Cognitive and symptom profiles in Asperger’s syndrome and high-functioning autism
Tomonori Koyama, MS, Department of Mental Health Administration, National Institute of Mental Health, National Center for Neurology and Psychiatry, 4-1-1 Ogawa-Higashi, Kodaira, Tokyo 187-8553, Japan. Email: email@example.com
Abstract Asperger syndrome (AS) and autistic disorder are two subtypes of pervasive developmental disorders (PDD), but there has been considerable debate over whether AS and autistic disorder without mental retardation (IQ ≥ 70), called high-functioning autism (HFA), are distinct conditions or not. The aim of the present paper was to clarify this issue through a comparison of cognitive function and autistic symptom profiles. Based on the DSM-IV and ICD-10 definitions of language acquisition, 36 age- and IQ-balanced subjects with AS (mean age, 12.8 years; mean full-scale IQ, 98.3) were compared with 37 subjects with HFA (mean age, 12.6 years; mean full-scale IQ, 94.6) on the Japanese version of the Wechsler Intelligence Scales and the Childhood Autism Rating Scale-Tokyo Version (CARS-TV). Compared with the HFA subjects, the AS subjects scored significantly higher on Verbal IQ, Vocabulary, and Comprehension, but scored significantly lower on Coding. Although the total CARS-TV score did not differ significantly between the two groups, AS subjects scored significantly lower (i.e. less abnormal) on Verbal communication and Non-verbal communication than did the HFA subjects. A history of normal language acquisition in early childhood could predict his/her better verbal ability in mid-childhood or later. Autistic cognitive characteristics shared by both AS and HFA subjects appear to support the validity of the current diagnostic classification of PDD.
Asperger’s syndrome (AS) and autistic disorder (autism) are two subtypes of pervasive developmental disorders (PDD) in which persons demonstrate qualitative impairments in social interaction and communication; and stereotyped/repetitive behaviors.1,2 The major diagnostic difference between the two disorders is that persons with AS have a history of normal language acquisition defined in DSM-IV and ICD-10 as expressing single words by age 2 and communicative phrases by age 3,1,2 although they usually demonstrate difficulties in pragmatic and/or non-verbal aspects of communication.
In contrast, most persons with autism have significant developmental delay of speech. However, some persons with autism are not mentally retarded (i.e. IQ ≥ 70) and this is termed high-functioning autism (HFA). According to Chakrabarti and Fombonne, the rate of HFA was 30.8% in children with autism,3 which is slightly higher than previously thought. Because almost all persons with AS are also not mentally retarded, there has been considerable debate over whether AS and HFA are distinct conditions or not.
Because psychometric assessment is a crucial part of the diagnostic practice of developmental disorders, some previous studies compared Wechsler intellectual profile between AS and HFA, although results have not been conclusive.4–8 Szatmari et al. compared 26 subjects with AS (mean age [age], 14.3 years; mean full-scale IQ [IQ], 86.6) according to the Wing criteria9 and 17 subjects with DSM-III10 HFA (age, 22.8 years; IQ, 82.2) on Wechsler Intelligence Scale for Children Revised (WISC-R)/Wechsler Adult Intelligence Scale Revised (WAIS-R).4 Although they found a dominance of AS over HFA on Similarities, they concluded that in terms of level of performance, the AS and HFA groups had significant impairments on both verbal and performance subtests and were quite similar to each other.4
Ehlers et al. compared 40 children with AS (age, 9.8 years; IQ, 102.5) who satisfied the Gillberg and Gillberg criteria11 and 40 children with DSM-III HFA (age, 9.9 years; IQ, 78.8) on WISC-R and found dominance of AS over HFA in many respects.5 By comparing subjects with full-scale IQ ranging from 71 to 105, they found that 21 AS children (age, 9.9 years; IQ, 89.7) scored significantly higher on Verbal IQ, Verbal Comprehension, Vocabulary and Comprehension than 20 HFA children (age, 9.6 years; IQ, 84.2). They also found troughs on Object Assembly and Coding in AS children and a peak on Block Design in HFA children.5
Manjiviona and Prior compared 35 AS children (age, 10.4 years; IQ, 102.6) who met modified ICD-102 criteria (i.e. ‘no delay in language development’ was judged not in early childhood but at the time of study in middle childhood) and 21 children with DSM-III-R12 HFA (age, 11.6 years; IQ, 88.6) on WISC-R/WAIS-R.6 By comparing overall differences on the verbal and performance subtests with Hotelling’s t-squared test, they found no significant difference between the two groups except for AS dominance over HFA on full-scale IQ. In addition, they regrouped their subjects into a no language delay group (n = 26; IQ, 102.1) and a language delay group (n = 30; IQ, 93.3) based on language delay as defined in DSM-IV1 and ICD-10,2 but they did not find any significant difference between the two groups. They concluded that the presence or absence of language delay did not appear to have any relationship to the neurocognitive profiles of those children.6
Ozonoff et al. compared 12 children with DSM-IV1 AS (age, 13.9 years; IQ, 115.6) and 23 children with DSM-IV HFA (age, 13.3 years; IQ, 108.9) on Wechsler Intelligence Scale for Children Third Edition (WISC-III)/Wechsler Adult Intelligence Scale Third Edition (WAIS-III).7 They found that AS children scored significantly higher on Comprehension than HFA children and that both groups had troughs on Coding.7
Ghaziuddin and Mountain-Kimchi compared 22 subjects with DSM-IV AS (age, 12.2 years; IQ, 103.3) and 12 subjects with DSM-IV HFA (age, 12.4 years; IQ, 92.2) on WISC-III/WAIS-R.8 They found that the AS subjects scored significantly higher on Verbal IQ, Information, Arithmetic and Vocabulary than the HFA subjects. They noted that higher scores in the AS subjects may be explained on the basis of their higher full-scale and verbal IQ.8
The inconclusiveness of previous studies might have resulted from various factors, such as the small sample size, difference in diagnostic criteria especially for AS, and lack of well-balanced ages and full-scale IQs between groups of subjects. By addressing those methodological issues, we attempted to clarify whether AS and HFA are distinct conditions or not through a comparison of cognitive function and autistic symptom profiles.
This study was conducted as a part of the comprehensive study of the prognosis of PDD approved by the ethics committee of the Tokyo University Graduate School of Medicine. The subjects were 36 AS subjects (mean age, 12.8 ± 6.4 years; range, 5.6–30.5 years; 33 male, three female) and 37 HFA subjects (mean age, 12.6 ± 5.8 years; range, 5.4–30.3 years; 33 male, four female), with no significant differences in chronological age or sex ratio. All subjects had full-scale IQ ≥70.
All subjects whose relevant data were derived from their clinical records were recruited from three clinics in and near Tokyo (Child Guidance Clinic affiliated with the National Welfare Foundation for Disabled Children, Nerima Welfare Center for the Mentally and Physically Handicapped, and Kawasaki Central District Center for Remedial Therapy for Handicapped Children), well known in Japan for their specialty in developmental disorders and related conditions. In each clinic, a clinical team consisting of experienced professionals (i.e. child psychiatrists, clinical psychologists, and pediatric neurologists), led by the same child psychiatrist (H.K.) having 30 years’ clinical experience of developmental disorders including PDD, diagnosed children by consensus according to relevant DSM or ICD schemes depending on the year of visit of children. Based on detailed clinical examination at first visit and follow-up observations as well as questionnaires on development and symptoms of children filled out by parents, the clinical teams diagnosed the subjects according to DSM-IV1 as follows.
If an individual’s behavior and history exhibited characteristics of PDD, the diagnosis of autistic disorder was considered first regardless of his/her intellectual ability. To meet DSM-IV criteria for autistic disorder, individuals must show six symptoms or more from the list of 12 including at least two from (i) the social domain; and at least one each from (ii) the communication domain and (iii) the restricted behaviors/interests domain (criterion A); delays or abnormal functioning in at least one area documented as present before age 3 (criterion B); and which are not better explained by Rett’s disorder or childhood disintegrative disorder (criterion C). If an individual’s history revealed normal language acquisition (i.e. expressing single words by age 2 and communicative phrases by age 3) and non-retarded cognitive and adaptive functioning, the diagnosis of AS was considered. To meet DSM-IV criteria for AS, individuals have to demonstrate at least two DSM-IV symptoms in (i) the social domain and at least one in (iii) the restricted behaviors/interests domain (criteria A,B); have a history of normal language acquisition (criterion D) as aforementioned; is functioning intellectually and adaptively in the non-retarded range (criterion E); and do not meet criteria for another specific PDD or schizophrenia (criterion F). If an individual did not meet the criteria for either disorder, he/she was excluded from the study.
Japanese version of Wechsler Intelligence Scales
The Japanese version of the Wechsler Intelligence Scales (i.e. WISC-R, WISC-III, WAIS-R) used in the present study to grasp cognitive function of the subjects were standardized for sufficient number of Japanese children or adults, with good reliability and validity.13–15 Wechsler Intelligence Scales consist of verbal subtests (Information, Similarities, Arithmetic, Vocabulary, Comprehension, and Digit Span) and performance subtests (Picture Completion, Coding/Digit Symbol, Picture Arrangement, Block Design, and Object Assembly) and yield verbal IQ (VIQ), performance IQ (PIQ), and full-scale IQ. Performance subtests of Symbol Search and Mazes were not used in the present study because they were not administered to every subject. In this study experienced psychologists administered WISC-R to 10 AS and 11 HFA, WISC-III to 21 AS and 20 HFA, and WAIS-R to five AS and six HFA subjects, with no significant differences in the rate of the diagnostic groups among the three scales.
Childhood Autism Rating Scale-Tokyo Version
The Childhood Autism Rating Scale-Tokyo Version (CARS-TV) with satisfactory reliability and validity16,17 used in the present study is the Japanese version of the CARS,18 one of the most widely used scales to evaluate the degree and profiles of autism in children. CARS-TV consists of 15 items (Relationships with people, Imitation, Affect, Use of body, Relation to non-human objects, Adaptation to environmental change, Visual responsiveness, Auditory responsiveness, Near receptor responsiveness, Anxiety reaction, Verbal communication, Non-verbal communication, Activity level, Intellectual functioning, and General impressions) and each item scores from 1.0 (normal) to 4.0 (severely abnormal) in units of 0.5. The total CARS-TV score is obtained by summing up each item score and thus ranges from 15.0 to 60.0. Based on behavior observation, experienced psychologists also administered CARS-TV to the present study’s subjects except for four each of the AS and HFA subjects, who could not be assessed for practical reasons.
Differences of mean Wechsler IQs and subtest scores between AS and HFA were tested by t-test. We calculated an effect size (ES) of the diagnosis on each score by dividing the difference of a mean score between the two groups by the pooled standard deviation of the score. According to Cohen, values of this type of ES of 0.2, 0.5, and 0.8 indicate small, moderate, and large effect sizes, respectively.19 Paired t-test was used to evaluate the discrepancy between VIQ and PIQ in AS and HFA.
For CARS-TV scores (total score and 15 item scores), t-test was used to determine the difference of a mean score between the two groups.
All statistical analyses were performed with SPSS 13.0 J for Windows and a significant level was set at P < 0.05 (two-tailed test).
Cognitive function profile
Table 1 shows that the AS subjects scored significantly higher on VIQ (moderate ES, 0.48) than the HFA subjects. Although the differences did not reach significance, VIQ tended to be higher (P < 0.10) than PIQ in the AS subjects (t = 1.78, d.f. = 35, P = 0.084), whereas PIQ was higher than VIQ in the HFA subjects (t = 1.34, d.f. = 36, P = 0.190).
Table 1. Wechsler profile in AS and HFA
| Digit span||11.8||(3.0)||11.9||(3.4)||0.19||0.05|
| Picture completion||9.3||(3.2)||9.0||(3.6)||0.32||0.07|
| Picture arrangement||9.3||(3.3)||8.5||(3.1)||0.99||0.23|
| Block design||11.5||(3.5)||12.0||(3.9)||0.61||0.14|
| Object assembly||9.6||(3.0)||10.0||(3.7)||0.59||0.13|
The AS subjects scored significantly higher on Vocabulary (large ES, 0.75) and Comprehension (large ES, 0.86), but scored significantly lower on Coding (moderate ES, 0.50) than the HFA subjects. The difference on the other eight subtests was not significant. Both groups scored low on Comprehension and Picture Arrangement and scored high on Digit Span and Block Design.
Autistic symptom profile
The total CARS-TV score did not differ significantly between the AS subjects (mean, 22.22 ± 3.57) and the HFA subjects (mean, 23.61 ± 3.42; t = 1.60, d.f. = 63, P = 0.114).
Among the 15 items of the CARS-TV, the AS subjects scored significantly lower (i.e. less abnormal) than the HFA subjects on Verbal communication (AS vs HFA: 1.31 vs 1.59; t = 3.34, d.f. = 63, P = 0.001) and Non-verbal communication (1.48 vs 1.67; t = 2.05, d.f. = 63, P = 0.044). In addition, the AS subjects tended to score lower (P < 0.10) on Relationships with people (1.42 vs 1.56; t = 1.90, d.f. = 63, P = 0.061) than the HFA subjects.
Consistent with previous studies,4–8 the present AS subjects had higher VIQ and excelled significantly over the HFA subjects on verbal subtests of Vocabulary and Comprehension. Moreover, the AS subjects had less significant communicative abnormality on CARS-TV compared with the HFA subjects, indicating that a history of normal language acquisition in early childhood could predict his/her better verbal ability in mid-childhood or later.
As found in many previous HFA studies,5,20,21 the present HFA subjects scored lowest on Comprehension and scored highest on Block Design, suggesting the unique intellectual structure in autism (i.e. superior in visuospatial ability but inferior in skills related to so-called social intelligence22). Even though milder compared with the HFA subjects, the present AS subjects also showed this autistic cognitive pattern. These cognitive characteristics shared by both groups appear to support the validity of the current diagnostic classification of PDD.
In the present study, the AS subjects scored significantly lower on Coding than the HFA subjects, consistent with the Ehlers et al. findings that children with AS had troughs on Coding.5 In other previous studies, however, the low score on Coding was found in both AS and HFA,7,8 or even lower in HFA.4,6 As Ghaziuddin and Mountain-Kimchi noted, higher score of AS in such previous study may be explained by their higher IQs.8 The present HFA subjects did not show such weakness, but according to the Siegel et al. summary (table 1; pp. 390–391), the low score of Coding was one of the common characteristics in persons with HFA.21 Because the concept of AS became popular only after the publication of ICD-10 and DSM-IV in the middle 1990s, studies before then did not necessarily differentiate AS from HFA. In other words, the subjects with HFA in some previous studies might involve subjects with the current AS. Future studies using rigorous diagnostic criteria to differentiate between AS and HFA in terms of the presence or absence of delay in language acquisition might clarify this issue. As Ehlers et al. noted, the low score of AS on Coding might reflect their distractibility, the extreme slowness, circumstantiality and/or drive for perfection.5
VIQ-PIQ discrepancy was one of the most important topics in this field,8 but we could not simply compare our findings with those in previous studies due to the diagnostic reason mentioned here. Although the differences did not reach significance, the present subjects had the familiar VIQ-PIQ pattern, namely, the AS subjects had higher VIQ than PIQ while the HFA subjects had higher PIQ than VIQ. However, recent studies based on DSM-IV criteria do not necessarily support this intellectual pattern.8,23 This discordance, which might result from the differences in ages of subjects and/or the definition of high-functioning, needs further studies for clarification.
Although both groups showed similar degree of autism on CARS-TV, the HFA group was slightly more abnormal in social interaction and communication. Compared with the Kurita finding that children with AS scored significantly lower than children with high-functioning ICD-10 atypical autism on total and four-item scores (i.e. imitation, visual responsiveness, auditory responsiveness and non-verbal communication),24 the differences between AS and HFA were more apparent in core autistic items. Nonetheless, it may be difficult to differentiate between adolescent AS and HFA subjects solely based on autistic symptoms without any information of his/her development in infancy and early childhood.
The results of the present study should be interpreted carefully. We unified results obtained from the three different versions of the Wechsler intelligence scales to maximize the sample size, because they have basically the same structure, and small sample size was often the case in high-functioning PDD studies. Although the present study was relatively larger in sample size than previous studies4–8 and used demographic and intellectual balancing, which might strengthen our findings, a more extensive and detailed study is needed to test the present findings. The present clinic-based study may have involved more severely impaired persons than a study based on a sample from the general population. Such bias if any in the present sample, however, does not seem inappropriate to obtain clinical suggestions.
In conclusion, compared with HFA subjects, the AS subjects were superior in verbal functioning but were inferior in processing speed, as indicated by Coding. A history of normal language acquisition in early childhood could predict his/her better verbal ability in mid-childhood or later. Autistic cognitive characteristics shared by both AS and HFA subjects appear to support the validity of the current diagnostic classification of PDD.
This study was supported in part by the Research Grant for Nervous and Mental Disorders from the Ministry of Health and Welfare, Japan. We would like to thank Ms Yoko Hayashi, Ms Keiko Shimoyamada, Ms Mika Tobari, and Mr Hiromi Ishida for their help in data collection.