Individual- and population-level associations of mental disorders with intentional self-harm

Registry-based studies can inform suicide prevention by identifying mental disorders with the highest risk. Previous studies focused on severe disorders and suicide, with limited data on non-lethal self-harm or population impact. We quantified individual- and population-level associations of 32 mental disorders with non-lethal intentional self-harm (NLISH) and suicide.

Registry-based cohort study representative for all residents of Catalonia (Spain) aged ≥10 years (2014-2019; n = 645,571). Cause-specific Cox models estimated individual (hazard ratios [HRs]) and population-level (population attributable fractions [PAFs]) associations with NLISH and suicide, stratified by sex and adjusted for age, socioeconomic status, and nationality.

Individual-level associations with NLISH were strongest for borderline personality disorder (BPD; females HR = 26.9 [95%CI 24.9-29.0]; males HR = 18.9 [95%CI 16.7-21.4]). Associations with suicide were strongest for BPD in females (HR = 40.9 [95%CI 28.5-58.8]) and obsessive-compulsive disorder in males (HR = 17.4 [95%CI 5.3-56.5]). Associations with suicide were stronger among females, and those aged 10-44 across mood, substance use, dissociative, borderline personality, and psychotic disorders. Substantial proportions of outcomes were associated with common disorders: depressive episodes (PAFs 29.8-49.8%), substance use disorders (PAFs 25.1-48.7%), mixed anxiety-depressive disorders (PAFs 19.7-53.2%), and adjustment disorders (PAFs 10.6-44.6%).

Depressive, anxiety, adjustment, and substance use disorders are associated with large shares of self-harm and suicide, whereas BPD confers particularly high individual risk. Our findings support multilevel prevention strategies, especially among young people, including improved risk assessment, collaborative care, and timely access to specialized interventions.