Cognitive Profiles and ASD Symptomatology

Background: Numerous studies have reported enhanced nonverbal relative to verbal IQ (NV > V) in children with ASD, and that children with a NV > V cognitive split display greater social impairment than children with other cognitive profiles (Tager-Flausberg & Joseph, 2003; Joseph, Tager-Flausberg, & Lord, 2002). Furthermore, recent genetic research has identified a quantitative trait locus influencing nonverbal-verbal IQ discrepancy in multiplex ASD families suggesting that IQ discrepancy may be a potential ASD phenotype (Chapman, et al., 2011). These findings support the hypothesis that children with this IQ discrepancy represent a unique subtype within ASD.

Objectives: We aimed to explore cognitive profiles in children with ASD and assess the relationship of nonverbal-verbal discrepancy to ASD symptomatology. We hypothesized 1) a distinct NV > V cognitive split would be observed in children with ASD with greater frequency than normative samples and 2) the presence of a NV > V split would be associated with greater ASD symptomatology and adaptive functioning impairment.

Methods: Participants included 1949 children (1710 males; 244 females) with ASD between the ages of 4 years and 17 years, 11 months old (M = 8.83 years, SD = 3.49) from the Simons Simplex Collection. Children were assigned to three different groups (NV > V, V > NV, and no discrepancy) based on the degree of split between their nonverbal and verbal IQ as measured by the Differential Ability Scales – 2nd edition (DAS-2). ASD symptomatology was measured using the Calibrated Severity Score (CSS) from the Autism Diagnostic Observation Schedule (ADOS) and domain scores from the Autism Diagnostic Interview-Revised (ADI-R). Adaptive behavior was measured using the Vineland Adaptive Behavior Scales (VABS).

Results: In our sample, 1150 (58.9%) showed no cognitive split, 535 (27.4%) showed a NV > V cognitive split, and the remaining 269 (13.8%) displayed a V > NV cognitive split. Results indicate a greater rate of NV > V IQ discrepancy relative to the DAS-2 normative sample (X²(2) = 236.31, p< .001). MANCOVA, with age and gender covaried, revealed significant group differences in ASD symptomatology. The NV > V group showed greater impairment than the other groups on the ADOS CSS (F(2, 1845 ) = 12.62, p < .001) and the ADI-R social (F(2, 1845) = 8.75, p < .001) and communication (F(2, 1845) = 11.23, p < .001) domains. No significant group differences were found in adaptive functioning or ADI-R restricted and repetitive behaviors.

Conclusions: Our findings indicate that, within our large cohort of children with ASD, significantly more children exhibited the NV > V IQ discrepancy profile than would be expected in a neurotypical population. Children with the NV > V profile displayed elevated ASD symptomatology per clinician observation as well as parent interview compared to those with the V > NV profile and with no discrepancy. These findings are consistent with previous work identifying higher rates of a NV > V discrepancy in children with ASD and support the existence of an ASD subtype that may be a valuable phenotype for future genetic studies.