Poole, R., et al, BMJ 359:J5024, November 22, 2017
BACKGROUND: Studies examining the benefits versus harms of coffee consumption have yielded conflicting results.
METHODS: The authors, from the United Kingdom, present an umbrella review of meta-analyses published up to 2017 to determine the associations between coffee consumption and any health outcome in adults.
RESULTS: This review includes 201 meta-analyses of observational studies with 67 health outcomes and 17 meta-analyses of randomized trials with 9 health outcomes according to coffee consumption defined as high versus low, any versus none, or per each extra cup per day. Coffee was associated with statistically significant protective effects against several diseases including type 2 diabetes, renal stones, Parkinson’s disease, Alzheimer’s disease, depression, leukemia, gout, colorectal cancer, liver cancer, chronic liver disease, cirrhosis, and endometrial cancer (odds ratios of 0.35-0.94). For some outcomes including all-cause mortality, cardiovascular mortality and cardiovascular disease, a nonlinear association was found whereby 3-4 cups per day (versus 0) conferred the greatest risk reduction (by 17%, 19% and 15%, respectively). High versus low consumption reduced the risk of incident cancers by 18%. Significant harms included lung cancer, urinary tract cancer, pregnancy loss, low birth weight, preterm birth, acute leukemia in childhood, and fracture risk in women (odds ratios of 1.03-1.57). Many of the harms were mitigated after adjustment for smoking, except the risks during pregnancy. This analysis of observational trials cannot prove causality.
CONCLUSIONS: Moderate levels of coffee consumption appear to be safe or even beneficial, except during pregnancy and in women at high fracture risk. 132 references (email@example.com – no reprints)