BACKGROUND: How much the association between depressive symptoms (DSs) and all-cause mortality depends on cardiovascular disease (CVD) events is poorly known. We aimed to prospectively quantify the association between DSs at repeated study visits and all-cause and cause-specific mortality, and the influence of incident CVD on this association. METHODS: The Three-City Study has included adults 65 years and older, who were examined at baseline between 1999 and 2001 and after 2, 4, 7, and 10 years of follow-up. At each visit, a score of 16 or greater on the 20-item Center for Epidemiologic Studies Depression Scale defined the presence of DSs. DS status and incident coronary heart disease or stroke events were used as time-dependent variables in a Cox proportional hazard model of mortality. RESULTS: We studied 7377 participants (63.7% females) aged 73.8 years (SD = 5.4 years) without a history of CVD at baseline examination. DSs were present in 19% to 22% of subjects at each study visit. During a median follow-up of 9.4 years, 650 subjects developed a first CVD, and 1255 had died. After adjustment for baseline sociodemographic variables, vascular risk factors, impairment in daily life activities, and antidepressants, time-dependent DSs were associated with a 28% increased risk of mortality (hazard ratio [HR] = 1.28; 95% confidence interval [CI] = 1.06-1.55), and incident CVD event was associated with a 63% increased risk (HR = 1.63; 95% CI = 1.30-2.04). However, the association between DSs and mortality was not influenced by the occurrence of CVD (HR for DS and CVD interaction = 1.03; 95% CI = 0.66-1.61). A mediation analysis confirmed that incident CVD only explained 6.9% of the excess of mortality associated with DSs. CONCLUSION: In older participants, the increased risk of all-cause mortality associated with the presence of DSs at baseline and during follow-up is not modified by and is moderately mediated by incident CVD. J Am Geriatr Soc 00:1-7, 2019.< Réduire