auditory event-related potentials as indices of language...

AUDITORY EVENT-RELATED POTENTIALS AS INDICES OF LANGUAGE IMPAIRMENT IN CHILDREN BORN PRETERM AND WITH ASPERGER SYNDROME

EIRAJANSSON-VERKASALO

Department of Finnish, Saami and Logopedics,University of Oulu

Department of Clinical Neurophysiology,Department of Paediatrics,

Oulu University HospitalCognitive Brain Research Unit,

Department of Psychology,University of Helsinki

OULU 2003

EIRA JANSSON-VERKASALO

AUDITORY EVENT-RELATED POTENTIALS AS INDICES OF LANGUAGE IMPAIRMENT IN CHILDREN BORN PRETERM AND WITH ASPERGER SYNDROME

Academic Dissertation to be presented with the assent ofthe Faculty of Humanities, University of Oulu, for publicdiscussion in the Auditorium 12 of the Department ofPaediatrics, Oulu University Hospital, on December 19th,2003, at 12 noon.

OULUN YLIOPISTO, OULU 2003

Copyright © 2003University of Oulu, 2003

Supervised byProfessor Pirjo KorpilahtiProfessor Matti LehtihalmesAcademy Professor Risto NäätänenDocent Ville JänttiDocent Leena Vainionpää

Reviewed byPh.D. Nicole BruneauDocent Teija Kujala

ISBN 951-42-7244-7 (URL: http://herkules.oulu.fi/isbn9514272447/)

ALSO AVAILABLE IN PRINTED FORMATActa Univ. Oul. B 54, 2003ISBN 951-42-7243-9ISSN 0355-3205 (URL: http://herkules.oulu.fi/issn03553205/)

OULU UNIVERSITY PRESSOULU 2003

Jansson-Verkasalo, Eira, Auditory event-related potentials as indices of languageimpairment in children born preterm and with Asperger syndrome Department of Finnish, Saami and Logopedics, University of Oulu, P.O.Box 1000, FIN-90014University of Oulu, Finland; Department of Clinical Neurophysiology, Oulu University Hospital,P.O.Box 50, FIN-90029 Oulu University Hospital, Finland, Department of Paediatrics, OuluUniversity Hospital, P.O.Box 23, FIN-90029 Oulu University Hospital, Finland; Cognitive BrainResearch Unit, Department of Psychology, University of Helsinki, P.O.Box 9, FIN-0014, Finland Oulu, Finland2003

Abstract

The main objective of the present follow-up study was to investigate auditory processing by usingauditory event related potentials (ERPs), and language development to determine whether acorrelation exists between auditory ERPs and language development.

Auditory processing was investigated in very low birth weight (VLBW) preterm children andmatched controls at mean ages of 4 and 6 years to determine whether there are differences in ERPsbetween VLBW preterm children and controls. Language development was measured at the meanages of 2, 4 and 6 years to investigate the developmental course of language learning and to determinewhether a relationship exists between ERPs, especially mismatch negativity (MMN), and languagedevelopment. Auditory ERPs were also measured in children with AS (mean age 9;1 years) andmatched controls to assess whether differences can be found between these two groups of children.Language development in children with AS was not investigated for this study.

VLBW preterm children exhibited difficulties in the auditory processing at the level of obligatoryERPs, MMN, late MMN (lMMN) and behavioural tests. Both language comprehension andproduction were deficient in the preterm group compared to their controls. Lexical development wasthe most prominent phenomenon differentiating preterm children from their controls. MMN andlMMN amplitudes were attenuated most in children with naming difficulty at the ages of 4 and 6years. Weak or totally missing MMN at the age of 4 years was mainly found in children with namingdifficulties.

Children with AS also displayed abnormalities in auditory processing, as indexed by delayedMMN latency. MMN was most delayed in the right hemisphere and specifically for tones.

In conclusion: VLBW preterm children and children with AS exhibited difficulties in auditoryprocessing. MMN correlated well with language development in preterm children. Therefore,auditory ERPs, especially MMN, should be used in combination with language measures to identifythe children at a risk for deficient auditory processing and language delays.

Keywords: Asperger syndrome, auditory processing, language, late mismatch negativity,mismatch negativity, preterm, very low birth weight

To my family

Acknowledgements

This work was carried out at the Department of Finnish, Saami and Logopedics, University of Oulu, in collaboration with the Department of Clinical Neurophysiology, Paediatrics, Child Psychiatry, Otorhinolaryngology, and Diagnostic Radiology, Oulu University Hospital, and Cognitive Brain Research Unit, Department of Psychology, University of Helsinki during the years 1999-2003 as part of the Langnet, Finnish National Graduate School for Language Studies.

I wish to express my warmest and deepest thanks to Professor Pirjo Korpilahti and Matti Lehtihalmes, my supervisors in the Department of Finnish, Saami and Logopedics. Without your scientific enthusiasm and broad-mindedness this thesis would not have been possible. You gave me the possibility to test my scientific limits, and lead me gently back to the field of logopedics.

I owe my gratitude to my supervisor, Academy Professor Risto Näätänen, PhD, Head of the Cognitive Brain Research Unit, University of Helsinki. You made my research possible by giving me the opportunity to use the most objective method for measuring auditory discrimination in children. It has always been possible for me to rely on your support, and your gentle way of teaching the scientific writing during these years.

I would also like to thank my supervisor, Docent Ville Jäntti, for trusting in my capability as a collaborator in starting MMN measurements in the Cognitive Laboratory, Oulu University Hospital, and for giving me all the facilities needed for the measurements. I also thank warmly Docent Leena Vainionpää, my supervisor from the Department of Paediatrics, for her support during these years, and Anna-Liisa Saukkonen, MD, PhD, for introducing me to the topic of prematurity.

I wish to thank Docent Teija Kujala, Helsinki Collegium for Advanced Studies, University of Helsinki, and Nicole Bruneau, PhD, INSERM, Tours, for their valuable comments and constructive criticism. I highly appreciate your humane attitude towards me as a junior researcher at the final steps of this thesis project.

I would like to express my special thanks to Marita Valkama, MD, PhD, who primarily collected the present preterm group. Without your help this study would not have been possible. I highly respect your reliability and collegiate way of working right from the beginning of this research.

My warmest thanks are due to Kalervo Suominen, Phil.Lic., phycisist. You gave me the technical support for the measurements. You also developed an excellent EEG-View program for the MMN analyses and repeatedly modified it to meet my needs. Even more, you were my support during the times of despair.

I acknowledge my sincere gratitude to Rita Čeponienė, MD, PhD, who taught me the basics of ERP measurement and helped with the initiation of ERP measurements in the Cognitive Laboratory, Oulu University Hospital. I have learned a lot from your scientific thinking and way of working during the years of collaboration

I want to thank Professor Irma Moilanen, MD, PhD, Head of the Department of Child Psychiatry, for giving me the opportunity to do research in collaboration with your team. Your research group together with Marko Kielinen, MA, Sirkka-Liisa Linna, MD, PhD, and Terttu Tapio, MA, guided me into an intriguing world that was so unknown to me.

I owe my sincere thanks to the heads and staff of the departments where the research projects were carried out. Especially, I would like to thank Professor Mikko Hallman, who is such an expert in research on premature infants. I also thank all the co-writers of my articles: Docent Paavo Alku, Docent Eero Ilkko, Docent Kyösti Laitakari and Docent Eija Pääkkö. Without your support, this multidiciplinary research would not have been possible.

I want to express my warmest thanks to the staff of the Department of Clinical Neurophysiology. You provided me with a pleasant research environment. Raija Jakkula, Head Nurse, my warmest thanks for the flexible arrangements of the working day. Raija Remes, EEG technician, I wish to thank you so much for your help with the measurements. I would also like to thank Sanna Fält and Pasi Lepola for their help as well as for the many nice moments that we have shared during the years.

I would like to express my thanks to many people in the Cognitive Brain Research Unit, University of Helsinki, who have helped me during the years. Without your support and help this thesis would not have been possible.

I wish to express my thanks also to the Department of Finnish, Saami and Logopedics. My sincere thanks to Merja Karjalainen, PhD, who taught me how to analyze video- recorded data. I warmly thank Seija Pekkala for your friendship during the graduate school and for so many moments of joy. I also thank Sonja Haverinen, Nina Kuusinen, Sanna Perenius, Marja-Leena Rautava, and Päivi Viinikka for helping me with the data collection and data analyses. I also thank Niina Korhonen and Iri Patronen for their collaboration during the years.

I am very grateful to Risto Bloigu, MSc, and Esa Lepola, MSc, for their support with the statistics and Sirkka-Liisa Leinonen for revising the English of the manuscript.

I am deeply grateful to all the children who participated to this study and to their parents. This study would not haven been possible without you.

I owe my warmest thanks to my friends for years, Tuula Tervo-Määttä and Tuomo Määttä, Satu Aaltonen, Eeva-Liisa Övermark, Leena Raveikko and Leena Rantala. Your friendship has given me the possibility to examine the finer nuances of life and to grow up as a person.

My parents, Kerttu and Martti Jansson, I thank you for the loving care through all my life. Whenever, day or night, I needed you, you were there. My sister Eija and her husband Reijo, you are so close and dear to me. You, together with your children, Osku and Sandra, have brought so much delight into my life.

This work would not have been possible without my loving family. I would like to express my deepest and warmest gratitude to my husband Raimo for his endless support and understanding. My children Sini-Maria, Samu-Pekka and Linda, together with Jaakko, Sini-Maria´s husband, and my tiny grand-daugher Isla, you have kept me as a living person. By being near me you have reminded me of the real values of life.

This research was supported financially by Alma och K.A.Snellman Foundation, Oulu University Hospital and Oulu University Scolarship Foundation.

Abbreviations

ABR auditory brainstem response AEP auditory evoked potential ADHD attention deficit hyperactivity disorder AS Asperger syndrome BAEP brainstem auditory evoked potential CATCH 22 acronym for Cardiac defects, Abnormal facies, Thymic hypoplasia, Cleft palate and Hypocalcemia CP cerebral palsy EEG electroencephalogram ELBW extremely low birth weight ERP event-related potential HFA high-functioning autism HL hearing level ISI interstimulus interval IVH intraventricular hemorrhage LDN late discriminative negativity lMMN late mismatch negativity MCDI MacArthur Communicative Developmental Language Inventory MLU mean length of utterance MMN mismatch negativity MRI magnetic resonance imaging NLD nonverbal learning disability PVL periventricular leukomalacia SLI Specific Language Impairment SOA stimulus onset asynchrony TsC Tuberous Sclerosis VLBW very low birth weight

List of original papers

This thesis is based on the following articles, which are referred to in the text by their Roman numerals:

I Jansson-Verkasalo, E., Valkama, M., Vainionpää, L., Pääkkö, E., Ilkko, E. & Lehtihalmes, M. (2003) Language development in very low birth weight preterm children: a follow-up study. Folia Phoniatrica et Logopaedica, in press.

II Jansson-Verkasalo, E., Čeponienė, R., Valkama, M., Vainionpää, L., Laitakari, K., Alku, P., Suominen, K. & Näätänen, R. (2003) Deficient speech-sound processing, as shown by the electrophysiologic brain mismatch negativity response, and naming ability in prematurely born children. Neurosience Letters, 348, 5-8.

III Jansson-Verkasalo, E., Korpilahti, P., Jäntti, V., Valkama, M., Vainionpää, L., Alku, P., Suominen, K. & Näätänen, R. (2003) Neurophysiologic correlates of deficient phonological representations and object naming in prematurely born children. Clinical Neurophysiology, 115, 179-187.

IV Jansson-Verkasalo, E., Čeponienė, R., Kielinen, M., Suominen, K., Jäntti, V., Linna, S-L., Moilanen, I. & Näätänen, R. (2003) Deficient auditory processing in children with Asperger Syndrome, as indexed by event-related potentials. Neuroscience Letters, 338, 197-200.

Contents

Abstract Acknowledgements Abbreviations List of original papers Contents 1 Introduction ...................................................................................................................17 2 Review of the literature .................................................................................................19

2.1 Auditory processing and language development ....................................................19 2.2 Auditory event-related potentials (ERPs) in children .............................................20

2.2.1 Exogenous (obligatory) ERPs..........................................................................21 2.2.2 Mismatch negativity (MMN) and late mismatch negativity (lMMN) .............23

2.2.2.1 Mismatch negativity............................................................................23 2.2.2.2 Late mismatch negativity (lMMN)......................................................24

2.3 Preterm birth and language development ...............................................................25 2.3.1 Preterm birth and neurobiological alterations..................................................25 2.3.2 Language impairment in children born preterm with VLBW..........................26

2.4 Neurobiological findings and language in individuals with Asperger syndrome....32 2.4.1 Neurobiological findings in individuals with Asperger syndrome...................32 2.4.2 Language in individuals with Asperger syndrome...........................................33

2.5 Auditory event-related potentials in children at risk for or with speech and language impairment..............................................................................................34

2.5.1 Obligatory ERPs..............................................................................................34 2.5.2 Mismatch negativity (MMN) and late mismatch negativity (lMMN) .............35

3 Aims of the study...........................................................................................................42 4 Materials and methods...................................................................................................43

4.1 Subjects...................................................................................................................43 4.2 Event-related potential measurements ....................................................................47

4.2.1 Stimuli .............................................................................................................47 4.2.2 Recordings and data analysis...........................................................................48

4.3 Evaluation of language skills (Study I)...................................................................50 4.4 Statistical analyses ..................................................................................................51

5 Results ...........................................................................................................................52 5.1 Auditory ERPs in children born preterm with VLBW (Studies II and III) .............52

5.1.1 Obligatory responses .......................................................................................52 5.1.2 Mismatch negativity and late mismatch negativity(Studies II and III) ............56

5.2 Language development in VLBW children (Study I) .............................................61 5.3 Auditory ERPs and naming ability in children born preterm with VLBW (Studies II and III) ..................................................................................................64 5.4 Visual inspection of MMN (Study III)....................................................................68 5.5 Auditory ERPs in children with Asperger syndrome ..............................................69

6 Discussion .....................................................................................................................74 6.1 Auditory ERPs in children born preterm with VLBW............................................74 6.2 Language development in VLBW preterm children...............................................76 6.3 Auditory ERPs and language development on a single-subject level .....................77 6.4 Auditory ERPs in children with Asperger syndrome ..............................................78

7 Conclusions ...................................................................................................................80 References Original papers

1 Introduction

Healthy infants possess well-developed acoustic capabilities from birth, allowing the perception of speech and nonspeech sounds, as shown by behavioural (Eimas et al., 1971; Jusczyk, 1999; Jusczyk & Hohne, 1997; Swingley & Aslin, 2000) and electrophysiologic methods of testing (Cheour et al., 1998a; Čeponienė et al., 2002a; Kushnerenko, 2002b). Well-functioning auditory skills constitute a solid foundation for the acquisition of speech and language (Benasich & Tallal, 2002; Cunningham et al., 2000; Kraus et al., 1996; Locke 1994, 1997).

Although most children develop language skills easily, a subset of children have great difficulty in acquiring language. Accumulating evidence suggests that the main reason for these speech and language impairments may lie in auditory perceptual deficits related to the processing of complex signals, such as speech (Bradlow et al., 1999; Kraus et al., 1996; Tallal, 1980; Uwer et al., 2002). The exact nature of the deficit, however, remains controversial. Some researchers (Benasich & Tallal, 2002; Benasich et al., 2002; Merzenich et al., 1996; Tallal et al., 1996) suggest that children with speech and language deficits suffer from temporal processing deficits, whereas others (e.g. Studdert-Kennedy & Mody, 1995) propose that the deficits arise from impaired phonologic coding.

An infant’s capacity to develop normal speech and language largely depends upon the integrity of the auditory system from the ear through the cerebral cortex (Kurzberg et al., 1988). Cortical auditory event-related potentials (ERPs) permit the assessment of several well-defined stages of central auditory processing, on which language acquisition and everyday functioning are based (Kurtzberg et al., 1988; Stapells & Kurtzberg, 1991). The mismatch negativity (MMN) component of cortical auditory ERPs reflects auditory discrimination and memory (Näätänen, 1992, 1999; Näätänen & Winkler, 1999). MMN is preattentively elicited (Näätänen, 1992) and can be recorded quite early in life (Čeponienė et al., 2002a). MMN is a valuable tool in studying the neural mechanisms underlying speech perception (Kraus & Cheour, 2000).

The present study investigated auditory processing by using auditory event-related potentials, mainly MMN, in two groups of children with developmental brain disorders with different types of language impairment, i.e., children born preterm with very low birth weight (VLBW) and children with Asperger syndrome (AS). The aim was to determine whether auditory processing, as indexed by ERPs, was deficient in these two

18

populations and, further, whether deficient auditory processing was linked to language development in the preterm population. Language development of VLBW preterm children was followed up for 4 years and investigated thoroughly. Lexical score was used as an index of language development, since lexical development is suggested to be central to language development (Bates et al., 1995; Wolf, 1991; Wolf & Obregon 1991) and to closely interact with the development of the auditory domain (Locke, 1994, 1997).

Auditory processing deficits have been reported in both groups under study, based on behavioral measurements (Menyuk et al., 1995) and cortical auditory event-related potential recordings (Kurzberg et al., 1988) in a preterm population and on autobiographies of individuals with AS (Williams, 1992). To my knowledge, there are neither studies where language development would have been followed up in young VLBW children nor follow-up studies on language development including auditory ERPs. Furthermore, although individuals with autism have been studied for years using auditory ERPs (Bruneau et al., 1987, 1999; Courchesne et al., 1985a, b; Dawson et al., 1986; Gomot et al., 2002; Lincoln et al., 1993), this is the first study on auditory processing in children with AS by using auditory ERPs.

2 Review of the literature

2.1 Auditory processing and language development

During the first few months of their lives, infants discriminate speech sounds from each other regardless of whether or not these sounds belong to the surrounding adult language (Cheour et al., 1998a; Werker, 2000). During the second half of their first year of life, infants learn to discriminate better the sounds that are typical of their native language than non-native sounds (Cheour et al., 1998a; Jusczyk et al., 1994; Werker, 2000; Werker & Desjardins, 2001). This is the point at which children begin to recognize the prosodic and intonational contours typical of their own language (Jusczyk, 1999), to understand meaningful speech (Bates et al., 1995; Jusczyk & Hohne, 1997; see also Jusczyk, 1999) and to produce their first words (Bates et al., 1995; Kunnari, 2000) and sounds of their native language in their own pre-speech babble (Kunnari, 2000; Oller et al., 1994).

Auditory discrimination of discrete speech sounds is an important but not a sufficient prerequisite for language development. The child must also learn that the speech sounds of its native language may get different manifestations in different contexts (Bishop, 1997; see also Cutler & Clifton, 1999), and that the same word varies, depending on the speaker, rate of speech, context and several other factors (Swingley & Aslin, 2000; see also Cutler & Clifton, 1999). It is assumed that the child stores phonological representations of familiar sequences of speech sounds in its long-term memory, so that these sound patterns are recognized as familiar words when encountered again (Swingley & Aslin, 2000; see also Cutler & Clifton, 1999).

The relationship between auditory discrimination and language development is interactive. The acceleration in lexical development (mostly occurring from 18 months onwards) improves the infant’s auditory discrimination skills as well as his/her efficient storage and retrieval of linguistic information (Locke, 1994, 1997). In addition, auditory discrimination and vocabulary together substantially contribute to the development of morphology (Bates et al., 1995) as well as to the development of reading and other academic abilities (Swan & Goswami, 1997a, b; Wolf, 1991; Wolf & Obregon, 1991).

The suprasegmental features of language co-occur with strings of sound segments and thus constitute a phonetic overlay that is superimposed on the segmental speech stream

20

(Harley, 2001; Snow, 2000). The suprasegmental aspects of language, called prosody (pitch, length and loudness), have several functions in speech: they segment words from the speech stream (Jusczyk, 1999), they convey the meanings of words and sentences (Lyons, 1977), and they signal the speaker’s attitude and affect (Cutler & Clifton, 1999; Snow, 2000). Hence, an ability to discriminate the suprasegmental features of language is of crucial importance. During the first year of life, healthy children are well tuned to the suprasegmental features of their native language, e.g., they listen longer to the prosody of their native language than to that of a non-native language at the age of nine months already (Jusczyk, 1999).

Auditory processing plays a crucial role in language development. Still, there are several other factors, such as visual and somatosensory perception, with emotions, that interactively contribute to the development of language skills (Clancy & Finlay, 2001). Furthermore, although neural development plays a role in language learning, learning itself affects the ability of the brain to learn something new. The relationship between neural development and the learning experience thus is bi-directional (Clancy & Finlay, 2001). The complexity of the interactions and the role of different deficiencies remain to be investigated.

2.2 Auditory event-related potentials (ERPs) in children

Auditory event-related potentials are minute and discrete electrical potentials in the electroencephalogram (EEG) (Näätänen, 1992; Regan, 1989; Stapells & Kurtzberg, 1991). The EEG represents the spontaneous, ongoing electrical activity of the brain. In contrast, ERPs are manifestations of neural activity that is specifically related, or time-locked, to sensory stimulation (Stapells & Kurzberg, 1991). ERPs are small in amplitude. Therefore, they are usually obscured by the spontaneous background EEG activity. Numerous EEG epochs must be summated and averaged to extract the ERPs from the spontaneous EEG non-time-locked to stimuli (Stapells & Kurzberg, 1991).

The ERP waveform consists of a sequence of positive and negative deflections or peaks that are named according to their polarity (positive/negative) and latency (timing relative to the stimulus onset), their serial order (Picton et al., 1974) or cognitive meaning. Although the peaks provide a useful way of measurement, they are composites of several partly overlapping components that are generated by different cerebral sources (Näätänen and Picton, 1987). Each of the auditory ERPs has unique characteristics that are age-specific. Therefore, individual responses should be compared with age-matched normative data (Stapells & Kurzberg, 1991).

When classified on the basis of their latency, auditory ERPs are called short-, middle-, and long-latency responses (Picton et al., 1974). The brainstem auditory evoked potentials (BAEP) or auditory brainstem responses (ABR) are the shortest-latency responses recordable from the scalp in response to auditory stimuli. They consist of deflections peaking within the first 10 ms following the stimulus onset (Doyle, 1999; Kurzberg et al., 1988; Näätänen, 1992) and reflect the arrival of sensory input into the cochlea and the various nuclei of the brainstem (Doyle, 1999). BAEP is widely used to

21

assess auditory sensitivity, the integrity of the peripheral and brainstem auditory system and the neural plasticity of the central auditory pathway (Doyle, 1999; Hayes et al., 2003; King et al., 2002; Kurzberg et al., 1988).

The brainstem responses are followed by middle-latency responses within a time-window of 10 to 50-60 ms from the stimulus onset (McGee & Kraus, 1996; Picton et al., 1974) and, further, by long-latency responses, peaking from about 50-60 ms onwards (Näätänen, 1992), also called cortical responses (Kurzberg et al., 1988). Long-latency ERPs provide the main ERP method used in this study.

The other way to classify auditory ERPs is to categorize them into exogenous (obligatory or sensory) and endogenous components. The obligatory ERP components are elicited by any appropriate auditory stimulus (Näätänen, 1992). They represent the brain’s response to the occurrence of the stimulus and index the physical features of the stimulus (Näätänen, 1992). Endogenous ERPs mainly reflect internally generated mental events that are related to the cognitive processes (Donchin et al., 1978). BAEPs as well as middle-latency and long-latency ERPs (P1, N1, P2, N2, N4) are classified as obligatory ERPs (Čeponienė, 2001; Näätänen, 1992). The long-latency obligatory ERPs that are relevant to the present study (P1, N2, N4) and mismatch negativity (MMN), one of the endogenous ERPs, together with late MMN (lMMN) are introduced in the following chapters. The main focus is on the data obtained from young and school-aged children, i.e. the age-range covered in the present study.

2.2.1 Exogenous (obligatory) ERPs

During childhood (1-10 years), long-latency obligatory auditory ERPs are dominated by positivity at about 85-120 ms and negativity at about 200-240 ms (Kurzberg et al., 1986), called the P100 and N250 (Korpilahti, 1996; Korpilahti & Lang, 1994) or P1 and N2 peaks, respectively (Čeponienė et al., 2001, 2002b, 2003a, b; Kushnerenko et al., 2002a; Ponton et al., 2000, 2002, fig. 1). A later range of negativity, called N4 or N450, has also been reported in children (Čeponienė et al., 1998, 2003a, b; Kushnerenko et al., 2002a).

A P1-N1b-P2-N2 complex is typical of adults (Courchesne, 1990; Ponton et al., 2002). N1b is elicited in children from about the age of 3 onwards (Paetau et al., 1995), but with interstimulus intervals of over 1 second only (Bruneau & Gomot, 1998; Čeponienė et al., 1998, 2002b). N1b becomes progressively consistent (Courchesne, 1990; Ponton et al., 2002) from the age of 10 years onwards (Ponton et al., 2002). Therefore, N1b could not be objectively defined in the present study groups, as was also the case with P2. These obligatory ERPs will thus not be described here.

P1 is the dominant peak of the obligatory ERPs in children from early childhood (Kushnerenko et al., 2002a) to school age (Cunningham et al., 2000; Korpilahti, 1996; Korpilahti and Lang, 1994; Ponton et al., 2000, 2002). P1 latency and amplitude change as a function of age (Čeponienė et al., 2002b; Cunningham et al., 2000; Korpilahti, 1996; Ponton et al., 2000, 2002; Sharma et al., 1997) and the features of the acoustic stimulus (Čeponienė et al., 2002b; Cunningham et al., 2000; Ponton et al., 2000; Sharma et al., 1997) and as a function of the interstimulus interval (ISI) (Čeponienė et al., 2002b). P1

22

peak latency decreases from about 150 ms in 1-year-old children (Kushnerenko et al., 2002a) to about 100 ms in young and school-aged children (Čeponienė et al., 2002b; Korpilahti, 1996) and further to about 60 ms in adults (Čeponienė et al., 2002b). P1 amplitude decreases along with increasing age (Čeponienė et al., 2002b; Korpilahti, 1996).

Fig. 1. A schematic illustration of the obligatory P1, N2 and N4 ERPs in children.

Based on the intra-cortical recordings of Liégois-Chauvel et al. (1994), the generator complex for P1 has been attributed to the lateral portion of Heschl´s gyrus (likely in the secondary auditory cortex) and its early-maturing cortical layers III and IV (Ponton et al., 2002). P1 has been proposed to index the transient encoding of the acoustic features of sound (Čeponienė et al., 2002b). Findings on the functional significance of this component are rather scarce, however.

N2 is a large, dominant negativity that is seen at approximately 240-280 ms in young and school-aged children (Čeponienė et al., 2002b; Cunningham et al., 2000). During the school age (8-15 years), N2 latency and amplitude remain stable (Cunningham et al., 2000; Korpilahti, 1996). In adolescence, both latency and amplitude decrease, the latency in adults being about 200-220 ms and the amplitude about 1 µV. The N2 parameters change as a function of stimulus type (Čeponienė et al., 2001), which probably explains why Ponton et al. (2002) obtained somewhat contradictory results in response to clicks compared to the findings of Cunningham et al. (2000) and Korpilahti (1996). N2 latency increased as a function of age up to the age of 14 years, whereas amplitude increased in 10-year-olds, and decreased thereafter, reaching adult values at the age of 17 (Ponton et al., 2002). Single-sweep recordings showed that N2 amplitude increased with stimulus repetition (Karhu et al., 1997), and this was suggested to index the building-up of a neural representation, or a sensory memory trace, of the repeated stimulus. The neural generators of N2 in children are not well known.

N4, a negativity at about 380-460 ms in children (Čeponienė et al., 1998, 2003a; Kushnerenko et al., 2002a), is included in the obligatory ERPs, for it is elicited in the same kind of paradigm as N2 in children (Čeponienė et al., 1998, 2002b, 2003a; Kushnerenko et al., 2002a). N4 becomes robust during infancy (Kushnerenko et al.,

23

2002a) and diminishes in amplitude thereafter (Čeponienė et al., 2002b). The functional significance of N4 is unknown.

2.2.2 Mismatch negativity (MMN) and late mismatch negativity (lMMN)

2.2.2.1 Mismatch negativity

Mismatch negativity (MMN) is a negative waveform of endogenous auditory ERPs that is elicited when any discriminable change (deviant) occurs in the sequence of repetitive (standard) sounds (Näätänen, 1992, 2003; Näätänen et al., 1978; Näätänen & Winkler, 1999). In addition to simple stimulus features, such as the frequency and duration of the sine tone (Čeponienė et al., 2002c, 2003b; Cheour et al., 1997; Gomot et al., 2000; Korpilahti, 1995, 1996; Korpilahti & Lang, 1994; Korpilahti et al., 2001; Schulte-Körne et al., 1998; Uwer et al., 2002), more complex and abstract sound changes, such as tone pairs (Kujala et al., 2001b), phonemes (Čeponienė et al., 2003a, b; Cheour et al., 2002), syllables (Kraus et al., 1996; Schulte-Körne et al., 1998; Uwer et al., 2002) and even words, elicit MMN in children (Korpilahti et al., 2001) and in adults (Pulvermüller et al., 2001; Shtyrov et al., 2002). The value of MMN is usually obtained by subtracting the ERP to a standard stimulus from that to a deviant stimulus.

The elicitation of MMN is based on the presence of a memory trace (cortical representation) caused by a preceding stimulus. The memory traces on which MMN is based can be either short-term (Yabe et al., 1997) or long-term (Cheour et al., 1998a; Näätänen et al., 1997b; Winkler et al., 1999; see also Cowan et al., 1993) in nature. One can probe these traces by presenting deviant stimuli of different magnitudes of deviation and thus indirectly determine the accuracy of these central sound representations (Näätänen, 2003). MMN is a preattentive neural correlate of stimulus change (Kraus et al., 1995; Näätänen et al., 2001; Näätänen & Winkler, 1999), which indexes individual discrimination accuracy (Näätänen, 1992; Näätänen & Alho, 1997) and correlates well with the behavioural discrimination ability in children (Kraus et al., 1996; Lang et al., 1989) as well as in adults (Kujala et al., 2001a; Lang et al., 1990).

MMN is well-established in infants (Cheour, 1998; Kushnerenko, 2003; Kushnerenko et al., 2002b), children (Čeponienė, 2001; Čeponienė et al., 2003a; Gomot et al., 2000; Korpilahti, 1996; Uwer et al., 2002) and adults (Näätänen & Alho, 1995a, b; Näätänen & Winkler, 1999). In children, the MMN component peaks between 130 ms (Gomot et al., 2000) and 400 ms (Čeponienė et al., 2003a), depending on age and the stimuli and paradigm used. The MMN response is rather stable in terms of latency and amplitude by school age (Csépe, 1995). A recent study (Shafer et al., 2000) showed that when a similar paradigm was used for subjects whose ages ranged from 4 to 10 years, only a slight decrease was seen in MMN latency. Furthermore, there were no significant changes in

24

MMN amplitude. In adults, however, MMN latency was significantly shorter and amplitude significantly smaller than in children (Shafer et al., 2000).

Studies concerning MMN generators in children are scarce. Gomot et al. (2000) showed that MMN in children includes both bilateral temporal and frontal components, as in adults (Alho, 1995; Giard et al., 1999; Hari et al., 1984). In children, a parietal component was also present (Gomot et al., 2000). MMN may, in addition, have subcortical components, as revealed by animal studies (Csépe, 1995; Kraus et al., 1994). The exact location of the MMN generators varies in adults as a function of, for example, the stimuli (Alho, 1995; Kraus et al., 1994; Rinne et al., 1999; Shtyrov et al., 2000), the masking effect (Shtyrov et al., 1998), the sensory functions and their re-organization (Kujala et al. 1995; see also Kujala et al., 2000).

In adults, the auditory cortex activity is assumed to reflect the automatic pre-perceptual change detection process, i.e., the comparison of a new stimulus with the stimulus that is represented in the auditory sensory memory (Näätänen, 1992). The activation of the frontal generator has been proposed to play a role in the initiation of an involuntary attention switch to sound change that is preperceptually detected in the auditory cortices (Giard et al., 1990). The functional significance of the components found in children remains to be verified (Gomot et al., 2000).

MMN can be used to study the maturation and development of central auditory processing (Čeponienė, 2001; Cheour, 1998; Kushnerenko, 2003). Furthermore, MMN is a powerful tool towards understanding the neural bases of speech perception (Kraus and Cheour, 2000) underlying language deficits (Korpilahti, 1995, 1996; Korpilahti & Lang, 1994; Kraus et al., 1996; Kujala & Näätänen, 2001; Leppänen et al., 2002; Schulte-Körne et al., 1998; Uwer et al., 2002; see also Cheour et al., 2001). Futhermore, MMN can be used as an index of recovery of auditory discrimination (Ilvonen et al., 2003) and to assess neural plasticity following learning in healthy children (Cheour et al., 2002) and rehabilitation in clinical populations (Kujala et al., 2001b).

2.2.2.2 Late mismatch negativity (lMMN)

A longer-latency ERP response, elicited together with MMN, was termed as late mismatch negativity (lMMN) by Korpilahti (1995, 1996: Korpilahti et al., 2001), as MM4 by Kraus et al. (1993) and as late difference negativity (LDN) by Čeponienė et al. (1998) and Cheour et al. (2001). Similarly to MMN, lMMN can also be elicited from birth (Čeponienė et al., 2002a; Kushnerenko, 2001; Kushnerenko et al., 2002b), and it has been mainly reported in children (Čeponienė et al., 1998, 2002a, 2003a; Korpilahti, 1996; Korpilahti et al., 2001; Kraus et al., 1993; Schulte-Körne et al., 1998). In contrast to MMN, lMMN greatly diminishes in amplitude during adolescence, being only rarely observed in adults (Cheour et al., 2001). In children, lMMN peaks in the latency range of 350-550 milliseconds in response to pure tones and vowels (Čeponienė et al., 2003a; Korpilahti, 1996; Korpilahti et al., 2001), the latency being longer for more complex stimuli, such as words (Korpilahti, 1996; Korpilahti et al., 2001).

25

The functional significance of lMMN is rather unclear. Recently, lMMN has been reported in infants in response to harmonic tones (Čeponienė et al., 2002a; Kushnerenko et al., 2002). Poor sensitivity to stimulus features was reported by Čeponienė et al. (2002c). However, Korpilahti et al. (2001) found that the lMMN amplitude was more pronounced for words compared to pseudowords and tones in school-aged children. Therefore, it was suggested that lMMN may reflect automatic detection of lexical differences. Although very little is known about lMMN, its amplitude was reported to be diminished in different clinical child populations (Kilpeläinen et al., 1999; Schulte-Körne et al., 1998). For this reason, lMMN was also included in the present study.

2.3 Preterm birth and language development

2.3.1 Preterm birth and neurobiological alterations

An infant who is born before 37 completed gestational weeks is defined as preterm, and infants born before 32 completed gestational weeks are called very preterm. Further, an infant is extremely preterm when he/she is born before 28 completed gestational weeks (ICD-10, 1999). A preterm infant with birth weight of less than 1500 g is classified as a very low birth weight (VLBW) preterm infant. When birth weight is less than 1000 g, the preterm infant is called extremely low birth weight (ELBW) preterm infant (ICD-10, 1999). In addition, preterm children with birth weight 2 or more standard deviations (SD) or 10 percentiles below the mean value for the infant’s gestational age (ICD-10, 1999) are classified as small for gestational age (SGA).

Hemorrhage, hypoxia, ischemia (Inder & Volpe, 2000; Ment et al., 2002) and fetal inflammation (Damman et al., 2001) have been linked with brain injury in preterm children. Periventricular white matter damage or periventricular leukomalacia (PVL) (Inder et al., 1999), which often coexist with intraventricular hemorrhage (IVH) (Inder & Volpe, 2000), have been suggested by Inder et al. (1999) as being the dominant form of brain injury in premature infants. PVL affects the early development of cortical grey matter, increases the ventricular and extracerebral cerebro-spinal fluid volume (Inder et al., 1999; Inder & Volpe, 2000) and disrupts the callosal fibres that mediate higher cortical functions (Whitaker et al., 1996). Changes in the thalamus have also been related to PVL (Krägeloh-Mann et al., 1999; Lin et al., 2001).

PVL is associated with language and cognitive impairments (Frisk & Whyte, 1994; Holling & Leviton, 1999; Krägeloh-Mann et al., 1999; Whitaker et al., 1996), cerebral palsy (CP) (Holling & Leviton, 1999; Ment et al., 2002), visual (Jacobsen et al., 1996; Krägeloh-Mann et al., 1999) and hearing impairments (Dusik, 1997) as well as attention deficits (Krägeloh-Mann et al., 1999). Similarly, IVH contributes to CP (Vohr et al., 2003), hearing impairment (Vohr et al., 2003) and cognitive (Ross et al., 1996; Vohr et al., 2003) as well as language development (Frisk & Whyte, 1994; Grunau et al., 1990; Ment et al., 2003; Vohr et al., 2003). However, the pathogenic pathways for these impairments are not fully understood.

26

VLBW infants account for approximately 1 percent of all live births (Greene, 2002). The median incidence of CP among VLBW children who survive is 7-24 % (Escobar et al., 1991; Valkama, 2001). The incidence of hearing impairment varies from about 1 percent up to about 12 percent (Dusick, 1997; Valkama, 2000a; Veen et al., 1991; Wood et al., 2000). Conductive losses are more common than sensorineural hearing loss (Dusick, 1997). Up to 40 percent of preterm children exhibit language impairments and disabilities (Veen et al., 1991).

2.3.2 Language impairment in children born preterm with VLBW

During the first two years of life, language comprehension and production have been suggested to be equal in VLBW preterm children and their controls when assessed by word comprehension (Menyuk et al., 1991, 1995) and the production of the first words (Jennische & Sedin, 1999b; Menyuk et al., 1991, 1995) (table 1, see also for the other references here).

At the ages of 2 (Vohr et al., 1988) and 3 years (Grunau et al., 1990; Menyuk et al., 1991; Singer et al., 2001), the scores of compound measures of language comprehension were, however, lower in VLBW preterm children than in full-term controls. Word comprehension was deficient in VLBW preterm children at the age of 3 years (Briscoe et al., 1998; Grunau et al., 1990) and, more specifically, in VLBW preterm children with bronchopulmonary dysplasia (BPD) (Singer et al., 2001) and IVH (Ment et al., 2003). Deficient word comprehension persists up to the age of 5 (Frisk & Whyte, 1994; Luoma et al., 1998) and 6 years (Jennische & Sedin, 2001b).

Deficits in language comprehension also manifest at the sentence level. VLBW preterm children cope less well in sentence comprehension tasks from 3 years (Grunau et al., 1990) up to 5 (Luoma et al., 1998) and 6 years of age (Frisk & Whyte, 1994; Jennische & Sedin, 2001b; Wolke & Meyer, 1999). Deficient language comprehension was reported up to the age of 8 years in a preterm population (Yliherva, 2002).

According to parent reports, the speech of VLBW preterm children is less intelligible for the parents than the speech of healthy children. In addition, short sentences were produced later by preterm children than by healthy children (Jennische & Sedin, 1999b). Similarly, compound measures of speech production have revealed that VLBW preterm children score lower in speech production than healthy controls at the age of 2 years (Menyuk et al., 1991; Riitesuo, 2000; Vohr et al., 1988). Deficits in lexical development have been reported at the ages of 5 (Luoma et al., 1998) and 6 years (Frisk & Whyte, 1994; Jennische & Sedin, 1998). In grammatical development, the deficit manifests as a less complex sentence structure from the age of 3 years (Grunau et al., 1990) up to the age of 6 years (Jennische & Sedin, 1998, 2001a; Wolke & Meyer, 1999). At school age, deficient language development manifests in multiple academic domains (Grunau et al., 2002)

The results concerning language development in VLBW preterm children are not conclusive, however. Age correction for prematurity has a substantial effect on the results (Menyuk et al., 1995). At present, consensus has not yet been reached with regard to the

27

length of time for which age correction for prematurity is needed. Furthermore, the standardized test norms may be outdated (Wolke et al., 1994). Therefore, the use of age correction and reliance on test norms without a matched control group (Ment et al., 2003) may lead to underestimation of the deficiency that may manifest in preterm children (Wolke et al., 1994). Furthermore, the composite measures reported in many studies (Fawer et al., 1995; Koller et al., 1997; Wood et al., 2000) may mask a specific deficiency. A specific deficiency can only be found by specific tests (Siegel, 1994). Accordingly, Luoma et al. (1998) found that, although the preterm group did not differ from the matched control group in the global verbal measures of an intelligence (IQ) test, the preterm group exhibited difficulties in naming and in the comprehension of relative concepts. Moreover, it is also possible that language deficits become more prominent when the child grows up.

Deficient auditory processing may contribute to the kinds of language deficits encountered in VLBW preterm populations (Kurzberg et al., 1988; Menyuk et al., 1995). It is possible that the effect of auditory dysfunction is specifically seen in lexical development, since lexical acquisition is, presumably, a process of storing representations of familiar sequences of speech sounds in a mental lexicon and associating these with specific meanings (Bishop, 1997). The early representations of words may not be specified in full phonetic detail (Bishop, 1997; Locke, 1997; Swingley & Aslin, 2000). However, if the child remains at the stage of immature perceptual categorization of speech input and unable to identify subsyllabic elements, such as phonemes, he/she may have difficulties in remembering novel phonological strings (Bishop, 1997). This may lead to difficulties in learning new words and, further, to limited lexical development and complicate word production more than word comprehension. The comprehension of words is possible on the basis of a less than complete phonological representation (Bishop, 1997). However, partial phonological representation is not sufficient for accurate production (Bishop, 1997). While lexical development was suggested to be a prerequisite for other language skills, such as morphological development (Bates et al., 1995), the main focus of the present thesis was on lexical development and auditory processing.

Biological factors, including PVL (Frisk & Whyte, 1994; Krägeloh-Mann et al., 1999) and IVH (Frisk & Whyte, 1994; Ment et al., 2003; Vohr et al., 2003), significantly contribute to the language deficiencies encountered in VLBW preterm children. However, social and environmental factors, including the maternal level of education and the primary language spoken at home, are also important contributors (Ment et al., 2003; Vohr et al., 2003).

Studies of serial language testing of VLBW preterm infants are scarce (Ment et al., 2003; Menyuk et al., 1991, 1995; Vohr et al., 1988, 1989). In addition, most studies have not followed up the same cohort of children or used the same tests for their evaluation (Vohr et al., 1988, 1989). Therefore, very little is known about the course of development of special language areas in VLBW children. Ment et al. (2003) proposed that serial studies should be conducted to study the language development of VLBW infants with appropriate control-matched participants. This was done in the present study.

28

Table 1. Studies of language development among very low birth weight and very preterm children from the year 1990 onwards. Only studies on language development during the first 6 years of life are included.

Reference Subjects of the study

Age of assessment

Methods used to assess cognitive and language development

Main results concerning language development

1.Grunau et al., 1990

23 ELBW preterm children, 23 matched controls

3;0-3;4 years CCA

PPVT MLU SALT Stanford Binet

ELBW children scored lower in PPVT, SALT, language comprehension and memory for sentences in Stanford-Binet. No difference in MLU.

2. Menyuk et al., 1991*

26 preterm, 12 VLBW preterm infants, 27 matched controls

0;1-3;0 years CCA and CA

Audio and video recordings, diaries, Uzgiris and Hunt, MacCarthy Scales, Bayley, SICDR & E, PPVT, NSST, DSS, Reynell

Lexical development was the main focus of the study. The first 10 words were learned later by the VLBW group. VLBW children scored lower on SICD, but were in the normal range. No differences in PPVT or other lexical measures.

3. Frisk & Whyte, 1994

68 ELBW children (26 with mild PVL, 15 with severe PVL), 20 controls.

6 years PPVT, EOWPVT, McCarthy Scales, TRG, Token

Trend towards lower scores on PPVT and significantly lower scores on EOWPVT, TRG, Token and short-term memory test among ELBW children with PVL.

4. Menyuk et al., 1995*

28 preterm children and 28 controls, 12 VLBW children

0;1-3;0 years Audio and video recordings (e.g. cry, vocalization, sounds, vocabulary, syntax, sentence comprehension and production, language functions, conversation), auditory discrimination, diaries, CDI, Uzgiris and Hunt, MacCarthy Scales, Bayley, SICDR & E, DSS, PPVT, NSST, Reynell

VLBW children had significantly lower scores on the SICD receptive and expressive vocabulary test. Tendency for VLBW children to be slightly behind other preterm children and controls in most tests. Still, their performance was within the normal range. VLBW children learned later to complete the auditory discrimination task.

5. Briscoe et al., 1998

26 very preterm children (GA<32 weeks), 32 controls

3;3-4;2 years CCA

BPVT (receptive vocabulary), MacCarthy Scales (expressive vocabulary), Bus Story Test, Phonological short-term memory (digit span, nonword repetition)

Very preterm children scored significantly lower on BPVT and the Bus Story Information test. No other differences.

29


Age of assessment



6. Luoma et al., 1998

55 very preterm children (GA<32 weeks), 55 matched controls

5;0-5;6 years NEPSY, WPPSI, Token, RAN

Preterm children scored significantly less well on all measures. When children with neurological disabilities were excluded, the preterm group still exhibited difficulties in naming and sentence comprehension.

7. Jennische & Sedin, 1998

284 children who required NIC, 81 very preterm children (GA<32 weeks), 40 controls

Mean 6;6 years CCA

Audio-recording (information score, speech motor, sound pattern, word finding, word selection, grammar, pragmatics)

Very preterm children scored significantly lower in all other measured areas but information giving. Most preterm (GA <28) children showed significantly better spontaneous speech than children born at the GA of 28-32 weeks.

8. Jennische & Sedin, 1999a


Mean 6;6 years CCA

Motor functions, imitation (articulatory positions, articulatory pattern, sentence), language comprehension (logical grammatical constructions, retelling story, following instructions), auditory discrimination, phoneme inventory, auditory short-term memory, PPVT

Very preterm children scored significantly lower in all other measures except word fluency. PPVT score correlated significantly with imitation tasks, auditory discrimination and auditory short-term memory. Preterm children born at the GA of 28-32 weeks had more pronounced difficulties than children born at <28 weeks.

9. Jennische & Sedin, 1999b


Mean 6;6 years CCA at the time of interview

Parent interview concerning early language development

Short sentences and intelligible speech were reached later by very preterm children than controls. Absence of babbling was common in most preterm children (GA < 28 weeks). First words were learned equally early by preterm and control children.

30


Age of assessment



10. Wolke & Meyer, 1999

264 very preterm children (GA <32 weeks), 264 matched controls. 311 normative controls

6;3 years Kaufman Assessment Battery (IQ), HSET (grammatical rules, language comprehension and production), articulation test, videorecording (quality of speech, e.g., grammatical correctness, voice quality, melody, volume, tempo of speech), pre-reading skills.

Very preterm children had significantly poorer language abilities in all subtests. Most problems pertained to grammatical rules and detecting semantically incorrect sentences. Significant deficits also in pre-reading skills.

11. Riitesuo, 2000*

24 very preterm infants (GA<32 weeks)

0:0-2;0 years CCA and CA

Videorecording, ASQ, Bayley, REEL-2, Reynell

Speech production tended to develop more slowly than other skills. Children born at GA of 24-28 performed under the expected levels in different areas of development at CCA of 2 years.

12. Jennische & Sedin, 2001a


Mean CCA 6;5 years

Same as in study 7. Significantly lower scores for the very preterm group in speech motor skills, sounds, grammar and motivation for conversation compared with matched controls. Word finding and selection scores for most preterm children (GA 23-27 weeks) significantly lower compared to matched controls.

13. Jennische & Sedin, 2001b


Mean CCA 6;5 years

Same as in study 8 and digit recall from ITPA

Very preterm children scored significantly lower in mouth and articulatory positions, the comprehension of logical grammatical structures, phoneme inventory and word fluency when compared with matched controls. PPVT score was also significantly lower in the preterm group.

14. Singer et al., 2001

90 VLBW children with BPD and 65 without BPD, 91 term controls

3;0 years Bayley, Communication Domain Subscale of the Battelle Development Inventory (receptive and expressive language, communication)

Receptive language score significantly poorer in the preterm groups. In addition, VLBW children with BPD achieved significantly lower scores than controls in expressive language and in communication.

31


Age of assessment



15. Ment et al., 2003*

296 VLBW infants

36, 54, 72, 96 months

Stanford-Binet, WPPSI-R, PPVT

Most children showed improvement in PPVT-R scores with increasing CA. VLBW children with IVH and subsequent central nervous system injury had lower PPVT-R scores, and their scores declined over time.

16. Jennische & Sedin, 2003


Mean CCA 6;5 years

Same as in study number 13.

Language development of very preterm boys was less affected than that of very preterm girls. Very preterm girls had significantly lower scores than controls in speech motor functions and in communication. Similarly, very preterm girls lagged behind controls in the imitation of mouth positions and articulatory patters, language comprehension, auditory discrimination, word fluency and PPVT. Very preterm boys were poorer than controls in information score, speech motor, articulatory positions, phoneme inventory and auditory memory.

ASQ= Ages and stages Questionnaire; BPD=bronchopulmonary dysplasia; BPVT=British Picture Vocabulary Test; CA=chronological age; CCA=age corrected for prematurity; CDI=MacArthur Communicative Development Inventories; DSS=Developmental Sentence Scoring; ELBW= Extremely low birth weight (birth weight <1000 g); EOWPVT=Expressive One Word Picture Vocabulary Test; GA=gestational age; HSET=Heidelberger Sprachentwicklungstest; ITPA=Illinois Test of Psycholinguistic Abilities; IVH= intraventricular hemorrhage; McCarthy Scales= McCarthy Scales of Children´s Ability; MLU=Mean Length of Utterance; NEPSY= Neuropsychological Investigation of Children; NIC=neonatal intensive care; NNST=Northwestern Syntax Screening Test; PPVT =Peabody Picture Vocabulary Test; PVL=periventricular leukomalacia; RAN=Rapid Automatic Naming Test; REEL-2=Receptive-Expressive Emergent Language Scale; Reynell=Reynell Developmental Language Scales; SALT=Systematic Analysis of Language Transcript (sentence complexity); SICDR & E=Sequenced Inventory in receptive and expressive language; Token=Token Test for Children; TRG=Test for the Reception of Grammar; Uzgiris and Hunt= Uzgiris and Hunt scales of psychological development; VLBW=very low birth weight; WPPSI=Wechsler Preschool and Primary Scale of Intelligence; * follow-up study.

32

2.4 Neurobiological findings and language in individuals with Asperger syndrome

2.4.1 Neurobiological findings in individuals with Asperger syndrome

Asperger syndrome (AS) is a pervasive developmental disorder (PDD) that is diagnosed on the basis of a cluster of cognitive, social and motor signs (Gillberg & Gillberg, 1989; ICD-10, 1999). The characteristic features of AS are deficits in social interaction and unusual, rigidly ritualistic interests (Gillberg & Gillberg, 1989; ICD-10, 1999). Prevalence data on AS are scarce. In Sweden, 0.71 % of children were diagnosed as having AS (Ehlers & Gillberg, 1993), while the prevalence of AS in Finland remains to be determined.

AS shares some essential symptomatology with autism, and it is hence not clear whether autism and AS are distinct diagnostic entities (Schultz et al., 2000; Szatmari et al., 2000) or whether AS is equivalent to high-functioning autism (HFA) (Casanova et al., 2002; Howlin, 2003; Ramberg et al., 1996). The main distinguishing features used to discriminate between AS and HFA are the lack of clinically significant language and cognitive delay in AS (ICD-10, 1999) and the motor clumsiness present in children with AS but not in those with HFA (Gillberg & Gillberg, 1989; Schulz et al., 2000). Recent studies have shown, however, that although the early language history may have been different in AS and autism, the speech and prosody characteristics of individuals with AS and HFA may be equal in adolescence and adulthood (Gilchrist et al., 2001; Shriberg et al., 2001). The exact mechanism underlying the difference in the outcome remains to be determined. It is possible that AS and autism may represent different, but potentially overlapping, developmental trajectories that are influenced by similar, though not identical, prognostic factors (Starr et al., 2003; Szatmari et al., 2000)

Another disorder that bears striking similarity to AS is the nonverbal learning disability syndrome (NLD) (Ellis & Gunter, 1999; Gunter et al., 2002; Rourke, 1987, 1988, 1995). The typical features of NLD are deficits in non-verbal problem solving, visual-spatial organization, psychomotor coordination and the understanding and expression of the pragmatic and prosodic aspects of language, difficulty in adapting to novel and complex situations and superior verbal memory (Rourke, 1995).

Attention deficit hyperactivity disorder (ADHD) is often seen in individuals with AS (Ghaziuddin et al., 1998; Ramberg et al., 1996; Schatz et al., 2002). It is to be determined by further studies whether ADHD is a comorbidity of AS or whether ADHD is secondary to a deficit intrinsic to AS, including the unusual use of language, odd socialization, obsessive thought, etc. (Schatz et al., 2002).

The neural underpinnings of AS are not yet understood (Casanova et al., 2002; McKelvey et al., 1995; Schulz et al., 2000). There are probably separate neural systems that mediate different deficits encountered in individuals with AS (Schultz et al., 2000). Brain imaging data strongly suggest that the pathobiology of social awareness involves the limbic system, as previously assumed to be the case in autism (Damasio & Maurer, 1978), including the amygdala and the cortical system functionally connected with it, i.e.

33

mainly the prefrontal and temporal cortices (Baron-Cohen et al., 1999; Schultz et al., 2000), with a bias towards right-hemisphere dysfunction (Ellis & Gunter, 1999; Gunter et al., 2002; McKelvey et al., 1995). It has also been proposed that, similarly to NLD, the main cause of AS is a dysfunction of white matter (Ellis & Gunter, 1999; Volkmar et al., 2000), which might affect all aspects of brain activity, but, to an even greater extent, activities exclusively or mainly carried out in the right hemisphere (Ellis & Gunter, 1999). Further studies are, however, needed to explore more conclusively both the symptomatology and the etiology of AS.

2.4.2 Language in individuals with Asperger syndrome

AS is characterized by marked difficulties in reciprocal social interaction with fluent but pragmatically impaired speech (ICD-10, 1999). Language development is assumed to be fairly normal (ICD-10, 1999; Gilchrist et al., 2001) and literal language skills well-developed (ICD-10, 1999). Howlin (2003) stresses that it may be misleading to say that children with AS do not have significantly delayed language skills, since their adulthood level of language is markedly impaired (Howlin, 2003). More studies are needed concerning language, as there are only very few studies on language in children with AS where standardized language measures have been used (Dennis et al., 2001; Howlin, 2003; Ramberg et al., 1996; Szatmari et al., 2003).

Individuals with AS display substantial difficulties in the comprehension and production of prosodic features, manifested as, for instance, pedantic, affectless and monotonous speech with peculiar voice characteristics (Ghaziuddin & Gerstein, 1996; Gillberg & Gillberg, 1989). It was recently shown that not only affective prosody but also other suprasegmental aspects of speech, i.e., grammatical and pragmatic prosody, were affected in individuals with AS (Shriberg et al., 2001). Individuals with AS talk more than controls or individuals with HFA. In addition, inappropriate or nonfluent phrasing was typical of individuals with AS together with inappropriate stress patterns and too loud and too high a voice (Shriberg et al., 2001). The authors (Shriberg et al., 2001) proposed that these deficits are possibly due to speech-motor control deficits, including difficulties in self-monitoring associated with poor auditory or proprioceptive feedback systems.

Individuals with AS also have outstanding deficits in understanding hidden meanings, other people’s emotions (Gillberg & Gillberg, 1989) as well as the intonation and prosody they hear (Attwood, 1998). Accordingly, a recent study showed that individuals with AS had difficulty in extracting mental state information from the speech they heard. It was also shown that the difficulty was not due to deficient vocabulary (Rutherford et al., 2002).

The theory of mind (ToM) explanation has been applied to the communicative deficits encountered in individuals with AS (Baron-Cohen et al., 1999; see also Volkmar, 1998). In the present work, the main focus is on the auditory domain and language. Therefore, studies concerning ToM are not quoted.

34

Speech perception and production are monitored by the auditory domain. Auditory processing in individuals with autism has been studied for a long time using auditory ERPs (Bruneau et al., 1987, 1999; Courchesne et al., 1985a, b; Dawson et al., 1986, 1988; Dunn et al., 1999; Gomot et al., 2002; Lincoln et al., 1993). However, very little is known about auditory processing in children with AS. Autobiographies of several persons with AS refer to problems with auditory perception (Williams, 1992; see also Attwood, 1998). More research is needed on the interrelations of language and auditory processing in children with AS.

2.5 Auditory event-related potentials in children at risk for or with speech and language impairment

2.5.1 Obligatory ERPs

Studies involving obligatory ERPs in children with language deficits are scarce (Cunningham et al., 2000), being mainly reports of an oddball paradigm and along with MMN parameters (Korpilahti, 1995, 1996; Korpilahti & Lang, 1994; Ceponiene et al., 2003b).

Cunningham et al. (2000) studied obligatory ERPs together with behavioural measures. They found that the auditory-based learning measures (incomplete words, sound blending, sound pattern, listening comprehension, auditory processing) strongly predicted N2 latency in a group of children with learning problems, but not in a group without learning problems. In addition, the P1-N1-N2 parameters correlated with intelligence (IQ) in both groups. In contrast, these auditory ERPs did not correlate with performance in fine-grained speech discrimination tasks (Cunningham et al., 2000), like MMN (Kraus et al., 1996). Therefore, it was suggested that P1-N1-N2 and MMN responses reflect different aspects of auditory function. The obligatory ERPs P1-N1-N2 may reflect the neural encoding of sound features in the thalamo-cortical segment of the auditory pathway (Cunningham et al., 2000). The basic encoding, however, serves a wide range of auditory perceptual functions and contributes to listening comprehension.

Changes in N2 have also been reported in children with specific language impairment (SLI) (Korpilahti & Lang, 1994; Tonnquist-Uhlen, 1996). Korpilahti and Lang (1994) found that the N2 amplitude was smaller and latency longer in children with SLI than in their controls. These findings were supported by the results of Tonnquist-Uhlen (1996).

In autistic children, N2 latency and amplitude were found to be similar (Čeponienė et al., 2003b; Gomot et al., 2002), whereas P1 amplitude was smaller in response to vowels, or to complex and simple tones, than in the controls (Čeponienė et al., 2003b). The attenuation of P1 was suggested to index dysfunction of early sensory processes in children with autism (Čeponienė et al., 2003b).

To my knowledge, there are no studies investigating auditory ERPs together with language measures in VLBW preterm children or in children with AS.

35

2.5.2 Mismatch negativity (MMN) and late mismatch negativity (lMMN)

Auditory processing deficits, such as difficulty in auditory discrimination or memory, may be captured by behavioural methods (Benasich et al., 2002; Briscoe et al., 1998; Čeponienė et al., 1999b; Cunningham et al., 2000; Kraus et al., 1996). Although a relationship exists between the behavioural measures of auditory discrimination, short-term memory and MMN (Bradlow et al. 1999; Čeponienė et al., 1999b; Kraus et al., 1996), they represent different levels of processing. The behavioural response involves participation, focused attention and later phonetic processing, whereas MMN is a preattentive, neural correlate of stimulus change (Bradlow et al., 1999; Kraus et al., 1995; Näätänen & Winkler, 1999). As such, MMN can be used to study the neurobiological bases of speech and language deficits (Kraus et al., 1995, 1998; Kraus & Cheour, 2000; Kujala & Näätänen, 2001) and maturation as well as neural plasticity (Bradlow et al., 1999; Kraus et al., 1998; Kujala et al., 2001b) in infants and children at risk for or with speech and language deficits. Recent studies have shown that MMN can also be used to identify infants at risk for speech and language deficits (Leppänen et al., 2002; Lyytinen et al., 2001).

The MMN component of auditory event-related potentials has mostly been investigated in children at risk for or with SLI (Holopainen et al., 1996, 1997; Korpilahti, 1995, 1996; Korpilahti & Lang, 1994; Uwer et al., 2002), autism (Čeponienė et al., 2003b; Gomot et al., 2002; Kemner et al., 1995; Seri et al., 1999), dyslexia (Kujala et al., 2001b; Leppänen et al., 2002; Schulte-Körne et al., 1998), oral clefts (Čeponienė et al., 1999a, 2000; Cheour et al., 1997, 1998a, 1999) or learning disabilities (Bradlow et al., 1999; Kraus et al., 1996). Studies where MMN has been used to investigate children with or at risk for speech and language deficits are listed in table 2.

36

Table 2. Mismatch negativity (MMN) and late mismatch negativity (lMMN) in infants and children at risk for or with speech and language deficits.

Diagnostic group/Reference

Subjects Methods used to assess language development

ERP methodology (stimuli, ISI/SOA)

Main results concerning MMN and lMMN

Specific language impairment (SLI)

Korpilahti & Lang, 1994

14 children with SLI (8-13 years, 12 controls (7-13 years)

Boston naming test, Boehm, ITPA

Frequency: standard 500 Hz, deviant 553 Hz; 100 ms. Duration: standard 50 ms, deviants 110 or 500 ms; 1000 Hz; ISI 350 ms.

MMN for the frequency and long durational change significantly attenuated in the SLI group. Negative correlation between the peak latency of frequency MMN and age in controls, but not in SLI children.

Korpilahti, 1995 10 children with SLI (8-13 years), 10 controls (7-13 years)

Boehm, ITPA, Seashore, phoneme discrimination, sentence comprehension and instruction test by Korpilahti (1990, 1991, 1993), the rhythm subtst of BMA

Frequency: standard 500 Hz, deviant 553 Hz; 100 ms; ISI 350 ms.

MMN attenuated in SLI children. Maximal MMN amplitude at F4 in controls. SLI children showed no hemisphere predominance. Specific word finding related to phoneme discrimination defect. None of the behavioural discrimination tests alone predictive for MMN amplitude.

Holopainen et al., 1996

10 children with SLI (3-6 years), 14 controls (3-7 years)

Reynell, ITPA, Bo Ege naming test


MMN attenuated in SLI group. No difference in MMN latency. In controls, the maximal MMN amplitude located frontally with a trend to right-hemisphere dominance. In the SLI group, the maximal MMN located centrally in the midline area.

37





Holopainen et al., 1997

13 children with SLI (5-9 years), 12 mentally retarded (MR) (5-8 years), 10 controls (5-9 years)

Reynell, ITPA, Bo Ege naming test


MMN attenuated in SLI and MR groups. No difference between the SLI and MR groups. Maximal MMN amplitude in the MR group mostly in the left hemisphere, in the dysphasic group in the central region, and in controls in the right hemisphere and in the central regions.

Uwer et al., 2002

21 children with expressive and 21 with receptive SLI, 21 controls (mean age of all 8; 2 years)

HSET, imitating grammatical structures, comprehending grammatical structures, auditory discrimination test

Frequency: standard 1000 Hz, deviant 1200 Hz, 175 ms. Duration: standard 175 ms, deviant 100 ms, both 1000 Hz. Syllables: standard /da/, deviants /ga/ and /ba/; 175 ms. SOA 1 second.

MMN attenuated in SLI to speech stimuli, but no difference between the two SLI groups. MMN to tone stimuli did not differ between the groups. Two negativities, but not separately reported. Discrimination scores did not correlate with MMN amplitudes, although the SLI group had more errors in the discrimination task.

Autism Kemner et al., 1995

20 autistics children (mean age 9;8 years), 20 controls (mean age 10;6 years); 20 children with ADDH (mean age 9;9 years), 20 dyslectics (mean age 10 years)

Standardized reading tests (BRUS-1 or AVI cards) for dyslectic children.

Syllable: standard /oy/, deviant /ay/; 300 ms; novel /bbrrzzz/; 360 ms; ISI 4-6 seconds

No differences in MMN latency or amplitude between the groups.

Seri et al., 1999 7 autistics with TsC, 7 with TsC (mean age 8;6 years)

- Frequency: standard 1000 Hz, deviant 1500 Hz; 60 ms; ISI 800-1000 ms.

MMN latency increased and amplitude smaller in the autistic group.

Gomot et al., 2002

15 autistics, 15 matched controls (5-9 years)

- Frequency: standard 1000 Hz, deviant 1100 Hz; 50 ms; ISI 700 ms.

MMN latency was shorter in autistic children at Fz. Additional current sources were evident at midline in children with autism.

38





Čeponienė et al., 2003b

9 autistics (6;3-12;4 years), 10 controls (6;6-12.4 years)

Reynell Frequency: standard 458 Hz; vowel; standard /ö/, harmonics; made to resemble the standard /ö/. Deviants were created by raising 10 % the frequencies of the standards; SOA 700 ms.

No differences in MMN latency or amplitude between the groups.

Dyslexia Schulte-Körne et al., 1998

19 children with spelling disability (mean age 12;5 years), 15 controls (mean age 12; 6 years)

- Frequency; standard 1000 Hz, deviant 1050 Hz, 90 ms; ISI 590 ms. Syllables: standard /da/, deviant /ba/.

MMN attenuated for speech stimuli, but only in the time window typical of lMMN. No differences in the frequency change between the groups.

Kujala et al., 2001b

48 children with dyslexia, 22 of the 48 children participated in ERP measurements and training

Finnish battery of diagnostic reading tests, auditory discrimination measurements

Frequency; first tone 500 Hz, second tone 750 Hz in standards, separated with 10-ms silent gap, in deviants pairs in reversed order, 40 ms each tone; SOA 600 between the pairs

ERPs of two groups of dyslectic children similar before training. MMN increased in dyslectic children with auditory training, but not in those who did not participate. A significant correlation between MMN amplitude change and change in reading performance

Leppänen et al., 2002

37 children at risk for dyslexia, 39 controls (mean age 6 months)

- Pseudoword: ata 300 ms, atta 460 ms; both used as standard and deviant; ISI 610 ms.

MMN-like response smaller in the risk group at C3 when /atta/ as deviant. No differences over the right hemisphere. Larger positivity preceded the MMN-like negativity in the children at risk. When /ata/ was deviant, there were two negativities in the time window of MMN and lMMN in the control group but not in the risk group.

39





Oral clefting Cheour et al., 1997

11 children with CATCH syndrome (6-10 years), 11 age-matched controls

ITPA, NEPSU, phoneme discrimination, phonological awareness, pseudoword reading and learning

Frequency; standard 1000 Hz, deviant 1100 Hz, 100 ms; SOA 450, 800, 1500 ms.

MMN smaller in CATCH children with 800 ms and 1500 ms SOA. Significant MMN was only obtained with 450 ms SOA in children with CATCH syndrome.

Cheour et al., 1998b

11 children with CATCH syndrome (6-10 years), 11 age-matched children with cleft palate, 11 age-matched controls

As above for children with CATCH syndrome and controls.

Frequency; standard 1000 Hz, deviant 1100 Hz, 100 ms; SOA 450, 800, 1500 ms.

MMN smaller in children with CATCH with 800 ms and 1500 ms SOA. MMN amplitude smaller only with 1500 ms SOA in children with a cleft. A significant difference between the control, CATCH and cleft groups with 1500 ms SOA.

Cheour et al., 1999

9 neonates with cleft palate, 8 healthy controls

- Frequency; standard 1000 Hz, deviant 1100 Hz, 100 ms; SOA 800 ms.

Prominent MMN in all controls, but only in 3 infants with cleft palate. Positivity at MMN latency in infants with cleft palate. In controls, negativity at MMN latency.

Čeponienė et al., 1999a

78 children with oral clefts (7;4-9;8 years), 32 age-matched controls (7;3-9;9 years)

- Frequency; standard 1000 Hz, deviant 1100 Hz, 100 ms; ISI 350, 700, 1400 ms.

MMN smaller in children with oral clefts and diminished with longer ISIs in them. Children with only CL comparable to controls.

Čeponienė et al., 2000

32 neonates with oral clefts, 12 controls; 15 of the 32 children with oral clefts, 8 of the 12 controls re-measured at the age of 6 months.

- Frequency; standard 1000 Hz, deviant 1100 Hz, 100 ms; ISI 700 ms

Two negativities in neonates in time windows typical of MMN and lMMN. Amplitudes at both time windows were smaller in infants with CPO. No difference between controls and infants with CLP. At 6 months, two negativities were found again. No differences between the groups at the first time window. The amplitude at the second time window was smaller in the CPO group.

40





Learning disabled children

Kraus et al., 1996 91 children with LD, 90 controls (6-15 years) measured behaviourally. 42 children included in ERP measurements; 21 “good”, 21 “poor” perceivers based on behavioural discrimination tests

Listening comprehension, visual speed of processing, sound blending, auditory processing, reading, spelling, auditory memory for words (names of the tests not given), auditory discrimination test with stimuli used in ERP measurements.

Syllables: standard /da/, deviant /ga/; standard /ba/, deviant /wa/, 90 ms; ISI 1 second

MMN robust in “good” perceivers and absent in “poor” perceivers. MMN area and duration smaller in “poor” perceivers. A correlation between /da/-/ga/ discrimination scores and MMN area and duration. In contrast, no difference in MMN duration and area between subgroups of “poor” and “good” perceivers.

ADDH=attention deficit disorder with hyperactivity; BMA=Borell-Maisonny-Alahuhta test, Alahuhta; Boehm=Boehm test of basic concepts; CL=cleft lip; CLP=cleft lip and palate; CPO=cleft palate only; HSET=Heidelberger Sprachentwicklungstest; ISI=interstimulus interval; ITPA=Illinois Test of psycholinguistic abilities; LD=learning disabled; lMMN=late mismatch negativity; MR=mentally retarded; NEPSY= The Neuropscyhological Investigation of Children; Reynell=Reynell Developmental Language Scales; Seashore=The Seashore test of musical talents; SLI=Specific language impairment; SOA=stimulus onset asynchrony; TsC=Tuberous sclerosis; WJ-R=Woodcock-Johnson-Revised; WRAT=Wide Range Achievement Test-3.

The results concerning MMN in children with SLI are quite conclusive. MMN is attenuated in response to pure tones (Holopainen et al., 1996, 1997; Korpilahti, 1995; Korpilahti & Lang, 1994). Only Uwer et al. (2002) reported no significant difference in the MMN amplitude in response to pure tones. They found the only difference between their SLI and control groups to be the attenuated lMMN in the time window typical of the lMMN in response to speech syllables. However, Uwer et al. (2002) used higher frequencies, the difference between standard and deviant stimuli was larger, and the stimulus onset asynchrony (SOA = time from the onset of the stimulus to the onset of the next stimulus) was longer than in other studies. It is possible that the difference in pure tone was also easy to discriminate for children who may have difficulties in auditory discrimination. Furthermore, when ISI or SOA grows, the ERP waveform changes in children and new ERP components can be seen (Bruneau & Gomot, 1998; Čeponienė et al., 1998, 2002b; Korpilahti, 1996; Paetau et al., 1995). The obligatory ERP components also contribute to the difference waveform from which the MMN amplitude is measured, and thus the difference between the earlier findings and those of Uwer et al. (2002) might be mainly due to the long SOA.

The MMN results in children with autism are contradictory. While Seri et al. (1999) reported an increased MMN latency and a smaller amplitude in children with tuberous

41

sclerosis (TsC) and autism than in children with TsC but with no autism, Gomot et al. (2002) reported shortened MMN latencies in children with autism compared with healthy controls. Furthermore, Čeponienė et al. (2003b) and Kemner et al. (1995) did not find any differences in MMN latency or amplitude in children with autism and their controls. The reasons for the different results remain to be studied.

Auditory ERPs already differ in infancy in children at risk for language deficits (Kurzberg et al., 1988; Molfese & Molfese, 1985, 1997) and dyslexia (Guttorm et al., 2001; Leppänen et al., 1999; Molfese, 2000; Pihko et al., 1999) compared with healthy controls. These difficulties are seen as an attenuated MMN, mainly in the left hemisphere, in infancy (Leppänen et al., 2002) and as an attenuated MMN in response to speech stimuli in the latency range of lMMN at school age (Schulte-Körne et al., 1998).

The CATCH 22 syndrome (acronym for Cardiac defects, Abnormal facies, Thymic hypoplasia, Cleft palate, and Hypocalcemia) has been studied recently using ERPs, since children with CATCH 22 syndrome often exhibit difficulties in language development and learning (Kok & Solman, 1995). The results concerning children with CATCH 22 syndrome are conclusive, demonstrating a smaller MMN amplitude in children with CATCH syndrome compared with that of controls and more specifically with long SOAs (Cheour et al., 1997, 1998b).

An attenuated MMN amplitude has also been found in children with oral clefts (Čeponienė et al., 1999a, 2000; Cheour et al., 1999), although not in all subgroups of oral cleft patients (Čeponienė et al., 1999a, 2000). Similarly, school-aged children with learning disabilities exhibited MMN attenuation compared with healthy controls, together with behaviourally measured difficulties in auditory processing (Bradlow et al., 1999; Kraus et al., 1996).

Two groups of children with developmental brain disorders, i.e, children born preterm with VLBW and children with AS, exhibit different types of language impairment. There are no earlier studies reporting on MMN in VLBW preterm children together with language measures. Furthermore, no earlier study has followed up both language development and MMN in children born preterm. No earlier studies exist where auditory ERPs have been recorded in children with definite diagnosis of AS.

3 Aims of the study

The main cause of speech and language impairments may lie in auditory perceptual deficits. Therefore, the purpose of the present follow-up study was to investigate auditory processing by using ERPs, especially MMN, and language development to determine whether a correlation exists between auditory ERPs and language development. Two groups of children with developmental disorders were selected as subjects: children born preterm with VLBW and children with AS. The language development of VLBW preterm children was studied at the ages of 2, 4 and 6 years. ERP recordings were done at the ages of 4 and 6 years. Since very little is known about language development and auditory processing in children with AS, a pilot study was carried out to examine auditory processing by using auditory ERPs. Instead, language development in children with AS was not assessed for this study. The specific aims were as follows:

1. To investigate auditory processing in VLBW preterm children and matched controls at the ages of 4 (mean 4;4) (Study II) and 6 (mean 5;7) years (Study III) for possible differences in ERPs.

2. To investigate the developmental course of language learning in children born preterm with VLBW and in matched controls (Study I).

3. To determine whether a relationship exists between ERPs, more specifically MMN, and language development as indexed by object naming ability (studies II and III).

4. To investigate auditory ERPs in children with AS and matched controls for possible differences between children with AS and their controls (Study IV).

Since the main topic of the thesis was auditory processing, the results concerning auditory processing in VLBW children (Studies II and II) precede the results of language development in VLBW preterm children (Study I).

4 Materials and methods

4.1 Subjects

The very low birth weight (VLBW) preterm population was drawn from a birth cohort of all preterm infants born at Oulu University Hospital between 1st November, 1993 and 31st October, 1995 (fig. 2). The VLBW preterm infants with birth weight less than 1500 g and gestational age less than 34 weeks at birth were included in the primary study as described by Valkama et al. (2000b). A total of 29 VLBW preterm infants under the age of 30 months in March 1996 were eligible for inclusion in this series, and informed consent was obtained from 17 parents (including those of two pairs of twins) (fig. 2). Two children were excluded, one because of a profound hearing loss and one because of missing data on language development. The final population thus consisted of 17 VLBW preterm children (mean birth weight 1049 g, range 660-1445 g; mean gestational age 28 weeks, range 24-33 weeks; 8 girls) (table 3).

A neonatologist performed neurological follow-up examinations on all infants every 3 months up to 18 months of corrected age and at the chronological age of 6 years using the Griffiths Developmental Scales (Griffiths, 1954). At the chronological age of 18 months, the total score (GQ) was <90 for 4 children with CP. The subquotient of hearing and speech (SQ) was slightly abnormal for two preterm children (normal > 90) (table 3). At the age of 6 years, GQ was <90 for 4 children, 2 of these 4 children having CP, whereas SQ was slightly abnormal in 9 preterm children.

Infants with delayed motor development and abnormal muscle tone and motor function as an impairment or disability at a corrected age of 18 months were defined as having CP in the sense of Hagberg et al. (1975). The neurological outcome of all the VLBW children was confirmed by the paediatric neurologist at the chronological age of 2 years. The neurological outcome did not differ from the previous measurements made by the neonatologist. Five children were defined as having CP (table 3), and their rehabilitation was arranged at Oulu University Hospital.

Brain magnetic resonance imaging (MRI) at term age, as described by Valkama et al. (2000b), was performed on all the preterm children using a 1.0 T scanner (Magneton, Siemens). Fifteen of the 17 children had changes in their neonatal brain MRI. The

44

changes mostly consisted of dilated ventricles and sulci with an increase in cerebrospinal fluid volume (table 3).

Table 3. Clinical characteristics of the preterm group at birth, results of neonatal brain MRI and performance on the Griffiths Developmental Scales (GQ) and its sub-quotient of hearing and speech (SQ) at the ages of 18 months and 6 years and neuromotor outcome at the age of 18 months.

Clinical characteristics

MRI abnormality at term age Griffith Developmental Scales

Case and sex

GW at birth

BW (g) Parenchymal lesion

Cerebrospinal space dilatation

SQ/ GQ at 18 months

SQ/GQ at 6 years

Neuromotor outcome

1 M 33 1210 no no 128/125 114/128 normal 2 M 24 760 yes V + S 95/83 101/94 mild spastic

diplegia 3 M 29 1230 yes V + S 98/109 82/104 normal 4 M 29 920 no V + S 95/108 82/102 normal 5 M** 25 690 yes V + S 87/67 - severe spastic

diplegia 6 F 27 730 no V + S 125/116 78/81 normal 7 M 31 1300 yes V + S 123/127 83/99 normal 8 M 29 1430 no V + S 128/123 81/105 normal 9 F 28 1170 yes S 117/108 115/121 mild spastic

diplegia 10 F* 31 1250 no no - - normal 11 F 30 1145 no V 87/107 78/87 normal 12 F 26 660 yes V 121/115 88/103 normal 13 F 28 760 yes no 107/97 90/75 severe spastic

diplegia 14 F* 26 790 yes V + S 111/114 - normal 15 F 30 1445 no S 134/118 99/114 normal 16 M 31 1375 no V 118/111 84101 normal 17 M 26 960 yes V + S 102/84 85/78 mild spastic

diplegia * dropped out at the age of 4 years; ** dropped out at the age of 6 years; M = male, F = female; GW = gestational weeks at birth, BW = birth weight (g), V = ventricles, S = sulci

All children had normal hearing as tested by otoacoustic emissions (OAE) and brainstem auditory evoked potentials (BAEP) during the neonatal period (Valkama, 2001). Their hearing ability was followed up at Oulu University Hospital. One child had a hearing aid in one ear at the age of 4 years, and another child used hearing aids at the age of 6 years. These 2 children were excluded from the ERP measurements. However, the language development of these children was assessed, because, with their hearing aids, these children were capable of normal speech communication.

Control children were recruited from kindergartens (mean birth weight 3617 g, range 2840-4700 g; mean gestational age 40 weeks, range 38-42 weeks) and matched to the index cases by age, gender and the mother’s educational level. The ages of the control

45

children differed by no more than 2 months from the corresponding ages of their pairs at the time of the language measurements and EEG recordings. All the control children were healthy without any congenital neurological deficits. Their hearing and development were assessed and followed up at family clinics in health care centres.

At the age of 4 years, two VLBW preterm subjects dropped out (fig. 2). One VLBW preterm girl was excluded because of cancer treatment, and another on her mother’s request. In addition, tree more VLBW preterm children dropped out of the ERP recordings: one child who used hearing aids as mentioned above, and one VLBW preterm child because of refusal. The third child had such noisy EEG data that his ERPs could not be analyzed. Two control children were excluded because of abnormal neurological development, and 3 new control children were recruited to replace them.

At the age of 6 years, one VLBW preterm child declined from the study because of recent surgical operations (fig. 2). ERP measurements were not carried out on 2 further VLBW preterm children because of hearing impairment, as mentioned above. In addition, 4 controls dropped out on their mother’s request. Twelve matched pairs were included into the statistical analysis of language development and into ERP recordings.

Children with Asperger syndrome (AS) were recruited from a large research going on in the Department of Child Psychiatry, Oulu University Hospital. Twelve children (age range 7-12 years, mean 9;1 years) who met 5 or 6 of the six Gillberg’s criteria (Gillberg & Gillberg, 1989) for AS and all the criteria of the ICD-10 (ICD-10, 1999) were examined. All the children had normal hearing as indexed by audiometry. The verbal IQ of all the subjects was > 80. Twelve healthy controls, matched by age and gender, were selected from primary schools by a school psychologist. Two children with AS and one control child were excluded from the ERP analyses because of noisy EEG data.

Written consents were obtained from the parents of all children. The studies were approved by the Ethical Committee of Oulu University Hospital.

46

Fig. 2. Population of the preterm study with drop-outs.

29 consecutive VLBW children eligible for this study in February 1996. Informed consent from 17 parents in March 1996 (2 twins).

Drop-out: 1 child because of profound hearing loss, 1 child because of missing language data

Drop-out: 1 VLBW preterm child because of cancer treatment, 1 for the mother’s request, 2 developmentally abnormal controls

15 VLBW preterm children studied at the age of 4 years in May-June 1998. Language measurements: ERP recordings: 15 VLBW preterm children 12 VLBW preterm 15 matched controls children, 12 matched controls

Drop-out: 1 preterm child because of recent surgical operations, 4 controls on mother’s request

Birth cohort: 51 consecutive VLBW preterm infants November 1993-October 1995

17 VLBW preterm children studied at the age of 2 years in March–July 1996 and 17 matched controls

14 VLBW preterm children studied at the age of 6 years in March-May 2000.Language measurements: ERP recordings: 12 VLBW preterm children 12 VLBW preterm 12 matched controls children, 12 matched controls

47

4.2 Event-related potential measurements

4.2.1 Stimuli

The stimuli used in the present study were semi-synthetic Finnish syllables and pure tones. The syllables were used to elicit auditory ERPs in children born preterm with VLBW and with AS (Studies II, III, and IV) (table 4). The tone stimuli were only used in Study IV for investigating auditory ERPs in children with AS (table 4). Study I investigated language development, and it is not included in table 4.

Table 4. Number and age of subjects, stimuli, and SOAs in Studies II-IV.

Study II Study III Study IV Subjects 12 children born preterm

with VLBW, 12 matched controls

12 children born preterm with VLBW, 12 matched controls

12 children with AS, 12 matched controls

Age of the subjects 4 years (SD 5 months) 6 years (SD 5 months) 8-12 years Stimuli /taa/ standard; /kaa/ and /ta/

deviants /taa/ standard; /kaa/ and /ta/ deviants

/taa/ standard; /kaa/ and /ta/ deviants sine tone 1000 Hz standard; 1100 Hz deviant

SOA ms 650 ms 650 ms 650 ms for syllables, 600 for tones

AS= Asperger syndrome; SOA=stimulus onset asynchrony; VLBW very low birth weight

In the syllable condition, the standard syllable was /taa/ (190 ms in duration, probability 80 %), and the 2 deviant stimuli were /ta/ (duration change; 90 ms in duration, probability 10 %) and /kaa/ (consonant change; 190 ms in duration, probability 10 %). The stimuli were synthesized using a computer-based stimulus-generation method, Semisynthetic Speech Generation (SSG) (Alku et al., 1999). The consonant change (/t/, /k/) is typical of the Finnish language, unlike the consonants /d/, /g/ used in some previous studies (Kraus et al., 1996). We also used vowel duration decrement (/taa/ vs /ta/), which is semantically meaningful and very common in Finnish words. Syllables were used as stimuli, since in the speech stream, syllables are more discrete elements of speech and may be more precisely separated than single phonemes.

The syllables were binaurally presented through headphones in separate sequences of 400 stimuli. At least 4 sequences were presented to every subject. The syllables (75 dB HL) were presented with a SOA of 650 ms.

The tone stimuli were simple sinusoidals, the standard tone being 1000 Hz in frequency (100 ms in duration, probability 90 %) and the deviant tone being 1100 Hz in frequency (100 ms in duration, probability 10 %). The tones (65 dB HL) were presented binaurally through headphones with 600 ms SOA. In both tests, the syllable and tone conditions, standards and deviants, were presented in a random order, and in the string of stimuli, at least 3 standards always separated the consecutive deviants. In each condition, the total amount of data included at least 100 deviants collected from every subject.

48

4.2.2 Recordings and data analysis

Auditory ERPs were recorded at the sites F4, C4, P4 (right hemisphere, fig. 3), Fz, Cz, Pz (midline) and F3, C3 and P3 (left hemisphere), according to the International 10-20 system, using the NeuroScan Synamps amplifier and the NeursoScan 4.0 software (on-line bandpass 0.05-70 Hz, sampling rate 500 Hz). Electro-oculogram (EOG) was recorded with 2 electrodes: one attached below the left eye and the other above the right eye. During data acquisition, the common reference was at FCz, and after averaging, the data were re-referenced to the ipsilateral mastoids. Midline electrodes were re-referenced to the left mastoid. Epochs contaminated by artifacts exceeding 200 µ V (Study II) or 150 µ V (Studies III and IV) at any electrode were omitted from the averaging.

Fig. 3. Electrode placement used in the study.

The ERP-data was analyzed by using the EEG-View program. The ERPs were digitally filtered off-line with a 0.5-15 Hz bandpass filter, using a finite-duration impulse response

49

(FIR) filter of the order of 920, and averaged separately for all stimulus types. The time windows for measuring the different ERP peaks were determined on the basis of the grand-average latencies and the across-subjects latency distributions. In standard-stimulus ERP (obligatory responses), P1 was defined as the most prominent positive peak within 50-150 ms. The most negative peak within 180-300 was defined as N2, and that within 350-500 ms (Studies II and III) or 320-500 ms (Study IV) as N4. The peak latencies and amplitudes were used in the statistical analyses.

MMN and lMMN were obtained by subtracting the standard-stimulus ERP from that to deviant stimuli. In the resulting difference waves, the most negative peak within 200-340 ms was defined as MMN in the preterm studies (Studies II and III) and that within 340-550 as lMMN in the consonant change condition. In Study IV, the time window for MMN was 200-380 ms in the syllable condition and 150-320 in the tone condition. For MMN and lMMN, large ERP peak latency jitter is typical in children. Therefore, the mean amplitudes calculated over a 50 ms period centred at the peak latency of the MMN and lMMN of each participant were used for statistical analyses, as was done earlier by Čeponienė (2001). The jitter for lMMN was so pronounced in the children with AS that the data concerning lMMN were excluded from statistical analyses. Late MMN was also weak or absent in response to vowel duration change and tone change, and it was therefore left out of the statistical analyses.

During the EEG recording, each child, having been instructed to ignore the stimuli, watched a silent, self-chosen cartoon on video. A parent or research assistant was present with the child during the measurements to keep him/her relaxed and calm. The vigilance and movements of the subjects were monitored by observing their EEG and behaviour through a TV monitor. Short breaks were given at every 10-15 minute, or when needed, between the stimulus blocks. The children with AS were informed about the study by a personal letter including a full description of the measurements with photographs.

In the preterm study (Study III), the ERP waveform and the difference wave were visually checked by the author and one additional experienced researcher from the 4-year-old ERP data for the presence of MMN, and consensus was negotiated. The second tester was completely unaware of the clinical status of the children. MMN was defined subject by subject from all of the 9 electrodes used in the statistics. For the visual analysis, a three-category rating scale for MMN (present, absent, questionable) was used. The researchers recorded, electrode by electrode, whether MMN was present, missing or questionable at each electrode site. MMN was assessed to be present when a well-defined negative peak was found in the selected time window, whereas it was considered to be absent when the MMN peak did not reach negativity (did not cross the baseline) and/or could not be defined in the time window selected for MMN. Further, MMN was questionable when it was under 1 µV in amplitude. The results of the visual rating were compared with naming ability at the age of 6 years to determine whether there was a connection between the MMN difference waveform at the age of 4 years and naming ability at the age of 6 years (Study III). In the statistical analyses the most negative peak was taken to represent MMN, as desribed earlier.

50

4.3 Evaluation of language skills (Study I)

The evaluation of the language skills of the VLBW preterm children and their matched controls was done at the mean chronological age of 2 (mean 2;4 years; SD=11 months), 4 (mean 4;4 years; SD=10 months) and 6 (mean 5;7years; SD=5 months) years (table 5).

Table 5. Behavioural measures of language development.

Age 2 years Age 4 years Age 6 years Constituent part of language measured Measure Measure Measure Language comprehension Reynell Developmental

Language Scales Reynell Developmental

Language Scales

Language production Vocabulary MCDI Boston Naming Test Boston Naming Test Morphological development

MCDI, Basic sentence types, MLU

Morphological Test Morphological Test

Phonology Pseudoword imitation Auditory discrimination - Auditory Discrimination

Test Auditory Discrimination

Test MCDI=MacArthur Communicative Developmental Inventory, MLU=mean length of utterance

Language comprehension was assessed by using the Reynell Developmental Language Scales (Reynell &Huntley, 1985, table 5). The Reynell test was excluded from the 6-year-old measurements since the norms of the test did not reach up to that age range in the version used in this thesis.

Lexical development was assessed by MCDI at the age of 2 years (Fenson et al., 1994; Lyytinen, 1999, table 5) and by the Boston naming test (Kaplan et al., 1983; Laine et al., 1988) at the ages of 4 and 6 years. The child was assessed as having difficulty in object naming when the raw score of the Boston Naming Test result was at least 1 standard deviation below the mean of the control group.

Morphological development at the mean age of 2 years was assessed by MCDI. Furthermore, spontaneous speech samples were collected from a 15-minute video of free play interaction between the child and the caregiver (table 5). The recordings were transcribed, and unclear utterances were assessed by 2 trained investigators to reach consensus. The Mean Length of Utterance (MLU, Brown, 1973), a widely used index of early grammatical development, was measured in terms of the numbers of morphemes in the last 100 utterances that the child had produced during free play, and all the utterances were included when needed. Furthermore, the basic sentences used by the child were classified in accordance with Hakulinen and Karlsson (1979) and Karvonen (1985). The basic sentences are idealistic forms of kernel sentences (Hakulinen & Karlsson, 1979), and they are based on the transformation theory by Chomsky (1957). In the Finnish language, there are at least 6 types of kernel sentences (Hakulinen & Karlsson, 1979), whereas in the English language, there are at least 11 types (Lyons, 1977). The sentences produced by children differ from those used by adults. Therefore, the basic sentences produced by children must be classified differently (Karvonen, 1985). In this study, we rely on the findings of Karvonen (1985) since there are no other data concerning the basic

51

sentences produced by Finnish-speaking children. Later, at the ages of 4 and 6 years, morphological development was assessed using the Morphological Test (Lyytinen, 1988).

Phonological development was assessed by pseudoword imitation (table 5). Pseudowords were taken from the Morphological Test (Lyytinen, 1988). The score of phonological mistakes was based on the number of articulatory errors in pseudoword imitation. The method was not sensitive, and it was therefore omitted from the 6-year-old measurements.

Auditory discrimination ability was tested using the Auditory Discrimination Test (Korpilahti, 1991) at the ages of 4 and 6 years (table 5). This test determines how well the subject discriminates between Finnish vowels and consonants. The test includes a consonant discrimination battery of 50 minimal pairs of syllables and a vowel discrimination battery of 32 minimal pairs. The number of correct judgements in each was included in the analyses. If the child was unwilling to perform the test, this was treated as a case of missing data, and if the child ceased to co-operate halfway through the test, the quotient was taken to be that reached by the child up till that point. Auditory discrimination was not assessed at the age of 2 years, since there are no available standardized tests to be used at that age for measuring auditory discrimination.

Language skills were not evaluated in children with AS for the present study. The main focus of the present study was on the piloting of ERP measurements in that population.

4.4 Statistical analyses

Mean values were used when tested differences in the ERP data. The inter-group differences in ERP data were tested using repeated-measures ANOVAs (Studies II and III) and three-way mixed ANOVAs (Study IV) with Group as a between-subject factor and Hemisphere [Right (F4, C4, P4) x Midline (Fz, Cz, Pz) x Left (F3, C3, P3)] x Anterior-Posterior [Frontal (F3, Fz, F4) x Central (C3, Cz, C4) x Parietal (P3, Pz, P4)] as a within-subject factors. Assumption about the variance-covariance matrices of the dependent variables were checked. The sphericity assumption was slightly fulfilled. Therefore, the least conservative Huynh-Feldt adjustment method was used when evaluating the statistical significance of the within-subjects effects. (SPSS Advanced Models 10.0, 1999). One-way between-group ANOVA was used to measure differences at the single-electrode level between the groups. The statistical difference of the ERP peak values was tested using two-tailed t-tests, comparing the ERP amplitudes with the 0 µV baseline (Study IV).

Paired samples t-test was used to compare the language test scores in the follow-up of language development (Study I). The independent-samples t-test was used to compare the Boston Naming Test results between the preterm and the control group (Studies II and III).

The correlations between the language test results (Study I) and between the Boston Naming Test results and the MMN parameters were tested using Spearman’s Correlation Coefficients (Study III).

5 Results

5.1 Auditory ERPs in children born preterm with VLBW (Studies II and III)

5.1.1 Obligatory responses

The standard syllable /taa/ elicited a clear P1-N2-N4 waveform both at the age of 4 and at the age of 6 years. The obligatory ERPs, elicited by standard stimuli in VLBW children and in their controls (Studies II and III), are presented in the figures 4 and 5.

There were no significant differences between the preterm group and the control group in P1 latency and amplitude over the 9 electrodes at the ages of 4 and 6 years (table 6, figs. 4, 5). For P1 latency, there were no Right-Left x Anterior-Posterior x Group interactions at the ages of 4 and 6 years. Nor was there any Right-Left x Anterior-Posterior x Group interactions for P1 amplitude at the age of 4 years. However, at the age of 6 years, a significant Right-Left x Anterior-Posterior x Group interaction was found for P1 amplitude (F(3.0, 66.0)=8.86, p=0.000, Huynh-Feldt correction), being due to the smaller P1 amplitude at the frontal and central electrodes in the right hemisphere in preterm children (fig 5). The difference between the preterm and control groups was most pronounced at the C4 electrode (F(1,22)=7.61, p=0.011, one-way between-group ANOVA, fig. 5). P1 amplitude at C4 was 8.9 µV in the control group, while in the preterm group it was 5.2 µV (fig. 5).

53

Table 6. The peak latencies and amplitudes of the obligatory responses P1, N2 and N4. The values represent the corresponding means over 9 electrodes (F4, C4, P4, Fz, Cz, Pz, F3, C3, P3) in VLBW preterm children and their controls at the ages of 4 and 6 years. The p-values represent the results of the repeated measures ANOVAs over 9 electrodes.

Latency ms Amplitude µV Response Age Preterms Controls p -value Preterms Controls p -value

4 years 120 ms 120 ms Ns. 7.6 µV 8.9 µV Ns. P1 6 years 112 ms 113 ms Ns. 7.4 µV 8.3 µV Ns. 4 years 234 ms 246 ms Ns. -4.1 µV -3.3 µV Ns. N2 6 years 232 ms 245 ms 0.035 -4.6 µV -6.5 µV Ns. 4 years 378 ms 377 ms Ns. -4.9 Μv -5.1 µV Ns. N4 6 years 357 ms 360 ms. Ns. -6.9 µV -8.4 µV Ns.

54

Fig. 4. Obligatory responses to standards in VLBW preterm children and their controls at the age of 4 years in the left- (on the left) and right-hemisphere (on the right) electrodes.

At the age of 4 years, N2 latency and amplitude did not differ significantly over the 9 electrodes between the VLBW preterm group and their control group (table 6, fig. 4). No Right-Left x Anterior-Posterior x Group interactions were found for N2 latency or amplitude. In contrast, at the age of 6 years, N2 latency was significantly shorter in the preterm group compared to the control group (F(1,22)=5.02, p=0.035, fig. 5). There were no Right-Left x Anterior-Posterior x Group interactions for N2 latency between the two groups at that age.

55

Fig. 5. Obligatory responses to standard syllables in VLBW children and in their controls at the age of 6 years in the left- (on the left) and right-hemisphere (on the right) electrodes.

At the age of 6 years, N2 amplitude did not differ significantly between the preterm and control groups over the 9 electrodes (table 6). However, at the age of 6 years, the Right-Left x Anterior-Posterior x Group interaction for N2 amplitude was significant (F(2.6, 57.9)=5.29, p=0.004, Huynh-Feldt correction), being due to the smaller N2 amplitude in the frontal and midline electrodes in the right hemisphere in the preterm group compared to the control group (F(1,22)=3.49, p=0.075, fig. 5). At the level of single electrodes, the most significant difference was encountered at the C4 electrode (F(1,22)=8.71, p=0.007, one-way between-group ANOVA, fig 5). N2 amplitude in the preterm group was -3.9 µV at the C4 electrode, while in the control group it was -7.7 µV.

56

N4 latency and amplitude over the 9 electrodes did not differ significantly between the two groups at the ages of 4 or 6 years (table 6, figs. 4, 5). There were no Right-Left x Anterior-Posterior x Group interactions for N4 latency. At the age of 4 years, no significant Right-Left x Anterior-Posterior x Group interactions were found for N4 amplitude. In contrast, at the age of 6 years, there was a significant Anterior-Posterior x Group interaction for N4 amplitude (F(2.0, 44.0)=7.36, p=0.002, Huynh-Feldt correction), being due to the smaller N4 amplitude at the frontal and central electrodes in the right hemisphere in the preterm group (F(1,22)=6.41, p=0.019, fig. 5). The difference between the preterm and control groups was most pronounced at the C4 electrode (F(1,22)=8.70, p=0.007, one-way between group ANOVA, fig. 5).

5.1.2 Mismatch negativity and late mismatch negativity (Studies II and III)

The MMN difference waveform showed two distinct negative deflections in the consonant change condition; an earlier peak in the time window of 200-340 ms (defined as MMN) and a later peak in the time window of 340-550 ms (termed as lMMN). In the vowel duration change condition, only one distinct negative deflection was reliably identified, occurring in the time window of 200-340 ms, and being defined as MMN.

At the age of 4 years, MMN latency and amplitude in response to consonant change did not significantly differ between the preterm and control groups over all of the 9 electrodes (table 7, fig. 6), which was at least partly due to the large standard deviation in both groups. At the age of 6 years, however, a significant difference was seen in both MMN latency and amplitude between the two groups. MMN latency was shorter (F(1,22)=4.47, p=0.046) in the preterm group than in the control group (fig. 7). There was a significant Right-Left x Group interaction (F(1.8, 38.8)=3.61, p=0.042, Huynh-Feldt correction), being due to the shorter latencies at the left-hemisphere and midline electrodes in the preterm group than in the control group. The MMN latency at the left-hemisphere was 267 ms in the preterm group and 287 ms in the control group. The latencies at the midline were 270 ms and 287 ms, respectively. MMN amplitude was smaller in the preterm group compared with the control group (F(1,22)=4.61, p=0.043). No Right-Left x Anterior-Posterior x Group interactions were found for MMN amplitude.

57

Table 7. MMN latencies and amplitudes over 9 electrodes in VLBW children and in their controls at the ages of 4 and 6 years in response to consonant and vowel duration change. Late MMN latency and amplitude in response to consonant change are reported here. The values represent the corresponding means and standard deviations over 9 electrodes (F4, C4, P4, Fz, Cz, Pz, F3, C3, P3) in VLBW preterm children and their controls at the ages of 4 and 6 years. The p-values represent the results of the repeated measures ANOVAs over 9 electrodes.

Latency ms Amplitude µV Response/ condition

Age in years Preterms

Mean (SD) Controls

Mean (SD)

p -value Preterms Mean (SD)

Controls Mean (SD)

p -value

4 297 ms (30) 292 ms (30) Ns. -2.2 µV (3.0) -2.7 µV (1.7) Ns. MMN Consonant change

6 269 ms (9.2) 284 ms (23) 0.046 -1.8 µV (2.6) -3.6 µV (1.3) 0.043

4 293 ms (18) 294 ms (19) Ns. -2.1 µV (2.0) -4.3 µV (2.6) 0.035 MMN Vowel duration change

6 276 ms (12) 273 ms (11) Ns. -3.4 µV (2.5) -4.4 µV (2.5) Ns.

4 431 ms (43)

414 ms (25) Ns. -2.8 µV (2.5) -5.0 µV (2.0) 0.035 lMMN consonant change 6 408 ms (33) 415 ms (38) Ns. -3.0 µV (2.2) -4.8 µV (2.2) 0.058

Late MMN latency in response to consonant change did not differ between the preterm and control groups at either age (table 7, figs. 6, 7). In contrast, at the age of 4 years, lMMN amplitude was significantly smaller in the preterm group than in the control group (F(1,21)=5.08, p=0.035, table 7, fig. 6). There was no Right-Left x Anterior-Posterior x Group interactions. Still, at the age of 6 years, there was a tendency for the lMMN to be smaller in amplitude in the preterm group than in the control group (F(1,22)=4.01, p=0.058, table 7, fig. 7). Similarly as at the age of 4 years, no Right-Left x Anterior-Posterior x Group interactions were found at the age of 6 years.

58

Fig. 6. MMN in response to consonant change in VLBW preterm children and in their controls at the age of 4 in the left- (on the left) and right-hemisphere (on the right) electrodes

59

Fig. 7. The MMN in response to consonant change in VLBW preterm children and in their controls at the age of 6 years in the left- (on the left) and right-hemisphere (on the right) electrodes

MMN latency in response to vowel duration did not differ between the two groups at the age of 4 years (table 7, fig. 8), whereas MMN amplitude was significantly attenuated in the preterm group (F(1,21)=5.09, p=0.035). There were no Right-Left x Anterior-Posterior x Group interactions. At the age of 6 years, no significant differences were found in MMN latency or amplitude in response to vowel duration change between the two groups (fig. 9)

60

Fig. 8. MMN in response to vowel duration change in VLBW preterm children and in their controls at the age of 4 years in the left- (on the left) and right-hemisphere (on the right) electrodes

61

Fig. 9. MMN in response to vowel duration change in VLBW preterm children and in their controls at the age of 6 years in the left- (on the left) and right-hemisphere (on the right) electrodes.

5.2 Language development in VLBW children (Study I)

At the age of 2 years, the VLBW children attained significantly lower scores in the language comprehension subtest of Reynell Developmental Language Scales than the controls (table 8). The maximum sentence length in MCDI was significantly shorter in the preterm group than in the control group (p=0.010). A delay in grammatical development was also evident in the form of a smaller number of basic (p=0.063) and predicative sentences (p=0.018). The number of existential sentences was also slightly lower (p=0.053). However, no significant differences between the preterm and control

62

groups were found in the size of the vocabulary, in the number of transitive, intransitive or marginal sentences or in the mean length of the utterance.

Table 8. Means, standard deviations (SD), and 95-% confidence intervals (CI) of the difference in language scores between VLBW preterm children and their controls at the age of 2 years. The p-values represent the results of the paired samples t-test.

Preterm children Control children N=17 N=17

Test

Mean SD Mean SD

Mean of the difference

95-% CI p -value

Reynell -0.2 1.1 0.8 0.7 -1.1 -1.7, -0.4 0.003 MCDI Vocabulary 336 138 369 149 -28.4 .104.1, 47.2 0.436 Plural and word endings

5.3 3.2 7.1 2.8 -1.9 -4.0, 0.3 0.081

Verb forms 4.8 1.5 5.7 1.5 -0.8 -1.8, 0.1 0.091 Maximal sentence length

6.1 2.1 9.3 4.3 -3.1 -5.4, -0.9 0.010

Free play situation MLU 2.6 1.4 3.1 1.1 -0.5 -1.3, 0.3 0.196 Types of basic sentences

Transitive 12.8 12.5 15.5 11.3 -2.6 -11.2, 5.9 0.522 Intransitive 7.2 6.4 10.4 7.4 -3.1 -7.8, 1.6 0.180 Predicative 5.1 3.7 10.4 7.6 -5.4 -9.7, -1.0 0.018 Existential 3.3 3.8 7.0 5.4 -3.7 -7.5, 0.0 0.053 Marginal 1.1 1.1 2.1 2.1 -1.1 -2.4, 0.3 0.111 Number of all basic sentences

29.5 19.2 45.4 23.1 -15.9 -32.7, 0.9 0.063

Reynell = Reynell Developmental Language Scales, comprehension subscore, MCDI=MacArthur Communicative Development Inventory, MLU=Mean Length of Utterance

At the age of 4 years, the VLBW preterm group again scored significantly lower than their matched controls in the Reynell Developmental Language Scales comprehension subtest (p=0.020, table 9) and the Auditory Discrimination Test (vowel discrimination p=0.018, consonant discrimination p=0.000). In addition, the VLBW preterm children exhibited difficulties in the Boston Naming Test relative to their controls (p=0.010). They also scored lower than their controls in the Morphological Test, but the difference was not significant. Nor was the difference in the number of phonological mistakes significant.

63

Table 9. Means, standard deviations (SD), and 95-% confidence intervals (CI) of the difference in language scores between VLBW preterm children and their controls at the age of 4 years. The p-values represent the results of the paired samples t-test.


Test

Mean SD Mean SD


95-% CI p -value

Reynell -0.2 1.2 0.7 0.6 -0.9 -1.7, -0.2 0.020 Auditory Discrimination Test

Vowel discrimination

15 15.4 27 11.1 -10.9 -19.6, -2.2 0.018

Consonant discrimination

16 17.1 39 17.3 -23.3 -33.8, -12.8 0.000

Boston Naming Test

17 6.9 23 6.1 -6.9 -10.3, -1.7 0.010

Morphological Test

22 24.3 35 18.7 -13.1 -29.0, 2.8 0.100

Phonological mistakes

7 7.3 5 3.5 1.9 -3.2, 7.0 0.429

Reynell = Reynell Developmental Language Scales, comprehension subscore

At the age of 6 years, the VLBW preterm children achieved significantly lower scores in the Auditory-Discrimination Test (consonant discrimination p=0.010, vowel discrimination p=0.065, table 10) and in the Boston Naming Test (p=0.014) than their controls. There was also a tendency in the Morphological Test for the preterm group to score lower than the control group (p=0.057).

Table 10. Means, standard deviations (SD), and 95-% confidence intervals (CI) of the difference in language scores between VLBW preterm children and their matched controls at the age of 6 years. The p-values represent the results of the paired samples t-test.


Test

Mean SD Mean SD


95-% CI p -value

Auditory Discrimination Test

Vowel discrimination 23 10 28 3.6 -11 -11.5, .429 0.065 Consonant discrimination

34 13.2 45 5.4 -11.0 -18.8, -3.30 0.010

Boston Naming Test 29 8.5 40 7.2 -10.4 -18.3, -2.5 0.014 Morphological Test 35 26.9 59 16.6 -23.8 -48.5, .90 0.057

Language test results at the age of 2 years significantly correlated with those at the age of 4 years (table 11). The main correlations were found between the vocabulary score at the age of 2 years and the Auditory-Discrimination Test results at the age of 4 years.

64

Furthermore, the vocabulary score at the age of 2 years correlated with the Boston Naming Test score at the age of 4 years. There was also a significant correlation between the Boston Naming Test scores at the ages of 4 and 6 years.

Table 11. Correlations between language test results at the ages of 2, 4 and 6 years as shown by Spearman´s Correlation coefficient.

2 years 4 years 6 years Test I II III IV V VI VII VIII IX X

I 0.689** 0.367 0.511** 0.529** 0.498** 0.084 0.277 0.329 0.096

II 0.241 0.541** 0.577** 0.270 0.245 0.411 0.472* 0.153

III 0.500** 0.417* 0.470* 0.510* 0.107 0.084 0.068

IV 0.899** 0.554** 0.442* 0.584** 0.522* 0.314

V 0.453* 0.364 0.522* 0.469* 0.275

VI 0.490* 0.175 0.158 0.551**

VII 0.355 0.343 0.594**

VIII 0.723** 0.293

IX 0.174

I=MCDI/Vocabulary at 2 years; II=MCDI/Maximum sentence length at 2 years; III=Morphological Test at the age of 4 years; IV=Consonant discrimination at the age of 4 years; V=Vowel discrimination at the age of 4 years; VI=Boston Naming Test at the age of 4 years; VII Morphological Test at the age of 6 years; VIII=Consonant discrimination at the age of 6 years; IX= Vowel discrimination at the age of 6 years; X=Boston Naming Test at the age of 6 Years; **Correlation is significant at the 0.01 level (2-tailed), * Correlation is significant at the 0.05 level (2-tailed)

5.3 Auditory ERPs and naming ability in children born preterm with VLBW (Studies II and III)

At the age of 4 years, 6 of the 12 VLBW preterm children, and at the age of 6 years, 8 of the 12 VLBW preterm children whose ERPs were recorded were identified as having naming difficulty. In the results, the preterm children with naming difficulty are compared both with their matched controls and with the preterm children without naming difficulty at the age of 4 years. At the age of 6 years, most preterm children exhibited naming difficulty. There was only a small subgroup of preterm children with normal

65

naming skills. Hence, no comparisons between the preterm children with and without naming difficulty were made.

The obligatory responses showed no significant differences between the VLBW preterm group with naming difficulty and their control group at the age of 4 or 6 years over 9 electrodes (table 12). At the age of 4 years, no significant Right-Left x Anterior-Posterior x Group interactions were found for P1, N2 and N4 latencies or amplitudes. Nor were there any significant Right-Left x Anterior-Posterior x Group interactions for P1, N2 and N4 latencies at the age of 6 years. However, at the age of 6 years, significant Right-Left x Anterior-Posterior x Group interactions were found for P1 (F(3.5, 49.0)=7.84, p=0.000, Huynh-Feldt correction), N2 (F(2.7, 37.5)=4.67, p=0.009, Huynh-Feldt correction) and N4 amplitudes (F(2.8, 39.4)=3.96, p=0.016, Huynh-Feldt correction). P1 tended to be smaller at the central electrodes in the preterm group with naming difficulties compared to their control group (F(1,14)=4.06, p=0.064, one-way between-group ANOVA).). The most pronounced difference was encountered at the C4 electrode (F(1,14)=6.61, p=0.022, one-way between-group ANOVA). N2 amplitude tended to be smaller at the frontal and central electrodes in the right hemisphere in the preterm group with naming difficulties than in their control group (F(1,14)=3.93, p=0.067, one-way between-group ANOVA). At a single-electrode level, the most significant difference between the two groups was encountered at the C4 electrode (F(1,14)=6.03, p=0.028, one-way between group ANOVA). Similarly, N4 amplitude was smaller at the frontal and central electrodes in the right hemisphere in the preterm group with naming difficulties compared to their controls (F(1,14)=5.74, p=0.031, one-way between-group ANOVA). The most pronounced difference was seen at C4 (F(1,14)=8.74, p=0.010, one-way between group ANOVA).

Table 12. Peak latencies and amplitudes of the obligatory responses P1, N2 and N4 in VLBW preterm children with naming difficulty and their controls at the ages of 4 and 6 years. The values represent the corresponding means over 9 electrodes (F4, C4, P4, Fz, Cz, Pz, F3, C3, P3) in VLBW preterm children and their controls at the ages of 4 and 6 years. The p-values represent the results of the repeated measures ANOVAs over 9 electrodes.

Latency ms Amplitude µV Response Age Preterms Controls p -value Preterms Controls p -value

4 years 120 ms 120 ms Ns. 7.3 µV 10.2 µV Ns. P1 6 years 112 ms 115 ms Ns. 7.3 µV 9.1 µV Ns. 4 years 234 ms 242 ms Ns. -3.6 µV -3.0 µV Ns. N2 6 years 238 ms 243 ms Ns. -4.9 µV -7.3 µV Ns. 4 years 370 ms 373 ms Ns. -3.9 µV -5.0 µV Ns. N4 6 years 357 ms 360 ms. Ns. -7.5 µV -9.3 µV Ns.

At the age of 4 years, there was no significant difference in MMN latency in response to consonant change between the VLBW preterm group with naming difficulty and their control group (table 13, fig. 10). MMN latency was also equal in the preterm group without naming difficulties. At 6 years of age, MMN latency was, however, significantly shorter in the preterm group with naming difficulty than in the control group

66

(F(1,14)=8.20, p=0.013). No Right-Left x Anterior-Posterior x Group interactions were found.

Fig. 10. MMN in response to consonant change in VLBW preterm children with naming difficulty and without naming difficulty and in controls at the age of 4 years (at the top). At the bottom, MMN in VLBW children with naming difficulty and in their controls at the age of 6 years.

At the age of 4 years, MMN amplitude in response to consonant change was significantly smaller in the preterm children with naming difficulty than in their matched controls (F(1,10)=7.79, p=0.021, table 13, fig. 10). In addition, there was a trend for the consonant change MMN to be smaller in amplitude in the preterm group with naming difficulty than in the preterm group without naming difficulty (F(1,10)=3.78, p=0.081). In contrast, no significant MMN differences were found between the preterm children without naming difficulty and their matched controls. Nor was there any significant Right-Left x Anterior-Posterior x Group differences in the consonant change condition. At the age of 6 years, the MMN amplitude in response to consonant change still tended to be smaller in the

67

preterm group with naming difficulty than in their matched controls (F(1,14)=4.51, p=0.052, table 13).

At the age of 4 years, lMMN amplitude in response to consonant change was significantly smaller in VLBW children with naming difficulty than in their matched controls over 9 electrodes (F(1,10)=33.2, p=0.000, table 13, fig. 10). There was no significant difference in lMMN latency between the VLBW preterm group with or without naming difficulties and their control group at the age of 4 years. At the age of 6 years, no significant difference was found in lMMN amplitude or latency in response to consonant change between the preterm group with naming difficulty and their control group. No Right-Left x Anterior-Posterior x Group interactions were obtained at either age.

Table 13. MMN latencies and amplitudes in VLBW children with naming difficulty and in their controls at the ages of 4 and 6 years. Late MMN latency and amplitude in response to consonant change are reported here. The values represent the corresponding means and standard deviations over 9 electrodes (F4, C4, P4, Fz, Cz, Pz, F3, C3, P3) in VLBW preterm children with naming difficulty and their controls at the ages of 4 and 6 years. The p-values represent the results of the repeated measures ANOVAs over 9 electrodes.

Latency ms Amplitude µV Response/ condition

Age in years Preterms

Mean (SD) Controls

Mean (SD) p -value Preterms

Mean (SD) Controls

Mean (SD) p -value

4 303 ms (25) 293 ms (34) Ns. -0.7 µV (1.3) -3.1 µV (1.4) 0.021 MMN Consonant change

6 272 ms (10) 289 ms (15) 0.013 -1.6 µV (2.4) -3.9 µV (1.4) 0.052

4 288 ms (11) 289 ms (13) Ns. -1.0 µV (1.7) -4.1 µV (2.4) 0.031 MMN Vowel duration change

6 273 ms (14) 273 ms (10) Ns. -3.5 µV (2.9) -5.3 µV (2.3) Ns.

4 414 ms (28) 428 ms (49) Ns. -1.1 µV (1.4) -5.6 µV (1.2) 0.000 lMMN consonant change

6 409 ms (39) 413 ms (34) Ns. -3.4 µV (2.7) -5.1 µV (2.5) Ns.

At the age of 4 years, no significant differences in MMN latency in response to vowel duration change were seen between the preterm group with naming difficulty, their controls or the preterm children without naming difficulty (table 13, fig. 11). Further, no significant Right-Left x Anterior-Posterior x Group interactions were found.

MMN amplitude in response to vowel duration change at the age of 4 years was significantly smaller in the preterm children with naming difficulty than in their matched controls (F(1,10)=6.56, p=0.031, table 13, fig. 11) or in the preterm group without naming difficulty (F(1,10)=5.38, p=0.043). No significant Right-Left x Anterior-Posterior x Group interactions were found. Further, at the age of 6 years, no significant differences were found in MMN parameters between the preterm group with naming difficulties and their matched controls (table 13, fig. 11). Nor were there any Right-Left x Anterior-Posterior x Group interactions.

68

Fig. 11. At the top, MMN elicited by vowel duration change in VLBW children with naming difficulty and without naming difficulty and in the controls of naming-deficient VLBW preterm children at the age of 4 years. At the bottom, MMN in response to vowel duration change in VLBW preterm children with naming difficulty and in their controls at the age of 6 years.

5.4 Visual inspection of MMN (Study III)

The ERP waveforms of 19 children who participated in ERP recordings and language measurements at the ages of both 4 and 6 years were inspected child by child by the author and another experienced researcher (table 14). It was found that MMN in response

69

to consonant change was absent or questionable at 6 or more electrodes in 5 preterm children (Subjects 4, 6, 7, 8, 9) at the age of 4 years. These children exhibited object naming difficulty both at the ages of 4 and 6 years. This was also the case with one control child (Subject 17). One VLBW preterm child with normal MMN at the age of 4 years had naming difficulty at the age of 6 years (Subject 19).

Table 14. Categorization of MMN for consonant change at the age of 4 years in 19 children who were measured at the ages of 4 and 6 years. The subjects 1-10 are prematurely born children and the subjects 11-19 belong to the controls group.

Naming score Electrodes

Age

Subject

4 6 F4 C4 P4 Fz Cz Pz F3 C3 P3

1 27 40 Absent P P Absent Absent Absent Absent P P 2 20 34 P P P P P P Absent P P 3 19 34 P P P P P P P P P 4* 17 20 Q P Q Absent Absent Absent Q Q Absent 5 24 40 P Q Q P P Q Absent Absent P 6* 17 27 Absent Absent Absent Absent Absent Absent P Absent Absent 7* 7 16 Absent Absent Absent Absent Absent Q Absent Absent Absent 8* 5 26 Q Absent Absent Absent Absent Absent Absent Absent Absent 9* 12 28 Q Q P Absent Absent Absent Absent Absent Absent 10* 27 32 P P P P P P P P P 11 28 45 P P Absent P P P P P P 12 33 39 P P P P P P P P P 13 25 41 P P P Q P P Absent P P 14 23 51 Q P P Absent Q Q P P Q 15 34 44 P P P P P P P P P 16* 23 31 Absent P P Q P P Absent Q Q 17* 11 29 P Q Absent Absent Absent Absent Absent Absent Absent 18 22 40 P P P Absent Absent P Absent Q P

19 27 42 Absent P P P P P Absent P Q

* a child having naming difficulty at the ages of 4 and 6 years; P = MMN present; Q = MMN questionable

5.5 Auditory ERPs in children with Asperger syndrome (Study IV)

The standard syllables and tones elicited a clear P1-N2 wave complex both in children with AS and in their controls (fig. 12). The P1 amplitude to standards was significantly smaller in the AS group than in the control group in both syllable (F(1,19)=4.93, p=0.040) and tone (F(1,19)=4.24, p=0.050) conditions (fig. 12). In contrast, no between-group differences were found in P1 latency. Also, N2 was smaller in amplitude in the AS group than in the control group in both stimulus conditions. However, the between-group

70

difference was statistically significant only in the tone condition at the right central electrode C4 (F(1,20)=2.89, p<0.010, one-way between group ANOVA). In addition, N2 latency was delayed in the AS group at the frontal and central electrodes for the tone stimuli compared with the controls (F(1,20)=4.32, p=0.050). No hemispheric differences were found for the P1 or N2 peaks between the groups. No significant differences were encountered for N4 over the 9 electrodes between the two groups.

Fig. 12. Obligatory ERPs elicited by syllables (top) and tones (bottom) in children with AS and in their controls.

MMN latency or amplitude did not differ significantly over 9 electrodes in response to consonant between children with AS and their controls. However, MMN latency was longer in the AS group than in the controls in response to consonant change in the right hemisphere (p<0.002, post hoc), whereas no significant differences between the two groups were seen over the left hemisphere (Right-Left x Group interaction F(1,19)=5.13, p=0.067, fig. 13). For MMN amplitude in response to consonant change, there was a

71

trend for Right-Left x Group interaction (F(1,19)=3.77, p=0.067). MMN amplitude was larger over the right hemisphere in the AS group, whereas in the control group it was larger over the left hemisphere (figs 13, 14). There were no significant differences in the MMN parameters between the two groups in response to vowel duration change.

Table 15. Mean latencies and mean amplitudes (SD) of MMN in children with AS and in their controls. The values represent the corresponding means and standard deviations over three electrodes on the right and left sides. The p-values represent the post hoc results of the significant Right-Left x Group interactions of ANOVA analysis, indicating the significance of the between-group differences.

Latencies (ms) Amplitudes (µV) Condition AS Mean

(SD) Controls

Mean (SD) p-value AS Mean

(SD) Controls

Mean (SD) p-value

Consonant change

Right array 324 (36) 297 (34) 0.002 -3.7 (2.1) -2.9 (2.5) Ns. Left array 309 (44) 307 (34) Ns. -2.9 (2.1) -3.6 (2.2) Ns. Tone change Right array 215 (42) 184 (12) 0.000 -5.4 (2.8) -4.7 (1.8) Ns. Left array 206 (37) 194 (15) 0.001 -5.3 (3.1) -3.9 (2.0) Ns.

In the tone condition, there was no significant difference between the two groups over 9 electrodes in the MMN latency. However, MMN latency was longer in the AS group than in the control group over the right hemisphere (Right-Left x Group interaction (F(1,19)=6.23, p=0.020), but the post hoc test revealed that MMN latency was longer in the AS group than in the control group over both the right (p=0.000) and the left (p=0.001) hemispheres (table 15). There was no significant difference between the AS and control groups in MMN amplitude.

72

Fig. 13. MMN in response to consonant change (top) and to tone change (bottom) in children with AS and in their controls.

73

Fig. 14. MMN amplitudes elicited by consonant change (on the left) and tone change (on the right) in children with AS and in their controls.

0

1

2

3

4

5

6

Left hemisphere Right hemisphere

-uV

Children with AS Controls

0,0

1,0

2,0

3,0

4,0

5,0

6,0

Left hemisphere Right hemisphere

-uV

Children with AS Controls

6 Discussion

The present study investigated auditory processes that can be tapped by using auditory ERPs, mainly MMN, in children born preterm with VLBW and children with AS. The purpose was to determine whether auditory ERPs, specifically MMN, correlate with naming ability, one index of language development, in VLBW preterm children. Furthermore, the present study aimed to investigate whether ERPs can be used as indices of language impairment. In children with AS, the main aim was to study auditory processing.

The language development of VLBW preterm children and their matched controls was followed up from the age of 2 years up to the age of 6 years, with auditory ERPs being recorded twice, at the ages of 4 and 6 years. In children with AS, ERPs were recorded once, with the age of children varying from 8 to 12 years.

6.1 Auditory ERPs in children born preterm with VLBW

Cortical auditory ERPs permit the assessment of several well-defined stages of central auditory processing. (Kurzberg et al., 1988; Näätänen, 1992; Stapells & Kurzberg, 1991). The present study showed that VLBW preterm children exhibit difficulties in auditory processing at the level of obligatory ERPs and MMN. Although obligatory ERPs in the preterm group and their control group were mainly equal over 9 electrodes, the P1, N2 and N4 amplitudes were smaller in the preterm group than in the control group frontally and centrally in the right hemisphere. Although very little is known about the functional significance of obligatory ERPs for language development, the P1 and N2 parameters have been found to correlate with the standardized tests of Spelling, Auditory processing and Listening comprehension (Cunningham et al., 2000). In contrast, no such correlation was found by Cunningham et al. (2000) between the P1 and N2 parameters and fine-grained auditory discrimination tasks. Therefore, it was suggested that P1 and N2 may reflect neural encoding of sound features in the thalamo-cortical segment of the auditory pathway, serving a wide range of auditory perceptual functions and contributing to listening comprehension, but reflecting different aspects of auditory function compared to

75

MMN. Therefore, it is possible that deficits in sensory processing, as indexed by the attenuated P1 and N2 amplitudes frontally and centrally in the right hemisphere in VLBW preterm children compared to controls, may contribute to deficient language learning.

Preattentive auditory discrimination was also deficient in the preterm group, as revealed by the attenuated MMN amplitude in response to vowel duration change at the age of 4 years and to consonant change at 6 years of age. During 2 years, the heterogeneity within the groups was reduced, resulting in more prominent MMN in response to consonant change in the control group, whereas MMN amplitude in the preterm group diminished during that time. The shortened MMN latency seen in the preterm group was linked with attenuated MMN amplitude and therefore seems to index deficient auditory discrimination. MMN amplitude was attenuated even more in the VLBW preterm children with naming difficulty at the ages of both 4 and 6 years. Therefore, it seems that deficient auditory processing and lexical development are linked to each other, as suggested earlier (Locke, 1994, 1997) .

The results of the behavioral tests were very similar. In accordance with earlier studies (Jennische & Sedin, 1999a, 2001b; Menyuk et al., 1995), the preterm group scored significantly lower than the control group in the auditory discrimination test. At the age of 4 years, the preterm group achieved significantly lower scores in both the consonant and the vowel discrimination subtests, the difference being more significant in the consonant discrimination subtest. Similarly, at the age of 6 years, the preterm group had significantly lower scores in the consonant discrimination test. In vowel discrimination, there was a tendency for the preterm group to score lower than the control group.

The findings of behavioural tests and ERP recordings are consistent. It has been shown earlier that the auditory ERPs of VLBW children already differ from those of healthy controls in infancy (Kurzberg et al., 1988). Here, it was shown that the deficiency in auditory discrimination is also evident later in childhood, as indexed by both attenuated MMN amplitude and behavioural tests. It is possible that this deficiency hampers the later development of preterm children, for the ability to process and categorize auditory stimuli occurring within tens of milliseconds, such as consonant changes, is one of the critical skills needed for language development (Benasich et al., 2002, Benasich & Tallal, 2002).

Similarly to MMN, lMMN amplitude was significantly smaller in the preterm group than in the control group at the age of 4 years in response to consonant change. This was the case to an even greater extent in the preterm children with naming difficulty compared to their controls. At the age of 6 years, a similar tendency was still found. Attenuated lMMN amplitude has earlier been reported in children with ADHD in response to tones (Kilpeläinen, 1999) and in children with dyslexia in response to speech syllables (Schulte-Körne, 1998). Yet, the functional significance of lMMN is rather unknown. Late MMN is mainly found in children and in the same paradigm as MMN. Therefore, lMMN may be linked to auditory discrimination as is MMN. Much of the sensory processing necessary for speech discrimination occurs within a brief window of time (Benasich et al., 2002; Pulvermüller, 2002). Within that time window, several temporally co-ordinated processes generated by separate subcortical and cortical areas take place. At least two different explanations may be found for lMMN. Firstly, it is possible that the temporal co-ordination does not function equally well in children as in

76

adults. Thatcher et al. (1987) showed that EEG coherence and phase in the frontal and temporal lobes develop markedly up till adolescence. Secondly, children may use more extensive or even different subcortical or cortical areas for the discrimination process. These differences may result in an additional component, called lMMN. In VLBW preterm children, the temporal organization may not function well and the use of subcortical or cortical areas typical in healthy children may be hampered because of early insults to the developing brain.

It is not possible to tell from our results why the auditory processing, as indexed by obligatory ERPs, MMN and behavioural measurements, was deficient in the preterm population. However, it has been shown earlier that difficulties in auditory processing may arise as a consequence of a brain lesion (Isaacs et al., 2000; Peterson et al., 2000, 2002). Peterson et al. (2002) showed, by using fMRI, that preterm children used such brain areas to process meaningful speech (semantics) as were used by term children for processing strings of meaningless sounds. The greater the resemblance in the preterm group, the more impaired was their understanding of the story they heard, and the lower was their verbal IQ. Ten of the present VLBW preterm children had pathological findings in the parietal area that could be defined as sulcal dilatation, and 12 had dilatation of the lateral ventricles or a reduction of cerebral white matter neonatally. It has been shown that preterm birth impairs brain maturation most frequently in the mid-temporal and parieto-occipital cortices (Peterson et al., 2000), which are important for auditory processes, for language and other cognitive processes and for the integration of the information received (Seikel et al., 2000). It has also been shown that periventricular white matter lesions affect the early development of cortical grey matter, increase the ventricular and extracerebral cerebrospinal fluid volume (Inder et al., 1999; Inder & Volpe, 2000) and disrupt the callosal fibres that mediate higher cortical functions (Whitaker et al., 1996). To summarize, it is possible that the neonatal brain lesions found in the present population may have contributed both to deficient auditory sensory processing, as indexed by obligatory ERPs, and to preattentive auditory discrimination, as indexed by MMN

6.2 Language development in VLBW preterm children

Both language comprehension and language production were deficient in preterm children. Intellectual disabilities could be the primary or only cause of deficient language comprehension and production. This may not be the case here, however. The general development quotient on Griffiths Developmental Scales (Griffiths, 1954) was mildly abnormal in only one child with CP at the age of 18 months, and the result was partly due to motor disability, while the subquotient for hearing and speech was within normal limits in all cases. At the age of 6 years, too, GQ was mildly abnormal in 4 children. However, SG was abnormal in 9 children. Therefore, it seems that VLBW preterm children, even according to the Griffiths Developmental Scales (Griffiths, 1954), exhibit mainly difficulties in the area of language development. Although the Griffiths Developmental Scales (Griffiths, 1954) are not equivalent to psychological tests, the scale identifies the

77

children with the most severe disabilities, and it is often used to assess the development of young children.

Deficient language production manifested as deficient development of the lexicon and morphology. At the age of 2 years, lexical development did not differ significantly between the preterm and control groups. It is possible that this result is due to the marked variability within the two groups. In addition, there is also the possibility that the children in the control group already knew so many words that MCDI failed to tap the potential difficulty encountered by the preterm group. Thereafter, however, lexical development seems to be the most prominent phenomenon differentiating between the groups. It has been shown earlier that lexical development precedes morphological development in the early phase of language development (Bates et al., 1995). Furthermore, acceleration in lexical development improves the infant’s auditory discrimination skills as well as efficient storage and retrieval of linguistic information (Locke, 1994, 1997). Hence, deficient morphological development may be a reflection of deficient lexical and auditory discrimination. Bishop (1997) states that lexical development may be seen as a process of storing representations of familiar sequences of speech sounds in a mental lexicon and associating these with specific meanings. If the child is not able to discriminate and identify the components of words in detail, both word comprehension and word production become more difficult, with word production, naming being one index of it, being hampered even more (Bishop, 1997).

The number of changes in neonatal MRI was strikingly high in the present study group. Therefore, the present results mainly represent the outcome of the VLBW children with structural changes in neonatal brain MRI and should not be generalized to all VLBW preterm children. Furthermore, it is possible that the parents of children with normal language development at the age of 2 years were not willing to participate in this study. In addition, the study series was small. For these reasons, the results must be considered with caution. However, the drop-out rate of the preterm children was rather low, although not all of the measurements were done at all age levels. In addition, the results were consistent during the 4-year follow-up time, and the results of the preterm group were compared with those of matched controls.

6.3 Auditory ERPs and language development on a single-subject level

On the basis of the visual ratings, MMN was weak or totally missing in 5 VLBW preterm children. These 5 children exhibited naming difficulty at the ages of both 4 and 6 years. Furthermore, MMN was absent at 6 or more electrodes in one control child. Naming difficulty was also seen in that child. These results suggest that an attenuated MMN response may be a precursor of naming difficulty even on the single-subject level, as earlier found on the group level (Leppänen et al., 2002; Lyytinen et al., 2001).

It has been suggested earlier that MMN is not reliable and sensitive enough for individual diagnostics, because of its poor reliability and individual stability (Uwer et al., 2002). It may, indeed, not be possible to rely on amplitude and latency. MMN amplitude

78

and latency may change as functions of many factors, including developmental stage and study design (Korpilahti, 1996). Furthermore, there are not enough data to assess what is normal and abnormal at each age and in each study paradigm. However, we here compared preterm children with healthy controls with normal language development. The study showed that, although amplitude and latency may change, only children with poor lexical development have a missing or questionable MMN at 6 or more of the 9 electrodes at the age of 4 years. There was also one preterm and one control child with naming difficulty, who were not identified on the basis of the MMN results. It is possible that the auditory difficulties exhibited by these children were not detected by MMN. It is, however, also possible that the naming deficiency in these children was due to some other reason than poor auditory discrimination.

At the age of 2 years, both the preterm and the control groups were heterogeneous, including children with slow language development and very fast language development. By the age of 4 years, most of the control children progressed fast, which resulted in more homogeneous results in this group. Two of the control children continued to have difficulties at the age of 6 years, manifested as poor naming ability. However, fast development was not typical of the VLBW children with delayed language development at the age of 2 years. The same VLBW children continued to have difficulties throughout the 4-year follow-up period. In addition, most VLBW preterm children scored on the lower limit of normal standard deviation in the naming test at the age of 6 years. Variations in the rate and style of learning are common in healthy children (Bates et al., 1995). However, special attention should be paid to children with brain damage (Bates et al., 1995).

6.4 Auditory ERPs in children with Asperger syndrome

Autobiographies of individuals with AS refer to problems with auditory perception (Williams, 1992). However, studies on the accuracy of central auditory processing in AS are sparse, if any. The present study showed that children with AS display abnormalities in auditory sensory processing as indexed by long-latency obligatory ERPs and MMN. The main finding was the delayed MMN latency, which was more obvious over the right hemisphere than the left hemisphere. In addition, MMN in response to tone pitch was more deficient than that in response to syllables, tone pitch processing being mainly carried out by the right-hemisphere cortex (Zatorre et al., 1992, 2002). Together, these findings suggest a possibly predominant right-hemisphere dysfunction in AS, which corroborates the existing behavioural and brain-imaging results (Ellis & Gunter, 1999; McKelvey et al., 1995).

However, deficiency was also encountered in the left hemisphere function. The lateralization of syllable discrimination was reversed in the AS group, as indexed by the larger MMN amplitude over the right than the left hemisphere. Most often, speech sounds produce a larger MMN amplitude over the left hemisphere (Rinne et al., 1999; Shtyrov et al., 2000; Tervaniemi et al., 2000), which was also the case in the present control group. Thus, it is possible that not only right-hemisphere functioning is impaired in children

79

with AS, but at a more general level, inter-hemispheric specialization as such is abnormal.

We may wonder whether the prolongation of MMN latency in the AS group is due to N1 enhancement. However, the time window for MMN occurs later than that of typical N1. Furthermore, N1 is only observed at slow stimulus rates in children in the age range of 7-12 years (Bruneau & Gomot, 1998; Paetau, 1995), which was the age range of the present study. It is also unlikely that the differences in obligatory ERP peaks would have contributed to the results, since the latency of MMN differed from those of N2 or N4. Also, there were no group differences in N2 or N4 latencies or amplitudes in response to syllable change. Further, there were no hemispheric differences in either N2 or N4 in either condition. Therefore, delayed MMN latency is suggested to be a real phenomenon.

Individuals with AS have considerable difficulties in interpreting affective intonation of speech (Gillberg & Gillberg, 1989). Affective intonation is mainly carried out by tone pitch (Snow, 2000). In the present study, children with AS mainly exhibited difficulties in discriminating tone pitch, as indexed by the most prolonged MMN latency in response to tones. It is not possible to compare the MMN results with the behavioural measurements in the present study, but it is possible that the difficulties encountered in speech comprehension are linked to the kind of deficient auditory processing reported here.

While the present results await replication, it can be suggested that the prolongation of MMN latency found in both hemispheres in the children with AS might reflect a general developmental delay in the auditory cortical system. The present study leaves open the question of whether this delay is present from birth and, further, whether it continues to be present later in adolescence and adulthood. Furthermore, to verify whether peculiarities in speech and language can be linked to deficient auditory processing, language development together with auditory ERPs should be studied in children with AS. The parents and siblings of children with AS may be studied together with the children with AS to investigate the familial background of AS.

7 Conclusions

1. Preattentive auditory processing was deficient in children born preterm with VLBW, as indexed by the diminished MMN amplitude.

2. Auditory processing deficit in preterm children was encountered at different levels of auditory processing at the age of 4 and 6 years. At the age of 4 years, the MMN and lMMN were attenuated. At the age of 6 years, the attenuation was evident also at the level of obligatory ERPs, mainly in the right hemisphere.

3. VLBW preterm children lagged significantly behind their controls in language development from the age of 2 years up to the age of 6 years. Deficient language development manifested as poor language comprehension, poor lexical development, poor morphological development as well as poor auditory discrimination ability.

4. The MMN response of auditory ERPs was found to be an index of language impairment among prematurely born children and their controls at the ages of both 4 and 6 years. MMN amplitude was found to be attenuated in children with poor lexical development but not in children with normal lexical development.

5. Children with AS displayed abnormalities in auditory processing, as indexed by long-latency obligatory auditory ERPs (P1 and N2) and by MMN. MMN latency was more delayed over the right hemisphere than the left hemisphere, and more so for tones than for syllables, suggesting possibly predominant right-hemisphere dysfunction in children with AS.

References

Alho, K. (1995) Cerebral generators of mismatch negativity (MMN) and its magnetic counterpart (MMNm) elicited by sound changes. Ear & Hearing, 16, 38-51.

Alku, P., Tiitinen, H. & Näätänen, R. (1999) A method for generating natural-sounding speech stimuli for cognitive brain research. Clinical Neurophysiology, 110, 1329-1333.

Attwood, T. (1998) Asperger´s syndrome: A guide for parents and professionals. London: Jessica Kingsley Publishers.

Baron-Cohen, S., Ring, H., Wheelwright, S., Bullmore, E., Brammer, M., Simmons, S. & Williams, S. (1999) Social intelligence in the normal and autistic brain: an fMRI study. European Journal of Neuroscience, 11, 1891-1898.

Bates, E., Dale, P. & Thal, D. (1995) Individual differences and their implications for theories of language development. In: P. Fletcher & B. MacWhinney (eds) The handbook of child language. Oxford: Blackwell publishers, 96-151.

Benasich, A. & Tallal, P. (2002) Infant discrimination of rapid auditory cues predicts later language impairment. Behavioral Brain Research, 136, 31-49.

Benasich, A., Thomas, J., Choudhury, N. & Leppänen P. (2002) The importance of rapid auditory processing abilities to early language development: evidence from converging methodologies. Developmental Psychobiology, 40, 278-292.

Bishop, D. (1997) Uncommon understanding. Development and disorders of language comprehension in children. Hove: Psychology Press.

Bradlow, A., Kraus, N., Nicol, T., McGee, T., Cunningham, J., Zecker, S. & Carrell, T. (1999) Effects of lengthened formant transition duration on discrimination and neural representation of synthetic CV syllables by normal and learning-disabled children. Journal of the Acoustic Society of America, 106, 2086-2096.

Briscoe, J., Gathercole, S. & Marlow, N. (1998) Short-term memory and language outcomes after extreme prematurity at birth. Journal of Speech and Hearing Research, 41, 654-666.

Brown, R. (1973) A first language. The early stages. London: George Allen & Unwin Ltd. Bruneau, N., Garreaus, B., Roux, S. & Lelord, G. (1987) Modulation of auditory evoked potentials

with increasing stimulus intensity in autistic children. In: Johnson, R., Parasuraman, R., Rohrbaugh, J. (eds) Current Trends in Event-Related Potential Research (EEG. Suppl. 40). Amsterdam: Elsevier, 584-589.

Bruneau, N. & Gomot, M. (1998) Auditory evoked potentials (N1 wave) as indices of cortical development. In: B. Garreau (Ed.) Neuroimaging in Child Neuropsychiatric Disorders. New York: Springer Verlag, 113-123.

Bruneau, N., Roux, S., Adrien, JL. & Barthélémy, C. (1999) Auditory associative cortex dysfunction in children with autism: evidence from late auditory evoked potentials (N1 wave-T complex). Clinical Neurophysiology, 110, 1927-1934.

82Casanova, M., Buxhoeveden, D., Switala, A. & Roy, E. (2002) Asperger´s syndrome and cortical

neuropathology. Journal of Child Neurology, 17, 142-145. Čeponienė, R. (2001) Event-related potential (ERP) indices of central auditory development in

healthy children and in children with oral clefts. Doctoral dissertation. Helsinki: Yliopistopaino. Čeponienė, R., Cheour, M. & Näätänen, R. (1998) Interstimulus interval and auditory event-related

potentials in children: evidence for multiple generators. Electroencephalography and Clinical Neurophysiology, 108, 345-354.

Čeponienė, R., Hukki, J., Cheour, M., Haapanen, M-L., Ranta, R. & Näätänen, R. (1999a) Cortical auditory dysfunction in children with oral clefts: relation with cleft type. Clinical Neurophysiology, 110, 1921-1926.

Čeponienė, R., Hukki, J., Cheour, M., Haapanen, M-L., Koskinen, M., Alho, K. & Näätänen, R. (2000) Dysfunction of the auditory cortex persists in infants with certain cleft types. Developmental Medicine and Child Neurology, 42, 258-265.

Čeponienė, R., Kusnherenko, E., Fellman, V., Renlund, M., Suominen, K. & Näätänen, R. (2002a) Event-related potential features indexing central auditory discrimination by newborns. Cognitive Brain Research, 13, 101-113.

Čeponienė, R., Lepistö, T., Alku, P., Aro, H. & Näätänen, R. (2003a) Event-related potential indices of auditory vowel processing in 3-year-old children. Clinical Neurophysiology, 114, 652-661.

Čeponienė, R., Lepistö, T., Shestakova, A., Vanhala, R., Alku, P., Näätänen, R. & Yaguchi, K. (2003b) Speech-sound-selective auditory impairment in children with autism: They can perceive but do not attend. Proceedings of the National Academy of Sciences, 29, 5567-5572.

Čeponienė, R., Rinne, T. & Näätänen, R. (2002b) Maturation of cortical sound processing as indexed by event-related potentials. Clinical Neurophysiology, 113, 870−882.

Čeponienė, R., Service, E., Kurjenluoma, S., Cheour, M. & Näätänen, R. (1999b) Children’s performance on pseudoword repetition depends on auditory trace quality: Evidence from event-related potentials. Developmental Psychology, 35, 709-20.

Čeponienė, R., Shestakova, A., Balan, P., Alku, P., Yaguchi, K. & Näätänen, R. (2001) Children´s auditory event-related potentials index stimulus complexity and ”speechness”. International Journal of Neurocience, 109, 245-260.

Čeponienė, R., Yaguchi, K., Shestakova, A., Alku, P., Suominen, K. & Näätänen R. (2002c) Sound complexity and ”speechness” effects on pre-attentive auditory discrimination in children. International Journal of Psychophysiology, 43, 199-211.

Cheour, M. (1998) Mismatch negativity (MMN) as a tool for investigating auditory discrimination in infants and school-age children. Doctoral dissertation. Helsinki; Yliopistopaino.

Cheour, M., Čeponienė, R., Lehtokoski, A., Luuk, A., Allik, J., Alho, K. & Näätänen R. (1998a) Development of language-specific phoneme representations in the infant brain. Nature Neuroscience, 1, 351-353.

Cheour, M., Čeponienė, R., Hukki, J., Haapanen, M-L., Näätänen, R. & Alho, K. (1999) Brain dysfunction in neonates with cleft palate revealed by the mismatch negativity. Clinical Neurophysiology, 110, 324-328.

Cheour, M., Haapanen, M-L., Čeponienė, R., Hukki, J., Ranta, R. & Näätänen, R. (1998b) Mismatch negativity (MMN) as an index of auditory sensory memory deficit in cleft-palate and CATCH syndrome children. NeuroReport, 9, 2709-2712.

Cheour, M., Haapanen, M-L., Hukki, J., Čeponienė, R., Kurjenluoma, S., Alho, K., Tervaniemi, M., Ranta, R. & Näätänen, R. (1997) The first neurophysiological evidence for cognitive brain dysfunctions in children with CATCH. NeuroReport, 8, 1785-1787.

Cheour, M., Korpilahti, P., Martynova, O. & Lang, H. (2001) Mismatch negativity and late discriminative negativity in investigating speech perception and learning in children and infants. Audiology & Neuro Otology, 6, 2-11.

Cheour, M., Shestakova, A., Alku, P., Čeponienė, R. & Näätänen, R. (2002) Mismatch negativity shows that 3-6-year-old children can learn to discriminate non-native speech sounds within two months. Neurosciece Letters, 325, 187-190.

Chomsky, N. (1957) Syntactic structures. Hague: Mouton.

83Clancy, B. & Finlay, B. (2001) Neural correlates of early language learning. In: M. Tomasello & E.

Bates (eds) Language development. The essential readings. Oxford: The Blackwell Publishers, 307-330.

Courchesne, E. (1990) Chronology of postnatal human brain development: Event-related potential, positron emission tomography, myelinogenesis, and synaptogenesis studies. In: R. Rohrbaugh, R Parasuraman & R. Johnson (Eds.) Event-related brain potentials: Basic issues and application. Oxford: Oxford University Press, 210-241.

Courchesne, E., Courchesne, R., Hicks, G. & Lincoln, A. (1985a) Functioning of the brainstem auditory pathway in non-retarded autistic individuals. Electroencephalography and Clinical Neurophysiology, 61, 491-501.

Courchesne, E., Lincoln, A., Kilman, B. & Galambos, R. (1985b) Event-related brain potential correlates of the processing of novel visual and auditory information in autism. Journal of Autism and Developmental Disorders, 15, 55-76.

Cowan, N., Winkler, I., Teder, W. & Näätänen, R. (1993) Memory prerequisites of mismatch negativity in the auditory event-related potential (ERP). Journal of Experimental Psychology, Learning, Memory, and Cognition, 19, 909-921.

Csépe, V. (1995) On the origin and development of the mismatch negativity. Ear & Hearing, 16, 91-104.

Cunningham, J., Nicol, T., Zecker, S. & Kraus, N. (2000) Speech-evoked neurophysiologic responses in children with learning problems: Development and behavioral correlates of perception. Ear & Hearing, 21, 554-568.

Cutler, A. & Clifton, C. (1999) Comprehending spoken language: a blueprint of the listener. In: C. Brown & P. Hagoort (eds) The Neurocognition of Language. Oxford: University Press, 123-166.

Damasio, A. & Maurer, R. (1978) A neurological model for childhood autism. Archives of Neurology, 35, 777-786

Dammann, O., Durum, S. & Leviton, A. (2001) Do white cells matter in white matter damage? Trends in Neurosciences, 24, 320-324.

Dawson, G., Finley, C., Phillips, S. & Galpert, L. (1986) Hemispheric Specialization and the language abilities of autistic children. Child Development, 57, 1440-1453.

Dawson, G., Finley, C., Phillips, S. & Galpert, L. (1988) Reduced P3 amplitude of the event-related brain potential: Its relationship to language ability in autism. Journal of Autism and Developmental Disorders, 18, 493-504.

Dennis, M., Lazenby, A. & Lockyer, L. (2001) Inferential language in high-functioning children with autism. Journal of Autism and Developmental Disorders, 31, 47-54.

Donchin, E., Ritter, W. & McCallum, W. (1978) Cognitive psychology: The endogenous components of the ERP. In: E. Callaway, P. Tueting & S. Koslow (eds) Event-Related Brain Potentials in Man. New York: Academic Press, 349-441.

Doyle, J. (1999) Forms of hearing assessment. In J. Doyle (ed) Practical audiology for speech-language therapists. London: Whurr Publishers Ltd., 58-99.

Dunn, M., Vaughan H., Kreuzer, J. & Kurzberg, D. (1999) Electrophysiologic correlates of semantic classification in autistic and normal children. Developmental Psychology, 16, 79-99.

Dusik, A. (1997) Medical outcomes in preterm infants. Seminars in Postnatology, 21, 164-177. Ehlers, S. & Gillberg, C. (1993) The epidemiology of Asperger Syndrome. A total population

study. Journal of Child Psychology and Psychiatry, 34, 1327-1350. Eimas, P., Siqueland, E., Juyczyk, P. & Vigorito, J. (1971) Speech perception in infants. Science,

171, 303-306. Ellis, H. & Gunter, H. (1999) Asperger syndrome: a simple matter of white matter? Trends in

Cognitive Sciences, 3, 192-200. Escobar, G., Littenberg, B. & Petitti, D. (1991) Outcome among surviving very low birthweight

infants: a meta-analyses. Archives of Disease in Childhood, 66, 204-211. Fawer, C., Besnier, S., Forcada, M., Buclin, T. & Calame, A. (1995) Influence of perinatal,

developmental and environmental factors on cognitive abilities of preterm children without major impairments at 5 years. Early Human Development, 43, 151-164.

84Fenson, L., Dale, P., Reznick, J., Bates, E., Thal. D. & Pethick, S. (1994) Variability in early

communicative development. Monograps of the Society for Research in Child Development, 59, 242.

Frisk, V. & Whyte, H. (1994) The long-term consequences of periventricular brain damage on language and verbal memory. Developmental Neuropsychology, 10, 313-333.

Ghaziuddin, M. & Gerstein, L. (1996) Pedantic speaking style differentiates Asperger syndrome from high-functioning autism. Journal of Autism and Developmental Disorders, 26, 585-595.

Ghaziuddin, M., Weidmer-Mikhail, E. & Ghaziuddin, N. (1998) Comorbidity of Asperger syndrome: a preliminary report. Journal of Intellectual Disability Research, 42, 279-283.

Giard, M., Perrin, F., Pernier, J. & Bouchet, P. (1990) Brain generators implicated in the processing of auditory stimulus deviance: A topographic event-related potential study. Psychophysiology, 27, 627-640.

Gilchrist, A., Green, J., Cox, A., Burton, D., Rutter, M. & Le Couteur, A. (2001) Development and current functioning in adolescents with Asperger syndrome: A comparative study. Journal of Child Psychology and Psychiatry, 42, 227-240.

Gillberg, C. and Gillberg, I.C. (1989) Asperger syndrome – some epidemiological considerations: a research note. Journal of Child Psychology and Psychiatry, 30, 631-638.

Gomot, M., Giard, M-H., Adrien, J-L., Barthelemy, C. & Bruneau, N. (2002) Hypersensitivity to acoustic change in children with autism: Electrophysiological evidence of left frontal cortex dysfunction. Psychophysiology, 39, 577-584.

Gomot, M., Giard, M-H., Roux, S., Barthélémy, C. & Bruneau, N. (2000) Maturation of frontal and temporal components of mismatch negativity (MMN) in children. NeuroReport, 11, 3109-3112.

Greene, M. (2002) Outcomes of very low birth weight in young adults. New England Journal of Medicine, 346, 146-148.

Griffiths, R. (1954) The abilities of young children. A comprehensive system of mental measurement for the first eight years of life. London: McGraw Hill.

Grunau, R., Kearney, S. & Whitfield, M. (1990) Language development at 3 years in pre-term children of birth weight below 1000 g. British Journal of Disorders of Communication, 25, 173-182.

Grunau, R., Whitfield, M. & Davis, C. (2002) Pattern of Learning disabilities in children with extremely low birth weight and broadly average intelligence. Archives of Pediatrics & Adolescence Medicine, 156, 615-620.

Gunter, J., Ghaziudding, M. & Ellis, H. (2002) Asperger syndrome: Tests of right hemisphere functioning and interhemispheric communication. Journal of Autism and Developmental Disorders, 32, 263-281.

Guttorm, T., Leppänen, P., Richardson, U. & Lyytinen, H. (2001) Event-related potentials and consonant differentiation in newborns with familial risk for dyslexia. Journal of Learning Disabilities, 34, 534-544.

Hagberg, B., Hagberg, G. & Olow, I. (1975) The changing panorama of cerebral palsy in Sweden 1954-1970. Acta Paediatrica Scandinavica, 64, 187-192.

Hakulinen, A. & Karlsson, F. (1979) Nykysuomen lauseoppia. Jyväskylä: Gummerus. Hari, R., Hämäläinen, M., Ilmoniemi, R., Kaukoranta, E., Reinikainen, K., Salminen, J., Alho, K.,

Näätänen, R. & Sams, M. (1984) Responses of the primary auditory cortex to pitch changes in a sequence of tone pips: Neuromagnetic recordings in man. Neuroscience Letters, 50, 127-132.

Harley, T. (2001) The psychology of language. Hove: Psychology Press. Hayes, E., Warrier, C., Nicol, T., Zecker, S. & Kraus, N. (2003) Neural plasticity following

auditory training in children with learning problems. Clinical Neurophysiology 114, 673-684. Holling, E. & Leviton, A. (1999) Characteristics of cranial ultrasound white-matter echolucencies

that predict disability: a review. Developmental Medicine and Child Neurology, 41, 136-139. Holopainen, I., Korpilahti, P., Juottonen, K., Lang, H. & Sillanpää. M. (1996) Abnormal frequency

mismatch negativity in mentally retarded children and in children with developmental dysphasia. Journal of Child Neurology, 13, 178-183.

Holopainen, I., Korpilahti, P., Juottonen, K., Lang, H. & Sillanpää. M. (1997) Attenuated auditory event-related potential (mismatch negativity) in children with developmental dysphasia. Neuropediatrics, 4, 253-256.

85Howlin, P. (2003) Outcome in high-functioning adults with autism with and without early language

delays: Implications for the differentiation between autism and Asperger syndrome. Journal of Autism and Developmental Disorders, 33, 3-13.

ICD-10 Tautiluokitus (1999) Finnish version of the International Statistical Classification of Diseases and Related Health Problems, Tenth Revision in Geneva 1992. Sosiaali ja terveysalan tutkimus- ja kehittämiskeskus. 2nd Ed. Turenki: Ro-Offsed.

Ilvonen T-M., Kujala, T., Kiesiläinen, A., Salonen, O., Kozou, H., Pekkonen, E., Roine, R., Kaste, M. & Näätänen, R. (2003) Auditory discrimination after left-hemisphere stroke. A mismatch negativity follow-up study. Stroke, 34, 1746-1753.

Inder, T., Huppi, P., Warfield, S., Kikinis, R., Zientara, G., Barnes, P., Jolesz, F. & Volpe, J. (1999) Periventricular white matter injury in the premature infant is followed by reduced cerebral cortical gray matter volume at term. Annals of Neurology, 46, 755-760.

Inder, T. & Volpe, J. (2000) Mechanisms of perinatal brain injury. Seminars in Neonatology, 5, 3-16.

Isaacs, E., Lucas, A., Chong, W., Wood, S., Johnson, C., Marshall, C., Vargha-Khadem, F. & Gadian, D. (2000) Hippocampal volume and everyday memory in children of very low birth weight. Pediatric Research, 47, 713-720.

Jacobsen, L., Ek, U., Fernell, E., Flodmark, O. & Broberger, U. (1996) Visual impairment in preterm children with periventricular leukomalacia – visual, cognitive and neuropediatric characteristics related to cerebral imaging. Developmental Medicine and Child Neurology, 38, 724-735.

Jennische, M. & Sedin, G. (1998) Speech and language skills in children who required neonatal intensive care. I. Spontaneous speech at 6.5. years of age. Acta Paediatrica, 87, 654-666.

Jennische, M. & Sedin, G. (1999a) Speech and language skills in children who required neonatal intensive care II. Linguistic skills at 6½ years of age. Acta Paediatrica, 88, 371-383.

Jennische, M. & Sedin, G. (1999b) Speech and language skills in children who required neonatal intensive care: evaluation at 6.5 y of age based on interviews with parents. Acta Paediatrica, 88, 975-982.

Jennische, M. & Sedin, G. (2001a) Spontaneous speech at 6½ years of age in children who required neonatal intensive care in 1986-1989. Acta Paediatrica, 90, 22-33.

Jennische, M. & Sedin, G. (2001b) Linguistic skills at 6½ years of age in children who required neonatal intensive care in 1986-1989. Acta Paediatrica, 90, 199-212.

Jennische, M. & Sedin, G. (2003) Gender differences in outcome after neonatal intensive care: speech and language skills are less influenced in boys than in girls at 6.5 years. Acta Paediatrica, 92, 364-378.

Jusczyk, P. (1999) How infants begin to extract words from speech. Trends in Cognitive Sciences, 3, 323-328.

Jusczyk, P. & Hohne, E. (1997) Infants’ memory for spoken words. Science, 277, 1984-1986. Jusczyk, P., Luce, P. & Charles-Luce, J. (1994) Infants’ sensitivity to phonotactic patterns in the

native language. Journal of Memory and Language, 33, 630-645. Kaplan, E., Goodglass, H. & Weintraub, S. (1983) The Boston Naming Test. Philadelphia: Lea &

Febiger. Karhu, J., Herrgård, E., Pääkkönen, A., Luoma, L., Airaksinen, E. & Partanen, J. (1997) Dual

cerebral processing of elementary auditory input in children. NeuroReport, 8, 1327-1330. Karvonen, A. (1985) Oululaislasten syntaksia. Havaintoja alle 2½ vuotiaiden lauseista. MA thesis

from the Department of Finnish, Saami and Logopedics. Oulu: University Press. Kemner, C., Verbaten, M., Cuperus, J., Camfferman, G. & van Engeland, H. (1995) Auditory

event-related brain potentials in autistic children and three different control groups. Biological Psychiatry, 38, 150-165.

Kilpeläinen, R., Partanen, J. & Karhu, J. (1999) Reduced mismatch negativity (MMN) suggests deficits in preattentive auditory processing in distractible children. NeuroReport, 10, 3341-3345.

King, C., Warrier, C., Hayes, E. & Kraus, N. (2002) Deficits in auditory brainstem pathway encoding of speech sounds in children with learning problems. Neuroscience Letters, 319, 111−115.

86Kok, L. & Solman, R. (1995) Velocardiofacial syndrome: Learning difficulties and intervention.

Journal of Medical Genetics, 32, 612-618. Koller, H., Lawson, K., Rose, S., Wallace, I. & McCarton, C. (1997) Patterns of cognitive

development in very low birth weight children during the first six years of life. Pediatrics, 99, 383-389.

Korpilahti, P. (1991) Comprehension problems in developmental dysphasia. In: M. Laine, J. Niemi & P. Koivuselkä-Sallinen (Eds), Proceedings of the Fourth Conference of Neurolinguistics. Turku. Joensuu Papers in Neuropsychology and Neurolinguistics, 23, 75-81

Korpilahti, P. (1995) Auditory discrimination and memory functions in SLI children: A comprehensive study with neurophysiological and behavioral methods. Scandinavian Journal of Logopedics and Phoniatrics, 20, 131-139.

Korpilahti, P. (1996) Electrophysiological correlates of auditory perception in normal and language impaired children. Doctoral dissertation. Annales Universitas Turkuensis D 232.

Korpilahti, P., Krause, C., Holopainen, I. & Lang, H. (2001) Early and late mismatch negativity elicited by words and speech-like stimuli in children. Brain & Language, 76, 332-339.

Korpilahti, P. & Lang, H. (1994) Auditory ERP components and mismatch negativity in dysphasic children. Electroencephalography and Clinical Neurophysiology, 91, 256-264.

Kraus, N. & Cheour, M. (2000) Speech sound representation in the brain. Audiology & Neuro-Otology, 5, 1140-150.

Kraus, N., McGee, T., Carrell, T. & Sharma, A. (1995) Neurophysiologic bases of speech discrimination. Ear & Hearing, 16, 19-37.

Kraus, N., McGee, T., Carrell, T., Sharma, A., Micco, A. & Nicol, T (1993) Speech-evoked cortical potentials in children. Journal of American Academy of Audiology, 4, 238-248.

Kraus, N., McGee, T., Carrell, T., Zecker, S., Nicol, T. & Koch, D. (1996) Auditory neurophysiologic responses and discrimination deficits in children with learning problems. Science, 273, 971-973.

Kraus, N., McGee T. & Burton Koch, D. (1998) Speech sound representation, perception, and plasticity: a neurophysiologic perspective. Audiology & Neuro-Otology, 3, 168-182.

Kraus, N., McGee T., Littman, T., Nicol, T. & King, C. (1994) Nonprimary auditory thalamic representation of acoustic change. Journal of Neurophysiology, 72, 1270-1277.

Krägeloh-Mann, I., Toft, P., Lunding, J., Andresen, J., Pryds, O. & Lou, H. (1999) Brain lesions in preterms: origin, consequences and compensation. Acta Paediatrica, 88, 897-908.

Kujala, T., Alho, K. & Näätänen, R. (2000) Cross-modal reorganization of human cortical functions. Trends in Neurosciences, 23, 115-120.

Kujala, T., Huotilainen, M., Sinkkonen, J., Ahonen, A., Alho, K., Hämäläinen, M., Ilmoniemi, R., Kajola, M., Knuutila, J., Lavikainen, J., Salonen, O., Simola, J., Standertskjöld-Nordenstam, CG., Tiitinen, H., Tissari, S. & Näätänen, R. (1995) Visual cortex activation in blind humans during sound discrimination. Neuroscience Letters, 183, 143-146.

Kujala, T., Kallio, J., Tervaniemi, M. & Näätänen, R. (2001a) The mismatch negativity as an index of temporal processing in audition. Clinical Neurophysiology, 112, 1712-1719.

Kujala, T., Karma, J., Čeponienė, R., Belitz, S., Turkkila, P., Tervaniemi, M. & Näätänen, R. (2001b) Plastic neural changes and reading improvement caused by audiovisual training in reading-impaired children. Proceedings of the National Academy of Sciences, 98, 10509-10514.

Kujala, T. & Näätänen R. (2001) The mismatch negativity in evaluating central auditory dysfunction in dyslexia. Neuroscience and Behavioral Reviews, 25, 535-543.

Kunnari, S. (2000) Characteristics of early lexical and phonological development in children acquiring Finnish. Doctoral dissertation. Acta Universitas Ouluensis, B Humaniora 34. Oulu: Oulu University Press.

Kurzberg, D., Stapells, D., & Wallace, I. (1988) Event-related potential assessment of auditory system integrity: Implications for language development. In: P. Vietze & H. Vaughan (eds) Early identification of infants with developmental disabilities. Philadelphia: Grune & Stratton, 160-180.

Kurtzberg, D., Stone, C. & Vaughan, H. (1986) Cortical responses to speech sounds in the infant. In: R. Gracco & I. Bodis-Wollner (eds) Frontiers of clinical neuroscience. Evoked potentials. New York: Alan R. Liss, 513-520.

87Kushnerenko, E. (2003) Maturation of the cortical auditory event-related brain potentials in

infancy. Doctoral Dissertation. Helsinki: Yliopistopaino. Kushnerenko, E., Čeponienė, R., Balan, P., Fellman, V., Huotilainen, M. & Näätänen, R. (2002a)

Maturation of the auditory event-related potentials during the first year of life. NeuroReport, 13, 47−51.

Kushnerenko, E., Čeponienė, R., Balan, P., Fellman, V. & Näätänen R. (2002b) Maturation of the auditory change detection response in infants: a longitudinal ERP study. NeuroReport, 13, 1843-1848.

Laine, M., Koivuselkä-Sallinen, P., Hänninen, R. & Niemi, J. (1988) The Boston Naming Test. Finnish version. Helsinki: Psykologien Kustannus Oy.

Lang, H., Nyrke, T., Ek, M., Aaltonen, O., Raimo, I. & Näätänen, R. (1990) Pitch discrimination performance and auditory event-related potentials. In: A. Kok, G. Mulder & M. Verbaten (eds) Psychophysiological brain research. Vol1. Tilburg: Tilburg University Press, 294-298.

Lang, H., Nyrke, T. & Näätänen, R. (1989) The correlation between pitch discrimination performance and auditive ERPs. Electroencephalography and Clinical Neurophysiology, 73, 95.

Leppänen, P., Pihko, E., Eklund, K. & Lyytinen, H. (1999) Cortical responses of infants with and without a genetic risk for dyslexia: II. Group effects. NeuroReport, 10, 969-973.

Leppänen, P., Richardson, U., Pihko, E., Eklund, K., Guttorm, T., Aro, M. & Lyytinen, H. (2002) Brain responses to changes in speech sound durations differ between infants with and without familial risk for dyslexia. Developmental Psychology, 22, 407-422.

Liégeois-Chauvel, C., Musolino, A., Badier, P., Marquis, P. & Chauvel, P. (1994) Evoked potentials recorded from the auditory cortex in man: evaluation and topography of the middle latency components. Electroencephalography and Clinical Neurophysiology, 92, 204-214.

Lin, Y., Okumara, A., Hayakawa, F., Kato, T., Kuno, K. & Watanabe, K. (2001) Quantitative evaluation of thalami and basal ganglia in infants with periventricular leukomalacia. Developmental Medicine and Child Neurology, 43, 481-485.

Lincoln, A., Courchesne, E., Harms, L. & Allen, M. (1993) Contextual probability evaluation in autistic, receptive developmental language disorder, and control children: event-related brain potential evidence. Journal of Autism and Developmental Disorders, 23, 37-58.

Locke, J. (1994) Gradual emergence of developmental language disorders. Journal of Speech and Hearing Research, 7, 608-616.

Locke, J. (1997) A theory of neurolinguistic development. Brain and Language, 58, 265-326. Luoma, L., Herrgård, E., Martikainen, A. & Ahonen, T. (1998) Speech and language development

of children born at <32 weeks’ gestation: a 5-year prospective follow-up study. Developmental Medicine and Child Neurology, 40, 380-387.

Lyons, J. (1977) Semantics: 2. Cambridge: University Press. Lyytinen, H., Ahonen, T., Eklund, K., Guttorm, T., Laakso, M-L., Leinonen, S., Leppänen, P.,

Lyytinen, P., Poikkeus, A-M., Puolakanaho, A., Richardson, U. & Viholainen, H. (2001) Developmental Pathways of children with and without familial risk for dyslexia during the first years of life. Developmental Neuropsychology, 20, 535-554.

Lyytinen, P. (1988) Morphological test. Jyväskylä: Reports from the Department of Psychology, 298.

Lyytinen, P. (1999) Communicative and language development inventory. Finnish version. Jyväskylä: Niilo Mäki Institute.

McGee, T. & Kraus, N. (1996) Auditory development reflected by middle latency response. Ear & Hearing, 17, 419−429.

McKelvey, J.R., Lambert, R., Mottron, L. & Shevell, M. (1995) Right-hemisphere dysfunction in Asperger´s syndrome. Journal of Child Neurology, 10, 310-314.

Ment, L., Bada, H., Barnes, P., Grant, P., Hirtz, D., Papile, L., Pinto-Martin, J., Rivkin, M. & Slovis, T. (2002) Practice parameter: Neuroimaging of the neonate. Report of the quality standards subcommittee of the American Academy of Neurology and the Practice Committee of the Child Neurology Society. Neurology, 58, 1726-1738.

Ment, L., Vohr, B., Allan, W., Katz, K., Schneider, K., Westerveld, M., Duncan, C. & Makuch, R. (2003) Change in cognitive function over time in very low-birth-weight infants. JAMA, 289, 705-711.

88Menyuk, P., Liebergott, J., Schultz, M., Chesnick M. & Ferrier, L. (1991) Patterns of early lexical

and cognitive development in premature and full-term infants. Journal of Speech and Hearing Research, 34, 88-94.

Menyuk, P., Liebergott, J. & Schultz, M. (1995) Early language development in full-term and premature infants. Hillsdale: Lawrence Erlbaum Associates.

Merzenich, M., Jenkins, W., Johnston, P., Schreiner, C., Miller, S. & Tallal, P. (1996) Temporal processing deficits of language impaired children ameliorated by training. Science, 271, 77-80.

Molfese, D. (2000) Predicting dyslexia at 8 years of age using neonatal brain responses. Brain and Language, 72, 238-245.

Molfese, D. & Molfese, V. (1985) Electrophysiological indices of auditory discrimination in newborn infants. The basis for predicting later language performance? Infant Behavior and Development, 8, 197-211.

Molfese, D. & Molfese, V. (1997) Discrimination of language skills at five years of age using event-related potentials recorded at birth. Developmental Neuropsychology, 13, 135-156.

Näätänen, R. (1992) Attention and brain function. New Jersey: Lawrence Erlbaum Associates. Näätänen, R. (1999) Phoneme representations of the human brain as reflected by event-related

potentials. In: C. Barber, G. Celesia, I. Hashimoto & R. Kakigi (eds) Functional Neuroscience: Evoked Potentials and Magnetic Fields. (EEG Suppl. 4), Amsterdam: Elsevier, 170-173.

Näätänen, R. (2003) Mismatch negativity: clinical research and possible applications. International Journal of Psychophysiology, 48, 179-188.

Näätänen, R. & Alho, K. (1995a) Mismatch negativity – A unique measure of sensory processing in audition. International Journal of Neuroscience, 80, 317-337.

Näätänen, R. & Alho, K. (1995b) Event-related potentials in human selective attention research. In: F. Boller & J. Grafman (eds) Handbook of Neuropsychology. Elsevier Science, 75-104.

Näätänen, R. & Alho, K. (1997) Mismatch negativity – the measure for central sound representation accuracy. Audiology & Neuro-Otology, 2, 341-353.

Näätänen, R., Gaillard, A. & Mäntysalo, S. (1978) Early selective attention effect on evoked potential reinterpreted. Acta Psychologica, 42, 313-329.

Näätänen, R., Lehtokoski, A., Lennes, M., Cheour, M., Huotilainen, M., Iivonen, A., Vainio, M., Alku, P., Ilmoniemi, R., Luuk, A., Allik, J., Sinkkonen, J. & Alho, K. (1997b) Language specific phoneme representations revealed by electric and magnetic brain responses. Nature, 385, 432-434.

Näätänen, R. & Picton T. (1986) N2 and automatic versus controlled processes. Electroencephalography and Clinical Neurophysiology, Supplement, 38, 169−186.

Näätänen, R., Tervaniemi, M., Sussman, E., Paavilainen, P. & Winkler, I. (2001). ”Primitive Intelligence” in the auditory cortex. Trends in Neurosciences, 24, 283-288.

Näätänen, R. & Winkler, I. (1999) The concept of auditory stimulus representation in cognitive neuroscience. Psychological Bulletin, 125, 826-859.

Oller, D.K., Eilers, R., Steffens, M., Lynch, M. & Urbano, R. (1994) Speech-like vocalizations in infancy: an evaluation of potential risk factors. Journal of Child Language, 21, 33-58.

Paetau, R., Ahonen, A., Salonen, O. & Sams, M. (1995) Auditory evoked magnetic fields to tones and pseudowords in healthy children and adults. Journal of Clinical Neurophysiology, 12, 177−185.

Peterson, B., Vohr, B., Kane, M., Whalen, D., Schneider, K., Katz, K., Zhang, H., Duncan, C., Makuch, R., Gore, J. & Ment, L. (2002) A functional magnetic resonance imaging study of language processing and its cognitive correlates in prematurely born children. Pediatrics, 110, 1153-1162.

Peterson, B., Vohr, B., Staib, L., Cannistraci, C., Dolberg, A., Schneider, K., Katz, K., Westerveld, M., Sparrow, S., Anderson, A., Duncan, C., Makuch, R., Gore, J. & Ment, L. (2000) Regional brain volume abnormalities and long-term cognitive outcome in preterm infants. JAMA, 284, 1939-1947.

Picton, T., Hillyard, S., Krausz, H. & Galambos, R. (1974) Human auditory evoked potentials, I. Evaluation of components. Electroencephalography and Clinical Neurophysiology, 36, 179-190.

89Pihko, E., Leppänen, P., Eklund, T., Cheour, M., Guttorm, T. & Lyytinen, H. (1999) Cortical

responses of infants with and without a genetic risk for dyslexia: I. Age effects. NeroReport, 10, 901-905.

Ponton, C., Eggermont, J., Khosla, D., Kwong, B. & Don, M. (2002) Maturation of human central auditory system activity: separating auditory evoked potentials by dipole source modeling. Clinical Neurophysiology, 113, 407-420

Ponton, C., Eggermont, J., Kwong, B. & Don, M. (2000) Maturation of human central auditory system activity: evidence from multi-channel evoked potentials. Clinical Neurophysiology, 111, 220−236.

Pulvermüller, F. (2002) A brain perspective on language mechanism: from discrete neuronal ensembles to serial order. Progress in Neurobiology, 67, 85-111.

Pulvermüller, F., Kujala, T., Shtyrow, Y., Simola, J., Tiitinen, H., Alku, P., Alho, K., Martinkauppi, S., Ilmoniemi, R. & Näätänen, R. (2001) Memory traces for words as revealed by the mismatch negativity. NeuroImage, 14, 607-616.

Ramberg, C., Ehlers, S., Nydén, A., Johansson, M. & Gillberg, C. (1996) Language and pragmatic functions in school-age children on the autism spectrum. European Journal of Disorders of Communication, 31, 387-413.

Regan, D. (1989) Human brain electrophysiology. Evoked potentials and evoked magnetic fields in science and medicine. New York: Elsevier Science Publishing.

Reynell, J. & Huntley, M. (1985) The Reynell Developmental Language Scales. 2nd edition. Windsor: The NFER-NELSON.

Riitesuo, A. (2000) A preterm child grows. Focus on speech and language during the first two years. Doctoral dissertation. Jyväskylä: Jyväskylä University Printing House.

Rinne, T., Alho, K., Alku, P., Holi, M., Sinkkonen, J., Virtanen, J., Bertrand, O. & Näätänen, R. (1999) Analysis of speech sounds is left-hemisphere predominant at 100-150 ms after sound onset. NeuroReport, 10, 1113-1117.

Ross, G., Boatright, S., Auld, P. & Nass, R. (1996) Specific cognitive abilities in 2-year-old children with subependymal and mild intraventricular hemorrhage. Brain and Cognition, 32, 1-13.

Rourke, B. (1987) Syndrome of non-verbal learning disabilities: the final common pathway of white matter disease/dysfynction? The Clinical Neuropsychologist, 1, 209-234.

Rourke, B. (1988) The syndrome of nonverbal learning disabilities: developmental manifestations in neurological diseases disorder and dysfunction. The Clinical Neuropsychologist, 4, 293-330.

Rourke, B. (1995) Syndrome of nonverbal learning disabilities; Neurodevelopmental manifestations. New York: Guilford Press.

Rutherford, M., Baron-Cohen, S. & Wheelwright, S. (2002) Reading the mind in the voice: a study with normal adults and adults with Asperger syndrome and high functioning autism. Journal of Autism and Developmental Disorders, 32, 189-194.

Schatz, A., Weimer, A. & Trauner, D. (2002) Brief report: Attention Differences in Asperger syndrome. Journal of Autism and Developmental Disorders, 32, 333-336.

Schulte-Körne, G., Deimel, W., Bartling, J. & Remschmidt, H. (1998) Auditory processing and dyslexia: evidence for a specific speech processing deficit. NeuroReport, 9, 337-340.

Schultz, R., Romanski, L. & Tsatsanis, K. (2000) Neurofunctional Models of Autistic Disorder and Asperger Syndrome. In: A. Klin, F. Volkmar, F. & S. Sparrow (eds) Asperger Syndrome. New York: The Guilford Press, 172-209.

Seikel, J., King, D. & Drumright, D. (2000) Anatomy and Physiology for Speech, Language, and Hearing. San Diego: Singular Publishing Group.

Seri, S., Cerquiglini, A., Pisani, F. & Curatolo, P. (1999) Autism in tuberous sclerosis: evoked potential evidence for a deficit in auditory sensory processing. Clinical Neurophysiology, 110, 1825-1830.

Shafer, V., Morr, M., Kreuzer, J. & Kurzberg, D. (2000) Maturation of mismatch negativity in school-age children. Ear & Hearing, 21, 242-251.

Sharma, A., Kraus, N., McGee, T. & Nicol, T. (1997) Developmental changes in P1 and N1 central auditory responses elicited by consonant-vowel syllables. Electroencephalography and Clinical Neurophysiology, 104, 540-545.

90Shriberg, L., Paul, R., McSweeny, J., Klin, A., Cohen, D. & Volkmar, F. (2001) Speech and

prosody characteristics of adolescents and adults with high-functioning autism and Asperger syndrome. Journal of Speech, Language, and Hearing Research, 44, 1097-1115.

Shtyrov, Y., Kujala, T., Ahveninen, J., Tervaniemi, M., Alku, P., Ilmoniemi, R. & Näätänen, R. (1998) Background acoustic noise and the hemispheric lateralization of speech processing in the human brain: magnetic mismatch negativity study. Neuroscience Letters, 251, 141-144.

Shtyrov, Y., Kujala, T., Palva, S., Ilmoniemi, R. & Näätänen, R. (2000) Discrimination of speech and of complex nonspeech sounds of different temporal structure in the left and right cerebral hemispheres. NeuroImage, 12, 657-663.

Shtyrov, Y. & Pulvermüller, F. (2002) Neurophysiological evidence of memory traces for words in the human brain. NeuroReport, 13, 521-525.

Siegel, L. (1994) The long-term prognosis of pre-term infants. Human Nature, 5, 103-126. Singer, L., Siegel, A.C., Lewis, B., Hawkins, S., Yamashita, T. & Baley, J. (2001) Preschool

language outcomes of children with history of bronchopulmonary dysplasia and very low birth weight. Developmental and Behavioral Pediatrics, 22, 19-26.

Snow, D. (2000) The emotional basis of linguistic and nonlinguistic intonation: Implications for hemispheric specialization. Developmental Neuropsychology, 17, 1-28.

SPSS Advanced Models 10.0 (1999) Chigago: SPSS. Stapells, D. & Kurtzberg, D. (1991) Evoked potential assessment of auditory system integrity in

infants. Clinics in Perinatology, 18, 497-518. Starr, E., Szatmari, P., Bryson, S. & Zwaigenbaum, L. (2003) Stability and change among high-

functioning children with pervasive developmental disorders: A 2-year outcome study. Journal of Autism and Developmental Disorders, 33, 15-22.

Studdert-Kennedy, M. & Mody, M. (1995) Auditory processing deficits in the reading impaired: a critical review of the evidence. Psychonometric Bulletin Review, 2, 508-514.

Swan, D. & Goswami, U. (1997a) Picture naming deficits in developmental dyslexia: The phonological representation hypothesis. Brain and Language, 56, 334-353.

Swan, D. & Goswami, U. (1997b) Phonological awareness deficits in developmental dyslexia and the phonological representations hypothesis. Journal of Experimental Child Psychology, 66, 18-41.

Swingley, D. & Aslin, R. (2000) Spoken word recognition and lexical representation in very young children. Cognition, 76, 147-166.

Szatmari, P., Bryson, S., Boyle, M., Streiner, D. & Duku, E. (2003) Predictors of outcome among high functioning children with autism and Asperger syndrome. Journal of Child Psychology and Psychiatry, 44, 520-528.

Szatmari, P., Bryson, S., Streiner, D., Wilson, F., Archer, L. & Ryerse, C. (2000) Two-year outcome of preschool children with autism and Asperger´s syndrome. American Journal of Psychiatry, 157, 1980-1987.

Tallal, P. (1980) Auditory temporal perception, phonics, and reading disabilities in children. Brain and Language, 9, 182-198.

Tallal, P., Miller, S., Bedi, G., Byma, G., Wang, X., Nagarajan, S., Schreiner, D., Jenkins, W. & Merzenich, M. (1996) Language comprehension in language-learning impaired children improved with acoustically modified speech. Science, 271, 81-84.

Tervaniemi, M., Medvedev, S., Alho, K., Pakhomow, S., Roudas, M., van Zuijen, T. & Näätänen, R. (2000) Lateralized automatic auditory processing of phonetic versus musical information: a PET study. Human Brain Mapping, 10, 74-79.

Thatcher, R., Walker, R. & Giudice, S. (1987) Human cerebral hemispheres develop at different rates and ages. Science, 236, 1110-1113.

Tonnquist-Uhlen, I. (1996) Topography of auditory evoked long-latency potentials in children with severe language impairments: The P2 and N2 components. Ear & Hearing, 17, 314-326.

Uwer, R., Albrecht, R. & von Suchodoletz, W. (2002) Automatic processing of tones and speech stimuli in children with specific language impairment. Developmental Medicine and Child Neurology, 44, 527-532.

Valkama, M. (2001) Predictors of neurosensory disability in very low birth weight preterm infants. Doctoral dissertation. Acta Universitas Ouluensis, D Medica 617. Oulu: Oulu University Press.

91Valkama, A.M., Laitakari, K., Tolonen, E.U., Väyrynen, M., Vainionpää, L. & Koivisto, M.

(2000a) Prediction of permanent hearing loss in high-risk preterm infants at term age. European Journal of Pediatrics, 159, 459-464.

Valkama, M., Pääkkö, E., Vainionpää, L., Lanning, F.P., Ilkko, E. & Koivisto, M. (2000b) Magnetic resonance imaging at term and neuromotor outcome in preterm infants. Acta Paediatrica, 89, 348-355.

Veen, S., Ens-Dokkum, M., Schreuder, A., Verloove-Vanhorick, S., Brand, R. & Ruys, J. (1991) Impairments, disabilities, and handicaps of very preterm and very-low-birthweight infants at five years of age. Lancet, 338, 33-36.

Vohr, B., Allan, W., Westerveld, M., Schneider, K., Katz, K., Makuch, R. & Ment, L. (2003) School-age outcomes of very low birth weight infants in the indomethacin intraventricular hemorrhage prevention trial. Pediatrics, 111, 340-346.

Vohr, B., Garcia-Coll, C. & Oh, W. (1988) Language development of low-birthweight infants at two years. Developmental Medicine and Child Neurology, 30, 608-615.

Vohr, B., Garcia-Coll, C. & Oh, W. (1989) Language and neurodevelopmental outcome of low-birthweight infants at three years. Developmental Medicine and Child Neurology, 31, 582-590.

Volkmar, F. (1998) Autism and pervasive developmental disorders. Cambridge: Cambridge University Press.

Volkmar, F., Klin, A., Schultz, R., Rubin, E. & Bronen, R. (2000) Asperger´s disorder. American Journal of Psychiatry, 157, 262-267.

Werker, J. (2000) Becoming a native listener. In: D. Muir & A. Slater (eds) Infant Development. The Essential Readings. Oxford: Blackwell Publishers, 149-162.

Werker, J. & Desjardins, R. (2001) Listening to speech in the 1st year of life: Experiential influences on phoneme perception. In: M. Tomasello & E. Bates (eds) Language development. The essential readings. Oxford: The Blackwell Publishers, 26-33.

Whitaker. A., Feldman, J., van Rossem, R., Schonfeld, I., Pinto-Martin, J., Torre, C., Blumenthal, S. & Paneth, N. (1996) Neonatal cranial ultrasound abnormalities in low birth weight infants: relation to cognitive outcomes at six years of age. Pediatrics, 98, 719-729.

Williams, D. (1992) Nobody Nowhere. New York: Times Books. Winkler, I., Kujala, T., Tiitinen, H., Sivonen, P., Alku, P., Lehtokoski, A., Czigler, I., Csepe, V.,

Ilmoniemi, R. & Näätänen, R. (1999) Brain responses reveal the learning of foreign language phonemes. Psychophysiology, 36, 638-642.

Wolf, M. (1991) Naming, speech and reading: The contribution of the cognitive neurosciences. Reading Research Quarterly, 26, 123-141.

Wolf, M. & Obregon, M. (1991) Early naming deficits, developmental dyslexia, and a specific deficit hypothesis. Brain and Language, 42, 219-247.

Wolke, D., Ratschinski, G., Ohrt, B. & Riegel, K. (1994) The cognitive outcome of very preterm infants may be poorer than often reported: an empirical investigation of how methodological issues make a big difference. European Journal of Pediatrics, 12, 906-915.

Wolke, D. & Meyer, R. (1999) Cognitive status, language attainment, and prereading skills of 6-year-old very preterm children and their peers: the Bavarian longitudinal study. Developmental Medicine and Child Neurology, 41, 94-109.

Wood, N., Marlow, N., Costeloe, K., Chir, B., Gibson, A. & Wilkinson, A. (2000) Neurologic and developmental disability after extremely preterm birth. The New England Journal of Medicine, 343, 378-384.

Yabe, H., Tervaniemi, M., Reinikainen, K. & Näätänen, R. (1997) Temporal window of integration revealed by MMN to sound omission. NeuroReport, 8, 1971-1974.

Yliherva, A. (2002) Speech and language abilities of preterm and low birthweight children at 8 years of age in the northern Finland 1-year birth cohort for 1985-1986. Doctoral dissertation. Acta Universitas Ouluensis B44. Oulu: Oulu University Press.

Zatorre, R., Evans, A., Meyer, E. & Gjedde, A. (1992) Lateralization of phonetic and pitch discrimination in speech processing. Science, 256, 846-849.

Zatorre, R., Belin, P. & Penhune, V. (2002) Structure and function of auditory cortex: music and speech. Trends in Cognitive Neuroscience, 6, 37-46.

auditory event-related potentials as indices of language...

Documents