Oxytocin pathway-specific polygenic risk scores are associated with autistic traits in a population-based cohort

Oxytocin is a key neuromodulator and hormone involved in social behavior, attachment, and affiliative processes. Genetic variation in oxytocin-related genes, particularly the oxytocin receptor gene, has been implicated in autism spectrum disorder (ASD). However, the contribution of genetic variation across the broader oxytocin signaling pathway to individual differences in autistic traits remains unclear. In our recent publication, we demonstrated that the therapeutic efficacy of oxytocin on core social symptoms in ASD is associated with variation in the RYR2 gene, an oxytocin signaling pathway gene, suggesting that oxytocin pathway-related genes may modulate social functioning [1]. Nevertheless, the extent to which cumulative genetic variation within the oxytocin pathway influences ASD-related behavioral phenotypes in the general population has not been systematically investigated. Given that autistic traits are continuously distributed, examining these associations in population-based samples may provide important insights into the genetic architecture underlying social behavior.

Participants were drawn from the Hamamatsu Birth Cohort for Mothers and Children (HBC study), a population-based longitudinal cohort in Japan. The present analysis included 774 children (383 males, 391 females) with available genetic and behavioral data. Autistic traits at age 13 were assessed using the Social Responsiveness Scale, Second Edition (SRS-2), which provides total and subscale scores reflecting multiple domains of social functioning. Genomic DNA was obtained from umbilical cord blood at birth and genotyped using the Japonica array, followed by standard quality control and imputation using the 1000 Genomes Project Phase 3 reference panel. Polygenic risk scores (PRSs) were calculated using summary statistics from the iPSYCH-PGC ASD genome-wide association study. Two types of PRSs were constructed: (i) genome-wide PRSs using all available SNPs and (ii) oxytocin pathway-specific PRSs restricted to 92 genes identified from the KEGG database. Best-fit PRSs were determined across a range of p-value thresholds. Linear regression analyses were conducted with adjustment for sex and the top 10 principal components.

Genome-wide PRSs derived from ASD GWAS summary statistics showed no significant associations with SRS-2 total or subscale scores. In contrast, oxytocin pathway-specific PRSs demonstrated significant associations with multiple domains of autistic traits. Specifically, higher oxytocin pathway PRSs were associated with lower SRS-2 total scores (maximum R² = 0.011, p = 0.012), indicating reduced social impairment. Significant negative associations were also observed for the Awareness, Cognition, Communication, and Motivation subscales (maximum R² range: 0.007–0.014, p = 0.004–0.043). Conversely, a modest positive association was found for the Restricted and Repetitive Behaviors subscale (maximum R² = 0.007, p = 0.049).

In conclusion, oxytocin pathway-specific polygenic risk scores, but not genome-wide PRSs, were associated with autistic traits in a population-based sample. These findings suggest that biologically informed, pathway-restricted PRSs may capture meaningful variance in social functioning that is not detected by genome-wide approaches. The observed pattern of associations may reflect dimensional aspects of ASD-related traits in the general population, including potentially protective effects on social domains. Further studies in independent cohorts and comparative analyses across neurobiological pathways are warranted to clarify the specificity and robustness of these findings and to explore their implications for biomarker development and individualized intervention strategies.

References

1. Kuwabara H, Kojima M, Benner S, Otowa T, Watanabe T, Kuroda M, Owada K, Yassin W, Hamada J, Kano Y, Uno Y, Kushima I, Mori D, Arioka Y, Munesue T, Kasai K, Higashida H, Abe O, Takao H, Wakuda T, Kameno Y, Inoue J, Harada T, Yamauchi A, Ogawa N, Honda N, Kikuchi S, Seto M, Tomita H, Miyoshi N, Matsumoto M, Kawaguchi Y, Kanai K, Ikeda M, Nakamura I, Isomura S, Hirano Y, Onitsuka T, Takahashi N, Nakashima M, Saitsu H, Kondo K, Ikeda M, Iwata N, Shimada M, Sasaki T, Takei N, Ozaki N, Kosaka H, Okada T, Yamasue H., 2026. A key gene modulating oxytocin efficacy in autism: genome-wide discovery and verification in randomized controlled trials datasets. Mol. Psychiatry. https://www.nature.com/articles/s41380-026-03562-y .