Cognitive Profiles of Adolescents with Parent-Reported Extraordinary Skills and Relative Strengths

Background: Parent-report studies suggest 26-63% of individuals with ASD have “savant” skills - extraordinary skills (ES) that stand out relative to the general population (e.g., Meilleur et al., 2015; Howlin et al., 2009). Studies suggest that only a small subset of individuals with a parent-reported ES also show evidence of an exceptional cognitive skill on an IQ test (Howlin et al., 2009; Bolte & Poustka, 2004); however, research in this area is limited. Moreover, few studies have explored relative strengths (RS; i.e., strengths exceeding the individual’s own general ability level, but not general population norms). Understanding the correspondence between parent-reported and directly assessed skills is important to inform future studies of strengths in ASD.

Objectives: To investigate whether autistic adolescents with and without parent-reported ES or RS show differences in cognitive profiles. We expect that adolescents with ES will exhibit significantly higher scores on the cognitive domain most relevant to that skill (see Table 2) than peers without ES. In addition, we expect that adolescents with RS, will not differ from adolescents without parent-reported strengths, but will show within-subject strengths in the most relevant cognitive domain).

Methods: Participants were 322 adolescents (ages 13-17; 84% male) from the Simons Simplex Collection who completed the Differential Ability Scale-II, School Age form and the Autism Diagnostic Interview-Revised (ADI-R). ADI-R items 88-93 were used to group adolescents: 0=no strength (NS), 1=RS, 2=ES in six areas: visuospatial, memory, drawing, musical, reading, and computational ability. Analyses were conducted separately for adolescents with NVIQ above and below 70 (>70=77%). ANOVA compared the three groups for each area on Verbal, Nonverbal, and Spatial Domain standard scores. Paired-sample t-tests made within-subject score comparisons in adolescents with RS.

Results: Forty-seven percent of adolescents had at least one ES and 22% a RS; proportions varied by specific skill/strength across NVIQ groups (Table 1). Within NVIQ<70, neither the ES or RS groups showed the expected group differences or within-subject strengths. The following results focus on NVIQ>70. Adolescents with visuospatial or drawing ES had significantly higher Spatial scores than adolescents without strengths in these areas, but did not differ on other domains (Table 2). Adolescents with reading ES had higher Verbal and Nonverbal scores and those with computation ES had higher Nonverbal and Spatial scores, relative to the NS group for those areas. Memory and musical ES groups did not differ from NS groups. The drawing RS group had lower Nonverbal scores and the computational RS had higher Nonverbal scores, relative to the NS group in each area; other RS groups did not differ on any cognitive domain (Table 2). There was some evidence for expected within-subject domain differences: visuospatial RS=Spatial>Verbal (t(21)=-4.00, p=.001), drawing RS=Spatial>Nonverbal (t(10)=-2.33, p<.042), computational RS=Nonverbal>Spatial (t(20)=.44, p<.662).

Conclusions: Findings lend support for the validity of parent-reported ES. Nearly one-quarter of adolescents exhibited RS, suggesting a need for additional research to understand strengths at all levels, not just the extraordinary. Understanding the neural basis of these strengths may advance understanding of cognitive processing in individuals with ASD.