Depressive symptoms and mortality risk in a national sample: confounding effects of health status

We examined the association between depressive symptoms and all-cause mortality in a population sample. Published findings on the relation between depressive symptoms and mortality risk point to an inconsistent association and one that is likely influenced by health status. Few studies have assessed this relation in randomly selected population samples.

Participants were 3617 noninstitutionalized adults, age 25 years or older, from the Americans' Changing Lives Study, an ongoing longitudinal study of a nationally representative sample. Depressive symptoms were measured by the 11-item version of the Center for Epidemiological Studies Depression Scale (CES-D). Cox proportional hazards models estimated the relative risk of mortality as a function of the CES-D scores at baseline.

In 7.5 years of follow-up, 542 deaths occurred. Each 1-standard unit increase on the CES-D predicted a 21% increased risk of all-cause mortality, adjusting for age, gender, and race (hazard ratio = 1.21, 95% confidence interval = 1.08 to 1.36, p = .001). This association was weakened somewhat following adjustment for education, income, body mass index, smoking and alcohol consumption (hazard ratio = 1.13, 95% confidence interval = 0.99 to 1.28, p = .06). However, control for self-reported functional limitations or chronic health conditions at baseline effectively eliminated the relationship. Analyses limited to participants with good to excellent health or no functional impairments at baseline showed no association between depressive symptoms and subsequent mortality risk. Secondary analyses showed no association between depressive symptoms and cardiovascular mortality.

These findings from a randomly selected, nationally representative sample do not support the hypothesis that depressive symptoms are independently related to mortality in the general population, after adequate adjustment for the confounding effects of physical health status.