Resveratrol: the ingredient behind the headlines

A bottle says "red wine molecule". A friend mentions mice that lived longer. Somewhere at the back of your head sits a half-remembered "French paradox". Resveratrol has had an excellent press agent for thirty years, so let's find out how much of the story survives contact with human trials.

Resveratrol is a real, heavily studied compound, but the dramatic claims come from yeast, cells and mice. In people the trials are mostly small and short, and the results are mixed. There are mixed blood-pressure signals in some pooled analyses, inconsistent glucose results, no steady effect on memory in older adults, and no human data on lifespan at all. It is an interesting ingredient with a great origin story; the human evidence for anti-ageing has not arrived.

What resveratrol is, and why anyone sold it as a longevity ingredient

Resveratrol is a polyphenol, a plant-made compound that the 2003 Howitz paper described as found in red wine, and one that supplement makers commonly extract from Japanese knotweed. Walle's 2004 absorption study calls it a dietary polyphenol. It got its longevity reputation when lab work suggested it mimicked calorie restriction in yeast, a story we unpack below, and it has been on shelves ever since.

The story in six lines

• Origin: the 1992 "French paradox" paper pointed at wine and alcohol, not at resveratrol. The molecule joined the story later.
• Lab and animal work: resveratrol extended yeast lifespan and improved the health of mice on a high-calorie diet. A large US mouse programme found no lifespan effect in genetically varied mice.
• Mechanism: the headline idea that resveratrol directly switches on the sirtuin enzyme SIRT1 is contested.
• Human trials: small, short and inconsistent. Where effects appear, they tend to be in people with metabolic problems and at higher doses.
• Absorption: your gut takes it up well, then your body converts almost all of it into other compounds very quickly. Very little unchanged resveratrol reaches the blood.
• Safety: well tolerated at supplement doses in the trials covered here, with gastrointestinal upsets at high doses and a documented effect on drug-processing enzymes.

A paradox, a yeast and a very lucky mouse

The story starts in 1992, when Renaud and de Lorgeril noticed that France combined a high intake of saturated fat with relatively low coronary heart disease deaths. Their Lancet paper suggested that wine, and specifically the alcohol in it acting on platelets, might be part of the explanation. Resveratrol does not feature in that argument. The "French paradox" is a dietary observation, not a resveratrol finding.

Resveratrol entered the longevity conversation in 2003, when Howitz and colleagues reported that it activated SIRT1 in a test tube and extended the lifespan of yeast by 70%. In 2006, Baur and colleagues showed that resveratrol improved health and survival in middle-aged mice fed a high-calorie diet. Note the setting: these were mice on a diet designed to make them unhealthy, and resveratrol moved them back towards the physiology of mice on a standard diet. That is a different claim from "resveratrol makes healthy mice live longer".

Then the scrutiny arrived:

• In 2005, Kaeberlein and colleagues found that resveratrol boosted SIRT1 activity in the lab only when the test peptide carried a fluorescent tag, and had no detectable effect on Sir2 or lifespan in three yeast strains.
• In 2010, Pacholec and colleagues reported that resveratrol did not directly activate SIRT1 with native substrates, again seeing activation only with the fluorophore-tagged peptide.
• In the US National Institute on Aging's Interventions Testing Program, Strong and colleagues (2013) gave resveratrol to genetically varied mice from four months of age and found no statistically significant effect on lifespan in males or females at the concentrations tested.

None of this proves resveratrol does nothing. Cells and animals do respond to it, and other mechanisms (such as AMPK signalling, which the human trials below touched on) are still being explored. But "activates the longevity gene" is a hypothesis that has been argued about for twenty years, not a settled fact.

What happened when people took it

Human trials are where a good story usually meets a calendar. Here are the most informative ones, with the populations and doses as studied.

• Trial: Timmers 2011; Who and what: 11 obese men, 150 mg/day for 30 days, crossover design; Result: Lower resting metabolic rate, liver fat, fasting glucose, triglycerides and inflammation markers; systolic blood pressure and HOMA index improved. Tiny and short, and the authors framed it as calorie-restriction-like. European Research Council grant listed.
• Trial: Poulsen 2013; Who and what: 24 obese men, 500 mg three times daily (1.5 g/day) for 4 weeks, placebo-controlled; Result: No effect on the primary outcome (insulin sensitivity by clamp) or on blood pressure, body fat or inflammatory markers. Investigator-initiated; funded by public and charitable foundations.
• Trial: Yoshino 2012; Who and what: Non-obese postmenopausal women with normal glucose tolerance, 75 mg/day for 12 weeks; Result: No change in insulin sensitivity, body composition, lipids or inflammation. The putative molecular targets (AMPK, SIRT1 and others) did not change either.
• Trial: Witte 2014; Who and what: 46 healthy, overweight adults aged 50 to 75 (23 per group), 200 mg/day for 26 weeks; Result: Better 30-minute word retention (p = 0.038), more hippocampal connectivity and lower HbA1c than placebo. Small, with several outcomes tested.
• Trial: Turner 2015; Who and what: 119 people with mild to moderate Alzheimer's disease, 500 mg/day escalating to 1,000 mg twice daily over 52 weeks; Result: Safe and tolerable, with nausea, diarrhoea and weight loss reported. Some biomarker levels shifted, and brain volume loss was greater with resveratrol than placebo; the authors said the biomarker changes need further study. Supported by US National Institute on Aging grants.

Read across the rows and a pattern emerges. The better-looking results come from people with a metabolic problem (obesity, diabetes) or from small trials with many outcomes. The null results come from healthy or normal-glucose participants and from the most carefully controlled measure of insulin sensitivity. The result that deserves the most caution is not a benefit at all: more brain volume loss in a trial of people with Alzheimer's. Nobody knows yet what that means, and it is the reason we won't treat "more is better" as a safe assumption.

What the pooled analyses say

Meta-analyses combine trials. They are only as good as the trials that go into them, and for resveratrol those are small, varied and use many different doses.

• Outcome: Blood pressure, pooled trials; Pooled finding: Liu 2015, six randomised trials with 247 people: no significant overall effect on systolic or diastolic pressure. In the subgroup taking 150 mg/day or more, systolic pressure was 11.9 mmHg lower (95% CI -20.99 to -2.81), a large drop, but diastolic pressure didn't change, the meta-regression showed no dose effect, and the authors asked for more high-quality studies.
• Outcome: Blood pressure, type 2 diabetes; Pooled finding: Nyambuya 2020, five trials with 388 people, most already on metformin: systolic pressure lower with resveratrol, with high heterogeneity between trials.
• Outcome: Blood glucose, type 2 diabetes; Pooled finding: Conflicting. Abdelhaleem 2022 (17 trials, 871 people) found lower fasting glucose at 500 mg or more per day. García-Martínez 2022 found effects that depended on age and dose and said the evidence was not enough to propose a therapeutic dose. An umbrella review of ten meta-analyses (Ghalichi 2026) found no significant overall effect on glycaemic indices and concluded resveratrol could not be considered beneficial for them, given high heterogeneity.
• Outcome: Cognition in older adults; Pooled finding: Sánchez-Nieto 2023, six studies: no significant effect on memory or processing speed. Khorshidi 2021 found that animal studies mostly reported benefits, while pooled human data showed no statistically significant effect on cognition, mood, grey matter volume or blood pressure, and described any human effect as likely small.

For ageing itself, there is nothing to report. No trial has measured whether resveratrol changes human lifespan, and none is likely to. What exists are short trials of markers, and markers are not outcomes.

Absorption: swallowed is not the same as in your blood

Resveratrol has a famous absorption puzzle. In a 2004 study by Walle and colleagues, six volunteers took radiolabelled resveratrol. At least 70% of a 25 mg oral dose was absorbed, yet only trace amounts of unchanged resveratrol (below 5 ng/ml) could be found in plasma. The rest circulated as sulphate and glucuronide conjugates, with total resveratrol-related material peaking around 491 ng/ml and a half-life of about 9 hours. The authors describe "extremely rapid" sulphate conjugation in the gut and liver as the step that limits how much free resveratrol gets through. In other words, the gut takes it up well and the body rewrites it fast.

In a dose-escalation study of 40 healthy volunteers taking 0.5 to 5 g daily for 29 days, Brown and colleagues (2010) found the metabolites reached exposures up to about 20 times higher than the parent compound. Whether those metabolites do anything useful is a separate, open question. Studies in tissue and cells are not the same as studies in people who swallowed a capsule.

This matters when you read a label. Products with micronised, liposomal or "enhanced absorption" formats may raise blood levels, and that may be real. But higher blood levels of resveratrol are not the same as better health outcomes. The trials above used plain preparations, so a claim about an enhanced product needs its own published human data. The sources behind this page contain no such data for any specific product format.

Reading a label: trans-resveratrol, knotweed and the arithmetic

Resveratrol exists in two shapes (isomers), trans and cis. The form studied in most supplement research and the form authorised in the EU is trans-resveratrol, and EU rules require the ingredient to be labelled "trans-resveratrol".

Many products use an extract from Japanese knotweed, which you may also see labelled with the older Latin name Polygonum cuspidatum. The extract is a plant powder containing a stated percentage of resveratrol, so these two label lines are very different:

• "Knotweed extract 500 mg", with no resveratrol amount, tells you how heavy the capsule fill is, not how much resveratrol you get.
• "Knotweed extract 500 mg, standardised to 50% trans-resveratrol" would mean 250 mg of trans-resveratrol per serving (our illustrative arithmetic, not a real product).

Then compare that figure to the dose studied. The trials in the table above used anything from 75 mg to 5 g daily, and EFSA's safety assessment of the EU-authorised ingredient covered up to 150 mg/day for adults. The mismatch is worth noticing: many trial doses sit well above what the EU safety assessment covered, and "the dose in the study" is not a dose we can endorse.

Our guide to reading supplement labels covers the general method of checking forms, serving sizes and extract ratios.

Where resveratrol stands in the EU and the Netherlands

As of 8 October 2026:

• The European Commission authorised synthetic trans-resveratrol as a novel food in Implementing Decision (EU) 2016/1190 for food supplements in capsule or tablet form for adults, with a maximum dose of 150 mg per day. The Decision requires labels on supplements containing it to say that people using medicines should only consume the product under medical supervision.
• Regulation (EU) 2022/672 amended the specifications of trans-resveratrol from a microbial source (a genetically modified strain of the yeast Saccharomyces cerevisiae). Its recitals say the microbial version was authorised on the basis of substantial equivalence to resveratrol with a history of consumption before 15 May 1997, extracted from Japanese knotweed (Fallopia japonica), and that Regulation (EU) 2021/51 later lifted restrictions on the delivery formats of supplements containing it.
• Later amendments may change the conditions of use, so check the current Union list. We couldn't confirm the EU novel-food status of a standalone Japanese knotweed extract, so we make no statement about it. Check the Commission's Novel Food Status Catalogue before buying.
• These are EU-wide authorisations, and they apply in the Netherlands. No separate Dutch ruling specific to resveratrol turned up; rules change, so check with the Dutch food authority (NVWA) before relying on this.

Safety and interactions

At the doses sold as supplements, resveratrol was tolerated in the trials covered here. Higher doses are different.

• Digestive effects: in the 29-day study, the 2.5 g and 5 g daily doses caused mild to moderate gastrointestinal symptoms (Brown 2010). EFSA noted diarrhoea or other gastrointestinal symptoms in four uncontrolled studies at 1 g/day or more. In the Alzheimer's trial, nausea, diarrhoea and weight loss were the most common adverse events.
• Drug-processing enzymes: in a study of 42 healthy volunteers, 1 g of resveratrol daily for four weeks inhibited indices of CYP3A4, CYP2D6 and CYP2C9 activity and induced CYP1A2. The authors said pharmacological doses could lead to adverse drug interactions, and suggested lower doses be considered.
• EFSA's assessment: EFSA's 2016 opinion concluded that 150 mg/day was safe for adults. It also noted that a resveratrol metabolite could inhibit CYP enzymes and may interact with medicines mainly metabolised by CYP2C9.
• Anticoagulants: the only evidence on warfarin here is from rats, where resveratrol increased warfarin exposure and its anticoagulant effect (Huang 2020). No human interaction study is cited, so treat this as a reason to ask your pharmacist or prescriber rather than a measured human risk.
• Who it is authorised for: the EU authorisation is for adults only. The sources behind this page have no data in pregnancy or breastfeeding, which is a reason to skip it, not a green light.

If you take prescription medication, speak to a pharmacist or doctor first. Our guide to supplement and drug interactions explains how to bring the right questions to that conversation.

What are you hoping resveratrol will do?

The answer changes with the hope.

• If the hope is "slow ageing" or "activate longevity genes", the human evidence does not support it.
• If the hope is a measurable change in blood pressure or glucose, there are pooled analyses pointing both ways, mostly in people with metabolic conditions. That is a conversation for a doctor, not a reason to self-treat.
• If you are curious and healthy and just want to try something, the main practical questions are label arithmetic (how many milligrams of trans-resveratrol), how that compares to the studied doses, and whether you take any medicine that shares the CYP pathways above.

Resveratrol is one of the most studied compounds in longevity research, and its yeast, cell and mouse stories are fascinating. For people, the research hasn't yet delivered an anti-ageing result.

Where to read next

Resveratrol shares a shelf, and a story, with several other longevity names. A few guides to browse next:

• NMN and NAD: understand the relationship, for how NAD+, NMN, NR and niacin relate and what human trials found.
• Spermidine: why the longevity interest?, for the autophagy story, what food studies and supplement trials found, and the EU rules.
• Fisetin and quercetin: what does senolytic mean?, for what 'senolytic' means and what mice and the first human trials showed.
• Longevity: mechanisms versus human outcomes, for how to read a longevity label as a description of a mechanism rather than an outcome.

Written by Smart Supplements with AI assistance from the sources linked in the text. It has not been reviewed by a clinician.

Last updated: 12 October 2026

This article is for informational purposes only and is not intended as medical advice. Always consult a healthcare professional before starting any new supplement, especially if you take prescription medication.