OBJECTIVE: To determine if a question about symptoms of depression in a mail survey predicts mortality after adjusting for a large number of covariates. DESIGN: National cross-sectional survey of 141,589 enrollees in Medicare, age 65 and older. Analyses used multivariate logistic regression models with death as the outcome. RESULTS: Response to a question about sadness or anhedonia was associated with death in 2 years (OR = 1.32; 95% CI = 1.2, 1.4). Results were consistent across age, gender, and presence/absence of known heart disease. Other responses associated with death were older age, male gender, and self-reported cancers, shortness of breath, heart failure, smoking, and other characteristics. Higher education and being married appeared to protect from death. DISCUSSION: A single survey question about feelings of sadness or anhedonia is predictive of death in 2 years.
OBJECTIVE: To determine if a question about symptoms of depression in a mail survey predicts mortality after adjusting for a large number of covariates. DESIGN: National cross-sectional survey of 141,589 enrollees in Medicare, age 65 and older. Analyses used multivariate logistic regression models with death as the outcome. RESULTS: Response to a question about sadness or anhedonia was associated with death in 2 years (OR = 1.32; 95% CI = 1.2, 1.4). Results were consistent across age, gender, and presence/absence of known heart disease. Other responses associated with death were older age, male gender, and self-reported cancers, shortness of breath, heart failure, smoking, and other characteristics. Higher education and being married appeared to protect from death. DISCUSSION: A single survey question about feelings of sadness or anhedonia is predictive of death in 2 years.