The Pervasive Developmental Disorders Checklist: Development and preliminary validation

Although there is substantial support for the notion that autism represents a biologically based, neurodevelopmental disorder, uncovering specific and replicable neurobiological substrates for this problem has proven difficult. Sources for this problem include the behavioral heterogeneity of autistic samples, and the lack of metrics for adequately quantifying symptoms diagnostic of and associated with the disorder, and that can be useful in investigating neurobiological correlates of autism.; The present study aimed to develop and examine preliminary data for the Pervasive Developmental Disorders Checklist (PDDC), a measure that comprehensively quantifies symptoms commonly occurring in autistic spectrum disorders including those symptoms central to diagnosis (e.g., "autistic triad") as well as associated symptoms (e.g., sensory problems). The essential feature of the measure is the empirical derivation of specific scales that separately quantify all the relevant symptoms rather than sum symptoms to a single index, in order to instigate potentially useful clinical descriptions of an individual child's behavior, as well as allow the investigation of the relationship between symptoms (or symptom patterns) and external criteria (e.g., neurobiological correlates). The PDDC measures symptoms or behaviors along 16 domains (29 subdomains), including four competence domains, 11 deviance domains, and one response bias domain.; Sixty-two school-aged children (mean age = 104.78 months +/- 31.52) with confirmed diagnoses of autism spectrum disorders and their caregivers were included in the study. Data were collected on demographic characteristics, autistic symptomatology, intellectual, language, adaptive behavior and behavioral functioning. AMT PET data were available for a portion of the sample (n = 29). All 16 domains were demonstrated to have adequate to excellent internal consistency and stability. Correlations with measures of similar constructs indicated that twelve of the domains have good convergent and, for most of the domains, adequate discriminant validity. Cluster analyses of eight of the domains resulted in a four-group solution that was demonstrated to have promising external validity based on group differences on demographic, neurodevelopmental, and AMT PET variables.