Mental Health
The associations between major depressive disorder and the risks of mortality and hospitalization Sonali Amarasekera* Sonali Amarasekera Eo Rin Cho Susan Bondy Trevor Young Patrick Brown Prabhat Jha
Background: Major depressive disorder (MDD) may increase the risk of mortality and hospitalization from both suicide and chronic diseases. However, these associations must be interpreted cautiously due to potential confounding effects from demographic (age, sex), modifiable (smoking, alcohol, body mass index, physical activity) and social (socioeconomic status, marital status, family history of mental illness) factors. Objective: To estimate associations between lifetime MDD and cause-specific mortality and hospitalization by sex using UK Biobank data. Design: We conducted a retrospective cohort study of 113,196 participants (aged 40-69) assessed for lifetime MDD, with no medical or psychiatric comorbidities at baseline. Methods: Proportional hazards models assessed the effect of MDD on mortality and hospitalization rates due to suicide, vascular disease, respiratory disease, and cancer. We adjusted for age, modifiable and social factors. Residual confounding was evaluated by examining hazard ratios (HRs) and changes in the log-hazard after adjusting for confounders. Results: Lifetime MDD significantly increased the risk of suicide mortality in both sexes, with stronger associations observed in women (HR = 7.91; 95% CI = 3.05-20.54) than men (HR = 4.25; 2.00-9.06). Further, the log-hazard increased with confounder adjustment in women, and decreased in men (12%). MDD also significantly increased the risk of hospitalization from suicide attempts, but these associations were more sensitive to residual confounding as indicated by greater reductions in the log-hazard (14% in women, 33% in men). In both sexes, MDD was positively associated with hospitalization from cardiac and respiratory diseases. Associations with stroke were significant only in women. Conclusion: Lifetime MDD appears to increase risk of mortality and hospitalization due to suicide and certain chronic diseases. The strength of the associations and confounding effects vary by sex.
