RESUMEN
BACKGROUND: The association of depression with mortality and the significance of explanatory factors, in particularly gender, have remained an issue of debate. We therefore aimed to estimate the effect of depression on all-cause mortality, to examine potential explanatory factors and to assess effect modification by gender. METHODS: We used Cox regression models to estimate the effect of depression on mortality based on data from the Gutenberg Health Study, which is a prospective cohort study of the adult population in the districts of Mainz and Mainz-Bingen, Germany. Baseline assessment was between 2007 and 2012. Effect modification by gender was measured on both additive and multiplicative scales. RESULTS: Out of 14,653 participants, 7.7% were depressed according to Patient Health Questionnaire 9 (PHQ-9), and 1,059 (7.2%) died during a median follow-up of 10.7 years. Depression elevated the risk of mortality in men and women in age-adjusted models (HR: 1.41, 95%-CI: 1.03-1.92; resp. HR: 1.96, 95%-CI: 1.43-2.69). Adjustment for social status, physical health and lifestyle covariates attenuated the effect and in the fully-adjusted model the hazard ratio was 0.96 (95%-CI: 0.69-1.33) in men and 1.53 (95%-CI: 1.10-2.12) in women. For effect modification by gender, the measure on multiplicative interaction was 0.68 (95%-CI 0.44-1.07) and on additive interaction was RERI=-0.47 (95%-CI -1.24-0.30). LIMITATIONS: The PHQ-9 is a single self-report measure of depression reflecting symptoms of the past two weeks, limiting a more detailed assessment of depression and course of symptoms, which likely affects the association with mortality. CONCLUSIONS: Depression elevates mortality by multifactorial pathways, which should be taken into account in the biopsychosocially informed treatment of depression. Effect modification by gender was not statistically significant.