Head circumference in autism, Asperger syndrome, and ADHD: a comparative study

Transcription

1 Head circumference in autism, Asperger syndrome, and ADHD: a comparative study Christopher Gillberg* MD PhD; Department of Psychiatry, St George s Hospital Medical School, London; Department of Child and Adolescent Psychiatry, Göteborg University; Linda de Souza MB, Department of Child and Adolescent Psychiatry, Göteborg University, Göteborg, Sweden. *Correspondence to first author at Department of Child and Adolescent Psychiatry, Göteborg University, Kungsgatan 12, SE , Göteborg, Sweden. ann.nordstrom@pediat.gu.se This study was undertaken to test the hypothesis that children with autistic spectrum disorders often have macrocephalus, and that those without comorbid learning disability* are most frequently affected. Fifty consecutive children with Asperger syndrome (45 males, five females; mean age 9 years, range 1 year 6 months to 16 years) without indications of underlying medical disorders were matched for birth year and sex with 50 children (45 males, five females; mean age 6 years 4 months, range 1 year 4 months to 13 years 11 months) who met criteria for autistic disorder (a lower-functioning disorder within the autism spectrum) and with 50 children (45 males, five females; mean age 8 years 4 months, range 1 year 6 months to 15 years 5 months) who met criteria for attentiondeficit hyperactivity disorder. Birth and neuropsychiatric follow-up records were examined and data relating to occipitofrontal circumference, weight, and height were detailed. The group with Asperger syndrome included a subset of individuals with macrocephalus recorded both at birth and at follow-up after the first year of life. Another subgroup developed macrocephalus during early childhood. Autistic spectrum disorders include a subgroup with macrocephalus characterized by a relatively high level of functioning and a clinical presentation most often consistent with a diagnosis of Asperger syndrome. *UK usage. North American usage: mental retardation. Autistic disorder remains elusive with regard to basic pathophysiology, in spite of strong indirect evidence of neurobiological aetiologies (Gillberg and Coleman 2000). Recent studies of Asperger syndrome (Klin 1995, Gillberg 1998) have suggested the existence of a wider category of autism spectrum disorders (Wing 1996). In this spectrum, autistic disorder belongs in the lower and middle sections, whereas Asperger syndrome is conceptualized as a high-functioning variant. Autistic disorder is usually associated with marked intellectual impairment. Asperger syndrome, on the other hand, is diagnosed almost exclusively in individuals with normal or high levels of intelligence. Asperger syndrome/high-functioning autism is believed by many authorities to represent the core syndrome in the autistic spectrum: its symptoms showing in crystallized form, not confounded by the effects of concomitant learning disability. Macrocephalus has been proposed to be overrepresented in autism (Bailey et al. 1993, Woodhouse et al. 1996, Lainhart et al. 1997, Fombonne et al. 1999, Fidler et al. 2000, Miles et al. 2000) with several studies indicating rates of head enlargement of 14 to 37%. However, one study (Ghaziuddin et al. 1999) showed a high rate of macrocephalus in attentiondeficit hyperactivity disorder (ADHD) casting doubt on the specificity of the association with autism. One study (Davidovitch et al. 1996) found macrocephalus in autism to be strongly related to increased height and weight, possibly indicating a generalized increased growth rate in autism. One epidemiological study (Rodier et al. 1997) failed to confirm a high rate of abnormal head circumference in autism as compared with other kinds of developmental disabilities. However, a few studies have found evidence for microcephaly being overrepresented in autism (Fombonne et al. 1999, Miles et al. 2000), and this has been taken as a possible explanation for the fact that a link between macrocephalus and autism has not been noted until fairly recently. Theoretically, a link between autism and macrocephalus could be accounted for by overproduction (and crowding?) of nerve cells, synapse pruning failure during development of the CNS, or other factors affecting growth. Indirect support for such hypotheses has been found in autopsy studies showing too many cells too small (Bauman 1991) in certain brain areas, and in neurochemical research reporting indirect evidence of high synapse turnover in autism (Ahlsén et al. 1993, Nordin et al. 1998). Interestingly, in this context Stevenson and coworkers (1997) suggested that macrocephalus in autism may not be present at birth but develops during early childhood. Progressive macrocephalus has also been noted in Cole Hughes syndrome, which comprises a subgroup of individuals with autistic features (Naqvi et al. 2000). On the other hand, Sotos syndrome, another behavioural phenotype syndrome with a high rate of associated autism spectrum disorder (Zappella 1990), often presents with macrocephalus from birth. The present study set out to examine (1) whether there is a high rate of macrocephalus in children with autistic spectrum disorders, and (2) whether children with the high-functioning variant of autism, not compounded by comorbid learning disability, have a particularly high rate of macrocephalus. In order to examine whether or not a possible link with macrocephalus is specific to autistic spectrum disorders, a contrast group of children with ADHD was included for comparison. Head size was compared with weight and height in order to 296 Developmental Medicine & Child Neurology 2002, 44:

2 avoid overdiagnosing tall children and overweight children, and underdiagnosing those who were small and thin. Method The study included 150 child psychiatric patients with Asperger syndrome, autistic disorder, or ADHD. All had occipitofrontal circumference (OFC) measured at birth and OFC data had been recorded for 88% of the sample at or after 16 months of age. OFC data were compared across diagnostic groups and against standardized age- and sex-matched normative values. PARTICIPANTS Asperger syndrome Fifty consecutive patients (45 males, five females) born from 1982 to 1991 with Asperger syndrome not associated with a known medical disorder, and attending the state Child Neuropsychiatry Department in Göteborg, Sweden were selected. All met criteria for Asperger syndrome outlined by Gillberg (Gillberg et al. 1989, Gillberg 1991), Szatmari and colleagues (1989), and the DSM-IV (American Psychiatric Association 1994), except that normal early development information as required by the DSM-IV had not been or could not be documented in all patients. This was not because parents were not interviewed in depth about early development but because they could not remember enough detail to validate this specific criteria. Several authors have testified to the difficulties of eliciting appropriate information in this domain (Miller and Ozonoff 1997, Leekam et al. 2000). Mean age at follow-up OFC measurement for the group was 9 years (range 1 year 6 months to 16 years). None of the individuals in the group with Asperger syndrome met the full diagnostic criteria for autistic disorder (but all met four or more symptom criteria for that disorder). All children with Asperger syndrome who were tested had a total IQ of 81 or above. Autistic disorder Fifty patients diagnosed with autistic disorder, according to DSM-IV criteria and not associated with a known medical disorder were selected from the register of the Child Neuropsychiatry Department. They were matched with the Asperger syndrome group for birth year and sex. Mean age at follow-up OFC measurement was 6 years 4 months (range 1 year 4 months to 13 years 11 months). All but two of the children with autistic disorder, who had their IQ tested had a total IQ of 80 or below. The other two had Performance IQs above 110, but Verbal IQs of less than 80. The non-tested children were all judged to have severe learning disability. ADHD Fifty patients with ADHD according to DSM-IV criteria were selected from the register of the Child Neuropsychiatry Department. They were matched with the Asperger syndrome group for birth year and sex. All individuals in this group also met DSM-IV criteria for developmental coordination disorder. Mean age at follow-up OFC measurement was 8 years 4 months (range 1 year 6 months to 15 years 5 months). PROCEDURE All three diagnostic groups had been clinically examined in great detail by a team of experts in child neuropsychiatry and child neuropsychology. Diagnostic investigations included the Childhood Autism Rating Scale (Schopler et al. 1980) or the Autism Diagnostic Interview Revised (Lord et al. 1994) in the Asperger syndrome and autistic disorder groups. The DSM-III-R/DSM-IV ADHD, autistic disorder, and Asperger syndrome checklists and questionnaires (Conners 1967, Landgren et al. 1996) were used in the group with ADHD. Standardized IQ tests, usually the WISC-R (Wechsler 1974), had been used to estimate patients level of intellectual functioning. However, about 35% of the autistic disorder group had a very low cognitive age and had not been given the WISC-R. Their IQ was estimated on the basis of developmental tests or the interview from the Vineland Adaptive Scales (Sparrow et al. 1984). All the children with Asperger syndrome and autistic disorder had received a comprehensive medical diagnostic assessment comprising karyotype, DNAtesting, and neuroimaging. None of the children had any of the identifiable medical disorders sometimes known to be associated with autism. Mean IQ in those children who had been given formal IQ tests was: 97 in the Asperger syndrome group (n=47; three parents did not want their child to be tested for IQ); 69 in the group with autistic disorder (n=32); and 90 in the ADHD group (n=46; four parents in the group refused testing of their child s IQ). It should be noted again that the group with autistic disorder included a subgroup with severe learning disability (n=18), who were not included among those who had formal IQ tests. If the subgroup with severe learning disability were to be included and the mean IQ in this subgroup estimated to be 25, the estimated mean IQ of individuals with autistic disorder would be 53. Occipitofrontal circumference OFC was detailed at birth and in 88% of patients on at least one other occasion after the age of 16 months. Data for OFC at birth was obtained from the standardized records for newborn infants that have been used in Sweden from the 1970s (Karlberg et al. 1976). Such measurements taken during routine newborn examination procedures have provided the databases for the general population normative values currently in use. Data for OFC after 16 months of age were taken from the records of the Child Neuropsychiatry Department. OFC measurement as well as measurement of height and weight are part of the routine medical examination at the Department. However, there were patients (12% of the total sample of 150), for whom no OFC follow-up data were available. Doctors and nurses are trained to take OFC measurements in a uniform fashion. Over 90% of all measurements were performed by one of two nurses who had worked in the Department throughout the time the participants had been assessed. Height and weight Height and weight were obtained from birth records and from the child s neuropsychiatry records after 16 months of age. Measurements performed at the Child Neuropsychiatry Clinic were carried out by one of the two highly experienced nurses. Swedish and Norwegian normative data There are sex and age normative values for weight and height from birth to 15 years for Swedish children (Karlberg et al. Head Circumference in Autism, Asperger Syndrome, and ADHD Christopher Gillberg and Linda de Souza 297

3 1976). Swedish normative OFC values are available only for children aged 0 to 2 years (Karlberg et al. 1988). For children older than 2 years, sex and age normative values for Norwegian children were used instead (Knudtzon et al. 1988). ANALYSIS OF DATA As follow-up data were not collected when the children were of the same age, a standard deviation (SD) from norm procedure was employed. To allow comparison, each OFC, weight, and height measurement (at birth and after age 16 months) was plotted onto an enlarged (400%) growth chart and deviance from the population mean (+ or ) was shown in tenths of SDs. Genuine macrocephalus was considered to be present in children whose OFC surpassed the age and sex normative data by 2SDs or more and whose height and weight were both 1SD or more under that expected on the basis of the OFC. It was argued that genuine macrocephalus would not be an appropriate label for children with large heads proportional to overall body size. Relative macrocephalus was considered to be present (1) in children whose OFC surpassed the age and sex normative values by 2SDs or more and whose height or weight, but not both, was 1SD or more under that expected on the basis of the OFC, and (2) in patients whose OFC was above the age and sex normative values and whose height was 1.5SDs or more below the level expected on the basis of the OFC. Microcephaly was considered to be present in any child whose OFC after age 16 months was under 2SDs from the normative values of the general population. Apgar scores An Apgar score of 5 or under at 5 minutes (range 0 to 10, 10 optimal) as recorded in the birth records was used as an index of perinatal asphyxia. STATISTICAL METHODS χ 2 tests were used for comparison of group rates. Group means were compared using Fisher s non-parametric pertamution test (Bradley 1968). Spearman s rank correlations were also used in some instances. Statistical significance was set at p<0.05. Results MEAN OFC All three groups (Asperger syndrome, autistic disorder, and ADHD) had mean OFCs at birth that were larger than normative values (Table I). The Asperger syndrome group mean was significantly greater than that of the group with autistic disorder (35.2 cm versus 34.7 cm, corresponding to SDs of 1.02 and 0.63 respectively, p<0.01). All three groups also had mean OFCs above the age and sex norms when examined after age 16 months. The Asperger syndrome group mean was significantly greater than that of the ADHD group (SD 0.48 versus SD 0.32, p<0.05). RATE OF MACROCEPHALUS Genuine macrocephalus was present in 11 of 43 of the Asperger syndrome group at birth (see Table I) and in nine of 43 after the age of 16 months. Corresponding rates in the group with autistic disorder were four of 42 (p<0.01) and 4 (ns), and in the ADHD group, seven of 47 (ns) and three of 47 (ns) respectively. In the Asperger syndrome group, 10 of 43 childrenshowed genuine or relative macrocephalus both at birth and after age 16 months. This proportion tended to be greater than that in the ADHD group (4 of 47, p<0.06). The corresponding proportion in the autistic disorder group was four of 42 children. A majority of those with Asperger syndrome showing macrocephalus (genuine or relative) after age 16 months had been macrocephalic at birth also (10 of 13). The case was rather different in the group with autistic disorder (4 of 11) and the ADHD group (4 of 8). Conversely, 10 of 24 of those with Asperger syndrome with macrocephalus at birth (6 11 of those with genuine macrocephalus) were still macrocephalic after age 16 months. The corresponding rate in the group with autistic disorder was four out of 17 (2 out of 4 of those with genuine macrocephalus). RATE OF MICROCEPHALY Four children in the group with Asperger syndrome, one in the group with autistic disorder, and two in the group with ADHD had microcephaly. Table I: Macrocephalus at birth and after age 16 months Finding Asperger syndrome Autistic disorder ADHD (n=43/50) (n=42/50) (n=47/50) Occipitofrontal circumference at birth, mean (SD), cm 35.2 (1.8) 34.7 (1.9) 35.1 (1.5) a Genuine macrocephalus at birth, n (%) 11 (22) b 4 (9.5) 7 (14) Relative macrocephalus at birth, n (%) 13(26) 13 (26) 16 (32) Genuine macrocephalus after age 16 months, n (%) 9 (21) 4 (9.5) 3 (6) at birth also, n (%) 6 (14) c 2 (5) 0 (0) Relative macrocephalus after age 16 months, n (%) 4 (9) 7(17) 4 (9) at birth also, n (%) 4 (9) 2 (5) 4 (9) Genuine and relative macrocephalus after age 16 months, n (%) 13 (30) 11 (26) 7 (15) at birth also, n (%) 10 (23) 4(9.5) 4 (9) a OFC birth normative value (Karlberg et al. 1988) of 34 (SD 1.6). b p<0.01 Asperger syndrome versus autistic disorder. c p<0.05 Asperger syndrome versus ADHD. 298 Developmental Medicine & Child Neurology 2002, 44:

4 CORRELATION OF OFC WITH WEIGHT Significant correlations were found between OFC and birthweight in all groups (Asperger syndrome, autistic disorder, ADHD; rho=0.58, 0.66, and 0.78 respectively, p<0.001). OFC AND HEIGHT There was a significant correlation between OFC and height at birth in all groups (Asperger syndrome, autistic disorder, and ADHD; (rho=0.49, 0.53, and 0.67, p<0.001). Correlations between OFC and height at birth were also significant in the autistic disorder and Asperger syndrome groups (rho=0.53 and 0.49, p<0.001). OFC AT BIRTH AND AFTER AGE 16 MONTHS There were significant correlations between OFC at birth and after age 16 months in the group with Asperger syndrome (rho=0.54, p<0.001), but no significant correlations were found in the autistic disorder and ADHD groups. OFC AND IQ There was no significant correlation of OFC and IQ across or within diagnostic groups. Mean OFC at birth and at followup did not differ significantly across the IQ-tested autistic disorder subgroup and the subgroup in which no formal testing had been achieved. OFC AND PERINATAL ASPHYXIA Nine children (6% of the total sample) had perinatal asphyxia with Apgar scores of 5 or under at 5 minutes after birth. Two of these both in the Asperger syndrome group had microcephaly. Otherwise, there was no clear association between low Apgar scores and OFC. Discussion On the basis of the results obtained in this clinical study, it would appear that one in four to one in five of children with Asperger syndrome, but only one in 10 of those with autistic disorder, have macrocephalus ascertained after age 16 months, which had been present from birth. The findings corroborate those of previous studies, even though it seems that the macrocephalus is rather more typical of the highest functioning variants of autistic spectrum disorders, currently referred to as Asperger syndrome, and not a striking feature of classic moderate- to low-functioning autistic disorder. However, contrary to some previous findings (e.g. Lainhart et al. 1997, Stevenson et al. 1997), there were indications in our study that about half of those with Asperger syndrome (and 50 to 75% of those with autistic disorder) who were macrocephalic after 16 months of age had developmental macrocephalus, i.e. the large head was not present from birth. Nevertheless, a substantial proportion of all the patients with autistic spectrum disorders in our study also had developmental macrocephalus. It is possible that a small fraction of the overrepresentation of macrocephalus encountered in the Asperger syndrome could have been caused by a few individuals in that group being somewhat older at latest OFC measurement than in the other two groups. It is important to note that the rate of macrocephalus at birth was very high in all diagnostic groups, that it was lower at follow-up, but that new cases emerged (particularly in the ADHD and autistic disorder groups) from birth to follow-up. Thus, having access only to one OFC measurement in a clinical setting may yield a misleading picture. Combining the follow-up measurement with that of the newborn period led to the identification of a considerable subgroup with consistent macrocephalus in the Asperger syndrome category only. As in the study by Ghaziuddin and coworkers (1999), at least on a cross-sectional basis, ADHD and not just autistic spectrum disorders seems to be quite often associated with enlarged head size, although this finding may not be stable over time. However, it is not absolutely certain that the age and sex normative values can be regarded as truly reflecting growth parameters in the background population of children included in the present study. No control group without neuropsychiatric disorder was included, and so the results should be interpreted with caution. Nevertheless, the finding that the group with Asperger syndrome differed significantly from the other groups on important parameters suggests that this group does have a high rate of macrocephalus, possibly of two different types: one which is present from birth and one which emerges only after some years of development. The present study does not address the issue of causality as regards macrocephalus. Our index of perinatal adversity (low Apgar scores at 5 minutes after birth) was not (negatively or positively) correlated with head size, except possibly in a few instances of microcephaly. It is possible that a familial disposition to large head size may exist, but this is speculative at the present time and needs to be studied separately in future research. Conclusion Findings of this study suggest that on a group-wise basis, high-functioning autistic spectrum disorder may differ from low-functioning forms of the disorder both with regard to rate and type of macrocephalus. This could be taken to support separate neurobiological mechanisms in a subgroup of high-functioning patients as compared with low-functioning patients. This would lend some support for the notion of Asperger syndrome and autistic disorder being different conditions that should continue to be treated separately in research. Nevertheless, the majority of those with Asperger syndrome and those with autistic disorder did not differ with regard to head size. This underscores the need to look for an aetiological subgrouping within the autistic spectrum rather than simply grouping all patients with high-functioning autistic disorder under one label and patients with the low-functioning variant under another. Accepted for publication 22nd October References Ahlsén G, Rosengren L, Belfrage M, Palm A, Haglid K, Hamberger A, Gillberg C. (1993) Glial fibrillary acidic protein in the cerebrospinal fluid of children with autism and other neuropsychiatric disorders. Biological Psychiatry 33: American Psychiatric Association. (1994) Diagnostic and Statistical Manual of Mental Disorders. 4th edn (DSM-IV). Washington, DC: American Psychiatric Association. Bailey A, Luthert P, Bolton P, Le Couteur A, Rutter M, Harding B. (1993) Autism and megalencephaly. Lancet 341: (Letter). Bauman ML. (1991) Microscopic neuroanatomic abnormalities in autism. Pediatrics 87: Bradley JV. (1968) Distribution-free Statistical Tests. Englewood Cliffs, NJ: Prentice-Hall. Head Circumference in Autism, Asperger Syndrome, and ADHD Christopher Gillberg and Linda de Souza 299

5 Conners CK. (1967) The syndrome of minimal brain dysfunction: psychological aspects. Pediatric Clinics of North America 14: Davidovitch M, Patterson B, Gartside P. (1996) Head circumference measurements in children with autism. Journal of Child Neurology 11: Fidler DJ, Bailey JN, Smalley SL. (2000) Macrocephaly in autism and other pervasive developmental disorders. Developmental Medicine & Child Neurology 42: Fombonne E, Roge B, Claverie J, Courty S, Fremolle J. (1999) Microcephaly and macrocephaly in autism. Journal of Autism and Developmental Disorders 29: Ghaziuddin M, Zaccagnini J, Tsai L, Elardo S. (1999) Is megalencephaly specific to autism? Journal of Intellectual Disability Research 43: Gillberg C. (1991) Clinical and neurobiological aspects of Asperger syndrome. In: Frith U, editor. Autism and Asperger Syndrome. Cambridge: Cambridge University Press. p Gillberg C. (1998) Asperger syndrome and high-functioning autism. British Journal of Psychiatry 72: Gillberg C, Coleman M. (2000) The Biology of the Autistic Syndromes. 3rd edn. Clinics in Developmental Medicine No. 153/4 London: Mac Keith Press. Gillberg IC, Gillberg C. (1989) Asperger syndrome some epidemiological considerations: a research note. Journal of Child Psychology and Psychiatry 30: Karlberg P, Taranger J, Engström I, Karlberg J, Landström T, Lichtenstein H, Lindström B, Svennberg-Redegren I. (1976) Physical growth from birth to 16 years and longitudinal outcome of the study during the same age period. Acta Paediatrica Scandinavica 258: (Suppl.): Karlberg P, Kristiansson B. (1988) A revision of growth charts of infants is justified. The growth pattern, not weight or length, should be the basis. Läkartidningen 85: (In Swedish). Klin A, Volkmar FR, Sparrow SS, Cicchetti DV, Rourke BP. (1995) Validity and neuropsychological characterization of Asperger syndrome: convergence with non-verbal learning disabilities syndrome. Journal of Child Psychology and Psychiatry 36: Knudtzon J, Waaler PE, Skjaerven R, Solberg LK, Steen J. (1988) New Norwegian percentage charts for height, weight, and head circumference for age groups 0 17 years. Tidsskr Nor Laegeforen 108: (In Norwegian). Lainhart JE, Piven J, Wzorek M, Landa R, Santangelo SL, Coon H, Folstein SE. (1997) Macrocephaly in children and adults with autism. Journal of the American Academy of Child and Adolescent Psychiatry 36: Landgren M, Pettersson R, Kjellman B, Gillberg C. (1996) ADHD, DAMP, and other neurodevelopmental/neuropsychiatric disorders in six-year-old children. Epidemiology and comorbidity. Developmental Medicine &Child Neurology 38: Leekam S, Libby S, Wing L, Gould J, Gillberg C. (2000) Comparison of ICD-10 and Gillberg s criteria for Asperger syndrome. Autism 4: Lord C, Rutter M, Le Couteur A. (1994) Autism Diagnostic Interview-Revised: a revised version of a diagnostic interview for caregivers of individuals with possible pervasive developmental disorders. Journal of Autism and Developmental Disorders 24: Miles JH, Hadden LL, Takahashi TN, Hillman RE. (2000) Head circumference is an independent clinical finding associated with autism. American Journal of Medical Genetics 11: Miller JN, Ozonoff S. (1997) Did Asperger s cases have Asperger disorder? A research note. Journal of Child Psychology and Psychiatry 38: Naqvi S, Cole T, Graham Jr. JM. (2000) Cole-Hughes macrocephaly syndrome and associated autistic manifestations. American Journal of Medical Genetics 11: Nordin V, Lekman A, Johansson M, Fredman P, Gillberg C. (1998) Gangliosides in cerebrospinal fluid in children with autism spectrum disorders. Developmental Medicine &Child Neurology 40: Rodier PM, Bryson SE, Welch JP. (1997) Minor malformations and physical measurements in autism: data from Nova Scotia. Teratology 55: Schopler E, Reichler RJ, DeVellis RF, Daly K. (1980) Toward objective classification of childhood autism: Childhood Autism Rating Scale (CARS). Journal of Autism and Developmental Disorders 10: Sparrow SS, Balla DA, Cicchetti DV. (1984) Vineland Adaptive Behavior Scales. Interview Edition, Vol. 1, Survey Form Manual. Circle Pines, MN: American Guidance Service. Stevenson RE, Schroer, Skinner C, Fender D, Simensen RJ. (1997) Autism and macrocephaly. Lancet 14: Szatmari P, Brenner R, Nagy J. (1989) Asperger s syndrome: a review of clinical features. Canadian Journal of Psychiatry 34: Wechsler D. (1974). Manual of the Wechsler Intelligence Scale for Children Revised. New York: Psychological Corporation. Wing L. (1996) Autism spectrum disorders. No evidence for or against an increase in prevalence. British Medical Journal 312: Woodhouse W, Bailey A, Rutter M, Bolton P, Baird G, Le Couteur A. (1996) Head circumference in autism and other pervasive developmental disorders. Journal of Child Psychology and Psychiatry 37: Zappella M. (1990) Autistic features in children affected by cerebral gigantism. Brain Dysfunction 3: Developmental Medicine & Child Neurology 2002, 44: