Sex-based differences in the longitudinal association between autistic traits and positive psychotic experiences in adolescents: A population-based cohort study

Abstract

Previous reports have suggested a cross-sectional association between autistic traits and psychotic experiences (PEs) in adolescents. However, while both autistic traits and PEs show sex-related differences, no studies have directly assessed whether such differences exist in the longitudinal association between autistic traits and PEs. Using a population-based adolescent cohort sample (n = 3007), we tested whether the longitudinal association between autistic traits and positive PEs was affected by sex-based differences using regression analyses. Autistic traits were assessed at 12 years old (timepoint 1 [T1]), and PEs were assessed at 12 and 14 years old (T1 and T2). Subsequently, we tested whether subdomains of autistic traits (difficulties in social interaction, communication, imagination, attention to detail, and attention switching) were associated with subtypes of PEs (auditory hallucinations, visual hallucinations, and delusions) using structural equation modeling, after controlling for PEs at T1, socio-economic status, school performance and parents' psychiatric disorders. After controlling for PEs at T1, we did not find any associations between autistic traits at T1 and PEs at T2 in both sexes. There was no significant positive or negative association between all subdomains of autistic traits and subtypes of PEs in both sexes. Autistic traits do not seem to predict future PEs in general adolescents regardless of sex.