A Systematic Review and Meta-Analysis: Research Using the Autism Polygenic Score

Objective Genetic factors play a substantial role in the etiology of autism and its co-occurrence with other conditions and traits. The autism polygenic score, derived from the latest autism case-control meta-genome-wide association studies, captures some of the accumulated influence of common genetic variants on autism. We reviewed and meta-analyzed published studies that assessed the relationship between this autism polygenic score and autism diagnosis, and autistic, behavioral and neurobiological traits. Method Systematically searching public databases, we identified 72 studies and > 750 outcome measures. Included studies received a quality assessment. Results The majority of included studies were rated as good quality. The autism polygenic score was most strongly associated with autism diagnosis (meta-analytic r = .162, 95% CI .066 – .258). The autism polygenic score was also significantly associated with autistic traits but to a lesser degree than for autism (meta-analytic r = .042 (95% CI .004 – .081). Associations with other outcomes were inconsistent and meta-analytic effect sizes were generally small (median r = .03). Conclusion We conclude that the current autism polygenic score is consistently associated with autism diagnostic status and autistic traits, but overlap between autism and other traits and conditions is not, from publications to date, explained significantly by the autism polygenic score. When compared to other mental conditions, autism is phenotypically and etiologically heterogeneous, which might drive the relatively modest associations observed with the autism polygenic score to date.