Using the Autism Diagnostic Interview—Revised to Increase Phenotypic Homogeneity in Genetic Studies of Autism

Background Many chromosomal regions for susceptibility to autism spectrum disorders (ASDs) have been identified, but few have reached genomewide significance. In response, researchers have attempted to increase the power of their analyses by stratifying samples to increase phenotypic homogeneity. Although homogeneity has typically been defined by a single variable, resultant groups often differ in other dimensions that may be directly pertinent. Group differences in age, gender, IQ, and measures of autism severity are examined as related to Autism Diagnostic Interview—Revised (ADI-R) domains previously used for subsetting or Quantitative Trait Analysis (QTL). Methods Participants were research participants and clinic referrals for assessment of possible autism. Assessments included the ADI-R, Autism Diagnostic Observation Schedule, Vineland Adaptive Behavior Scales, and a developmental or cognitive test. Data were collected for 983 individuals, ages 4 to 52 years, with diagnoses of autism and ASDs. Results Findings suggest that, of several potential grouping variables, only restricted and repetitive behaviors associated with Insistence on Sameness were independent of age, IQ, and autism severity. Conclusions Results emphasize the potential unintended effects of stratification and the importance of understanding such interrelationships between phenotypic characteristics when defining subgroups or performing QTL.