behavioral/emotional status. Thecorrelation between sleep problems and depres-sion/anxiety increased significantly from age4 years to midadolescence.10

Medications appear to be effective in treatingnight-time awakening in the short term, butlong-term efficacy has not been demonstrated.In contrast, specific behavioral interventionshave consistently shown both short-term effi-cacy and the possible long-term effects of deal-

ing with settling problems and night-time awak-ening.11

Hyperactivity is associated with both increasedbody movement during sleep and sleepwalking;behavioral problems are related to bedtime strug-gles. The symptoms of depression and anxietyare associated with night terrors, difficulty fallingasleep, and daytime somnolence.12

Behavioral extinction techniques as well asparent education are effective in preventing sleepproblems. They are now considered to be wellestablished treatments. Graduated extinction andscheduled awakenings appear to be effectivebehavioral treatments.13

Less sleep at night is associated with a psy-chiatric diagnosis in children, and less night-timesleep and less sleep in a 24-hour period are asso-ciated with increased behavior problems andmore externalization of problems.14 Others havefocused on the relationship between the regula-tion of sleep and the control of attention, emo-tion, and behavior.15

Confusional Arousal

No controlled studies of the effects of medica-tion have been performed. In severe cases, clini-cal reviews indicate that benzodiazepines andtricyclics are effective.16 Because antihistaminesare sedating and do not consistently have anxi-olytic effects, they may increase the likelihoodof parasomnia.

Night Terrors

Night terrors should be distinguished from night-mares (i.e., anxiety dreams) and sleep-relatedpanic attacks. Night terrors occur early in thenight with no morning recollection and are char-acterized by intense autonomic arousal and dis-oriented behavior. Nightmares typically occurlater in the sleep period but are more variable intheir time of occurrence, and typically a vividdream with a narrative plot is reported. Nightterrors emerge from stages 3 and 4 sleep, andnightmares emerge from REM sleep. Sleep-relatedpanic attacks are similar to night terrors in theseverity of perceived dread but are characterizedby morning recall and the concurrent develop-ment of fear of the bedroom and sleep, which areassociated in the child’s mind with the possibil-ity of recurrence.


Ch28.qxd 1/25/05 3:35 AM Page 333

A sleep and behavioral/psychiatric disorderquestionnaire showed that adolescents withsleep terrors and sleepwalking have an increasedprevalence of neurotic traits, psychiatric disor-ders, and behavioral problems. Sleep terrors andsleepwalking in childhood are probably princi-pally related to genetic and developmental fac-tors and their continuation; in particular, theironset in adolescence may be related to psycho-logical factors.17

It is not clear whether the hypnotic or anxi-olytic properties of benzodiazepines constitutethe mechanism of action of these drugs in treat-ing night terrors. There are no controlled studiesof the effects of these medications. Diazepamat bedtime is the standard treatment, but mida-zolam, oxazepam, lorazepam, or clonazepam, issometimes preferred.16

One study of night terror episodes18 featuredconfused behaviors, motor activity, and absentor fragmented recall. Polysomnography docu-mented arousal from SWS in 9 of 10 patients. Allof the original patients reported psychiatricsymptoms. All 6 patients who received the sub-sequent structured evaluation met the lifetimecriteria for Axis I conditions (most commonlyaffective and substance abuse disorders) and hadelevated scores on the personality scale of theMillon Clinical Multiaxial Inventory–II. Nightterrors were not limited to psychiatric episodes.The episodes that occur in adults are similarto those that are described in children. Whilethey are distinct from sleep-related panic attacks,night terrors appear to occur in adults with his-tories of psychopathology.18

Sleep-Related Panic Attacks

In this placebo-controlled trial, clonazepam wasan efficacious and safe short-term treatment forpanic disorder. Discontinuation during and afterslow tapering was well tolerated.19

Panic attacks may be induced during stages3 and 4 sleep by the administration of high dosesof caffeine to sleeping subjects.20

Fluoxetine has been found to be effective inthe treatment of a variety of anxiety disorders inchildren who were not responsive to psychother-apy, including separation anxiety, social phobia,generalized anxiety disorder, and panic disorder.Improvement occurs over a 5-week period. Dosesvary and should be individualized for each clin-ical situation.21

Children with panic disorder show greaterchanges in heart and respiratory rates when lac-tate is infused during sleep.22 Patients with a his-tory of sleep panic have higher rates of generalizedanxiety disorder, social phobia, and majordepression. The presence of sleep-related panicattacks may delineate a subgroup of panic disor-der patients who have early difficulties with anx-iety and comorbid mood and anxiety disordersas adults.23

Panic disorder frequently appears first duringadolescence and may occur concurrent with sleep-related panic attacks or night terrors. Besidespanic attacks and avoidance behavior, adoles-cents often have sleep initiation and continua-tion disturbances. They suffer from insomnia,nocturnal panic attacks, and fear of going to bedor falling asleep.24,25

Concurrent acute onset of night terrors, som-nambulism, and spontaneous daytime panicattacks meeting the criteria for panic disorderis reported in a 10-year-old-boy with a familyhistory of panic disorder (unpublished anec-dote). Both the parasomnias and the panic disor-der were fully responsive to therapeutic doses ofimipramine. A second case of night terrors andinfrequent full-symptom panic attacks is notedin another 10-year-old boy whose mother haspanic disorder with agoraphobia. The clinicalresemblance and reported difference betweennight terrors and panic attacks are described.The absence of previous reports of this comor-bidity is notable. It is hypothesized that nightterrors and panic disorder involve a similar con-stitutional vulnerability to dysregulation ofbrainstem-altering systems.

Sleepwalking

There are no controlled studies of the effects ofmedication. Clinical reviews indicate triazolam,lorazepam, diazepam, and clonazepam or imipra-mine, desipramine, and clomipramine may beused at bedtime. Another medication that hassome demonstrated efficacy is carbamazepine.16

Nocturnal Enuresis

In this disorder, only the results of long-term fol-lowup are meaningful. Most studies of desmo-pressin (DDAVP) show it to be statistically moreeffective than placebo after 2 to 4 weeks of use.This finding is not clinically useful, as symptom


Ch28.qxd 1/25/05 3:35 AM Page 334

resumption frequently follows withdrawal. BothDDAVP and imipramine are effective in short-term studies.26 Those most likely to be perma-nently dry with DDAVP are older children whorespond to 20 μg and do not wet the bed fre-quently.27 Oxybutynin (Ditropan) for bladderspasms is the most effective when nocturnal enure-sis is combined with daytime incontinence.28

Treatment with tricyclic antidepressant drugs(imipramine, amitriptyline, lomipramine and des-ipramine) reduces bedwetting by about one nightper week. DDAVP and tricyclic antidepressantsappear to be equally effective while on treatment,but this effect is not sustained after treatmentis stopped. Alarms appear to be more effective inthe long term.29

Treatment with DDAVP is more effective thanretention control training in decreasing the num-ber of wet nights but not effective in childrenwith a low functional bladder capacity. Daytimefunctional bladder capacity predicts a response toDDAVP but not to retention control training.30

In a 9-week study, there is a greater effect ofDDAVP combined with alarm therapy than eitheralone. However, combined treatment and eitherDDAVP or alarm therapy for 9 weeks have a6 month follow-up success rate of about 37%.31

Treatment of Insomnia with Moderateto Severe Developmental Disabilities

Eighteen of 20 children with developmental dis-abilities had shorter sleep latencies when receiv-ing melatonin than placebo in a double-blind,6-week trial. The greater the sleep latency atbaseline, the more it decreased with melatonin.However, melatonin does not increase sleep dura-tion, nor does it decrease the number of awak-enings. No side effects have been reported whenmelatonin is used in this manner.32

Sleeplessness after PinealTumor Extraction

Melatonin replacement therapy is beneficial forthose patients who have deficient melatonin syn-thesis from both DIMS and circadian rhythm dis-orders.33

Sleepless Children Who Do NotTolerate Oral Medication

Eight children who were inpatients at Children’sHospital in Boston or at Children’s Hospital at

Stanford (aged 5 to 16.6 years) and were unableto tolerate oral medications received intravenousamitriptyline for neuropathic pain, depression,and sleep disturbance and as an adjuvant agentfor opioid analgesia. One patient experiencedan extrapyramidal reaction that was successfullymanaged with diphenhydramine.34

Insomnia in Children Secondaryto Anxiety Disorders

Listed above are settling disorders, sleep onsetinsomnia, and sleep phobia. In contrast to eachof these sleep problems, insomnia secondary toan anxiety disorder entails generalized anxiety ina wide variety of circumstances, chronic worry-ing, and frequently oversensitivity to the feelingsof others. These children may have a variety ofthe symptoms of anxiety and a family historyof anxiety disorders.

Currently favored medications for anxietydisorders, including generalized anxiety disor-der, panic disorder, phobias, and phobic disorders,are antidepressants, especially SSRIs, and bu-spirone. The heterocyclic doxepin and the SSRIparoxetine are both approved for anxiety disor-ders and have sedating side effects. In addition,anxiety disorders frequently overlap withdepressive disorders, varying from major depres-sive disorder with subsyndromal anxiety in gen-eralized anxiety disorder with dysthymia. Bothdepressive symptoms and anxiety level may varyindependent of each other.

If the onset of insomnia is coincident with anexacerbation of anxiety symptoms, then it shouldfirst be addressed with antianxiety medications,frequently SSRIs such as paroxetine, which issedating, and the heterocyclic doxepin. For gen-eralized anxiety disorder, the serotonergic 5-HT1Aagonist buspirone is a first-line treatment.Clonidine is helpful in blocking the noradrener-gic aspects of anxiety, such as tachycardia, sweat-ing, and tremor.

Benzodiazepines are useful in the treatment ofthe subjective aspects of anxiety, especially whena rapid response is required. They are helpful intreating residual insomnia after symptom relieffrom the underlying anxiety disorder has beenaddressed. The risk–benefit ratio must be con-sidered carefully before prescribing an agentfrom this class to a child because of the possibil-ity of daytime sedation, abuse, and addiction.Short–half-life benzodiazepine receptor agonists


Ch28.qxd 1/25/05 3:35 AM Page 335

such as zaleplon are preferred because of lessdaytime sedation. Long–half-life medications inthis class, such as clonazepam, are preferred whentheir daytime anxiolytic effects are sought.

Long-term Medication for DifficultiesInitiating and Maintaining Sleepin Children

In the treatment of long-standing DIMS in chil-dren, antidepressants are the most frequentlyused class of medication. Brief treatment mostfrequently uses antihistamines. The most sedat-ing antidepressants are amitriptyline, doxepin,nefazodone, and trazodone.

In laboratory studies in adults, SSRIs (fluoxe-tine, paroxetine, fluvoxamine, and sertraline) sig-nificantly disrupt sleep, with the exception ofcitalopram (Celexa).

Zolpidem (Ambien) and zaleplon (Sonata) arenonbenzodiazepines that are active at the benzo-diazepine receptor site. They have less morningsedation and result in fewer withdrawal symp-toms than classic benzodiazepines. Zolpidemat bedtime is indicated for late-night sleep main-tenance and early morning awakening; bothstart at 5-mg doses for adults. Zaleplon, with itsultrashort half-life, is indicated for sleep onsetinsomnia. The use of these medications in chil-dren has not been studied extensively, but theirfavorable side effect profiles, including less day-time sedation, suggest that they might be welltolerated.

All of the antidepressants that increase the sleeptendency suppress REM sleep, with the exceptionof nefazodone and trazodone. The behavioral,developmental, and physiologic effects of long-term benzodiazepine administration in childrenhave not been studied.

Treatment of Resistant, Severe DIMS

Medications with different mechanisms of action,including antihistaminergic, anticholinergic,serotonergic, benzodiazepine, and antiseizureagents, should be rotated. A combination ofmedications, each at low doses, from two or threeclasses may be preferred over increasing thedose.

It is becoming more common to prescribeatypical antipsychotics with sedating profiles

to children and adolescents who have agitationand DIMS that do not respond to other treat-ments. Unlike the classic neuroleptics, thesemedications have fewer extrapyramidal sideeffects, especially at low doses. These agentsinclude ziprasidone (Geodon), quetiapine(Seroquel), and olanzapine (Zyprexa) but notrisperidone (Risperdal), which has insomnia asa side effect.

Disrupted Sleep in Children withAttention Deficit Hyperactivity Disorder

The parents of children with ADHD state thattheir children show a high rate of sleep prob-lems, more than those described by theparents of normal control subjects. Sleep dif-ferences in ADHD are not verified by actigra-phy or sleep diary data, with the exceptionof longer sleep duration and increased bed-time resistance. Interactions during bedtimeroutines are more challenging in ADHD,which may be the basis of the perceived sleepproblems.35

Most sleep disturbances in ADHD are secondaryto stimulant treatment and comorbid anxiety andbehavior disorders. ADHD cannot be explainedas a consequence of sleep disruption.36 Childrenwith ADHD are not different from control sub-jects on polysomnographic sleep variables, butvideo analysis shows more upper and lower limbmovements.37

Although no polysomnographic differencesare found, boys with ADHD fall asleep more rap-idly on the multiple sleep latency test than chil-dren without symptoms of ADHD. Its resultscorrelate with hyperactivity, impulsivity, and inat-tentiveness. Children with ADHD are more sleepyduring the day due to a deficit in alertness andnot poorer sleep quality.38

Children with bipolar disorder differ fromthose with ADHD in the presence of mania-spe-cific symptoms, including grandiosity, and theytypically show ultrarapid or ultradian cycling.The sleep disorders of bipolar children are besttreated with mood stabilizers.39

Sleep disturbances in children with ADHDshould be addressed behaviorally and/or bymodifying stimulant schedules and doses andnot by sedatives/hypnotics unless an independ-ent sleep diagnosis is verified.


Ch28.qxd 1/25/05 3:35 AM Page 336

References

1. Broughton RJ: NREM arousal parasomnias. InKryger MH, Roth T, Dement WC (eds): Principlesand Practice of Sleep Medicine, 3rd ed.Philadelphia, WB Saunders, 2000.

2. Gottesmann C, Gandolfo G, Arnaud C, GauthierP:The intermediate stage and paradoxical sleepin the rat: Influence of three generations of hyp-notics. Eur J Neurosci 1998;10:409-414.

3. Israel A, Kramer J: Safety of zaleplon in the treat-ment of insomnia. Ann Pharmacother 2002;36:852-859.

4. Noguchi H, Kitazumi K, Mori M, Shiba T: Bindingand neuropharmacological profile of zaleplon,a novel nonbenzodiazepine sedative/hypnotic.Eur J Pharmacol 2002;434:21-28.

5. Crestani F, Martin JR, Mohler H, Rudolph U:Mechanism of action of the hypnotic zolpidemin vivo. Br J Pharmacol 2000;131:1251-1254.

6. Johnson EO, Chilcoat HD, Breslau N: Troublesleeping and anxiety/depression in childhood.Psychiatry Res 2000;94:93-102.

7. Smits MG, Nagtegaal EE, van der Heijden J, et al:J Child Neurol 2001;16:86-92.

8. Ashton H: Guidelines for the rational use of ben-zodiazepines: When and what to use. Drugs1994;48:25-40.

9. Thunstrom M: Severe sleep problems in infancyassociated with subsequent development ofattention-deficit/hyperactivity disorder at 5.5years of age. Acta Paediatr 2002;91:584-592.

10. Gregory AM, O’Connor TG: Sleep problems inchildhood: A longitudinal study of developmen-tal change and association with behavioral prob-lems. J Am Acad Child Adolesc Psychiatry2002;41:964-971.

11. Ramchandani P, Webb VV, Stores G: A systematicreview of treatment of settling problems and nightwaking in young children. West J Med 2000;173:33-38.

12. Smedje H, Broman JE, Hetta J: Associationsbetween disturbed sleep and behavioral diffi-culties in 635 children aged six to eight years:A study based on parent’s perceptions. Eur ChildAdolesc Psychiatry 2001;10:1-9.

13. Mindell JA: Empirically supported treatments inpediatric psychology: Bedtime refusal and nightwakings in young children. J Pediatr Psychol1999;24:465-481.

14. Lavigne JV, Arend R, Rosenbaum D, Smith A, et al:Sleep and behavior problems among preschoolers.J Dev Behav Pediatr 1999;20:164-169.

15. Dahl RE: The development and disorders of sleep.Adv Pediatr 1998;45:73-90.

16. Broughton R: Principles and Practice of SleepMedicine. 2000.

17. Gau SF, Soong WT: Psychiatric comorbidity ofadolescents with sleep terrors or sleepwalking:A case-control study. Aust NZ J Psychiatry 1999;33:734-739.

18. Llorente MD, Currier MB, Norman SE, et al:Night terrors in adults: Phenomenology and rela-tionship to psychopathology. J Clin Psychiatry1992;53:392-394.

19. Moroz G, Rosenbaum JF: Efficacy, safety, andgradual discontinuation of clonazepam in panicdisorder: A placebo-controlled, multicenterstudy using optimized dosages. J Clin Psychiatry1998;60:604-612.

20. Koenigsberg HW, Pollak CP, Ferro D: Can panicbe induced in deep sleep? Examining the neces-sity of cognitive processing for panic. DepressAnxiety 1998;8:126-130.

21. Fairbanks JM, Pine DS, Tancer NK, et al: Openfluoxetine treatment of mixed anxiety disordersin children and adolescents. J Child AdolescPsychopharmacol 1997;7:17-29.

22. Koenigsberg HW, Pollak CP, Fine J, et al: Cardiacand respiratory activity in panic disorder: Effectsof sleep and sleep lactate infusions. AmJ Psychiatry 1994;151:1148-1152.

23. Labbate LA, Pollack MH, Otto MW, et al: Sleeppanic attacks: An association with childhood anx-iety and adult psychopathology. Biol Psychiatry1994;36:57-60.

24. Lepola U, Koponen H, Leinonen E: Sleep in panicdisorders. J Psychosom Res 1994;38(Suppl 1):105-111.

25. Garland EJ, Smith DH: Simultaneous prepubertalonset of panic disorder, night terrors, and som-nambulism. J Am Acad Child Adolesc Psychiatry1991;30:553-555.

26. Yang S, Chiou YH, Lin CY, et al: Treatment guide-lines of enuresis in Taiwan. Acta Paediatr Taiwan2001;42:271-277.

27. Kruse S, Hellstrom AL, Hanson E, et al: Treatmentof primary symptomatic nocturnal enuresis withdesmopressin: Predictive factors. BJU Int 2001;88:572-576.

28. Neveus T: Oxybutynin, desmopressin andenuresis. J Urol 2001;166:2459-2462.

29. Glazener CM, Evans JH, Peto RE: Tricyclic andrelated drugs for nocturnal enuresis in children.Cochrane Database Syst Rev 2000;CD002117.

30. Hamano S, Yamanishi T, Igarashi T, et al:Functional bladder capacity as predictor ofresponse to desmopressin and retention controltraining in monosymptomatic nocturnal enure-sis. Eur Urol 2000;37:718-722.

31. Leebeek-Groenewegen A, Blom J, Sukhai R, vander Heijden B: Efficacy of desmopressin com-bined with alarm therapy for monosymptomaticnocturnal enuresis. J Urol 2001;166:2456-2458.


Ch28.qxd 1/25/05 3:35 AM Page 337

32. Dodge NN, Wilson GA: Melatonin for treatmentof sleep disorders in children with developmen-tal disabilities. J Child Neurol 2001;16:581-584.

33. Jan J, Tai J, Hahn G, Rothstein RR: Melatoninreplacement therapy in a child with a pineal tumor.J Child Neurol 2001;16:139-140.

34. Collins JJ, Kerner J, Sentivany S, Berde CB:Intravenous amitriptyline in pediatrics. J PainSymptom Manage 1995;10:471-475.

35. Corkum P, Tannock R, Moldofsky H, et al:Actigraphy and parental ratings of sleep in chil-dren with attention-deficit/hyperactivity disor-der (ADHD). Sleep 2001;24:303-312.

36. Mick E, Biederman J, Jetton J, Faraone SV: Sleepdisturbances associated with attention deficithyperactivity disorder: The impact of psychiatric

comorbidity and pharmacotherapy. J ChildAdolesc Psychopharmacol 2000;10:223-231.

37. Konofal E, Lecendreux M, Bouvard MP, Mouren-Simeioni MC: High levels of nocturnal activity inchildren with attention-deficit hyperactivity dis-order: A video analysis. Psychiatry Clin Neurosci2001;55:97-103.

38. Lecendreux M, Konofal E, Bouvard M, et al: Sleepand alertness in children with ADHD. J ChildPsychol Psychiatry 2000;41:803-812.

39. Geller B, Williams M, Zimerman B, et al:Prepubertal and early adolescent bipolarity differ-entiate from ADHD by manic symptoms, grandiosedelusions, ultra-rapid or ultradian cycling. J AffectDisord 1998;51:81-91.


Ch28.qxd 1/25/05 3:35 AM Page 338

Acetaminophen, after adenotonsillectomy, 256Achondroplasia, sleep disorders in, 23tAcrophase, circadian, 91Actigraphy, 62, 68Active sleep, in neonates

body movements during, 60–61electroencephalogram during, 52, 54f

Adaptive theory of sleep, 2Adenoma, pituitary, narcolepsy and, 175Adenotonsillar hypertrophy

chronic, 251diagnosis of, 250–251obstructive sleep apnea syndrome and, 211, 224–225

Adenotonsillectomyadditional surgical procedures and, 258–259adenoidal regrowth after, 237ambulatory, 255behavior disorders and, 163–164cognitive changes and, 164–165complications after, 256–257efficacy of, 235–236follow-up after, 238growth after, 213in obstructive sleep apnea syndrome, 235–238, 237t,

257–258in upper airway resistance syndrome, 253–254, 257–258indications for, 253–254morbidity and mortality of, 237–238, 237tneuromuscular disorders and, 258obesity and, 258outcomes of, 257–258polysomnography in

follow-up, 260postoperative, 256preoperative, 252–253

postoperative management in, 255–256preoperative evaluation in, 254respiratory compromise after, 237–238, 237t, 255surgical techniques in, 254–255tonsillectomy alone vs., 236–237

ADHD. See Attention-deficit/hyperactivity disorder.Adolescents

circadian rhythm in, 88, 96delayed sleep phase syndrome in, 105, 145–146excessive daytime sleepiness in, 183sleep problems in, impact of, 30

α-Adrenergic antagonists, 327Advanced sleep phase syndrome, 109–110

differential diagnosis of, 110

Advanced sleep phase syndrome (Continued)epidemiology of, 109–110insomnia in, 145, 153–154prevalence of, 102temperature curves in, 101, 102ftreatment of, 110, 153–154

Affective disorders. See Mood disorders.Aggressiveness, in partial arousal disorders, 284Airflow

absence of. See Apnea.limitation of, 206toronasal

in obstructive sleep apnea syndrome, 202, 203fmeasurement of, 65, 67

Airwaynasal obstruction of, 250upper. See also Pharynx.

abnormalities of, obstructive sleep apnea syndromeand, 275

obstruction of. See High upper airway resistance;Obstructive sleep apnea syndrome; Upper airwayresistance syndrome.

respiratory functions of, 275–276Starling resistor model of, 223, 224f

Alcohol, in case-based analysis of sleep problems, 44, 46tAllergic rhinitis, nasal obstruction from, 250Allergy

cow’s milk, insomnia in, 138–139in obstructive sleep apnea syndrome, 243sleep problems in, 32

Alpha rhythms, development of, 56Amitriptyline, 330, 335Amnesia, in partial arousal disorders, 284Amoxapine, 330Amphetamine

in idiopathic hypersomnia, 187tin Kleine-Levin syndrome, 195in narcolepsy, 178t, 179in post-traumatic hypersomnia, 191

Analgesics, after adenotonsillectomy, 256Anterior tibial muscle electromyogram, 60, 65, 68Antibiotics

in acute tonsillitis, 251in chronic adenotonsillar hypertrophy, 251insomnia from, 142

Anticholinergic medications, 327Anticonvulsants

hypersomnia from, vs. idiopathic hypersomnia, 185sleep and, 331

Index

Note: Page numbers followed by f and t refer to figures and tables, respectively.

339

Index.qxd 1/25/05 3:36 AM Page 339

Antidepressantsin anxiety disorders, 335in insomnia, 336insomnia from, 19, 20noradrenergic and specific serotonergic, 331tricyclic

in arousal disorders, 302in confusional arousals, 333in sleep-related enuresis, 323, 335sleep and, 330

Antidiuretic hormone, enuresis and, 232, 318Antiepileptics. See Anticonvulsants.Antihistamines

adverse effects of, 142sedating, 327

Antipsychotics, atypicalin insomnia, 336sleep and, 331

Antiseizure medications. See Anticonvulsants.Anxiety

dream, 286–287, 310–311, 311tsleep-related

differential diagnosis of, 331–332treatment of, 108t, 332

vs. delayed sleep phase syndrome, 103tAnxiety disorders

insomnia in, 335–336treatment of, 335

Anxiolytics, 329Apnea

central, 206t, 272, 274fexpiratory, 273, 276fmixed, 206t, 272, 275fmyelomeningocele with, respiratory patterns in, 289–290obstructive, 206t, 272, 273f, 276of prematurity, 276–278post-sigh, 273–274, 277fsleep, obstructive. See Obstructive sleep apnea syndrome.

Apnea-hypopnea index, 69, 204, 204tApnea index, 69, 204, 204tArousal(s)

confusional, 283–284, 295t, 296treatment of, 302, 333vs. somnambulism, 285

in attention-deficit/hyperactivity disorder, 161in obstructive sleep apnea syndrome, 201respiratory effort–related, 206t

Arousal disorders, 283–286, 288, 293–302. See also Sleepterrors; Somnambulism.

clinical description of, 293–296clinical evaluation of, 299, 300t, 301tconditions that mimic or trigger, 299, 301tgenetics of, 298partial, 283–284, 288pathophysiology of, 296–298polysomnography in, 298–299, 300–301sleep history in, 299, 300tspectrum of, 293, 295timing of, 293, 294ftreatment of, 301–302

Asperger’s syndromehypersomnia in, 22sleep disorders in, 23t

Associative memory, sleep and, 4, 6Asthma, sleep problems in, 20, 32Atrial natriuretic peptide, in obstructive sleep apnea

syndrome, enuresis and, 232Attention deficit

after sleep deprivation, 162in delayed sleep phase syndrome, 166

Attention-deficit/hyperactivity disorderhabitual snoring and, 163, 164thypoarousal in, 161in adenotonsillectomy patients, 163–164insomnia in, 141–142narcolepsy and, 166neurologic and EEG abnormalities in, 4obstructive sleep apnea syndrome and, 215–216restless legs syndrome/periodic limb movement disorder

and, 165–166, 165fsleep disorders and

pathophysiologic links between, 161–163, 162ftreatment of, 108t, 336

stimulants in, 141–142vs. delayed sleep phase syndrome, 103tvs. idiopathic hypersomnia, 185

Audio-video monitoringin obstructive sleep apnea syndrome, 200in polysomnography, 65

Autisminsomnia in, 21sleep disorders in, 23tsleep patterns in, 5

Autonomic nervous systemheart rate variability and, 64in obstructive sleep apnea syndrome, 214

Awakeningsnight

colic and, 119–120in attention-deficit/hyperactivity disorder, 141in narcolepsy, 173incidence and prevalence of, 128–129seizures and. See Seizures, sleep-related.sensory thresholds and, 130temperament and, 129–130unusual. See Arousal disorders.

scheduledin confusional arousals, 302in sleep-onset insomnia, 108t

Barbiturates, sleep and, 329Basal ganglia, in sleep-wake cycle, 37Basic rest and activity cycle (BRAC), 62Bath, hot, sleep hygiene and, 107Bayley scores, REM storms and, 271BEARS mnemonic, 22, 43Beclomethasone, in nasal obstruction, 250Bedtime

advancing, in delayed sleep phase syndrome, 107, 146

consistent, sleep hygiene and, 106delaying. See Chronotherapy.scheduled, in sleep-onset insomnia, 108tSunday night, in delayed sleep phase syndrome, 105

Bedtime struggles, differential diagnosis of, 144Bedwetting. See Enuresis.

340 Index


Behavior(s)impact of sleep problems on, 29–31sleep-related

abnormal, 22–23normal, 27–28variables affecting, 28–29

Behavior therapyin arousal disorders, 302in insomnia, 132–134in sleep-onset association disorder, 150in sleep-onset insomnia, 108t, 332in sleep-related enuresis, 321–322

Behavioral considerationsin case-based analysis of sleep problems, 44, 46tin delayed sleep phase syndrome, 104–105in insomnia, 129–130, 134–138in sleep, 11

Behavioral disordershypersomnia in, 21in breathing-related sleep disorders, 163–164in obstructive sleep apnea syndrome, 215–216, 226insomnia in, 19REM-sleep, 288, 312

post-traumatic, 189vs. nightmares, 287

sleep disorders and, 161–167, 162fvs. idiopathic hypersomnia, 185–186

Benzodiazepine receptor site, 327Benzodiazepines

adverse effects of, 142in anxiety disorders, 335–336in confusional arousals, 333in partial arousal disorders, 288in sleep-onset insomnia, 108t, 332in sleep terrors, 334sleep and, 329

Bilevel ventilation, vs. nasal CPAP, 239Biofeedback, in arousal disorders, 302Biological clocks, 85–86, 101Biological day and night, 87f, 88–89, 89f

rate of change from, 87f, 90, 90fBiological parameters, circadian rhythm of, 91Bipolar disorder, vs. idiopathic hypersomnia, 186Birds, sleep in, 10Bladder

distention ofarousal effect of, 136somnambulism and, 285

functional capacity of, enuresis and, 317–318pressure on, enuresis from, 232, 319

Bladder-training exercises, in sleep-related enuresis, 321Blindness

insomnia and, 140non–24-hour sleep-wake disorder and, 110, 154sleep problems and, 31sleep-wake cycle timing problems and, 20–21

Blood pressureduring sleep, 77–78, 77televated, in obstructive sleep apnea syndrome, 214–215heart rate and, 65

Body movementsduring sleep, 60–63, 65, 66, 68

development of, 60–61

Body movements (Continued)evolution of, 61in dystonia, 62–63in hydranencephaly, 62in sleep-disturbed children, 62state dependency of, 61

electromyographic recording of. See Electromyogram.monitoring of, 62, 68

Body position, during sleepshifts in, 60sleep state organization and, 61–62

Body rocking, 283, 307–308, 309fBody-shuttling, 283Body temperature

circadian rhythm of, 86f, 90, 95in circadian rhythm sleep disorders, 101, 102f, 105–106regulation of, during sleep, 74–75, 75t

Bottle, weaning from, in excessive nocturnal fluids,150–151

Brain. See also Central nervous system.electrical activity of. See also Electroencephalogram.

development of, 49–56injury to. See Head injury.structural development of, 49temperature of, during sleep, 73, 74ttumors of, hypersomnia in, 195

Brainstemischemic disturbances of, narcolepsy and, 175lesions of, sleep abnormalities in, 270reticular activating system of. See Reticular activating

system.Breast-feeding, weaning from, in excessive nocturnal fluids,

150–151Breathing. See also Respiratory entries.

increased work of, in obstructive sleep apnea syndrome,213

paradoxical, quantification of, 202periodic, 79, 206t, 274, 278f

Breathing-related sleep disorders, 272–281. See alsoObstructive sleep apnea syndrome; Snoring.

behavioral disorders in, 163–164cognitive changes in, 164–165diagnosis of, 280–281diagnostic classification of, 205–206, 206t, 207tenuresis in, 231–232, 232f, 232t, 319executive function deficits in, 216, 217fhypersomnia in, 21–22hypoventilation in, 278–280. See also

Hypoventilation.insomnia in, 20obstructive, spectrum of, 197, 198fotolaryngologic management of, 249–260periodic leg movements with, hyperactivity and, 166respiratory pauses in, 272–278, 273f–278f. See also

Apnea.vs. idiopathic hypersomnia, 185

Bright-light box, in delayed sleep phase syndrome, 107Bright-light therapy

in advanced sleep phase syndrome, 110in delayed sleep phase syndrome, 107, 154in sleep-onset insomnia, 108t

Bruxism, sleep, 312–314, 313f, 314fBupropion, sleep and, 330

Index 341


Burn injury, sleep problems in, 32Buspirone, in anxiety disorders, 335

Caffeinein apnea of prematurity, 277in arousal disorders, 302in case-based analysis of sleep problems, 44, 46tsleep hygiene and, 106

Capnography, 67, 203Carbamazepine, 331

in paroxysmal hypnogenic dystonia, 282in sleep-related enuresis, 323

Carbon dioxideend-tidal, 67

in obstructive sleep apnea syndrome, 203partial pressure of

cerebral blood flow and, 74during sleep, 78, 78tin obstructive sleep apnea syndrome, 203

Cardiac outputduring sleep, 77, 77theart rate and, 64–65

Cardiorespiratory processes, in case-based analysis of sleepproblems, 44, 46t

Cardiovascular systemof fetus, 63sleep-related variations in, 77–78, 77t

Case-based analysis, of sleep disorders, 43–45, 46tCataplexy

in narcolepsy, 171, 172, 178t, 179in Niemann-Pick disease, 175post-traumatic, 189

Cats, sleep in, 10–11Central apnea, 206t, 272, 274fCentral hypoventilation, 280

congenital, breathing-related sleep disorders in, 22Central nervous system. See also Brain.

development ofcorrelates of, 269–270disorders of, sleep abnormalities in, 270–271polysomnographic correlates of, 271–272sleep and, 4

disorders ofabnormal sleep-related behaviors in, 23hypersomnia in, 22insomnia in, 21

sleep-inducing structures in, 12sleep-related variations in, 73–74, 74ttumors of, hypersomnia in, 195

Cephalic flexure, 269Cerebellar vermis, sleep spindles and, 37Cerebellum

dysfunction of, in dyslexia, 4–5hemangioblastoma of, narcolepsy and, 175in sleep-wake cycle, 35, 37

Cerebral blood flow, during sleep, 73–74, 74tCerebral hemispheres

development of, 49, 270during sleep, in dolphins, 2

Cerebral palsyobstructive sleep apnea syndrome in, craniofacial surgery

for, 240

Cerebral palsy (Continued)otolaryngologic surgery in, 258, 259sleep problems in, 288–289

Cervical flexure, 269Chiari malformation, sleep disorders in, 23tChin muscle electromyogram, 60, 68Chloral hydrate

adverse effects of, 142in insomnia, 133in neurologic disorders, 140, 151

Cholinergic disturbances, in narcolepsy, 175Chronic fatigue/fibromyalgia syndrome, vs. idiopathic

hypersomnia, 185Chronic illness

insomnia in, 142–143, 157–158sleep problems in, 32

Chronobiology of sleep, 85–97. See also Circadian rhythm.Chronophysiologic factors, in insomnia, 145–146Chronotherapy

in advanced sleep phase syndrome, 110, 145in delayed sleep phase syndrome, 107, 109

Circadian acrophase, 91Circadian nadir, 91Circadian pacemaker, 43, 86, 94Circadian rhythm, 85–97

amplitude of, 86, 86fbiological clocks and, 85–86, 101biological day and night in, 87f, 88–89, 89f

rate of change from, 87f, 90, 90fconsolidated sleep period in, 90, 90fdefinition of, 85development of, 85, 93–96

in first year of life, 94–95in neonates, 94–95in preschool children, 95in school-aged children, 95–96in utero, 93–94melatonin in, 94nonphotic zeitgebers in, 94suprachiasmatic nucleus in, 94

disorders of. See Circadian rhythm sleep disorders.entrainment of, 92forbidden sleep zone in, 88f, 91, 101–102in case-based analysis of sleep problems, 43, 46tmanifestations of, 91, 297masking of, 92melatonin and, 91–92. See also Melatonin.of body movements during sleep, 62of body temperature, 86f, 90, 95of wakefulness, 95period length (tau) of, 86, 86f–87fphase of, 86, 87f, 88

effect of indoor light on, 92–93morningness-eveningness preference and, 88

phase shifts in, 88f, 91in adolescents, 88, 96

point of singularity in, 88f, 91properties of, 85–93resetting of

intensity of light in, 93timing of light in, 93

zeitgebers in, 101

342 Index


Circadian rhythm (Continued)bright light as, 88f, 91nonphotic, 91, 94weak, 91

Circadian rhythm sleep disorders, 101–111consequences of, 101differential diagnosis of, 102, 103tin neurologic disorders, 290non–24-hour, 101, 102, 110, 154, 186phase-advanced. See Advanced sleep phase

syndrome.phase-delayed. See Delayed sleep phase syndrome.post-traumatic, 189–190prevalence of, 102temperature curves in, 101, 102fvs. idiopathic hypersomnia, 186

Circadian systems, homeostatic drive and, synchronizationof, 298

Citalopram, 330Clitoral erection, during sleep, 80Clock, internal (biological), 85–86, 101. See also Circadian

rhythm.Clomipramine, in narcolepsy, 178t, 179Clonazepam

in anxiety disorders, 336in arousal disorders, 302in neurologic disorders, 152in parasomnias, 286in partial arousal disorders, 288in sleep-related panic attack, 334

Clonidinein anxiety disorders, 335in neurologic disorders, 153

Clozapine, 331Coagulation studies, preoperative, 254Codeine, after adenotonsillectomy, 256Cognitive deficits

attentional. See Attention-deficit/hyperactivity disorder.

cerebellar-vestibular dysfunction and, 4–5in breathing-related sleep disorders, 164–165in obstructive sleep apnea syndrome, 216

model for, 216, 217fin upper airway resistance syndrome, 263–264sleep disorders and, 161–167sleep problems and, 31, 162f

Cognitive procedural tasks, memory processing of, sleepand, 4

Colic, 113–123, 137–138alternative view of, 121–123crying and, 114–115definition of, 113–114etiology of, 114fussing and, 115natural history of, 113–114sleep in infants with, 115–117, 137–138sleep problems after, 120, 138summary of, 120–121temperament and, 117–118, 121–123wakefulness and, 117

Combativeness, in partial arousal disorders, 284Confusion, in Kleine-Levin syndrome, 193

Confusional arousals, 283–284, 295t, 296treatment of, 302, 333vs. somnambulism, 285

Congenital disorders, breathing-related sleep disorders in, 22

Conjugate eye movements, development of, 59Consolidated sleep period, in circadian rhythm, 90, 90fContinuous positive airway pressure (CPAP), nasal

compliance with, 239efficacy of, 238in apnea of prematurity, 278in obstructive sleep apnea syndrome, 238–240, 240tinstitution of, 239method of action of, 238side effects of, 239–240, 240t

Cor pulmonale, in obstructive sleep apnea syndrome, 227

Corneal reflection technique, 57Cortex, in sleep-wake cycle, 35, 36f, 38f, 39fCortical laminar necrosis, sleep abnormalities in, 271Corticosteroids

in acute tonsillitis, 251in chronic adenotonsillar hypertrophy, 251in nasal obstruction, 250in obstructive sleep apnea syndrome, 242–243

Cortisolduring sleep, 75t, 76in Kleine-Levin syndrome, 194secretion of, colic and, 114

Cosleepinginsomnia and, 130–131patterns of, 27

Coughing, during sleep, 79Counseling, in arousal disorders, 302Cow’s milk allergy, insomnia in, 138–139CPAP. See Continuous positive airway pressure (CPAP).Craniofacial abnormalities

obstructive sleep apnea syndrome and, 224otolaryngologic surgery in, 258

Craniofacial surgery, in obstructive sleep apnea syndrome,240–241, 241f

Craniopharyngioma, narcolepsy and, 175Creativity, REM sleep and, 4Cricoarytenoid muscles, 275, 276Crying, colic and, 114–115Cultural practices, sleep behavior and, 27–28Cystic fibrosis, insomnia in, 20

Daytime functioningin obstructive sleep apnea syndrome, 226–227sleep problems and, model for, 162, 162f

Daytime sleepiness. See also Hypersomnia.behavioral manifestations of, 30differential diagnosis of, 18t, 21–22, 184t, 185impact of, 30in adolescents, 183in circadian rhythm disorders, 101in obstructive sleep apnea syndrome, 197–198, 226in upper airway resistance syndrome, 263

Deafness, insomnia and, 140Declarative learning tasks, memory processing of, sleep

and, 4

Index 343


Delayed sleep phase syndrome, 102, 104–107, 109age of onset in, 105attention deficit in, 166diagnostic criteria for, 104differential diagnosis of, 102, 103tevaluation of, 105family dynamics in, 104in adolescents, 105in neurologic disorders, 290insomnia in, 145–146, 154interpersonal dynamics in, 104presenting complaints in, 104prevalence of, 102psychiatric/behavioral symptoms coexisting with,

104–105sleep-wake cycle timing problems in, 20Sunday night bedtime in, 105temperature curves in, 101, 102ftreatment of, 108t, 154

advancing bedtime in, 107bright-light box in, 107bright-light therapy in, 107, 154chronotherapy in, 107, 109maintenance phase in, 109melatonin in, 107phase shifting approach in, 106sleep hygiene in, 106–107successful, 109temperature nadir and, 104–105

vs. idiopathic hypersomnia, 186vs. narcolepsy, 177

Delta brushes, on electroencephalogram, 50, 51fDepression

differential diagnosis of, 20, 22insomnia in, 143–144

combination medications for, 331maternal, childhood insomnia and, 143sleep problems and, 31

Desipramine, 330Desmopressin, in sleep-related enuresis, 322,

334–335Detrusor muscle, spontaneous contractions of, enuresis

and, 318Developmental changes

in case-based analysis of sleep problems, 44in ultradian cycles, 44

Developmental delay, pervasive, sleep problems and, 31Developmental disabilities, moderate to severe, insomnia in,

pharmacologic treatment of, 335Dexamethasone

in acute tonsillitis, 251in chronic adenotonsillar hypertrophy, 251perioperative, 255

Dextroamphetamine, in idiopathic hypersomnia, 187t

Diabetes, sleep-related enuresis in, 319Diazepam

in neurologic disorders, 152in parasomnias, 286in partial arousal disorders, 288in sleep-onset insomnia, 332sleep and, 329

Diencephalonabnormalities of, in Kleine-Levin syndrome, 194development of, 49, 270

Diphenhydraminein neurologic disorders, 151in primary insomnia, 158

Disorders of initiating or maintaining sleep (DIMS). SeeInsomnia.

Disorientation, in partial arousal disorders, 284Dissociated state, in arousal disorders, 297Diuretics, enuresis from, 320Dolphins, sleep in, 2, 11Dopamine, in narcolepsy, 175Down syndrome

obstructive sleep apnea syndrome in, craniofacial surgeryfor, 240

otolaryngologic surgery in, 258sleep disorders in, 23t

Doxapram, in apnea of prematurity, 278Doxepin, 330

in anxiety disorders, 335Dream, anxiety, 286–287, 310–311, 311tDrug(s). See also Pharmacotherapy.

heart rate and, 64hypersomnia from, 21in case-based analysis of sleep problems, 44, 46tinsomnia from, 20, 142

Dyslexia, cerebellar-vestibular dysfunction in, 4–5Dyssomnia. See Sleep disorders.Dystonia

body movements during sleep in, 62–63paroxysmal hypnogenic, 282

EEG. See Electroencephalogram.Electroacupuncture, in sleep-related enuresis, 323Electrocardiogram

description of, 68during sleep, 63–65, 68

Electrocautery, in adenotonsillectomy, 254Electroencephalogram, 49–57, 66, 67t

brain maturation and, 271delta brushes on, 50, 51fdescription of, 68discontinuous pattern on, 50, 50f, 51felectrodes in

arrays for, 66, 67t, 68placement of, 56, 58f

equipment for, 66, 67tin cerebral palsy, 289in early and middle childhood, 56–57, 57f–58in fetus, 49–51

at weeks 24 to 28, 50, 50fat weeks 28 to 32, 50, 51fat weeks 32 to 36, 51at weeks 36 to 40, 51

in neonates and infantsat term to 3 months, 51–53, 52f–54fat 3 to 12 months, 53–56

in obstructive sleep apnea syndrome, 201in parasomnias, 306–307, 307f–309fK-complexes on, 37, 55normative data for, 204, 204t

344 Index


Electroencephalogram (Continued)sleep spindles on, 54–55sleep stage scoring of, 57–58slow-waves on

during sleep, 51–52, 53f, 54during wakefulness, 53

sucking artifact on, 52, 55ftheta activity on

during early and middle childhood, 56hypersynchronous, 56, 57fin infants, 53

theta-delta pattern on, 56, 58fin parasomnias, 307, 308f, 309f

trace-alternant pattern on, 51, 52fvertex waves on, 55

Electromyogram, 60–63, 65, 66, 68anterior tibial muscle, 60, 65, 68brain maturation and, 271–272chin muscle, 60, 68equipment for, 66intercostal, 67leg muscle, 65

Electro-oculogram, 57, 58–60, 66equipment for, 66in infants, limitations of, 59

Electrophysiologic cycles, during sleep, 11–12Electrosurgery, in adenotonsillectomy, 254EMG. See Electromyogram.Emotional problems

in narcolepsy, 178t, 179in sleep-related enuresis, 318

Emotional statecircadian rhythm of, 91sleep problems and, 328

Encephalitis, viral, narcolepsy and, 175Encephalomyelitis, vs. idiopathic hypersomnia, 186Encephalopathy, static, insomnia in, 140Endocrine system

during sleep, 75–77, 75tin Kleine-Levin syndrome, 194

Endothermic animals, sleep patterns in, 2Energy conservation theory of sleep, 2–3Energy expenditure, in obstructive sleep apnea syndrome,

213Enuresis

in obstructive sleep apnea syndrome, 217, 226, 231–232,232f, 232t

respiratory disturbance index and, 231sleep-related, 317–323

diagnosis of, 320etiology of, 317–319organic factors associated with, 319–320primary vs. secondary, 317treatment of, 320–323, 334–335

Enuresis alarms, 321–322Environmental conditions, in obstructive sleep apnea

syndrome, 243Environmental manipulation, in sleep-onset association

disorder, 150Epiglottoplasty, in laryngomalacia, 241Epilepsy. See also Seizures.

benign rolandic, 306f

Epilepsy (Continued)sleep-related, 281, 301

vs. parasomnias, 306, 306fvs. sleep terrors, 286

vs. idiopathic hypersomnia, 185Epithalamus, development of, 49Esophageal manometry

in obstructive sleep apnea syndrome, 202normative data for, 204, 204t

Esophagitis, in gastroesophageal reflux, pathogenesis of, 81Eveningness preference, 88Evolutionary theory of sleep, 2Executive function deficits. See also Cognitive deficits.

in breathing-related sleep disorders, 216, 217fsleep problems and, 162, 162f

Expiratory apnea, 273, 276fExpiratory braking, 276Eye closure, persistent, during neonatal sleep, 52Eye movements

brain maturation and, 271conjugate, development of, 59density of, sleep deprivation and, 59–60during REM sleep

analysis of, 59, 60in dyslexia, 5

microtremor, 60recording of. See Electro-oculogram.saccadic, during REM sleep, 59

Face, structure of, obstructive sleep apnea syndrome and,224

Face perceptions, in newborns, 3–4Faded-response program, in sleep-onset association

disorder, 150Failure to thrive, in obstructive sleep apnea syndrome, 213,

226Family dynamics, in delayed sleep phase syndrome, 104Family values, sleep behavior and, 27–28Fatigue, chronic, vs. idiopathic hypersomnia, 185Fetus

cardiovascular system of, 63circadian rhythm development in, 93–94electroencephalogram in, 49–51

at weeks 24 to 28, 50, 50fat weeks 28 to 32, 50, 51fat weeks 32 to 36, 51at weeks 36 to 40, 51

Fever, somnambulism and, 285Fibromyalgia

in chronic fatigue syndrome, vs. idiopathic hypersomnia,185

sleep problems in, 32First-night effect, in polysomnography, 205Fluoroscopy, in obstructive sleep apnea syndrome, 200Fluoxetine, 330

in narcolepsy, 178t, 179in sleep-related panic attack, 334

Fluticasone, in nasal obstruction, 250Fluvoxamine, 330Follicle stimulating hormone (FSH), during sleep, 75t,

76–77Forbidden sleep zone, 88f, 91

Index 345


Forebrain, in sleep-wake cycle, 37Formula, hydrolyzed milk-protein, in cow’s milk allergy,

139Fussing

colic and, 115fatigue-driven, 122

GABA allosteric modulators, 327, 329–330Gabapentin, 331Gas exchange, in obstructive sleep apnea syndrome,

202–203, 205Gastric acid secretion, during sleep, 80–81Gastroesophageal reflux, 81, 250Gastrointestinal system, sleep-related variations in,

80–81Gastrolaryngopharyngeal reflux, 250Genetic factors

in arousal disorders, 298in sleep-related enuresis, 318–319in somnambulism, 298

Genitourinary system, sleep-related variations in, 80Glioma, ventricular, narcolepsy and, 175Growth hormone

during sleep, 75–76, 75tin Kleine-Levin syndrome, 194

Growth retardationin breathing-related sleep disorders, 22in obstructive sleep apnea syndrome, 213

Guided imagery, in sleep-related enuresis, 321Guinea pigs, sleep in, 10–11

H-reflex, in cerebral palsy, 288–289Habit training. See Behavior therapy.Hallucinations

hypnagogicbefore sleep starts, 308in narcolepsy, 171, 173

hypnopompic, in narcolepsy, 173in Kleine-Levin syndrome, 193

Head-banging, 283, 307–308, 309fHead injury

cognitive improvement after, REM sleep and, 6–7hypersomnia after, 189–191narcolepsy and, 175

Head-rolling, 283Headache, sleep problems and, 32Heart rate

blood pressure and, 65cardiac output and, 64–65during sleep, 63–65, 77, 77tmeasurement of, 68. See also Electrocardiogram.respiratory modulation of, 64state-dependency of, 64variability of

autonomic nervous system and, 64during REM sleep, 63in obstructive sleep apnea syndrome, 214maturation of, 64

Heart rhythm, measurement of, 68Hemangioblastoma, cerebellar, narcolepsy and, 175Hematoma, subdural, vs. idiopathic hypersomnia, 185Hemispheric sleep, in dolphins, 2

Hemoglobin-oxygen saturation, 67Hemoglobinopathies, sleep-related enuresis in, 319Hemorrhage, after adenotonsillectomy, 256High upper airway resistance, 278–280, 279fHirschsprung’s disease, sleep disorders in, 23tHLA (histocompatibility leukocyte antigen)

in narcolepsy, 174in sleepwalking, 298serologic testing for, 176

Homeostatic drivecircadian systems and, synchronization of, 298in arousal disorders, 297sleep deprivation and, 298

Homeostatic processes, in case-based analysis of sleep prob-lems, 43, 46t

Hormone secretion, during sleep, 75–77, 75tHydranencephaly, body movements during sleep in, 62Hydrocodone, after adenotonsillectomy, 256Hydroxyzine

in neurologic disorders, 152in primary insomnia, 158

Hyperactivityattention deficit and. See Attention-deficit/hyperactivity

disorder.breathing-related sleep disorders and, 163–164habitual snoring and, 163, 164thypersomnia as, 183obstructive sleep apnea syndrome and, 166periodic leg movements and, 166

Hypercapnia, in obstructive sleep apnea syndrome, 218Hypercapnic ventilatory response, during sleep, 78t, 79–80Hyperkinetic children, sleep patterns in, 5Hyperphagia, in Kleine-Levin syndrome, 193Hypersexuality, in Kleine-Levin syndrome, 193Hypersomnia. See also Daytime sleepiness.

confusional arousals in, 284differential diagnosis of, 18t, 21–22idiopathic, 183–187

clinical presentation in, 184–185differential diagnosis of, 183, 184t, 185–186evaluation of, 183–184, 184ttreatment of, 186, 187ttypes of, 183vs. narcolepsy, 178

menstruation-associated, 193–194post-traumatic, 189–191recurrent, 193–195. See also Kleine-Levin syndrome.

Hypertensionpulmonary, in obstructive sleep apnea syndrome, 214systemic, in obstructive sleep apnea syndrome, 214–215

Hypnagogic jerks, 308Hypnagogic sleep paralysis, 287, 311Hypnic myoclonia, 308Hypnogenic dystonia, paroxysmal, 282Hypnopompic sleep paralysis, 287, 311Hypnotics. See Sedative/hypnotic medications.Hypnotoxins, 7Hypocretin-1, cerebrospinal fluid

assay for, 175, 176–177in human narcolepsy, 174, 174f

Hypocretin receptor-2 gene mutation, in animal narcolepsy,173

346 Index


Hypomania, from bright-light therapy, 107Hypopharyngeal airway expansion–related procedures, 259Hypopnea, 206tHypothalamic-pituitary-adrenal (HPA) axis, in Kleine-Levin

syndrome, 194Hypothalamus

development of, 49in sleep-wake cycle, 37lesions of, sleep abnormalities in, 271thermoregulatory control by, during sleep, 75

Hypoventilation, 206tcentral, 280

congenital, breathing-related sleep disorders in, 22during sleep, 78in obstructive sleep apnea syndrome, 218obstructive, 197, 278–280, 279f

Hypoxia/hypoxemia, in obstructive sleep apnea syndrome,163, 214–216

measurement of, 202–203neurocognitive and behavioral consequences of,

215–216, 215t, 217fHypoxic ventilatory response, during sleep, 78t, 79

Ibuprofen, after adenotonsillectomy, 256Imagery

guided, in sleep-related enuresis, 321mental, in arousal disorders, 302

Imipramine, 330in neurologic disorders, 151–152in sleep-related enuresis, 323, 335

Incontinence, urinary. See Enuresis.Indeterminate sleep, 52Indoleacetic acid, in narcolepsy, 175Infants. See also Neonates.

electroencephalogram in, 51–53, 52f–54fat term to 3 months, 51–53, 52f–54fat 3 to 12 months, 53–56

electro-oculogram in, limitations of, 59polysomnography in, indications for, 66premature

apnea in, 276–278periodic breathing in, 79

Infectionurinary tract, enuresis and, 319vs. idiopathic hypersomnia, 186

Insomnia, 127–158behavioral considerations in, 129–130, 134–138blindness and, 140chronophysiologic factors causing, 145–146cosleeping and, 130–131deafness and, 140differential diagnosis of, 17, 18t, 19–21, 327–328drug-induced, 20, 142evaluation of, 131–132, 148–149general considerations in, 127–128impact of, 148in advanced sleep phase syndrome, 145, 153–154in affective disorders, 143–144in anxiety disorders, 335–336in attention-deficit/hyperactivity disorder, 141–142in behavioral and psychophysiologic disorders, 19in breathing-related sleep disorders, 20

Insomnia (Continued)in central nervous system disorders, 21in chronic illness, 142–143, 157–158in colic, 137–138in cow’s milk allergy, 138–139in delayed sleep phase syndrome, 145–146, 154in depression, combination medications for, 331in excessive nocturnal fluids, 136–137, 150–151in irregular sleep schedule, 146in limit setting disorders, 19, 144–145in mood disorders, 143–144in movement disorders, 20in neurologic disorders, 139–141, 151–153in night-time fears, 144in non–24-hour sleep-wake cycle schedule, 154in otitis media, 139in parasomnias, 21in psychiatric disorders, 19–20in regular but inappropriate sleep-wake cycle, 146in sleep-onset association disorder, 19, 134–136,

149–151in sleep-wake cycle timing disorders, 20–21incidence and prevalence of, 128–129, 148management of

general concepts in, 132–134, 133tin specific conditions, 149–158

medical disorders causing, 138–143pediatric

definition of, 147–148parental concerns in, 127–128

pharmacologic treatment ofafter pineal tumor extraction, 335general recommendations for, 155–157in special needs children, 157–158intravenous medications in, 335long-term medications in, 336secondary to anxiety disorders, 335–336with moderate to severe developmental disabilities,

335primary

childhood onset of, 147psychiatric disorders and, 31treatment of, 108t, 155–157, 158vs. delayed sleep phase syndrome, 103t

psychogenic, post-traumatic, 189, 190psychological factors causing, 143–145referral indications in, 149resistant, severe, 336screening for, 148short sleeper and, 147sleep-onset

differential diagnosis of, 331–332treatment of, 108t, 332

stress and, 130, 143temperament and, 129–130

Insulin, fasting, in obstructive sleep apnea syndrome, 226Insulin-like growth factor-1, in obstructive sleep apnea syn-

drome, 213Intercostal electromyogram, 67Intercostal muscle, hypotonia of, during REM sleep, 79International 10-20 System of EEG electrode placement, 56,

58f

Index 347


Interpersonal dynamics, in delayed sleep phase syndrome,104

Intestinal smooth muscle contraction, colic and, 114Intracranial pressure

during sleep, 74, 74tincreased, vs. idiopathic hypersomnia, 185

Iron loss, in movement disorders, 20Irritability, in Kleine-Levin syndrome, 193

Jerks, hypnagogic, 308

K-complexes, 37, 55Kleine-Levin syndrome

clinical characteristics of, 193–195evaluation of, 195recurrent hypersomnia in, 22, 193–195treatment of, 195vs. idiopathic hypersomnia, 186vs. narcolepsy, 177–178

Landau-Kleffner syndrome, 281Laryngomalacia, epiglottoplasty in, 241Larynx, respiratory functions of, 275–276Learning deficits. See Cognitive deficits.Learning theory of sleep, 3–7Leg muscle electromyogram, 65Legs, restless. See Restless legs syndrome.Light

brightas zeitgeber, 88f, 91entrainment of circadian rhythms by, 92in advanced sleep phase syndrome, 110in delayed sleep phase syndrome, 107, 154in sleep-onset insomnia, 108t

for resetting circadian rhythmintensity of, 93timing of, 93

indooras weak zeitgeber, 91circadian rhythm phase shifts influenced by, 92–93

night, in sleep-onset association disorder, 150Light receptors, retinal, 90, 90fLimb movements, periodic. See Periodic limb movement

disorder.Limit-setting sleep disorders, insomnia in, 19, 144–145Lomipramine, 330Loprazolam, in sleep-onset insomnia, 332Lorazepam

in neurologic disorders, 153in parasomnias, 286in partial arousal disorders, 288in primary insomnia, 158

Lormetazepam, in sleep-onset insomnia, 332Luteinizing hormone (LH), during sleep, 75t, 76Lymphoma, temporal lobe B-cell, narcolepsy and, 175

Macroglossia, surgery for, 241Magnetic resonance imaging, in obstructive sleep apnea

syndrome, 200Maintenance of wakefulness test (MWT), in idiopathic

hypersomnia, 183–184Malnutrition, sinus pauses and, 65

Mammals, sleep in, 10–11Mandibular distraction, 259

for micrognathia, 241, 241fMaxillary expansion, 259

in obstructive sleep apnea syndrome, 242Medical disorders

in case-based analysis of sleep problems, 45, 46tinsomnia in, 138–143sleep problems in, 32vs. idiopathic hypersomnia, 186

Medullary lesions, sleep abnormalities in, 270Melatonin, 91–92

administration of, phase-response curve to, 92, 93colic and, 114in advanced sleep phase syndrome, 153–154in cerebral palsy, 289in delayed sleep phase syndrome, 107, 154in insomnia with moderate to severe developmental

disabilities, 335in neurologic disorders, 153in non–24-hour sleep-wake disorder, 101, 102, 110, 154in sleep-onset insomnia, 108t, 332secretion of

at dusk, 90in development of circadian rhythm, 94, 270maternal, 93period of, duration of biological night and, 87f, 89plot of, 90, 90fregulation of, 91

Memory processing, during REM sleep, 3–7Meningitis, vs. idiopathic hypersomnia, 186Menstruation, hypersomnia associated with, 193–194Mental imagery, in arousal disorders, 302Mental retardation, sleep problems and, 31Mesencephalon, lesions of, sleep abnormalities in, 270Metabolic disorders, vs. idiopathic hypersomnia, 186Metabolic syndrome, in obstructive sleep apnea syndrome,

226Methylenedioxymethamphetamine (MDMA, “Ecstasy”),

20Methylphenidate

in idiopathic hypersomnia, 187tin Kleine-Levin syndrome, 195in narcolepsy, 178–179, 178tin post-traumatic hypersomnia, 191

Methylxanthines, in apnea of prematurity, 277Micrognathia

mandibular distraction for, 241, 241fobstructive sleep apnea syndrome in, 224

Microsleepsdevelopment of, 56during sleep deprivation, 9–10

Microtremor eye movements, 60Midazolam, sleep and, 329Middle ear disease, insomnia in, 139Middle fossa tumors, non–24-hour sleep-wake disorder

and, 154Midfacial advancement, 241Midfacial hypoplasia, obstructive sleep apnea syndrome

and, 224Migraine, sleep problems and, 32Milk, warm, sleep hygiene and, 106

348 Index


Minute ventilation, during sleep, 78, 78tMirtazapine, 331Mitosis, sleep and, 2Modafinil (Provigil)

in idiopathic hypersomnia, 187tin Kleine-Levin syndrome, 195in narcolepsy, 178t, 179in post-traumatic hypersomnia, 191

Mometasone, in nasal obstruction, 250Monoamine disturbances, in narcolepsy, 175Mood

circadian rhythm of, 91impact of sleep problems on, 29–31

Mood disorders. See also Anxiety disorders; Depression.insomnia in, 143–144sleep-onset insomnia in, 108tvs. delayed sleep phase syndrome, 103tvs. idiopathic hypersomnia, 185–186

Morningness preference, 88Motor procedural tasks, memory processing of, sleep and, 4Mouthguard, in sleep bruxism, 314Movement activity. See also Body movements; Eye

movements.during sleep, 60–63, 65, 66, 68

Movement disorders. See also Restless legs syndrome.hypersomnia in, 22in sleep bruxism, 312–314, 313f, 314finsomnia in, 20iron loss in, 20periodic limb. See Periodic limb movement disorder.rhythmic, 282–283, 307–308, 309fsleep-related, 19t, 23

MSLT. See Multiple sleep latency test (MSLT).Mucociliary clearance, during sleep, 79Mucopolysaccharide storage disorders, sleep disorders in,

23tMultiple sclerosis, narcolepsy and, 175Multiple sleep latency test (MSLT)

in idiopathic hypersomnia, 183–184in Kleine-Levin syndrome, 195in narcolepsy, 176, 177tin post-traumatic sleep disorders, 190serial, in narcolepsy, 176, 177f

Muscle weakness, vs. idiopathic hypersomnia, 186Myelomeningocele, with apnea, respiratory patterns in,

289–290Myoclonia, hypnic, 308

Nadircircadian, 91temperature, in delayed sleep phase syndrome, 104–105

Nap studies, in obstructive sleep apnea syndrome, 204–205Napping

benefits of, 1involuntary, in idiopathic hypersomnia, 184planned

in narcolepsy, 179in post-traumatic hypersomnia, 191

scheduled, normal trends in, 27Narcolepsy, 171–179

attention-deficit/hyperactivity disorder and, 166cataplexy in, 171, 172, 178t, 179

Narcolepsy (Continued)clinical features of, 171–173confusional arousals in, 284daytime sleepiness in, 171, 172, 178–179, 178tdiagnosis of, 176–177, 177f, 177tdifferential diagnosis of, 19–20, 22, 177–178epidemiology of, 171, 172fhallucinations in, 171, 173histocompatibility antigens in, 174hypocretins and, 173, 174, 174fin animals, 173–174in preschool-aged children, 171–172in school-aged children, 172management of, 178–179, 178tmonoamine disturbances in, 175pathophysiology of, 173–175post-traumatic, 189secondary, 175sleep drunkenness in, 284sleep paralysis in, 171, 173, 287, 312two-hit hypothesis of, 175vs. idiopathic hypersomnia, 185

Narcolepsy Network, 179Nasal cannula pressure recordings, in obstructive sleep

apnea syndrome, 202, 203fNasal continuous positive airway pressure. See Continuous

positive airway pressure (CPAP), nasal.Nasal obstruction, treatment of, 250Nasal/oral airflow, measurement of, 65, 67

in obstructive sleep apnea syndrome, 202, 203fNasal pressure transduction, 67Nasopharyngeal stenosis, after adenotonsillectomy, 257Nefazodone, 331Neonates. See also Infants.

active sleep in, 52, 54f, 60–61circadian rhythm development in, 94–95electroencephalogram in, 51–53, 52f–54fface-processing abilities of, 3–4polysomnography in, indications for, 66quiet sleep in, 51–52, 52f, 60

Neural tube, 269Neurodevelopmental disorders, sleep problems and, 31Neurologic correlates, of sleep disorders, 270–271Neurologic disorders

apnea in, 276, 289–290circadian rhythm disorders in, 290delayed sleep phase syndrome in, 290insomnia in, 139–141, 151–153sleep in, 269–290sleep-related enuresis in, 319–320sleep-wake cycle timing problems in, 20–21vs. idiopathic hypersomnia, 185

Neurologic processes, in case-based analysis of sleepproblems, 44, 46t

Neuromuscular disorders, adenotonsillectomy and, 258Neuromuscular tone, obstructive sleep apnea syndrome

and, 225Neuronal activity, maturation of, 269–270Neuronal susceptibility, to hypoxia, 217–218Neuropsychological manifestations, in narcolepsy, 173Neurosomnology, 271–272Neurotransmitter systems, sleep pharmacology and, 327

Index 349


Niemann-Pick disease, cataplexy in, 175Night awakenings. See Awakenings, night.Night light, in sleep-onset association disorder, 150Night terrors. See Confusional arousals; Sleep terrors.Night-time fears, insomnia in, 144Nightmares, 286–287, 310–311, 311tNocturnal fluids, excessive, insomnia in, 136–137, 150–151Non–24-hour sleep-wake cycle, 101, 102, 110, 154, 186Noradrenergic antidepressants, sleep and, 331Noradrenergic disturbances, in narcolepsy, 175Nortriptyline, 330NREM sleep. See Slow-wave sleep.

Obesityadenotonsillectomy and, 258in obstructive sleep apnea syndrome, 225, 226, 249–250

Obstructive apnea, 272, 273f, 276Obstructive hypoventilation, 197, 278–280, 279fObstructive sleep apnea syndrome, 274–276

adenotonsillar hypertrophy and, 211ancillary diagnostic studies in, 199–200, 200tarousal from sleep in, 201attention-deficit/hyperactivity disorder and, 215–216behavior disorders in, 163–164, 215–216characteristics of, 197, 211, 212f, 274–275clinical features of, 225–227, 225tcognitive changes in, 164–165, 216

model for, 162, 162f, 216, 217fconsequences of, 211–218daytime functioning in, 226–227diagnosis of, 141, 197–206, 206t, 207t, 252

delay in, 272domiciliary and nap studies in, 204–205electroencephalogram in, 201enuresis in, 217, 226, 231–232, 232f, 232tepilepsy and, 281failure to thrive in, 213, 226gas exchange in, 202–203, 205history in, 197–199, 199thome audio-video monitoring in, 200hyperactivity and, 166hypersomnia in, 21–22hypoventilation in, 218hypoxemia in, 163, 214–216

measurement of, 202–203neurocognitive and behavioral consequences of,

215–216, 215t, 217fneuronal susceptibility to, 217–218

imaging studies in, 200, 200tincreased work of breathing in, 213mortality in, 227nasal obstruction in, 250obesity in, 225, 226, 249–250pathophysiology of, 223–225, 224f, 249–250physical examination in, 199, 199tpolysomnography in, 252

during daytime nap, 205night-to-night variability in, 205overnight, 200–203, 203f

respiratory effort in, 201–202restless sleep in, 226risk factors for, 253, 253t

Obstructive sleep apnea syndrome (Continued)school performance in, 165, 226–227severity classification of, 207tsleep fragmentation in, 217–218snoring in, 225–226somnambulism in, 285treatment of, 108t, 235–243

adenotonsillectomy in, 235–238, 237t, 257–258continuous positive airway pressure in, 238–240, 240tcraniofacial and other surgeries in, 240–241, 241f,

258–260emergency, 235environmental conditions in, 243indications for, 235oral orthodontic appliances in, 242oxygen supplementation in, 242, 242fpharmacologic, 242–243tracheostomy in, 243, 259–260uvulopalatopharyngoplasty in, 240weight loss in, 241–242

upper airway abnormalities associated with, 275vs. delayed sleep phase syndrome, 103tvs. narcolepsy, 177vs. primary snoring, 211, 225–226

Olanzapine, 331Operant conditioning, in sleep-related enuresis, 321–322Oppositional defiant disorder, in adenotonsillectomy

patients, 164Oral appliances, in obstructive sleep apnea syndrome, 242Orexin-2 gene mutation, in animal narcolepsy, 173Oronasal airflow, measurement of, 65, 67

in obstructive sleep apnea syndrome, 202, 203fOronasal mask, in obstructive sleep apnea syndrome, 202Orthodontic appliances, in obstructive sleep apnea

syndrome, 242Otitis media, insomnia in, 139Otolaryngologic surgery. See also specific procedures, e.g.,

Adenotonsillectomy.in breathing-related sleep disorders, 249–260

Over-the-counter medications, insomnia from, 142Overeating, in Kleine-Levin syndrome, 193Oxybutynin chloride, in sleep-related enuresis, 322–323,

335Oxygen

partial pressure of, during sleep, 78, 78tsupplementation of, in obstructive sleep apnea syndrome,

242, 242fOxygen saturation

during sleep, 74in obstructive sleep apnea syndrome, 203pulse oximetry monitoring of, 65, 67

Pacemaker, circadian, 86, 94Pain conditions, sleep problems in, 32Panic attack, sleep-related, 334Paradoxical breathing, quantification of, 202Paradoxical sleep, phylogenetic development of, 10Parasomnias, 18t–19t, 282–288, 305–314

abnormal sleep-related behaviors in, 23arousal-related. See Arousal disorders.associated with REM sleep, 286–288, 310–314clinical and laboratory evaluation of, 282

350 Index


Parasomnias (Continued)enuresis in, 319etiology of, 305evaluation of, 305–307, 306f–309finsomnia in, 21management of, 286nonepileptic stereotypic, 282–283, 307–308, 309fpolysomnography in, 306–307, 307f–309fprevalence of, 295tsubcategories of, 293, 305

Parental concerns, in pediatric insomnia, 127–128Parenting practices, colic and, 121–123Parents, sleep problems in, impact of, 30Paroxetine, 330

in anxiety disorders, 335Paroxysmal disorders, sleep-related, 281–282Paroxysmal hypnogenic dystonia, 282Pemoline, in idiopathic hypersomnia, 187tPenile tumescence, during sleep, 80Perceptual learning, sleep and, 3–4Performance

circadian rhythm of, 91impact of sleep problems on, 29–31impairment of, after sleep deprivation, 8–9school, in obstructive sleep apnea syndrome, 165,

226–227Period length (tau), of circadian rhythm, 86, 86f–87fPeriodic breathing, 79, 206t, 274, 278fPeriodic limb movement disorder

attention-deficit/hyperactivity disorder and, 165–166,165f

hyperactivity and, 166in narcolepsy, 173, 178tinsomnia in, 20, 108tvs. delayed sleep phase syndrome, 103tvs. idiopathic hypersomnia, 185

Pharmacotherapy, 327–336of confusional arousals, 333of enuresis, 334–335of insomnia

after pineal tumor extraction, 335general recommendations for, 155–157in primary insomnia, 158in resistant, severe insomnia, 336in special needs children, 157–158intravenous medications in, 335long-term medications in, 336secondary to anxiety disorders, 335–336with moderate to severe developmental disabilities, 335

of obstructive sleep apnea syndrome, 242–243of sleep disorders, 327–336

agents used in, 329–331caution in, 328–329combination therapy in, 331critical aspects of, 327–331in attention-deficit/hyperactivity disorder, 336

of sleep-onset insomnia, 108t, 332of sleep phobia, 332–333of sleep-related anxiety, 332of sleep-related panic attack, 334of sleep terrors, 333–334of somnambulism, 334

Pharyngeal mass, nasal obstruction from, 250Pharynx

compliance of, factors affecting, 225obstructive sleep apnea syndrome and, 223–225, 224fpatency of, 223respiratory functions of, 275–276size of, factors affecting, 223–225Starling resistor model of, 223, 224f

Phenergan, in insomnia, 133Phenobarbital

adverse effects of, 142heart rate and, 64

Phylogenetic considerations, in sleep, 10–11Physical therapy, in neurologic disorders, 153Physiologic considerations, in sleep, 11–12Physiologic variations, during sleep, 73–79Pierre Robin syndrome

otolaryngologic surgery in, 258sleep disorders in, 23t

Piezocrystal belts, for evaluation of respiratory effort, 66–67

Piezoelectric strain gauge transducers, for evaluation of eyemovements, 60

Pineal body, 49, 270Pineal gland, melatonin secretion by, 92Pituitary adenoma, narcolepsy and, 175Plethysmography, for evaluation of respiratory effort,

66–67Pneumography, in breathing-related sleep disorders,

280–281Poikilothermic animals, sleep patterns in, 2Point of singularity, in circadian rhythm, 88f, 91Polysomnography, 49–70. See also specific parameters, e.g.,

Electroencephalogram.audio-video monitoring in, 65brain maturation and, 271–272description of, 68–69duration of study in, 68electrocardiogram in, 63–65, 68electroencephalogram in, 49–57, 66, 67telectromyogram in, 60–63, 65, 66, 68electro-oculogram in, 57, 58–60, 66equipment for, 66, 67tfuture usage of, 66heart rate/rhythm monitoring in, 68in adenotonsillectomy

follow-up, 260postoperative, 256preoperative, 252–253

in arousal disorders, 298–299, 300–301in breathing-related sleep disorders, 280in idiopathic hypersomnia, 183in Kleine-Levin syndrome, 195in narcolepsy, 176in neonates and infants, indications for, 66in obstructive sleep apnea syndrome, 252

during daytime nap, 205night-to-night variability in, 205overnight, 200–203, 203f

in parasomnias, 282, 306–307, 307f–309fin post-traumatic sleep disorders, 190in primary snoring, 266

Index 351


Polysomnography (Continued)in REM sleep behavior disorder, 288in rhythmic movement disorders, 307–308, 309fin sleep bruxism, 313, 313f, 314fin sleep-related enuresis, 320movement activity in, 60–63, 65, 66, 68normative data for, 204, 204tpatient preparation for, 69respiratory activity in, 65, 66–68safety of, 69sleep laboratory environment in, 69–70techniques in, 66–68

Polyuria, sleep-related enuresis in, 319Pons, lesions of, sleep abnormalities in, 270Pontine flexure, 269Porpoises, sleep in, 11Positive airway pressure

bilevel, vs. nasal CPAP, 239continuous. See Continuous positive airway pressure

(CPAP).Posterior cricoarytenoid muscles, 275, 276Prader-Willi syndrome

sleep disorders in, 23tvs. idiopathic hypersomnia, 186

Prednisone, in obstructive sleep apnea syndrome, 242–243

Prefrontal cortex, dysfunction of, sleep problems and, 162,162f

Premature infantsapnea in, 276–278periodic breathing in, 79

Problem solving, REM sleep and, 4Prolactin, during sleep, 75t, 76Protein synthesis, during REM sleep, 2Protriptyline, 330

in narcolepsy, 178t, 179Psychiatric disorders

hypersomnia in, 21in case-based analysis of sleep problems, 44, 46tinsomnia in, 19–20sleep problems and, 31, 328vs. idiopathic hypersomnia, 186

Psychiatric symptoms, delayed sleep phase syndrome and,104–105

Psychogenic insomnia, post-traumatic, 189, 190Psychological factors

in insomnia, 143–145in sleep-related enuresis, 318insomnia from, 143–145

Psychophysiologic disordersabnormal sleep-related, 22hypersomnia in, 21insomnia in, 19

Psychotherapyin arousal disorders, 302in narcolepsy, 179in sleep-onset insomnia, 108t

Pulmonary artery pressure, in obstructive sleep apneasyndrome, 214

Pulmonary edema, after adenotonsillectomy, 256Pulmonary function testing, in obstructive sleep apnea

syndrome, 200

Pulmonary hypertension, in obstructive sleep apneasyndrome, 214

Pulmonary secretions, mucociliary clearance of, duringsleep, 79

Pulse oximetry monitoring, 65, 67in obstructive sleep apnea syndrome, 202–203, 205

Q-T interval, state-dependency of, 64Quetiapine, 331

in insomnia, 336Quiet sleep

body movements during, in neonates, 60electroencephalogram during, in neonates, 51–52, 52fheart rate during, 63respiratory rate during, 63

Radiography, in obstructive sleep apnea syndrome, 200Rats, sleep in, 10–11Rebreathing, 206tReferral to sleep specialist, indications for, 149Relaxation techniques

in arousal disorders, 302in sleep-onset insomnia, 108t

REM latency, definition of, 68REM sleep

as state of being, 296–297dream anxiety attacks in, 287during early and middle childhood, 56electroencephalogram during

in infants, 56in neonates, 52

electromyogram during, 60, 61eye movements during, 59, 60heart rate variability during, 63in learning and memory, 3–7in restoration of brain tissue, 2mechanisms generating, 39fmemory processing during, 3–7parasomnias associated with, 286–288, 310–314phylogenetic development of, 10respiratory disturbance index during (REM RDI), 69

REM-sleep behavioral disorder, 288, 312post-traumatic, 189vs. nightmares, 287

REM storms, Bayley scores and, 59, 271Renal function, during sleep, 80Reptiles, sleep in, 10Respiratory activity, during sleep, 65, 66–69Respiratory compromise, after adenotonsillectomy,

237–238, 237t, 255Respiratory disturbance index

definition of, 69during REM sleep, 69enuresis and, 231

Respiratory effortin obstructive sleep apnea syndrome, 201–202normative data for, 204, 204trecording of, 65, 66–67

Respiratory inductive plethysmography (RIP), in obstruc-tive sleep apnea syndrome, 202

Respiratory muscles, hypotonia of, during REM sleep, 79Respiratory pattern scoring, 206, 206t

352 Index


Respiratory pausesdiagnosis of, 280–281during sleep, 272–278, 273f–278f. See also Apnea.

Respiratory rateduring quiet sleep, 63, 64during sleep, 78–79, 78t

Respiratory system, sleep-related variations in, 78–80, 78tRestless legs syndrome

attention-deficit/hyperactivity disorder and, 165–166,165f

insomnia in, 20, 108tvs. delayed sleep phase syndrome, 103tvs. idiopathic hypersomnia, 185

Restless sleepin attention-deficit/hyperactivity disorder, 141in obstructive sleep apnea syndrome, 226

Restoration theory of sleep, 1–2Reticular activating system, 12

in sleep-wake cycle, 35, 36f, 37, 40Retinohypothalamic tract, 90, 90fRett syndrome, insomnia in, 157Rheumatoid arthritis, juvenile, sleep problems in, 32Rhinitis, allergic, nasal obstruction from, 250Rhythmic movement disorders, 282–283, 307–308,

309fRisperidone, 331

Saccadic eye movements, during REM sleep, 59Saline irrigation, in nasal obstruction, 250Salivary flow, during sleep, 81Sarcoidosis, narcolepsy and, 175Schizophrenia, versus narcolepsy, 19–20School performance, in obstructive sleep apnea syndrome,

165, 226–227Sedative/hypnotic medications. See also specific classes and

agents , e.g., Benzodiazepines.contraindications to, 156–157in children, adverse effects of, 142in insomnia, 133, 329–331

general recommendations for, 155–157in special needs children, 157–158

in neurologic disorders, 140–141in primary insomnia, 158indications for, 156pediatric, ideal, 149

Seizuresabnormal sleep-related behaviors in, 23enuresis and, 320sleep-related, 281, 301

vs. parasomnias, 306, 306fvs. sleep terrors, 286

vs. idiopathic hypersomnia, 185Sensory thresholds, night awakenings and, 130Septoplasty, 259Serotonergic antidepressants, specific, sleep and, 331Serotonin, colic and, 114Serotonin and norepinephrine reuptake inhibitors, sleep

and, 331Serotonin antagonists/reuptake inhibitors, sleep and,

331Serotonin 5-hydroxytryptophan antagonists, 327

in sleep-onset insomnia, 108t

Serotonin reuptake inhibitors, selectivein anxiety disorders, 335in sleep-onset insomnia, 108tsleep and, 330

Sertraline, 330in narcolepsy, 178t, 179

Settling disordersdifferential diagnosis of, 331–332treatment of, 108t, 332, 333vs. delayed sleep phase syndrome, 103t

Sexually acting-out behavior, in Kleine-Levin syndrome,193

Shivering, during sleep, 74–75, 75tShort sleeper, 147Sickle cell disease

preoperative screening for, 254sleep problems in, 32sleep-related enuresis in, 319

Sinonasal mass, nasal obstruction from, 250Sinus pauses, malnutrition and, 65Sinus surgery, 259Sinusitis, nasal obstruction from, 250Skeletal muscle activity, during sleep, 80Sleep

anatomy of, 35–40behavioral considerations in, 11chronobiology of, 85–97. See also Circadian rhythm.colic and, 113–123disturbances of. See Sleep disorders; Sleep problems.function of, 1–7in neurologic disorders, 269–290insufficient nocturnal

vs. idiopathic hypersomnia, 185vs. narcolepsy, 177

NREM. See Slow-wave sleep.paradoxical, phylogenetic development of, 10pharmacology of, 327–336phylogenetic considerations in, 10–11physiologic considerations in, 11–12physiologic parameters monitored during. See

Polysomnography.physiologic variations during, 73–79REM. See REM sleep.slow-wave. See Slow-wave sleep.temperament and, 118–119, 130theories of, 1–2toxins produced during, 7urinary incontinence during. See Enuresis.

Sleep apnea, obstructive. See Obstructive sleep apnea syndrome.

Sleep architecture, analysis of, 68Sleep bruxism, 312–314, 313f, 314fSleep cycle

development of, 56heart rate and, 63physiology of, 11–12

Sleep deprivation, 8–10arousal disorders and, 299, 302attention deficit after, 162eye movement density and, 59–60from circadian rhythm disorders, 101homeostatic drive and, 298

Index 353


Sleep deprivation (Continued)microsleeps during, 9–10partial, 9–10somnambulism and, 285total, 8

Sleep diaryin idiopathic hypersomnia, 183in insomnia, 132in pharmacologic therapy, 328

Sleep disorders. See also Sleep problems.arousal-related. See Arousal disorders; Parasomnias.attention-deficit/hyperactivity disorder and, 161–167,

162f. See also Attention-deficit/hyperactivity disorder.breathing-related. See Breathing-related sleep disorders.case-based analysis of, 43–45, 46tcircadian rhythm. See Circadian rhythm sleep disorders.cognitive and behavioral changes in, 161–167complex syndromes associated with, 23, 23tdiagnosis of, delay in, 272differential diagnosis of, 17–23, 18t–19tepidemiology of, 27–32evaluation of, 17hypersomnolent. See Hypersomnia.narcoleptic. See Narcolepsy.neurologic correlates of, 270–271of initiating or maintaining sleep. See Insomnia.pharmacologic treatment of, 327–336. See also

Pharmacotherapy.agents used in, 329–331caution in, 328–329combination therapy in, 331critical aspects of, 327–331

primary, vs. idiopathic hypersomnia, 185Sleep drunkenness, in narcolepsy, 284Sleep duration

television viewing and, 128temperament and, 130

Sleep efficiency, calculation of, 68Sleep history

in arousal disorders, 299, 300tin idiopathic hypersomnia, 183in parasomnias, 305–306

Sleep hygieneabnormal, vs. narcolepsy, 177in delayed sleep phase syndrome, 106–107in idiopathic hypersomnia, 186in sleep-onset association disorder, 149–150principles of, 133t

Sleep-induced respiratory impairment. See alsoBreathing-related sleep disorders.

hypersomnia in, 21–22insomnia in, 20

Sleep laboratory environment, 69–70Sleep latency

definition of, 68testing of. See Multiple sleep latency test (MSLT).

Sleep log. See Sleep diary.Sleep onset, wake after, 68Sleep-onset association disorder, 134–136

clinical presentation and diagnosis of, 134–135insomnia in, 19treatment of, 135–136, 149–151

Sleep-onset insomniadifferential diagnosis of, 331–332treatment of, 108t, 332

Sleep-onset REM period, in narcolepsy, 176Sleep paralysis, 287–288

familial, 287isolated, 311–312

Sleep patterns and behaviorsabnormal, 22–23normal, 27–28variables affecting, 28–29

Sleep phobia, treatment of, 332–333Sleep physiology, after sleep deprivation, 8–9Sleep problems. See also Sleep disorders.

daytime effects of, model for, 162, 162fgeneral

behavioral difficulties and, 130impact of, 29–31nosology of, 28prevalence of, 29temperament and, 129variables affecting, 28–29

postcolic, 120Sleep process matrix, 45, 46tSleep schedule, irregular, insomnia in, 146Sleep specialist, referral to, indications for, 149Sleep spindles

asymmetrical, in infants, 54cerebellar vermis and, 37development of, 54–55learning experience and, 4thalamus and, 37

Sleep starts, 308Sleep talking, 295t, 310Sleep terrors, 285–286, 296

panic disorder and, 334treatment of, 333–334vs. nightmares, 287, 310, 311tvs. somnambulism, 285

Sleep-wake cyclebasal ganglia in, 37cerebellum in, 35, 37cortex in, 35, 36f, 38f, 39fdelayed maturation of, in colic, 117desynchronization of, in arousal disorders, 297during early and middle childhood, 56forebrain in, 37hypothalamus in, 37in infants, 53in neonates, 52irregular, arousal disorders and, 299, 302regular but inappropriate, 146regulation of, 270

factors influencing, 43–45, 46tthalamus in, 37

Sleep-wake cycle timing disordershypersomnia in, 22insomnia in, 20–21

Sleep-wake transition disorders, 307–310, 309fSleep zone, forbidden, 88f, 91, 101–102

colic and, 117Sleepiness. See Daytime sleepiness; Hypersomnia.

354 Index


Sleeplessness. See Insomnia.Sleepwalking. See Somnambulism.Slow-wave brain activity

during sleepin infants, 54in neonates, 51–52, 53f

during wakefulness, in infants, 53high-amplitude, 51–52, 53f

Slow-wave sleep, 2as state of being, 296–297electroencephalogram during

in infants, 55in middle childhood, 56in neonates, 51–52, 53f

electromyogram during, 60, 61energy conservation during, 2–3in repair of somatic tissue, 1–2instability of, in arousal disorders, 299mechanisms generating, 38fmemory processing during, 4restorative value of, 1–2theta-delta pattern during, 56, 58f

Smith-Magenis syndrome, insomnia in, 157Smoking

maternal, colic and, 114obstructive sleep apnea syndrome and, 243

Snoring, 206tcognitive changes associated with, 164growth retardation and, 213habitual

attention-deficit/hyperactivity disorder and, 163, 164t

prevalence of, 264, 265tin obstructive sleep apnea syndrome, 225–226in upper airway resistance syndrome, 263primary, 263–266

characteristics of, 197definition of, 263diagnosis of, 207t, 264, 266epidemiology of, 264, 265tfuture directions in, 266natural history of, 264polysomnography in, 266treatment of, 266vs. obstructive sleep apnea syndrome, 211vs. upper airway resistance syndrome, 263

Social contact, as nonphotic zeitgeber, 94Sodium oxybate, in narcolepsy, 178t, 179Somatomedin, levels of, sleep and, 2Somnambulism, 284–285, 295, 295t

confusional arousals and, 296genetics of, 298treatment of, 334

Somniloquy, 295t, 310Sound machines, in sleep-onset association disorder,

150Spastic diplegia, physical activity in children with, 289Spinal reflexes, during sleep, 80Starling resistor model, of pharynx, 223, 224fState of being

determination of, 296–297mixed (dissociated), in arousal disorders, 297

Stereotypiesbasic rest and activity cycle and, 62nonepileptic, parasomnias with, 282–283, 307–308,

309fSteroids. See Corticosteroids.Stimulants

in attention-deficit/hyperactivity disorder, 141–142in idiopathic hypersomnia, 186, 187tin narcolepsy, 178–179, 178tin post-traumatic hypersomnia, 191in sleep-onset insomnia, 108tinsomnia from, 20

Stressinsomnia and, 130, 143sleep problems and, 31

Subcortical arousals, in obstructive sleep apnea syndrome,201

Subdural hematoma, vs. idiopathic hypersomnia, 185Substance abuse

hypersomnia from, 21in case-based analysis of sleep problems, 44, 46tinsomnia from, 20

Sucking artifact, on electroencephalogram, 52, 55fSunday night bedtime, in delayed sleep phase syndrome,

105Suprachiasmatic nucleus

development of, 94input to, 90, 90fmelatonin secretion and, 92pacemaker properties of, 86, 94

Sweating, during sleep, 74, 75tSympathetic tone, in obstructive sleep apnea syndrome, 214Syringomyelia, sleep disorders in, 23t

Teeth grinding, 312–314, 313f, 314fTelevision viewing, total sleep time and, 128Temazepam, in sleep-onset insomnia, 332Temperament

colic and, 117–118, 121–123insomnia and, 129–130sleep and, 118–119sleep duration and, 130

Temperaturebody

circadian rhythm of, 86f, 90, 95in circadian rhythm sleep disorders, 101, 102f, 104–105regulation of, during sleep, 74–75, 75t

brain, during sleep, 73, 74tTemporal lobe B-cell lymphoma, narcolepsy and, 175Thalamus

development of, 49, 270in sleep-wake cycle, 37lesions of, sleep abnormalities in, 271

Theophylline, in apnea of prematurity, 277Thermistors, oronasal airflow measurement with, in

obstructive sleep apnea syndrome, 202Theta activity

during early and middle childhood, 56hypersynchronous, 56, 57fin infants, 53

Theta-delta pattern, 56, 58fin parasomnias, 307, 308f, 309f

Index 355


Third ventricle, 270Thoracoabdominal asynchrony, in obstructive sleep apnea

syndrome, 202Thyroid-stimulating hormone (TSH), during sleep,

75t, 76Toilet training, rigid, sleep-related enuresis and, 319Tongue reduction surgery, 241Tonic electrophysiologic activity, during sleep, 62Tonsillar hypertrophy. See Adenotonsillar hypertrophy.Tonsillectomy. See also Adenotonsillectomy.

in obstructive sleep apnea syndrome, 235–238, 237tsubtotal/partial (intracapsular), 255

Tonsillitis, acute, 251Tourette’s syndrome, insomnia in, 157Toxins, sleep-producing, 7Trace-alternant pattern, of quiet sleep, 51, 52fTracheostomy, 259–260

in obstructive sleep apnea syndrome, 243Traumatic life events

hypersomnia after, 189–191psychogenic insomnia after, 189, 190sleep problems and, 31

Trazodone, 331Triazolam, sleep and, 329, 330Tricyclic antidepressants

in arousal disorders, 302in confusional arousals, 333in sleep-related enuresis, 323, 335sleep and, 330

Trimipramine, 330Tumors

central nervous system, hypersomnia in, 195middle fossa, non–24-hour sleep-wake disorder and, 154

Turbinate reduction, 259

Ultradian cycledevelopmental changes in, 44in arousal disorders, 293, 294f, 297in body movements during sleep, 62in case-based analysis of sleep problems, 43–44, 46t

Unlearning theory of sleep, 7Upper airway resistance, high, 278–280, 279fUpper airway resistance syndrome

adenotonsillectomy in, 253–254, 257–258characteristics of, 197, 198fdefinition of, 263diagnosis of, 207t, 252, 266symptoms of, 263–264vs. idiopathic hypersomnia, 185vs. narcolepsy, 177vs. primary snoring, 263

Urinalysis, in sleep-related enuresis, 320Urinary incontinence. See Enuresis.Urinary tract

anatomic abnormalities of, enuresis in, 319infection of, enuresis and, 319

Urine culture, in sleep-related enuresis, 320Uvulopalatopharyngoplasty, 259

in obstructive sleep apnea syndrome, 240

Vagal tone, heart rate variability and, 64Valproic acid, 331Vasoconstriction, during sleep, 78Vasomotor activity

during sleep, 75, 75tperipheral, during sleep, 77t, 78

Velopharyngeal insufficiency, after adenotonsillectomy, 257

Venlafaxine, 331Ventilation

bilevel, vs. nasal CPAP, 239positive pressure. See Continuous positive airway

pressure (CPAP).Ventricular glioma, narcolepsy and, 175Vertex waves, on electroencephalogram, 55Vestibular dysfunction, in dyslexia, 4–5Vestibular-oculomotor split, in dyslexia, 5Video monitoring

in obstructive sleep apnea syndrome, 200in parasomnias, 282, 306in polysomnography, 65in rhythmic movement disorders, 307–308

Vitamin B12, in non–24-hour sleep-wake disorder, 154

Wakefulnessas state of being, 296–297circadian rhythm of, 95colic and, 117maintenance of, in idiopathic hypersomnia, 183–184mechanisms generating, 36f

WASO (wake after sleep onset), 68Weakness, muscle, vs. idiopathic hypersomnia, 186Weight loss, in obstructive sleep apnea syndrome,

241–242White noise, in sleep-onset association disorder, 150Williams syndrome, insomnia in, 157

Zaleplonin anxiety disorders, 335–336in insomnia, 336in sleep-onset insomnia, 332sleep and, 329

Zeitgebers, 101bright light as, 88f, 91nonphotic, in development of circadian rhythm, 94weak, 91

Ziprasidone, 331in insomnia, 336

Zolpidemin insomnia, 336sleep and, 329–330

356 Index
