Adaptive Behavior Scores As an Indicator of Severity of Functional Limitations in Children with Autism Spectrum Disorder: Results from a Population-Based Study

Background: Few epidemiologic studies of autism spectrum disorder (ASD) include information on functional status or severity of functional limitations, despite the increased emphasis in autism and disability classification systems on variations in functioning and levels of support required over the life course. Measures of adaptive behavior such as the Vineland Adaptive Behavior Scales were developed to measure levels of functioning in individuals with intellectual disability (ID), but their usefulness for measuring level of functioning in individuals with ASD, including those without co-occurring ID, is unclear.

Objectives: The aims of this study were to (a) describe variations in adaptive behavior scores in a large, population-based sample of children with ASD, including those with and without co-occurring ID; and (b) evaluate whether variations in adaptive behavior scores are associated with ratings of the level of impairment by clinician reviewers and other clinical features that might be related to ASD severity.

Methods: The Autism and Developmental Disabilities Monitoring (ADDM) Network is a multisite surveillance system that incorporates information from clinical and educational records of all 8-year-old children meeting diagnostic criteria for ASD in selected populations in the United States. For the present study, we combined cross-sectional data from the 15 ADDM Network sites that contributed ASD surveillance data for the period 2000-2012. Adaptive behavior, IQ scores and clinician ratings of autism symptom severity and other clinical data were available for 10,795 (47.3% of) children with ASD (8,843 boys, 1,952 girls). Adaptive behavior scores were classified based on population normative data as ≥average; borderline (between 1 and 2 standard deviations (sd) below the population mean); mild limitations (2-3 sd below the mean); and moderate to profound limitations (≥3 sd below the mean). Correlation and categorical (chi-square) analyses were used to evaluate associations between adaptive behavior results and indicators of autism symptom severity, overall and in analyses stratified by IQ categories and sex.

Results: Adaptive behavior scores were positively correlated with IQ (Spearman rho=0.5, p<0.0001), as expected, and children with ASD and co-occurring ID were more likely (29.1%) to have ≥moderate limitations in adaptive behavior than were those without ID (6.6%, p<0.0001). Among the 3,639 children with ASD and IQ >85, adaptive behavior scores indicated functional limitations that were borderline for 40.0%, mild for 31.1% and moderate to profound for 5.3%. In this group, adaptive behavior scores were significantly (p<0.0001) inversely associated with the number of developmental concerns recorded, number of ASD diagnostic criteria met and clinician ratings of severity of autism impairments, but the strength of these associations were weak to moderate (rho=0.04 to 0.260). Stratified analyses showed similar patterns for boys and girls.

Conclusions: Adaptive behavior scores available in clinical and educational records may provide a reliable indicator of severity of functional limitations in children for use in epidemiologic studies of ASD. This analysis suggests that adaptive tests are capturing some of the variation in functional limitations in children with ASD that is not fully accounted for by variations in IQ or clinician ratings of impairment based on autism symptoms.