Coffee consumption and all-cause and cause-specific mortality: a meta-analysis by potential modifiers

Abstract

Coffee consumption has been associated with decreased mortality in previous studies. As aging, obesity, and lifestyle factors affect the risk of mortality, the association between coffee and mortality needs to be examined in various subpopulations by characteristics of subjects. To quantitatively assess this association, we conducted an updated meta-analysis including stratified analyses by potential modifiers. We searched in the PubMed and Web of Science databases through March 8, 2019, and conducted meta-analysis including linear and non-linear dose-response analyses. We identified 40 studies including 3,852,651 subjects and 450,256 all-cause and cause-specific deaths. Non-linear inverse associations between coffee consumption and mortality from all-causes, cardiovascular disease (CVD), and cancers were found. The lowest relative risk (RR) was at intakes of 3.5 cups/day for all-cause mortality (RR = 0.85, 95% CI 0.82-0.89), 2.5 cups/day for CVD mortality (RR = 0.83, 95% CI 0.80-0.87), and 2 cups/day for cancer mortality (RR = 0.96, 95% CI 0.94-0.99), while additional intakes were not associated with further lower mortality. An inverse association between coffee consumption and all-cause mortality was maintained irrespective of age, overweight status, alcohol drinking, smoking status, and caffeine content of coffee. By region, Europe and Asia showed stronger inverse associations than US. A non-linear inverse association was found for mortality from respiratory disease and diabetes, while linear inverse association was found for mortality from non-CVD, non-cancer causes. Moderate coffee consumption (e.g. 2-4 cups/day) was associated with reduced all-cause and cause-specific mortality, compared to no coffee consumption. The inverse association between coffee and all-cause mortality was consistent by potential modifiers except region.

Keywords: Age; Alcohol consumption; BMI; Coffee; Meta-analysis; Mortality.