RESUMO
BACKGROUND: It is unclear whether post-traumatic stress disorder [PTSD] is associated with suicide risk in the general population, whether this differs by sex, or what the population impact of PTSD is for suicide. METHODS: We constructed a nationwide cohort of all people living in Sweden, born 1973-1997, followed from their 14th birthday (or immigration, if later) until suicide, other death, emigration or 31 December 2016. We used Cox proportional hazards regression to estimate hazard ratios [HR], and calculated the population impact of PTSD on suicide. We included sensitivity analyses to explore effects of outcome and exposure definitions, and to account for potential competing risks. RESULTS: Of 3,177,706 participants, 22,361 (0â¢7%) were diagnosed with PTSD, and 6,319 (0â¢2%) died by suicide over 49â¢2 million person-years. Compared with women and men without PTSD, suicide rates were 6â¢74 (95%CI: 5â¢61-8â¢09) and 3â¢96 (95%CI: 3â¢12-5â¢03) times higher in those with PTSD, respectively, after sociodemographic adjustment. Suicide rates remained elevated in women (HR: 2â¢61; 95%CI: 2â¢16-3â¢14) and men (HR: 1â¢67; 95%CI: 1â¢31-2â¢12) after adjustment for previous psychiatric conditions; attenuation was driven by previous non-fatal suicide attempts. Findings were insensitive to definitions or competing risks. If causal, 1â¢6% (95%CI: 1â¢2-2â¢1) of general population suicides could be attributed to PTSD, and up to 53.7% (95%CI: 46.1-60.2) in people with PTSD. LIMITATIONS: Residual confounding remains possible due to depressive and anxiety disorders diagnosed in primary care but unrecorded in these registers. CONCLUSIONS: Clinical guidelines for the management of people with PTSD should recognise increased suicide risks.