- Top of page
Aim: Asperger's disorder (Asperger syndrome, AS) and pervasive developmental disorder not otherwise specified (PDD-NOS) are different subtypes of mild pervasive developmental disorders (PDD).
Methods: Using the Japanese version of the Wechsler Intelligence Scale for Children–Third Edition (WISC-III), 28 AS children (mean age, 9.3 years, 24 male) were compared with 78 PDD-NOS children (mean age, 7.6 years, 64 male) with normal intelligence (IQ ≥ 85), using analysis of covariance (ancova) with the chronological age of a child as a covariate.
Results: Verbal IQ tended to be higher in the AS children than in the PDD-NOS children (mean raw scores, AS vs PDD-NOS: 103.9 vs 99.6; P < 0.10), although full-scale and performance IQ did not differ significantly. Compared with the PDD-NOS children, the AS children scored significantly higher on Freedom from Distractibility index (110.1 vs 104.5; P < 0.05) consisting of Arithmetic (11.0 vs 9.9, P = 0.04) and Digit Span (12.4 vs 11.6, P = 0.051), but tended to score lower on Coding (8.5 vs 9.8, P = 0.08).
Conclusion: The typical cognitive profile of PDD (i.e. low score on Comprehension and high score on Block Design) was shared by both groups, which may support the validity of the current diagnostic classification of PDD. Relatively better verbal ability in AS children seems to reflect their normal language acquisition in infancy, and strong numeric interest may produce the AS children's mathematical excellence over PDD-NOS children. A low score on Coding in AS children might reflect their extreme slowness, circumstantiality and/or drive for perfection.
PERVASIVE DEVELOPMENTAL DISORDERS (PDD) or autism spectrum disorders had been thought to be rare until the middle of the 1990s. According to Fombonne's comprehensive review of 23 epidemiological surveys of autism published between 1966 and 1998, the prevalence of PDD was estimated at 0.143%.1 For the last 10 years, however, several epidemiological studies have reported a much higher prevalence of PDD. For example, Baird et al. reported a prevalence of PDD of 1.161% (95% confidence interval [CI]: 0.904–1.418), and this was the first epidemiological study in the world to report a prevalence of PDD of >1%.2 Such a significant rise in the PDD prevalence has been possibly brought about by broadened recognition and early detection of mild PDD variants with normal intellectual functioning.
Asperger's disorder (Asperger syndrome, AS) is a mild PDD subtype characterized by qualitative impairment in social interaction and stereotyped/repetitive interests/behaviors that are not accompanied by significant delay in speech and cognitive development.3,4 Chakrabarti and Fombonne's epidemiological study in England found 25 cases of AS in 26 403 children (prevalence, 0.095%; 95%CI: 0.061–0.140), none of whom had mental retardation.5
Another mild PDD subtype is PDD not otherwise specified (PDD-NOS), which is currently thought of as comprising the majority of PDD.6 According to the Chakrabarti and Fombonne epidemiological study, PDD-NOS consisted of 51.6% of the total PDD and the prevalence was as high as 0.314% (95%CI: 0.250–0.390).5 In addition, according to their study, 88.8% of the PDD-NOS children were not mentally retarded (high-functioning). Given the high prevalence of PDD-NOS itself and the high incidence of high-functioning in it, studying the difference in cognitive ability between high-functioning PDD-NOS children and AS children is important to deepen the understanding of both conditions, which are still much under-studied compared to autistic disorder or childhood autism.
The Wechsler intelligence test is one of the most widely used scales to examine the cognitive ability of children/adults and has been translated into many languages including Japanese. Some previous studies have examined the Wechsler intelligence test profiles of AS7–12 in comparison with high-functioning autism (HFA). Most of these previous studies found a dominance of AS over HFA on any verbal subtest,7,8,10–12 although higher full-scale IQ (FIQ) in subjects with AS might exaggerate the results in some studies. Such verbal cognitive advantage of AS, however, might not emerge when compared with PDD-NOS, which usually involves milder autistic symptoms in childhood than autistic disorder.13
To our knowledge, only two studies examined the Wechsler intelligence test profile in PDD-NOS children.14,15 De Bruin et al. compared the Wechsler Intelligence Scale for Children–Revised (WISC-R) profile between AS and PDD-NOS, but they studied children with a broad IQ range (FIQ range, 48–152) and had unbalanced FIQ between the groups compared (autism, 83.6; AS, 102.9; PDD-NOS, 89.6).14 Because cognitive profiles specific to PDD seem to manifest more clearly in PDD children with normal intelligence than in those with a mental handicap, the study of such children is also important to clarify the difference in cognitive ability between AS and PDD-NOS subjects.
For all of the aforementioned reasons, we attempted to clarify the difference in the cognitive profile on the Japanese version of the Wechsler Intelligence Scale for Children–Third Edition (WISC-III) between normally intelligent (IQ ≥ 85) children with AS and those with PDD-NOS in the present study.
- Top of page
As shown in Table 1, VIQ tended to be higher in the AS children than in the PDD-NOS children (P < 0.10). FIQ and PIQ did not differ significantly between the two groups.
Table 1. WISC-III Scores in AS and PDD-NOS
|WISC-III||AS (n = 28) (Mean ± SD)||PDD-NOS (n = 78) (Mean ± SD)||F(1,103)|
|IQ|| || || |
| Full-scale||102.1 ± 9.3||101.0 ± 12.0||1.25|
| Verbal||103.9 ± 15.2||99.6 ± 16.5||3.13†|
| Performance||99.8 ± 10.3||102.4 ± 12.3||0.40|
|Index|| || || |
| Verbal comprehension||103.0 ± 15.7||99.6 ± 17.0||2.24|
| Perceptual organization||102.1 ± 10.2||103.1 ± 12.8||0.02|
| Freedom from distractibility||110.1 ± 16.9||104.5 ± 16.0||5.27*|
| Processing speed||93.0 ± 12.7||99.4 ± 13.2||2.48|
As to the indices, the AS children scored significantly higher on Freedom from Distractibility index than the PDD-NOS children. No significant intergroup difference was found on Verbal Comprehension, Perceptual Organization or Processing Speed indices.
As shown in Fig. 1, compared with the PDD-NOS children, the AS children scored significantly higher on Arithmetic (F = 4.24, d.f. = 1,103, P = 0.04) and tended to score higher on Digit Span (F = 3.89, d.f. = 1,103, P = 0.051). The differences on the other four verbal subtests were not significant, and both groups scored the lowest on Comprehension.
Figure 2 shows that compared with the PDD-NOS children, the AS children tended to score lower on Coding (F = 3.03, d.f. = 1,103, P = 0.08). The differences on the other five performance subtests were not significant, and the two groups scored the highest on Block Design.
- Top of page
To our knowledge, this is the first study that compared WISC-III profiles between AS and PDD-NOS children with normal intelligence. Both groups in the present study had low scores on Comprehension and high scores on Block Design, consistent with almost all of the previous studies of high-functioning autistic individuals.7–12,15,18–22 This cognitive profile, superior visual processing and inferior social intelligence,23 seems to be specific to autistic individuals, and the fact that it was shared by both of the AS and PDD-NOS children may support the validity of the current diagnostic classification of PDD.
Consistent with previous studies that found a dominance of AS over HFA on some verbal subtests,7,8,10–12 the verbal advantage of AS was sustained even when compared to PDD-NOS, a milder PDD variant than autistic disorder or childhood autism, confirming that a history of normal language acquisition in infancy for an AS subject could predict his/her better verbal intelligence in childhood.
Among the six verbal subtests, the difference between the two groups was clearest on Arithmetic and Digit Span, both of which constitute Freedom from Distractibility index and require careful attention to numbers. As to those subtests, most previous studies could not find a significant difference between AS and HFA.7–10,12 A possible explanation for the differences in the present study is that numeric interest frequently seen in autistic children might be strong enough in AS children to produce mathematical excellence over PDD-NOS children, who usually had generally milder autistic symptomatology than AS and HFA children.13
Compared with the PDD-NOS children, the AS children in the present study tended to score lower on Coding (constituting Processing Speed index). The trough regarding Coding has been one of the common cognitive characteristics of autistic individuals,20 although studies done before the current concept of AS became popular may not have clearly differentiated AS from other PDD subtypes. These findings and our previous finding that AS subjects scored significantly lower on Coding than HFA subjects12 suggest that a low score on Coding may be specific to AS. Although we have no convincing evidence for that, the low score on Coding in AS children might reflect their extreme slowness, circumstantiality and/or drive for perfection, as Ehlers et al. noted.8
The present results should be interpreted with caution due to some methodological limitations. The number of AS children was not necessarily large enough, although the whole sample size was larger than that used in similar previous studies, and the well-balanced FIQ between the two groups in the present study allows comparison of cognitive profiles between AS and PDD-NOS. The present findings on clinic-based samples may not be generalized because the children could have more severe autistic symptomatology than those in non-clinic situations. A more extensive study with a larger sample is necessary to replicate the present findings, and possible behavioral and/or biological correlates of those cognitive characteristics need to be studied further.
In conclusion, the present study clarified the differences in cognitive profile on the WISC-III between normally intelligent children with AS and those with PDD-NOS. The typical cognitive characteristics of PDD (i.e. strength in visual processing and weakness in social intelligence) were shared by both groups, which may support the validity of the current diagnostic classification of PDD. Relatively better verbal ability in AS children seems to reflect their normal language acquisition in infancy, and strong numeric interest may produce the AS mathematical excellence over PDD-NOS. A low score on Coding in AS children might reflect their extreme slowness, circumstantiality and/or drive for perfection.