Back

Pathways from Polygenic Risk to Suicidality: Effects of Alcohol Use Disorder and Childhood Adversity

Wu, V.; Qin, X.-J.; Ashley-Koch, A.; Kimbrel, N. A.; Gelernter, J.; Docherty, A. R.; Kranzler, H.; Feinn, R.; Davis, C. N.

2026-02-12 genetic and genomic medicine
10.64898/2026.02.10.26345999 medRxiv
Show abstract

BackgroundThe prevalences of suicidal ideation (SI) and suicide attempt (SA) are influenced by genetic, behavioral, and environmental factors. Alcohol use disorder (AUD) and adverse childhood experiences (ACEs) may mediate or moderate genetic liability for suicidality. MethodsUsing data from 10,275 participants (43.8% female; 47.2% African-like genetic ancestry [AFR], 52.8% European-like genetic ancestry [EUR]), we tested whether polygenic scores (PGS) for SI and SA predicted lifetime SI or SA. We also evaluated whether alcohol use disorder (AUD) mediated these associations and whether adverse childhood experiences (ACEs) moderated the direct and indirect pathways. ResultsAlthough there were significant direct associations of the SA PGS with SA (AFR: b = 0.36, SE = 0.01; EUR: b = 0.17, SE = 0.01; both ps < 2e-16), the SI PGS did not predict SI (p > 0.55). AUD mediated SA genetic risk (average causal mediation effect (ACME): AFR = 0.01, 95% CI [0.01-0.01]; EUR = 0.02, 95% CI [0.01-0.02]; both ps < 2e-16). Moderation analyses indicated that indirect effects were attenuated by ACEs score ({Delta}ACME: AFR = 0.02, p < 2e-16; EUR = 0.01, p = 0.03). There was neither mediation nor moderated mediation for SI. ConclusionsGenetic liability to SA operates partly through AUD, particularly among individuals with lower childhood adversity. Under higher adversity, alternative pathways to SA likely predominate. These findings highlight the need to consider distinct etiological pathways to the development of suicidality and the relevance of AUD as a modifiable target for suicide prevention among individuals at high genetic liability.

Matching journals

The top 6 journals account for 50% of the predicted probability mass.

50% of probability mass above

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.