MEASUREMENT AND CLASSIFICATION OF THE HETEROGENEOUS AUTISM PHENOTYPE

Abstract

<p>Autism Spectrum Disorder (ASD) is a heterogeneous disorder with a high burden of suffering and economic cost to society. The current Thesis represents a systematic attempt to investigate ASD heterogeneity, as it relates to the measurement and classification of the clinical phenotype. The Thesis integrates information from multiple constructs (symptoms, traits, behaviours), methods (factor analysis, cluster analysis, and factor mixture modeling), populations (clinical and high-risk samples) and time points (at diagnosis and at age 6) for the investigation of the underlying structure of the ASD phenotype in young children. The Thesis consists of four interrelated empirical studies and one Editorial. Results can be organized into three overarching themes: 1) in preschool children with ASD core diagnostic symptoms (social communication deficits and repetitive behaviours) appear to overlap with other emotional/behavioural problems (attention, withdrawal, anxiety, aggression, emotional reactivity); 2) along the heterogeneous autism spectrum there appear to be distinct, relatively homogeneous subgroups of children; on average, children across these subgroups differ in their levels of symptom severity, adaptive skills, and emotional/behavioural problems; 3) the underlying structure of the ASD symptom phenotype changes as children grow and develop. Thesis findings lend support to a much-needed shift in our conceptual and methodological approach to the study of measurement and classification of autism pathology: that is, instead of a set of categorical symptoms that present early in childhood and remain static over the life span, ASD might be better understood as a complex and dynamic disorder, structured on both categorical and dimensional constructs that vary not only across individuals at any given point, but also within individuals across time.</p>