Is coffee good for your health?

Most meta-analyses published since 2015 (BMJ, JAMA Internal Medicine, Harvard T.H. Chan School of Public Health) report that moderate intake — 3 to 5 cups per day, up to 400 mg caffeine — is associated with lower all-cause mortality, lower risk of type 2 diabetes and of several liver diseases. Coffee is neither a drug nor a cure, but for most healthy adults the benefit-risk balance sits on the positive side when consumed in moderation.

The strongest evidence comes from large epidemiological cohorts and pooled meta-analyses. The BMJ 2017 umbrella review by Poole and colleagues screened 201 meta-analyses and concluded that 3-4 cups per day was associated with the greatest relative benefit: a 17 % lower all-cause mortality, a 19 % lower cardiovascular mortality, and reduced incidence of type 2 diabetes, liver disease (cirrhosis, hepatocellular carcinoma) and some cancers (endometrial, melanoma). A Harvard T.H. Chan meta-analysis (Circulation, 2015) further showed that moderate coffee drinking correlated with reduced mortality in smokers and non-smokers alike.

The likely mechanisms are plural. Coffee is the leading source of polyphenols in the European diet — chlorogenic, caffeic and ferulic acids — with antioxidant and anti-inflammatory activity documented in vitro. It also contains diterpenes (cafestol and kahweol, concentrated in unfiltered brews such as French press and Turkish coffee), whose effect on LDL cholesterol is less favourable; this is why cardiologists generally prefer paper-filtered coffee. Caffeine itself temporarily improves insulin sensitivity in some people, which may partly explain the lower type-2-diabetes risk.

Several caveats apply. These studies are largely observational: they reveal correlation, not causation. Genetics modulates the picture — in certain studies (JAMA, 2006), CYP1A2 slow metabolisers showed a slight uptick in cardiovascular risk at very high doses, while fast metabolisers did not. Preparation style matters: unfiltered coffee consumed in volume can raise LDL, while paper-filtered coffee does not. And benefits evaporate if the cup is loaded with sugar, syrup or industrial cream — what was studied is coffee, not a dessert latte.

In Belgium, where generations have drunk 4-6 cups of paper-filtered coffee daily — still common in the Walloon Brabant countryside — the evidence is broadly reassuring, provided sleep and underlying health stay in good shape. This FAQ summarises the web-verifiable literature (EFSA, BMJ, JAMA, Harvard, Mayo Clinic) but is not medical advice: please speak to a healthcare professional about any personal concern.

Coffee and health: what meta-analyses report

Summarising the evidence: what the science confidently says

The most defensible summary of coffee's health evidence as of 2026 is: moderate habitual coffee consumption (3–5 cups daily) is associated with net health benefits for the majority of healthy adults, is not associated with increased risk of most common diseases, and the specific populations for whom restrictions are warranted (pregnant women, those with specific cardiac or anxiety conditions, children) are identifiable and the restrictions are proportionate to documented risks. This summary would have been controversial 20 years ago — early epidemiological research was dominated by confounded positive associations between heavy coffee consumption and smoking, which created a misleading picture of coffee as a cardiovascular risk factor. Decades of improved research methodology have substantially reversed this picture.

The evidence is strongest for several specific health associations: reduced risk of type 2 diabetes (the most robustly replicated finding across the largest number of studies and populations), reduced risk of several liver conditions (cirrhosis, non-alcoholic fatty liver disease, hepatocellular carcinoma — for which the evidence is remarkably consistent), reduced risk of Parkinson's disease, and reduced all-cause mortality at 3–5 cups daily. These associations have survived multiple methodological improvements, correction for confounders, and replication across diverse populations with different genetic backgrounds, dietary patterns and healthcare systems. The biological mechanisms (anti-inflammatory effects of polyphenols, anti-fibrotic effects of coffee compounds in liver tissue, neuroprotective adenosine antagonism) are plausible and partially elucidated.

Going deeper

The areas where coffee's health evidence remains genuinely uncertain include: effects on bone density (evidence is mixed and may depend on calcium intake), effects on fertility in both men and women (weak, inconsistent associations across studies), effects on mental health beyond the acute anxiety question (some evidence for reduced depression risk, but causation uncertain), and optimal dose for specific health outcomes (the dose-response curves suggest diminishing returns above 3–5 cups but are noisy at higher intake levels). Honest communication about coffee's health effects requires representing both the confident findings (the liver and diabetes associations are solid) and the uncertain ones (bone density, fertility) rather than either dismissing coffee health benefits as marketing or overstating them as established medicine.