Classifying Autism Spectrum Disorder By Severity of Functional Limitations: Results from a Large Population-Based Study

Background: With the shift to DSM-5 criteria for autism spectrum disorder (ASD) and adoption of the World Health Organization’s International Classification of Functioning, Disability and Health, there is increased interest in characterizing severity of functional limitations for individuals with ASD and other disabilities. Yet most epidemiologic studies of ASD incorporate little or no information to allow monitoring of the severity of functional limitations of individuals meeting diagnostic criteria for ASD. Adaptive behavior tests, which measure functioning in communication, daily living and motor and social skills, have been shown to be inversely correlated with both the DSM-5 ASD severity categories and the Autism Classification System of Functioning – Social Communication.

Objectives: The aims of this study were to evaluate: (a) the potential to classify ASD cases identified by the Autism and Developmental Disabilities Monitoring (ADDM) Network by severity of functional limitations based on routinely collected adaptive behavior test results; and (b) whether the increase in ASD prevalence among U.S. children between 2000 and 2012 occurred equally across levels of severity of functional limitations.

Methods: We combined cross-sectional data from all ADDM Network sites contributing ASD surveillance data for the period 2000-2012. The ADDM Network is a multiple source surveillance system incorporating health and educational records and clinician reviews to determine the number of 8-year-old children in the population meeting DSM criteria for ASD. Among 22,802 ASD cases, the records of 12,739 (55.9%) included composite and domain scores from one or more adaptive behavior tests. Based on composite score categories from the latest adaptive behavior test administered, we classified cases into three levels of adaptive functioning: (1) no significant limitations; (2) mild limitations; and (3) moderate to profound limitations.

Results: Among 12,739 ASD cases with adaptive behavior information, composite scores indicated no significant functional limitations for 37.1%, mild limitations for 44.2%, and moderate, severe or profound limitations for 18.8%. The percentage classified as having no significant functional limitations increased each surveillance year, from 21.3 in 2000 to 44.5 in 2012, while the percentage with at least moderate functional limitations decreased from, 31.3 in 2000 to 10.6 in 2012 (test for trend p<0.001). Applying the percentages in each functional category to the overall prevalence of ASD, we estimate that the prevalence of ASD per 1,000 8-year-old children with no significant functional limitations increased over time from 1.4 in 2000 to 6.5 in 2012, the prevalence of ASD with mild functional limitations increased from 3.2 in 2000 to 6.6 in 2012, and the prevalence of ASD with moderate to profound functional limitations remained relatively constant over time (2.1 in 2000, 1.6 in 2012). Analyses with imputation of missing data will also be presented.

Conclusions: Tests of adaptive behavior provide an existing source of data for classifying severity of specific functional limitations in epidemiologic studies of ASD. Available data suggest the increase in ASD prevalence between 2000 and 2012 is explained by increases in cases with mild or no significant functional limitations. Additional research could assess limitations not measured by adaptive tests.