university of groningen the infant motor profile heineman, … · 2016-03-06 · in the neuronal...
TRANSCRIPT
University of Groningen
The infant motor profileHeineman, Kirsten Roselien
IMPORTANT NOTE: You are advised to consult the publisher's version (publisher's PDF) if you wish to cite fromit. Please check the document version below.
Document VersionPublisher's PDF, also known as Version of record
Publication date:2010
Link to publication in University of Groningen/UMCG research database
Citation for published version (APA):Heineman, K. R. (2010). The infant motor profile: a standardized and qualitative assessment of motorbahaviour in infancy. Groningen: [s.n.].
CopyrightOther than for strictly personal use, it is not permitted to download or to forward/distribute the text or part of it without the consent of theauthor(s) and/or copyright holder(s), unless the work is under an open content license (like Creative Commons).
Take-down policyIf you believe that this document breaches copyright please contact us providing details, and we will remove access to the work immediatelyand investigate your claim.
Downloaded from the University of Groningen/UMCG research database (Pure): http://www.rug.nl/research/portal. For technical reasons thenumber of authors shown on this cover page is limited to 10 maximum.
Download date: 01-06-2020
The Infant Motor Pro�leA standardized and qualitative assessment of
motor behaviour in infancy
Kirsten Heineman
ISBN: 978-90-367-4532-1 (digital version)
© 2010, K.R. HeinemanNo parts of this thesis may be reproduced or transmitted in any forms or by any means, electronicor mechanical, including photocopying, recording or any information storage and retrieval system,without permission of the author
Cover: Willeke van der Broek, oktober 2007
Lay-out: Peter van der Sijde, Groningen
Printed by: drukkerij van Denderen, Groningen
ISBN: 978-90-367-4492-8 (printed version)
RIJKSUNIVERSITEIT GRONINGEN
The Infant Motor Pro�leA standardized and qualitative assessment of
motor behaviour in infancy
Proefschrift
ter verkrijging van het doctoraat in deMedische Wetenschappen
aan de Rijksuniversiteit Groningenop gezag van de
Rector Magni!cus, dr. F. Zwarts,in het openbaar te verdedigen op
woensdag 8 september 2010om 16.15 uur
door
Kirsten Roselien Heineman
geboren op 13 november 1978te Nijmegen
Promotores: Prof. dr. M. Hadders-AlgraProf. dr. A.F. Bos
Beoordelingscommissie: Prof. dr. J.G. BecherProf. dr. O.F. BrouwerProf. dr. L.S. de Vries
Paranimfen: Layla Ben-ZviKarin Middelburg
CONTENTS
Chapter 1 General introduction 9
Chapter 2 Evaluation of neuromotor function in infancy – 19a systematic review of available methodsJournal of Developmental and Behavioral Pediatrics 2008;29:315-23
Chapter 3 The Infant Motor Pro!le: a standardized and qualitative 37method to assess motor behaviour in infancyDevelopmental Medicine and Child Neurology 2008;50:275-82
Chapter 4 Development of adaptive motor behaviour in typically 51developing infantsActa Paediatrica 2010;99:618-24
Chapter 5 Construct validity of the Infant Motor Pro!le: 65Relation with prenatal, perinatal and neonatal risk factorsDevelopmental Medicine and Child Neurology 2010, epub ahead of print
Chapter 6 Concurrent and predictive validity of the Infant Motor Pro!le 77Submitted
Chapter 7 General discussion 95
Summary 107
Appendices 111
Appendix I: Summary on the domains and items of the Infant Motor Pro!leAppendix IIa: Validity of the testsAppendix IIb: Reliability of the testsAppendix III: Infant Motor Pro!le
Nederlandse samenvatting 129
Dankwoord 133
Curriculum Vitae 137
List of publications 141
9
General introduction
1CHAPTER
10
GENERAL INTRODUCTION
Theories on motor development
During the �rst two years of life, children show an impressive development of motor skills such
as reaching, grasping, sitting, crawling, standing and walking. Rapidly, they change from supine
infants to walking toddlers who are able to explore the world around them. Motor development is
a complex process in which many factors play a role. The limited knowledge on the exact processes
that are involved in motor development induced a wide range of theories. Traditionally, the process
of motor development was explained by the Neuronal Maturation Theory (NMT). According to this
theory, at birth all programs involved in motor development are present. As a result of maturation
of the central nervous system, motor development is brought about by a gradual unfolding of
these programs1-3. This unfolding process is not dependent on environmental factors and there
is only a limited role for exercise3,4. Because not all parts of the central nervous system mature at
the same time, motor development takes place from proximal to distal and from cranial to caudal
direction1,3-5. Based on these general developmental rules, Gesell and Amatruda described all motor
milestones and the ages at which they emerge in their book ‘Developmental Diagnosis’2.
In the Dynamic Systems Theory (DST)6,7, motor development is considered as a self-organizing
process in which all components are equally involved. Motor behaviour is determined by an
interplay between many subsystems, such as muscle force, muscle tone, the nervous system, body
weight, state and motivation of the infant, context variables and task characteristics. Development
is described as ‘a series of states of stability, instability, and phase shifts in the attractor landscape,
re!ecting the probability that a pattern will emerge under particular constraints’6. Phase shifts are
brought about by changes in one or more of the involved subsystems, leading to a forced transition
of the system to another state. Attractors states are preferred patterns of behaviour, for example
patterns of locomotion such as rolling, crawling or walking. Depending on the characteristics of all
factors involved, one way of locomotion is the most stable attractor in a certain situation and this
attractor has the highest probability of occurring. For example, a child who has been able to walk
independently for some weeks will choose the locomotion pattern ‘walking’ in easy circumstances,
but will probably switch to crawling when the surface is slippery. In this simpli�ed example, the
surface characteristic is regarded as the control parameter, that is the changed factor that makes
the system shift to another state6,7. In contrast with NMT, in which the central nervous system has
a leading role in the process of motor development, in DST the nervous system is only one of the
elements involved.
In the Neuronal Group Selection Theory (NGST)8-12 on motor development both genetic, innate
factors and environmental in!uences play a role. According to NGST, typical motor development is
characterized by two phases of variability. During the phase of primary variability, variation in motor
behaviour is not geared to external conditions. By means of spontaneous, self-generated activity,
the primary neuronal networks that are already present at birth and the corresponding motor
Chapter 1
1
General introduction
11
repertoires are explored. This can for example be observed in the so-called general movements
(GMs)13. General movements are the most frequent used movement pattern of the fetus and the
young infant until the age of approximately four months post term. They are characterised by
abundant variation and complexity and are not adapted to the environment nor goal-directed. From
three to four months onwards, these general movements are gradually displaced by goal-directed
motility, such as reaching movements.This goal-directed motility is at !rst characterised by abundant
variation, for example variation in reaching trajectories, amplitude of movements and movement
speed. Gradually, this seemingly random variation decreases and the second phase of motor
development, the phase of secondary variability, takes over. In the phase of secondary variability,
the child develops the ability to select out of the extensive motor repertoire the best motor strategy
in a certain situation, which gives motor behaviour a more e"cient impression. Development of this
ability to select adaptive motor strategies is dependent on the a#erent information brought about
by the self-generated activity. In other words, by means of a process of exploration and selection
the child develops the capacity to adapt motor behaviour to environmental constraints. This ability
to select adaptive motor strategies out of the motor repertoire occurs at function-speci!c ages10,11.
The process of typical motor development is disturbed in developmental motor disorders, such
as cerebral palsy (CP) or developmental coordination disorder (DCD). See below for descriptions of
these motor disorders. According to NGST, children with pre or perinatally acquired lesions of the
brain, such as children with CP and some children with DCD, exhibit stereotyped motor behaviour
with reduced variability. This might be related to a reduction of the repertoire of primary cortical-
subcortical neuronal networks. Besides the reduced neural repertoire, motor behaviour of these
children is also a#ected by problems in adaptive selection of the most suitable motor strategies
in a certain situation. This might be due to problems in processing a#erent, sensory information
that are often encountered in children with CP and DCD14,15, as adaptive selection relies on a#erent
feedback11,16.
Developmental motor disorders
Cerebral palsy (CP) is de!ned as a group of disorders of motor function, movement and posture; it
is permanent but not unchanging and is caused by non-progressive lesions or brain abnormalities
in the developing/immature brain17. It is often associated with severe intellectual de!cit (31%),
severe visual impairment (11%) and/or seizures (21%)18. CP is classi!ed as spastic CP, ataxic CP or
dyskinetic CP. Spastic CP is the most frequent form of CP (85-90%)19 and is characterised by at least
two of the following three characteristics: abnormal pattern of posture and/or movement, increased
tone and pathological re$exes. Spastic CP may be unilateral, involving limbs on one side of the
body, or bilateral with involvement of limbs on both sides of the body. Ataxic CP is characterized
by abnormal pattern of posture and/or movement in combination with a loss of orderly muscular
coordination so that movements are performed with abnormal force, rhythm and accuracy. In
dyskinetic CP, abnormal pattern of posture and/or movement is accompanied by involuntary,
12
Chapter 1
uncontrolled, recurring, occasionally stereotyped movements18. Prevalence of CP is 2 to 3 per 1000
live births18. Infants with very low birth weight (below 1500 gram) and/or very preterm birth are at
higher risk for development of CP. Due to improved prenatal and neonatal care, the survival of this
group of vulnerable infants has increased in the last decennia, possibly leading to an upward trend
in the overall prevalence of CP during the 1970s and 1980s. In the recent years trends have been
less clear18,19, but the prevalence seems to have stabilized or even decreased20. CP is caused in more
than 80% by brain lesions that can be visualised by brain imaging with magnetic resonance imaging
(MRI) or neonatal ultrasound scans. Type and localisation of the brain damage determine the clinical
subtype of CP and are related to the presence and severity of associated disabilities20. Often, no
speci!c cause can be identi!ed and it is assumed that a combination of factors is responsible for the
brain damage. For a long time, only complications of labour and delivery were considered causal risk
factors for CP, but current evidence suggests that intrauterine exposure to infection or in"ammation
and coagulation disorders could be more important causes21. At present, knowledge on the precise
contributors to the risk for CP is still not complete.
Compared to CP, developmental coordination disorder (DCD) is a much less disabling
developmental motor disorder, but it still has major impact on daily life and academic achievements.
Diagnosis of DCD can be made according to the following DSM-IV criteria22: performance in daily
activities that require motor coordination is substantially below that expected, given the person’s
chronological age and measured intelligence. This may be manifested by marked delays in achieving
motor milestones, dropping things, clumsiness, poor performance in sports, or poor handwriting.
The disturbance signi!cantly interferes with academic achievement and/or activities of daily living
and is not due to a general medical condition such as CP or muscular dystrophy. If intellectual delay
is present, the motor di#culties are greater than would be expected, given the level of delay22. DCD
is a common condition with an estimated prevalence of 5 to 6%22. It is often associated with the
presence of signs of minor neurological dysfunction (MND)16.
Assessment of neuromotor function in infancy
Infants born very preterm are especially at risk for developmental motor disorders such as CP or
DCD23-26. Due to improved prenatal and neonatal care the group of surviving very preterm infants
has increased in the last decennia23,26. Follow-up of preterm infants aims at the detection of those
infants that could bene!t from early intervention at young age, when the brain is characterised by
high plasticity27,28. In order to detect those infants, clinicians such as physical therapists, occupational
therapists, paediatricians and paediatric neurologists have a wide range of instruments that assess
neuromotor function in infancy at their disposal. In chapter 2, a systematic review of existing
methods and their psychometric properties is presented. We found that instruments that assess
qualitative aspects of motor behaviour, such as the general movement (GM) method13,29 and the
Test of Infant Motor Performance (TIMP)30, are most promising in terms of prediction of future
developmental outcome. However, these two methods are only suitable for infants until the age
13
1
General introduction
of four months. Therefore, we developed the Infant Motor Pro�le, a qualitative assessment which is
applicable throughout infancy until the age of 18 months.
The Infant Motor Pro�le
In the development process of the Infant Motor Pro�le, we started with precise observation and
analysis of video-recordings of spontaneous motor behaviour of typically developing infants from
three months until 2 years of age. In addition, we observed video-recordings of infants at high
risk for developmental motor disorders such as cerebral palsy. We tried to discern which aspects
of motor behaviour di!ered between typically developing and high-risk infants. With the NGST as
theoretical background, we categorized the observed variables into various domains. We de�ned
the items on these variables as precisely as possible and wrote guidelines for the examiner on how
to perform the assessment. We viewed and assessed a heterogeneous group of video-recordings
of typically and non-typically developing infants at di!erent ages, and step by step we re�ned the
items and the classi�cation into domains. In the end, the Infant Motor Pro�le had been developed
to the form as it is now and the process of validation was started.
The Infant Motor Pro�le is a video-based assessment of motor behaviour of infants aged 3 to 18
months. It consists of 80 items (see Appendix 1). These items constitute �ve domains. Two domains
are based on ideas of the NGST as described above10,11, the �rst one is on size of motor repertoire
(variation) and the second one is on the ability to select adaptive motor strategies (variability). These
domains consist of respectively 25 and 15 items. The other three domains are on movement "uency
(seven items), movement symmetry (10 items) and motor performance (23 items), the last one being
a more quantitative assessment of motor milestones. Details on the scoring of the items and on
calculation of the domain scores are described in chapter 3. The total IMP-score is computed as the
mean of the �ve domain scores. A general principle in calculating the scores on the various domains
is that only items that re"ect observed motor behaviour are taken into account. This means that for
example an item on variability of leg movements during independent walking is only scored if the
child indeed is able to walk independently. If not, the item is not taken into account in computing
the domain-score. An exception to this rule are the items of the performance domain, which can
always be scored. For example, if the child is not able to walk, item 5 of standing and walking is
scored as ‘1 = cannot walk’.
Aim of this thesis
The aim of this thesis was threefold: �rst to review existing methods to assess neuromotor function
in infancy. The second aim was development of the Infant Motor Pro�le and the third aim was to
investigate its psychometric properties. In order to be able to investigate reliability and validity of
the Infant Motor Pro�le, we performed a longitudinal prospective study on a heterogeneous group
of term and preterm born infants from 3 to 18 months. In addition, another group of term born
infants had cross-sectional assessments at the same ages.
Chapter 1
14
Main focus is on:
1. Intra- and interobserver reliability of the IMP.
2. Construct validity of the IMP, operationalized as relations between IMP-scores throughout
infancy and prenatal, perinatal and neonatal risk factors, including the presence of brain
pathology on neonatal ultrasound scans.
3. Concurrent validity of the IMP with the Alberta Infant Motor Scale (AIMS31) and the Touwen
Infant Neurological Examination (TINE32,33).
4. Predictive validity of the IMP-scores throughout infancy for neurological outcome at corrected
age of 18 months.
METHODS
Study group
The study sample consisted of a heterogeneous group of term and preterm infants, which is
a valuable type of sample when assessing validity of a new instrument. Thirty term born infants with
a median gestational age of 40.1 weeks (range 37.6-42 weeks) and a median birth weight of 3588
grams (range 2730-4470 grams) were followed longitudinally and assessed at ages 3,4,5,6,8,10,12,15
and 18 months. These infants were recruited amongst colleagues and acquaintances of the
researchers. A group of 59 preterm infants was followed longitudinally at corrected ages 4,6,10,12
and 18 months. They had all been admitted to the neonatal intensive care unit (NICU) of the Beatrix
Children’s Hospital of the University Medical Center Groningen (UMCG) between December 2003
and January 2005. Inclusion criteria were gestational age below 35 weeks, singleton or twin,
parents with appropriate understanding of the Dutch language and travel distance between the
child’s home and the hospital of approximately less than one hour. Infants with severe congenital
anomalies were excluded from the study. Median gestational age was 29.7 weeks (range 25-34.7
weeks) and median birth weight was 1285 grams (range 630-2180 grams). Another group of 116
term infants was recruited at Well Child Centres. These infants had a median gestational age of 40.1
weeks (range 37-43 weeks) and a median birth weight of 3500 grams (range 1960-4660 grams).
These infants had cross-sectional assessments at one (n=102), two (n=13) or three (n=1) of the ages
of 4, 6, 10, 12 or 18 months. All parents gave informed consent and the project was approved by the
local Ethics Committee.
For all infants, socio-economic, prenatal and neonatal data were collected on standardized
forms by means of an interview with the parents and consultation of NICU discharge certi!cates. For
the preterm infants, neonatal ultrasounds of the brain were assessed with respect to periventricular
leukomalacia (PVL)34 and intraventricular haemorrhages (IVH)35. Details on socio-economic and
neonatal characteristics of the study groups will be presented in the following chapters.
1
15
General introduction
Assessments
At all ages, the assessment consisted of a video-recording of motor behaviour of approximately 15
minutes. Motor behaviour was assessed in several positions, depending on the age and functional
capacities of the child: supine, prone, sitting, standing and walking and reaching, grasping and
manipulation of objects. Motor behaviour was spontaneous or was elicited by presenting toys to
the infant. On the basis of the video-recording, both the Infant Motor Pro�le and the Alberta Infant
Motor Scale (AIMS)31 were scored at a later time. The assessment consisted further of an age-speci�c
neurological examination that was carried out after the video-recording. At all ages, except at 18
months, this consisted of theTouwen Infant Neurological Examination (TINE)32,33. InTINE, neurological
signs are grouped according to age-speci�c norms into �ve possible clusters of dysfunction,
namely reaching and grasping, gross motor function, brain stem function, visuomotor function and
sensorimotor function (consisting of re!exes and muscle tone). Neurological condition is classi�ed
as abnormal if there is a distinct neurological syndrome, such as a hemisyndrome, irrespective of
number of deviant clusters. Neurological condition is classi�ed as minor neurological dysfunction
(MND) if there are two or more clusters of dysfunction. Neurological condition is considered normal-
suboptimal when one or two clusters are deviant and normal when no clusters are deviant33. At the
age of 18 months, the Hempel assessment36 was used instead of TINE, as TINE is only applicable
until the age of 12 to 15 months. The Hempel examination is suitable for preschool children from
18 months until 4 years of age. Similar to TINE, the Hempel assessment classi�es neurological signs
into clusters of dysfunction, namely �ne motor dysfunction, gross motor dysfunction, dysfunctional
muscle tone regulation, re!ex abnormalities and visuomotor dysfunction. Neurological condition
can be classi�ed as abnormal, complex MND (when there is more than one dysfunctional cluster),
simple MND (when one cluster of dysfunction is present) or normal (no deviant clusters or the
isolated presence of re!ex abnormalities)16.
Outline of the thesis
Chapter 2 consists of a systematic review of available methods to evaluate neuromotor function in
infancy and their psychometric properties. The review forms the background to which the Infant
Motor Pro�le was developed. Chapter 3 presents the pilot study on the IMP with description of the
instrument and �rst data on intra and inter observer reliability and concurrent validity. In chapter
4, a small side step is made, and development of adaptive motor behaviour in typically developing
infants is investigated with the NGST as theoretical background. Chapter 5 examines construct
validity of the IMP, operationalized as the relation of IMP-scores throughout infancy with prenatal,
perinatal and neonatal risk factors. Chapter 6 re-assesses reliability and investigates concurrent
validity of the IMP with the Alberta Infant Motor Scale (AIMS)31 and with Touwen Infant Neurological
Examination (TINE)32,33 and predictive validity for neurological outcome at 18 months. Chapter 7
consists of a general discussion on the IMP and its psychometric properties as presented in this
thesis.
16
Chapter 1
REFERENCES
1. Gesell A. The Embryology of Behavior. New York: Harper & Row, 1945.
2. Gesell A, Amatruda CS. Developmental Diagnosis. Normal and Abnormal Child Development, 2nd ed. NewYork: Harper & Row, 1947.
3. Illingworth RS. The Development of the Infant and Young Child: Normal and Abnormal Development, 3rded. Edinburgh: Churchill Livingstone, 1966.
4. Peiper A. Cerebral Function in Infancy and Childhood. 3rd ed. New York: Consultants Bureau, 1963.
5. McGraw MB. The Neuromuscular Maturation of the Human Infant. Classics in Developmental Medicine, No.4. London: Mac Keith Press, 1989.
6. Thelen E. Motor development. A new synthesis. American Psychologist 1995;50:79–95.
7. Ulrich BD. Dynamic systems theory and skill development in infants and children. In: Connolly KJ, ForssbergH, editors. Neurophysiology and Neuropsychology of Motor Development. Clinics in DevelopmentalMedicine No. 143/144, London: Mac Keith Press, 1997, p 319–45.
8. Edelman GM. Neural Darwinism: selection and reentrant signaling in higher brain function. Neuron1993;10:115-125.
9. Sporns O, Edelman GM. Solving Bernstein’s problem: a proposal for the development of coordinatedmovement by selection. Child Development 1993;64: 960–81.
10. Hadders-Algra M. The Neuronal Group Selection Theory: a framework to explain variation in normal motordevelopment. Dev Med Child Neurol 2000;42:566-572.
11. Hadders-Algra M. The Neuronal Group Selection Theory: promising principles for understanding andtreating developmental motor disorders. Dev Med Child Neurol 2000;42:707-715.
12. Hadders-Algra M. Early brain damage and the development of motor behavior in children: clues fortherapeutic intervention? Neural Plast 2001;8:31-49.
13. Hadders-Algra M. General movements: A window for early identi!cation of children at high risk fordevelopmental disorders. J Pediatr 2004;145:S12-18.
14. Wilson PH, McKenzie BE. Information processing de!cits associated with developmental coordinationdisorder: a meta-analysis of research !ndings. J Child Psychol Psychiatry 1998;39:829-840.
15. Cooper J, Majnemer A, Rosenblatt B, Birnbaum R. The determination of sensory de!cits in children withhemiplegic cerebral palsy. J Child Neurol 1995;10:300-309.
16. Hadders-Algra M. Developmental coordination disorder: is clumsy motor behaviour caused by a lesion ofthe brain at early age? Neural Plast 2003;10:39-50.
17. Surveillance of cerebral palsy in Europe: a collaboration of cerebral palsy surveys and registers. Surveillanceof Cerebral Palsy in Europe (SCPE). Surveillance of Cerebral Palsy in Europe. Dev Med Child Neurol2000;42:816-24.
18. Prevalence and characteristics of children with cerebral palsy in Europe. Surveillance of Cerebral Palsy inEurope. Dev Med Child Neurol 2002;44:633-40.
19. Krägeloh-Mann I, Bax M. Cerebral Palsy. In: Aicardi J , ed. Diseases of the nervous system in childhood. 3rdEdition. Clinics in Developmental Medicine. London: Mac Keith Press, 2009:210-242.
20. Krägeloh-Mann I, Cans C. Cerebral palsy update. Brain Dev 2009 Apr 20. Epub ahead of print.
21. Nelson KB, Grether JK. Causes of cerebral palsy. Curr Opin Pediatr 1999;11:487-91.
22. American Psychiatric Association. Diagnosis and Statistical Manual of Mental Disorders. 4th edn, TextRevision. Washington, DC: American Psychiatric Association, 2000.
17
1
General introduction
23. Wilson-Costello D, Friedman H, Minich N, Fanaro� AA, Hack M.. Improved survival rates with increasedneurodevelopmental disability for extremely low birth weight infants in the 1990s. Pediatrics 2005;115:997-1003.
24. Marlow N, Wolke D, Bracewell MA, Samara M; EPICure Study Group. Neurologic and developmentaldisability at six years of age after extremely preterm birth. N Engl J Med. 2005;352:9-19.
25. Davis NM, Ford GW, Anderson PJ, Doyle LW; Victorian Infant Collaborative Study Group. Developmentalcoordination disorder at 8 years of age in a regional cohort of extremely-low-birthweight or very preterminfants. Dev Med Child Neurol 2007;49:325-30
26. Larroque B, Ancel PY, Marret S, Marchand L, André M, Arnaud C, Pierrat V, Rozé JC, Messer J, Thiriez G,Burguet A, Picaud JC, Bréart G, Kaminski M; EPIPAGE Study group. Neurodevelopmental disabilities andspecial care of 5-year-old children born before 33 weeks of gestation (the EPIPAGE study): a longitudinalcohort study. Lancet. 2008;371:813-20.
27. Kolb B, Brown R, Witt-Lajeunesse A, Gibb R. Neural compensations after lesion of the cerebral cortex.Neural Plast 2001;8:1-16.
28. De Graaf-Peters VB, Hadders-Algra M. Ontogeny of the human central nervous system: what is happeningwhen? Early Hum Dev 2006;82:257-266.
29. Einspieler C, Prechtl HFR, Bos AF, et al. Prechtl’s method on the qualitative assessment of GeneralMovements in preterm, term and young infants. Clinics in Developmental Medicine No. 167. London: MacKeith Press; 2004.
30. Campbell SK, Kolobe TH, Osten ET, Girolami GL, et al. Construct validity of the test of infant motorperformance. Phys Ther 1995;75:585-596.
31. Piper MC, Darrah J. Motor assessment of the developing infant. Philadelphia: Saunders, 1994.
32. Touwen BCL. Neurological development in infancy. Clinics in Developmental Medicine No. 58. London:Mac Keith Press, 1976.
33. Hadders-Algra M, Heineman KR, Bos AF, Middelburg KJ. The assessment of minor neurological dysfunctionin infancy using the Touwen Infant Neurological Examination: strengths and limitations. Dev Med ChildNeurol, in press.
34. De Vries LS, Eken P, Dubowitz LM. The spectrum of leukomalacia using cranial ultrasound. Behav Brain Res1992;31;49:1-6.
35. Volpe JJ. Neurology of the Newborn. 4th ed. Philadelphia PA: WB Saunders, 2001.
36. Hempel MS. The neurological examination for toddler-age. PhD thesis. University of Groningen, 1993.
18
19
Evaluation of neuromotor function in infancy –a systematic review of available methods
Kirsten Heineman1,2 and Mijna Hadders-Algra1
1Department of Paediatrics, Institute of Developmental Neurology2Department of Neurology
University Medical Center Groningen
Journal of Developmental and Behavioral Pediatrics 2008;29:315-23
2CHAPTER
20
Chapter 2
ABSTRACT
Background: Neuromotor function in infancy can be evaluated in various ways. Assessment
instruments are used for early detection of children with a high risk for developmental disorders.
Early detection enables clinicians to provide intervention at a young age when plasticity of the
nervous system is high. The assessments may also be used to monitor intervention. The present
paper will review the psychometric properties of methods to assess neuromotora function in
infancy.
Methods: A literature search was performed in PubMed, Medline and PsycINFO (1966 to March
2007) on instruments to assess neuromotor functioning of infants.
Results: Fifteen instruments were included and classi!ed into four groups: 1) Comprehensive
neurological examinations (n=4). These techniques are widely used, though little is known about
their reliability. Their validity in predicting major developmental disorders such as cerebral palsy
is good; their predictive validity for minor motor disorders is moderate at best. 2) Procedures with
standardized scoring (n=7). These have good reliability, but only moderate predictive validity for
major developmental disorders. No data available for prediction of minor developmental disorders.
3) Observation of milestones (n=2). Its predictive validity for major developmental disorders is only
moderate, while reliability is good. 4) Assessment of quality of motor behavior or motor patterns
(n=2). These instruments have the best predictive validity for major and minor developmental
motor disorders, but current methods are only useful under the age of four months.
Conclusion: Prediction of developmental outcome at an early age is di"cult. In medical evaluations
of high-risk infants the best predictions are achieved through a combination of multiple,
complementary tools, that is, achieved milestones, neurological examination and assessment of the
quality of motor behavior.
2
21
Evaluation of neuromotor function in infancy
INTRODUCTION
As the chances of survival of preterm and high-risk full-term infants have increased1,2, extensive
follow-up programs have been developed to determine which of these infants need intervention.
Recent studies suggest that intervention may be most e�ective when it is applied during infancy
when there is high plasticity of the brain3,4. A prerequisite for early intervention is early detection of
infants with a high risk for major developmental disorders such as cerebral palsy (CP) and minor motor
disorders such as developmental coordination disorder (DCD) and minor neurological dysfunction
(MND). It appears that parents of children with developmental disorders are concerned signi!cantly
later than physicians are about the developmental status of their children and therefore, in general,
cannot be relied on for early recognition of infants who are likely to bene!t from early intervention5.
Physiotherapists, occupational therapists, pediatricians and other clinicians in primary health care
settings play an important role in early detection. They have a heterogeneous group of instruments
at their disposal for the detection of early evidence of motor dysfunction in high-risk infants. In
general, these instruments are not only used for detection but also for the evaluation of the
e�ectiveness of an intervention. Usually the instruments are chosen based on habit and for practical
reasons, and not on the basis of information regarding test accuracy and utility and theoretical basis6.
In fact, a primary selection criterion should be: “Has the instrument been designed for the task at
hand?” Kirshner and Guyatt7 classi!ed health measure instruments into three categories according
to the goals they served. The !rst one is discrimination. In the !eld of neuromotor assessment this
implies making a distinction between children who show features of a deviant neuromotor function
compared to the general, healthy population. The second purpose is prediction; that is, instruments
are used as a diagnostic tool to predict developmental outcome, for example, the likeliness that
a child will develop CP. The third purpose is evaluation, the measurement of longitudinal change
of an individual or group over time, for example, changes in motor function of infants enrolled in
early intervention programs. Instruments are generally validated for only one of the three goals. This
means that the instruments cannot automatically be used for other purposes6.
The aim of this paper is to present a systematic review of the instruments used for the evaluation
of neuromotor function and motor behavior in infancy.The contents of the methods will be reviewed,
while special attention will be paid to psychometric properties, that is, reliability and validity.
METHODS
Selection procedure
A literature search in the following databases was performed: PubMed, Medline (1966 to March
2007) and PsycINFO (1967 to March 2007). Keywords used were“neuromotor,”“motor development,”
“motor behavior,” “assessment,” “neurological examination,” “evaluation,” “instrument,” “method,”
“infants,” “neonatal,” “preschool” and “review.” All articles with a name of an assessment in the title
and/or abstract and reviews on one or more methods were selected. Further searches with names
22
Chapter 2
of assessments and authors were performed and references to the articles were studied to �nd
information on reliability and validity. Manuals were obtained when available in the Netherlands.
Instruments were included when they could be applied to infants aged three to eighteen
months and also when the age range of application was more extensive. We focused on the age
range from three to eighteen months, as it has been relatively neglected. Reviews on neonatal
neurological evaluation8,9, instruments to assess motor function of children diagnosed with CP6,10
and assessments of motor development and function in preschool children aged eighteen months
to four years11 were available. Methods were selected if they focused on neurological condition
or motor performance, or if they combined items on neuromotor function with items on other
developmental domains such as mental development, speech or behavior. In the latter case, only
data on reliability and validity of the motor subscale were reviewed. Instruments were only included
if they had been described in at least two English-language peer-reviewed papers.
Instruments used only for screening purposes were excluded, since 1) an overview of frequently
used general developmental screening instruments was provided by Glascoe12 and 2) our main
interest was “full” assessment of the infant’s neuromotor functioning. Screening was de�ned as “the
application to all children born of certain procedures that can be carried out in a short time by the
less specialized members of sta! and that will give indication of the presence and absence of certain
disabilities” (WHO 1967)13. Screening is important in clinical practice. This is illustrated, for instance,
by the considerable power of infant motor screening tests to predict CP, such as the Early Motor
Pattern Pro�le14 and Capute’s motor quotient15. Instruments were considered screening instruments
if the words screen or screening were part of the instrument’s name and/or if the authors mentioned
“screening”as the main purpose of the instrument. Fifteen instruments ful�lled the selection criteria
and were included in the review.
Evaluation procedure
The selected instruments were systematically evaluated with a focus on population, age, purpose
of instrument (discrimination, prediction or evaluation), type of instrument, test description, type
of test data (categorical vs. continuous), test construction and standardization, training required
to become an assessor and time needed to administer the test. For the classi�cation of the type of
instrument, the instrument’s dominant features were used.
The various types of validity and reliability were evaluated by the criteria presented in Table I.
Validity is the extent to which an instrument measures what it is intended to measure. We addressed
construct validity, concurrent validity and predictive validity. Construct validity is the extent to which
items re"ect the theoretical construct of interest, in this case neuromotor functioning. For tests of
neuromotor function, it may be assumed, for instance, that relationships of pre-, peri- and neonatal
adversities, such as preterm birth, and results of brain imaging with test scores may contribute
to construct validity. Concurrent validity is the extent to which scores relate to scores on another
measure on the same construct, ideally a gold standard. If there is no gold standard available,
23
2
Evaluation of neuromotor function in infancy
correlation with other established instruments is assessed. Predictive validity of an instrument is
the extent to which the scores on the instrument now predict future outcome11,17. In the present
review, we concentrated on predictive validity for developmental motor disorders; we distinguished
predictive validity for major motor disorders such as CP from predictive validity for minor disorders
such as DCD and MND. Factors that were taken into consideration in the judgment of the predictive
validity of the various instruments were type of population (typically developing children vs.
high-risk population), age at follow-up and tests used at follow-up. Reliability is the ability of a
Table I: Evaluation criteria
Construct validity
++
+
+
!
nda
very good
good
good
moderate
moderate
poor
no data
Scoring “good” twice, that is, good for relationship between scores and pre, peri- andneonatal adversities, and good for correlation of scores with results of brain imagingindicates very good construct validity
Scores negatively a!ected by pre-, peri- and neonatal adversities such as PT birth,IUGR, other medical complications, or
Scores correlate with visible pathology on brain imaging (US, MRI), or withelectrophysiological parameters of brain function (CFM, EEG)
Scores are correlated with level of motor development or presence of motor delaysnot documented with a reliable assessment tool
Scores show a signi"cant increase with increasing age
Scores are not related to any of the above-mentioned factors
No data available
Concurrent validity and predictive validity
++
+
!
nda
very good
good
moderate
poor
no data
Cohen’s kappa or Spearman’s ρ > 0.80a
Cohen’s kappa or Spearman’s ρ 0.61-0.80
Cohen’s kappa or Spearman’s ρ = 0.40 – 0.60
Cohen’s kappa or Spearman’s ρ < 0.40
no data available
Intra-observer agreement and inter-observer agreement
++
+
!
nda
very good
good
moderate
poor
no data
Cohen’s kappa or Spearman’s ρ > 0.80a
Cohen’s kappa or Spearman’s ρ 0.61-0.80; agreement >80%
Cohen’s kappa or Spearman’s ρ = 0.40 – 0.60; agreement 60-80%
Cohen’s kappa or Spearman’s ρ < 0.40; agreement < 60%
no data available
a Criteria for reliability according to Landis and Koch16, extended to Spearman’s ρ and % agreementPT = preterm, IUGR = intrauterine growth restriction, US = ultrasound, MRI = magnetic resonance imaging, CFM= cerebral function monitor, EEG = electroencephalography and PPV = positive predictive value
24
Chapter 2
measurement to give consistent scores on repeated assessments in the absence of change in the
characteristics being studied18. Intra-observer agreement is the stability of the observer’s ratings on
the same behavior. Often videotapes are used, which are scored twice by the same observer after
a pre-determined time interval. Interobserver reliability is the stability of ratings across di!erent
evaluators8. The ages used in the review imply that preterm infants are assessed at ages corrected
for preterm birth.
RESULTS
Description of instruments
Fifteen methods to assess neuromotor function in infancy ful"lled the inclusion criteria and were
included in the review. Their main characteristics are described in Table II. The age range in which
the instruments can be applied varied between birth to four months and zero to six years. Four
types of instruments were discerned: 1) comprehensive neurological examination, 2) procedures
with standardized scoring (i.e. condensed neurological assessment with or without observation
of motor behavior with standardized scoring), 3) observation of milestones and speci"c aspects
of motor behavior, and 4) quality of motor behavior. Four instruments were classi"ed as
comprehensive neurological examinations with a focus on cranial nerves, posture, muscle tone,
re#exes and reactions19-22. Seven instruments were procedures with standardized scoring24,25,29,31-34.
Two instruments focused on observation of motor milestones and speci"c aspects of motor
behavior35,36. Two instruments focus on assessing the quality of motor behavior, including postural
adjustments40,41,37.
The authors of all the instruments stated that the main purpose of their instrument was to
discriminate between infants with a deviant neuromotor condition and infants falling within
the range of typical development. Additional purposes have been described for six instruments:
three40,41,21,22 aimed at prediction of future neuromotor development and another three29,32,34 at
evaluation of changes in motor function over time or during intervention. Test construction di!ered
considerably for the various methods, but most assessments took elements from pre-existing
methods, adapted these and/or combined them with other test elements. For four instruments test
construction was not described. Data on standardization was available for seven out of the "fteen
instruments. These standardization procedures were carried out in strikingly heterogeneous ways
with respect to sample sizes (ranging from 35 to 2202 children) and types of populations (typically
developing infants, neonatal intensive care unit (NICU) graduates or infants with a high risk for
developmental delay). Assessors were pediatricians, neonatologists, psychologists, occupational
therapists, nurses or other clinicians working in NICU follow-up programs.Time needed to administer
the test varied from a few minutes to one hour.
25
2
Evaluation of neuromotor function in infancy
Tabl
eII:
Des
crip
tion
ofth
ein
stru
men
ts
Ass
essm
ent
(Aut
hor)
Shor
tna
me
Popu
lati
onA
gegr
oup
Purp
ose
Test
desc
ript
ion
-as
sess
men
tof:
Type
ofva
riab
les
Test
cons
truc
tion
/st
anda
rdiz
atio
nA
sses
sors
/tim
eto
adm
inis
ter
Neu
rolo
gica
lexa
min
atio
ns
Touw
enin
fant
neur
olog
ical
exam
inat
ion
(Tou
wen
19)
Touw
enin
fant
s0
mos
.to
inde
pend
ent
wal
king
DCr
ania
lner
ves,
post
ure,
tone
,re�
exes
and
reac
tions
,tr
unk
coor
dina
tion
and
gros
san
d�n
em
otor
func
tions
cat.
Base
don
trad
ition
alne
urop
edia
tric
conc
epts
(“G
roni
ngen
scho
ol”)
/nd
a
Pedi
atric
ians
and
othe
rsw
ithsp
eci�
cne
urod
evel
opm
enta
ltr
aini
ng/1
5m
inut
es
Am
iel-T
ison
neur
olog
ical
exam
inat
ion
(Am
iel-
Tiso
nan
dG
osse
lin20
)
Am
iel-
Tiso
nat
-ris
kin
fant
s0
to6
yrs.
DAc
tive
and
pass
ive
mus
cle
tone
,cra
nial
nerv
es,m
otor
mile
ston
es,s
pont
aneo
usm
otor
activ
ity,r
e�ex
esan
dre
actio
ns,q
ualit
ativ
eab
norm
aliti
es
cat.
Base
don
trad
ition
alne
urop
edia
tric
conc
epts
(“Fr
ench
scho
ol”)
/nda
Neo
nato
logi
sts,
pedi
atric
ians
with
neur
odev
elop
men
tal
trai
ning
,occ
upat
iona
lth
erap
ists
/10
min
utes
Activ
ean
dpa
ssiv
em
uscl
epo
wer
(de
Gro
ot21
)
Mus
cle
pow
erhi
gh-r
isk
infa
nts
3to
12m
os.
D,P
Spec
iale
mph
asis
onre
latio
nshi
pbe
twee
nac
tive
and
pass
ive
mus
cle
pow
er;
both
com
pone
nts
shou
ldbe
inba
lanc
ein
orde
rto
crea
tea
stab
lepo
stur
ean
d�u
entm
otili
ty
cat.
nda
/nda
nda
/nda
Ham
mer
smith
Infa
ntN
euro
logi
cal
Exam
inat
ion
(Haa
taja
etal
.22)
HIN
Ein
fant
s2
to24
mos
.D
,PCr
ania
lner
vefu
nctio
n,po
stur
e,m
ovem
ents
,ton
e,re
�exe
san
dre
actio
n(2
6ite
ms)
,dev
elop
men
tal
mile
ston
es(8
item
s),s
tate
ofbe
havi
or(3
item
s)
cont
.Ba
sed
onD
ubow
itzan
dD
ubow
itzm
etho
dfo
rne
urol
ogic
asse
ssm
ent
ofth
ene
wbo
rn23
/st
anda
rdiz
edin
alo
w-r
isk
popu
latio
nof
135
infa
nts
atag
e12
and
18m
os.
Rout
ine
clin
ical
prac
tice,
rela
tivel
yin
expe
rienc
edst
a!/n
da
26
Chapter 2
Ass
essm
ent
(Aut
hor)
Shor
tna
me
Popu
lati
onA
gegr
oup
Purp
ose
Test
desc
ript
ion
-as
sess
men
tof:
Type
ofva
riab
les
Test
cons
truc
tion
/st
anda
rdiz
atio
nA
sses
sors
/tim
eto
adm
inis
ter
Proc
edur
esw
ithst
anda
rdiz
edsc
orin
g
Prim
itive
Re�e
xPr
o�le
(Cap
ute
etal
.24)
PRP
infa
nts
0to
2yr
s.D
9pr
imiti
vere
�exe
sco
nt.
Re�e
xes
wer
ese
lect
edfr
omcl
inic
alob
serv
atio
nsof
path
olog
ical
lype
rsis
ting
prim
itive
re�e
xes
inch
ildre
nw
ithCP
/nda
Dev
elop
men
tal
pedi
atric
ians
/nda
Infa
ntN
euro
logi
cal
Inte
rnat
iona
lBa
tter
y(E
lliso
net
al.25
)
Infa
nib
at-r
isk
infa
nts
inN
ICU
follo
w-u
ppr
ogra
ms
1to
18m
os.
D20
item
sgr
oupe
din
to5
cate
gorie
s:sp
astic
ity,
vest
ibul
arfu
nctio
n,he
adan
dtr
unk
cont
rol,
rest
ing
tone
and
desc
riptio
nof
mot
orbe
havi
orof
the
legs
cat.
Base
don
fact
oran
alys
esof
neur
omot
orbe
havi
oron
32ite
ms
sele
cted
from
four
exis
ting
asse
ssm
ents
24,2
6-
28in
308
NIC
Ugr
adua
ted
infa
nts
/st
anda
rdiz
atio
nba
sed
onth
ese
308
infa
nts
Prof
essi
onal
sw
orki
ngin
NIC
Ufo
llow
-up
prog
ram
s/a
few
min
utes
Bayl
eySc
ales
ofIn
fant
Dev
elop
men
t,2nd
/ /3d
Ed.
(Bay
ley29
,30 )
BSID
-II/
IIIch
ildre
n1
mo.
to3.
5yr
s.D
,EM
otor
scal
e(8
1ite
ms,
gros
san
d�n
em
otor
beha
vior
)(in
addi
tion:
men
tals
cale
and
beha
vior
ratin
gsc
ale)
cont
.G
ener
alm
atur
atio
nalis
tpr
inci
ples
/st
anda
rdiz
atio
non
1700
US
child
ren
aged
1to
42m
os.
Psyc
holo
gist
s/1
5-20
min
utes
(mot
orsc
ale
only
)tot
alte
st:2
5-60
min
utes
Peab
ody
Dev
elop
men
tal
Mot
orSc
ales
,2nd
ed.
(Fol
ioan
dFe
wel
l31)
PDM
S-II
child
ren
0to
6yr
s.D
Gro
ssan
d�n
em
otor
scal
essu
btes
ts:r
e�ex
es,
stat
iona
ry,l
ocom
otio
n,ob
ject
man
ipul
atio
n,gr
aspi
ngan
dvi
sual
-mot
orin
tegr
atio
n
cont
.nd
a/s
tand
ardi
zed
on20
03ch
ildre
nO
ccup
atio
nalt
hera
pist
s/4
5-60
min
utes
27
2
Evaluation of neuromotor function in infancy
Ass
essm
ent
(Aut
hor)
Shor
tna
me
Popu
lati
onA
gegr
oup
Purp
ose
Test
desc
ript
ion
-as
sess
men
tof:
Type
ofva
riab
les
Test
cons
truc
tion
/st
anda
rdiz
atio
nA
sses
sors
/tim
eto
adm
inis
ter
Mov
emen
tA
sses
smen
tof
Infa
nts
(Cha
ndle
ret
al.32
)
MA
Ihi
gh-r
isk
infa
nts
inN
ICU
follo
w-u
pcl
inic
s
0to
12m
os.
D,E
65te
stite
ms
grou
ped
into
4se
ctio
ns:m
uscl
eto
ne,p
rimiti
vere
�exe
s,au
tom
atic
reac
tions
and
volit
iona
lmov
emen
t
cont
.nd
a/a
pro�
lefo
rty
pica
lmot
orbe
havi
orat
4m
os.h
asbe
ende
velo
ped
base
don
35in
fant
s
Phys
ical
ther
apis
ts,
occu
patio
nalt
hera
pist
s,ph
ysic
ians
,nur
ses,
psyc
holo
gist
san
dot
hers
who
have
spec
ializ
edkn
owle
dge
and
expe
rienc
ein
infa
ntde
velo
pmen
t/20
to30
min
utes
Neu
rom
otor
Beha
vior
alIn
vent
ory
(Gor
gaan
dSt
ern33
)
NBI
infa
nts
0to
12m
os.
D12
0ite
ms
grou
ped
into
cate
gorie
son
mus
cle
tone
,dev
elop
men
tal
mot
orab
ilitie
s,qu
ality
ofm
ovem
ent,
re�e
xes
and
reac
tions
,ora
l-mot
orbe
havi
or
cont
.nd
a/n
dand
a/n
da
Todd
lera
ndIn
fant
Mot
orEv
alua
tion
(Mill
eran
dRo
id34
)TI
ME
child
ren
4m
os.t
o3.
5yr
s.D
,EFi
vepr
imar
ysu
btes
tson
mob
ility
,mot
oror
gani
zatio
n,st
abili
ty,
soci
al/e
mot
iona
lab
ilitie
san
dfu
nctio
nal
perf
orm
ance
and
thre
ecl
inic
alsu
btes
ts:q
ualit
yra
ting,
atyp
ical
posi
tions
and
com
pone
ntan
alys
is;
focu
son
tran
sitio
nsbe
twee
nm
ovem
ent
patt
erns
cont
.Su
bdom
ains
deve
lope
dby
apa
nelo
fexp
erts
,�e
ldte
stin
gan
dex
pert
cons
ulta
tion
/sta
ndar
diza
tion
on14
4ch
ildre
nw
ithm
oder
ate
tose
vere
deve
lopm
enta
ldel
ays
and
731
TDch
ildre
n
Phys
ical
and
occu
patio
nalt
hera
pist
sor
othe
rclin
icia
nsw
ithex
pert
ise
inas
sess
men
tof
mot
orsc
ales
/10-
55m
inut
es
28
Chapter 2
Ass
essm
ent
(Aut
hor)
Shor
tna
me
Popu
lati
onA
gegr
oup
Purp
ose
Test
desc
ript
ion
-as
sess
men
tof:
Type
ofva
riab
les
Test
cons
truc
tion
/st
anda
rdiz
atio
nA
sses
sors
/tim
eto
adm
inis
ter
Obs
erva
tion
ofm
otor
beha
vior
Alb
erta
Infa
ntM
otor
Scal
e(P
iper
and
Dar
rah35
)
AIM
Sin
fant
s0
toin
depe
nden
tw
alki
ng
D58
item
sin
four
post
ural
posi
tions
:pro
ne,s
upin
e,si
ttin
gan
dst
andi
ng,
test
item
sar
esc
ored
asob
serv
edor
noto
bser
ved
base
don
draw
ings
cont
.Ev
alua
tion
ofth
ese
quen
tial
deve
lopm
ento
fpo
stur
alco
ntro
lre
lativ
eto
four
post
ural
posi
tions
/st
anda
rdiz
atio
non
2202
sex
and
age
stra
ti�ed
Cana
dian
FTin
fant
s
Pedi
atric
phys
ical
ther
apis
ts/1
5m
inut
es
Stru
ctur
edob
serv
atio
nof
mot
orpe
rfor
man
ce(P
erss
onan
dSt
röm
berg
36)
SOM
P-I
high
-ris
kin
fant
s0
to10
mos
.D
13as
cend
ing
scal
esof
mot
orde
velo
pmen
tfor
each
body
part
insu
pine
and
pron
epo
sitio
nan
din
the
who
lebo
dyw
ithth
ein
fant
sitt
ing,
stan
ding
and
inlo
com
otio
n
cont
.A
sses
smen
tofl
evel
ofm
otor
deve
lopm
ent
and
qual
ityof
mov
emen
ts/n
da
Phys
ical
ther
apis
ts,
neon
atol
ogis
ts,
paed
iatr
icne
urol
ogis
ts/n
da
Qua
lity
ofm
otor
beha
vior
Test
ofIn
fant
Mot
orPe
rfor
man
ce(C
ampb
ell37
)
TIM
Pin
fant
sbi
rth
(32
wks
.PM
A)t
o4
mos
.
D42
item
sgr
oupe
din
totw
ose
ctio
ns:o
bser
ved
and
elic
ited
sect
ions
cont
.So
me
item
sfr
omne
urol
ogic
al,
neur
obeh
avio
ral
and
mot
oras
sess
men
ts23
,37-
39/
nda
Phys
ical
and
occu
patio
nalt
hera
pist
s/n
da
Gen
eral
Mov
emen
ts(E
insp
iele
reta
l.40,
Had
ders
-Alg
ra41
)
GM
sin
fant
sbi
rth
to4
mos
.D
,PA
sses
smen
tofv
aria
bilit
yan
dco
mpl
exity
ofsp
onta
neou
sm
otor
beha
vior
insu
pine
posi
tion
cat.
Neu
rona
lgro
upse
lect
ion
theo
rypr
inci
ples
/nda
No
spec
i�c
prof
essi
on,
butt
rain
ing
cour
ses
are
requ
ired
tobe
com
ea
skill
edob
serv
er/
3-m
inut
evi
deo
cat.
=ca
tego
rical
data
,con
t.=
cont
inuo
usda
ta,D
=di
scrim
inat
ive,
E=
eval
uativ
e,FT
=fu
ll-te
rm,m
os.=
mon
ths,
nda
=no
data
avai
labl
e,N
ICU
=ne
onat
alin
tens
ive
care
unit,
P=
pred
ictiv
e,PT
=pr
eter
m,T
D=
typi
cally
deve
lopi
ng,U
SA=
Uni
ted
Stat
esof
Am
eric
a,yr
s.=
year
s.
29
2
Evaluation of neuromotor function in infancy
Validity and reliability
In Table III data on the di�erent kinds of validity and reliability of the selected methods are
presented. Extended versions of Table III on validity and reliability can be found in Appendix II or
on the journal’s website. Construct validity, the extent to which the items of the instruments re�ect
neuromotor function, was moderate to very good for most instruments. For three instruments no
data were available. No data were available on concurrent validity with other methods to assess
neuromotor function for eight instruments; for the other seven the range was from moderate to
very good. Studies on the predictive validity of the instruments for CP or minor developmental
disorders showed good predictive validity for six instruments, moderate predictive validity for four
instruments and no data were described for �ve instruments. Data were available on intra-observer
reliability for only four assessments. Two instruments had a very good and two had a moderate
intra-observer reliability. For eleven instruments information on interobserver agreement was
available: it ranged from moderate to very good.
Table III: Validity and reliability of the tests
Assessment(short name)
Constructvalidity
Concurrentvalidity
Predictivevalidity
Intra-observeragreement
Inter-observeragreement
Touwen19,74 nda nda + nda nda
Amiel-Tison20,42-46 + + + nda
Muscle power47-49 ++ nda + nda nda
HINE22,50-52 + nda + nda ++
PRP24,53 ! ! nda nda +
Infanib25,54,55 + nda nda nda
BSID-II29,56 nda nda +
PDMS-II31,56 nda ++
MAI32,57-62 nda nda
NBI54,63 + nda nda nda nda
TIME64 nda nda nda ++
AIMS65-67 ++ ++ ++
SOMP-I68,69 + nda nda +
TIMP37,70-73 + + + nda +
GMs40,41,74-78 ++ + + ++ ++
++ = very good, + = good, = moderate,!= poor, nda = no data available, see Table I for description of evaluationcriteria. Extended versions of Table III on validity and reliability can be found in Appendix II.
30
Chapter 2
DISCUSSION
We will discuss the �fteen instruments according to their classi�cation into 1) comprehensive
neurological examination, 2) procedures with standardized scoring, 3) observation of milestones
and speci�c aspects of motor behavior and 4) assessment of the quality of motor behavior.
Comprehensive neurological examinations19-22 are mainly based on traditional neuropediatric
concepts. They have a good construct validity and a good predictive validity for the development
of major motor disorders such as CP or the prediction of future locomotor function. Virtually no
information was available on the predictive validity of neurological examinations for more subtle,
minor developmental disorders such as MND and DCD. The exception to this rule was the Touwen
examination for which some data were available showing that MND at school age may be predicted
to a limited extent74. It is interesting to note that little information was available on the reliability of
these frequently used assessments.
The second group of instruments consists of procedures with standardized scoring, which are
quite heterogeneous and comprise the assessment of primitive re!exes (Primitive Re!ex Pro�le
(PRP)24), a compilation of tests of pre-existing neurological assessments (Infant Neurological
International Battery (Infanib)25) and more extensive test batteries like the Motor Scales of the Bayley
(BSID-II/III)29,30 and the Peabody Developmental Motor Scales (PDMS-II)31. Three tests, the Movement
Assessment of Infants (MAI)32, Neuromotor Behavioral Inventory (NBI)33 and Toddler and Infant Motor
Evaluation (TIME)34, combine neurological test items with observation of speci�c aspects of motor
behavior. In general, the procedures with standardized scoring have a poor to moderate construct
validity and concurrent validity. Predictive validity was either only moderate25,31,32 or no data were
available. This might be a point of concern, but it should be realized that these tests have not been
developed with the aim of predicting future motor disorders. The Bayley Scales, MAI and TIME can
be used for the evaluation of changes in neuromotor functioning. Another attractive feature of the
Bayley and PDMS-II is their standardization for very large groups of children. Interobserver reliability
of most of these assessments is good to very good.
Two instruments consist mainly of observation of milestones and speci�c aspects of motor
behavior: the Alberta Infant Motor Scale (AIMS)35 and the Structured Observation of Motor
Performance (SOMP)36. Construct validity for both instruments is acceptable. For the SOMP,
additional validity data are lacking. Concurrent validity for the AIMS is good, but predictive validity
for major developmental disorders is only moderate. Reliability of these observational instruments
is satisfactory.
The last two assessments, the General Movement method (GM)40,41 and the Test of Infant Motor
Performance (TIMP)37 share the feature that they both assess quality of motor behavior or motor
patterns. The GM method assesses quality of spontaneous motor behavior in supine position. The
TIMP di"ers from the GM method in that not only does it focus on spontaneous movements but
also mainly assesses quality of postural adjustments elicited by handling the infant. Interestingly,
31
2
Evaluation of neuromotor function in infancy
both methods have a good predictive validity, the TIMP for major developmental disorders such as
CP, and the GM method for both major and minor developmental disorders such as MND. Construct
and concurrent validity and reliability of the GMs and TIMP are also satisfactory. These last two
assessments are only useful under the age of four months.
CONCLUDING REMARKS
The main issue in choosing a suitable instrument in a certain situation is de!ning the goal that
the instrument needs to serve. If the main goal is discriminating between infants with deviant
neuromotor function and infants falling in the range of typical development, all !fteen reviewed
instruments can be used. However, only three of the instruments (BSID-II29,30, MAI32 and TIME34) can
be used for evaluation of the e"ect of intervention. Data on predictive validity are available for ten
instruments. The predictive validity for most of them is moderate at best. Only the two instruments
that assess qualitative aspects of motor behavior (TIMP37 and GM-method40,41) show good predictive
validity. Prediction of developmental outcome at an early age will never be perfect. This is inherent
in the developmental characteristics of the young brain4. Therefore, in medical evaluation of high-
risk infants, the best prediction is achieved when multiple, complementary clinical tools are used.
A good combination is medical history including achieved milestones, physical and neurological
examination, a speci!c assessment of the quality of motor behavior and results of neuroimaging
such as ultrasound or MRI assessment.
32
Chapter 2
REFERENCES
1. Lorenz JM, Wooliever DE, Jetten JR, et al. A quantitative review of mortality and developmental disabilityin extremely premature newborns. Arch Pediatr Adolesc Med 1998;152:425-435.
2. Clark SL, Hankins GDV. Temporal and demographic trends in cerebral palsy – Fact and �ction. Am J ObstetGynecol 2003;188:628-633.
3. Blauw-Hospers CH, Hadders-Algra M. A systematic review of the e�ects of early intervention on motordevelopment. Dev Med Child Neurol 2005;47:421-432.
4. De Graaf-Peters VB, Hadders-Algra M. Ontogeny of the human central nervous system: what is happeningwhen? Early Hum Dev 2006;82:257-66.
5. Ehrmann Feldman D, Couture M, Grilli L, et al. When and by whom is concern �rst expressed for childrenwith neuromotor problems? Arch Pediatr Adolesc Med 2005;159:882-886.
6. Ketelaar M, Vermeer A, Helders PJ. Functional motor abilities of children with cerebral palsy: a systematicliterature review of assessment measures. Clin Rehabil 1998;12:369-380.
7. Kirshner B, Guyatt G. A methodological framework for assessing health indices. J Chronic Dis 1985;38:27-36.
8. Majnemer A, Mazer B. Neurologic evaluation of the newborn infant: de�nition and psychometricproperties. Dev Med Child Neurol 1998;40:708-715.
9. Majnemer A, Snider L. A comparison of developmental assessments of the newborn and young infant.Ment Retard Dev Disabil Res Rev 2005;11:68-73.
10. Boyce WF, Gowland C, Rosenbaum PL, et al. Measuring quality of movement in cerebral palsy: a review ofinstruments. Phys Ther 1991;71:813-819.
11. Tieman BL, Palisano RJ, Sutlive AC. Assessment of motor development and function in preschool children.Ment Retard Dev Disabil Res Rev 2005;11:189-196.
12. Glascoe FP. Screening for developmental and behavioral problems. Ment Retard Dev Disabil Res Rev2005;11:173-9.
13. World Health Organization (WHO). 1967. The early detection and treatment of handicapping defectsin young children. Report on a working party convened by the regional office for Europe and WHO.Copenhagen: WHO Regional Office for Europe.
14. Morgan AM, Aldag JC. Early identi�cation of cerebral palsy using a pro�le of abnormal motor patterns.Pediatrics 1996;98:692-7.
15. Capute AJ, Shapiro BK. The motor quotient. A method for the early detection of motor delay. Am J Dis Child1985;139:940-2.
16. Landis JR, Koch GG. The measurement of observer agreement for categorical data. Biometrics 1977;33:159-174.
17. Baxter P. Invalid measurement validity. Editorial. Dev Med Child Neurol 2005;47:291-291.
18. Mitchell SK. Interobserver agreement, reliability, and generalizability of data collected in observationalstudies. Psychol Bull 1979;86:376-390.
19. Touwen BCL. Neurological development in infancy. Clin Dev Med No. 58. Philadelphia: Lippincott; 1976.
20. Amiel-Tison C, Gosselin J. Neurological development from birth to six years. Guide for examination andevaluation. Baltimore and London: The Johns Hopkins University Press; 2000.
21. De Groot L, Hopkins B, Touwen BC. A method to assess the development of muscle power in preterms afterterm age. Neuropediatrics 1992;23:172-179.
22. Haataja L, Mercuri E, Regev R, et al. Optimality score for the neurologic examination of the infant at 12 and18 months of age. J Pediatr 1999;135:153-161.
33
2
Evaluation of neuromotor function in infancy
23. Dubowitz L, Dubowitz V. The neurological assessment of the preterm and full-term newborn infant. Clinicsin Developmental Medicine No 79. London: Heinemann; 1981.
24. Capute AJ, Accardo PJ, Vining EP, et al. Primitive re�ex pro�le. A pilot study. Phys Ther 1978;58:1061-1065.
25. Ellison PH, Horn JL, Browning CA. Construction of an Infant Neurological International Battery (Infanib) forthe assessment of neurological integrity in infancy. Phys Ther 1985;65:1326-1331.
26. Milani-Comparetti A, Gidoni EA. Routine developmental examination in normal and retarded children.Dev Med Child Neurol 1967;9:631-638.
27. Paine R, Oppe T. Neurologic examination of infants and children. Clinics in Developmental Medicine No.20/21. Philadelphia: JB Lippincott; 1966.
28. Amiel-Tison C, Grenier A. Neurologic evaluation of the newborn and the infant. New York: Masson; 1983.
29. Bayley N. Manual for the Bayley Scales of Infant Development: Second Edition. San Antonio: ThePsychological Corporation; 1995.
30. Bayley, N. (2006). Bayley scales of infant and toddler development–Third edition. San Antonio, TX: HarcourtAssessment, Inc.
31. Folio MR, Fewell RR. Peabody Developmental Motor Scales: Examiner’s Manual. 2nd ed. Texas: Pro-ED; 2000.
32. Chandler LS, MS Andrews, MW Swanson. Movement Assessment of Infants (MAI). Rolling Bay, WA: InfantMovement Research; 1980.
33. Gorga D, Stern F. The neuromotor behavioural inventory. New York: New York Hospital – Cornell MedicalCenter; 1979.
34. Miller LJ, Roid GH. The TIME Toddler and Infant Motor Evaluation, a standardized assessment. San Antonio,Texas: Therapy Skill Builders; 1994.
35. Piper MC, Darrah J. Motor assessment of the developing infant. Philadelphia: Saunders; 1994.
36. Persson K, Strömberg B. Structured Observation of Motor Performance (SOMP-I) applied to neonatallyhealthy fullterm infants at the ages of 0-10 months. Early Hum Dev 1995;40:127-143.
37. Campbell SK, Kolobe TH, Osten ET, Girolami GL, et al. Construct validity of the test of infant motorperformance. Phys Ther 1995;75:585-596.
38. Amiel-Tison C, Grenier A. Neurologic evaluation of the newborn and the infant. New York: Masson; 1983
39. Brazelton TB, Nugent JK. Neonatal behavioral assessment scale. (3rd ed.) Clinics in DevelopmentalMedicine No. 137, London: Mac Keith Press; 1995.
40. Einspieler C, Prechtl HFR, Bos AF, et al. Prechtl’s method on the qualitative assessment of GeneralMovements in preterm, term and young infants. Clinics in Developmental Medicine No. 167. London: MacKeith Press; 2004.
41. Hadders-Algra M. General movements: A window for early identi�cation of children at high risk fordevelopmental disorders. J Pediatr 2004,145:S12-18.
42. Paro-Panjan D, Neubauer D, Kodric J, et al. Neurological assessment at term age: clinical application,correlation with other methods, and outcome at 12 to 15 months. Dev Med Child Neurol 2005;47:19-26.
43. Paro-Panjan D, Sustersic B, Neubauer D. Comparison of two methods of neurologic assessment in infants.Pediatr Neurol 2005;33:317-324.
44. Stewart AL, Costello AM, Hamilton PA, et al. Relationship between neurodevelopmental status of verypreterm infants at one and four years. Dev Med Child Neurol 1989;31:756-765.
45. Roth SC, Baudin J, Pezzani-Goldsmith M, et al. Relation between neurodevelopmental status of verypreterm infants at one and eight years. Dev Med Child Neurol 1994;36:1049-1062.
46. Roth S, Wyatt J, Baudin J, et al. Neurodevelopmental status at 1 year predicts neuropsychiatric outcome at14-15 years of age in very preterm infants. Early Hum Dev 2001;65:81-89.
34
Chapter 2
47. Samsom JF, De Groot L, Hopkins B. Muscle power in‘high-risk’preterm infants at 12 and 24 weeks correctedage: a measure for early detection. Acta Paediatr 2001;90:1160-1166.
48. Samsom JF, de Groot L, Hopkins B. Muscle power and medical history in high risk preterm infants at 3months of corrected age. Neuropediatrics 1998;29:127-132.
49. Samsom JF, Sie LT, de Groot L. Muscle power development in preterm infants with periventricular �aring orleukomalacia in relation to outcome at 18 months. Dev Med Child Neurol 2002;44:735-740.
50. Haataja L, Mercuri E, Guzzetta A, et al. Neurologic examination in infants with hypoxic-ischemicencephalopathy at age 9 to 14 months: use of optimality scores and correlation with magnetic resonanceimaging �ndings. J Pediatr 2001;138:332-337.
51. Frisone MF, Mercuri E, Laroche S, et al. Prognostic value of the neurologic optimality score at 9 and 18months in preterm infants born before 31 weeks’ gestation. J Pediatr 2002;140:57-60.
52. Ricci D, Cowan F, Pane M, et al. Neurological examination at 6 to 9 months in infants with cysticperiventricular leukomalacia. Neuropediatrics 2006;37:247-252.
53. Bartlett D. Primitive re�exes and early motor development. J Dev Behav Pediatr 1997;18:151-157.
54. Gorga D, Stern FM, Ross G, et al. Neuromotor development of preterm and full-term infants. Early Hum Dev1988;18:137-149.
55. Pedersen SJ, Sommerfelt K, Markestad T. Early motor development of premature infants with birthweightless than 2000 grams. Acta Paediatr 2000;89:1456-1461.
56. Palisano RJ. Concurrent and predictive validities of the Bayley Motor Scale and the Peabody DevelopmentalMotor Scales. Phys Ther 1986;66:1714-1719.
57. Harris SR, Swanson MW, Andrews MS, Sells, et al. Predictive validity of the Movement Assessment ofInfants. J Dev Behav Pediatr 1984;5:336-342.
58. Piper MC, Pinnell LE, Darrah J, et al. Early developmental screening: sensitivity and speci�city ofchronological and adjusted scores. J Dev Beh Pediatr 1992;13:95-101.
59. Harris SR. Early neuromotor predictors of cerebral palsy in low-birthweight infants. Dev Med Child Neur1987;29:508-519.
60. Deitz JC, Crowe TK, Harris SR. Relationship between infant neuromotor assessment and preschool motormeasures. Phys Ther 1987;67:14-17.
61. Haley SM, Harris SR, Tada WL, et al. Item reliability of the Movement Assessment of Infants. Phys OccupTher Pediatr 1986;6:21-39.
62. Harris SR, Haley SM, Tada WL, et al. Reliability of observational measures of the Movement Assessment ofInfants. Phys Ther 1984;64:471-477.
63. Gorga D, Stern FM, Ross G. Trends in neuromotor behavior of preterm and fullterm infants in the �rst yearof life: a preliminary report. Dev Med Child Neurol 1985;27:756-766.
64. Rahlin M, Rheault W, Cech D. Evaluation of the primary subtests of Toddler and Infant Motor Evaluation:implications for clinical practice in pediatric physical therapy. Pediatr Phys Ther 2003;15:176-183.
65. Bartlett DJ, Fanning JE. Use of the Alberta Infant Motor Scale to characterize the motor development ofinfants born preterm at eight months corrected age. Phys Occup Ther Pediatr 2003;23:31-45.
66. Darrah J, M Piper, MJ Watt. Assessment of gross motor skills of at-risk infants: predictive validity of theAlberta Infant Motor Scale. Dev Med Child Neurol 1998;40:485-491.
67. Piper MC, Pinnell LE, Darrah J, Maguire T, Byrne PJ. Construction and validation of the Alberta Infant MotorScale (AIMS). Can J Public Health 1992;83:S46-50.
68. Persson K, Strömberg B. Structured Observation of Motor Performance (SOMP-I) applied to preterm andfullterm infants who needed neonatal intensive care. A cross-sectional analysis of progress and quality ofmotor performance at ages 0-10 months. Early Hum Dev 1995;43:205-224.
35
2
Evaluation of neuromotor function in infancy
69. Hammarlund K, Persson K, Sedin G, et al. A protocol for structured observation of motor performancein preterm and term infants. Interobserver agreement and intraobserver consistency. Ups J Med Sci1993;98:77-82.
70. Barbosa VM, Campbell SK, Berbaum M. Discriminating Infants From Di�erent Developmental OutcomeGroups Using the Test of Infant Motor Performance (TIMP) Item Responses. Pediatr Phys Ther 2007;19:28-39.
71. Kolobe TH, Bulanda M, Susman L. Predicting motor outcome at preschool age for infants tested at 7, 30,60, and 90 days after term age using the Test of Infant Motor Performance. Phys Ther 2004;84:1144-1156.
72. Campbell SK, Kolobe TH, Wright BD, et al. Validity of the Test of Infant Motor Performance for predictionof 6-, 9- and 12-month scores on the Alberta Infant Motor Scale. Dev Med Child Neurol 2002;44:263-272.
73. Campbell SK. Test-retest reliability of the Test of Infant Motor Performance. Ped Phys Ther 1999;11:60-66.
74. Hadders-Algra M, Mavinkurve-Groothuis AM, Groen SE, et al. Quality of general movements and thedevelopment of minor neurological dysfunction at toddler and school age. Clin Rehabil 2004;18:287-299.
75. Cioni G, Ferrari F, Einspieler C, et al. Comparison between observation of spontaneous movements andneurologic examination in preterm infants. J Pediatr 1997;130:704-711.
76. Cioni G, Prechtl HF, Ferrari F, et al. Which better predicts later outcome in full-term infants: quality ofgeneral movements or neurological examination? Early Hum Dev 1997;50:71-85.
77. Prechtl HF, Einspieler C, Cioni G, et al. An early marker for neurological de�cits after perinatal brain lesions.Lancet 1997;10;349:1361-1363.
78. Groen SE, de Blecourt AC, Postema K, et al. General movements in early infancy predict neuromotordevelopment at 9 to 12 years of age. Dev Med Child Neurol 2005;47:731-738.
36
37
3CHAPTER
The Infant Motor Pro�le:a standardized and qualitative method to assess motor
behaviour in infancy
Kirsten Heineman1,2, Arend Bos3, Mijna Hadders-Algra1
1Department of Paediatrics, Institute of Developmental Neurology2Department of Neurology
3Department of Paediatrics, Division of NeonatologyUniversity Medical Center Groningen
Developmental Medicine and Child Neurology 2008;50:275-82
38
Chapter 3
ABSTRACT
Background: A reliable and valid instrument to assess neuromotor condition in infancy is a
prerequisite for early detection of developmental motor disorders. We developed a video-based
assessment of motor behaviour, the Infant Motor Pro�le (IMP), on motor abilities, movement
variability, ability to select motor strategies, movement symmetry and �uency. The IMP consists
of 80 items and is applicable from 3 to 18 months. The present study aims at teasting intra and
interobserver reliability and concurrent validity of the IMP with the Alberta Infant Motor Scale and
Touwen neurological examination.
Methods: The study group consisted of 40 low-risk at term born (median gestational age (GA) 40
weeks, range 38-42) and 40 high-risk preterm born infants (median GA 29.6, range 26-33) with
corrected ages 4 to 18 months (31 girls, 49 boys).
Results: Intra and interobserver agreement of the IMP were satisfactory (Spearman ρ=0.9).
Concurrent validity of IMP and AIMS was good (Spearman ρ=0.8, p<0.0005) . The IMP was able
to di"erentiate between infants with normal neurological condition, simple minor neurological
dysfunction (MND), complex MND and abnormal neurological condition (p<0.0005).
Conclusion: The IMP may be a promising tool to evaluate neurological integrity during
infancy –a suggestion which needs con�rmation by means of assessment of larger groups of infants
with heterogeneous neurological conditions.
39
3
The infant Motor pro�le
INTRODUCTION
Assessment of motor dysfunction in young children is still in its infancy. This is regrettable because
a sensitive, reliable and valid instrument is a prerequisite for early detection of infants with
developmental motor disorders, such as cerebral palsy (CP) and developmental co-ordination
disorder (DCD). Early detection is needed to warrant the provision of intervention at young ages,
i.e. at ages when brain development is characterized by high degrees of plasticity1-3. Adequate
assessment techniques are also indispensable for the evaluation of the e!ectiveness of intervention
strategies.
Recently, we reviewed "fteen instruments to assess neuromotor condition in infancy with
respect to their characteristics and psychometric properties (unpublished data). It was concluded
that the traditional forms of neurological examination showed good predictive validity for major
developmental disorders such as CP. However, instruments that assess quality of motor behaviour
(TIMP4, General Movement method5,6) turned out to be most promising in terms of predictive validity,
as they have good predictive value for both major and minor (minor neurological dysfunction
(MND), DCD) developmental disorders.
The "nding that quality of motor behaviour and variation in particular is an essential parameter
to assess the condition of the young nervous system, "ts well to the concepts of the Neuronal Group
Selection Theory (NGST)7-10. According to NGST, normal motor development is characterized by two
phases of variability. During the phase of primary variability variation in motor behaviour is not
geared to external conditions. Next, at function-speci"c ages, the phase of secondary variability
takes over. The child learns to select the best motor strategy for each situation on the basis of
a!erent information resulting from self-generated motor activity, from active trial and error. In other
words, the child develops the capacity to adapt motor performance to environmental constraints.
Children with pre or perinatally acquired lesions of the brain, such as children with CP and some
children with DCD, exhibit stereotyped motor behaviour. This might be related to a reduction of the
repertoire of primary cortical-subcortical neuronal networks. Motor behaviour of these children is
not only a!ected by the reduced size of the repertoire of motor strategies, but also by problems
in the process of selection of the most e#cient strategy for each situation. The latter problems
may be attributed to the absence of speci"c strategies which forces the child to choose in certain
conditions amongst non-optimal strategies and the presence of de"cits in the processing of sensory
information9-11.
Having noticed that application of motor variability as a means to evaluate neuromotor
condition functions so well in early infancy, we embarked on the development of a similar
assessment instrument for neuromotor condition at ages between 3 and 18 months. This resulted
in the development of the Infant Motor Pro"le (IMP).
40
Chapter 3
The Infant Motor Pro�le (IMP)
The IMP evaluates spontaneous motor behaviour of infants aged 3 to 18 months, or rather until the
age at which the child has a couple of months experience with walking independently. This means
that the IMP can be applied in infants with a more or less typical development until approximately
the age of 18 months. In children with moderate to severe developmental motor disorders
the IMP can be used beyond the age of 18 months. The IMP is a video based assessment. Motor
behaviour is evaluated in the following conditions: supine, prone, sitting, standing and walking. In
addition, reaching, grasping and manipulation of objects are evaluated in supine and in supported
sitting conditions, i.e. while sitting on the parent’s lap. The child’s motor activity may be entirely
spontaneous or may be elicited by the presentation of interesting objects. The order in which
the various conditions are evaluated depends on the child’s age, functional capacities, mood and
interest. In the youngest infants, assessment in general starts with an observation of behaviour in
supine condition for 5 minutes. In older children, it is more common to start with a sitting condition,
either on the parent’s lap or in a freely sitting condition.
The IMP consists of 80 items classi!ed into !ve subscales (for details see Appendix). The !rst
subscaleaddressesmovementvariability; itassessesthesizeofthemovementrepertoire.The25items
of this subscale are scored as ‘insu"ciently variable’ denoting the presence of a limited repertoire of
ways in which the child is able to perform the task, or as ‘su"ciently variable’ denoting the presence
of a variable repertoire of motor solutions. The second subscale also addresses movement variability
and comprises 15 items; it evaluates the child’s ability to select adaptive motor strategies from his or
her movement repertoire, an ability which develops during the phase of secondary variability. The
items of this subscale are scored as ‘no selection’ or as ‘adaptive selection’. ‘No selection’ means that
the child does not select from his or her repertoire the movement strategy which !ts the situation
best, but uses various motor strategies. If the child chooses for most of the time the most suitable
motor strategy from his or her motor repertoire for a speci!c situation ‘adaptive selection’ is scored.
The other subscales of the IMP are movement symmetry, #uency, and performance. The subscale
symmetry contains 10 items which address the presence or absence of stereotyped asymmetries.
In fact, movement symmetry may be regarded as a speci!c form of movement variability, but we
decided for a separate subscale on presence and absence of symmetry as it might have speci!c
diagnostic value12. Movement #uency re#ects the ability of the infant to !ne-tune motor output.
The subscale consists of 7 items, including two items on the presence of tremors. Loss of movement
#uency is one of the !rst signs of a non-optimal neurological condition6,13. The subscale performance
of 23 items is a more or less traditional one on achieved motor milestones. This means that the
IMP does not only address quality of motor behaviour but also evaluates what the child is able to
achieve.
Aim of the study
The aim of the present study is to assess psychometric properties of the IMP. First, inter and
3
41
The infant Motor pro�le
intraobserver reliability of the IMP are addressed. Second, relations between IMP scores and age,
gestational age at birth and ultrasound scans of the brain made during the neonatal period are
evaluated. These relations might shed light on the construct validity of the IMP. We hypothesize
that total IMP scores and scores on the subscales variability-ability to select and performance are
positively correlated with corrected age as these capacities develop with increasing age, whereas
such relationships between corrected age and the subscores variability-size of repertoire, symmetry
and �uency are absent. In addition, we assume that low gestational age leads to lower IMP scores,
especially on the subscales variability-size of repertoire, symmetry and �uency. Infants born
preterm have a higher risk for brain lesions which can lead to a reduced, less variable movement
repertoire, including stereotyped asymmetries. As loss of �uency is one of the �rst signs of a non-
optimal neuromotor condition, we expect preterm infants to have lower �uency scores. We expect
that infants with serious brain lesions on ultrasound scans show lower IMP scores.
Third, concurrent validity of the IMP with the AIMS14 and the neurological examination according to
Touwen15 is assessed. To this end, a group of 80 infants with corrected ages from 4 to 18 months was
evaluated by means of the IMP, the AIMS and the neurological examination according to Touwen.
We expect a signi�cant positive correlation between AIMS and total IMP score and IMP performance
score, and weakly to moderately positive correlations between AIMS and other subscale scores of
the IMP. In addition, we hypothesize that the degree of neurological dysfunction shows a clear
inverse relationship with the total IMP score and IMP subscores.
METHODS
Participants
The study group consisted of 80 infants (31 girls and 49 boys). Forty infants were born at term
(gestational age at birth 40 weeks (median value; range 38 to 42 weeks) without pre or perinatal
complications with a median birth weight of 3550 grams (range 2730 to 4220 grams). Forty infants
were born preterm with a gestational age at birth of 26-33 weeks (median value 29.6 weeks) and
a median birth weight of 1180 grams (range 585 to 2120 grams). The preterm infants had been
admitted to the neonatal intensive care unit (NICU) of the Beatrix Children’s Hospital of the University
Medical Center (UMC) in Groningen during the years 2003 and 2004.
Infants were assessed cross-sectionally at the corrected ages of 4, 6, 10, 12 and 18 months
(Table I). All parents of the infants signed an informed consent form. The project was approved by
the Ethics Committee of the UMC in Groningen.
Procedures
Each assessment started with a video recording of spontaneous motor behaviour in the following
conditions: supine, prone, sitting with or without support, standing with or without support, walking
with or without support and sitting on the parent’s lap. In the supine position and when seated
42
Chapter 3
on the parent’s lap small attractive objects were presented to assess the infant’s ability to reach,
grasp and manipulate objects. Toys were also used to elicit rolling, crawling, standing up behaviour
and trunk rotation during sitting and standing. In the youngest children recording always started
with the infant lying in supine for 5 minutes. The order of the conditions was not �xed but was
adapted to the child’s interest. In general, video recording lasted about 15 minutes. On basis of the
video recording both the Infant Motor Pro�le scores and AIMS14 were determined. Total AIMS scores,
instead of percentiles, were used in the data processing, as Canadian reference values are currently
inappropriate for Dutch children16. O!-line scoring of the IMP took approximately 10 minutes per
video recording. Calculation of IMP scores will be explained in the next paragraph.
After the video recording of motor behaviour, a neurological examination according to
Touwen15 was carried out. The neurological condition of the child was classi�ed as normal, as the
simple or complex form of minor neurological dysfunction (MND) or as de�nitely abnormal. A child
was classi�ed as de�nitely abnormal, when he or she exhibited a clear neurological syndrome such
as a clear hemisyndrome, a marked hypertonia or hypotonia in combination with clearly abnormal
re"exes or a hyperexcitability syndrome. MND denoted the presence of one or more clusters of
mild signs of neurological dysfunction, such as mild abnormalities in muscle tone regulation, mildly
abnormal re"exes, mild visuomotor dysfunction or mild dysfunction in gross motor or �ne motor
performance11. Simple MND indicates the presence of one cluster of mild signs of neurological
dysfunction and can be regarded as a normal, but non-optimal condition of the nervous system.
Complex MND denotes the presence of two or more clusters of dysfunction and is related to early
brain damage17.
Data analyses
IMP scores were separately calculated for each subscale according to the following formula:
Sum of item scores
Score = * 100%
(no. of items of scale or subscale – no. of items NA) * max score of items
‘No’ stands for number and ‘NA’ means not applicable (for instance the item ‘variability of arm
movements during walking’when a child is not able to walk) and ‘max score of items’ is the maximum
Table I: Number and age distribution of infants included in the validity study
Age (months) 4 6 10 12 18 Total
Term 7 7 7 8 11 40Preterm 7 7 7 8 11 40
Total 14 14 14 16 22 80
43
3
The infant Motor pro�le
score amongst the items of the subscale. For the subscales variability-size of repertoire, variability-
ability to select and �uency the maximum score for each item is two and for the subscale symmetry,
three. Because the performance subscale contains items with di�erent numbers of maximum scores
(see Appendix I), the scores are weighed before computing the performance subscale score. If a
child scores 3 points on the item ‘reaching and grasping of an object in supine position’ where the
maximum score is 7 points, the child is given 3/7 = 0.43 points. For an item with a 4-point scale,
the score would be 3/4 = 0.75 points. This means that all performance items contribute to a similar
extent to the performance subscore and the total IMP score (see Appendix I). The total IMP score is
computed by summing the scores on the �ve subscales and dividing this number by �ve. All scores
on subscales and the total IMP score are expressed as a percentage with a maximum score of 100%.
To assess the interobserver agreement of the IMP, 38 video recordings were scored independently
by two assessors. The sample was created by randomly selecting from each of the age groups
recordings of three preterm and three full-term infants. Both assessors evaluated eight videos
from the 12-month age group (four full term and four preterm infants) and twelve videos from the
18-month group (six full term and six preterm infants) in order to obtain su�cient data on standing
and walking items to determine reliability. One of the assessors (KRH) knew whether an infant was
born preterm or at term, but she was not aware of the details of the infant’s clinical history. The
other observer (MHA) was blind with respect to the clinical status of the infants. In order to establish
intraobserver agreement of the IMP scores, KRH evaluated the same 38 video recordings twice with
an interval of one month. Inter and intraobserver reliability at subscale level were analysed by means
of the Spearman correlation coe�cient ( ). Interpretation of correlation coe�cients was as follows:
< 0.49: weak relationship, 0.50 < < 0.75: moderate relationship and > 0.75: strong relationship18.
The relationships between IMP scores, AIMS scores and corrected age were evaluated with
Spearman rank correlation and associated con�dence intervals. The relation between IMP and
neurological classi�cation and status at birth (preterm or full-term) was assessed with help of the
non-parametric Kruskal-Wallis and Mann Whitney U test, respectively. Mann Whitney U test was also
used to determine di�erences in IMP scores between infants with and without serious brain lesions.
Throughout the analyses di�erences and correlations with a p-value < 0.05 were considered to be
statistically signi�cant (two-tailed testing).
RESULTS
Reliability
Intra and interobserver reliability of total IMP score were strong (Table II). Intraobserver agreement
was strong for subscales variability-repertoire, variability-ability to select and performance and was
moderate for subscales �uency and symmetry. Interobserver reliability of the subscales variability-
ability to select and performance was strong. Interobserver reliability was moderate for subscales
variability-repertoire and �uency and weak for subscale symmetry.
44
Chapter 3
Validity
The IMP total score varied between 66 and 97 (median value 88). No signi�cant di�erences in total
IMP scores and all subscales were found between girls and boys. The IMP total score was signi�cantly
related to corrected age (ρ = 0.68 (95% CI 0.54-0.78), p < 0.0005). The two subscales variability-ability
to select and performance, which on theoretical assumptions were expected to correlate with
age, were signi�cantly related to corrected age (variability-ability to select: ρ = 0.68 (95% CI 0.54-
0.78), p < 0.0005; performance: ρ = 0.91 (95% CI 0.86-0.94), p < 0.0005). Also, a positive correlation
between age and symmetry score was found, indicating that the infants showed less asymmetric
behaviour with increasing age (ρ = 0.40 (95% CI 0.20-0.57), p< 0.0005). The subscales variability-size
of repertoire and �uency were not related to corrected age (ρ = 0.19 and ρ = 0.07 resp.).
Preterm infants had signi�cantly lower total IMP scores than infants born at term (median IMP
score FT 89 vs. PT 82; Mann-Whitney: p < 0.0005). The di�erence was particularly present in the
subscales variability-size of repertoire (p < 0.0005), �uency (p < 0.0005) and symmetry (p = 0.021).
No statistically signi�cant di�erence between the two groups was found for the other two subscales
variability-ability to select (Mann-Whitney: p = 0.064) and performance (p = 0.126). Also, preterm
infants had signi�cantly lower AIMS scores than full term infants (p = 0.046).
Repeated brain ultrasound scans made during the �rst weeks of life indicated serious brain
lesions in each of three preterm infants: one infant had a unilateral periventricular haemorrhage
due to a venous infarction19, one had bilateral periventricular echodensities evolving into extensive
periventricular cysts20 and one infant had middle cerebral artery infarction on the right side.
Compared to the other 37 PT infants with no or milder brain lesions such as transient periventricular
echodensities, these three infants showed signi�cant lower IMP scores (Mann-Whitney: p = 0.006)
and AIMS scores (Mann-Whitney: p = 0.041).
The AIMS scores of the various assessments varied from 8 to 58 with a median value of 46. The
AIMS score showed a high correlation with corrected age (ρ = 0.91 (95% CI 0.86-0.94), p < 0.0005).
The total IMP score was related to the AIMS score (ρ = 0.78 (95% CI 0.68-0.85), p < 0.0005). This
correlation was brought about in particular by the high correlation between the conceptually
Table II: Intra and interobserver reliability of IMP: Spearman correlation coe�cients
IMP-subscale Spearman ρ (95% con"dence interval)
Intraobserver reliability Interobserver reliability
Total IMP score 0.9 (0.8-0.9) 0.9 (0.8-1.0)IMP-subscale:Variability-repertoire 0.8 (0.6-0.9) 0.7 (0.4-0.8)Variability-ability to select 0.8 (0.7-0.9) 0.8 (0.6-0.9)Fluency 0.6 (0.3-0.8) 0.7 (0.4-0.8)Symmetry 0.6 (0.4-0.8) 0.4 (0.0-0.6)Performance 1.0 (0.9-1.0) 0.9 (0.8-1.0)
45
3
The infant Motor pro�le
similar AIMS score and IMP subscore on performance (ρ = 0.94 (95% CI 0.91-0.96), p < 0.0005). Also
the subscales variability-size of repertoire, variability-ability to select and symmetry were related to
the AIMS (variability-size of repertoire: ρ = 0.42 (95% CI 0.22-0.59), p < 0.0005; variability-ability to
select: ρ = 0.69 (95% CI 0.56-0.79), p < 0.0005; symmetry: ρ = 0.46 (95% CI 0.27-0.62), p < 0.0005),
but the subscale on �uency was not (ρ = 0.20, n.s.). When the e�ect of the confounding factor of age
was partialled out, the correlations between the various IMP scores and the AIMS score remained
statistically signi�cant (ρ-values varying from 0.31 (95% CI 0.10-0.50), p = 0.006 (�uency) to 0.72
(95% CI 0.60-0.81), p < 0.0005 (performance); median ρ-value 0.338).
Twenty-seven of the 80 infants had a normal neurological condition, 19 were classi�ed as
having simple MND, 32 as having complex MND and 2 as neurologically abnormal. Neurological
condition was not related to corrected age. The total IMP score showed a highly signi�cant relation
with neurological condition (Kruskal-Wallis: p < 0.0005, Figure 1). Also the various subscales of the
IMP were related to neurological condition: variability-size of repertoire (Kruskal-Wallis: p < 0.0005),
variability-ability to select (p = 0.013), �uency (p< 0.0005), symmetry (p = 0.023), and performance
(p = 0.024; Figure 2A-E). It is interesting to note that di�erences in IMP total score and in scores on
the subscales of variability and performance are found in particular between infants with complex
MND or an abnormal neurological condition and those with a normal neurological condition or
simple MND.
TotalIMP-score
100
90
80
70
60
normal S-MND C-MND abnormal
-
- -
-
- -
Figure 1: Relationship between neurological condition and total IMP score. The data are presented bymedian values (horizontal bars) and interquartile ranges (boxes) and ranges (vertical lines). N = neurologicallynormal, S-MND = simple MND, C-MND = complex MND. Kruskal Wallis: ** p < 0.0005
46
Chapter 3
Varia
bilit
y-r
eper
toire
100
90
80
70
60
normal S-MND C-MND abnormal
A-
- -
-
- -
Varia
bilit
y–
abili
tyto
sele
ct
100
90
80
70
60
normal S-MND C-MND abnormal
B
–
- -
Flue
ncy
100
90
80
70
normal S-MND C-MND abnormal
C
- -
- -- -
Sy
mm
etry
100
90
80
70
60
normal S-MND C-MND abnormal
D
- -
-
Perf
orm
ance
100
90
80
70
60
normal S-MND C-MND abnormal
50
40
E
Figure 2: Relationship between neurologicalcondition and the �ve subscales of the IMP.A) variability-repertoire, B) variability-ability to selectC) �uency, D) symmetry, E) performance. KruskalWallis: * p < 0.05, ** p < 0.0005
47
3
The infant Motor pro�le
DISCUSSION
The present study indicates that the IMP can be performed in a reliable way. In addition, the study
suggests that the IMP has a good concurrent validity with the AIMS and with the infant’s neurological
condition.
The intra and interobserver agreement on the total IMP score was high. Agreement on
the subscales variability-repertoire, variability-ability to select, performance and �uency was
satisfactory. Intraobserver reliability on the subscale symmetry was moderate, but the interobserver
reliability on this subscale was only poor. This was probably caused by the virtual absence of serious
asymmetries in the group used for reliability testing: 34 out of the 38 infants obtained symmetry
scores of 100% from both assessors. The other four infants were given somewhat lower symmetry
scores, but the degree of reduction di!ered slightly between the assessors. This suggests that the
population studied suited the purpose of assessing the reliability of the symmetry subscale only to
a moderate extent.
In line with our assumptions, the IMP total score increased signi"cantly with increasing age.
This e!ect was brought about by the relationship between age and the subscores performance and
variability. Our "nding parallels those on the AIMS; the AIMS score is signi"cantly related to age14.
The positive correlation between age and the subscore variability-ability to select underlines the
notion that infants learn to select a preferred strategy from the motor repertoire when they get
older9. In our population, symmetry scores also correlated with age with older infants showing less
asymmetries than younger infants. This is probably due to the fact that in our study group only one
infant had a serious asymmetry. This girl was assessed at 10 months corrected age; later follow-up
assessments revealed that she developed a unilateral spastic cerebral palsy. Young infants quite
often show mild to moderate asymmetries which usually resolve spontaneously with increasing
age21. The low prevalence of serious unilateral disorders in the present sample precludes a proper
evaluation of the e!ect and utility of the symmetry subscale. As expected, movement �uency was
not related to age.
Preterm infants showed lower total IMP scores and lower scores on the subscales variability-
repertoire, symmetry and �uency, as was expected. Infants born preterm are at risk for brain
lesions, which can lead to a reduction of movement repertoire and the presence of stereotyped
asymmetries10,12. Loss of movement �uency is one of the "rst signs associated with non-optimal
neurological condition6,13. Prematurity was not related to the scores on the subscales variability-
ability to select and performance. Infants with serious brain lesions as seen on ultrasound, such as
haemorrhage grade IV or periventricular cystic lesions, indeed showed signi"cantly lower IMP scores
than infants with no or mild brain lesions. Altogether, construct validity of the IMP, operationalized as
relation of IMP scores with age, gestational age and brain lesions on ultrasound, seems satisfactory.
The IMP total score and the "ve subscales of the IMP correlated with the AIMS scores. Especially
the performance subscale showed high correlation with the AIMS, which is not surprising, as both
48
Chapter 3
assess achievements in motor behaviour. Also if correction for age is applied, these correlations
still subsist, which shows that both AIMS and IMP measure the same construct, i.e. integrity of the
brain. Nevertheless, it is interesting to note that the IMP is more sensitive in picking up the e!ect of
preterm birth and serious brain lesions than the AIMS.
Di!erences in total IMP scores and scores on the "ve subscales were found between the four
neurological conditions, with an inverse relationship between IMP scores and degree of neurological
dysfunction. Especially the subscales variability-repertoire and #uency showed clear di!erences.
Di!erences in total IMP scores, scores on variability subscales and on performance subscale were
in particular found between infants with complex MND or abnormal neurological condition and
those with a normal neurological condition or simple MND. Complex MND is the form of minor
neurological dysfunction which in contrast to simple MND has clinical relevance. It probably is
caused by adversities in the pre or perinatal period, and in some cases it might be considered as a
borderline form of cerebral palsy11,22. All in all, concurrent validity of the IMP with the AIMS and with
the neurological examination according to Touwen can be regarded as good.
A limitation of this pilot study is the small sample size of 80 infants of di!erent ages and with
di!erent risk for developmental motor disorders. Present data on reliability and concurrent validity
with AIMS and theTouwen neurological examination are promising, but further research on construct
and predictive validity is necessary. Extrapolation of the di!erence in predictive power of General
Movement-assessment5,6 and the neonatal neurological assessment22 generates the hypothesis
that predictive power of the IMP surpasses that of the AIMS and the neurological examination. The
next step in IMP-development will be the generation of norms and the determination of clinical
applicability.
In conclusion: the Infant Motor Pro"le is a reliable video based assessment of motor behaviour
in infancy which does not only address the child’s motor abilities but also movement variation,
ability to select movement strategies, symmetry and #uency. The current study suggests that the
latter characteristics are promising parameters of neurological integrity – a suggestion which needs
con"rmation by means of assessment of larger groups of infants with a heterogeneous neurological
condition.
ACKNOWLEDGEMENTSWe thank Cornill Blauw-Hospers, MSc , Victorine de Graaf-Peters, MSc and Karin Middelburg, MD, for performinga part of the neurological assessments, and Ms Lidy Kingma-Balkema and Ms Ineke Bakker for skilled technicalassistance. The study was supported by a Junior Scienti"c Masterclass grant of the post-graduate schoolBehavioral and Cognitive Neurosciences (BCN), University of Groningen.
49
3
The infant Motor pro�le
REFERENCES
1. Kolb B, Brown R, Witt-Lajeunesse A, Gibb R. Neural compensations after lesion of the cerebral cortex.Neural Plasticity 2001;8:1-16.
2. De Graaf-Peters VB, Hadders-Algra M. Ontogeny of the human central nervous system: what is happeningwhen? Early Hum Dev 2006;82:257-266.
3. Staudt M. (Re-)organization of the developing human brain following periventricular white matter lesions.Neurosci Biobehav Rev 2007;31:1150-6.
4. Campbell SK, Kolobe TH, Osten ET, Girolami GL, Lenke M. Construct validity of the test of infant motorperformance. Phys Ther 1995;75:585-596.
5. Einspieler C, Prechtl HF, Bos A, Ferrari F, Cioni G. Prechtl’s method on the qualitative assessment of generalmovements in preterm, term and young infants. Clin Dev Med No. 167, London: Mac Keith Press, 2004.
6. Hadders-Algra M. General movements: A window for early identi�cation of children at high risk fordevelopmental disorders. J Pediatr 2004;145:S12-18.
7. Edelman GM. Neural Darwinism: selection and reentrant signaling in higher brain function. Neuron1993;10:115-125.
8. Sporns O, Edelman GM. Solving Bernstein’s problem: a proposal for the development of coordinatedmovement by selection. Child Dev 1993;64:960-981.
9. Hadders-Algra M. The Neuronal Group Selection Theory: a framework to explain variation in normal motordevelopment. Dev Med Child Neurol 2000;42:566-572.
10. Hadders-Algra M. The Neuronal Group Selection Theory: promising principles for understanding andtreating developmental motor disorders. Dev Med Child Neurol 2000;42:707-715.
11. Hadders-Algra M. Developmental coordination disorder: is clumsy motor behaviour caused by a lesion ofthe brain at early age? Neural Plast 2003;10:39-50.
12. Guzzetta A, Haataja L, Cowan F, Bassi L, Ricci D, Cioni G, Dubowitz L, Mercuri E. Neurological examination inhealthy term infants aged 3-10 weeks. Biol Neonate 2005;87:187-196.
13. Hadders-Algra M, Mavinkurve-Groothuis AM, Groen SE, Stremmelaar EF, Martijn A, Butcher PR. Quality ofgeneral movements and the development of minor neurological dysfunction at toddler and school age.Clin Rehabil 2004;18:287-299.
14. Piper MC, Darrah J. Motor assessment of the developing infant. Philadelphia: WB Saunders Company, 1994.
15. Touwen BCL. Neurological development in infancy. Clin Dev Med No. 58. Philadelphia: Lippincott, 1976.
16. Fleuren KM, Smit LS, Stijnen T, Hartman A. New reference values for the Alberta Infant Motor Scale need tobe established. Acta Paediatr 2007;96:424-427.
17. Soorani-Lunsing I, Woltil HA, Hadders-Algra M. Are moderate degrees of hyperbilirubinemia in healthyterm neonates really safe for the brain? Pediatr Res 2001;50:701-5.
18. Portney LG, Watkins MP. Part IV Data analysis: Correlation. Foundations of clinical research. Applications topractice. 2nd ed. Upper Saddle River, NJ: Prentice Hall Health; p. 494, 2000.
19. Volpe JJ. Intraventricular haemorrhage in the premature infant - current concepts. Part II. Ann Neurol1989;25:109-116.
20. De Vries LS, Eken P, Dubowitz LM. The spectrum of leukomalacia using cranial ultrasound. Behav Brain Res1992;31;49:1-6.
21. Michaelis R, Asenbauer C, Buchwald-Saal M, Haas G, Krageloh-Mann I. Transitory neurological �ndings in apopulation of at risk infants. Early Hum Dev 1993;34:143-153.
22. Hadders-Algra M. Two distinct forms of minor neurological dysfunction: perspectives emerging from areview of data of the Groningen Perinatal Project. Dev Med Child Neurol 2002;44: 561-571.
50
51
4CHAPTER
Development of adaptive motor behaviour in typicallydeveloping infants
Kirsten Heineman1,2, Karin Middelburg1, Mijna Hadders-Algra1
1Department of Paediatrics, Institute of Developmental Neurology2Department of Neurology
University Medical Center Groningen
Acta Paediatrica 2010;99:618-24
52
Chapter 4
ABSTRACT
Aim:During motor development, infants learn to select adaptive motor strategies out of their motor
repertoire. The aim of this study is twofold: �rst, to investigate whether the presence of adaptive
motor behaviour can be observed reliably, and second, to explore the ages at which clinically
observable transition to adaptive motility emerges for four speci�c motor functions: abdominal
progression, sitting motility, reaching, and grasping.
Methods: The reliability part of the study included 38 assessments of term and preterm infants
in the age range of four to 18 months. The longitudinal prospective study included 30 term born
typically developing infants with nine assessments between three and 18 months. On the basis of
standardized video-recordings of spontaneous motor behaviour, the presence of adaptive motor
strategies was scored.
Results: Intra- and interobserver reliability were good. Clinically observable transitions to adaptive
selection started to emerge from six months onwards and peaked between eight and 15 months.
Transitions developed gradually and occurred at speci�c ages for di!erent motor functions.
Conclusion: Transition to adaptive motor behaviour can be observed reliably. Adaptive motor
behaviour develops gradually from six months onwards at function-speci�c ages. Comparison of
our results to literature showed that changes measured by neurophysiologic methods precede
clinically observed transitions.
53
4
Development of adaptive motor behaviour in typically developing infants
INTRODUCTION
During the �rst years of life, children show an impressive development of motor skills like sitting,
crawling, standing, walking, reaching and grasping. Rapidly, the child’s motor repertoire expands,
which enables the child to start exploration of the world. Motor development is a complex
process in which many factors play a role. The limited knowledge on processes governing motor
development induced a wide range of developmental theories. Theories di�er especially with
respect to the roles of ‘nature’ and ‘nurture’. Neuromaturationist theories for instance attribute
major part of development to endogenous, genetic factors, whereas environmental and contextual
factors dominate in the Dynamic Systems Theory1,2. The Neuronal Group Selection Theory (NGST)
is a theory which stresses the complex, continuous and cascadic interaction between information
from the genetic script and environment1,3-5.
According to the Neuronal Group Selection Theory (NGST)1,3-5, motor development is
characterised by two phases of variability. Development starts with the phase of primary variability
which can be observed during fetal life and early infancy. During this phase, all possibilities of the
innate neural networks are explored by means of self-generated activity and movements are neither
adapted to environmental constraints nor dependent on sensory feedback. In the fetus, newborn
and young infant this can be observed in General Movements: movements with great complexity
and variation in which all body parts are involved6,7. These general movements are present until
about four months of age, after which goal-directed motility gradually takes over. The emerging
goal-directed motility such as reaching, grasping and crawling, is characterised by large variability at
�rst. For each motor function all possible motor strategies and their corresponding neural networks
are explored. At function-speci�c ages, infants start to select more adaptive motor strategies.
Motor behaviour becomes more e�cient and adapted to the task requirements and environmental
characteristics. This phase of adaptive selection is called the phase of secondary variability. In this
phase, there is still abundant variability in motor behaviour, which serves the purpose of adaptation
of the movements to speci�c task and environmental characteristics.
Age of transition from primary to secondary variability di�ers for di�erent motor functions.
Sucking patterns, for example, are already adaptive around term age, i.e. at 37 to 40 weeks
gestational age8,9, while transition from primary to secondary variability for postural adjustments
during supported sitting occurs between three and eight months10-12. Reaching movements of
infants aged four months are characterised by large variability in movement trajectories. Around
the age of seven months, infants start to select straighter and more e�cient reaching movements,
indicating that for reaching the transition from primary to secondary variability lies approximately
between four and seven months1,13. These examples illustrate that secondary variability for basic
motor functions emerges during infancy. However, development of secondary variability continues
until adolescence. During this period, the child gains the capacity to �ne-tune motor output to
speci�c requirements and characteristics of task and environment.
54
Chapter 4
The above presented data on ages of transition from primary to secondary variability are based
on exact registration techniques such as electromyography (EMG) and kinematical recordings.
These techniques allow precise measurement of timing of changes in muscle patterns involved in
postural adjustments or straightening of reaching movement trajectories. No research has been
conducted on when transition from primary to secondary variability becomes clinically observable.
We hypothesize that changes measured by neurophysiologic methods precede clinically observable
transitions. Changes in motor behaviour recorded with exact registration techniques represent the
early phases of selection, while clinical observation shows when a certain pattern is selected during
the majority of movements. Knowledge on clinical observable transitions is of clinical relevance.
We recently demonstrated that the assessment of variation and the ability to select adaptive motor
strategies is a useful tool to assess neuromotor integrity (Infant Motor Pro�le14).
Aim of the study
The aim of this study was twofold. First, to investigate whether the presence of adaptive motor
behaviour can be observed reliably (intra- and interobserver reliability). Second, to explore the ages
at which clinically observable transition to adaptive motility emerges in typically developing infants
for four speci�c motor functions: abdominal progression, sitting motility, arm movements during
reaching and hand motility during grasping and manipulation. The four motor functions are part of
the IMP assessment14. The emergence of adaptive motor behaviour in the four motor functions was
studied as part of entire IMP assessments.
METHODS
The reliability part of the study included 38 cross-sectional assessments of term and preterm infants
in the age range of four to 18 months. The longitudinal prospective study included 30 term born
typically developing infants with nine assessments between three and 18 months. The study groups
are described in detail below. On the basis of standardized video-recordings of spontaneous motor
behaviour, the presence of adaptive motor strategies was assessed.
Participants in the reliability study
We studied whether the presence of adaptive motor behaviour can be observed reliably in 19 term
infants and 19 preterm infants. The term group in the reliability study consisted of 6 girls and 13
boys. Median gestational age was 40 weeks (range 38 – 42 weeks) and median birth weight was
3588 grams (range 3070 – 4200 grams). The preterm group consisted of 6 girls and 13 boys, median
gestational age was 29.9 weeks (range 27.1 – 32 weeks) and median birth weight was 1165 grams
(range 585 – 2120 grams). Three of the preterm infants had serious brain lesions on neonatal brain
ultrasound: two had signs of cystic periventricular leukomalacia15 and one had intraventricular
haemorrhage with ventricular dilatation (grade III)16. The preterm infants had been admitted to the
55
4
Development of adaptive motor behaviour in typically developing infants
neonatal intensive care unit (NICU) of the Beatrix Children’s Hospital of the University Medical Center
(UMC) in Groningen between December 2003 and January 2005. The infants were assessed at one of
the following ages: 4, 6, 10, 12 or 18 months. At 4, 6 and 10 months three full term and three preterm
infants were included, at 12 months four full term and four preterm infants and at 18 months six full
term and six preterm infants. The 38 videotapes were scored by two observers independently (KRH
and MHA) to assess interobserver agreement. To determine intra-observer agreement, the sample
was scored twice by one observer (KRH) with an interval of one month.
Participants in the study on emergence of adaptive motor behaviour
Thirty typically developing infants were included in the study on the emergence of adaptive motor
behaviour (12 girls and 18 boys).They were recruited at well baby clinics and amongst acquaintances.
They were born at term (gestational age at birth 40.1 weeks (median value), range 37.6 to 42 weeks)
without pre or perinatal complications with a median birth weight of 3588 grams (range 2730 to
4470 grams). The infants were followed longitudinally and assessments took place at ages 3, 4, 5,
6, 8, 10, 12, 15 and 18 months. Of 19 of these infants one assessment was included in the reliability
study as described above. Attrition rate was very low, only at 5 and 8 months one of the 30 infants
had no assessment, due to minor illness of the child or scheduling problems with the parents. At 6
months arm and hand motility during reaching and grasping could not be assessed for two of the
30 infants due to fatigue of the child. All parents signed informed consent and the research project
was approved by the Ethics Committee of the University Medical Center Groningen.
Procedures
Each assessment consisted of a video recording of in total 15 minutes of spontaneous motor
behaviour in supine, prone, sitting, standing and walking, depending on the child’s age and
functional capacities14. The order of the conditions was not !xed, but was adapted to the child’s
interest. Motor functions were only assessed if the infant was able to perform it independently,
e.g. showed abdominal progression, sat independently, showed successful reaching or showed
successful grasping. Table I shows the four items of the Infant Motor Pro!le and the de!nitions
used for determining whether motor behaviour can be classi!ed as adaptive motor behaviour for
the four speci!c motor functions (abdominal progression, sitting motility, arm movements during
reaching and hand motility during grasping and manipulation)14. Scoring is dichotomous: either the
motor behaviour is scored as ‘no selection’ or as ‘adaptive selection’ (Table I).
Abdominal progression was elicited with toys. Besides crawling, wriggling and pivoting
movements were also considered as forms of abdominal progression. Wriggling movements
are small, quick, twisting and turning movements of the body without speci!c use of arms and/
or legs resulting in spatial displacement, while the abdomen remains in contact with the support
surface17,18. Pivoting movements result in spatial displacement around the centre of the body, i.e.
around a vertical axis through the umbilicus. Older infants who showed crawling were encouraged
56
Chapter 4
to follow a ball or a toy car.
Sitting motility was only assessed if infants were able to sit independently. To observe sitting
motility, reaching movements and rotation movements of the trunk were elicited by presenting
toys at various distances, in various directions and at various heights, e.g. close to the trunk of the
infant, at arm length and a bit further at di�erent heights in front and to the sides of the sitting
infant. Sitting behaviour of children who were able to sit up independently was observed during
spontaneous motor activity while they were moving into various body positions.
To assess the infant’s ability to reach, grasp and manipulate objects, small toys were presented
to the infant seated on the parent’s lap. If the child was able to grasp objects in the midline at arm
Table 1: De�nitions of non-adaptive and adaptive motor behaviour for the four motor functions
Abdominal progression: ability to make an adaptive selection
No selection: The child does not select speci!c crawling strategies out of its motor repertoire. Thechild explores various crawling strategies with variation in movement patterns andmovement ranges. These variations in movements occur in a random way, whichresults in relatively ‘ine"cient’ motor behaviour. Wriggling and pivoting are alwaysassigned ‘no selection’, as these can be considered exploratory and variable precursormovements of crawling.
Adaptive selection: The child selects speci!c, e"cient crawling strategies out of its motor repertoire. Thechild is able to crawl e"ciently with appropriately coordinated movements of armsand legs, allowing for e.g. smooth crawling trajectories while turning a corner.
Sitting motility: ability to make an adaptive selection
No selection: The child does not select speci!c motor strategies out of its motor repertoire. Thechild explores various postural adjustment strategies, with variation in combination ofrecruited muscles, movement direction and movement velocity. Sitting motility is note"cient and there are multiple minor and major swaying movements of the upperbody in all directions with or without balance disturbance.
Adaptive selection: The child selects speci!c, e"cient strategies of postural adjustments out of its motorrepertoire. The child is able to adjust sitting motility e"ciently and does not loosebalance while reaching for objects at various distances and heights.
Reaching movements: ability to make an adaptive selection
No selection: The child does not select speci!c reaching strategies out of its motor repertoire. Thechild explores various reaching movements with variable movement trajectories,movement ranges and movement velocities. Movements are not adapted to distance,position and other characteristics of presented objects.
Adaptive selection: The child selects speci!c, e"cient reaching strategies out of its motor repertoire.Reaching movements are goal-directed with more or less straight and e"cientmovement trajectories.
Hand motility during grasping: ability to make an adaptive selection
No selection: The child does not select speci!c motor strategies out its motor repertoire. The childexplores various grasping and manipulation strategies. The variations in movementsoccur in a random way. Movements are not adapted to the size and shape of thepresented objects.
Adaptive selection: The child selects speci!c, e"cient grasping and manipulation strategies out of itsmotor repertoire. Grasping is adapted to the size and shape of the presented objects.
57
4
Development of adaptive motor behaviour in typically developing infants
length distance, toys were presented in di�erent positions to elicit di�erent reaching and grasping
movements. Type of grasping and variation in grasping were assessed by presenting objects of
various sizes and forms. A small cupboard with drawers with small handles was presented to the
older infants to observe �ne manipulation skills. Toys used were not standardised, but commercially
available small (approximately the size of the infant’s hand) puppets, animals, toy cars and balls (See
�gure 1).
Figure 1: Examples of toys used in the Infant Motor Pro�le assessment.
Statistical analyses
To determine intra- and interobserver reliability, Cohen’s kappa’s were calculated. Criteria of Landis
and Koch19 were used which state that κ > 0.80 is considered very good, 0.61< κ < 0.80 is good, 0.40
< κ < 0.60 is moderate and κ < 0.40 is poor. To explore the ages of transition, i.e., the age at which
motor behaviour changed from no selection to adaptive selection, the non-parametric Sign or
Friedman tests for related samples were used. Throughout the analyses, di�erences with a p-value <
0.05 were considered to be statistically signi�cant (two-tailed testing).
RESULTS
Reliability
For the item on adaptive selection during abdominal progression, intra- and interobserver
agreement were very good (See Table II for Kappa and 95% con�dence intervals). This was also the
case for the item on sitting motility. The item on arm motility during reaching showed good intra-
and interobserver agreement. For the item on hand motility, intra-observer reliability was moderate,
while interobserver agreement was very good.
58
Figure 2: Variability of abdominal progression: presence of adaptive selection.Hatched parts of bars represent percentage of infants showing no selection of crawling patterns and black partsrepresent percentage showing adaptive selection of crawling patterns. White parts represent the percentagesof infants showing no abdominal progression. ** Signi!cant increase in adaptive selection between 8 and 15months (Friedman: p<0.0005). The three developmental steps from 8 to 10, 10 to 12 and 12 to 15 months werealso statistically signi!cant (Sign test: p = 0.03, p = 0.002 and p = 0.001, respectively).
Chapter 4
Emergence of adaptive motor behaviour
In the development of abdominal progression, the transition from variable, non-adaptive behaviour
to adaptive behaviour was observed between 8 and 15 months (Friedman, p < 0.0005). This means
that at the age of 15 months virtually all children selected adaptive crawling strategies (Figure 2).
For sitting motility (Figure 3) transition from exploratory, variable behaviour without selection to
adaptive selection started from 6 months onwards. At 8 months 33% of infants showed adaptive
sitting motility, which increased to almost 90% of infants at 10 months (Friedman test 6 to 10
months: p<0.0005). Transition to adaptive selection of arm movements during reaching occurred
in a majority of infants between 6 and 8 months (Sign test: p < 0.0005). However, in the period from
Table II: Intra- and interobserver agreement
Intra-observer agreement Interobserver agreement
Item Kappa 95% CI Kappa 95% CI
Abdominal progression 0.84 0.69-0.99 0.84 0.69-0.99
Sitting motility 1.0 1.0-1.0 0.90 0.77-1.0
Reaching movements 0.68 0.42-0.94 0.61 0.32-0.89
Hand motility during grasping 0.56 0.23-0.88 0.87 0.71-1.0
CI: con!dence interval
59
Figure 4: Variability of arm movements during reaching: presence of adaptive selection.Hatched parts of bars represent percentage of infants showing no selection of arm movements and blackparts represent percentage of infants showing adaptive selection of arm movements. White parts representthe percentages of infants showing no reaching movements. Asterisks indicate signi�cant increase in adaptiveselection (Sign test): * p = 0.02 and ** p < 0.0005.
4
Figure 3: Variability of sitting motility: presence of adaptive selection.Hatched parts of bars represent percentage of infants showing no selection of sitting motility and black partsrepresent percentage showing adaptive selection of sitting motility. White parts represent the percentagesof infants that are not able to sit independently. * Signi�cant increase in adaptive selection between 6 and10 months (Friedman: p<0.0005). The two steps from 6 to 8 and from 8 to 10 months were also statisticallysigni�cant (Sign test 6 to 8 and 8 to 10 months: p = 0.02).
Development of adaptive motor behaviour in typically developing infants
60
8 to 12 months about 10% to 25% of infants continued to show variable reaching movements. Not
until 15 months, all infants showed adaptive reaching movements (Sign test: p = 0.02; Figure 4). For
hand motility during grasping transition from variable, non-adaptive to adaptive motor behaviour
was observed between 15 and 18 months (Sign test: p < 0.0005; Figure 5).
DISCUSSION
This study illustrated that transition to selection of adaptive motor behaviour can be reliably
observed and that this transition emerges around the age of 6 months and peaks between 8 and 15
months. The transition develops gradually and occurs at speci!c ages for di"erent motor functions.
Findings
Intra and interobserver reliability of all items were good to very good, except intra observer reliability
for the item on hand motility during grasping, which was moderate. The latter may have been an
expression of a learning e"ect.
For abdominal progression, clinically observable transition to selection of adaptive crawling
strategies started at 8 months of age. At 15 months, almost all children showed adaptive crawling
patterns on hands and knees. Adolph et al.20 also found decreasing movement variability with
increasing weeks of crawling experience. More experience with prone position and practice of
crawling accelerates selection of adaptive crawling patterns20-22. Development of crawling continues
Figure 5: Variability of hand motility during grasping: presence of adaptive selection.Hatched parts of bars represent percentage of infants showing no selection of hand motility and black partsrepresent percentage of infants showing adaptive selection of hand motility. White parts represent thepercentages of infants showing no successful grasping movements. ** Signi!cant increase in adaptive selection(Sign test, p < 0.0005).
Chapter 4
61
4
Development of adaptive motor behaviour in typically developing infants
beyond infancy23.
With respect to sitting motility, we know from literature that selection of the complete EMG
pattern, in which all dorsal neck- and trunk muscles are activated synchronously, emerges between
4 and 6 months11,12. The present study showed that transition to adaptive sitting motility is clinically
observable between 6 and 10 months of age. This indicates that selection of a preferred postural
adjustment pattern measured by EMG precedes the clinical observable transition. Development
of postural control is not !nished during infancy; it has a protracted course that lasts at least until
adolescence12.
Our study showed that clinically observable transition from variable, exploratory reaching
movements to straighter, adaptive reaching movements emerged between six and eight months,
but not until 15 months all typically developing children showed adaptive reaching movements.
Kinematic data from literature showed rapid straightening of movement trajectories around the age
of four to six months13, again preceding the clinical observed transition we found. Stable patterns of
temporal coordination start to develop from 12 to 15 months onwards until the age of three years23,
illustrating the long course of development of secondary variability during which !ne-tuning of the
reaching movements to the speci!c task requirements occurs.
For hand motility during grasping, transition to secondary variability occurred for more than
70 percent of children between 15 and 18 months. Corticospinal connections are crucial for the
development of skilled hand and !nger movements. Direct cortico-motoneuronal projections
already develop in the !rst months of postnatal life and allow for voluntary independent !nger
movements25,26. Thereafter, corticospinal projections become !ne-tuned, a process that includes the
disappearance of the majority of ipsilateral connections which is !nished by the age of 24 months27.
During the process of activity-dependent remodeling of the corticospinal connections, adaptive
hand motility emerges from 15 months onwards. But again, !ne-tuning of grasping movements
continues at least until the age of eight to ten years26.
Strengths and limitations of the study
A strength of this study is its longitudinal prospective character. For closer monitoring of
developmental changes, it would have been better to have more frequent assessments, for example
each month. However, time intervals between assessments were especially short during the !rst
year of life when motor development has a rapid course. Another strength was the very low attrition
rate. A limitation of the reliability study is that only two assessors participated. They were not blinded
with respect to term or preterm status of the infant. However, they were unaware of any details of
the child’s clinical history or results of neonatal ultrasounds of the preterm infants. A weakness of the
longitudinal part of the study is the relatively small sample size, especially because variability within
and between typically developing infants can be quite large. Nevertheless, we were able to detect
the ages of transition. In the current study, we compared the clinically observed ages of transition
with data in the literature available on changes measured by neurophysiologic methods. It would
62
Chapter 4
have been interesting and possibly more exact if the same children had had the neurophysiological
assessments and the clinical observations. This could be done in further research.
CONCLUDING REMARKS
During the !rst half year of life, infant motor behaviour is characterised by variation and exploration.
Transition to clinically observable adaptive motor behaviour develops gradually from six months
onwards at function-speci!c ages. By comparing our results to literature, we found that changes
measured by neurophysiologic methods precede the clinically observed transitions. Changes
in motor behaviour recorded with exact registration techniques represent the early phases of
selection, while clinical observation shows when a certain pattern is used during the majority of
movements. It would be interesting to perform further research on ages of transition in infants
with a high risk for developmental motor disorders such as cerebral palsy (CP) or developmental
coordination disorder (DCD). These children can have di"culties with processing a#erent, sensory
information and experience problems in !ne-tuning and adapting motor behaviour5,28-29. This could
possibly be re$ected in delayed or impeded transition to secondary variability. If so, age of transition
could be a useful clinical parameter of neuromotor condition.
ACKNOWLEDGEMENTSWe kindly acknowledge Arend Bos, MD, PhD for critical comments on a previous version of the manuscript. Wethank Nicole Arink and Dineke Dijkhuizen for performing part of the assessments. The study was supported bya Junior Scienti!c Masterclass grant of the post-graduate school Behavioral and Cognitive Neurosciences (BCN),University of Groningen.
63
4
Development of adaptive motor behaviour in typically developing infants
REFERENCES1. Hadders-Algra M. The Neuronal Group Selection Theory: a framework to explain variation in normal motor
development. Dev Med Child Neurol 2000;42:566-572.
2. Thelen E. Motor development. A new synthesis. American Psychologist 1995;50:79–95.
3. Sporns O, Edelman GM. Solving Bernstein’s problem: a proposal for the development of coordinatedmovement by selection. Child Dev 1993;64:960-981.
4. Edelman GM. Neural Darwinism: selection and reentrant signaling in higher brain function. Neuron1993;10:115-125.
5. Hadders-Algra M. The Neuronal Group Selection Theory: promising principles for understanding andtreating developmental motor disorders. Dev Med Child Neurol 2000;42:707-715.
6. Hadders-Algra M. General movements: A window for early identi�cation of children at high risk fordevelopmental disorders. J Pediatr 2004;145:S12-18.
7. Einspieler C, Prechtl HFR, Bos AF, et al. Prechtl’s method on the qualitative assessment of GeneralMovements in preterm, term and young infants. Clinics in Developmental Medicine No. 167. London: MacKeith Press; 2004.
8. Wol� PH. The serial organization of sucking in the young infant. Pediatrics 1968;42:943-956.
9. Bu’Lock F, Woolridge MW, Baum JD. Development of co-ordination of sucking, swallowing and breathing:ultrasound study of term and preterm infants. Dev Med Child Neurol 1990;32:669-678.
10. Hedberg A, Carlberg EB, Forssberg H, Hadders-Algra M. Development of postural adjustments in sittingposition during the �rst half year of life. Dev Med Child Neurol 2005;47:312-320.
11. De Graaf-Peters VB, Bakker H, van Eykern LA, Otten B, Hadders-Algra M. Postural adjustments and reachingin 4- and 6-month-old infants: an EMG and kinematical study. Exp Brain Res 2007;181:647-56.
12. Hadders-Algra M. Development of postural control. In: M. Hadders-Algra and E. Brogren Carlberg, editors.Postural control: a key issue in developmental disorders. Clin Dev Med No. 179. London: Mac Keith Press,2008, pp 22-73.
13. Konczak J, Borutta M, Topka H, Dichgans J. The development of goal-directed reaching in infants: handtrajectory formation and joint torque control. Exp Brain Res 1995;106:156-68.
14. Heineman KR, Bos AF, Hadders-Algra M. The Infant Motor Pro�le: a standardized and qualitative method toassess motor behaviour in infancy. Dev Med Child Neurol 2008;50:275-82.
15. De Vries LS, Eken P, Dubowitz LM. The spectrum of leukomalacia using cranial ultrasound. Behav Brain Res1992;31;49:1-6.
16. Volpe JJ. Neurology of the Newborn. 4th ed. Philadelphia PA: WB Saunders, 2001.
17. Touwen BCL. Neurological development in infancy. Clin Dev Med No. 58. Philadelphia: Lippincott, 1976.
18. Piper MC, Darrah J. Motor assessment of the developing infant. Philadelphia: WB Saunders Compan, 1994.
19. Landis JR, Koch GG. The measurement of observer agreement for categorical data. Biometrics 1977;33:159-174.
20. Adolph KE, Vereijken B, Denny MA. Learning to crawl. Child Dev 1998;69:1299-1312.
21. Davis BE, Moon RY, Sachs HC, Ottolini MC. E�ects of sleep position on infant motor development. Pediatrics1998;102:1135-1140.
22. Lagerspetz K, Nygard M, Strandvik C. The e�ects of training in crawling on the motor and mentaldevelopment of infants. Scand J Psychol 1971;12:192-197.
23. Touwen BC, Hempel MS, Westra LC. The development of crawling between 18 months and four years. DevMed Child Neurol 1992;34:410-416.
24. Konczak J, Dichgans J. The development toward stereotypic arm kinematics during reaching in the �rst 3years of life. Exp Brain Res 1997;117:346-354.
64
Chapter 4
25. Armand J, Olivier E, Edgley SA, Lemon RN. Postnatal development of corticospinal projections from motorcortex to the cervical enlargement in the macaque monkey. J Neurosci 1997;1:251-266.
26. Forssberg H. Neural control of human motor development. Curr Opin Neurobiol 1999;9:676-682.
27. Eyre JA. Development and plasticity of the corticospinal system in man. Neural Plast 2003;10:93-106.
28. Wilson PH, McKenzie BE. Information processing de�cits associated with developmental coordinationdisorder: a meta-analysis of research �ndings. J Child Psychol Psychiatry 1998;39:829-840.
29. Cooper J, Majnemer A, Rosenblatt B, Birnbaum R. The determination of sensory de�cits in children withhemiplegic cerebral palsy. J Child Neurol 1995;10:300-309.
65
5CHAPTER
Construct validity of the Infant Motor Pro�le:Relation with prenatal, perinatal and neonatal risk factors
Kirsten Heineman1,2, Sacha la Bastide-van Gemert3, Vaclav Fidler3,Karin Middelburg1, Arend Bos4, Mijna Hadders-Algra1
1Department of Paediatrics, Institute of Developmental Neurology2Department of Neurology
3Department of Epidemiology4Department of Paediatrics, Division of Neonatology
University Medical Center Groningen
Developmental Medicine and Child Neurology 2010; Epub ahead of print
66
Chapter 5
ABSTRACT
Aim: The Infant Motor Pro�le (IMP) is a qualitative assessment of motor behaviour of infants aged
3 to 18 months. The aim of this study was to investigate construct validity of the IMP, through
the relation of IMP scores with pre-, peri- and neonatal variables, including the presence of brain
pathology indicated by neonatal ultrasound imaging of the brain.
Methods: A longitudinal prospective study was performed in a group of 30 term born (12 females
and 18 males; median gestational age 40.1wks, range 37.6-42wks) and 59 preterm infants (25
females and 34 males; median gestational age 29.7wks, range 25-34.7wks). IMP assessments were
performed at (corrected) ages of 4, 6, 10, 12 and 18 months. Socio-economic and perinatal data
were collected, which, in the case of preterm infants, included information on periventricular
leukomalacia and intraventricular haemorrhage based on neonatal cranial ultrasounds. Data were
analysed by �tting mixed-e!ects models.
Results: Gestational age, socio-economic status and 5-minute Apgar score were signi�cant
determinants of IMP scores in the total group of infants (p-values resp. <0.001, 0.002 and 0.042
respectively). In the subgroup of preterms, IMP scores were signi�cantly a!ected by brain lesions on
neonatal ultrasound (p<0.001) and socio-economic status (p=0.001).
Interpretation: The �ndings support the construct validity of the Infant Motor Pro�le: IMP scores
are clearly associated with relevant determinants of neuromotor function.
67
5
Construct validity of the Infant Motor Pro�le
INTRODUCTION
The Infant Motor Pro�le (IMP)1 is a recently developed qualitative assessment of motor behaviour
of infants aged 3 to 18 months. It may be used for early detection and developmental evaluation of
infants with a high risk for developmental motor disorders such as cerebral palsy or developmental
coordination disorder. Infants born very preterm are especially at risk for these developmental
motor disorders2-5. Owing to improved prenatal and neonatal care, the survival of very preterm
infants has increased in the last decennia2,3. Follow-up of preterm infants aims at detection of those
infants that could bene�t from early intervention at young age, when the brain is highly plastic6,7
The IMP is based on ideas of the Neuronal Group Selection Theory (NGST) of motor
development8-10. According to NGST, typical motor development starts with the phase of primary
variability with exploratory, variable motor behaviour. Children with pre- or perinatally acquired
brain damage show more stereotyped motor behaviour with considerably less variation. During
development, infants learn to select adaptive motor strategies out of their motor repertoire and
to adapt motor behaviour to the environment. This phase of adaptive selection is called secondary
variability. Children with developmental motor disorders often have problems in selecting
adaptive motor strategies9,10. Two domains of the IMP are based on the NGST-principles of motor
development; they deal with variation of motor behaviour (size of repertoire) and the ability to
select motor strategies (variability). Three additional domains assess movement �uency, movement
symmetry and motor performance.
The IMP assessment consists of video-recording of approximately 15 minutes of spontaneous
motor behaviour in supine, prone, sitting, standing, and walking condition, depending on age and
functional capacities of the infant. In addition, reaching, grasping and manipulation of objects are
tested in the supine and sitting positions. The 80 items of the IMP are scored based on the video
recording1. The 80 items constitute the subscores on the �ve domains: size of repertoire (variation),
ability to select (variability), movement �uency, movement symmetry and motor performance. The
mean of the �ve subscores is the total IMP score1 (Appendix I).
In our pilot study on the IMP1, we described the theoretical background of the method, its
domains and items, and details of scoring procedures. In addition, we have published �rst data on
the reliability and concurrent validity of the IMP with the Alberta Infant Motor Scale (AIMS)11 and
Touwen neurological examination12. Intra- and interobserver reliability for total IMP score were good
(Spearman’s rho 0.9 for both; 95% con�dence intervals 0.8–0.9 and 0.8-1.0 respectively). Concurrent
validity of the IMP with the AIMS was very high for the performance subscale and moderate for the
total IMP score. Concurrent validity of the IMP with the Touwen examination was very good, with a
clear inverse relationship between IMP scores and the degree of neurological dysfunction1.
The present study focuses on construct validity. Construct validity is the extent to which items
of the instrument re�ect the theoretical construct of interest, in this case neuromotor function13.
Parameters of construct validity with respect to neuromotor function are the relation of test scores
68
Chapter 5
with prenatal, perinatal and neonatal adversities and correlation with results from brain imaging. The
aim of the present study was to investigate construct validity of the IMP, through the relation of IMP
scores with prenatal, perinatal and neonatal variables, including the presence of brain pathology
indicated by neonatal ultrasound imaging of the brain. To this end, we performed a longitudinal
prospective study in a group of term and preterm infants from the age of 4 to 18 months.
METHODS
Participants
We included a longitudinal study group of 30 term and 59 preterm infants. The term infants
(12 females and 18 males) were recruited from amongst colleagues and acquaintances of the
researchers. The preterm infants had been admitted to the neonatal intensive care unit of the Beatrix
Children’s Hospital of the University Medical Center in Groningen between December 2003 and
January 2005. Inclusion criteria were: gestational age below 35 weeks, singleton or twin, parents
with appropriate understanding of the Dutch language, and travel distance between the child’s
home and the hospital of < 1 hour. Infants with severe congenital anomalies were excluded from
the study. During the time interval mentioned above, 148 infants were eligible for inclusion. Owing
to the limited capacity of our department, only a limited number of infants could be included in the
intensive follow-up scheme each month. Therefore, not all of the eligible infants were approached
for inclusion. By approaching parents at random, 59 infants were eventually included. The project
was approved by the local ethics committee. All parents of the infants gave informed consent.
Procedures
Longitudinal assessments were performed at (corrected) ages of 4, 6, 10, 12 and 18 months.
Assessments consisted of a video-recording of approximately 15 minutes of spontaneous motor
behaviour. Motor behaviour was recorded while supine, prone, sitting, standing, and walking,
depending on the age and functional capacity of the infant. Reaching, grasping and manipulation
of objects were evaluated in supine and in (supported) sitting positions. The IMP assessments were
carried out by one of the examiners (KRH or KJM). The examiners knew whether an infant was born
at term or preterm, but were not aware of any of the perinatal and neonatal details. The 80 items of
the IMP were scored on the basis of the video recording. In this paper, only the total IMP scores are
addressed.
Socio-economic, perinatal and neonatal data were collected for all infants on standardized
forms by means of an interview with the parents and by consultating neonatal intensive care unit
discharge certi!cates. Socio-economic status (SES) was de!ned by the sum score of four variables
describing educational and professional level of father and mother, all expressed on a scale from 0
(lowest) through 2 (highest). Neonatal ultrasounds of the brain of the preterm infants were assessed
with respect to periventricular leukomalacia (PVL) and intraventricular haemorrhage (IVH). PVL was
69
5
Construct validity of the Infant Motor Pro�le
classi�ed according to De Vries et al.14: grade I; transient periventricular echodensities persisting
for more than 7 days; grade II; transient periventricular densities, evolving into small localised
frontoparietal cysts; grade III; periventricular densities, evolving into extensive periventricular cystic
lesions; and grade IV; densities extending into the deep white matter evolving into extensive cystic
lesions. IVH was classi�ed according to Volpe15:grade I; haemorrhage con�ned to the subependymal
germinal matrix; grade II; haemorrhage into the lateral ventricles without ventricular dilatation;
grade III; IVH with ventricular dilatation; and grade IV; IVH with parenchymal involvement (venous
infarction). PVL grade II or worse and/or IVH grade III or worse were classi�ed as serious brain lesions.
‘Small for gestational age’ was de�ned as birthweight compared to gestational age below the 10th
percentile16.
Statistical analyses
For comparison of background characteristics of term and preterm groups we used the Mann-
Whitney U test or the Pearson chi-squared test. Analyses to test the association between risk variables
and IMP scores at di!erent ages were performed with the Mann-Whitney U test for categorical
variables and Spearman’s rho correlation for continuous variables. Variables that correlated with
IMP scores with a Spearman’s rho correlation of 0.20 or more at more than one of the �ve assessed
ages, or a Mann-Whitney U test with p-value of 0.05 at one or more of the �ve assessed ages
were entered in a mixed-e!ects analysis17. This analysis makes it possible to describe the change of
the IMP scores over time and in relation to perinatal and neonatal variables. It takes into account
the dependence between observations from the same infant, and it can also take into account
dependence between twins. Children for whom outcomes at some time-points were missing were
still included in the analysis by entering non-missing data at these time-points; the analysis assumes
that missing values do not a!ect the outcome (‘missing at random’ assumption18). The results of
neonatal sonography of the brain were investigated in a separate analysis, as these measurements
were available only in the preterm group. In selecting the best-�tting models, we used 0.05 as the
nominal level of signi�cance.
RESULTS
Table I shows socio-economic and neonatal characteristics of the term and preterm infants. Thirty-
�ve of the preterm infants were singletons and 24 were twins. Nine pairs of twins participated in the
study; the remaining six had lost their twin sibling. The attrition rate was low: in the preterm group
two children missed assessment at two of the �ve time-points and 11 children at one time-point.
Reasons for missed assessments were scheduling problems with the parents or minor illness of the
infant. In the term group no assessment was missed. The preterm infants generally had a higher
chance to be born small for gestational age, to be delivered by Caesarean section, more signs of fetal
distress at birth, and a lower 5-minute Apgar score than the infants born at term. Preterm and term
infants did not signi�cantly di!er in SES and maternal age.
70
Figure 1: Individual Infant Motor Pro!le curves in the term (left) and preterm (right) groups.IMP scores in the preterm group are, in general, lower and show a larger within-participant variability than in theterm group (p<0.001).
Chapter 5
Associations in the total group
Throughout infancy, preterm infants had signi�cantly lower IMP scores than term infants (Mann-
Whitney U; p<0.001). No signi�cant di�erence in IMP scores was found between males and females.
Figure 1 depicts the individual IMP curves in the term and preterm groups. Univariate analyses
showed a signi�cant relation between IMP and SES, gestational age, signs of fetal distress, Ceasarian
Table I: Socio-economic and neonatal characteristics of term and preterm groups
Term Preterm p-value
Number of children 30 59
Males/Females, n 18/12 34/15 0.831
Maternal age (years), mean (SD) 32.8 (4.1) 31.8 (5.2) 0.514
SESa, median (range) 5 (0-8) 4 (0-8) 0.265
Twins, n (%) 0 24 (41) <0.001
Gestational age wks, median (range) 40.1 (37.6-42) 29.7 (25-34.7) <0.001
Birthweight g, median (range) 3588 (2730-4470) 1285 (630-2180) <0.001
Small for gestational ageb, n (%) 2 (6.7) 20 (33.7) 0.005
Caesarian section, n (%) 3 (10) 33 (55.9) <0.001
Signs of fetal distressc, n (%) 5 (16.7) 30 (50.8) 0.002
Apgar score at 5 min,median (range)
10 (9-10) 9 (4-10) <0.001
a SES = socio-economic status, sum score of four variables describing educational and professional level of fatherand mother, all expressed on a scale from 0 (lowest) through 2 (highest).b Small for gestational age is de�ned as birthweight compared with gestational age below 10th centile16.c Presence of at least one of the following factors: meconium staining, cardiotocography abnormalities, acidaemiaduring delivery (arterial umbilical pH below 7.05).
! " #$ #% # #! #"
!$
&$
"$
'$
#$$
()* ,-./0123
4.0567892:.;*
!"# %&'(&)*
! " #$ #% # #! #"
!$
&$
"$
'$
#$$
()* ,-./0123
4.0567892:.;*
+"!)!"# %&'(&)*
71
5
Construct validity of the Infant Motor Pro�le
section, and 5-minute Apgar score at nearly all �ve time-points. These variables were entered into
the mixed-e�ects model. The data were well described by a model (see Table II) which included,
as �xed e�ects, a quadratic function of age (re�ecting a substantial increase in IMP scores with
increasing age), gestational age, SES and 5-minute Apgar score. A nearly equally well-�tting model
was obtained by replacing gestational age with birthweight or with preterm status. The best-�tting
model also included a random intercept, a random linear age e�ect, and a within-participant
variance that was larger in the preterm group than in the term group (p<0.001). We also �tted a
mixed model including an additional random factor for family to account for dependence of twins.
The results obtained with this multilevel model were rather similar to those obtained without taking
twin status into account. We have decided to present the simpler model that does not take twin
status into account.
Associations in the subgroup of preterm infants
Neonatal ultrasound was available for 57 of the 59 preterm infants. Serious brain pathology was
observed in six infants: one infant had cystic PVL, four infants had IVH grade IV, and one infant had
Table II: Regression analysis of Infant Motor Pro�le scores (term and preterm infants, n = 76a)
Regression coe!cient Standard error p-value 95%-con"dence interval
Intercept 47.6 2.22
Age (mo) 2.16 0.165 <0.001
Age2 (x10-2) -5.70 0.734 <0.001
Gestational age (wks) 0.478 0.068 <0.001 0.34 – 0.61
SESb 0.427 0.131 0.002 0.16 – 0.68
Apgar 5 min 0.535 0.258 0.042 0.02 – 1.05
a For 12 term infants 5-minute Apgar score was missing and for one infant no complete data were available onSES.b SES = socio-economic status, sum score of four variables describing educational and professional level of fatherand mother, all expressed on a scale from 0 (lowest) through 2 (highest).
Table III: Regression analysis of Infant Motor Pro�le scores in the preterm group (n = 56a)
Regression coe!cient Standard error p-value 95%-con"dence interval
Intercept 64.0 1.36
Age (mo) 2.52 0.223 <0.001
Age2 (x10-2) -7.22 0.981 <0.001
Brain lesions -6.33 1.28 <0.001 -8.9 – -3.8
SESb 0.588 0.173 0.001 0.24 – 0.93
a Two preterm infants did not undergo neonatal ultrasonography, and for one infant no complete data wereavailable on SES.b SES = socio-economic status, sum score of four variables describing educational and professional level of fatherand mother, all expressed on a scale from 0 (lowest) through 2 (highest).
72
Chapter 5
middle cerebral artery infarction on the right side. Infants with serious brain lesions had signi�cantly
lower IMP scores at ages 4, 6, 10, and 18 months than infants without serious abnormalities on
neonatal ultrasound (Mann-Whitney U test: p=0.002, 0.003, 0.01 and 0.004 respectively). The mixed-
e�ects analysis revealed that lower IMP scores in the preterm group were associated with serious
brain lesions on ultrasound and with SES (Table III).
DISCUSSION
Our study revealed clear associations between prenatal, perinatal and neonatal adversities and
IMP scores at ages 4 to 18 months in term and preterm infants. In preterm infants serious brain
pathology on neonatal ultrasound was strongly related to lower IMP scores. These �ndings support
the construct validity of the IMP.
The strengths of the study are its longitudinal character and the low attrition rate. The study
group was explicitly heterogeneous with respect to neuromotor function, which is valuable in
assessing validity of a new instrument. The weaknesses of the study are the relatively small sample
size of the term group and the missing Apgar scores for 12 term children. Therefore, we recommend
that the study be replicated with a larger sample of term infants. A further limitation of the study
is that the assessors were not blind with respect to term or preterm status of the infant. However,
the assessors were unaware of any details of the child’s clinical history or results of neonatal
ultrasonography.
To investigate whether the group of preterm infants included in this study was representative
for a Dutch neonatal intensive care unit population in a tertiary referral centre, we compared it with
established reference groups of preterm infants19,20. We found that gestational age, birthweight,
frequency of Apgar score at 5 minutes below 7 (15%) and the need for ventilatory support (68%)
were similar. Our sample showed a relatively high percentage of infants small for gestational age
and of Caesarean deliveries.19,20. The latter could be a result of the increased tendency over the years
to deliver very preterm infants by Caesarean section21.
The IMP scores were most strongly a�ected by preterm birth. Preterm birth is a major risk factor
for major and minor neurodevelopmental disabilities3-5. Both types of disability may largely interfere
with daily activities and academic achievements. The clear in!uence of this important risk factor on
IMP scores contributes to construct validity of the IMP. In the total group of infants, a low 5-minute
Apgar score was associated with low IMP scores. In fact, the Apgar score was never intended to
be used to predict neurodevelopmental outcome, but mainly the risk of neonatal death22. The
Apgar score re!ects the condition of the infant shortly after delivery, and could be interpreted as
an indicator of the infant’s ability to adapt to the novel situation of extrauterine life. In the preterm
group, the presence of a serious brain lesion had a greater e�ect than the Apgar score on later IMP
performance.
Preterm infants showed more individual variations in IMP scores over time than term infants. The
73
5
Construct validity of the Infant Motor Pro�le
more variable developmental trajectories in preterm infants presumably re!ect that developmental
progress in infants with atypical brain function is less stable than in infants with typical brain
function23. In other words, atypical brain function is expressed di"erently at di"erent ages. The
inconsistent expression of atypical motor behaviour emphasizes the need of repeated assessments
in high-risk infants.
Serious brain lesions on ultrasound, such as cystic PVL or IVH, were associated with lower
IMP scores throughout infancy. According to NGST, brain lesions lead to a reduction of available
neuronal networks, which results, in particular, in reduced variation of motor behaviour and
an impaired ability to select adaptive motor behaviour. The presence of serious brain lesions on
neonatal ultrasound is a strong predictor of later neurodevelopmental disability24.
We found a clear increase in IMP scores with increasing age, which implies that the IMP is able to
detect and re!ect age-related changes in motor development (i.e., higher levels of performance and
better abilities to select adaptive motor behavior). This contributes considerably to the construct
validity of an instrument for assessment of motor development25.
In our study, higher SES was associated with higher IMP scores in the group of infants as a
whole and in the preterm group separately. No relation between maternal education and motor
development in infancy was found in a study on healthy, term born infants26, but lower SES was
clearly associated with the presence of the complex form of minor neurological dysfunction at
school age27. Highly educated parents probably provide both favourable genetic factors and a
healthy and stimulating environment for the child. The latter is especially important for children at
high biological risk of developmental disorders, such as preterm infants28.
In conclusion, this study supports the construct validity of the IMP. IMP scores were clearly
related to relevant risk factors for developmental motor problems. These #ndings support the use
of the IMP in clinical follow-up of high risk infants.
ACKNOWLEDGEMENTSWe thank Nicole Arink, Hylco Bouwstra, Dineke Dijkhuizen and Anne Hoekstra for performing part of theassessments. The study was supported by a Junior Scienti#c Masterclass grant of the post-graduate schoolBehavioral and Cognitive Neurosciences (BCN), University of Groningen.
74
Chapter 5
REFERENCES1. Heineman KR, Bos AF, Hadders-Algra M. The Infant Motor Pro�le: a standardized and qualitative method to
assess motor behaviour in infancy. Dev Med Child Neurol 2008;50:275-82.
2. Wilson-Costello D, Friedman H, Minich N, Fanaro� AA, Hack M.. Improved survival rates with increasedneurodevelopmental disability for extremely low birth weight infants in the 1990s. Pediatrics 2005;115:997-1003.
3. Larroque B, Ancel PY, Marret S, Marchand L, André M, Arnaud C, Pierrat V, Rozé JC, Messer J, Thiriez G,Burguet A, Picaud JC, Bréart G, Kaminski M; EPIPAGE Study group. Neurodevelopmental disabilities andspecial care of 5-year-old children born before 33 weeks of gestation (the EPIPAGE study): a longitudinalcohort study. Lancet. 2008;371:813-20.
4. Marlow N, Wolke D, Bracewell MA, Samara M; EPICure Study Group. Neurologic and developmentaldisability at six years of age after extremely preterm birth. N Engl J Med. 2005;352:9-19.
5. Davis NM, Ford GW, Anderson PJ, Doyle LW; Victorian Infant Collaborative Study Group. Developmentalcoordination disorder at 8 years of age in a regional cohort of extremely-low-birthweight or very preterminfants. Dev Med Child Neurol 2007;49:325-30
6. Kolb B, Brown R, Witt-Lajeunesse A, Gibb R. Neural compensations after lesion of the cerebral cortex.Neural Plast 2001;8:1-16.
7. De Graaf-Peters VB, Hadders-Algra M. Ontogeny of the human central nervous system: what is happeningwhen? Early Hum Dev 2006;82:257-266.
8. Sporns O, Edelman GM. Solving Bernstein’s problem: a proposal for the development of coordinatedmovement by selection. Child Dev 1993;64:960-981.
9. Hadders-Algra M. The Neuronal Group Selection Theory: a framework to explain variation in normal motordevelopment. Dev Med Child Neurol 2000;42:566-572.
10. Hadders-Algra M. The Neuronal Group Selection Theory: promising principles for understanding andtreating developmental motor disorders. Dev Med Child Neurol 2000;42:707-715.
11. Piper MC, Darrah J. Motor assessment of the developing infant. Philadelphia: Saunders, 1994.
12. Touwen BCL. Neurological development in infancy. Clin Dev Med No. 58. Philadelphia: Lippincott, 1976.
13. Heineman KR, Hadders-Algra M. Evaluation of neuromotor function in infancy - A systematic review ofavailable methods. J Dev Behav Pediatr 2008;29:315-23.
14. De Vries LS, Eken P, Dubowitz LM. The spectrum of leukomalacia using cranial ultrasound. Behav Brain Res1992;31;49:1-6.
15. Volpe JJ. Neurology of the Newborn. 4th ed. Philadelphia PA: WB Saunders, 2001.
16. Kloosterman GJ. On intrauterine growth: the signi�cance of prenatal care. Int J Gynaecol Obstet 1970;8:895-912.
17. Pinheiro JS, Bates DM. Mixed-E�ects Models in S and S-plus. New York: Springer, 2000.
18. Fitzmaurice GM, Laird NM, Ware JH. Applied longitudinal analysis. Hoboken, New Jersey: John Wiley &Sons, Inc., 2004.
19. Stoelhorst GM, Rijken M, Martens SE, Brand R, den Ouden AL, Wit JM, Veen S; Leiden Follow-Up Project onPrematurity. Changes in neonatology: comparison of two cohorts of very preterm infants (gestational age<32 weeks): the Project On Preterm and Small for Gestational Age Infants 1983 and the Leiden Follow-UpProject on Prematurity 1996-1997. Pediatrics 2005;115:396-405.
20. de Kleine MJ, den Ouden AL, Kollée LA, van Baar A, Nijhuis-van der Sanden MW, Ilsen A, Brand R, Verloove-Vanhorick SP. Outcome of perinatal care for very preterm infants at 5 years of age: a comparison between1983 and 1993. Paediatr Perinat Epidemiol 2007;21:26-33.
21. Haque KN, Hayes AM, Ahmed Z, Wilde R, Fong CY. Caesarean or vaginal delivery for preterm very-low-birth weight (< or =1,250 g) infant: experience from a district general hospital in UK. Arch Gynecol Obstet.
75
5
Construct validity of the Infant Motor Pro�le
2008;277:207-12.
22. Casey BM, McIntire DD, Leveno KJ. The continuing value of the Apgar score for the assessment of newborninfants. N Engl J Med 2001;15;344:467-471.
23. Hadders-Algra M, Heineman KR, Bos AF, Middelburg KJ. The assessment of minor neurological dysfunctionin infancy using the Touwen Infant Neurological Examination: strengths and limitations. Dev Med ChildNeurol, 2009, e-pub ahead of print.
24. Weisglas-Kuperus N, Baerts W, Fetter WP, Sauer PJ. Neonatal cerebral ultrasound, neonatal neurology andperinatal conditions as predictors of neurodevelopmental outcome in very low birthweight infants. EarlyHum Dev 1992;31:131-148.
25. Campbell SK, Kolobe TH, Osten ET, Lenke M, Girolami GL. Construct validity of the test of infant motorperformance. Phys Ther. 1995;75:585-96.
26. Ravenscroft EF, Harris SR. Is maternal education related to infant motor development? Pediatr Phys Ther2007;19:56-61.
27. Hadders-Algra M, Huisjes HJ, Touwen BC. Perinatal correlates of major and minor neurological dysfunctionat school age: a multivariate analysis. Dev Med Child Neurol 1988;30:472-481.
28. Weisglas-Kuperus N, Baerts W, Smrkovsky M, Sauer PJ. E�ects of biological and social factors on thecognitive development of very low birth weight children. Pediatrics 1993;92:658-665.
76
77
6CHAPTER
Concurrent and predictive validity of theInfant Motor Pro�le
Kirsten Heineman1,2, Karin Middelburg1, Arend Bos3,Lieke Eidhof1, Mijna Hadders-Algra1
1Department of Paediatrics, Institute of Developmental Neurology2Department of Neurology
3Department of Paediatrics, Division of NeonatologyUniversity Medical Center Groningen
Submitted
78
Chapter 6
ABSTRACT
Background: The Infant Motor Pro�le (IMP) is a qualitative instrument to assess motor behaviour
of infants aged 3 to 18 months. The IMP consists of �ve domains: size of motor repertoire, ability to
select motor strategies (variability), �uency, symmetry and motor performance.
Objective: To assess inter-observer reliability, concurrent validity of the IMP with the Alberta Infant
Motor Scale and the Touwen Infant Neurological Examination, and predictive validity of the IMP for
neurological outcome at 18 months.
Design: A longitudinal prospective study was performed in a group of 30 term born and 59 preterm
infants. For the concurrent validity part of the study, a second group of term infants was added with
cross-sectional assessments.
Methods: Assessments were performed at corrected ages of 4,6,10,12 and 18 months and consisted
of the IMP, AIMS and neurological assessment. Socio-economic and perinatal data were collected.
Non-parametric statistics were used to analyze the data.
Results: Inter-observer reliability was high (intra-class coe!cient 0.95). Correlations between IMP
and AIMS scores varied across IMP domains; they were highest for the performance domain of
the IMP (Spearman’s rho 0.47-0.84). A clear relationship was found between total IMP score and
neurological condition (Kruskal-Wallis p < 0.001). Sensitivity for prediction of abnormal neurological
outcome at 18 months was 86 to 100% with speci�city ranging from 71 to 78%.
Limitations: Assessors were not blinded with respect to term or preterm status of the infants.
Follow-up did not extend beyond the age of 18 months.
Conclusions: Reliability of the IMP is good and concurrent and predictive validity are satisfactory.
These �ndings support the notion that the IMP is a promising and valuable instrument to assess
motor behaviour in infancy.
79
Concurrent and predictive validity of the Infant Motor Pro�le
6
INTRODUCTION
Prediction of neurological outcome in infants with a high risk for developmental motor disorders,
such as cerebral palsy (CP) or developmental coordination disorder (DCD), is di�cult. It appears
that instruments that assess qualitative aspects of motor behaviour, such as the General Movement
method (GM1,2) and the Test of Infant Motor Performance (TIMP3) are most promising as single
clinical neuromotor predictors4. However, the GM method and TIMP are only applicable until the
age of four months. Therefore, we developed the Infant Motor Pro!le5, a qualitative assessment of
motor behaviour that is applicable throughout infancy until the age of 18 months.
The IMP was developed for three purposes: !rst it may be used to detect infants with a high
risk for developmental motor disorders, such as CP or DCD. Infants born very preterm are especially
at risk for these developmental motor disorders6. Early detection of high-risk infants is important
to provide early intervention at young age when plasticity of the brain is still high7,8. Second, the
IMP may be used for evaluation of changes in neuromotor function, e.g. during or after early
intervention. The third aim of the IMP is prediction of future developmental outcome.
The IMP is based on ideas of the Neuronal Group Selection Theory (NGST) on motor
development9,10,11. According to NGST, typical motor development starts with the phase of
primary variability with exploratory, variable motor behaviour. Children with pre- or perinatally
acquired brain damage show more stereotyped motor behaviour with considerably less variation.
During development, infants learn to select adaptive motor strategies out of their primary motor
repertoire and to adapt motor behaviour to the environment. This phase of adaptive selection is
called secondary variability. Children with developmental motor disorders often have problems
in selecting adaptive motor strategies10,11. Two domains of the IMP are based on these principles
of motor development; they assess variation of motor behaviour and the ability to select motor
strategies. Three additional domains assess movement "uency, movement symmetry and motor
performance.
Two types of validity that are important in the development and validation of a new instrument
are concurrent and predictive validity. Concurrent validity is the extent to which scores on the
new instrument relate to scores on another measure of the same theoretical construct, ideally a
‘gold standard’. However, no gold standard for assessment of neuromotor function in infancy is
available. Therefore, concurrent validity of the new instrument with other established instruments is
assessed. Predictive validity is de!ned as the extent to which current scores on the new instrument
predict future developmental outcome. A distinction can be made between prediction of major
developmental disorders, such as cerebral palsy, and prediction of minor developmental motor
problems, such as minor neurological dysfunction and developmental coordination disorder.
In our pilot study5, we described the Infant Motor Pro!le and its theoretical background, its
domains and items, and details on scoring procedures. In addition, !rst data on reliability and some
data on validity were presented. Intra and inter observer reliability of scoring were satisfactory in the
80
Chapter 6
persons who developed the IMP. Concurrent validity of the IMP with the Alberta Infant Motor Scale
(AIMS12) and the Touwen Infant Neurological Examination (TINE13,14) was assessed in a relatively
small sample of infants.
Aim of the present study is threefold: �rst to examine inter-observer reliability of a newly
trained assessor without prior experience with the IMP. The second aim is to investigate concurrent
validity of the IMP with AIMS and TINE in a large sample of infants and assessments. Based on the
idea that the AIMS measures motor performance and the IMP assesses various aspects of motor
behaviour including motor performance, we expect a moderate correlation between the total IMP
score and the AIMS and a high correlation between the performance domain of the IMP and the
AIMS. As the IMP assesses several parameters of neurological integrity, we expect strong association
between IMP scores and neurological condition assessed with TINE. Third aim of this study is to
determine predictive validity of the IMP at 4, 6, 10 and 12 months for neurological outcome at 18
months measured by the Hempel assessment.
METHODS
Participants
We included a longitudinal study group of term and preterm infants and a cross-sectional study
group of only term infants. The longitudinal study group consisted of 30 term born and 59 preterm
infants. The term infants (12 girls and 18 boys) were recruited from amongst colleagues and
acquaintances of the researchers. Median gestational age of the longitudinal term group was 40.1
weeks (range 37.6-42 weeks), median birth weight was 3588 grams (range 2730-4470 grams) and
there had been no pre or perinatal complications. Fifty-nine infants were born preterm (25 girls
and 34 boys) with median gestational age 29.7 weeks (range 25 to 34.7 weeks) and median birth
weight of 1285 grams (range 630 to 2180 grams). The preterm infants had been admitted to the
neonatal intensive care unit of the Beatrix Children’s Hospital of the University Medical Center (UMC)
in Groningen between December 2003 and January 2005. Thirty-�ve of the preterm infants were
singletons and 24 were twins. Nine pairs of twins participated in the study; the remaining six had
lost their twin sibling. Neonatal ultrasound was available for 57 of the 59 preterm infants. Serious
brain pathology was observed in six infants: one infant had cystic PVL15, four infants had IVH grade
IV16 and one infant had middle cerebral artery infarction on the right side. The longitudinal study
group (term and preterm infants) was assessed at corrected ages 4, 6, 10, 12 and 18 months.
For the concurrent validity part of the study, another group of 116 term born infants (62
girls and 54 boys) was added. They were recruited at Well Child Centers and had cross-sectional
assessments at ages 4, 6, 10, 12 or 18 months. Median gestational age was 40.1 weeks (range 37-
43 weeks) and median birth weight was 3500 grams (1960-4660 grams). All parents of the infants
signed an informed consent form. The project was approved by the Ethics Committee of the UMC
in Groningen.
81
6
Concurrent and predictive validity of the Infant Motor Pro�le
Socio-economic, perinatal and neonatal data were collected for all infants on standardised
forms by means of an interview with the parents and consultation of neonatal intensive care unit
discharge certi�cates. Socio-economic status (SES) was operationalized as the sum score of four
variables describing educational and professional level of father and mother, all expressed on a
scale from 0 (lowest) through 2 (highest). ‘Small for gestational age’ (SGA) was de�ned as birth
weight below 10th percentile17. ‘Signs of fetal distress’ was de�ned as the presence of at least one
of the following factors: meconium staining, cardiotocography abnormalities and acidaemia during
delivery (arterial umbilical pH below 7.05).Table I shows socio-economic and neonatal characteristics
Table I: Socio-economic and neonatal characteristics of study groups
Term Preterm p-value
Number of children 146 59
Male gender, n (%) 72 (49) 34 (58) 0.282
Maternal age at child birth (years), mean SD 32 5.1 31.8 5.2 0.826
SESa, median (range) 6 (0-8) 4 (0-8) 0.002f
Twins, n (%) 2 (1.4) 24 (41) <0.001
Gestational age (weeks), median (range) 40.1 (37-43) 29.7 (25-34.7) <0.001
Birthweight (grams), median (range) 3500 (1960-4660) 1285 (630-2180) <0.001
Small for gestational ageb, n (%) 15 (10) 20 (34) <0.001
Caesarian section, n (%) 16 (11) 34 (58) <0.001
Signs of fetal distressc, n (%) 35 (24)d 30 (51)e <0.001
Apgar score at 5 minutes, median (range) 10 (7-10) 9 (4-10) <0.001
Art�cial ventilation, n (%) 1 (0.7) 40 (68) <0.001
a SES = socio-economic status, sum score of four variables describing educational and professional level of fatherand mother, all expressed on a scale from 0 (lowest) through 2 (highest).b Small for gestational age is de�ned as birthweight compared with gestational age below 10th centile17.c Presence of at least one of the following factors: meconium staining, CTG abnormalities, acidaemia duringdelivery (arterial umbilical pH below 7.05).d Data on signs of fetal distress were available for 144 of the term infants: 107 (73%) infants showed no signs offetal distress, 23 (16%) had meconium staining, 11 (7,5%) had CTG abnormalities and one had arterial umbilicalpH of 6.82. This infant had perinatal asphyxia with meconium aspiration for which ventilation was required. Inaddition, neonatal convulsions occurred. Brain MRI at day 6 was normal.e 30 (51%) of the preterm infants showed signs of fetal distress: 4 (6,8%) had meconium staining, 22 (37%)showed CTG abnormalities, 2 (3%) had acidaemia with low umbilical pH and 2 (3%) showed a combination ofsigns of fetal distress, both as a result of placental dysfunction.f SES of preterm group is lower than term group
of term and preterm groups. As no signi�cant di!erences in characteristics were found between the
longitudinal and cross-sectional term groups, we displayed them as one term group.
Procedures
Assessments were performed at (corrected) ages 4, 6, 10, 12 and 18 months for the longitudinal
82
Chapter 6
term group and the preterm group. The term cross-sectional group was assessed at one (n = 102
infants), two (n = 13 infants) or three (n = 1) of these ages. The actual number of infants assessed at
each age is displayed in Table II. Assessments consisted of a video-recording of approximately 15
minutes of spontaneous motor behaviour in supine, prone, sitting, standing, and walking condition,
depending on age and functional capacities of the infant. Furthermore, reaching, grasping and
manipulation of objects was tested in supine and in (supported) sitting position. The 80 items of the
IMP were scored on the basis of the video recording. These constitute the scores in �ve domains: size
of repertoire (variation), ability to select (variability), movement �uency, movement symmetry and
motor performance. The mean of the �ve domain-scores is the total IMP score5. IMP assessments of
the longitudinal term and preterm groups were carried out by KRH who knew whether an infant
was term or preterm born, but was not aware of any of the perinatal and neonatal details. IMP
assessments of the cross-sectional term group were scored by KRH and LE.
Table II: Number of assessments for term and preterm groups
Group Numberof infants 4 mo 6 mo 10 mo 12 mo 18 mo Number of
assessments
Term longitudinal 30 30a 30 30 30 30 150
Term cross-sectional 116 22 25 26 29 30 131
Preterm 59 58 57b 54 54 57 280
Total 205 110 112 110 113 117 561
mo = monthsa For 1 term infant data of neurological examination at 4 months were missing, only AIMS and IMP were assessed.b For 1 preterm infant AIMS score could not be determined at 6 months, because assessment in prone positionwas not performed.
Reliability assessment
LE, who was involved in this study as a master student, was trained in the assessment of the IMP.
During a training period of �ve weeks, 100 video’s of term and preterm infants at various ages were
assessed. After this training period, inter observer agreement between LE and KRH was investigated
on a sample of another 25 video’s consisting of �ve video’s at each of the �ve assessment ages
(4,6,10,12 and 18 months). The �ve video’s at each age consisted of two randomly selected video’s
of the term infant group and three of the preterm group.
Concurrent validity
At all ages the AIMS12 and Touwen Infant Neurological Examination (TINE13,14) were assessed, in order
to investigate concurrent validity of the IMP with these instruments. The AIMS was scored on the
basis of the video-recording of spontaneous motor behaviour. Total AIMS scores, instead of centiles,
were used in the data processing, as the Canadian reference values seem currently inappropriate
for Dutch children18. Reliability of the AIMS is good, but predictive validity for major developmental
83
6
Concurrent and predictive validity of the Infant Motor Pro�le
disorders is only moderate19,20. The Touwen Infant Neurological Examination (TINE13,14) was
performed at (corrected) ages of 4, 6, 10 and 12 months. In TINE, neurological signs are organized
according to age-speci!c norms into clusters of dysfunction. Five clusters are distinguished: reaching
and grasping, gross motor function, brain stem function, visuomotor function and sensorimotor
function (consisting of re"exes and muscle tone). Neurological condition is classi!ed as abnormal
if there is a distinct neurological syndrome, such as a hemisyndrome, irrespective of the number
of deviant clusters. An infant is classi!ed as having minor neurological dysfunction (MND) in case
of presence of more than two clusters of dysfunction. Two forms of typical neurological condition
are distinguished: normal – suboptimal when one or two clusters are deviant and normal when no
clusters ful!l criteria for dysfunction14. Reliability of TINE is good. Predictive validity is good for major
developmental motor disorders such as cerebral palsy and moderate for minor motor disorders4,14.
Predictive validity for major developmental motor disorders such as cerebral palsy is good, but for
minor motor disorders moderate at best4,14.
Predictive validity
For assessment of the predictive validity of the IMP, neurological outcome at the (corrected) age of
18 months was determined with the Hempel assessment21. This method is suitable for children of
pre-school age, from 18 months until four years of age. Similar to the TINE, the Hempel assessment
classi!es neurological signs into clusters of dysfunction, namely !ne motor dysfunction, gross
motor dysfunction, dysfunctional muscle tone regulation, re"ex abnormalities and visuomotor
dysfunction. Neurological condition is classi!ed into four categories: abnormal, complex MND
(denoting the presence of more than one dysfunctional cluster), simple MND (one cluster of
dysfunction) or normal (no deviant clusters or the isolated presence of re"ex abnormalities)22.
Statistical analyses
To analyze inter-observer reliability, intra-class correlation coe#cients (ICCs) for a two-way mixed
e$ects model with associated 95% con!dence intervals were used. Di$erences in IMP scores
between term and preterm groups and between the four neurological conditions were analyzed
by means of the non-parametric Mann-Whitney U test and Kruskal-Wallis test respectively. Relations
between IMP scores and AIMS scores and correlation between IMP scores throughout infancy and
neurological outcome at 18 months were assessed with Spearman’s rank correlation with associated
con!dence intervals. Interpretation of Spearman’s correlation coe#cient was as follows: rho < 0.50
weak relationship, 0.50 ≥ rho ≥ 0.75 moderate relationship, rho > 0.75 good relationship23. To assess
predictive validity of the IMP scores for the outcome at 18 months, a cut-o$ score below the 5th
percentile was used. Throughout the analyses, di$erences and correlations with a p-value < 0.05
were considered to be statistically signi!cant (two-tailed testing).
84
Chapter 6
RESULTS
At all ages total IMP scores did not signi�cantly di�er between girls and boys. The preterm group
consistently showed lower total IMP scores than the term group (Figure 1, Mann-Whitney U test p
values < 0.001). This was also the case for the scores on the domains size of repertoire (p < 0.001 at
all ages), !uency (p< 0.001 at 4, 6, 12 and 18 months, p = 0.028 at 10 months) and performance (p =
0.001 at 4 months and p < 0.001 at 6, 10, 12 and 18 months). Scores on adaptive selection (variability)
were signi�cantly lower for preterm infants compared to term infants from age 10 months onwards
(4 months p = 0.90, 6 months p = 0.16, 10 months p<0.001, 12 months p = 0.03, 18 months p =
0.001). Symmetry scores were signi�cantly lower for the preterm group at ages 4 and 18 months
(p-values 0.02 and 0.04 respectively), but not at 6,10 and 12 months.
Reliability
Interobserver reliability of the total IMP score yielded an intraclass correlation coe"cient (ICC) of
0.95 (95% con�dence interval 0.89-0.98), indicating good reliability. Reliability of IMP domains was
moderate to good with ICC from 0.74 to 0.99 (Table III).
Concurrent validity
Correlations between AIMS scores and total IMP scores were weak to moderate at all ages
Figure 1: Di!erences in total IMP scores between term and preterm infants at 4, 6, 10, 12 and 18 months.Data are presented as median values (horizontal bars), interquartile ranges (boxes) and ranges (vertical lines).Open circles and asterisks represent outliers. FT = full term group, PT = preterm group. At all ages di�erences intotal IMP scores between the two groups were signi�cant (Mann-Whitney U test p < 0.001).
60
70
80
90
Tota
lIM
P-sc
ore
(%)
100
FT PT4 months
FT PT6 months
FT PT10 months
FT PT12 months
FT PT18 months
60
70
80
90
Tota
lIM
P-sc
ore
(%)
100
FT PT4 months
FT PT6 months
FT PT10 months
FT PT12 months
FT PT18 months
6
85
Concurrent and predictive validity of the Infant Motor Pro�le
(Spearman’s rho 0.36 to 0.55, see Table IV). The performance domain of the IMP showed the strongest
correlations with the AIMS scores, especially at the age of 10 and 12 months (Spearman’s rho 0.84
and 0.81 respectively, Table IV). Correlations between the other domains and the AIMS were weak
(Spearman’s rho 0.01-0.41, Table IV). Preterm infants had signi�cantly lower AIMS scores than term
infants at ages 4, 10, 12 and 18 months (Mann-Whitney U test, p-values 0.001, < 0.001, < 0.001 and
<0.001 respectively). No di�erence was found at age 6 months (p = 0.32).
We found a clear relationship between the total IMP score and neurological condition at all
ages: infants with a normal neurological condition had highest IMP scores and infants with an
abnormal neurological condition had lowest IMP scores (Figure 2, Kruskal-Wallis p < 0.001 at all
ages). The domains size of repertoire, �uency, symmetry and performance were highly signi�cantly
related to neurological condition at all ages (p-values for variability, �uency and performance
all < 0.001, except for �uency at 10 months p = 0.008; p-values for symmetry respectively 0.025,
0.006, < 0.001, 0.025 and < 0.001 at 4, 6, 10, 12 and 18 months). Scores on the domain ability to
select (variability) were signi�cantly di�erent between neurological conditions at ages 10 and 12
months (p=0.021 and 0.008 respectively), but not at ages 4, 6 and 18 months (p= 0.49, 0.46 and 0.06
respectively).
Predictive validity
Neurological condition at 18 months was determined with the Hempel examination21. Of the
longitudinal term group, 23 children had normal neurological condition and 7 had simple MND.
None of the term children showed complex MND or abnormal neurological condition at 18 months.
Of the preterm group, eleven children had a normal neurological condition, 7 had simple MND, 31
had complex MND and neurological condition of 8 infants was considered as abnormal (14% of
preterm group), of which four infants had a unilateral spastic CP and four had bilateral spastic CP.
Two preterm children did not have follow-up at 18 months.
For the total longitudinal group of infants, correlation between total IMP scores throughout
infancy and neurological outcome at 18 months was moderate, with Spearman’s rho’s of -0.62
Table III: Inter observer reliability of total IMP score and IMP domains
Intraclass correlation coe!cient (ICC)(95% Con"dence Interval)
Total IMP score 0.95 (0.89-0.98)
IMP domain
Size of repertoire 0.91 (0.81-0.96)
Ability to select 0.79 (0.58-0.90)
Fluency 0.74 (0.49-0.87)
Symmetry 0.99 (0.98-1.0)
Performance 0.98 (0.95-0.99)
86
Chapter 6
4 mo
Nn=31
N-subn=34
MNDn=31
An=13
60
70
80
90
Tota
lIM
P-sc
ore
(%)4
mo
Neurological condi on 4 mo
100
-
-
-
-
-
-6 mo
Nn=33
N-subn=46
MNDn=27
An=5
70
80
90
Tota
lIM
P-sc
ore
(%)6
mo
Neurological condi on 6 mo
100
-10 mo
Nn=42
N-subn=38
MNDn=26
An=4
70
80
90
Tota
lIM
P-sc
ore
(%)1
0m
o
Neurological condi on 10 mo
100
70
80
90
-
100
12 mo
Nn=49
N-subn=38
MNDn=22
An=4
70
80
90
Tota
lIM
P-sc
ore
(%)1
2m
o
Neurological condi on 12 mo
100-
18 mo
Nn=62
N-subn=16
S-MNDn=31
An=8
60
70
80
90
Tota
lIM
P-sc
ore
(%)1
8m
o
Neurological condi on 18 mo
100
-
--
Figure 2: Relationship between total IMP scoresand neurological condition at the various ages.Data are presented as median values (horizontalbars), interquartile ranges (boxes) and ranges(vertical lines). N = normal neurological condition,N-sub = normal suboptimal neurological condition,MND = minor neurological dysfunction, A =abnormal neurological condition, mo = months, n= number of infants, S-MND = simple MND, C-MND= complex MND. At all ages di�erences in total IMPscores between the four neurological conditionswere signi�cant (Kruskal-Wallis p < 0.001).
87
6
Concurrent and predictive validity of the Infant Motor Pro�le
Table IV: Spearman’s correlation coe�cients of IMP scores and AIMS scores per age
AIMS 4 mo AIMS 6 mo AIMS 10 mo AIMS 12 mo AIMS 18 mo
Total IMP score 0.43**
(0.26-0.57)0.34**
(0.19-0.52)0.55**
(0.40-0.66)0.43**
(0.27-0.57)0.36**
(0.19-0.51)
IMP domain
Size ofrepertoire
0.39**
(0.22-0.54)0.29**
(0.11-0.45)0.36**
(0.19-0.51)0.37**
(0.19-0.52)0.41**
(0.25-0.55)
Ability toselect
0.07(-0.11-0.26)
0.01(-0.18-0.19)
0.22*
(0.04-0.39)0.11
(-0.07-0.29)0.04
(-0.14-0.22)
Fluency 0.27**
(0.09-0.43)0.24*
(0.06-0.41)0.08
(-0.11-0.26)0.15
(-0.03-0.33)0.23*
(0.05-0.39)
Symmetry 0.23*
(0.05-0.40)0.05
(-0.13-0.24)0.13
(-0.05-0.31)0.12
(-0.07-0.30)0.32**
(0.14-0.47)
Performance 0.56**
(0.42-0.67)0.56**
(0.41-0.67)0.84**
(0.78-0.89)0.81**
(0.73-0.86)0.47**
(0.31-0.60)
Spearman’s correlation coe!cients with associated 95% con"dence intervals between brackets, * p < 0.05, ** p< 0.01.
Table V: Predictive validity of total IMP score at ages 4, 6, 10 or 12 months for neurological outcome at 18months
Total IMP4 mo
Total IMP6 mo
Total IMP10 mo
Total IMP12 mo
Prediction of CP at 18 mo (IMP score < p5)
Sensitivity 100 88 100 86
Speci"city 71 78 78 77
PPV 26 29 26 26
NPV 100 98 100 98
Accuracy 73 79 79 78
Prediction of deviant neurologicaloutcome (complex MND or CP) at 18 mo (IMP score < p5)
Sensitivity 63 55 56 63
Speci"city 85 94 93 98
PPV 77 88 87 96
NPV 75 72 73 78
Accuracy 76 76 77 83
mo = months, p5 = 5th percentileSensitivity = true positives/(true positives + false negatives)Speci"city = true negatives/(true negatives + false positives)PPV = positive predictive value = true positives/all positivesNPV = negative predictive value = true negatives/all negativesAccuracy = (true positives + true negatives)/all subjects
88
Chapter 6
(CI -0.74—0.47), -0.57 (CI -0.70—0.41), -0.67 (CI -0.77—0.53) and -0.65 (CI -0.76—0.50) respectively
at ages 4, 6, 10 and 12 months. Predictive validity of total IMP scores (with cut-o� score below 5th
percentile) for abnormal neurological condition (CP) at 18 months varied with age of assessment
and showed sensitivity from 86 to 100% and speci�city from 71 to 78% (Table V). Predictive validity
for deviant neurological outcome (complex MND or CP) at 18 months demonstrated sensitivity from
55 to 63% and speci�city of 85 to 98% (Table V). If instead of 5th percentile the 15th percentile was
used as cut-o� point, sensitivity increased to 68 to 89%, but speci�city decreased to 72 to 87%.
Accuracy was around 80% and did not di�er between level of cut-o� point at 5th or 15th percentile.
DISCUSSION
Our study showed a good reliability of the IMP. Concurrent validity of the IMP with the AIMS met
the a priori theoretical expectations: the IMP performance domain correlated best with the AIMS.
Concurrent validity of the IMP with TINE and Hempel examination was very good. Predictive validity
of the IMP for neurological outcome at 18 months was satisfactory.
Strengths of this study are its predominantly longitudinal character and the low attrition rate
(3.5% over all longitudinal assessments, see Table II). A weakness of the study is that the assessors
were not blind with respect to term or preterm status of the infant. This may have in!uenced
scoring. However, the assessors were unaware of any details of the child’s clinical history or results
of neonatal ultrasounds. Besides, if assessors had been blinded to term or preterm status, it would
have been di"cult to conceal preterm status, because the infant’s appearance usually discloses
preterm birth. A second weakness of this study is the relatively short duration of follow-up. At 18
months, clinical signs of CP often are not yet fully expressed and signs of minor developmental
motor disorders may not be present until school age24,25.
The term group showed a relatively high percentage of infants with signs of fetal distress,
consisting mainly of meconium staining which is fairly common in term deliveries26. The group of
preterm infants included in this study may be considered as a representative sample of a Dutch
neonatal intensive care unit population in a tertiary referral centre with respect to gestational age,
birthweight, frequency of Apgar score at 5 minutes below 7 (15%) and ventilatory support (68%)27,28.
Our preterm sample showed a relatively high percentage of infants small for gestational age and
Caesarean deliveries27,28. The latter could be due to the increased tendency over the years to deliver
very preterm infants by Caesarean section29. The preterm group had lower socio-economic status
than the term group, in accordance with social disadvantage being considered a risk factor for
preterm birth30.
Interobserver reliability of total IMP score and the domains size of repertoire, symmetry and
performance was good to very good. Reliability of the IMP domains ability to select and !uency was
moderate. Interobserver reliability in this study was higher than in our pilot study5, probably as a
result of the more precise de�nitions and descriptions of the IMP items which had been developed
6
89
Concurrent and predictive validity of the Infant Motor Pro�le
in the meantime. With good training, IMP scoring can be learnt reliably without prior experience in
infant motor development.
The study showed clear di�erences in IMP scores between term and preterm infants for total
IMP scores and the domains size of repertoire (variability), �uency and performance at all ages.
Preterm birth is a major risk factor for developmental motor disorders6. Preterm infants are at risk
for brain lesions, which according to NGST lead to reduced variability of motor behaviour11. Loss
of �uency of motor behaviour is one of the !rst signs of non-optimal neurological condition1.
Delayed acquisition of motor milestones, in the IMP re�ected as lower performance scores, can be
a sign of developing CP31. For the domain ability to select (variability), scores signi!cantly di�ered
between term and preterm infants from 10 months onwards. In typically developing infants, the
ability to select suitable motor strategies gradually emerges at function-speci!c ages after the !rst
half year of life10. Infants with developmental motor disorders often have problems in processing
a�erent, sensory information and !ne-tuning and adapting motor behaviour11,32,33. This could delay
or hamper development of the ability to select, as adaptive selection relies on a�erent feedback.
Scores on the domain symmetry di�ered between term and preterm infants at ages 4 and 18
months. The asymmetries observed at 4 months could be transitory neurological !ndings that
resolve spontaneously34, whereas the asymmetries observed at later age of 18 months could be
signs of the development of unilateral spastic CP24.
As we expected, the total IMP score only correlated to a moderate extent with the AIMS.
Correlation between AIMS and IMP was highest for the performance domain of the IMP, especially at
the ages of 10 and 12 months. Both the performance domain of the IMP and the AIMS assess motor
achievements in a quantitative way. From the age of 10 months onwards, motor development is
characterized by a rapid gain in motor milestones, which is re�ected in both IMP and AIMS scores.
After the age of 14 months, the discriminative power of the AIMS is diminished12,35. The IMP domain
size of repertoire assesses another aspect of motor behaviour than the AIMS. The !nding that these
parameters were weakly but signi!cantly correlated at all ages indicates that both assess di�erent
aspects of the same underlying construct, being neuromotor integrity. The IMP domain ability to
select was only related to the AIMS score at age ten months, the !rst age in the present study at
which selection of adaptive motor strategies was present to some extent. The IMP domain �uency
was weakly related to the AIMS. Loss of movement �uency is one of the !rst signs of non-optimal
neurological condition1, but it is not speci!c for serious developmental motor disorders that are
associated with low motor performance. The IMP domain symmetry was related to the AIMS score
at the ages of 4 and 18 months, the latter probably representing the infants with a developing
unilateral CP, which besides a low symmetry score also leads to reduced motor performance31.
Concurrent validity of IMP and the age-speci!c neurological examination (TINE and Hempel)
was very good: at all ages infants with normal neurological condition had higher IMP scores than
infants with minor neurological dysfunction or abnormal neurological condition. Di�erences in
IMP scores were especially found between infants with (complex) MND or abnormal neurological
90
Chapter 6
condition and infants with normal or normal – suboptimal neurological condition (or simple MND
at 18 months). Complex MND, in contrast with simple MND, has clinical relevance and is associated
with pre- or perinatal adversities22,25. These di!erences in IMP scores between infants with di!erent
neurological conditions were found at all ages for the domains size of repertoire (variation),
"uency, symmetry and performance, supporting the notion that these are indeed parameters of
neuromotor integrity5. Scores on the domain ability to select were related to neurological condition
at ages 10 and 12 months, but not at 4, 6 and 18 months. This is in analogy with the di!erences we
found between term and preterm infants, except that they did di!er in adaptive selection scores at
18 months. The data indicated that the absence of a relation between neurological condition and
adaptive selection at 18 months was brought about by relatively good scores of infants with an
abnormal neurological condition. Motor behaviour of these children was characterized by a limited
motor performance; nevertheless they demonstrated, within the skills which they had developed, a
relatively good ability of adaptive section out of their limited motor repertoire.
Sensitivity of IMP scores throughout infancy for predicting CP at 18 months was very high.
Positive predictive value of the IMP for neurological outcome at 18 months was low, but negative
predictive value was high, implying that IMP scores above the 5th percentile almost certainly excluded
abnormal neurological outcome at 18 months. It is important to realize that predictive values of a
test strongly depend on the prevalence of the disorder, e.g. CP, in the study population36. Therefore,
predictive values observed in our sample cannot be extrapolated to the general population.
Prediction of developmental outcome at an early age is di#cult and will never be perfect,
because change is one of the main characteristics of the developing brain. To optimize prediction
of neuromotor outcome in children at high risk for developmental motor disorders, it is probably
best to combine multiple, complementary tools, such as neurological examination, assessment of
milestones and assessment of qualitative aspects of motor behaviour4 in addition to neuroimaging
and neurophysiological techniques.
CONCLUSION
The Infant Motor Pro$le is a qualitative assessment of motor behaviour based on the Neuronal Group
Selection Theory on motor development. Interobserver reliability of the IMP is good. Concurrent
validity of the IMP with the Alberta Infant Motor Scale was especially high for the performance
domain of the IMP. Concurrent validity of IMP with Touwen Infant Neurological Examination was
very good. With respect to the three purposes for which the IMP was developed, we can conclude
that the IMP is well able to discriminate between typically developing infants and infants with high
risk for developmental motor disorders. The ability of the IMP to evaluate motor function over time
should be further explored by applying the IMP in intervention studies. Prediction of neurological
outcome is very di#cult, due to change being one of the main characteristics of the developing
nervous system. In our study population, predictive validity of the IMP was satisfactory. Future
studies will aim at generating norm-scores and determining clinical applicability.
91
6
Concurrent and predictive validity of the Infant Motor Pro�le
ACKNOWLEDGEMENTSWe thank Nicole Arink, Hylco Bouwstra, Dineke Dijkhuizen and Anne Hoekstra for performing part of theassessments. The study was supported by a Junior Scienti�c Masterclass grant of the post-graduate schoolBehavioral and Cognitive Neurosciences (BCN), University of Groningen.
92
Chapter 6
REFERENCES1. Hadders-Algra M. General movements: A window for early identi�cation of children at high risk for
developmental disorders. J Pediatr 2004,145:S12-18.
2. Einspieler C, Prechtl HFR, Bos AF, et al. Prechtl’s method on the qualitative assessment of GeneralMovements in preterm, term and young infants. Clinics in Developmental Medicine No. 167. London: MacKeith Press; 2004.
3. Campbell SK, Kolobe TH, Osten ET, Girolami GL, et al. Construct validity of the test of infant motorperformance. Phys Ther 1995;75:585-596.
4. Heineman KR, Hadders-Algra M. Evaluation of neuromotor function in infancy - A systematic review ofavailable methods. J Dev Behav Pediatr 2008;29:315-23.
5. Heineman KR, Bos AF, Hadders-Algra M. The Infant Motor Pro�le: a standardized and qualitative method toassess motor behaviour in infancy. Dev Med Child Neurol 2008;50:275-82.
6. Larroque B, Ancel PY, Marret S, Marchand L, André M, Arnaud C, Pierrat V, Rozé JC, Messer J, Thiriez G,Burguet A, Picaud JC, Bréart G, Kaminski M; EPIPAGE Study group. Neurodevelopmental disabilities andspecial care of 5-year-old children born before 33 weeks of gestation (the EPIPAGE study): a longitudinalcohort study. Lancet. 2008;371:813-20.
7. Kolb B, Brown R, Witt-Lajeunesse A, Gibb R. Neural compensations after lesion of the cerebral cortex.Neural Plast 2001;8:1-16.
8. De Graaf-Peters VB, Hadders-Algra M. Ontogeny of the human central nervous system: what is happeningwhen? Early Hum Dev 2006;82:257-266.
9. Sporns O, Edelman GM. Solving Bernstein’s problem: a proposal for the development of coordinatedmovement by selection. Child Dev 1993;64:960-981.
10. Hadders-Algra M. The Neuronal Group Selection Theory: a framework to explain variation in normal motordevelopment. Dev Med Child Neurol 2000;42:566-572.
11. Hadders-Algra M. The Neuronal Group Selection Theory: promising principles for understanding andtreating developmental motor disorders. Dev Med Child Neurol 2000;42:707-715.
12. Piper MC, Darrah J. Motor assessment of the developing infant. Philadelphia: Saunders, 1994.
13. Touwen BCL. Neurological development in infancy. Clinics in Developmental Medicine No. 58. London:Mac Keith Press, 1976.
14. Hadders-Algra M, Heineman KR, Bos AF, Middelburg KJ. The assessment of minor neurological dysfunctionin infancy using the Touwen Infant Neurological Examination: strengths and limitations. Dev Med ChildNeurol, in press.
15. De Vries LS, Eken P, Dubowitz LM. The spectrum of leukomalacia using cranial ultrasound. Behav Brain Res1992;31;49:1-6.
16. Volpe JJ. Neurology of the Newborn. 4th ed. Philadelphia PA: WB Saunders, 2001.
17. Kloosterman GJ. On intrauterine growth: the signi�cance of prenatal care. Int J Gynaecol Obstet 1970;8:895-912.
18. Fleuren KM, Smit LS, Stijnen T, Hartman A. New reference values for the Alberta Infant Motor Scale need tobe established. Acta Paediatr 2007;96:424-427.
19. Darrah J, M Piper, MJ Watt. Assessment of gross motor skills of at-risk infants: predictive validity of theAlberta Infant Motor Scale. Dev Med Child Neurol 1998;40:485-491.
20. Bartlett DJ, Fanning JE. Use of the Alberta Infant Motor Scale to characterize the motor development ofinfants born preterm at eight months corrected age. Phys Occup Ther Pediatr 2003;23:31-45.
21. Hempel MS. The neurological examination for toddler-age. PhD thesis. University of Groningen, 1993.
22. Hadders-Algra M. Developmental coordination disorder: is clumsy motor behavior caused by a lesion ofthe brain at early age? Neural Plast 2003;10:39-50.
93
6
Concurrent and predictive validity of the Infant Motor Pro�le
23. Portney LG, Watkins MP. Part IV Data analysis: Correlation. Foundations of clinical research. Applications topractice. 2nd ed. Upper Saddle River, NJ: Prentice Hall Health; p. 494, 2000.
24. Krägeloh-Mann I, Cans C. Cerebral palsy update. Brain Dev 2009; Epub ahead of print.
25. Hadders-Algra M. Two distinct forms of minor neurological dysfunction: perspectives emerging from areview of data of the Groningen Perinatal Project. Dev Med Child Neurol 2002;44:561-71.
26. Tran SH, Caughey AB, Musci TJ. Meconium-stained amniotic �uid is associated with puerperal infections.Am J Obstet Gynecol 2003;189:746-50.
27. Stoelhorst GM, Rijken M, Martens SE, Brand R, den Ouden AL, Wit JM, Veen S; Leiden Follow-Up Project onPrematurity. Changes in neonatology: comparison of two cohorts of very preterm infants (gestational age<32 weeks): the Project On Preterm and Small for Gestational Age Infants 1983 and the Leiden Follow-UpProject on Prematurity 1996-1997. Pediatrics 2005;115:396-405.
28. de Kleine MJ, den Ouden AL, Kollée LA, van Baar A, Nijhuis-van der Sanden MW, Ilsen A, Brand R, Verloove-Vanhorick SP. Outcome of perinatal care for very preterm infants at 5 years of age: a comparison between1983 and 1993. Paediatr Perinat Epidemiol 2007;21:26-33.
29. Haque KN, Hayes AM, Ahmed Z, Wilde R, Fong CY. Caesarean or vaginal delivery for preterm very-low-birth weight (< or =1,250 g) infant: experience from a district general hospital in UK. Arch Gynecol Obstet.2008;277:207-12.
30. Zeka A, Melly SJ, Schwartz J. The e�ects of socioeconomic status and indices of physical environment onreduced birth weight and preterm births in Eastern Massachusetts. Environ Health. 2008;7:60.
31. Allen MC, Alexander GR. Using motor milestones as a multistep process to screen preterm infants forcerebral palsy. Dev Med Child Neurol 1997;39:12-6.
32. Wilson PH, McKenzie BE. Information processing de�cits associated with developmental coordinationdisorder: a meta-analysis of research �ndings. J Child Psychol Psychiatry 1998;39:829-840.
33. Cooper J, Majnemer A, Rosenblatt B, Birnbaum R. The determination of sensory de�cits in children withhemiplegic cerebral palsy. J Child Neurol 1995;10:300-309.
34. Michaelis R, Asenbauer C, Buchwald-Saal M, Haas G, Krägeloh-Mann I. Transitory neurological �ndings in apopulation of at risk infants. Early Hum Dev 1993;34:143-53.
35. Van Haastert IC, de Vries LS, Helders PJ, Jongmans MJ. Early gross motor development of preterm infantsaccording to the Alberta Infant Motor Scale. J Pediatr 2006;149:617-22.
36. Altman DG. Practical statistics for medical research. London: Chapman and Hall, p411-412, 1991.
94
95
7CHAPTER
General discussion
96
Chapter 7
GENERAL DISCUSSION
This thesis presents the Infant Motor Pro�le (IMP), a recently developed qualitative assessment
of motor behaviour in infancy. The IMP is based on ideas of the Neuronal Group Selection Theory
(NGST) which states that motor development is characterised by exploration of the primary variable
motor repertoire and selection of adaptive motor strategies out of this repertoire1,2,3. The IMP
consists of 80 items that are grouped into �ve domains: size of the repertoire (variation), adaptive
selection (variability), movement �uency, movement symmetry and motor performance.
This thesis had three aims: 1) review of existing methods for the assessment of neuromotor
function in infancy, 2) development of the IMP and 3) assessment of the psychometric properties
of the IMP. The �rst two aims are primarily outlined in chapters 1 and 2. The present discussion
will mainly focus on the third aim. First, the �ndings with respect to the psychometric properties
of the IMP will be commented. Next, some methodological considerations with respect to study
group and assessments will be discussed. Last, future research topics and concluding remarks will
be presented.
Psychometric properties of the IMP
To assess the psychometric properties of the IMP, we performed a longitudinal prospective study
on a heterogeneous group of term and preterm born infants aged 3 to 18 months. In addition,
another group of term born infants had cross-sectional assessments at the same ages. Assessments
consisted of the IMP, the Alberta Infant Motor Scale (AIMS4) and an age-speci�c neurological
examination (Touwen Infant Neurological Examination5,6 (TINE) at 4, 6, 10 and 12 months corrected
age and Hempel assessment at 18 months corrected age7).
Intra- and interobserver reliability of the IMP
Intra- and interobserver reliability of the IMP were assessed in the pilot study (chapter 3) and in the
study described in chapter 6. In the pilot study, intra- and interobserver reliability were satisfactory
in the persons who developed the IMP. To test whether it was possible to train a new, inexperienced
assessor, we trained a master student without prior experience in infant motor development.
Reliability of the total IMP score was very good and reliability of the di!erent domains varied from
moderate (ability to select and �uency) to good and very good (size of the repertoire (variation),
symmetry and performance). We found that scoring of the IMP can be learnt reliably after good
training; a typical training course would consist of basic instructions followed by assessment of
about 100 videorecordings with feedback by an experienced supervisor at regular intervals. No prior
experience with developmental assessment of infants is required. The amount of training needed is
comparable to for example the training required for the General Movement method8. Reliability of
the IMP is comparable to that of other methods for assessment of neuromotor function in infancy9,
such as the Touwen Infant Neurological Examination6, Alberta Infant Motor Scale4,10,11 (AIMS), Bayley
Variability in IMP scores at di#erent ages (within-participant variability) was especially large
in the preterm infants. It seems that infants with atypical motor development have less stable
developmental trajectories than typically developing infants, underscoring the need of repeated
assessments especially in high-risk groups, as atypical development can easily be missed with a
single assessment. This "nding is possibly not limited to the IMP assessment, but also valid for other
developmental motor assessments and has consequences for the organization of clinical follow-up
of high risk infants.
In the preterm group, higher socio-economic status was associated with higher IMP scores.
Highly educated parents probably provide both favourable genetic factors and a stimulating
environment. The latter is especially important for children at high biological risk of developmental
problems, such as preterm infants17. Early intervention should therefore especially be targeted at
this group of vulnerable children with both biological and social risk factors for developmental
problems, as these children are most likely to bene"t from it17. Unfortunately, it appears that
especially these children with high social risk are less likely to receive early intervention in practice18.
Figure 1 in chapter 5 shows large di#erences in IMP scores between term and preterm infants,
but even more pronounced di#erences within the preterm group. In general, a small part of the
preterm group has IMP scores comparable to the term infants. This small part is expected to have a
Scales of Infant Development12 and Peabody Developmental Motor Scale13. Intra- and interobserver
agreement of infant motor assessment are not perfect, which could give clinicians the inconvenient
impression of subjectivity. However, it should be kept in mind that other assessments that are used
in daily clinical routine do not have a perfect reliability as well (e.g. assessment of neonatal head
ultrasound scans)14,15 or are subject to an expectancy bias (e.g. interpretation of plantar re!exes in
adults16.
Construct validity of the IMP
Construct validity of the IMP was investigated in chapter 5. Construct validity is the extent to which
an instrument measures the theoretical construct of interest, in this case neuromotor function.
Construct validity was operationalized as the relation of IMP scores with pre-, peri- and neonatal
variables, including the presence of brain pathology on neonatal ultrasound scans of the brain.
A longitudinal prospective study in a group of term and preterm born infants from age 4 to 18
months was performed. Gestational age, socio-economic status and 5-minute Apgar score were
signi"cant determinants of IMP scores throughout infancy in the total group of infants. In the group
of the preterm infants, IMP scores were strongly a#ected by the presence of serious brain lesions on
neonatal ultrasounds and by socio-economic status. We also found a clear increase of IMP scores
with increasing age, which implies that the IMP is able to detect and re!ect age-related changes in
7
97
General discussion
motor development. These !ndings strongly contribute to construct validity of the IMP and make
it more plausible that the IMP indeed measures the construct of interest, namely neuromotor
function.
98
Chapter 7
more or less typical motor development. Many of the preterm infants have lower IMP scores than
the term infants throughout infancy, but they do develop most of the motor skills, though at a
later time, with a less variable repertoire to choose from than the typically developing infants. This
could possibly be the group of children that develops minor motor disorders such as DCD at a later
age, when more complicated �ne motor skills, such as writing or using scissors, and gross motor
skills involving �ne balancing are required in daily and school activities19. Whether these later motor
problems are already re!ected in the IMP scores in infancy has yet to be investigated.
A small number of the preterm infants had clearly low IMP scores from the start and did only
develop some basic motor skills throughout infancy, with IMP scores hardly rising or even lowering
throughout infancy (Figure 1 chapter 5). These are the children with a seriously hampered motor
repertoire due to brain lesions who were diagnosed with CP. It would be interesting to use the IMP in
this group of children during or after application of early intervention, in order to evaluate changes
in neuromotor function. This is currently done in a study on the e"ect of COPCA, a newly developed
physiotherapeutic early intervention programme, in a group of high risk infants20,21.
As mentioned above, the ability to detect age-related changes in motor development
contributes to the construct validity of an instrument for assessment of neuromotor function22.
In chapter 4 the development of adaptive motor behaviour in typically developing infants was
described. For this study, the longitudinal group of term infants had nine assessments between the
ages of 3 and 18 months. The emergence of adaptive selection of motor patterns was investigated
in four speci�c motor functions, namely abdominal progression, sitting motility, arm movements
during reaching and hand motility during grasping and manipulation. We found that transition to
clinically observable adaptive motor behaviour developed gradually from age 6 months onwards,
with a peak between 8 and 15 months. It occurred at speci�c ages for the di"erent motor functions.
It would be interesting to perform further research on ages of transition in infants with a high risk
for developmental motor disorders such as CP or DCD. In these children transition to secondary
variability could possibly be delayed or impeded, as they can have di#culties with processing
a"erent, sensory information and can experience problems in �ne-tuning and adapting motor
behaviour2,23,24. Further research on this subject is needed to make out whether age of transition to
secondary variability could be a useful clinical parameter of neuromotor condition.
Concurrent validity of the IMP
Concurrent validity of an instrument is the extent to which scores on the instrument relate to scores
on another instrument that assesses the same construct. It is important to note that instruments
available for infant motor assessment di"er in focus on the aspects of neuromotor function that
are assessed, depending on the theoretical background of the instrument (see review in chapter 2).
Good concurrent validity of a new instrument with established instruments contributes to the idea
that the new instrument indeed measures neuromotor function, as the established instruments are
supposed to do so. As the focus of the IMP is partly on di"erent aspects of neuromotor function than
99
7
General discussion
established instruments, namely size of repertoire (variation) and selection of motor behaviour,
we did not expect perfect concurrent validity between the IMP and established instruments. In
chapter 6, concurrent validity of the IMP with the Alberta Infant Motor Scale (AIMS), the Touwen
Infant Neurological Examination (TINE) and Hempel neurological examination (at 18 months)
was investigated. The total IMP score correlated to a moderate extent with the AIMS. Correlation
between AIMS and IMP was highest for the performance domain of the IMP, as both the performance
domain of the IMP and the AIMS assess motor achievements in a quantitative way. Concurrent
validity between IMP and neurological examination (TINE or Hempel) was very good: infants
with normal neurological condition had higher IMP scores than infants with minor neurological
dysfunction or abnormal neurological condition. Di�erences in IMP scores were especially found
between infants with complex MND or abnormal neurological condition and infants with normal
or normal - suboptimal neurological condition. Complex MND, in contrast with simple MND, has
clinical relevance and is associated with pre- or perinatal adversities25,26. All in all, the results on
concurrent validity of the IMP are satisfactory and therefore con�rm that the IMP, in line with other
infant motor assessments, is able to measure various aspects of neuromotor function.
In the study on concurrent validity of the IMP with the neurological examination, we noticed
that the total IMP score was not always the most optimal re�ection of the neuromotor functioning
of the child, especially for the infants with an abnormal neurological condition and a very limited
motor repertoire. In these infants, scores on the performance domain were generally low. As a result,
scores on the adaptive selection domain were relatively high, as only the limited motor performance
that the child did show was assessed with respect to adaptive selection. This leads to a relatively high
total IMP score, which is not a good re�ection of the overall poor neuromotor function. In this light,
in future studies for example during application of early intervention, especially in children with a
very high risk for developmental motor disorders such as CP, more attention should be paid to the
individual scores on the di�erent domains and not only to the total IMP scores. Possibly, it is better
to leave the score on the domain on adaptive selection out of the total IMP score during at least
during the �rst six month of age, as adaptive selection of motor behaviour for most of the motor
functions only starts to emerge from this age onwards. We think the studies described in this thesis
were not signi�cantly hampered by this, as only a limited number of infants in the preterm group
developed CP. However, this issue needs further research and it would be interesting to investigate
the development of IMP domain scores throughout infancy and the relation of domain scores with
pre, peri and neonatal variables, including results of brain imaging.
Predictive validity of the IMP
Predictive validity of an instrument is the extent to which the scores on the instrument now predict
future outcome27. Predictive validity of the IMP for neurological outcome at 18 months was described
in chapter 6. We found a very high sensitivity of IMP scores throughout infancy for predicting CP at
18 months. Positive predictive value of the IMP for neurological outcome at 18 months was low,
100
Chapter 7
but negative predictive value was high, implying that IMP scores above the 5th percentile almost
certainly excluded abnormal neurological outcome at 18 months. It is important to realize that
predictive values of a test strongly depend on the prevalence of the disorder, e.g. CP, in the study
population28. Therefore, predictive values observed in our sample cannot be extrapolated to the
general population. In addition, it makes comparison of predictive validity of di�erent infant motor
assessment methods di�cult, as investigations on predictive validity of di�erent tests are generally
done in di�erent study populations.
Prediction of developmental outcome at an early age is di�cult and will never be perfect,
because change is one of the main characteristics of the developing brain. To optimize prediction
of neuromotor outcome in children at high risk for developmental motor disorders, it is probably
best to combine multiple, complementary tools, which focus on di�erent aspects of neuromotor
function, such as neurological examination, assessment of milestones and assessment of qualitative
aspects of motor behaviour, in addition to neuroimaging and neurophysiological techniques9. As
mentioned above, repeated assessments are mandatory, especially in high-risk groups, because of
large variability in expression of neuromotor dysfunction over time6.
Methodological considerations
Study group
The study group consisted of three samples of infants: two longitudinal groups of respectively term
and preterm infants and one group of term infants with cross-sectional assessments.The longitudinal
term group was relatively small with only 30 infants that were assessed nine times between the ages
of 3 to 18 months. In spite of the small sample size, we were able to detect ages of transition for
di�erent motor functions in the study on development of adaptive selection in typically developing
infants (Chapter 4). The very low attrition rate (<1% of assessments were missed) was very valuable
in this respect. But it should be kept in mind that the inclusion procedure may have introduced a
selection bias. The infants were recruited amongst colleagues and acquaintances of the researchers:
socio-economic status was relatively high, neurological condition of the infants was remarkably
good and parents were highly motivated to participate which yielded the very low attrition rate. The
cross-sectional term group consisted of 116 infants who were recruited at several Well Child Centres
in a region were socio-economic living standards were relatively high as well. The percentage of
infants with signs of neonatal distress such as meconium staining (16%) or CTG abnormalities
(7.5%) was somewhat high, possibly because concerned parents of children with a history of minor
perinatal events have a stronger motivation to participate in developmental research.
The preterm group consisted of 59 infants who had been admitted to the neonatal intensive
care unit of the Beatrix Children’s Hospital of the University Medical Center in Groningen between
December 2003 and January 2005. Inclusion criteria were gestational age below 35 weeks, singleton
or twin, parents with appropriate understanding of the Dutch language and travel distance
between the child’s home and the hospital of <1 hour. Infants with severe congenital anomalies
101
7
General discussion
were excluded from the study. During the time interval mentioned above, 148 infants were eligible
for inclusion. Due to the limited capacity of our department for the intensive follow-up scheme,
there was a maximum number of infants that could be included per month for logistical reasons.
Therefore, not all of the eligible infants were approached for inclusion. By approaching parents at
random, 59 infants were eventually included. Unfortunately, it was not well recorded how many
parents of the eligible group actually were approached for the current study, how many refused
to participate and for which reasons. This may have introduced a non-response bias. Therefore,
we compared our sample with established reference groups of preterm infants to investigate
representativeness of the sample retrospectively29,30. We did �nd comparable gestational age, birth
weight, frequency of Apgar score at 5 minutes below 7 (15%) and the need for ventilatory support
(68%). Our preterm sample showed a relatively high percentage of infants small for gestational
age (34%) and Caesarean deliveries29,30. The latter could be due to the increased tendency over the
years to deliver very preterm infants by Caesarean section31. The high percentage of infants born
small for gestational age is common in tertiary referral centres, such as the NICU in the University
Medical Center (UMC) in Groningen. This is due to the preterm birth of intra-uterine growth retarded
children, in whom obstetrical interventions take place on the basis of fetal distress. For our study,
however, this may have induced selection of relatively more vulnerable children.
The preterm group had a lower socio-economic status than the term group, in accordance with
social disadvantage being considered a risk factor for preterm birth32. Attrition rate amongst the
preterm infants was higher than in the longitudinal term group, but still very acceptable (5% of
assessments were missed compared to <1% of assessments for the term group). All in all, strengths
of our study with respect to the study group were the mainly longitudinal character of the data and
the overall low attrition rate. Limitations were a possible selection bias of term infants with relatively
high socio-economic status and a fairly high percentage of small for gestational age preterm infants.
However, the study sample consisted of a heterogeneous group of infants, which is a valuable type
of sample when the study purpose is assessment of validity of a new instrument. We think that our
current study group does suit this goal, but should not be used e.g. to generate reference values for
IMP scores, because of a non-optimal recording of non-response in the preterm group and therefore
a possible bias.
Assessments
An important limitation of the study is that the assessors were not blinded to group allocation
with respect to term or preterm status of the infant. This was mostly for practically reasons, as
recruitment of infants, scheduling appointments and performing the assessments was for the
larger part done by KRH, with help of colleagues and master students. Even so, if assessors would
have been blinded to term or preterm status, group allocation could have become obvious to them
because of the characteristic appearance of preterm infants6. Still, as so many infants were enrolled
in the study, assessors were not aware of details on clinical and developmental history. During the
102
Chapter 7
scheduled assessment, a video-recording was made of spontaneous motor behaviour and an age-
speci�c neurological examination5-7 was carried out. IMP and AIMS were scored on the basis of the
video-recording in di!erent sessions, a long time after the data collection and the large part of
AIMS assessments were scored by master students (with adequate supervision). It is unlikely that
appraisal of the results of neurological examination, IMP and AIMS has mutually in"uenced each
other.
Future research
To date, the IMP has only been used in research settings. The results described in this thesis indicate
that the IMP is well able to discriminate between typically developing infants and infants at high
risk for developmental motor disorders. Predictive validity for neurological outcome at 18 months
was satisfactory in our study group. Due to these characteristics, addition of the IMP to clinical
follow-up of high-risk infants could be valuable. An important requirement for possible clinical
application of the IMP is the generation of norm-scores at di!erent ages in a large, representative
cohort of typically developing infants without relevant pre, peri and neonatal risk factors. Another
requirement is a precise description of all IMP items with corresponding de�nitions, which is also
indispensable for future training courses.
Some issues that need to be addressed in further research were already mentioned in the
earlier sections of this chapter. Two other interesting research topics are described below. First, it
would be of interest to investigate application of the IMP in other groups of infants with a high risk
for developmental motor problems, for example infants with term asphyxia or infants born late
preterm (gestational age between 34 and 36 weeks33). The last group is especially at risk for minor,
more subtle motor problems, which could possibly be detected already at young age with the IMP.
It would also be appealing to study how the IMP performs in other children with an atypical course
of motor development, e.g. children with inborn errors of metabolism34 or children with Prader-Willi
syndrome with related hypotonia35, and to investigate whether IMP scores in infancy can predict
future neurological and developmental outcome.
Second, it would be interesting to unravel the relation of IMP scores with (neonatal) brain
imaging. In the current thesis, the relation of IMP scores with results of neonatal brain ultrasounds
was examined, and we found that serious brain pathology on cranial ultrasound, such as cystic
periventricular leukomalacia (PVL36) or intraventricular haemorrhage (IVH37) grade III or IV was
clearly associated with lower IMP scores throughout infancy. Debate exists on whether magnetic
resonance imaging (MRI) is superior with respect to prediction of future developmental outcome
compared to neonatal cranial ultrasound scanning38. In a large study, abnormal �ndings on
magnetic resonance imaging (MRI) at term age were found to be strong predictors of adverse
neurodevelopmental outcome at two years of age39. Besides identifying di!use white matter injury,
which is common in very preterm infants and clearly associated with impaired motor outcome,
MRI is also able to detect speci�c abnormalities, for example abnormal myelination of the posterior
103
7
General discussion
limb of the internal capsule (PLIC), that are associated with impaired motor development38. Much
research on the relation between term MRI abnormalities and developmental outcome is still in
progress. It would be very interesting to use the IMP as a follow-up tool in this �eld, as we expect
that the IMP will not only be able to detect infants at high risk for major developmental disorders
such as CP, but could possibly also detect infants with minor developmental motor problems. Future
studies should investigate whether speci�c types of MRI abnormalities are associated with lower
IMP scores, especially lower scores on the variation domain. Low scores on the variation domain
might be a re�ection of impaired cerebral connectivity38, but further research is de�nitely needed
to elucidate this issue.
Concluding remarks
This thesis presents the Infant Motor Pro�le (IMP), a recently developed qualitative assessment of
motor behaviour for infants aged 3 to 18 months. Its psychometric properties including reliability,
construct validity, concurrent validity and predictive validity were investigated. With respect to
the three purposes for which the IMP was developed (chapter 6,), we can conclude that the IMP
is well able to discriminate between typically developing infants and infants with high risk for
developmental motor disorders. The ability of the IMP to evaluate motor function over time should
be further explored by applying the method in intervention studies. Prediction of developmental
motor outcome at an early age will always remain di!cult, since change is one of the main
characteristics of the developing nervous system. In our study population, predictive validity of the
IMP was satisfactory. Future studies should be directed at generating norm-scores for the IMP and
determining clinical applicability.
104
Chapter 7
REFERENCES1. Hadders-Algra M. The Neuronal Group Selection Theory: a framework to explain variation in normal motor
development. Dev Med Child Neurol 2000;42:566-72.
2. Hadders-Algra M. The Neuronal Group Selection Theory: promising principles for understanding andtreating developmental motor disorders. Dev Med Child Neurol 2000;42:707-15.
3. Hadders-Algra M. Variation and variability: keywords in human motor development. Submitted forpublication.
4. Piper MC, Darrah J. Motor assessment of the developing infant. Philadelphia: Saunders, 1994.
5. Touwen BCL. Neurological development in infancy. Clinics in Developmental Medicine No. 58. London:Mac Keith Press, 1976.
6. Hadders-Algra M, Heineman KR, Bos AF, Middelburg KJ. The assessment of minor neurological dysfunctionin infancy using the Touwen Infant Neurological Examination: strengths and limitations. Dev Med ChildNeurol, 2009, e-pub ahead of print.
7. Hempel MS. The neurological examination for toddler-age. PhD thesis. University of Groningen, 1993.
8. Hadders-Algra M. General movements: A window for early identi�cation of children at high risk fordevelopmental disorders. J Pediatr 2004;145:S12-8.
9. Heineman KR, Hadders-Algra M. Evaluation of neuromotor function in infancy - A systematic review ofavailable methods. J Dev Behav Pediatr 2008;29:315-23.
10. Piper MC, Pinnell LE, Darrah J, Maguire T, Byrne PJ. Construction and validation of the Alberta Infant MotorScale (AIMS). Can J Public Health. 1992;83 Suppl 2:S46-50.
11. Almeida KM, Dutra MV, Mello RR, Reis AB, Martins PS. Concurrent validity and reliability of the AlbertaInfant Motor Scale in premature infants. J Pediatr (Rio J) 2008;84:442-8.
12. Bayley N. Manual for the Bayley Scales of Infant Development: Second Edition. San Antonio: ThePsychological Corporation; 1995.
13. Folio MR, Fewell RR. Peabody Developmental Motor Scales: Examiner’s Manual. 2nd ed. Texas: Pro-ED; 2000.
14. Kuban K, Adler I, Allred EN, Batton D, Bezinque S, Betz BW, Cavenagh E et al. Observer variability assessingUS scans of the preterm brain: the ELGAN study. Pediatr Radiol 2007;37:1201-8.
15. Hintz SR, Slovis T, Bulas D, Van Meurs KP, Perritt R, Stevenson DK, Poole WK, et al. Interobserver reliabilityand accuracy of cranial ultrasound scanning interpretation in premature infants. J Pediatr 2007;150:592-6.
16. Van Gijn J, Bonke B. Interpretation of plantar re�exes: biasing e�ect of other signs and symptoms. J NeurolNeurosurg Psychiatry 1977;40:787-9.
17. Weisglas-Kuperus N, Baerts W, Smrkovsky M, Sauer PJ. E�ects of biological and social factors on thecognitive development of very low birth weight children. Pediatrics 1993;92:658-65.
18. Roberts G, Howard K, Spittle AJ, Brown NC, Anderson PJ, Doyle LW. Rates of early intervention services invery preterm children with developmental disabilities at age 2 years. J Paediatr Child Health 2008;44:276-80.
19. American Psychiatric Association. Diagnosis and Statistical Manual of Mental Disorders. 4th edn, TextRevision. Washington, DC: American Psychiatric Association, 2000.
20. Blauw-Hospers CH, de Graaf-Peters VB, Dirks T, Bos AF, Hadders-Algra M. Does early intervention in infantsat high risk for a developmental motor disorder improve motor and cognitive development? NeurosciBiobehav Rev 2007;31:1201-12.
21. Hielkema T, Blauw-Hospers CH, Dirks T, Bos AF, Hadders-Algra M: Does the type of intervention matter formotor development in infants at high risk for a developmental disorder? Revalidata 2009;151: 29.
22. Campbell SK, Kolobe TH, Osten ET, Lenke M, Girolami GL. Construct validity of the test of infant motorperformance. Phys Ther. 1995;75:585-96.
105
7
General discussion
23. Wilson PH, McKenzie BE. Information processing de�cits associated with developmental coordinationdisorder: a meta-analysis of research �ndings. J Child Psychol Psychiatry 1998;39:829-40.
24. Cooper J, Majnemer A, Rosenblatt B, Birnbaum R. The determination of sensory de�cits in children withhemiplegic cerebral palsy. J Child Neurol 1995;10:300-9.
25. Hadders-Algra M. Developmental coordination disorder: is clumsy motor behavior caused by a lesion ofthe brain at early age? Neural Plast 2003;10:39-50.
26. Hadders-Algra M. Two distinct forms of minor neurological dysfunction: perspectives emerging from areview of data of the Groningen Perinatal Project. Dev Med Child Neurol 2002;44:561-71.
27. Tieman BL, Palisano RJ, Sutlive AC. Assessment of motor development and function in preschool children.Ment Retard Dev Disabil Res Rev 2005;11:189-96.
28. Altman DG. Practical statistics for medical research. London: Chapman and Hall, p411-412, 1991.
29. Stoelhorst GM, Rijken M, Martens SE, Brand R, den Ouden AL, Wit JM, Veen S; Leiden Follow-Up Project onPrematurity. Changes in neonatology: comparison of two cohorts of very preterm infants (gestational age<32 weeks): the Project On Preterm and Small for Gestational Age Infants 1983 and the Leiden Follow-UpProject on Prematurity 1996-1997. Pediatrics 2005;115:396-405.
30. de Kleine MJ, den Ouden AL, Kollée LA, van Baar A, Nijhuis-van der Sanden MW, Ilsen A, Brand R, Verloove-Vanhorick SP. Outcome of perinatal care for very preterm infants at 5 years of age: a comparison between1983 and 1993. Paediatr Perinat Epidemiol 2007;21:26-33.
31. Haque KN, Hayes AM, Ahmed Z, Wilde R, Fong CY. Caesarean or vaginal delivery for preterm very-low-birth weight (< or =1,250 g) infant: experience from a district general hospital in UK. Arch Gynecol Obstet.2008;277:207-12.
32. Zeka A, Melly SJ, Schwartz J. The e�ects of socioeconomic status and indices of physical environment onreduced birth weight and preterm births in Eastern Massachusetts. Environ Health. 2008;7:60.
33. Davido� MJ, Dias T, Damus K, Russell R, Bettegowda VR, Dolan S, Schwarz RH, Green NS, Petrini J. Changesin the gestational age distribution among U.S. singleton births: impact on rates of late preterm birth, 1992to 2002. Semin Perinatol 2006;30:8-15.
34. Bruggink JL, van Spronsen FJ, Wijnberg-Williams BJ, Bos AF. Pilot use of the early motor repertoire in infantswith inborn errors of metabolism: outcomes in early and middle childhood. Early Hum Dev 2009;85:461-5.
35. Festen DA, Wevers M, Lindgren AC, Böhm B, Otten BJ, Wit JM, Duivenvoorden HJ, Hokken-Koelega AC.Mental and motor development before and during growth hormone treatment in infants and toddlerswith Prader-Willi syndrome. Clin Endocrinol 2008;68:919-25.
36. De Vries LS, Eken P, Dubowitz LM. The spectrum of leukomalacia using cranial ultrasound. Behav Brain Res1992;31;49:1-6.
37. Volpe JJ. Neurology of the Newborn. 4th ed. Philadelphia PA: WB Saunders, 2001.
38. Hintz SR, O’Shea M. Neuroimaging and neurodevelopmental outcomes in preterm infants. Semin Perinatol2008;32:11-9.
39. Woodward LJ, Anderson PJ, Austin NC, Howard K, Inder TE. Neonatal MRI to predict neurodevelopmentaloutcomes in preterm infants. N Engl J Med 2006;355:685-94.
106
107
Summary
108108
Summary
This thesis presents the Infant Motor Pro�le (IMP), a video-based qualitative assessment of motor
behaviour of infants between 3 and 18 months. The IMP was developed for three purposes:
discrimination between typically developing infants and infants with high risk for developmental
motor disorders, evaluation of motor function over time and prediction of developmental motor
outcome. The IMP is based on the Neuronal Group Selection Theory (NGST) which states that motor
development is characterised by exploration of the primary variable motor repertoire and selection
of adaptive motor strategies out of this repertoire. The IMP consists of 80 items that are grouped into
�ve domains: size of the repertoire (variation), adaptive selection (variability), movement �uency,
movement symmetry and motor performance.
Chapter 2 describes a systematic review of available methods to evaluate neuromotor function
in infancy and their psychometric properties. This review forms the background against which
the Infant Motor Pro�le was developed. We found that instruments that assess quality of motor
behavior or motor patterns, such as the General Movement method (GM) and the Test of Infant
Motor Performance (TIMP), have the best predictive validity for major and minor developmental
motor disorders. However, these methods are only applicable until the age of four months.
Chapter 3 presents the pilot study on the IMP with a description of the instrument and the �rst
data on reliability and validity. Concurrent validity of the IMP with the Alberta Infant Motor Scale
(AIMS) and with Touwen Infant Neurological Examination (TINE) was assessed in a heterogeneous
study group of low-risk term born and high-risk preterm born infants. Intra and interobserver
reliability of the IMP were satisfactory. Concurrent validity of the IMP with the AIMS was good.
The IMP was able to di!erentiate between infants with normal neurological condition, minor
neurological dysfunction and abnormal neurological condition.
In chapter 4, development of adaptive motor behaviour in typically developing infants is
investigated with the NGST as theoretical background. According to the NGST, infants learn to select
adaptive motor strategies out of their motor repertoire during motor development. This study
explored the ages at which clinically observable transition to adaptive motility emerged for four
speci�c motor functions: abdominal progression, sitting motility, reaching, and grasping in a group
of typically developing infants. In addition, reliability of observation of the presence of adaptive
motor behaviour was investigated. Results indicated that transition to adaptive motor behaviour
could be observed reliably. Clinically observable transitions to adaptive selection started to emerge
from six months onwards and peaked between eight and 15 months. Transitions developed
gradually and occurred at speci�c ages for di!erent motor functions. Comparison of our results to
literature showed that changes measured by neurophysiologic methods precede clinically observed
transitions.
The study described in chapter 5 investigates construct validity of the IMP, operationalized
as the relation of IMP scores throughout infancy with pre-, peri- and neonatal variables, including
the presence of brain pathology on neonatal ultrasound scans. In order to do so, a longitudinal
prospective study was performed in a mixed group of term born and preterm infants. Socio-
109109
Summary
economic and perinatal data were collected, which included for preterm infants information
on periventricular leukomalacia and intraventricular haemorrhage based on neonatal cranial
ultrasounds. Data were analyzed by �tting mixed e�ects models. In the total group of infants
gestational age, socio-economic status and 5-minute Apgar score were signi�cant determinants of
IMP scores throughout infancy. In the subgroup of preterms, IMP scores were signi�cantly a�ected
by brain lesions on neonatal ultrasound and socio-economic status. These �ndings support the
construct validity of the Infant Motor Pro�le: IMP scores were clearly associated with relevant
determinants of neuromotor function.
In chapter 6 reliability, concurrent validity of the IMP with the AIMS and withTINE and predictive
validity for neurological outcome at 18 months are assessed. A longitudinal prospective study was
performed in a heterogeneous group of term born and preterm infants. For the concurrent validity
part of the study, a second group of term infants was added with cross-sectional assessments. Inter-
observer reliability was high. Correlations between IMP and AIMS scores varied across IMP domains;
they were highest for the performance domain of the IMP. A clear relationship was found between
total IMP score and neurological condition. Sensitivity and speci�city for prediction of abnormal
neurological outcome at 18 months were satisfactory in the study sample.
With respect to the three purposes for which the IMP was developed, we can conclude that the
IMP is well able to discriminate between typically developing infants and infants with high risk for
developmental motor disorders. The ability of the IMP to evaluate motor function over time should
be further explored by applying the method in intervention studies. Prediction of developmental
motor outcome at an early age will always remain di!cult, since change is one of the main
characteristics of the developing nervous system. In our study population, predictive validity of the
IMP was satisfactory. Future studies should be directed at generating norm-scores for the IMP and
determining clinical applicability.
110
111
Appendices
112112
Appendix I
APPENDIX I: SUMMARY ON THE DOMAINS AND ITEMS OF THE INFANT MOTOR PROFILE
The Infant Motor Pro�le consists of �ve domains. The items and scoring methods of each domain
are listed below. IMP scores are calculated for each of the �ve domains according to the following
formula:
Sum of item scores
Score = * 100%
(number of items of domain – number of items not assessed) * maximum score of items
Items are not assessed when a certain motor behaviour is not observed (e.g. the item ‘variation of
arm movements during walking’ is not assessed when a child is not able to walk). ‘Maximum score
of items’ is the maximum score among the items of the domain. For the domains variation-size of
repertoire, variability-ability to select and �uency the maximum score for each item is 2 and for the
domain symmetry the maximum score is 3. Because the performance domain contains items with
di�erent numbers of maximum scores, scores are weighted before computing the performance
domain score. The total IMP score is computed by summing the scores on the �ve domains and
dividing this number by 5. All scores on the domains and the total IMP score are expressed as a
percentage, with a maximum score of 100%.
Domain 1: Variation – size of repertoire
Scoring for variation-items is as follows:
1. Insu!cient variation.
2. Su!cient variation.
Items of this domain:
1) Variation of head movements (Sup)
2 ) Posture, presence of ATNR (Sup) P/F N/Oa
3) Posture, presence of hyperextension of neck and trunk (Sup) P/F N/O
4) Variation of arm movements (Sup)
5) Variation of �nger movements (Sup)
6) Variation of leg movements (Sup)
7) Variation of toe movements (Sup)
8) Variation of reaching or prereaching movements of the arms (Sup)
9) Variation of hand motility during reaching, grasping and manipulation(Sup)
10) Variation of head movements (P)
11) Variation of pre-crawling movements of the legs (P)
12) Variation of wriggling, pivoting or crawling (P)
13) Variation in sitting motility (Sit)
14) Variation in sitting up behaviour (Sit)
15) Bottom shu"ing (Sit) P/F N/Oa
16) Variation in standing up behaviour (S&W)
113113
Appendix I
17) Variation of arm movements (S&W)
18) Variation of trunk movements (S&W)
19) Variation of leg movements (S&W)
20) Variation of placing of feet (S&W)
21) Variation of prereaching or reaching movements of the arms(RGM)
22) Variation of hand motility during reaching and grasping (RGM)
23) Facial expression (Gen)
24) Drooling (Gen) M Nb
25) Presence of stereotyped tongue protrusion (Gen) Yes No
Exceptions to the standard scoring are denoted in the second and third column: a P/F = persistent or
frequent, N/O = no or occasionally, b M = marked drooling, N = no or little drooling
Domain 2: Variability – ability to select
All items of this domain are scored as follows:
1. No selection.
2. Adaptive selection.
Items of this domain:
1) Variability of head movements: ability to make an adaptive selection (Sup)
2) Variability of reaching or prereaching movements of the arms: ability to make an adaptive
selection (Sup)
3) Variability of hand motility during reaching, grasping and manipulation: ability to make an
adaptive selection (Sup)
4) Variability of head movements: ability to make an adaptive selection (P)
5) Variability of wriggling, pivoting or crawling: ability to make an adaptive selection (P)
6) Variability in sitting motility: ability to make an adaptive selection (Sit)
7) Variability in sitting up behaviour: ability to make an adaptive selection (Sit)
8) Variability in standing up behaviour: ability to make an adaptive selection (S&W)
9) Variability of arm movements: ability to make an adaptive selection (S&W)
10) Variability of trunk movements: ability to make an adaptive selection (S&W)
11) Variability of leg movements: ability to make an adaptive selection (S&W)
12) Variability of placing of feet: ability to make an adaptive selection (S&W)
13) Variability of prereaching or reaching movements of the arms: ability to make an adaptive
selection (RGM)
14) Variability of hand motility during reaching and grasping: ability to make an adaptive selection
(RGM)
15) Facial expression: ability to make an adaptive selection (Gen)
114114
Appendix I
Domain 3: Symmetry
All items of this domain are scored as:
1. Strong asymmetry.
2. Moderate asymmetry.
3. No or mild asymmetry.
Items of this domain:
1) Position of head (Sup)
2) Reaching, grasping and manipulation of objects (Sup)
3) Position of head (P)
4) Arm posture and motility during activity in prone (P)
5) Position of head during sitting (Sit)
6) Position of trunk during sitting or supported sitting (Sit)
7) Arm posture and motility during sitting or supported sitting (Sit)
8) Arm posture and motility during independent walking (S&W)
9) Leg posture and motility during independent walking (S&W)
10) Reaching, grasping and manipulation of objects (RGM)
Domain 4: Fluency
All !uency items of this domain are scored as:
1. Majority of movements non-!uent.
2. Majority of movements !uent.
All items with presence or absence of tremor are scored as:
1. Frequently tremor present.
2. No or occasionally tremor present.
Items of this domain:
1) Tremor during reaching or prereaching (Sup)
2) Fluency of motor behaviour in supine (Sup)
3) Fluency of motility during independent walking (S&W)
4) Tremor during reaching or prereaching (RGM)
5) Fluency of motility during reaching or prereaching (RGM)
6) Tremor (Gen)
7) Fluency of motor behaviour (Gen)
115115
Appendix I
Domain 5: Performance
Performance items are not scored in a standard fashion, but are adapted to the speci�c motor
function and range from two to seven options.
Items of this domain:
1) Control of head movements (Sup)
2) Manipulative behaviour of hands and �ngers (Sup)
3) Tilting of pelvis (Sup)
4) Rolling from supine to prone (Sup)
5) Reaching, grasping and manipulation of objects (Sup)
6) Head lift in prone (P)
7) Functional ability of shoulder girdle (P)
8) Functional ability of hands (P)
9) Rolling from prone to supine (P)
10) Progression in prone: development of crawling (P)
11) Control of head movements (Sit)
12) Sitting ability (Sit)
13) Position of trunk, preference posture (Sit)
14) Need of arm support (Sit)
15) Sitting up (Sit)
16) Standing ability (S&W)
17) Standing up (S&W)
18) Walking (S&W)
19) Balance during independent walking (S&W)
20) Arm posture and motility (S&W)
21) Heel-toe gait (S&W)
22) Reaching, grasping and manipulation of objects (RGM)
23) Type of grasping (RGM)
Sup = supine, P = prone, Sit = sitting, S&W = standing and walking, RGM = reaching, grasping and
manipulation, Gen = general
116116
Appendix II
APP
END
IXIIA
:EX
TEN
DED
VER
SIO
NO
FTA
BLE
IIIA
(CH
APT
ER2)
:VA
LID
ITY
OF
THE
TEST
S
Ass
essm
ent
(sho
rtna
me)
Cons
truc
tval
idit
yCo
ncur
rent
valid
ity
Pred
icti
veva
lidit
yfo
rCP
orm
inor
deve
lopm
enta
ldi
sord
ers
(e.g
.DCD
,MN
D)
Touw
ennd
and
aG
roup
of37
high
and
low
risk
infa
nts:
sens
itivi
tyof
Touw
enas
sess
men
tat3
to4
mo
ford
evel
opm
ento
fco
mpl
exM
ND
atag
e1.
5yr
sis
75%
(3/4
),sp
eci�
city
94%
;gro
upof
52hi
ghan
dlo
wris
kin
fant
s:se
nsiti
vity
ofTo
uwen
asse
ssm
enta
t3to
4m
ofo
rde
velo
pmen
tofc
ompl
exM
ND
at4
to9
yrs
44%
(4/7
),sp
eci�
city
78%
19,7
4
+
Amie
l-Tis
onSe
nsiti
vity
tode
tect
child
ren
with
ultr
asou
ndab
norm
aliti
esof
the
brai
n0.
97(3
5/36
);ch
ildre
nw
ithab
norm
aliti
eson
cere
bral
func
tion
mon
itorin
g0.
89(2
3/26
);ch
ildre
nw
ithEE
Gab
norm
aliti
es0.
88(1
6/18
)42
+G
roup
of54
PTin
fant
s,as
sess
men
tat
term
age
and
3m
o,co
ncur
rent
valid
ityw
ithG
Mm
etho
d:ka
ppa
0.87
and
0.54
43
+D
ata
on15
2PT
child
ren:
sens
itivi
tyof
Am
iel-T
ison
neur
olog
ical
asse
ssm
enta
t1yr
and
cogn
itive
and
mot
orou
tcom
eat
4yr
sis
94%
(16/
17))
44,4
5 ,pr
obab
ility
ofth
ene
edfo
rext
raed
ucat
iona
lpr
ovis
ion
at8
yrs
9-87
%45
and
sens
itivi
tyfo
rne
urod
evel
opm
enta
lou
tcom
eat
14-1
5yr
s86
%(1
2/14
),sp
eci�
city
33%
(39/
120)
.46
45PT
infa
nts,
Am
iel-T
ison
neur
olog
ical
asse
ssm
ent
atte
rmag
ean
d3
mo,
neur
olog
ical
outc
ome
at12
-15
mo:
PPV
0.68
and
0.83
,sen
sitiv
ity0.
92an
d1.
00,
spec
i�ci
ty0.
45an
d0.
7543
+
Mus
cle
pow
erCl
eard
i!er
ence
sin
optim
ality
ofm
uscl
epo
wer
regu
latio
nbe
twee
nPT
and
FTin
fant
san
din
fant
sw
ithdi
!ere
ntgr
ades
ofris
ksc
ore
incl
udin
ggr
ade
ofPV
L47,4
8
++nd
a44
infa
nts
with
PVF
orPV
L,co
mbi
ned
resu
ltsof
mus
cle
pow
erde
velo
pmen
tin
shou
lder
san
dtr
unk
at3
mo
toge
ther
with
shou
lder
sat
6m
o:se
nsiti
vity
topr
edic
tneu
rom
otor
outc
ome
at18
mo
85%
and
spec
i�ci
ty79
%49
+
117117
Appendix II
Ass
essm
ent
(sho
rtna
me)
Cons
truc
tval
idit
yCo
ncur
rent
valid
ity
Pred
icti
veva
lidit
yfo
rCP
orm
inor
deve
lopm
enta
ldi
sord
ers
(e.g
.DCD
,MN
D)
HINE
Goo
dco
rrel
atio
nw
ith�n
ding
son
MRI
scan
50+
nda
Dat
aon
74PT
infa
nts,
asse
ssm
enta
tCA
6-15
mo;
53FT
infa
nts
with
hypo
xic-
isch
emic
ence
phal
opat
hy,
asse
ssm
enta
tage
9-14
mo;
24in
fant
sw
ithcy
stic
PVL,
asse
ssm
enta
tCA
6-9,
5m
o:go
odpr
edic
tion
oflo
com
otor
func
tion
atag
e2
yrs.
and
4yr
s.50-5
2
+
PRP
Theo
retic
alas
sum
ptio
n:se
lect
edpr
imiti
vere
!exe
sar
ere
late
dto
onse
tof
mot
orfu
nctio
nin
norm
alan
dbr
ain-
dam
aged
child
ren;
assu
mpt
ion
notv
alid
ated
.N
ore
latio
nbe
twee
npr
imiti
vere
!exe
san
dm
otor
deve
lopm
entw
asfo
und53
-G
roup
of16
5TD
child
ren:
noco
rrel
atio
nbe
twee
nA
IMS
and
PRP-
scor
eat
ages
6w
ks.,
3m
oan
d5
mo53
-nd
a
Infanib
Clea
rdi"
eren
ces
insc
ores
once
rtai
nne
urom
otor
cate
gorie
sbe
twee
nhe
alth
yFT
and
PTan
dill
PTin
fant
s54
+nd
a20
9PT
LBW
infa
nts:
spec
i�ci
tyof
mot
orev
alua
tion
with
Infa
nib
at7
mo
rega
rdin
gCP
atag
e36
mo
isun
satis
fact
ory,
norm
alne
urom
otor
asse
ssm
enta
t7m
ois
high
lypr
edic
tive
ofsu
bseq
uent
norm
alm
otor
func
tion
(spe
ci�c
ity10
0%(1
11/1
11))55
±
BSID-II
Cons
truc
tsm
ore
di"e
rent
iate
dw
ithag
e29
±G
roup
of44
TDan
dlo
w-r
isk
PTin
fant
s,as
sess
men
tat1
2,15
and
18m
o,co
ncur
rent
valid
ityw
ithPe
abod
ygr
oss
mot
orsc
ores
:r=
.78
tor=
.96
and
with
Peab
ody
�ne
mot
orsc
ores
:r=
.25
tor=
.5756
±nd
a
118118
Appendix II
Ass
essm
ent
(sho
rtna
me)
Cons
truc
tval
idit
yCo
ncur
rent
valid
ity
Pred
icti
veva
lidit
yfo
rCP
orm
inor
deve
lopm
enta
ldi
sord
ers
(e.g
.DCD
,MN
D)
PDMS-II
Subt
ests
core
sco
rrel
ate
with
age56
±G
roup
of44
TDan
dlo
w-r
isk
PTin
fant
s,as
sess
men
tat1
2,15
and
18m
o,co
ncur
rent
valid
ityw
ithBS
ID-II
ofgr
oss
mot
orsc
ores
:r=
.78
tor=
.96
and
of�n
em
otor
scor
es:r
=.2
5to
r=.5
756
±44
FTan
dlo
wris
kPT
infa
nts,
PDM
Sat
12m
o:pr
edic
tion
ofsc
ores
onBS
IDm
otor
scal
esat
18m
or=
.54
(PT)
and
r=.5
6(F
T),P
DM
Sat
15m
o:r=
.59
(FT)
and
r=.6
6(P
T)56
±
MAI
nda
nda
246
high
-ris
kin
fant
sM
AI4
mo,
corr
elat
ion
with
Bayl
eym
otor
scal
eat
age
2yr
sr=
-0.3
7(p
<0.0
01),
75PT
infa
nts
MA
Iat4
mo
CAse
nsiti
vity
and
spec
i�ci
tyfo
rdev
elop
men
talo
utco
me
at18
mo
resp
.61%
(14/
23)a
nd83
%(4
3/52
)57-6
2
77hi
ghris
kin
fant
s,M
AIa
t4m
oan
dgr
oss
and
�ne
mot
orev
alua
tions
(PD
MS
and
Fros
tigEy
eM
otor
Coor
dina
tion
subt
est)
at4,
5yr
s:Sp
earm
an’s
rho
resp
.-0.
12(n
.s.)a
nd-0
.23
(p≤0
.05)
60
±
NBI
Clea
rdis
tinct
ion
insc
ores
betw
een
FTan
dPT
infa
nts54
,63
+nd
and
a
TIME
Child
ren
with
mot
orde
lays
di"e
rfr
omch
ildre
nw
ithou
tdel
ays
inse
quen
ces
ofm
ovem
ents
whe
nas
sess
edfr
omse
vera
lsta
rtin
gpo
sitio
ns64
±nd
and
a
119119
Appendix II
Ass
essm
ent
(sho
rtna
me)
Cons
truc
tval
idit
yCo
ncur
rent
valid
ity
Pred
icti
veva
lidit
yfo
rCP
orm
inor
deve
lopm
enta
ldi
sord
ers
(e.g
.DCD
,MN
D)
AIMS
Ata
ge8
mos
.cle
ardi
�ere
ntia
tion
betw
een
neur
olog
ical
lyno
rmal
,su
spec
tora
bnor
mal
infa
nts65
±12
0TD
and
68de
laye
din
fant
s,co
ncur
rent
valid
ityw
ithBS
IDm
otor
scal
e:r=
.97
and
.93;
with
PDM
Sgr
oss
mot
orsc
ale:
r=.9
9an
dr=
.9566
++17
3hi
gh-r
isk
infa
nts,
pred
ictiv
eva
lidity
ofA
IMS-
scor
eat
4m
ofo
r18
mo
neur
olog
ical
outc
ome:
PPV
=40
%(s
ensi
tivity
77%
,spe
ci�c
ity82
%)
AIM
S-sc
ore
at8
mo:
PPV
=66
%(s
ensi
tivity
87%
,sp
eci�
city
93%
)66
±
SOMP-I
Clea
rdi�
eren
ces
inqu
ality
ofm
otor
perf
orm
ance
betw
een
FTan
dPT
infa
nts68
+nd
and
a
TIMP
TIM
Psc
ores
are
sens
itive
toch
ange
sin
infa
nts’
mot
orpe
rfor
man
cedu
eto
mat
urat
ion
and
med
ical
com
plic
atio
ns37
Goo
ddi
scrim
inat
ion
betw
een
TDin
fant
s,in
fant
sw
ithCP
and
infa
nts
with
deve
lopm
enta
ldel
ay70
+A
sses
smen
tat3
mos
.,TIM
Pid
enti�
ed80
%of
the
sam
ein
fant
sas
the
AIM
S72
+96
infa
nts,
TDan
dhi
gh-r
isk,
pred
ictiv
eva
lidity
ofTI
MP-
scor
eat
3m
ofo
rthe
AIM
S-sc
ore
at12
mo
r=
.5572
61TD
and
risk-
infa
nts,
pred
ictiv
eva
lidity
ofTI
MP-
scor
eat
3m
ofo
rPD
MS-
IIat
4-5
yrs.
r=.6
9,PP
V=
92%
(sen
sitiv
ity62
%,
spec
i�ci
ty97
%)71
+
120120
Appendix II
Ass
essm
ent
(sho
rtna
me)
Cons
truc
tval
idit
yCo
ncur
rent
valid
ity
Pred
icti
veva
lidit
yfo
rCP
orm
inor
deve
lopm
enta
ldi
sord
ers
(e.g
.DCD
,MN
D)
GMs
Pres
ence
ofab
norm
alG
Msi
slin
ked
todi
scer
nabl
ele
sion
sof
the
brai
n40an
dpr
e-,p
eri-
and
neon
atal
adve
rsiti
essu
chas
PTbi
rth
and
IUG
R41
,74
++M
ixed
grou
pof
80hi
ghan
dlo
wris
kin
fant
s,as
sess
men
tsat
0-4
mo,
conc
urre
ntva
lidity
with
neur
olog
ical
exam
inat
ion:
Spea
rman
rho’
s0.
32-0
.5541
58FT
infa
nts
with
HIE
and
66PT
:agr
eem
entb
etw
een
GM
and
neur
olog
ical
exam
inat
ions
80%
75,7
6
+52
low
and
high
risk
infa
nts,
GM
-ass
essm
enta
t2-4
mo,
neur
olog
ical
follo
w-u
pat
4to
9ye
ars:
abno
rmal
GM
sat
�dge
tyag
eha
vese
nsiti
vity
ford
evel
opm
ent
ofCP
88%
(7/8
)and
spec
i�ci
ty10
0%41
,mild
lyab
norm
alG
Ms
are
rela
ted
tode
velo
pmen
tof
MN
D,A
DH
Dan
dag
gres
sive
beha
viou
rat4
-9yr
s(s
peci
�city
58%
)41an
dne
urol
ogic
alco
nditi
on(n
orm
al,s
impl
eM
ND
and
com
plex
MN
D)a
t9-1
2yr
s(r
ho0.
46,p
<0.0
1)78
130
FTan
dPT
infa
nts,
GM
asse
ssm
enta
t6-2
0w
ks,
neur
olog
ical
outc
ome
at2
yrs:
sens
itivi
tyfo
rCP
98%
(43/
44),
spec
i�ci
ty96
%(6
7/70
)40,7
7
+
BSID
=Ba
yley
Scal
esof
Infa
ntD
evel
opm
ent,
CP=
cere
bral
pals
y,H
IE=
hypo
xic
isch
emic
ence
phal
opat
hy,I
UG
R=
intr
aut
erin
egr
owth
reta
rdat
ion,
MN
D=
min
orne
urol
ogic
aldy
sfun
ctio
n,m
o=
mon
th(s
),nd
a=
noda
taav
aila
ble,
PDM
S=
Peab
ody
Dev
elop
men
talM
otor
Scal
e,PP
V=
posi
tive
pred
ictiv
eva
lue,
PVF
=pe
riven
tric
ular
�arin
g,PV
L=
periv
entr
icul
arle
ukom
alac
ia,T
D=
typi
cally
deve
lopi
ng,V
LBW
=ve
rylo
wbi
rth
wei
ght,
yrs.
=ye
ars,
++=
very
good
,+=
good
,±=
mod
erat
e,-=
poor
121121
Appendix II
APPENDIX IIB: EXTENDED VERSION OF TABLE IIIB (CHAPTER 2): RELIABILITY OF THE TESTS
Assessment Intra-observer agreement Interobserver agreement
Touwen nda nda
Amiel-Tison nda Only data available on passive muscle toneitems: reliability excellent to poor for variousmanoeuvres20:
Muscle power nda nda
HINE nda Correlation coe!cient close to 122: ++
PRP nda Agreement 72 to 95%24: +
Infanib nda nda
BSID-II nda r = 0.75 (motor), r = 0.96 (mental)29: +
PDMS-II nda r = 0.9631: ++
MAI Kappa: 10% of items excellentreliability, 42% fair to good, 48%poor reliability61:
Overall score: r = 0.7262 item reliability: kappa: 2%of items excellent reliability, 58% fair to good, 40%poor reliability61:
NBI nda nda
TIME nda r = 0.88 – 0.9964: ++
AIMS r > .9967: ++ r > .9967: ++
SOMP-I Agreement 62%69: Agreement 80%69: +
TIMP nda Agreement 95%73: +
GMs Kappa 0.841,74: ++ Kappa 0.8 – 0.940,41,74: ++
nda = no data available, r = Spearman’s rho
122122
Appendix III
APPENDIX III: INFANT MOTOR PROFILE
!"
#%&'
()*+,*(++-./01234"567(++-8#9#:01;#<
6!:=
#>30:0"2.5661?&&@
@@AB#9#:01;#<
6!:<#>
30:0"2AC0;
D>14<#
D>14<#
D>14<#
D>14<#
!"
#$
&'()*+!)#$
,--#--"
#.)+!)#$
/01((#2)#+3,4#$
,--#--1($
01.)(1516*#!+"17#"
#.)-
! !!!
!
"#$$
%&"%
$&'%
()*#
+,
%-*,
*$&.
/
"#$
"%$&
'%()
*#+
,%-
*,*$
&.&%
#(0,
0&*+
*1&*
$&2
"#$
34((5
"%$&
'%()
*#+
,%-
*,*$
&.
!8!('!9'5'):16*#!+
"17#"#.)-
"#
"#
"#
"#
!
0$.4
330"0
*$&(5
-#'0#
6(*
/
.4330
"0*$
&(5-#
'0#6(
*
"8!('!9'5'):16*#!+
"17#"#.)-$!9'5'):)1
"!;#!.
!+!<)'7#-#5#2)'1."$
"$
"$
"$
!
$%.*
(*"&
0%$
/
#+#7
&0-*
.*(*
"&0%
$
#=1-')'1.16*#!+><(#7!'5'.4
*#!+
<1-')'1.)11.#-'+#
% %%%
!
.&'%
$8(5
7'*-
#0(0$
8)*
#+7%
.0&0%
$&%
&)*
9:
;/
,%+
*'#&
*(5
7'*-
#0(0$
8)*
#+7%
.0&0%
$&%
&)*
9:
;2
$%%'
,0(+
(57'
*-#0
(0$8
)*#+
7%.0
&0%$
&%%$
*.0
+*
$=1-)?(#><(#-#.2#16,@ A
"#
"#
"#
"#
!
7*'.
0.&*
$&(5
#$+
%6(08
#&%'
5%"
"4''0
$8<
=>9
/
3'*?4
*$&(5
%""4
''0$
8<
=>9
@64&
0$3#
$&"#
$#(
.%.)
%A%&
)*'7
%.&4
'*.
2
$%<
=>9
%'%"
"#.0
%$#(
(5$%
$B%6
(08#&
%'5
<=>
9
%=1-)?(#><(#-#.2#16*:<#(#B)#.-'1.
16.#2;!.+)(?.;
"#
"#
"#
"#
!
7*'.
0.&*
$&)5
7*'*
1&*$
.0%$
/
3'*?4
*$&(5
%""4
''0$
8)5
7*'*
1&*$
.0%$
2
$%%'
'#'*
(5)5
7*'*
1&*$
.0%$
&C!.'<?5!)'7#9#*!7'1?(16*!.+-!.+
6'.4#(-
! !!!
!
$%,
#$07
4(#&
0-*
6*)#
-0%4
'@*1
"(4+
0$8
&)4,
6.4
"C0$
8/
,#$
074(
#&*.
"(%&
)*.
:A0&)
)#$+
.0$
,0+
(0$*
C$**
.:3
**&%
'74&
.)#
$+D.
E0$
&%,
%4&)
'8!('!9'5'):16!(""17#"#.)-
"#
"#
"#
"#
!
0$.4
330"0
*$&(5
-#'0#
6(*
/
.4330
"0*$
&(5-#
'0#6(
*
(8!('!9'5'):166'.4#("
17#"
#.)-
"#
"#
"#
"#
!
0$.4
330"0
*$&(5
-#'0#
6(*
/
.4330
"0*$
&(5-#
'0#6(
*
)@'5)'.4
16<#57'-
! !!!
!
$%&0(
&0$8
%37*
(-0.
/
&0(&.
7*(-
0.@6
4&$%
&0$
#.4
")#
A#5
&)#&
)#$+
.,
#56*
#6(*
&%&%
4")
C$**
.2
&0(&.
7*(-
0.0$
.4")
#A
#5&)
#&)#
$+.
,#5
6*#6
(*&%
&%4"
)C$
**.
F
)#$+
.7(
#5A
0&)3*
*&G
74&.
&%*.
0$&%
,%4
&)
*8!('!9'5'):165#4"17#"
#.)-
"#
"#
"#
"#
!
0$.4
330"0
*$&(5
-#'0#
6(*
/
.4330
"0*$
&(5-#
'0#6(
*
+8!('!9'5'):16)1#"17#"
#.)-
"#
"#
"#
"#
!
0$.4
330"0
*$&(5
-#'0#
6(*
/
.4330
"0*$
&(5-#
'0#6(
*
,A155'.46(1"
-?<'.#)1<(1.#
! !!!
!
$%&4
'$0$
8%'
'%((0$
8#&
&*,
7&.
/
,#C
*.HA
088(
0$8I
,%-
*,*$
&.A
0&)&)
*7*
(-0.
@64&
+%*.
$%&'
%((&
%.0
+*2
'%((.
&%.0
+*@4
$0(#
&*'#
((5
9:
;F
'%((.
&%.0
+*@6
0(#&*
'#((5
G
&4'$
.4$
0(#&*
'#((5
0$&%
7'%$
*
9:
;J
&4'$
.60
(#&*
'#((5
0$&%
7'%$
*
-A#!2*'.4>4(!-<'.4
!.+"!.'<?5!)'1.16
19D#2)-
! !!!
!
+%*.
$%&'
*#")
&%A
#'+.
%6K*
"&/
+%*.
$%&'
*#")
@64&
.)%A
.7'
*'*#
")0$
8,
%-*,
*$&.
2
'*#"
)*.
&%A
#'+.
%6K*
"&64
&+%
*.$%
&8'#
.70&
F
'*#"
)*.
&%A
#'+.
@8'#
.7.
#$+
)%(+
.%6
K*"&
@64&
+%*.
$%&,
#$07
4(#&
*%6
K*"&
G
'*#"
)*.
&%A
#'+.
@)%(
+.#$
+,
#$07
4(#&
*.!
%6K*
"&J
'*#"
)*.
&%A
#'+.
@)%(
+.#$
+,
#$07
4(#&
*./
%6K*
"&.
L
'*#"
)*.
&%A
#'+.
@)%(
+.#$
+,
#$07
4(#&
*.
2%6
K*"&
.
E<F!<6G0603 30F4:#
E<F!<6G0603 30F4:#
E<F!<6G0603 30F4:#
E<F!<6G0603 30F4:#
8#9#:01;#<
6!:=
#>30:0"2
H<49#3I462G#B
4C!:/#<
6#3J30<4<"#<
KL5#=#65#3:!<BI
123123
Appendix III
!"
#%&'
%()*+)%**,-./0123"456%**,7#8#9/0:#;
5!9<
#=2/9/"1-4550>&&?
??@A#8#9/0:#;
5!9;#=
2/9/"1@B/:
!
"#$%
&'()
(*#+
,&'(
#'-
.#'
&,/0
#1&2
'23
245"
$1+)
,*"+
"'$"
23#+
6..
"1*6
! !!!
!!"
#$%&
'(!)
*"+
$*$!
,-.#
$-$!
,"#
"!/0
.#$#
$%&'
(!)
*"+
$*$!
,-.#
$-$!
,1!
-,#"
!)%-
0**
$,#0
2!
34!
56
"#-,
-(7$
8!
*"7
$#%,
$%-
0**
$,#0
2!
34!
56
"#-,
-(7$
9!
!""#
*(/7
%-0*
*$,
#0
!7#
*&
0&16
23*"
#$%&
'(2*
,*"*
"#$%
&'(
.28
"."'
1+23
1%"
#*.
+"#
"#
"#
"#
!,'
$&'
(/77"
$-!"
,-'"
6.#
$#$%
&'(!
)"#
#$%&
'(!)
*"+
$*$!
,-1!
(!-:
;;(&(
$!,/0
+%#(%
</$
8!
-:;;(
&($!
,/0+%
#(%</
$
"7#
*&
0&16
23,*
"*"#
$%&'
(2*
*"#$
%&'(
.28
"."'
1+23
1%"
#*.
+9#4
&0&16
12.
#:"
#'#-
#,1&8
"+"
0"$1
&2'
"$
"$
"$
"$
!,'
$&'
(/77"
$-!"
,-'"
6.#
$#$%
&'(!
)"#
#$%&
'(!)
*"+
$*$!
,-1!
!"-$
/$&,
("!
8!
%7%.
,(+$
-$/$
&,("
!
#7#
*&
0&16
23%#
'-.
21&0&
16-/
*&'(
*"#$
%&'(
)(*#
+,&'
(#'
-.
#'&,
/0#1
&2'
"#
"#
"#
"#
!!"
#$%&
'(!)
*"+
$*$!
,-.#
$-$!
,"#
"!/0
.#$#
$%&'
(!)
*"+
$*$!
,-.#
$-$!
,1!
(!-:
;;(&(
$!,/0
+%#(%
</$
8!
-:;;(
&($!
,/0+%
#(%</
$
$7#
*&
0&16
23%#
'-.
21&0&
16-/
*&'(
*"#$
%&'(
)(*#
+,&'
(#'
-.
#'&,
/0#1
&2'9
#4&0&
1612
.#:
"#'
#-#,
1&8"
+"0"
$1&2
'"$
!!"
#$%&
'(!)
*"+
$*$!
,-.#
$-$!
,"#
"!/0
.#$#
$%&'
(!)
*"+
$*$!
,-.#
$-$!
,1!
!"-$
/$&,
("!
8!
%7%.
,(+$
-$/$
&,("
!
%;*
".2*
-/*&'
(*"
#$%&
'(2*
,*"*
"#$%
&'(
% %%%
!,'
$&'
(/77"
$-!"
,-'"
6.#
$#$%
&'(!
)"#
#$%&
'(!)
*"+
$*$!
,-1!
;#$=:
$!,/0
.#$-
$!,2
7$-&
#(<$
,0.$
>
??
??
??
??
??
??
?@@
8!
!""#
"&&%
-("!
%//0
,#$*
"#.#
$-$!
,27$
-&#(<
$,0
.$>
??
??
??
??
??
??
?@@
&<0
/"'$
623
.21
2*4"
%#8&
2/*&
'+/
,&'"
% %%%
1!
*%A
"#(,0
";*
"+$*
$!,-
!"!B
;/:$!
,>-,
(;;2
A$#C
02;/"
..04
-/:)
)(-'
2",'
$#6
(-$>
??
??
??
??
??
??
?@@
8!
*%A
"#(,0
";*
"+$*
$!,-
;/:$!
,
C=03;#B/;53;=#A
C=03;#B/;53;=#A
C=03;#B/;53;=#A
C=03;#B/;53;=#A
124124
Appendix III
!"
#%&'
()*+,*(++-./01234"567(++-8#9#:01;#<
6!:=
#>30:0"2.5661?&&@
@@AB#9#:01;#<
6!:<#>
30:0"2AC0;
!
"#$
%&'(&
)*+
,)"
! !!!
!
"#$%
&#'()*'#+
',+&
-$."
/!
"#$%
',+&
-$."
'#%)"$
0#%)')#&1)'-
.2)&)2.(()*'#
*'-$
-$."
3!
()*'%-$."
*#+.
*$1%$4#&"%5,'&
#(#&6
$+7!
2.)&'.)&%
-$."
()*'$"*#+.
'($.%' 8
%$4#&"%95,'-
.%%#2$")**)4,(':
)&(##;)&6.+#,&"
<!
2.)&'.)&%
-$."
()*'$".&
"(##;%.+#,
&"
!-,
.&(&,
),'
/"#$
0*+"
.")1
",'
*+"2
#&%&)
3/"
#$*,
.&(&,
)(,
,)"
.&$"
" """
!"#
$%&#
'+#'.'$-$
.")&'#.%)"$
0#%)')#&
!
%'+#&6
(:0+$=.)()&6-$."
0#%)')#&'#
'-$!
>?!
@/!
2#"$+.'$(:0+$=.)()&6-$
."0#%)')#&'#
'-$!
>?!
@3!
&##+
2)("(:0+$=.)()&6-$."
0#%)')#&'#
#&$%)"$
"4#
+&
%&(6
,'/"
#$7
,2"7
")(.
#$
#$
#$
#$
!
)&%,**)4)$&
'(:=.+).5($
/!
%,**)4)$&
'(:=.+).
5($
#4#
+&
%&(6
,'/"
#$7
,2"7
")(.
8#5&
%&(6
(,7
#9"
#)#$
#*(&2
"."
%"1(
&,)
#%
#%
#%
#%
!
&#%$($4')#&
/!
.".0')=$%$($4')#&
$:;
)1(&,
)#%#
5&%&(
6,'
./,;
%$"+
3&+$
%"&)
*+,)
"! !!!
!
"#$%
&#',%$
.+2%.&
"-.&"
%'#2#=$
-$."
.&"'-#+.A
,0/!
,%$%
.+2%.&
"-.
&"%'#2#=$-$."
.&"'-#+.A
,095
,'"#
$%&#'%,44$$"
)&.4')=$$(5#
1.&
"(#1$+
.+2%,00
#+'
3!
)%%,00#+'$"5:
$(5#
1%.&"(#1$+
.+2%
7!
()*'%,00$
+0.+'#*'-#+.A5:
B%'.&")&6C
#&-.&"
%.&
"$A'$&"
$".+2%
%:;
)1(&,
)#%#
5&%&(
6,'
/#)$
.&)
*+,)
"! !!!
!
-.%")**)4,(')$%
)&,%)&6.+2%.&"
-.&"%*#+0
#%',+.(4#&
'+#(.&"
"#$%
&#',%$
-.&"%*#+#
'-$+
.4')=)')$%
/!
,%$%
5#'-.+2%.&"-.&"%*#+0
#%',+.(
4#&'+#(9.&""#$%
&#',%$
-.&"%*#+
#'-$
+.4')=)')$%
3!
,%$%
5#'-.+2%*#+0
#%',+.(4#&
'+#(
1-)($-.
&"%%-#1
%#2$0(.:
.4')=)':
7!
,%$%
#&$.+2
*#+0
#%',+.(%,00#
+'91-)($4#&'+.(.'$+.(.+2
.&"-.&"
.+$
,%$"
*#+2
.&)0,(.')#&
&<
+7*,
.(;+
"#)
$7
,(&%&
(6$;
+&)3
#1(&2
&(6&)
*+,)
"0*+
".")
1",'
#.67
7"(
+6" """
!5#
'-.+2%+$2.)&)&0+#00$
"0#%)')#&
0+#=)"$"
5:$A.2
)&$+
!
%'+#&6
.%:2
2$'+:9
!>?!
@1#+%'%)"$
/!
2#"$+.'$.%:2
2$'+:9
!>?!
@1#+%'%)"$
3!
&##+
2)("
.%:2
2$'+:
'4#
+&
%&(6
,'*+
"=1+
#>%&)
37
,2"7
")(.
,'(/
"%"
3.#$
#$
#$
#$
!%-#1
%#&
(:4+.1
()&6)&0+#&
$#+
0+$*$+%
'#%)'9%'.&
"#+
1.(;.+#,
&" !
)&%,**)4)$&
'(:=.+).5($
/!
%,**)4)$&
'(:=.+).
5($
(?
,%%&)
3'+
,7*+
,)"
(,.;
*&)"
! !!!
!"#
$%&#
'%-#
1+#(()&
6.%
'-$4-)("
0+$*$+%'#4-.&6$
0#%)')#&1)'-
'-$
-$(0#*
%)'')&6
#+4+.1
()&6
!
&#',+&)&6#+
+#(()&
6.''$20'%91-)($&#
'.5
($'#
4-.&6$
0#%)')#&1)'-
'-$-$
(0#*
%)'')&6
#+4+.1
()&6
/!
+#((%
'#%)"$
9,&)(.'$+.((:
!>?!
@3!
+#((%
'#%)"$
95)(.'$+.((:
7!
',+&%,&)(.'$+.((:
)&'#%,0)&$
!>?!
@<!
',+&%5)(.'$+.((:
)&'#%,0)&$
)-+
,3+"
..&,
)&)
*+,)
"8$"
2"%,
*7")
(,'
1+#>
%&)3
! !!!
!
"#$%
&#'%-#
11+)6
6()&690)=#')&6#+
4+.1
()&6
/!
1+)6
6()&6#+
0)=#')&6
3!
.5"#2)&.(4+.1
()&69,%$%
#&(:
.+2%#+
($6%
7!
.5"#2)&.(4+.1
()&69,%$%
.+2%.&
"($6%
<!
4+.1
(%#&
-.&"
%.&
";&$$%9.5"#2$&
*+$$*+#
2%,00#+'%,+*.4$
*4#
+&
%&(6
,'>
+&33%
&)30
*&2,
(&)3
,+1+
#>%&)
3#$
#$
#$
#$
!"#
$%&#
'#+-
.+"(:%-#1
1+)6
6()&69
0)=#')&6#+
4+.1
()&6
!
)&%,**)4)$&
'(:=.+).5($
/!
%,**)4)$&
'(:=.+).
5($
+4#
+&
%&(6
,'>
+&33%
&)30
*&2,
(&)3
,+1+
#>%&)
38#5
&%&(6
(,7
#9"
#)#$
#*(&2
"."
%"1(
&,)
#%
#%
#%
#%
!"#
$%&#
'#+-
.+"(:%-#1
1+)6
6()&69
0)=#')&6#+
4+.1
()&6
!
&#%$($4')#&
/!
.".0')=$%$($4')#&
30<#
30<#
30<#
30<#
125125
Appendix III
!"
#%&'
(%)*+)(**,-./0123"456(**,7#8#9/0:#;
5!9<
#=2/9/"1-4550>&&?
??@A#8#9/0:#;
5!9;#=
2/9/"1@B/:
!
"#$%
"&"'
()*+
,"-
),)#
$.! !!!
!
"#$$%&"%$&'%()*#
+,%-*,
*$&.
/!
"#$"%$&'%()*#+,%-*,
*$&.&%#
(0,0&*+*1&*$&
2!
"#$34((5
"%$&'%()*#+,%-*,
*$&.
!/"
.0$0"
#"'
()*+
+1%0#
2.0
$$0#2
34%
).)#
5)"'
4%)-
*0&0#
2()
*+4"
.0$0"
#$"
"#)
.0+)
" """
!
.&'%$6
(57'*-#0(0$6)*#+
7%.0&0%$&%
&)*!
89!
:/!
,%+*'#&*(57'*-#0(0$6)*
#+7%.0&0%$&%
&)*!
89!
:2!
$%%'
,0(+(57'*-#0(0$6)*#+
7%.0&0%$&%
%$*.0+*
"60
$$0#2
*70&0
$8! !!!
!
"#$$%
&.0&0$+*
7*$+
*$&(5
/!
"#$.0&;
0&)*1&'*
,*7*(-0.#$
&*3(*10%$
<=*((5&%4")0$6
477*'(*6.>?;
0&)%'
;0&)%4
&#',
.0$7'%77*+7%.0&0%$@
"#$$%&
.0&47'06)&
2!
"#$.0&,
%'*%'
(*..
47'06)&
3%'#
3*;
.*"%$+.
A!
.0&.0$+*
7*$+*$
&(5?=4&"#$$
%&.)03&
;*06)
&%''%&#&*&'4
$B?%
'%$(5.0&.
=*&;**
$B$**
.C!
.0&.0$+*
7*$+*$
&(5?0.#=(*&%
.)03&
;*06)
&?=4
&'%&#&*.
&'4$B
&%#,0$0,#(
*1&*$&
%$(5
D!
.0&.0$+*
7*$+*$
&(5#$+0.#=
(*&%
.)03&
;*06)
&#$+
'%&#&*&'4
$B
#/"
.0$0"
#"'
$%1#
9+1
%0#2
0#+)
4)#+
)#$
.0$$0
#234
%)')
%)#5
)4"
.$1%
)! !!!
!
"#$$%&.0&0$+*7
*$+*
$&(5<."%'*+ %'
/#&
0&*,
2> !
,%.&(5
'%4$+=#"B
/!
,%.&(5
.&'#06)&
=#"B?$
%(%'+%.0.?%'
.0&.%3&*$=*&;**
$B$**
.2!
.&'#06)&=#"B
;0&)
.%,*(%'+%.0.;)*
$.0&&0$6
%$=4&&%"B.
$/"
.0$0"
#"'
$%1#
9+1
%0#2
.0$$0
#2"%
.144
"%$)
+.0
$$0#2
" """
!
.&'%$6
#.5,
,*&'5?"%((#7.*.
&%!
89!
:.0+*
/!
,%+*'#&*#.5,
,*&'5?"%((#7.*.
&%!
89!
:.0+*
2!
$%%'
,0(+
#.5,
,*&'5
%:
%,4"
.$1%
)*#
+,
"$0&0
$8+1
%0#2
.0$$0
#2"%
.144
"%$)
+.0
$$0#2
34%)
.)#5
)"'
*.8,
,)$
%8" """
!
.&'%$6
#.5,
,*&'5?
!89!
:;%'.&.0+*
/!
,%+*'#&*#.5,
,*&'5?
!89!
:;%'.&.0+*
2!
$%%'
,0(+
#.5,
,*&'5
&;
))+
"'*%
,.1
44"%
$+1%
0#2
.0$$0
#2! !!!
!
"#$$%&.0&0$+*7
*$+*
$&(5<."%'*+ E2#&
0&*,2>
!
$**+.=%&)#',.3%'7
%.&4'#(.47
7%'&
/!
$**+.%$*#',
3%'7
%.&4'#(.47
7%'&
2!
4.*.
=%&)#',.3%'-%(4$
&#'5#"&0-0&5?
+%*.
$%&4.*
#',.3%'7
%.&4'#(.477%
'&
'<*
%0*70
&0$8
0#.0
$$0#2
,"$
0&0$8
#$
#$
#$
#$
!"#$$
%&.0&0$+
*7*$
+*$&(5
<."%'*+ %'
/#&
0&*,2>
!
0$.4330"0*$
&(5-#'0#=(*
/!
.4330"0*$
&(5-#'0#
=(*
(<*
%0*70
&0$8
.0$$0
#2,
"$0&0
$8=*
70&0$
8$"
,*9
)*#
*+*4
$0-)
.)&)
5$0"
##%
#%
#%
#%
!"#$$
%&.0&0$+
*7*$
+*$&(5<."%'*+ %'
/#&
0&*,
2> !
$%.*(*"&0%$
/!
#+#7&0-*.*(*"&0%$
)60
$$0#2
14! !!!
!
"#$$%&.0&470$+*7*
$+*$
&(5/!
"#$.0&4
70$+*7*$+
*$&(5
+<*
%0*70
&0$8
0#.0
$$0#2
147)
(*-0
"1%#$
#$
#$
#$
!"#$$
%&.0&470$+*
7*$+
*$&(5
!+0+%$(5.0&47%$
"*+4
'0$6#..*..,*$
& !
0$.4330"0*$
&(5-#'0#=(*
/!
.4330"0*$
&(5-#'0#
=(*
,<*
%0*70
&0$8
0#.0
$$0#2
147)
(*-0
"1%=
*70&0
$8$"
,*9
)*#
*+*4
$0-)
.)&)
5$0"
##%
#%
#%
#%
!"#$$
%&.0&470$+*
7*$+
*$&(5
!+0+%$(5.0&47%$
"*+4
'0$6#..*..,*$
& !
$%.*(*"&0%$
/!
#+#7&0-*.*(*"&0%$
->
"$$"
,.(
1''&0
#2#$
#$
#$
#$
!"#$$
%&.0&0$+
*7*$
+*$&(5<."%'*+ E2#&
0&*,2>
!
=%&&%,.)433(0$6
7'*.*$&#.%$(5
.&'#&*65
&%,%-*#'%4$+
/!
$%=%&&%,
.)433(0$6
7'*.*$&%
'=%&&%,
.)433(0$6
7'*.*$&#.%$*%3.&'#&*60*.
&%,%-*#'%4$+
C3553;"
C3553;"
C3553;"
C3553;"
126126
Appendix III
!"
#%&'
()*+,*(++-./01234"567(++-8#9#:01;#<
6!:=
#>30:0"2.5661?&&@
@@AB#9#:01;#<
6!:<#>
30:0"2AC0;
!"
#$%&
$'#(
&)&"*
! !!!
!
"#$$%&'&#$(
)!
'&#$('
*+&,
,-./
0!
'&#$('
+$(-/-$(
-$&.1
2%3#
2-*'-"4
%$('
5!
'&#$('
+$(-/-$(
-$&.1
2%36
%3-&,#$
7'-"%$('89:&(
%-'$%
&',+2&*-+;,
&%3
3%&#&-&3:
$<=!
'&#$('
+$(-/-$(
-$&.18"#$
',+2&*-+;,
&+$2%3*#3(#$('+(-
*#3((+3-"&+%$8
9:&
,#'(+22+":.&+-'
+$3%&#&+$;&3:
$<>!
'&#$('
+$(-/-$(
-$&.1
#$(+'#9.-&%
3%&#&-&3:
$<
!!"
#$%&
$'+,
! !!!
!
"#$$%&$%
&'&#$(
:/)!
;-&'%$
<$--
'0!
'&#$('
:/+$(-/-
$(-$
&.1*+&,
&,-:'-
%2-?;?2:3$+&:3-
5!
'&#$('
:/+$(-/-
$(-$
&.18*
+&,%:
&:'+$;
2:3$+&:3-
"-#
.&#(&
)&"*
&$/"
#$%&
$'+,
(01#
2&3+
."#
"#
"#
"#
!"#$$
%&'&#$
(:/
!(+(%$.1'&#$(:/
%$"-
(:3+$
;#''-''4
6-$
& !
+$':22+"+-$
&.1@#3+#9.-
)!
':22+"+-$
&.1@#3+#
9.-
#-#
.&#(&
)&"*
&$/"
#$%&
$'+,
(01#
2&3+
.4#(
&)&"*
"35
#60
#$#%
#,"&2
0/0
)07"
&3$"$
"$
"$
"$
!"#$$
%&'&#$
(:/
!(+(%$.1'&#$(:/
%$"-
(:3+$
;#''-''4
6-$
& !
$%'-.-"&+%$
)!
#(#/&+@-'-.-"&+%$
$8
#)6&
$'! !!!
!
"#$$%&
*#.<
)!
*#.<'
*,-
$3-"-+@+$;'://%3&9
1&*%,#
$('
0!
*#.<'
*,-
$3-"-+@+$;'://%3&9
1%$
-,#$(
5!
*#.<'
+$(-
/-$(-$
&.1
%9
#)#$
70%+
.&$'
&$%0
,0$%
0$":
#)6&
$'! !!!
!
"#$$%&
*#.<+$(-/-
$(-$
&.1 !
/%%3
9#.#$"+$;"#/#"+&+-'
)!
6%(-3#&-9#.#$"+$;"#/#"+&+-'
0!
;%%(
9#.#$"+$;"#/#"+&+-'
&;
.5,3
/"+.
0#$
%5
3"&)&
"*%+
.&$'
&$%0
<,0
$%0$
":#)
6&$'
! !!!
!
"#$$%&
*#.<+$(-/-
$(-$
&.1 !
/3-(%6
+$#$
&.1,+;,
%3'-6+4,
#.2;:#
3()!
#39+&3#
31#36/%
'&:3-
';
.5,3
/"+.
0#$
%5
3"&)&
"*%+
.&$'
&$%0
<,0
$%0$
":#)
6&$'
=,.0
/0$7
03>
#/*5
50<
".*
% %%%
!"#$$
%&*#.<+$(-
/-$(-$
&.1 !
'&3%$;
#'16
6-&318
!AB!
C*%3'&'+(-
)!
6%(-3#&-#'16
6-&318
!AB!
C*%3'&'+(-
0!
$%%3
6+.(
#'16
6-&31
(-#
.&#(&
)&"*
3>#.
55
3205
0$"/
%+.&$
'&$
%0,0
$%0$
":#)
6&$'
"#
"#
"#
"#
!"#$$
%&*#.<+$(-
/-$(-$
&.1!
/3-(
%6+$#$&.1
'-6+4,#.2%3,
+;,;:
#3(
!
+$':22+"+-$
&.1@#3+#9.-
)!
':22+"+-$
&.1@#3+#
9.-
)-#
.&#(&
)&"*
3>#.
55
3205
0$"/
%+.&$
'&$
%0,0
$%0$
":#)
6&$'
4#(&
)&"*
"35
#60
#$#%
#,"&2
0/0
)07"
&3$
"$
"$
"$
"$
!"#$$
%&*#.<+$(-
/-$(-$
&.1!
/3-(
%6+$#$&.1
'-6+4,#.2%3,
+;,;:
#3(
!
$%'-.-"&+%$
)!
#(#/&+@-'-.-"&+%$
*-#
.&#(&
)&"*
3>".
+$6
532
050$
"/%+
.&$'
/"#$
%&$'
#$%
&$%0
,0$%
0$":
#)6&
$'"#
"#
"#
"#
7!
"#$$%&
*#.<+$(-/-
$(-$
&.1 !
+$':22+"+-$
&.1@#3+#9.-
)!
':22+"+-$
&.1@#3+#
9.-
+-#
.&#(&
)&"*
3>".
+$6
532
050$
"/%+
.&$'
&$%0
,0$%
0$":
#)6&
$'4#
(&)&"
*"3
5#6
0#$
#%#,
"&20
/0)0
7"&3
$"$
"$
"$
"$
!"#$$
%&*#.<+$(-
/-$(-$
&.1 !
$%'-.-"&+%$
)!
#(#/&+@-'-.-"&+%$
,?0
',3
/"+.
0#$
%5
3"&)&
"*%+
.&$'
&$%0
<,0
$%0$
":#)
6&$'
=,.0
/0$7
03>
#/*5
50<
".*
% %%%
!"#$$
%&*#.<+$(-
/-$(-$
&.1 !
'&3%$;
#'16
6-&318
!AB!
C*%3'&'+(-
)!
6%(-3#&-#'16
6-&318
!AB!
C*%3'&'+(-
0!
$%%3
6+.(
#'16
6-&31
--#
.&#(&
)&"*
3>)0
'5
3205
0$"/
%+.&$
'&$
<%0
,0$%
0$":
#)6&
$'"#
"#
"#
"#
!"#$$
%&*#.<+$(-
/-$(-$
&.1 !
+$':22+"+-$
&.1@#3+#9.-
)!
':22+"+-$
&.1@#3+#
9.-
.-#
.&#(&
)&"*
3>)0
'5
3205
0$"/
%+.&$
'&$
<%0
,0$%
0$":
#)6&
$'4#
(&)&"
*"3
5#6
0#$
#%#,
"&20
/0)0
7"&3
$"$
"$
"$
"$
!"#$$
%&*#.<+$(-
/-$(-$
&.1 !
$%'-.-"&+%$
)!
#(#/&+@-'-.-"&+%$
/@
00)<"
30'#
&"%+
.&$'
&$%0
,0$%
0$"
:#)
6<&$
'! !!!
!
"#$$%&
*#.<+$(-/-
$(-$
&.1 !
$%%3
%$.1%""#'+%$
#..1,--.4&%
-;#
+&)!
/3-(%6
+$#$
&.1,-
-.4&%
-;#+&
0-#
.&#(&
)&"*
3>,)
#7&$
'3>
>00"
%+.&$
'&$
%0<
,0$%
0$":
#)6&
$'"#
"#
"#
"#
!"#$$
%&*#.<+$(-
/-$(-$
&.1 !
+$':22+"+-$
&.1@#3+#9.-
)!
':22+"+-$
&.1@#3+#
9.-
1-#
.&#(&
)&"*
3>,)
#7&$
'3>
>00"
%+.&$
'&$
%0<
,0$%
0$":
#)6&
$'4#
(&)&"
*"3
5#6
0#$
#%#,
"&20
/0)0
7"&3
$"$
"$
"$
"$
!"#$$
%&*#.<+$(-
/-$(-$
&.1 !
$%'-.-"&+%$
)!
#(#/&+@-'-.-"&+%$
2A)
+0$7
*3>
53"
&)&"*
%+.&$
'&$
%0,0
$%0$
":
#)6&
$'& &&&
!"#$$
%&*#.<+$(-
/-$(-$
&.1 !
6#D%3+&1
%2*#.<+$;
6%@-6
-$&'$%$4
2.:-$&E'&+228D-3<182.%//
1B'.:;
;+',8
%&,-
3E
FFFFFFFFFFFFF??
)!
.#3;-6#D%3+&1
%2*#.<+$;
6%@-6
-$&'
2.:-$&
D6!<
B4<"!<
B@!:E4<"
D6!<
B4<"!<
B@!:E4<"
D6!<
B4<"!<
B@!:E4<"
D6!<
B4<"!<
B@!:E4<"
127127
Appendix III
!"
#%&'
()*+%*(++,-./0123"456(++,7#8#9/0:#;
5!9<
#=2/9/"1-4550>&&?
??@A#8#9/0:#;
5!9;#=
2/9/"1@B/:
!
"#$%
&'()
(*#+
,&'(
#'-
.#'
&,/0
#1&2
'23
245"
$1+
! !!!
!
"#$%
&#'($)*+
'#,)("%
#-.$*'
/!
"#$%
&#'($)*+0-1'%+#,
%2($($)*+3&45#6$5
$&'%
7!
($)*+$%'#,)("%
#-.$*'-1'
"#$%
&#'4()%2
3'8!
($)*+$%'#,)("%04()%2%)&
"+#9"%
#-.$*'0-1'
"#$%
&#'5
)&3219)'$#-
.$*'
:!
($)*+$%'#,)("%0+#9"%
)&"
5)&
3219)'$%
#-.$*'
;!
($)*+$%'#,)("%0+#9"%
)&"
5)&
3219)'$%
/#-.$*'%
<!
($)*+$%'#,)("%0+#9"%
)&"
5)&
3219)'$%
7#-.$*'%
!!
"#$%
&'()
(*#+
,&'(
#'-
.#'
&,/0
#1&2
'23
245"
$1+)
,*"+
"'$"
23#+
6..
"1*6
" """
!"#
$%&#
'($)*+
'#,)("%
#-.$*'
!
%'(#&4
)%=5
5$'(=0
!>?!
@,#(%'%3"$
/!
5#"$()'$)%=5
5$'(=0
!>?!
@,#(%'%3"$
7!
&##(
539"
)%=5
5$'(=
"7#
*&
0&16
23,*
"*"#
$%&'
(2*
*"#$
%&'(
.28
"."'
1+23
1%"
#*.
+#$
#$
#$
#$
!'+$*+39"
"#$%
&#'%+#
,2($($)
*+3&4#(
($)*+3&4
5#6$5
$&'%
!
3&%1AA3*3$&
'9=6)(3)-9$
/!
%1AA3*3$&
'9=6)(3)
-9$
#7#
*&
0&16
23,*
"*"#
$%&'
(2*
*"#$
%&'(
.28
"."'
1+9#
4&0&1
612
.#:
"#'
#-#,
1&8"
+"0"
$1&2
'#%
#%
#%
#%
!'+$*+39"
"#$%
&#'%+#
,2($($)
*+3&4#(
($)*+3&4
5#6$5
$&'%
!
&#%$9$*'3#&
/!
)")2'36$%$9$*'3#&
$;6
,"23
(*#+
,&'(
<+$2
*"4"
+1,"
*32*
.#'
$"=
%"'
-"#0
&'(
=&1%
#+.
#00
245"
$1>)
+&11&
'(2*
+/,&
'"! !!!
!"#
$%&#
'4()%2
#-.$*'
!
2)95)(
4()%2
/!
()"3)9B2)95)(
4()%2
7!
%*3%%#(4
()%2
8!
3&A$(3#(2
3&*$(4
()%2
:!
23&*$(
4()%2
%7#
*&
0&16
23%#
'-.
21&0&
16-/
*&'(
*"#$
%&'(
#'-
(*#+
,&'(
#$
#$
#$
#$
!&#
($)*+3&4
5#6$5
$&'%2($%$&'#
(#&
9=2($($)*+3&45#6$5
$&'%2($%$&'
!
3&%1AA3*3$&
'9=6)(3)-9$
/!
%1AA3*3$&
'9=6)(3)
-9$
&7#
*&
0&16
23%#
'-.
21&0&
16-/
*&'(
*"#$
%&'(
#'-
(*#+
,&'(
9#4&
0&16
12.
#:"
#'#-
#,1&8
"+"
0"$1
&2'
#%
#%
#%
#%
!&#
($)*+3&4
5#6$5
$&'%2($%$&'#
(#&
9=2($($)*+3&45#6$5
$&'%2($%$&'
!
&#%$9$*'3#&
/!
)")2'36$%$9$*'3#&
';*
".2*
-/*&'
(*"
#$%&
'(2*
,*"*
"#$%
&'(
& &&&
!'+$*+39"
"#$%
&#'%+#
,2($($)
*+3&4#(
($)*+3&4
5#6$5
$&'%
!
A($C1$&
'9=2($%$&'0"$
%*(3-$'=2$
D
EEEEEEEEEEEEEFF
/!
&##(
#**)%3#&)99='($
5#(
2($%$&'0
"$%*(3-
$'=2$
D
EEEEEEEEEEEEEFF
(?0
/"'$
623
.21
&0&16
-/*&'
(*"
#$%&
'(2*
,*"*
"#$%
&'(
& &&&
!'+$*+39"
"#$%
&#'%+#
,2($($)
*+3&4#(
($)*+3&4
5#6$5
$&'%
!
5).#(3'=
#A5#6$5
$&'%&#&BA91$&'D%'3AA0
.$(G=0A9#22
=?%91443%+0#'+$(,3%$D
EEEEEEEEEEEEEFF
/!
9)(4$5).#(3'=
#A5#6$5
$&'%A91$&'
C#!B43;"D"2!E03;"!;
A:!;
30=9!53/;/F/GH#B5E?439#
E3553;"/;0!2#;5E9!0
C#!B43;"D"2!E03;"!;
A:!;
30=9!53/;/F/GH#B5E?439#
E3553;"/;0!2#;5E9!0
C#!B43;"D"2!E03;"!;
A:!;
30=9!53/;/F/GH#B5E?439#
E3553;"/;0!2#;5E9!0
C#!B43;"D"2!E03;"!;
A:!;
30=9!53/;/F/GH#B5E?439#
E3553;"/;0!2#;5E9!0
128128
Appendix III
!"
#%&%
'()*+)'**,-./0123"456'**,7#8#9/0:#;
5!9<
#=2/9/"1-4550>&&?
??@A#8#9/0:#;
5!9;#=
2/9/"1@B/:
!"
#$"%
&'()
&**$
+,!"
!"
!"
!"
!
"#$%&&"'"(#
)*+,-."-/*(
0!
$%&&"'"(#
)*+,-."-
/*(
!!"
#$"%
&'()
&**$
+,-"
.$%$/
0/+
1"2
&",
"3"(
/$4&
*&%&
#/$+
,!#
!
#1$(*(')"1#
0!
-2-3)",($(*(')"1#
"5
)++%
$,6
!"
!"
!"
!"
!
4-.5(22.11*"#6
0!
#11.
*"))*(
2.11*"#6
#7)
&*&,
#&+8
*/&)
&+/0
(&3
/+,6
9&()
+/)9
*$+,
!"
!"
!"
!"
!
+($
0!
#1
$:)
&1+)
$ $$$
!
&.(7%(#
)*+3.($(#)82(
$'."/()+3(
90!
#11.
1''-$"1#-**+).(
41.
3.($(#)8
2($'."/
()+3(
9
:::::::::::::;;
%!%
9&,#
0+8
1+/
+).&
;"4$
+9)$ $$$
!
4-<1.")+
1&41,(4
(#)$#1#=&*%(#)9$)"&&8
<(.5+8&*133
+>$*%66"$?81)?(.@"$(9
:::::::::::::;;
0!
4-<1.")+
1&41,(4
(#)$&*%(#
)
C#;#2!9
C#;#2!9
C#;#2!9
C#;#2!9
<
9",/
$/0+8
1+/
$%$/0
!!
"!
#"!
##"!
###
!=
&;"4
$+9)
"%*/
"/&
$$
$$
$$
$$
$$
$$
$%%
$$
$$
$$
$$
$$
$$
$%%
$$
$$
$$
$$
$$
$$
$%%
$$
$$
$$
$$
$$
$$
$%%
$$
$$
$$
$$
$$
$$
$%%
$$
$$
$$
$$
$$
$$
$%%
$$
$$
$$
$$
$$
$$
$%%
$$
$$
$$
$$
$$
$$
$%%
">
&"%/;
#+,3
$/$+,
$$
$$
$$
$$
$$
$$
$%%
$$
$$
$$
$$
$$
$$
$%%
$$
$$
$$
$$
$$
$$
$%%
$$
$$
$$
$$
$$
$$
$%%
$$
$$
$$
$$
$$
$$
$%%
$$
$$
$$
$$
$$
$$
$%%
$$
$$
$$
$$
$$
$$
$%%
$$
$$
$$
$$
$$
$$
$%%
#?
/;&)
$$
$$
$$
$$
$$
$$
$%%
$$
$$
$$
$$
$$
$$
$%%
$$
$$
$$
$$
$$
$$
$%%
$$
$$
$$
$$
$$
$$
$%%
$$
$$
$$
$$
$$
$$
$%%
$$
$$
$$
$$
$$
$$
$%%
$$
$$
$$
$$
$$
$$
$%%
$$
$$
$$
$$
$$
$$
$%%
D#:!2EF/;
D#:!2EF/;
D#:!2EF/;
D#:!2EF/;
129
Nederlandse samenvatting
130130
Nederlandse samenvatting
De Infant Motor Pro�le (IMP) is een nieuw meetinstrument voor een kwalitatieve beoordeling van de
motoriek van zuigelingen van 3 tot en met 18 maanden. De IMP werd ontwikkeld met de volgende
drie doelstellingen: onderscheid maken tussen zuigelingen met een hoog risico op motorische
ontwikkelingsproblemen en zuigelingen met een normale motorische ontwikkeling, het evalueren
van motorische functies in de tijd en het voorspellen van motorische ontwikkelingsproblemen. De
IMP is gebaseerd op de Neuronale Groep Selectie Theorie die stelt dat motorische ontwikkeling
gekenmerkt wordt door exploratie van het primaire variabele motorische repertoire gevolgd door
de fase waarin het kind adaptieve motorische strategieën leert kiezen uit het repertoire. De IMP
bestaat uit 80 items die zijn gegroepeerd in 5 domeinen: bewegingsvariatie, vermogen tot selectie
van adaptieve motorische strategieën, vloeiendheid, symmetrie en motorische prestaties. De items
worden beoordeeld aan de hand van een video-opname.
In hoofdstuk 2 wordt een overzicht gepresenteerd van de bestaande methoden voor het
beoordelen van neuromotorische conditie op de zuigelingenleeftijd en de psychometrische
eigenschappen van deze methoden. Dit overzichtsartikel vormde de achtergrond waartegen
de Infant Motor Pro�le werd ontwikkeld. Het bleek dat meetinstrumenten die zich richten op
kwalitatieve aspecten van motoriek, zoals de General Movement methode (GM) en de Test of Infant
Motor Performance (TIMP) de beste voorspellende waarde hebben voor zowel lichte als ernstige
motorische ontwikkelingsproblemen op latere leeftijd. Deze bestaande methoden zijn echter
slechts toepasbaar tot de leeftijd van 4 maanden.
Hoofdstuk 3 bevat de pilot studie over de IMP met een beschrijving van de methode en de
eerste data over betrouwbaarheid en validiteit. Hiervoor werd de IMP toegepast in een heterogene
onderzoekspopulatie van op tijd en te vroeg geboren zuigelingen in de leeftijd van 3 tot en met
18 maanden. Intra- en inter-beoordelaarsbetrouwbaarheid bleken goed te zijn. De soortgenoot
validiteit met de Alberta Infant Motor Scale (AIMS) was goed. Er was een duidelijke relatie tussen
IMP scores en neurologische conditie, beoordeeld met behulp van het neurologisch onderzoek
volgens Touwen.
In hoofdstuk 4 wordt, met als theoretische achtergrond de Neuronale Groep Selectie Theorie,
een studie beschreven die gericht was op het vaststellen van de leeftijd waarop gezonde zuigelingen
het vermogen tot selecteren van adaptieve bewegingsstrategieën ontwikkelen. Er werd gekeken
naar vier speci�eke motorische functies: kruipen, zitten, reiken en grijpen. Ook werd onderzocht of
de deze beoordeling van de aanwezigheid van het selecteren van adaptieve bewegingsstrategieën
betrouwbaar gedaan kon worden. Dit bleek inderdaad het geval. Vanaf de leeftijd van 6 maanden
ontwikkelden zuigelingen het vermogen tot selectie, met een piek tussen de leeftijd van 8 en 15
maanden. Deze transities traden geleidelijk op op functiespeci�eke momenten. Toen we onze
resultaten vergeleken met resultaten in de literatuur van neurofysiologisch onderzoek, bleek dat
deze laatste in de tijd voorafgaan aan klinisch geobserveerde transities.
Hoofdstuk 5 gaat over de construct validiteit van de IMP. Construct validiteit was in deze
studie geoperationaliseerd als de relatie van IMP scores met belangrijke pre-, peri- en neonatale
131131
Nederlandse samenvatting
risicofactoren voor het ontwikkelen van motorische problemen. Om de construct validiteit te
onderzoeken werd een longitudinale, prospectieve studie in een groep op tijd en te vroeg geboren
zuigelingen uitgevoerd. Perinatale data, waaronder gegevens van neonatale schedelecho’s
in termen van intraventriculaire bloedingen en periventriculaire leukomalacie, en gegevens
over de sociaaleconomische situatie werden verzameld en hiermee werden statistische mixed
e�ects modellen gemaakt. In de totale groep zuigelingen bleken de IMP scores het sterkst te
worden beïnvloed door zwangerschapsduur, sociaaleconomische status en de Apgar score na 5
minuten. In de subgroep van de te vroeg geboren kinderen was de aanwezigheid van duidelijke
afwijkingen op de neonatale schedelecho’s de belangrijkste beïnvloedende factor, gevolgd door
sociaaleconomische status. Deze bevindingen ondersteunen de construct validiteit van de IMP: IMP
scores waren duidelijk gerelateerd aan belangrijke risicofactoren voor het optreden van motorische
ontwikkelingsproblemen.
Hoofdstuk 6 is gericht op de betrouwbaarheid, de soortgenoot validiteit met de AIMS en
het neurologisch onderzoek volgens Touwen en de predictieve validiteit van de IMP. Hiervoor
werd een prospectieve, longitudinale studie verricht in een heterogene groep van op tijd en te
vroeg geboren zuigelingen. Voor het deel van het onderzoek gericht op de soortgenoot validiteit
werd een groep op tijd geboren zuigelingen met cross-sectionele beoordelingen toegevoegd. De
inter-beoordelaarsbetrouwbaarheid bleek goed te zijn. De soortgenoot validiteit met de Alberta
Infant Motor Scale was het beste voor de domeinen motorische prestaties en bewegingsvariatie.
Er was een duidelijke relatie tussen IMP scores en neurologische conditie. De predictieve validiteit
van de IMP op de leeftijd van 4 tot en met 12 maanden voor het voorspellen van een afwijkende
neurologische conditie op de leeftijd van 18 maanden bleek redelijk goed te zijn.
Met betrekking tot de drie doelen waarvoor de IMP werd ontwikkeld, kunnen we stellen dat
de IMP goed in staat is om onderscheid te maken tussen zuigelingen met een goede motorische
ontwikkeling en zuigelingen met een hoog risico op motorische ontwikkelingsproblemen. Of
de IMP in staat is om de ontwikkeling van motorische functies in de tijd te evalueren, moet
verder onderzocht worden door de IMP toe te passen in interventiestudies. Het op jonge leeftijd
voorspellen van de motorische uitkomst zal altijd moeilijk blijven, aangezien het zich ontwikkelende
zenuwstelsel veranderlijk is. In onze onderzoekspopulatie was de predictieve validiteit van de IMP
redelijk goed. Vervolgonderzoek dient gericht te zijn op het genereren van normscores en op het
onderzoeken van de klinische toepasbaarheid van de IMP.
132
133
Dankwoord
134
Dankwoord
Dankwoord
De afgelopen jaren heb ik met veel plezier gewerkt aan mijn promotieonderzoek. Dit is voor een
belangrijk deel te danken aan de mensen die ik hieronder wil noemen.
Allereerst wil ik mijn eerste promotor prof. dr. M. Hadders-Algra bedanken. Beste Mijna, ik
herinner me nog goed dat ik bij je langs kwam om te praten over de mogelijkheden voor een
afstudeeronderzoek. Ik weet nog steeds niet waardoor het precies kwam, maar ik was meteen
geboeid door de ontwikkelingsneurologie en de manier waarop jij erover sprak. Ik ging naar huis
met twee artikelen van jou over de NGST en was meteen verkocht: hier wilde ik mee aan de slag!
Van deze keus heb ik nooit spijt gekregen en dat is grotendeels aan jou te danken. Ik heb veel
bewondering voor jou als wetenschapper: voor je originele ideeën, voor de manier waarop je die
overbrengt en voor je enorme doorzettingsvermogen en gedrevenheid. Ik heb ontzettend veel van
je geleerd en je hebt me steeds op een geschikte manier begeleid bij alle stappen op het pad van de
wetenschap. Het moet voor jou niet ideaal geweest zijn dat ik altijd meerdere dingen tegelijk deed,
eerst tijdens mijn MDPhD-traject en later tijdens het combineren van de opleiding neurologie met
mijn promotieonderzoek. Bedankt dat je me hiervoor de ruimte hebt gegeven. Daarnaast waardeer
ik je persoonlijke betrokkenheid en heb ik bewondering voor de manier waarop je in het leven staat.
Prof. dr. A.F. Bos, beste Arie, als tweede promotor was je voor mij het aanspreekpunt voor alle vragen
betre�ende de neonatologie. Toen ik net begon met mijn onderzoek heb je me rondgeleid over
de NICU en nog steeds ben ik, net als toen, iedere keer als ik daar kom weer onder de indruk van
de complexiteit van jullie vak. Ik ben je zeer dankbaar voor het gedetailleerde beoordelen van de
artikelen en je kritische, klinische blik daarbij.
De leescommissie bestaande uit prof. dr. L.S. de Vries, prof. dr. O.F. Brouwer en prof. dr. J.G. Becher wil
ik bedanken voor het beoordelen van dit proefschrift.
Dr. V. Fidler en dr. S. la Bastide-van Gemert van de afdeling Epidemiologie van het UMCG, beste
Vaclav en Sacha, bedankt voor jullie mooie statistische analyses uit hoofdstuk 5 en de geduldige
uitleg daarbij.
Beste collega’s van de ontwikkelingsneurologie: Cornill Blauw-Hospers, Hylco Bouwstra, Tineke
Dirks, Saskia van Goor, Victorine de Graaf-Peters, Tjitske Hielkema, Corina de Jong, Marjolein
Jongbloed-Pereboom, Hedwig Kikkert en Lieke Peters, een aantal van jullie heeft geholpen met
het �lmen van baby’s die soms in een afgelegen pittoresk Fries dorpje bleken te wonen. Veel dank
daarvoor. Maar vooral bedankt voor de leuke gesprekken over wetenschap en andere zaken en jullie
interesse in het onderzoek.
Loes de Weerd, secretaresse van de afdeling ontwikkelingsneurologie, heel veel dank voor je hulp
en ondersteuning bij duizend-en-één klussen die in de loop van het onderzoek voorbij kwamen.
Michiel Schrier, bedankt voor de mooie lay-out van het IMP scoreformulier en je technische
135
Dankwoord
ondersteuning.
In de loop van de jaren heeft een aantal studenten geneeskunde en bewegingswetenschappen in
het kader van hun afstudeeronderzoek meegewerkt aan het IMP-onderzoek. Nicole Arink, Dineke
Dijkhuizen, Lieke Eidhof, Emilie Groen en Anne Hoekstra: dank voor jullie hulp bij het werven, �lmen
en onderzoeken van de baby’s en vooral bedankt voor jullie vaak actieve meedenken. Ik heb veel
geleerd van jullie kritische vragen!
De Junior Scienti�c Masterclass wil ik bedanken voor de mogelijkheid om dit onderzoek in de vorm
van een MDPhD-traject uit te voeren.
Sta�eden en AIOS van de afdeling neurologie in het UMCG, �jn dat ik bij en met jullie mag werken.
Alle ouders en hun kinderen wil ik bedanken voor de deelname aan dit toch behoorlijk intensieve
follow-up onderzoek. Het is niet makkelijk om een kind te krijgen dat te vroeg geboren is, en ik vind
het dan ook extra bewonderenswaardig dat ouders de moeite wilden nemen om mee te doen.
Het was bijzonder en leuk om hun kinderen te zien opgroeien van baby’s tot pratende peuters van
anderhalf jaar.
Karin Middelburg en Layla Ben-Zvi, wat �jn dat jullie mijn paranimfen willen zijn.
Lieve Karin, je begon als nieuwe collega op de ontwikkelingsneurologie en werd al gauw een
vriendin. Bedankt voor al je hulp bij het �lmen van de baby’s als ik met de co-schappen bezig was
(je moest er voor je eigen onderzoek ook al zo veel…). Maar vooral dank voor je gezelligheid en de
vele winkel- en horeca-uitjes.
Lieve Layla, jij bent een enorm betrokken vriendin die altijd klaar staat met praktische adviezen.
Geweldig dat je zo gezellig om de hoek woont! Misschien moeten we weer eens wat meer gaan
dansen…
Familieleden en vrienden, ook al waren jullie niet betrokken bij de totstandkoming van dit
proefschrift, ik wil jullie toch niet overslaan, want voor allerlei andere leuke en gezellige dingen zijn
jullie heel belangrijk voor mij.
Lieve papa en David, wat hebben wij veel dokters in de familie… Gelukkig kan ik jullie over de
ontwikkelingsneurologie nog wel eens wat nieuws vertellen. Ik vind het �jn dat jullie er zijn.
Van mama heb ik de liefde voor het omgaan met kinderen en hun ouders meegekregen. Zij schreef
haar proefschrift op een bloknootje, op een bankje in de speeltuin waar wij speelden. Ze was erg
geïnteresseerd in mijn onderzoek en is zelfs een keer mee geweest om baby’s te �lmen. Wat jammer
dat ze niet meer bij de afronding van mijn proefschrift kan zijn.
Lieve Cyril, met mijn proefschrift heb je weinig te maken, maar met mij des te meer! Ik ben enorm
blij dat jij er bent en dat we samen van alle grote en kleine dingen in het leven kunnen genieten.
136
137
Curriculum vitae
138
139
CURRICULUM VITAE
Kirsten Heineman werd op 13 november 1978 geboren na een zwangerschapsduur van 31+1
weken met een geboortegewicht van 2030 gram. Zij volgde gymnasium klas 1 tot en met 5 aan het
Bernardinuscollege te Heerlen en behaalde in 1997 cum laude haar eindexamen aan het Willem
Lodewijk Gymnasium te Groningen. In 1998 behaalde ze cum laude de propedeuse Pedagogische
Wetenschappen aan de Rijksuniversiteit Groningen. Daarna begon ze met de bovenbouwstudie
Bewegingswetenschappen aan de RUG. In 1999 werd ze ingeloot voor Geneeskunde. In 2004
rondde ze de studie Bewegingswetenschappen af met een afstudeeronderzoek bij de afdeling
Ontwikkelingsneurologie in het Universitair Medisch Centrum Groningen. In datzelfde jaar startte
ze met een MDPhD-traject bij de afdeling Ontwikkelingsneurologie, waarbij ze de coschappen
combineerde met haar promotieonderzoek. Ze deed haar keuzecoschap bij de afdeling
Kinderneurologie in het UMCG en behaalde haar artsexamen in 2007. In december 2007 begon ze
met de opleiding tot neuroloog in het UMCG met als huidige opleider prof. dr. H.P.H. Kremer.
140
141
List of publications
142
143
LIST OF PUBLICATIONS
• Heineman KR, La Bastide-van Gemert S, Fidler V, Middelburg KJ, Bos AF, Hadders-Algra M.
Construct validity of the Infant Motor Pro�le: relation with prenatal, perinatal, and neonatal
risk factors. Dev Med Child Neurol 2010, epub ahead of print.
• Heineman KR, Middelburg KJ, Hadders-Algra M. Development of adaptive motor behaviour in
typically developing infants. Acta Paediatr 2010;99:618-24.
• Hadders-Algra M, Heineman KR, Bos AF, Middelburg KJ. The assessment of minor neurological
dysfunction in infancy using the Touwen Infant Neurological Examination: strengths and
limitations. Dev Med Child Neurol 2010;52:87-92.
• Heineman KR, Hadders-Algra M. Evaluation of neuromotor function in infancy-A systematic
review of available methods. J Dev Behav Pediatr 2008;29:315-23.
• Heineman KR, Bos AF, Hadders-Algra M. The Infant Motor Pro�le: a standardized and qualitative
method to assess motor behaviour in infancy. Dev Med Child Neurol 2008;50:275-82.
• Stevens M, Lemmink KAMP, De Jong J, Heineman K. Het e�ect van het GALM-
introductieprogramma op het dagelijkse energieverbruik van senioren van 55 - 65 jaar.
Geneeskunde en Sport 2003;6:170-5.
144