Grosso · Annual review of nutrition 2017 · umbrella review of meta-analyses · n=141 meta-analyses

Coffee, Caffeine, and Health Outcomes: An Umbrella Review.

Cited 537 times in the scientific literature.

Level 1 - systematic review of randomized trials

Umbrella review synthesizing meta-analyses of randomized controlled trials and observational studies

PubMed 28826374 · doi:10.1146/annurev-nutr-071816-064941 · record verified 2026-08-29

What was done

The authors conducted an umbrella review systematically evaluating evidence from published meta-analyses of observational studies and randomized controlled trials (RCTs) to assess associations between coffee and caffeine consumption and diverse health outcomes.

What was found

The abstract does not report numerical effect sizes, relative risks, or confidence intervals. Across 112 meta-analyses of observational studies (59 unique outcomes), coffee was associated with a probable decreased risk of breast, colorectal, colon, endometrial, and prostate cancers; cardiovascular disease and mortality; Parkinson's disease; and type-2 diabetes. Across 20 meta-analyses of observational studies (14 outcomes), caffeine was associated with a probable decreased risk of Parkinson's disease and type-2 diabetes, and an increased risk of pregnancy loss. Across 9 meta-analyses of RCTs (12 acute outcomes), coffee was associated with an increase in serum lipids (affected by significant heterogeneity), and caffeine was associated with an acute increase in blood pressure.

Why it matters

This review synthesizes meta-analytic evidence across dozens of clinical endpoints, indicating that habitual coffee intake is broadly associated with health benefits and compatible with a healthful diet in non-pregnant populations.

Limits

The abstract provides no numerical effect estimates, confidence intervals, or sample sizes. Most chronic disease outcomes rely exclusively on observational meta-analyses vulnerable to residual confounding and exposure misclassification. Randomized trial data were limited to short-term acute physiological outcomes rather than long-term clinical endpoints, and the lipid findings exhibited significant heterogeneity.

Cited by