Predictors of Adaptive Functioning in Verbal and Nonverbal Individuals with ASD

Background:

The relationship between adaptive functioning and other metrics of psychological functioning and psychopathology in individuals with autism spectrum disorders is poorly understood. Prior research suggests that this is due to the significant heterogeneity of individuals diagnosed with ASD, particularly the difference between those who are verbal and nonverbal.

Objectives:

The purpose of the current study is to examine the relationship between adaptive functioning and possible predictors in a sample of 1,380 individuals with ASD ages 4 to 17. Of these, 1,160 individuals are verbal and 220 are nonverbal. We explored the relationship between a summary measure of function on the Vineland Adaptive Behavior Scales, Second Edition (Vineland) and several possible predictors: age, gender, Verbal IQ, Nonverbal IQ, total scaled score on the Autism Diagnostic Observation Schedule (ADOS), subscales of the Autism Diagnostic Interview-Revised (ADI-R), and the Social Responsiveness Scale (SRS).

Methods:

All individuals participated in the Simons Simplex Collection (SSC), a North American multi- site, university-based research study that includes families with only one child with an ASD. A linear regression analysis was performed using SPSS to determine the most significant predictors of overall adaptive functioning, as measured by the Vineland Daily Living Scale (DLS). Analyses were performed separately for verbal and nonverbal subjects.

Results:

In the verbal subgroup, which comprised 84% of the SSC sample, as predicted, younger age, higher verbal and nonverbal IQ scores, and a lower score on the SRS were the best predictors of increased age-adjusted adaptive functioning. On the ADI-R a lower score on social difficulties predicted higher functioning, as expected; however, a higher ADI-R score on restricted and repetitive behaviors (RRB) predicted higher functioning. The ADOS was not a predictor of adaptive functioning. In the nonverbal SSC sample, younger age and lower SRS score were the best predictors of adaptive functioning.

Conclusions:

Though a range of assessment measures has improved our ability to understand the heterogeneity of the autism spectrum, predicting adaptive functioning in individuals with ASD remains a complicated challenge. Across both verbal and nonverbal samples, the SRS is the best specific measure for predicting functioning. In a verbal sample, higher verbal and nonverbal IQ and lower ADI-R social difficulties are also good predictors of functioning. The correlation of these measures is likely to be due to the fact that they are either parent-report (as is adaptive functioning) or they capture a directly practical set of abilities (IQ). On the other hand, the ADOS, an observational measure of social cognition, does not predict adaptive functioning. Two findings also suggest sampling or diagnostic bias. Younger children have higher age-adjusted adaptive functioning because children identified early on are selected more because of difficulties with social cognition than functional problems. Also, verbal individuals with more RRB have higher adaptive functioning, likely because presence of these traits make the diagnosis of ASD more likely even in the absence of more pervasive dysfunction. These findings help shed light on the phenotypic heterogeneity of the ASD population and the idiosyncrasies of an ASD diagnosis.