Examining the bidirectional association between emotion recognition and social autistic traits using observational and genetic analyses

Background: There is mixed evidence for an association between autism spectrum disorder (ASD) and emotion recognition deficits. We sought to assess the bidirectionality of this association using phenotypic and genetic data in a large community sample. Methods: Analyses were conducted in three stages. First, we examined the bidirectional association between social autistic traits at age 8 years and emotion recognition task (ERT) responses at age 24 years (Study 1; N = 3,562); and between Diagnostic Analysis of Non-Verbal Accuracy (DANVA) emotion recognition responses at age 8 years and social autistic traits at age 10 years (Study 2; N = 9,071). Next, we used genetic analyses (Study 3) to examine the association between polygenic risk scores for ASD and outcomes for the ERT and DANVA. The genetic correlation between ASD and ERT responses at age 24 was also estimated. Analyses were conducted in the Avon Longitudinal Study of Parents and Children. Results: Social autistic traits at age 8 years were negatively associated with later total correct responses on ERT in Study 1 (b = −0.18; 95% CI: −0.27 to −0.09). We also found evidence of an association in Study 2 (b = −0.04; 95% CI: −0.05 to −0.03). We found the opposite association, that is positive, between the ASD polygenic risk score and ERT (b = 0.40; 95% CI: 0.10 to 0.70); however, this association varied across different p-value thresholds and would not survive multiple testing, so should be interpreted with caution. We did not find evidence of a genetic correlation between ASD and ERT. Conclusion: We found an observational association between poorer emotion recognition and increased social autistic traits. Our genetic analyses may suggest a shared genetic aetiology between these or a potential causal pathway; however, future research would benefit from using better powered GWAS to examine this further. Our results may inform interventions targeting emotion recognition