Association of Coffee Consumption With Total and Cause-Specific Mortality Among Nonwhite Populations

Abstract

Background: Coffee consumption has been associated with reduced risk for death in prospective cohort studies; however, data in nonwhites are sparse.

Objective: To examine the association of coffee consumption with risk for total and cause-specific death.

Design: The MEC (Multiethnic Cohort), a prospective population-based cohort study established between 1993 and 1996.

Setting: Hawaii and Los Angeles, California.

Participants: 185 855 African Americans, Native Hawaiians, Japanese Americans, Latinos, and whites aged 45 to 75 years at recruitment.

Measurements: Outcomes were total and cause-specific mortality between 1993 and 2012. Coffee intake was assessed at baseline by means of a validated food-frequency questionnaire.

Results: 58 397 participants died during 3 195 484 person-years of follow-up (average follow-up, 16.2 years). Compared with drinking no coffee, coffee consumption was associated with lower total mortality after adjustment for smoking and other potential confounders (1 cup per day: hazard ratio [HR], 0.88 [95% CI, 0.85 to 0.91]; 2 to 3 cups per day: HR, 0.82 [CI, 0.79 to 0.86]; ≥4 cups per day: HR, 0.82 [CI, 0.78 to 0.87]; P for trend < 0.001). Trends were similar between caffeinated and decaffeinated coffee. Significant inverse associations were observed in 4 ethnic groups; the association in Native Hawaiians did not reach statistical significance. Inverse associations were also seen in never-smokers, younger participants (<55 years), and those who had not previously reported a chronic disease. Among examined end points, inverse associations were observed for deaths due to heart disease, cancer, respiratory disease, stroke, diabetes, and kidney disease.

Limitation: Unmeasured confounding and measurement error, although sensitivity analysis suggested that neither was likely to affect results.

Conclusion: Higher consumption of coffee was associated with lower risk for death in African Americans, Japanese Americans, Latinos, and whites.

Primary funding source: National Cancer Institute.

Figure.

Adjusted mortality in the Multiethnic Cohort, by age by category of coffee consumption, 1993–2012. Direct adjusted curves were generated using the Cox model as the average of the model-based curves at observed profiles for the following covariates: age at cohort entry, sex, ethnicity, smoking variables (smoking status; average number of cigarettes; squared average number of cigarettes; number of years smoking [time-dependent]; number of years since quitting [time-dependent]; and interactions between ethnicity and smoking status, average number of cigarettes, squared average number of cigarettes, and number of years smoking), body mass index, education, physical activity, alcohol consumption, total energy intake, energy from fat, and preexisting illness. For categorical covariates, dummy variables were used. The “None” and “1–3 cups/ mo” lines overlap.