Green tea and matcha: modest heart signal, and a liver caveat

Drinking green tea is associated with a modestly lower risk of dying from cardiovascular disease, but the evidence is observational and the trials that measured hard outcomes found none. Concentrated green tea extracts are a separate matter: they have caused liver injury at catechin doses people take deliberately, which is why the strongest claim the data support is for the drink, not the pill [s1][s3].

What the drinking studies show

The most influential evidence is a Japanese cohort. The Ohsaki study followed 40,530 adults aged 40 to 79 with no history of stroke, heart disease or cancer, for up to 11 years for all-cause mortality and 7 years for cause-specific death [s1]. Green tea intake tracked with lower mortality, and the association was stronger in women and stronger for cardiovascular than for all-cause death [s1]. Among women, all-cause mortality hazard ratios across rising intake were 1.00 for less than 1 cup a day, 0.98 (95 percent confidence interval 0.84 to 1.15) for 1 to 2 cups, 0.82 (0.70 to 0.95) for 3 to 4 cups, and 0.77 (0.67 to 0.89) for 5 or more cups (P < .001 for trend) [s1]. The gradient in men was shallower but pointed the same way — an all-cause hazard ratio of 0.88 (0.79 to 0.98) at 5 or more cups a day versus less than one (P = .03 for trend) [s1]. For cardiovascular death in women the association was steeper still — 0.69 (0.52 to 0.93) at 3 to 4 cups and 0.69 (0.53 to 0.90) at 5 or more (P = .004 for trend) — while cancer mortality showed no significant association at any intake level in either sex [s1].

That pattern is consistent and biologically plausible, but it is an association in people who chose how much tea to drink. Heavy tea drinkers may differ from light ones in ways a statistical model cannot fully remove, and the study's own null result for cancer is a reminder that the signal is specific, not a blanket health halo [s1].

What the trials add — and don't

Randomised trials are where an observational signal is meant to be confirmed, and here they thin out. A Cochrane review of green and black tea for the primary prevention of cardiovascular disease pooled 11 randomised trials with 821 participants, of which 7 tested green tea, and found that no study reported cardiovascular events at all — the trials were too short and small to measure them [s2]. What they could measure were risk markers. Green tea produced statistically significant reductions in total cholesterol (mean difference −0.62 mmol/L, 95 percent confidence interval −0.77 to −0.46), LDL cholesterol (−0.64 mmol/L, −0.77 to −0.52) and blood pressure (systolic −3.18 mmHg, −5.25 to −1.11; diastolic −3.42 mmHg, −4.54 to −2.30) [s2].

Those are real but modest changes in intermediate outcomes, and the reviewers cautioned that only a small number of trials contributed to each estimate, the studies were at risk of bias, and the results did not hold up under sensitivity analysis [s2]. A drop in LDL is not the same as a drop in heart attacks, a distinction that recurs across nutrition — as it does with potassium and blood pressure and the DASH diet. The honest summary is that the cohort signal for the drink is real but unconfirmed by trials, and the trials that exist measured cholesterol and blood pressure rather than the outcomes that matter to a reader — heart attacks, strokes and deaths [s1][s2].

Matcha, extracts and the liver

Matcha — whole powdered green tea leaf whisked into water — delivers a higher catechin dose per serving than a steeped cup, and it sits between the drink and the supplement. The concentrated end of that spectrum carries a documented harm. A United States Pharmacopeia expert review of green tea extracts concluded that they can cause a hepatocellular pattern of liver injury, with published case reports associating hepatotoxicity with intakes of the catechin EGCG ranging from 140 mg to roughly 1,000 mg a day, and with large person-to-person variation in susceptibility that may be partly genetic [s3]. Taking extract as a bolus on an empty stomach raises catechin absorption, apparently by saturating first-pass elimination, which is why USP added a cautionary label to its green tea extract monograph advising people to take it with food and to stop and seek care if they develop signs of liver trouble such as abdominal pain, dark urine or jaundice [s3]. It is the same supplement-versus-food lesson seen with turmeric and curcumin.

How to read this

For a drink, green tea's safety record is reassuring and its association with lower cardiovascular death is real if unproven as cause; the coffee literature shows the same shape, as our piece on coffee and type 2 diabetes sets out, and both beverages interact with the gut, covered in coffee, alcohol and the gut lining. The concentrated-extract capsule is where the risk-benefit balance shifts. This article is informational and is not medical or dietary advice.