Classification of children with autism spectrum disorders: A finite mixture modeling approach to heterogeneity

摘要

The heterogeneity found in autism and related disorders (i.e., "autism spectrum disorders") is widely acknowledged. Even within a specific disorder, such as Autistic Disorder, the range in abilities and clinical presentation is broad. The heterogeneity observed has prompted many researchers to propose subtypes beyond the commonly used DSM-IV-TR diagnoses. However, previous research has used analysis procedures that have significant limitations. This research examined the possible presence and structure of latent behavioral subgroups in young children. Multiple measures were utilized including direct observational measurement of social behavior. Finite mixture modeling analysis revealed evidence for two latent components or subgroups. The groups differed significantly on IQ, receptive language, and social interactions, but not on autism symptom severity. The two groups did not differ by age or proportion of males or females, but differed in terms of diagnoses. Results are compared to existing models of subgroups in autism spectrum disorders, and discussed in terms of implications for future research and treatment.