Tea
Daily tea is associated with modest reductions in cardiovascular mortality, dementia risk, and bone loss — smaller effects than coffee's, but consistent in direction across decades of studies and with an excellent safety profile. The one habit worth changing: let very hot tea stand for several minutes before drinking, because scalding-hot liquids raise the risk of esophageal cancer.
Tea — Camellia sinensis — is one of the most widely consumed beverages in the world, and the longevity literature on it has matured substantially over the past few years. The honest summary: tea is probably good for you, the effect sizes are smaller than coffee's, and most of the rest is observational and partly confounded by the fact that habitual tea drinkers in Western cohorts tend to be older, less likely to smoke, and more health-conscious than non-drinkers. This article walks the evidence by outcome, then by tea type, then by the practical points that matter — caffeine load, iron interaction, the green-tea-extract hepatotoxicity issue, and what to do with the L-theanine claims.
What's in a cup
Caffeine and bioactive content vary substantially by leaf processing.
| Type | Caffeine (240 mL / 8 oz cup) | Dominant polyphenols |
|---|---|---|
| Green tea | 25–50 mg | Catechins, particularly epigallocatechin gallate (EGCG) |
| Black tea | 40–70 mg | Theaflavins and thearubigins (oxidation products of catechins) |
| Oolong | 30–50 mg | Mixed catechins and partial oxidation products |
| White tea | 15–30 mg | Catechins (less processed than green) |
| Matcha (whole leaf, 1 tsp) | 60–80 mg | Catechins (higher density because the whole leaf is consumed) |
| Decaf tea | 2–5 mg | Catechin and theaflavin content not well characterised |
These bands are approximations, and the overlap between types is larger than the table implies: an analysis of 20 commercial teas found 14 to 61 mg per serving across white, green and black with no consistent trend by variety, because steeping time matters more than the type does.[1]
Other notable constituents:
- L-theanine — a non-protein amino acid found almost only in tea (roughly 5–30 mg per cup, varying with steeping time; black and green teas carry the most)[2]; crosses the blood-brain barrier and modulates glutamate, gamma-aminobutyric acid (GABA), and the alpha brain waves associated with relaxed attention.
- Tannins — a class of polyphenols that bind dietary minerals (especially non-heme iron from plant foods) in the gut. Black tea typically inhibits iron absorption more than green, mainly because it is usually brewed stronger and drunk in larger volume.
- Fluoride — tea plants accumulate fluoride from soil; brick teas (rare in Western consumption) can deliver substantial doses.
Caffeine in tea is metabolised by the liver enzyme CYP1A2 with the same half-life as in coffee — about 5 hours on average, but anywhere from roughly 1.5 to 9.5 hours between individuals — so the chronotype caveats from the Coffee article apply directly.
Mortality and health outcomes
All-cause and cardiovascular mortality
Recent large-cohort and meta-analytic data consistently show modest mortality reductions at moderate tea intake (typically defined as ~2 cups/day). Effect sizes are smaller and the statistical uncertainty greater than for coffee, but the direction is consistent. Moderate evidence.
- In the UK Biobank (about 498,000 adults, mostly black-tea drinkers, tracked for a median 11 years), drinking 2 or more cups per day was associated with 9–13% lower all-cause mortality, with the largest reduction — about 13% — at 2–3 cups per day.[3] Tellingly, this held regardless of genetic variation in caffeine metabolism, and regardless of milk or tea temperature — supporting the polyphenols-not-caffeine reading below. Added sugar is less settled: two 2025 re-analyses of the same cohort disagree, one finding the benefit only in unsweetened tea and the other finding it in both.[4][5]
- A 2024 meta-analysis pooling 38 cohort datasets (nearly 2 million people) put the highest-versus-lowest tea intake at about 10% lower all-cause mortality and 14% lower cardiovascular mortality; cancer mortality was not significantly reduced in that highest-versus-lowest comparison, though the dose-response analysis in the same paper found a significant reduction from about 1.5 cups a day.[6] The benefit plateaued around 1.5–3 cups per day for cardiovascular death, though the reduction kept deepening at higher intakes. The pooled figure also hides a sharp geographical split: all-cause mortality was 16% lower in Asian cohorts but showed no benefit at all in European ones, a difference that was borderline significant — which matters for a mostly black-tea-drinking Western readership.
- Black tea and coronary heart disease: risk keeps falling across the whole range studied, with no plateau — about 5% lower at 2 cups a day and about 16% lower at 10, though the overall highest-versus-lowest estimate is a more modest 11% and the association appeared in European but not US cohorts.[7]
- Green tea and coronary heart disease: the benefit is greatest at low-to-moderate intake — best around 600 to 900 mL a day, two to three of the 300 mL cups these studies measured, and no longer significant past about 1.5 litres — so the useful range tops out roughly where the advice below already sits.[8]
- The case for polyphenols rather than caffeine rests on the fact that the mortality benefit holds regardless of how fast a person metabolises caffeine, not on decaffeinated tea — too few people drink decaf tea for cohorts to analyse it separately.[9]
A causal caveat worth keeping in view: Mendelian-randomization studies, which use genetic variants as a natural experiment to test causation, have been conflicting for cardiovascular disease. Two found no causal effect of genetically predicted tea intake on coronary artery disease, myocardial infarction, atrial fibrillation or heart failure,[10][11] while a third found tea protective against high blood pressure, heart failure and ischaemic stroke, with blood pressure appearing to carry the effect.[12] All of them draw their genetic instrument from the same UK Biobank data, so they are not independent replications of one another. The genetic evidence does not settle the question either way, and some of the observational signal may still be confounding.
A common interpretive caveat applies: in Western cohorts, tea drinkers tend to be older, female, less likely to smoke, and more health-conscious; in Asian cohorts (where green tea dominates), the inverse holds in different ways. Most studies adjust for these, but residual confounding is plausible.
Dementia and cognitive decline
This is currently tea's strongest individual finding — with one counterweight: the only Mendelian-randomization study of tea and dementia points the other way, linking genetically predicted tea intake to higher Alzheimer's risk, though its genetic instruments are weak.[13]
- A 2026 analysis in JAMA — pooling the Nurses' Health Study and the Health Professionals Follow-Up Study, tracking 131,821 adults over up to 43 years of follow-up, with just over 11,000 dementia cases — found that the highest third of tea drinkers (a median of about one cup a day) had roughly 14% lower dementia risk than the lowest third (hazard ratio 0.86, 95% confidence interval 0.83 to 0.90; the interval is the range the true value is most likely to lie in, and because it sits entirely below 1.0 the reduction is unlikely to be chance); the widely repeated 18% figure applies to caffeinated coffee, not tea (hazard ratio 0.82), at 2–3 cups per day.[14] The tea benefit was strongest at 1–2 cups per day, and decaffeinated coffee showed no association with dementia. On the cognitive tests, given to the women of the Nurses' Health Study, the highest third of tea drinkers scored slightly better on all three objective measures while caffeinated coffee was null on two of them — though every difference was tiny. Moderate evidence: these are large, decades-long prospective cohorts, but observational, in health-professional populations, with self-reported intake.
- Older meta-analyses report tea-associated reductions of around 12% for Alzheimer's (relative risk 0.88) and 25% for vascular dementia (relative risk 0.75) — consistent with the 2026 finding, but methodologically weaker, and the vascular-dementia figure rests on only two cohorts. Whether green or black tea does more is untested: the studies pooled here mostly did not record which, and the newest synthesis finds comparable results across tea types.[15][16]
- 2025 brain-imaging cohorts report associations between habitual green-tea intake and lower cerebral white-matter lesion burden — biologically plausible (catechin antioxidant effects on cerebrovascular endothelium) but cross-sectional and exploratory, and in one of them, of about 8,800 older Japanese adults, only one of three imaging outcomes moved: hippocampal and total brain volume showed no association.[17]
Type 2 diabetes and metabolic syndrome
Habitual tea consumption — typically 3–4 cups per day — is associated with modest improvements in fasting glucose, lipid profile, and blood pressure across multiple cohorts. The type 2 diabetes signal is dose-dependent and concentrated at high intake: pooled cohort data show no overall effect, but 4 or more cups per day is associated with roughly 17% lower diabetes risk.[18] A 2025 UK Biobank analysis of 382,946 adults found the signal starting earlier and levelling off higher — about 7% lower risk at 2–3 cups a day and 14% at 4–5 — and traced about half of the association to differences in body weight, not to tea's effects on blood pressure or cholesterol, which mediated under 10% between them.[19] Randomised trials also show tea nudges cardiometabolic markers modestly: green-tea catechins lower systolic blood pressure by about 2 mmHg — larger in people whose pressure is already elevated, and larger when the catechins are taken as a concentrated extract than as a brewed cup,[20] with the largest and most recent synthesis putting the effect nearer 1 mmHg for both systolic and diastolic pressure,[21] and lower low-density lipoprotein (LDL) cholesterol by about 0.12–0.14 mmol/L, with high-density lipoprotein and triglycerides unchanged.[22][23] Moderate evidence for these small effects. EGCG has plausible mechanisms, though almost all of them are cell or rodent work: it activates the cellular energy sensor AMP-activated protein kinase (AMPK) in cultured cells, and improves insulin sensitivity in diabetic rats at doses equivalent to well over a gram of EGCG in a person. The one human demonstration is narrower — a concentrated green tea extract measurably slowed starch digestion in a 28-person crossover.[24] Human randomised-trial effect sizes are small — clinically marginal in people whose numbers are already normal. Tea is useful as part of a broader dietary pattern; it is not a stand-alone metabolic intervention.
Cancer
The cancer literature on tea is genuinely mixed. Green tea catechins inhibit tumor growth in cell and animal models, and the clearest of the human signals is for stomach cancer, where a pooling of 22 studies found regular tea drinkers at modestly lower risk overall and a stronger association in China and Japan specifically, where green tea is the norm[25] — but Western cohorts often fail to replicate, and high-dose green tea extract supplements have been linked to rare hepatotoxicity. A 2020 Cochrane review of 142 studies (over 1.1 million people) found the evidence for a green-tea cancer-prevention benefit limited and inconsistent, rating the certainty of every estimate as low — driven by the small number and size of the studies rather than by how they were run, and by results that flipped direction depending on study design for esophageal, prostate and urinary-tract cancer.[26] A separate dose-response meta-analysis even flagged a weak adverse signal — an 18% higher breast-cancer risk per 3 additional cups of black tea per day (relative risk 1.18) — though this was a four-study subgroup inside an analysis whose overall breast-cancer result was flatly null (relative risk 1.02 across 15 studies), and no later pooled analysis has reported the finding again. Worth noting rather than alarming over.[27] No strong, consistent cancer-prevention claim is supported for habitual tea drinking; treat any benefit as uncertain. Weak-preliminary evidence throughout.
Green vs. black: does it matter?
The two are produced from the same leaf at different oxidation levels. During oxidation (which makes a tea "black"), catechins polymerize into theaflavins and thearubigins. Both classes are bioactive antioxidants, but their pharmacology differs. Weak-to-moderate evidence — the split below reflects where each type was studied more than head-to-head comparison.
- Green tea (unoxidized): catechin-dominant. This is where most of the metabolic and cognitive research sits. EGCG is the most-studied tea catechin.
- Black tea (fully oxidized): theaflavin-dominant. This is where most of the vascular trial evidence sits — tea improves flow-mediated dilation, a measure of how well arteries widen, in randomised controlled trials (RCTs), and black tea is what those trials mostly used — and the meta-analysis pooling those trials was written by employees of a company that sells black tea.[28] No trial has shown black tea to beat green on a vascular endpoint.
- Oolong / white / pu-erh: intermediate or specialty processing; less data, no reason to think they're inferior on safety, weak evidence base for specific outcomes.
Practical view: if you drink one or the other for taste, the difference is unlikely to matter clinically. If you're choosing for a specific reason — green tea for metabolic/cognitive, black tea for vascular — the literature does not really support the conventional split; it reflects where each type happened to be studied. The handful of direct comparisons that exist are small, and they lean towards green rather than black.
The "Successful Aging Index" green tea claim (real but heavily confounded)
A frequently cited statistic from the 2019 ATTICA/MEDIS combined analysis (about 3,300 Greek adults over 50) reports that green tea drinkers had roughly 77% higher odds of scoring above the median on a validated Successful Aging Index than black tea drinkers (odds ratio 1.77).[29] The association is real, but weaker than it looks. It rests on 595 green-tea and 600 black-tea drinkers rather than the full sample; it is cross-sectional; the green tea group was also more physically active and less hypertensive at baseline — and neither of those was actually adjusted for, since the models controlled only for age, sex, smoking and coffee. Most of the gap comes from black tea drinkers scoring badly rather than green tea drinkers scoring well. The "successful aging" outcome is plausibly driven by a cluster of healthy lifestyle habits, not by green tea per se. Weak evidence — useful as a directional hint, not as a quantified causal claim. (The "36–50% more likely" figure circulating in popular write-ups is not in the source paper.)
L-theanine: real for acute focus, speculative beyond
L-theanine has a well-documented acute effect: in small brain-wave (electroencephalography, EEG) and behavioral trials, it boosts alpha brain waves and partially blunts caffeine-induced jitters, producing the "relaxed alertness" that tea drinkers describe and most coffee drinkers don't get. The attention effect is the best-supported part — pooled trials show faster reaction times after a single 200 mg dose — while the alpha-wave evidence rests on a handful of small single-lab studies, and pooled estimates for theanine combined with caffeine are less certain than for theanine alone; one review found the caffeine dose explained more of the effect than the theanine dose. The 100–200 mg doses these trials use are the equivalent of several to dozens of cups of tea, depending on strength. Moderate evidence for the acute focus effect; Weak / preliminary for anything longer-term.
The leap that often follows in popular writing — that this acute mechanism translates into long-term neuroprotection or dementia prevention — is not supported. The 2026 JAMA analysis could not separate caffeinated from decaffeinated tea, or green from black — so it cannot isolate what in the cup is doing the work. What it does show is that decaffeinated coffee carried no dementia benefit. No human trial has shown that tea's constituents protect the brain over years: a 12-month trial adding EGCG to a lifestyle programme in 129 people at genetic risk of Alzheimer's missed its primary endpoint,[30] and a 12-month matcha trial in 99 older adults with cognitive decline found no effect on cognitive screening scores or daily function.[31] Treat the acute focus benefit as the supported claim; treat long-term neuroprotection from theanine as a hypothesis.
For supplemental L-theanine specifically (sleep/anxiety dose, side effects), see Sleep supplements.
Iron and minerals
Tea is a stronger inhibitor of non-heme iron absorption than coffee — the relevant compounds are polyphenols, chiefly the galloylated catechins. Black tea inhibits more than green mainly because it is usually brewed stronger and drunk in larger volume, and peppermint tea inhibits iron absorption about as strongly.[32] Co-consumed with a meal or an oral iron supplement, strong black tea can cut iron uptake substantially.
Separating tea from iron-rich meals or supplements by at least an hour substantially reduces the interaction, though it does not abolish it: in a stable-isotope study in twelve women, tea drunk with a meal cut iron absorption by 37%, while tea drunk an hour afterwards cut it by 18% — roughly halving the effect.[33] Better still, drink it before the meal or between meals: the equivalent coffee experiment found that an hour beforehand removed the interaction entirely, while an hour afterwards was no better than drinking it with the meal.[34] The interaction is local to the gut lumen — tea does not deplete stored iron in healthy people on a varied diet — but matters for menstruating women, plant-based eaters, and anyone on iron therapy. For the opposite group — adults trying to avoid lifelong iron accumulation — the same effect is a mild dietary lever rather than a drawback; see Iron and aging. See the iron section in the coffee article for the same mechanism and the alternate-day iron-dosing strategy.
Bones: a 2025 meta-analysis of coffee and tea drinkers covering roughly 560,000 adults found tea drinkers had about 25% lower odds of osteoporosis (odds ratio 0.75) — but that figure rests on eight studies with wide disagreement between them, nine of the fourteen were cross-sectional, twelve came from China or Korea, and restricting the analysis to the stronger prospective designs halves the effect to about 13% (odds ratio 0.87). Coffee scored much the same in the same analysis (odds ratio 0.79), which is the pattern you would expect if the drinkers rather than the drink were doing some of the work.[35] A 2025 analysis of the Study of Osteoporotic Fractures (9,704 women aged 65+, followed with repeated bone scans over a decade) found higher total hip bone mineral density in tea drinkers, though the difference was small — 0.718 against 0.715 g/cm², under half a percent — and the abstract reports no femoral-neck result for tea.[36] The mechanism usually offered is that tea polyphenols restrain the bone-resorbing cells, but the one randomised human trial to look found the opposite pattern — a marker of bone formation rose while the resorption marker did not move.[37] The net signal is mildly bone-protective. Weak-to-moderate evidence: a twelve-month trial of high-dose green tea extract in postmenopausal women found no effect on bone density in its 121-woman scan substudy, and no trial has ever used fractures as an endpoint.[38]
Microbiome (plausible, mostly preclinical)
Green tea polyphenols are extensively metabolized by gut microbes and reciprocally shape microbial communities. Of particular interest is Akkermansia muciniphila, a mucin-layer commensal with strong links to metabolic health: a 2019 Nature Medicine study — a randomised pilot the authors themselves call proof-of-concept, with 32 people completing across three groups — found that pasteurized Akkermansia supplementation improved insulin sensitivity by about 29% and lowered total cholesterol by about 9%, while body weight and fat mass did not move. Two of its authors co-founded the company selling the product, and a published critique argues the metabolic results could reflect unadjusted confounding.[39]
The claim that habitual green tea drinking robustly proliferates Akkermansia in humans rests mostly on animal models and observational microbiome data. The one randomised test to look — 96 adults given green tea catechins with the prebiotic fibre inulin for twelve weeks — missed its primary endpoint (visceral fat), and no bacterial genus survived statistical correction, Akkermansia included.[40] The biology is plausible; the human effect size for tea specifically is unestablished. Weak / preliminary evidence.
Safety considerations
Very hot tea and esophageal cancer
This is the single most actionable tea-specific risk, and it is about temperature, not tea. In the Golestan cohort (about 50,000 adults, tea temperature measured objectively by trained staff), drinking tea at 60°C (140°F) or hotter carried a 41% higher risk of esophageal squamous-cell carcinoma than drinking it cooler, and a stated preference for "very hot" tea more than doubled the risk compared with drinking it cold or lukewarm — while the intermediate "hot" preference was not significantly associated at all.[41] Drinking large volumes very hot (700 mL or more per day at 60°C-plus) was linked to roughly 90% higher risk, and at 60°C-plus, smaller daily volumes showed no clear excess risk. The International Agency for Research on Cancer (IARC) classifies drinking beverages above 65°C (149°F) as "probably carcinogenic" for exactly this reason — thermal injury to the esophageal lining, not the drink itself.[42] The same review looked at coffee and moved it the other way, downgrading it to "not classifiable" after more than a thousand studies. Moderate evidence — IARC itself graded the human evidence "limited", the 60 °C threshold comes from a single cohort in a region where this cancer is 5 to 50 times more common than in Europe or the US (in Western countries most oesophageal cancer is a different cell type), and IARC notes the proportion of cases attributable to hot drinks is unknown. Practical takeaway: wait about six minutes after pouring before drinking — in the Golestan cohort, drinking within two minutes of pouring carried roughly 50% higher risk than waiting six minutes or more.[43] Comfortably warm is fine.
- Caffeine load. A heavy daily intake of strong tea (5+ cups of black tea or matcha) can put a sensitive person near or over the 400 mg/day caffeine limit recommended by both the European Food Safety Authority and the US Food and Drug Administration (FDA). In pregnancy, the 200 mg/day cap applies to total caffeine from all sources.
- Green-tea-extract supplements (Caution). Concentrated catechin pills have caused rare cases of idiosyncratic liver injury serious enough to require transplant. The European Food Safety Authority (EFSA) found that doses of 800 mg or more of EGCG per day raise liver enzymes in some people and could not identify any safe supplemental dose.[44] The injury is immune-mediated with a strong genetic susceptibility: about 72% of the genetically typed cases carried the HLA-B*35:01 immune variant, far above background.[45] Risk is highest when pills are taken fasted, which raises blood levels of EGCG sharply — and the same opinion found no evidence of liver harm below 800 mg a day for up to twelve months; brewed tea delivers far less.[46] Brewed tea itself is generally safe; EFSA notes rare liver reactions to it, most probably idiosyncratic. Mega-dose green-tea-extract pills are not equivalent and not recommended.
- Iron interaction — see above; matters only for those with marginal iron status (menstruating women, plant-based eaters, anyone on iron therapy).
- Fluoride. Typical brewed tea contributes a small amount of dietary fluoride and is fine at a few cups a day. Skeletal fluorosis has appeared only at extreme intakes — case reports involve one woman drinking tea brewed from 100 to 150 tea bags daily for seventeen years,[47] and another ingesting an estimated 37–74 mg of fluoride per day from gallons of double-strength instant tea,[48] against the US tolerable upper limit of 10 mg/day. Europe is more conservative — EFSA's 2025 reassessment set a safe intake of 3.3 mg/day, a level a heavy black-tea drinker could approach — though that figure is derived from effects on fetal brain development rather than from bone, and typical tea drinking is not considered a health risk. Brick tea (rare outside Tibet, Mongolia and western China) is the one genuinely concentrated source.
- Trace contaminants. Tea leaves accumulate metals from soil; analytical surveys find measurable lead and aluminium in brewed tea, occasionally above conservative guidelines — in a survey of 30 tea-bag products every brewed tea contained some lead, and 20% exceeded guidelines for aluminium, while arsenic and cadmium stayed within accepted standards.[49] The benchmark those lead figures are measured against is California's Proposition 65 threshold for reproductive toxicity, the most stringent in use, so "above guideline" here is a long way from "harmful". Countervailing work shows steeping actually pulls metals out of the water: brewing a single bag for three to five minutes removed about 15% of the lead from contaminated water.[50] At normal Western intakes this is not a clinical concern. Weak-preliminary evidence, no outcome data. A newer question is the bag rather than the leaf. Bags heat-sealed with polypropylene shed very large numbers of micro- and nanoplastic particles into the cup when brewed — roughly ten times more than plain cellulose bags — and those particles are taken up by human intestinal cells in the laboratory.[51][52] There is no human outcome data and no regulatory limit, so this sits at the same speculative level as the metals above — but the fix is free if it bothers you: loose leaf, a metal infuser, or a plain unbleached paper bag.
- Drug interactions. The one worth knowing about is real, well documented, and comes from ordinary brewed tea rather than supplements: drinking 700 mL of green tea daily for two weeks cut blood levels of the beta-blocker nadolol by about 85% and measurably blunted its blood-pressure effect, by blocking the intestinal transporter the drug needs to be absorbed.[53] Other drugs using the same transporter may behave similarly, though this has not been tested. Concentrated green tea polyphenols also antagonise bortezomib and related boronic-acid cancer drugs in laboratory and animal models — that is the cancer drug working less well, not a toxicity — with no human outcome data.[54] The often-repeated warfarin interaction rests on a single 1999 case report of a man drinking two to four litres of green tea a day, and was attributed to vitamin K rather than to catechins; it has never been reproduced.[55] If you take nadolol or a related beta-blocker, talk to your doctor before drinking green tea regularly: in a later study, even a single serving of brewed green tea an hour before the dose roughly halved how much drug was absorbed, so an hour's gap is not enough, and longer gaps have not been tested.
Practical recommendations
- 2–3 cups per day of green or black tea is the dose where most cohort signals converge. Higher (3–4 cups) is fine if tolerated.
- Match tea to taste, not to a specific health claim. The literature does not establish a meaningful green-vs-black difference; the larger benefit comes from drinking tea regularly versus not.
- Time around iron-rich meals or supplements if iron status matters to you (an hour or more before the meal or supplement, not with or after it; an hour after only halves the effect).
- Caffeine timing is the same as coffee: avoid within 8–10 hours of bedtime if you're a sensitive sleeper.
- Don't take green tea extract pills for longevity. Drink tea instead.
- Pregnancy: count tea toward the 200 mg/day caffeine ceiling.
- Don't start drinking tea purely for health benefits if you don't enjoy it — the effect sizes don't justify adoption against preference.
Further reading
- Zhang Y et al. Coffee and Tea Intake, Dementia Risk, and Cognitive Function — Nurses' Health Study and Health Professionals Follow-Up Study. JAMA 2026 (131,821 adults, up to 43-year follow-up).[56]
- Islami F et al. Tea drinking temperature and risk of esophageal squamous cell carcinoma — Golestan cohort. Int J Cancer 2020.[57]
- Inoue-Choi M et al. Tea consumption and all-cause and cause-specific mortality in the UK Biobank. Ann Intern Med 2022.[58]
- Kim Y & Je Y. Tea consumption and all-cause, cardiovascular, and cancer mortality — meta-analysis of 38 cohort datasets. Epidemiol Health 2024.[59]
- Filippini T et al. Green tea (Camellia sinensis) for the prevention of cancer. Cochrane Database Syst Rev 2020.[60]
- Li W et al. Coffee and tea consumption on the risk of osteoporosis — meta-analysis. Front Nutr 2025 (562,838 adults; tea odds ratio 0.75).[61]
- Liu RY & Liu E. Longitudinal Association of Coffee and Tea Consumption with Bone Mineral Density in Older Women — Study of Osteoporotic Fractures. Nutrients 2025 (9,704 women aged 65+).[62]
- Ahmad Fuzi SF et al. Tea timing and iron absorption. Am J Clin Nutr 2017.[63]
- Naumovski N et al. Green and black tea consumption and the Successful Aging Index — ATTICA / MEDIS combined analysis. Molecules 2019.[64]
- Depommier C et al. Pasteurised Akkermansia muciniphila in human metabolic syndrome. Nat Med 2019.[65]
For coffee-specific evidence (caffeine pharmacology, CYP1A2 metabolism, filtered-vs-unfiltered, microbiome, dementia), see Coffee →.