Contrasting Risk Profiles for Suicide Attempt and Suicide Using Danish Registers and Genetic Data.

Importance: Not all individuals who die by suicide have a history of nonfatal suicide attempt (SA); however, little is known about the extent to which the genetic and environmental etiologies of SA and suicide are shared or distinct. Objective: To examine shared and distinct risk factors for SA and suicide, focusing on clinically diagnosed health conditions and genetic factors. Design, Setting, and Participants: For health conditions, a nested case-control study was performed using data from the Danish registers. For genetic factors, a case-control analysis framework was used, with individual genotypes retrieved from the iPSYCH2015 dataset, which was nested within the entire Danish population. Individuals older than 10 years were included to minimize the risk of misclassification for SA and suicide. Data were analyzed from January 2024 to April 2025. Exposures: Twenty-eight health conditions and 35 polygenic scores (PGSs) for complex traits. Main Outcomes and Measures: The primary outcomes were SA, suicide, and cumulative SA burden. Associations between health conditions and the risk of SA and suicide were assessed using conditional logistic regression. PGSs for complex traits were calculated using LDpred2-auto, and their associations with SA and suicide were evaluated via logistic regression. To assess whether effect sizes differed significantly between SA and suicide, bayesian model-based classification and Cochran Q test were applied. Results: A total of 81 713 cases of SA (50 512 [61.8%] female; mean [SD] age, 32.3 [14.9] years), with 408 490 age-matched controls (252 525 [61.8%] female; mean [SD] age, 32.3 [14.9] years), and 9362 cases of suicide (2360 [25.2%] female; mean [SD] age, 45.1 [14.6] years), along with 46 749 matched controls (11 796 [25.2%] female; mean [SD] age, 45.1 [14.6] years) who were alive at the date of the case's death, were included in the health conditions analysis. The PGS analysis included 8221 cases of SA (5944 [72.3%] female; mean [SD] age, 19.7 [4.4] years) and 225 cases of suicide (80 [35.6%] female; mean [SD] age, 24.6 [5.0] years). Chronic diseases (eg, dyslipidemia or hearing problems) showed stronger associations with SA, while severe conditions (eg, cancer) were more strongly associated with suicide. Suicide was influenced only by PGSs for mental disorders, whereas SA was associated with both psychiatric and broader health-related genetic risk factors. Notably, dose-response associations were observed for most health conditions and PGSs in relation to cumulative SA burden. Conclusions and Relevance: A broad range of health conditions and genetic factors were associated with increased risk of both outcomes; however, their shared and distinct risk factors suggest that SA and death by suicide are not solely differentiated by liability severity.