Key points are not available for this paper at this time.
BACKGROUND: Although DSM-IV requires symptoms in three criterion domains for a diagnosis of autistic disorder, the extent to which those domains are phenotypically independent is an unanswered and important question. The identification of 'endophenotypes' of the autistic syndrome may be very useful for genetic and neurobiologic studies of autism, but only if they represent truly independent sub domains of the disorder. METHODS: In this study we examined the factor structure of autistic traits using data from 226 child psychiatric patients with and without pervasive developmental disorders, employing cluster analysis of data from the Autism Diagnostic Interview-Revised (ADI-r) and principal components factor analysis of data from the Social Responsiveness Scale (SRS, a quantitative genetic measure of autistic traits formerly known as the Social Reciprocity Scale). RESULTS: The results were consistent with the existence of a singular, continuously distributed underlying factor, resulting in disparate phenotypic manifestations across the three criterion domains for autistic disorder (social deficits, language deficits, and repetitive/stereotypic behaviors). CONCLUSION: The analyses generally failed to support the existence of independent sub domains of dysfunction in autism spectrum conditions. Future studies of the association between genetic/neurobiologic markers and autistic symptomatology may be enhanced by approaches which consider autistic symptoms as quantitative traits, and which are informed by ongoing research on the development and phenomenology of core deficiencies in reciprocal social behavior.
Constantino et al. (Tue,) studied this question.