← All sources View original paper →
Coffee consumption and health: umbrella review of meta-analyses of multiple health outcomes
Umbrella review re-synthesizing 201 meta-analyses of observational coffee studies plus 17 interventional meta-analyses: largest summary relative risk reductions for several outcomes at about three to four cups per day versus none (e.g., all-cause mortality RR 0.83, 95% CI 0.79–0.88; cardiovascular mortality RR 0.81, 0.72–0.90), with pregnancy-related harms at high versus low/no intake and a women's fracture signal; authors stress residual confounding and the need for RCTs.
Design
- Umbrella review of meta-analyses covering observational and interventional coffee–outcome links
- Corpus: 201 observational meta-analyses (67 unique outcomes) + 17 interventional meta-analyses (9 unique outcomes)
Headline pooled associations (summary estimates quoted in abstract)
- Largest RR reductions near ~3–4 cups/day vs none: all-cause mortality RR 0.83 (95% CI 0.79–0.88); cardiovascular mortality RR 0.81 (0.72–0.90); CVD RR 0.85 (0.80–0.90)
- High vs low coffee — incident cancer: RR 0.82 (95% CI 0.74–0.89)
- Pregnancy harms (high vs low/no): low birth weight OR 1.31 (95% CI 1.03–1.67); preterm birth ORs 1.22 (first trimester) and 1.12 (second); pregnancy loss OR 1.46 (95% CI 1.06–1.99)
- Women's fracture association reported in abstract narrative (directionally higher risk with higher coffee; not in men)
Evidence hygiene
- Dominated by observational meta-analyses → confounding (especially smoking) remains a central threat despite many adjustments.
- Not the same evidence stream as acute caffeine ergogenic dosing trials on
caffeine-ergogenic-supplementation— this row summarizes habitual coffee epidemiology.
Publication
Poole R, Kennedy OJ, Roderick P, et al. BMJ. 2017 Nov 22;359:j5024. doi: 10.1136/bmj.j5024. PMID 29167102.
Outcomes
- All-Cause Mortality RiskEvents: /
- Same umbrella summary band (~3–4 cups/day vs none): cardiovascular mortality RR 0.81 (95% CI 0.72–0.90) and cardiovascular disease RR 0.85 (95% CI 0.80–0.90) per abstract.