Abstract
OBJECTIVES: To understand which causes of death are higher in black than white community-dwelling older adults and determine whether differences in baseline risk factors explain racial differences in mortality. DESIGN: Longitudinal cohort study (Health, Aging, and Body Composition Study). SETTING: Pittsburgh, Pennsylvania; and Memphis, Tennessee. PARTICIPANTS: Black and white men and women aged 70 to 79 during recruitment (N=3,075; 48% men, 42% black) followed for a median of 13 years. MEASUREMENTS: A committee of physicians adjudicated cause of death, which was categorized as cardiovascular disease (CVD), stroke, cancer, dementia, pulmonary, infection, kidney, or other causes. Using competing risks regression, we examined whether known risk factors at baseline (demographic characteristics, smoking, body mass index, chronic diseases, physical function, cognition) could explain racial differences in cause-specific mortality risk. RESULTS: During follow-up, 1,991 (65%) participants died. Black participants died at higher rates from cancer (hazard ratio (HR)=1.36, 95% confidence interval (CI)=1.14-1.63), kidney disease (HR=2.09, 95% CI=1.16-3.74), stroke (HR=1.31, 95% CI=0.98-1.76); and CVD (HR=1.16, 95% CI=0.98-1.37). Poorer physical and cognitive performance at baseline among black participants explained most of the racial difference in risks of dying from kidney disease, stroke, and CVD but not cancer. When examining types of cancer deaths, black participants died at higher rates from multiple myeloma, pancreatic cancer, and prostate cancer, which baseline risk factors did not explain either. CONCLUSION: Factors contributing to poorer physical and cognitive performance in similarly aged black men and women could be targets to reduce excess mortality from CVD, stroke, and kidney disease. More work is needed to identify factors contributing to cancer mortality disparities.