The claim: resveratrol, the molecule in red wine, is a longevity supplement. It "activates the sirtuins," mimics calorie restriction, and slows aging. Search for it and you'll find it sold as an anti-aging capsule with mouse studies behind it.
Here is what the human evidence actually shows.
1. The famous studies were in mice, on a specific diet, at huge doses.
The hype starts in 2006: Baur et al. (Nature, PMID 17086191) reported that resveratrol "improves health and survival of mice on a high-calorie diet." That qualifier matters. The mice were on a high-calorie, obesogenic diet, and resveratrol shifted them back toward a normal-diet physiology. It was not shown to extend life in normal mice.
2. When the rigorous mouse test ran, resveratrol failed.
The US National Institute on Aging runs the Interventions Testing Program (ITP): the same compound, three independent sites, genetically heterogeneous mice, and they publish negatives on purpose. Miller et al. 2011 (J Gerontol A, PMID 20974732): "Rapamycin, but not resveratrol or simvastatin, extends life span of genetically heterogeneous mice." Resveratrol at 300 and 1200 ppm did not significantly affect survival in males or females.
Strong et al. 2013 (J Gerontol A, PMID 22451473) ran resveratrol alongside green tea extract, curcumin, oxaloacetic acid, and MCT oil, lifelong from 4 months. None of the five extended lifespan. So even before we get to humans, the lifespan signal did not survive better-designed mouse studies.
3. In humans, only metabolic surrogates have been tested, and mostly they come up empty.
No human trial has shown that resveratrol slows aging. What they measured was blood sugar, insulin, lipids. And the results are inconsistent at best.
- Timmers et al. 2011 (Cell Metab, PMID 22055504): 11 obese men, 150 mg/day for 30 days. It reduced resting and sleeping metabolic rate and improved muscle mitochondrial function. Real, but n=11, obese men only, a metabolic signal not an aging one.
- Yoshino et al. 2012 (Cell Metab, PMID 23102619): 12 weeks, 75 mg/day, nonobese postmenopausal women. No change in body composition, resting metabolic rate, lipids, or inflammatory markers, and no effect on the molecular targets (AMPK, SIRT1, NAMPT) in muscle or fat.
- Bo et al. 2016 (Pharmacol Res, PMID 27520400): the biggest one. 192 type 2 diabetics, 6 months, 40 and 500 mg/day. No significant difference in weight, BMI, waist, blood pressure, fasting glucose, HbA1c, insulin, or several inflammatory markers. Total cholesterol and triglycerides slightly increased at the 500 mg dose.
- Liu et al. 2014 (Am J Clin Nutr, PMID 24695890), meta-analysis of 11 RCTs: resveratrol reduced fasting glucose, insulin, HbA1c, and insulin resistance in diabetics, and had no significant effect in nondiabetics.
The newer pooled data keep landing on the same spot. Ghalichi et al. 2026 (Daru, PMID 42268476), an umbrella review of 10 meta-analyses in type 2 diabetes: no significant effect on fasting blood sugar, HbA1c, insulin, HOMA-IR, LDL, HDL, triglycerides, or total cholesterol. Simental-Mendia et al. 2026 (Nutr Metab Cardiovasc Dis, PMID 42471257), 7 RCTs: no effect on apoB or apoA-I, two standard cardiovascular risk markers.
One caution from a different disease: Turner et al. 2015 (Neurology, PMID 26362286) gave resveratrol to 119 people with mild-to-moderate Alzheimer's, escalated to 1000 mg twice daily for a year. Resveratrol and its metabolites reached the brain. But brain volume loss was increased in the resveratrol group compared with placebo. That is not an aging study, and the biomarker interpretation is genuinely contested, but it is the opposite of the direction the marketing implies.
4. The dose trap: you cannot drink your way there, and mouse doses do not translate.
Walle et al. 2004 (Drug Metab Dispos, PMID 15333514) fed six volunteers a dietary-relevant 25 mg of resveratrol. Absorption was at least 70%, but unchanged resveratrol in plasma was under 5 ng/mL. Almost everything circulating was sulfate and glucuronide metabolites, not resveratrol. That is the core problem: rapid conjugation by gut and liver limits how much active compound reaches tissues.
The wine pitch is worse. In one human study the wines used contained roughly 0.8 to 3.2 mg of free trans-resveratrol per liter (reported in a 2015 review, PMID 26221416). Trial doses ran from 75 mg/day up to 2000 mg/day. Even the lowest trial dose would mean drinking tens of liters of wine a day, which is not a health strategy.
And the mouse doses were enormous relative to body weight, which is exactly the translation problem Reagan-Shaw et al. 2008 (FASEB J, PMID 17942826) flagged: the animal doses behind the original headlines do not extrapolate to humans by simple body-weight scaling, and the popular-press concern was well founded.
5. Where that leaves it.
Resveratrol is safe and generally well tolerated at trial doses (nausea, diarrhea, and weight loss show up at the high end). In specific metabolic disease populations it may have small effects, and the evidence for even that is inconsistent. What the human data do not show is any effect on aging.
What would change the verdict: a human randomized trial with an actual aging endpoint (an epigenetic clock, frailty, or mortality), long enough and large enough to detect a real effect, rather than another 12-week metabolic surrogate. Until then, "slows aging" is a claim from mouse studies that the rigorous mouse studies already failed to replicate.
Bottom line: the claim overshoots. Resveratrol has real pharmacology and a real story behind it. That story is about obese mice, high doses, and a compound that barely reaches your bloodstream intact. It is not a human anti-aging result.
Sources, all free to read: PMIDs 17086191, 20974732, 22451473, 22055504, 23102619, 27520400, 24695890, 42268476, 42471257, 26362286, 15333514, 26221416, 17942826.
Got a receipt that says otherwise? Reply below (no account needed) or email 7ens@ilands.app. I check claims against the primary source and correct in public.