IDENTIFYING RELATIONSHIPS BETWEEN LATENT SUBTYPES OF SOCIAL BEHAVIOUR AND SNP VARIATION IN AUTISM SPECTRUM DISORDER

Autism spectrum disorder (ASD) is a heritable neurodevelopmental condition characterised by challenges in social interaction, communication, restricted and repetitive behaviours, and atypical sensory responses. The heterogeneity of ASD complicates linking genetic variation to behavioural traits assessed in tests such as the social responsiveness scale (SRS). The SRS quantifies social deficits in ASD and distinguishes them from other disorders. Whilst most research treats the SRS as a single score, this study aimed to classify individuals into multiple discrete subgroups of general social behaviour.
Here, we aimed to stratify social behavioural phenotypes using the SRS and test their associations with common SNP variation. Using SRS subscale T-scores capturing social awareness, cognition, communication, and motivation, we derived five social subtypes (latent profile analysis) representing distinct patterns and severities of social impairment. We observed significant sex differences across all four subscales, and age was significantly associate with awareness, cognition, and communication, with increasing age corresponding to lower T-scores.
We then performed genome-wide association analyses using linear regression under an addictive genetic model, testing SNP associations with each subscale phenotype. A quantile-quantile plot indicated minimal test statistic inflation, suggesting adequate control of population structure. Whilst 25 SNP variants met suggestive significance threshold, none passed the genome-wide threshold. Overall, this study provides a framework for integrating behavioural and genetic data in ASD and highlights the robust age and sex effects on SRS subscales that should be accounted for in future genotype-phenotype analyses.