Suicide-attempt polygenic scores modestly stratify risk among ideators

Researchers analysed genetic and self-reported data from the Australian Genetics of Depression Study and the Australian Genetics of Bipolar Disorder study. The analytic sample included 17,639 participants of European ancestry; 17,622 had complete data for multinomial analyses. Investigators modelled polygenic scores (PGS) for suicidal ideation, attempt and death across mutually exclusive phenotypes, and assessed symptom severity, age of onset and transition from ideation to attempt. Participants were recruited through antidepressant or lithium prescription histories, and outcomes were based on self-report.

Ideation- and attempt-related PGS showed partly distinct associations. Among participants reporting ideation, the attempt PGS was associated with 23% higher odds of an attempt; the absolute risk difference between the top and bottom PGS deciles was 9–12%. Adding PGS improved discrimination only incrementally (ΔAUC approximately 0.02–0.03; ΔC approximately 0.01). The death PGS showed no robust effects. The authors describe the findings as supporting partially distinct genetic liabilities, while noting that predictive utility remains modest. The treatment-enriched, European-ancestry cohorts may not represent the broader population, and the results do not establish clinical utility for individual risk prediction.