A good daily tea. The benefits are real but small, and the concentrated extract capsules carry a liver risk the tea itself doesn't.
Mechanisms, studies and evidence grades for each claim above.
Green tea is Camellia sinensis leaf that is steamed (Japan) or pan-fired (China) soon after picking. The heat kills the leaf enzymes, so the catechins stay intact instead of being oxidized into the theaflavins of black tea or fermented into the theabrownins of pu-erh. The main catechin is EGCG (epigallocatechin gallate), and nearly every mechanism claim about green tea is a claim about EGCG. That fits the rule this index follows elsewhere: split entries by compound class. The three teas come from one plant but deliver three different sets of molecules.
The most important distinction on this page is tea versus extract. A cup of brewed tea delivers a modest amount of EGCG with food-level exposure. An extract capsule can deliver the EGCG of many cups at once, often on an empty stomach. Most of the benefit data is on the lower-dose side, and all of the liver harm is on the higher-dose side (see Safety).
This is an adjudication entry, like orange juice and mastic. The source is a September 2026 thread by Siim Land (@siimland), a health author whose bio links a paid coaching community. I found no green-tea product tied to the thread. The opening post calls green tea "one of the healthiest beverages in the world" and lists six benefits, followed by six posts citing a study each. He says he drinks 1–3 cups a day himself. Every PMID he gave was checked against its own abstract.
| Claim | What the literature shows | Does it hold up? |
|---|---|---|
| "A 2023 meta-analysis of 55 RCTs… lowers cholesterol, LDL, blood sugar, HbA1c and blood pressure. All results were statistically significant." | Zamani et al. 2023 is real and says roughly this.1 Pooled effects: LDL −5.8 mg/dL, fasting glucose −1.7 mg/dL, HbA1c −0.15 points, HDL +1.9 mg/dL. For blood pressure the abstract reports only diastolic (−0.87 mmHg). Systolic is not among the significant results. The intervention was mostly green tea extract, with subgroups split at 1,000 mg/day. | Holds, but small and overstated. The direction is consistent. "All significant" is not quite right, since systolic pressure isn't listed. And these are extract doses, not cups of tea. |
| "A 2011 meta-analysis of 14 RCTs… lowers total and LDL cholesterol, independent of health status." | Zheng et al. 2011, AJCN, 1,136 people.2 Total cholesterol −7.2 mg/dL, LDL −2.2 mg/dL, HDL unchanged. The result was not affected by health status, dose, tea versus extract, or study quality. | Holds. The LDL drop is about 2 mg/dL, around 1–2% for most people. |
| "A 2014 meta-analysis of 13 RCTs… lowers systolic and diastolic blood pressure." | The PMID given (36704803) is the 2023 paper from the first row, not a 2014 study. The paper he describes exists: Khalesi et al. 2014, Eur J Nutr.3 Systolic −2.1 mmHg, diastolic −1.7 mmHg, larger in people starting at 130+ and in extract trials. | Holds, with a wrong citation. He calls it "modest" himself. Two mmHg is not a reason to drink anything. |
| "Exercise combined with green tea catechins results in significantly greater loss of visceral fat." | Maki et al. 2009, J Nutr, 12 weeks, 625 mg catechins in a drink, with an exercise program.4 Total abdominal fat area (−7.7% vs −0.3%) and subcutaneous abdominal fat fell more with catechins. Visceral fat did not differ significantly between groups. Total fat mass didn't differ either, and the weight difference was a non-significant trend. The trial was funded by Kao Corporation, which sells catechin-enriched tea (Healthya) in Japan, and five of its twelve authors were Kao employees. The PMID in the post is also missing its last digit (19074207). | Does not hold as stated. The one outcome in the headline, visceral fat, is the one the cited trial did not find. The real result is a small drop in belly fat under the skin, from an industry-run trial. |
| "A meta-analysis of 18 studies: 4 cups/day = 10–15% lower all-cause mortality." | Tang et al. 2015, Br J Nutr, 18 prospective cohorts.5 Highest vs. lowest green-tea intake: all-cause mortality RR 0.80, cardiovascular mortality RR 0.67. Each extra cup a day went with 4% lower all-cause and 5% lower cardiovascular mortality, so 4 cups lands near his figure. Cancer mortality: RR 1.06, no association. | Holds, as an association. It's observational, and the cohorts are mostly Japanese and Chinese. The missing cancer signal is worth noticing given the next claim. |
| "Targets all the hallmarks of aging" (genomic instability, telomeres, autophagy, mitochondria and so on) | The graphic comes from Yilmaz 2025, Aging and Disease, a single-author narrative review.6 The abstract says the evidence comes "from in vitro, animal, and human studies," and the human part is about outcomes such as cognition and cardiovascular health, not about the hallmarks themselves. | Mechanism only. "Targets" means EGCG changes these pathways in cells and mice. Nobody has measured all twelve hallmarks shifting in people drinking tea. See Slows cellular aging. |
| "Improves heart health" (the opening graphic) | The diagram comes from a September 2023 Preprints.org review (Akter et al., DOI 10.20944/preprints202309.0582.v1). It was not peer reviewed when posted and has no new data. It is a mechanism cartoon: NO, eNOS, antioxidant enzymes, then "no atherosclerosis." | Illustration, not evidence. The vessel mechanism is plausible and matches what black tea shows in FMD trials (see Assam), but the graphic proves nothing by itself. |
Overall: the thread is mostly right about direction and wrong about size.
Every lipid, glucose and blood-pressure effect he cites is real and small. The mortality association is real and is the strongest point in his favor. The one claim that clearly fails is the visceral-fat one, which his own citation contradicts. The "hallmarks of aging" claim is the one with the most reach and the least human evidence. What the thread leaves out is the fact that matters most for someone taking green tea as a supplement: high-dose extracts can injure the liver. Also, "one of the healthiest beverages in the world" isn't a green-tea-specific claim. Black tea shows a similar mortality association in the UK Biobank (see Assam).
↑ back to summaryThe best single cohort is the Ohsaki study: 40,530 Japanese adults followed for up to 11 years.7 Compared with less than a cup a day, 5+ cups went with lower all-cause mortality: hazard ratio 0.88 in men and 0.77 in women. The association with cardiovascular death was stronger, strongest of all for stroke. Cancer mortality was not lower in any category. The Tang meta-analysis above pools this with 17 other cohorts and gets the same shape.5 (Moderate — large, consistent cohorts; still observational)
Two caveats. First, in Japan, drinking a lot of green tea partly marks a traditional diet and lifestyle, and adjustment can't remove all of that. Second, the benefit concentrates in the cardiovascular and stroke deaths, which fits the vessel-function mechanism shared with black tea. It doesn't fit a general "anti-aging" story, which would predict a cancer signal too.
↑ back to summaryThe trial evidence here is broad and consistent: 14 trials in the 2011 meta-analysis, 55 in the 2023 one.2,1 The effects are small. LDL falls by about 2–6 mg/dL, fasting glucose by under 2 mg/dL, and HbA1c by 0.15 points. The larger numbers come from extract doses. (Moderate for the direction; the size is small)
On LDL specifically, the Statins & LDL paper argues LDL is causal but an incomplete decision rule. A 2–6 mg/dL change is small by either standard. These trials show green tea is metabolically friendly. They don't make it a cholesterol treatment.
↑ back to summaryThis is the most commercially motivated part of the literature. Catechin-enriched drinks were sold in Japan with a government health claim, and much of the trial evidence was run by or with the companies selling them. The Maki trial in the thread is one example (Kao-funded, see the claim table).4
The independent summary is unexciting. A Cochrane review split the trials in two because they disagreed too much to pool.8 Six trials outside Japan pooled to a weight difference of −0.04 kg, which is zero, and a waist difference of −0.2 cm. Eight Japanese trials ranged from −0.2 to −3.5 kg but were too different to pool. An earlier meta-analysis found about −1.3 kg overall, a smaller effect in white than Asian participants, and hints that heavy habitual caffeine users benefit less.9 (Moderate that a small effect exists; Contested on whether it matters outside Japanese trials)
The most likely mechanism is catechins and caffeine together slowing the breakdown of noradrenaline, which modestly raises fat burning. That also explains why habitual heavy caffeine users gain less. (Inferred; mechanism not cited here)
↑ back to summaryEGCG does a lot in a dish. It activates AMPK and sirtuins, inhibits mTOR, turns on Nrf2, turns down NF-κB, and promotes autophagy and mitophagy. That is the Nrf2-up/NF-κB-down theme that recurs across this index (see Black Seed). The Yilmaz review maps these effects onto all twelve hallmarks of aging.6 (Animal–in-vitro)
The problem is dose and exposure. EGCG is poorly absorbed and quickly conjugated. In a phase I study of single 200–800 mg EGCG doses, plasma exposure rose more than proportionally with dose (a 4× dose gave about 7× the exposure), and the other catechins circulated mostly as glucuronide and sulfate conjugates rather than the free forms tested in cell work.10 Many cell studies use concentrations that a cup of tea, or even a capsule, does not reach in blood. So a pathway shift in a dish is not evidence that tea shifts it in a person. (Inferred for any human anti-aging claim)
The human outcome data point the other way from "targets all hallmarks." Hallmarks such as genomic instability predict less cancer, and the cohorts show no cancer-mortality benefit from green tea.5,7 The benefit shows up where the vessel mechanism predicts it: heart disease and stroke.
↑ back to summaryThe clearest single trial is the Minnesota Green Tea Trial: 1,021 women randomized to green tea extract or placebo for a year.12 ALT rose by 5.4 U/L on extract. 5.1% of women developed moderate or worse liver-test abnormalities, an odds ratio of 7.0 versus placebo. Enzymes returned to normal when the extract was stopped and rose again on rechallenge, which is about as close to proof of cause as a supplement trial gets.
The US Pharmacopeia's review found case reports of liver injury at EGCG intakes as low as 140 mg a day, up to about 1,000 mg, with large differences between people that are probably genetic.13 It found that taking extract while fasting substantially raises EGCG exposure. USP now requires its green-tea-extract monograph to carry a label warning: take with food, don't use with liver problems, and stop if you get abdominal pain, dark urine or jaundice. Contaminants such as solvent residues and pesticides were looked for and not implicated. The catechin itself is the problem.
Caffeine: green tea usually has less than black tea, and gyokuro and matcha have the most. Cut back in the afternoon if sleep is fragile. Decaf extracts remove the caffeine but not the liver issue.
↑ back to summaryAll 13 references were checked against their PubMed abstracts in this session, including every PMID in the source thread. The Kao funding and the visceral-fat null in Maki 2009 come from the published paper's disclosures and results, not the abstract. Still uncited: the brewing temperature, the per-cup caffeine comparison, the noradrenaline mechanism for fat-burning, and the claim that many cell studies use EGCG concentrations above what blood reaches. The preprint behind the opening graphic (Akter et al. 2023) is identified by DOI only.