A Data-Driven Approach in an Unbiased Sample Reveals Equivalent Sex Ratio of Autism Spectrum Disorder-Associated Impairment in Early Childhood.

Background: Sex differences in the prevalence of neurodevelopmental disorders are particularly evident in autism spectrum disorder (ASD). Heterogeneous symptom presentation and the potential of measurement bias hinder early ASD detection in females and may contribute to discrepant prevalence estimates. We examined trajectories of social communication (SC) and restricted and repetitive behaviors (RRBs) in a sample of infant siblings of children with ASD, adjusting for age- and sex-based measurement bias. We hypothesized that leveraging a prospective elevated familial likelihood sample, deriving data-driven behavioral constructs, and accounting for measurement bias would reveal less discrepant sex ratios than are typically seen in ASD.

Methods: We conducted direct assessments of ASD symptoms at 6 to 9, 12 to 15, 24, and 36 to 60 months of age (total n = 1254) with infant siblings of children with ASD (n = 377) and a lower ASD-familial-likelihood comparison group (n = 168; n = 527). We established measurement invariance across age and sex for separate models of SC and RRB. We then conducted latent class growth mixture modeling with the longitudinal data and evaluated for sex differences in trajectory membership.

Results: We identified 2 latent classes in the SC and RRB models with equal sex ratios in the high-concern cluster for both SC and RRB. Sex differences were also observed in the SC high-concern cluster, indicating that girls classified as having elevated social concerns demonstrated milder symptoms than boys in this group.

Conclusions: This novel approach for characterizing ASD symptom progression highlights the utility of assessing and adjusting for sex-related measurement bias and identifying sex-specific patterns of symptom emergence.