Death of a Close Relative and the Risk of Suicide in Sweden—A Large Scale Register-Based Case-Crossover Study

BACKGROUND: Bereavement is thought to be a risk factor for suicide but the association has not been thoroughly investigated according to specific sensitive time periods and risk groups using a self-matched design. We aimed to 1) determine the risk of suicide within the first year after death of a close relative, 2) investigate if and how the risk changes within this time window and 3) determine if sex, age, and type of relationship, affect this association. METHODS: A self-matched, case-crossover study was performed by linking Swedish registers. In total, 31 059 individuals with suicide between 1990 and 2011 were included. Different periods within the year prior to the suicide were compared with corresponding periods one year earlier in the same individual’s life. Conditional logistic regression was used to calculate odds ratios (OR) and 95% confidence intervals (CI) for suicide after death of a close relative. RESULTS: Increased ORs of suicide were seen during the first month, OR 1·77 (95% CI 1·35–2·34), and the first half-year, 1·27 (1·13–1·43). An even higher OR was found within the first week, 3·43 (1·89–6·22). Patterns were similar for women and men and across age groups. Death of a partner or child but not death of a sibling or parent was associated with a significantly increased suicide risk. The strongest association was seen after death of a partner in individuals aged 45 and older. DISCUSSION: These findings provide knowledge of sensitive time periods and at-risk groups in the early period of bereavement. Due to the use of a self-matched study design, methodological challenges of unmeasured residual confounding could be overcome.