Coffee's sweet spot: 4 cups a day linked to 16% lower death risk

Coffee is one of the most widely consumed drinks in the world, so even modest associations with mortality carry real weight at population level. Individual studies had variously linked coffee to higher risk, lower risk and no difference, leaving the shape of the relationship unclear. In particular, nobody had firmly established whether the association changes with dose: whether one cup differs from four, and whether heavy consumption erases any benefit. This meta-analysis set out to trace that dose-response curve for deaths from all causes, cardiovascular disease and cancer.

The researchers identified prospective studies published between 1966 and 2013 through PubMed and the reference lists of selected articles, including only those reporting mortality risks across at least 3 categories of coffee consumption. That requirement is what makes a dose-response analysis possible. Twenty-one prospective studies qualified, together covering 997,464 participants and 121,915 deaths. Results from individual studies were pooled using a random-effects model, and separate dose-response curves were estimated for all-cause, cardiovascular and cancer mortality.

The associations were strongly non-linear for all-cause and cardiovascular mortality, meaning risk did not simply fall in a straight line as consumption rose. The largest reduction in all-cause mortality, 16 percent, was observed at 4 cups per day. For cardiovascular mortality the low point came earlier, at 3 cups per day, with a 21 percent reduction. Cancer mortality showed no association with coffee consumption at any level. Overall, coffee consumption was inversely associated with death from all causes and from cardiovascular disease.

This pooled analysis helped settle a long-running back and forth in which coffee was cast alternately as vice and virtue, and the data lean toward lower mortality among drinkers. The non-linear shape is the notable contribution: the association peaks at moderate intakes rather than growing with every extra cup. Because all the included studies followed people rather than assigning coffee at random, coffee drinkers may differ from abstainers in ways the studies cannot fully account for. The null result for cancer mortality is also informative, given persistent public suspicion on that front.

If you drink coffee in moderate amounts, this analysis is broadly reassuring: habitual consumption was associated with lower, not higher, mortality across nearly a million people. The pattern bottoms out at a few cups a day, so it offers no support for the idea that more is always better. It is equally not a reason for non-drinkers to start, since an association of this kind cannot establish cause. On cancer specifically, the data showed no relationship in either direction.