This page has affiliate links to Renue By Science: we earn a commission if you buy through them.
Resveratrol in human trials: what changed, and what did not
Four human trials of resveratrol, with doses, sample sizes, and where the results disagreed. Not a recommendation.
Resveratrol is a polyphenol found in grapes and other plants. Capsules of it are sold with a long list of hopes. This note stays inside four human studies and asks a plainer question: at the doses and durations those trials actually used, which measurements moved, and which did not?
The four papers do not tell the same story. One 30-day crossover in 11 men reported changes in metabolic rate, muscle enzymes, liver fat, glucose, triglycerides, and blood pressure [1]Citation 1. A 4-week trial at a much higher dose, in 24 men, did not find those metabolic changes and did not improve insulin sensitivity [2]Citation 2. A single-dose study found changes in a brain blood-flow signal and no change on the cognitive tasks that day [3]Citation 3. A 26-week comparison in older adults reported a difference on one memory score, along with a lower glycated hemoglobin and less body fat [4]Citation 4. That is a small evidence base. Treating it as settled would misread it.
This is a reading note. It is not a suggestion to take resveratrol.
A 30-day crossover in 11 men
Grade A. Human RCT or meta-analysis
Timmers and colleagues ran a randomized, double-blind crossover in 11 healthy obese men. Each man took placebo and, in the other period, 150 mg per day of resveratrol (resVida) for 30 days [1]Citation 1. A crossover of this size can detect within-person shifts. It cannot describe women, lean adults, or a year of use. The grade is Grade A because the design is a randomized trial. The grade is not a measure of how sure the result is. Eleven men is a small trial.
The paper reports that resveratrol lowered sleeping and resting metabolic rate. In muscle, the authors describe activation of AMPK, higher SIRT1 and PGC-1α protein, higher citrate synthase activity without a change in mitochondrial content, and higher mitochondrial respiration when the substrate was derived from fatty acids. They also report higher intramyocellular lipid, lower intrahepatic lipid, and lower circulating glucose, triglycerides, alanine aminotransferase, and markers they classed as inflammation markers. Systolic blood pressure fell, and the HOMA index improved. After a meal, adipose lipolysis and plasma fatty acids and glycerol were lower [1]Citation 1.
Those are laboratory and imaging endpoints, listed here because they are what the trial measured. They are not a catalogue of benefits a reader should expect. The authors described the pattern as resembling calorie restriction. This note does not repeat that as a claim about dieting. The trial measured 30 days in 11 men.
The supplement in that trial was a named material, resVida, at 150 mg per day. A shopper’s capsule with a different preparation is not automatically the same exposure. The paper does not test that comparison.
Four weeks at 500 mg three times a day
Grade A. Human RCT or meta-analysis
Poulsen and colleagues asked whether a higher dose would move insulin sensitivity and fuel use. Twenty-four obese but otherwise healthy men were randomized, double-blind, to resveratrol or placebo for 4 weeks, in parallel groups. The tablets contained 500 mg of trans-resveratrol, taken three times daily [2]Citation 2. That schedule is 1,500 mg per day, from the methods of the same paper. Before and after, the investigators measured glucose turnover and insulin sensitivity with a hyperinsulinemic euglycemic clamp.
Insulin sensitivity, the primary outcome, worsened by a non-significant amount in both groups. Endogenous glucose production did not change. Turnover and oxidation of glucose did not change. The paper reports no effect on blood pressure, resting energy expenditure, lipid oxidation, ectopic or visceral fat, or the inflammatory and metabolic biomarkers they assayed [2]Citation 2.
Set that list next to Timmers. Similar volunteers, in the sense of obese men. A dose about ten times higher. A shorter exposure, four weeks rather than thirty days, and a parallel design rather than a crossover. The high-dose trial did not confirm a blood-pressure drop, an energy-expenditure drop, or a shift in the fat depots and biomarkers it measured. Poulsen and colleagues wrote that the absence of an effect did not match the animal literature and that it raised doubt about using resveratrol as a nutritional supplement in that setting [2]Citation 2. This note quotes that doubt as the authors’ conclusion about their own trial. It is not a diagnosis and not an instruction.
Two trials in men, 11 and 24 people, cannot cancel each other in a formal way. They can show that the cheerful reading of the first paper was premature. A result that depends on one crossover of 11 men, and that a second randomized trial does not see, is a result that needs a third trial before anyone treats it as stable.
One afternoon, two doses, and a blood-flow signal
Grade A. Human RCT or meta-analysis
Kennedy and colleagues studied the hours after a dose, not a month of capsules. Twenty-two healthy adults took placebo, 250 mg, and 500 mg of trans-resveratrol on separate days, in random order, double-blind. After 45 minutes they did cognitive tasks that use the frontal cortex, for a further 36 minutes. Near-infrared spectroscopy tracked oxygenated and deoxygenated hemoglobin in the frontal cortex. A separate group of nine adults had blood drawn so the investigators could confirm that resveratrol and its conjugates were in plasma [3]Citation 3.
Total hemoglobin during the tasks rose in a dose-related way, which the authors use as an index of cerebral blood flow. Deoxyhemoglobin also rose after both doses, which they read as greater oxygen extraction. The cognitive scores did not change. Metabolites were present in plasma through the task window [3]Citation 3.
So a single dose can change a non-invasive blood-flow signal within an hour, in healthy adults, without a change on the thinking tests given the same afternoon. That split matters. A change in a monitor is not the same sentence as a change in performance. The study does not show what happens if those doses are repeated for weeks. It does not include people with a neurological condition. It is a crossover of 22 people plus a blood-sampling group of nine.
Twenty-six weeks in older adults
Grade B. Small, pilot or observational human study
Witte and colleagues report a 26-week comparison of 200 mg per day of resveratrol with placebo in healthy overweight older adults. Twenty-three people who completed the resveratrol period were pairwise matched to 23 people who received placebo: 46 people in total, 18 women, ages 50 to 75 years. They took memory tests and had brain imaging aimed at the hippocampus, plus blood measures that included glycated hemoglobin [4]Citation 4.
The abstract of the paper, which is the text this note is using, does not itself use the word randomized. This note therefore does not grade the study as a randomized trial. Grade B here means a human study that this note is not calling an RCT or a meta-analysis. If a fuller reading of the methods shows a randomized assignment, the grade can be revised. The numbers below are the ones the paper reports, either way.
Retention of words over 30 minutes differed between resveratrol and placebo, with a p value of 0.038. The paper also reports higher functional connectivity of the hippocampus, lower glycated hemoglobin, lower body fat, and higher leptin, each with a p value under 0.05. The rise in connectivity between the left posterior hippocampus and medial prefrontal cortex lined up with the memory scores and with the fall in glycated hemoglobin [4]Citation 4.
A p value of 0.038 on one memory score, in 46 people, is a difference the authors judged unlikely under a null model. It is also one test, in one sample, over half a year, in overweight older adults. It is not a finding that “memory” in general changed, and it is not evidence about dementia. The authors’ closing sentence looks toward maintaining brain health during aging. This note stops at the measurements: word retention, a connectivity measure, glycated hemoglobin, body fat, and leptin. It does not adopt a claim about aging.
The imaging and the memory score moved together in the analysis the authors present. A correlation inside one trial is not a mechanism that has been shown to operate in other groups. No second human trial in this set of four repeats the 26-week memory result.
What the four trials do not settle
The disagreements are the point of reading them together.
Metabolic rate and blood pressure moved in the 150 mg, 30-day crossover [1]Citation 1 and did not move in the 500 mg three-times-daily, 4-week trial [2]Citation 2. Liver fat and several blood lipids are reported as changed in the first paper and not as changed in the second. Insulin sensitivity was not the same measurement in both papers. Poulsen’s primary outcome, the clamp, did not improve [2]Citation 2. Timmers reported a better HOMA index [1]Citation 1. Those are related ideas and different tests. A reader who collapses them into “blood sugar improved” is ahead of both papers.
Thinking tests did not move on the day of a 250 or 500 mg dose [3]Citation 3. A 30-minute word list did differ after 26 weeks at 200 mg [4]Citation 4. Different tests, different ages, different durations. There is no basis in these four papers for a sentence such as “resveratrol helps cognition.”
No trial here followed people long enough, or in large enough numbers, to say anything about fractures, heart attacks, memory clinics, or years of life. Sample sizes of 11, 22, 24, and 46 are hypothesis-generating sizes. Three of the four used randomized designs [1]Citation 1[2]Citation 2[3]Citation 3. Randomization does not enlarge a sample. It only, when it works, balances the groups you have.
Preparations differed: resVida at 150 mg [1]Citation 1, trans-resveratrol tablets at 500 mg [2]Citation 2, trans-resveratrol at 250 and 500 mg [3]Citation 3, and 200 mg per day in the longer comparison [4]Citation 4. Particle size, co-ingested food, and the rest of the capsule are not standardized across these papers. A label that says “resveratrol” and a number of milligrams is not a guarantee that the exposure matches any one of these trials.
Label amounts and price
Prices checked 20 Nov 2026. Figures are US store list prices in USD, before shipping. The code HEALTHSPAN is 10 percent off that list price. These are label amounts, not instructions.
Resveratrol Pure Powder is loose powder. One 0.5 g scoop is 500 mg of trans-resveratrol, and the page states more than 98 percent trans. The pack is 200 servings at $64.00 list, which is $57.60 with the code. That is $0.32 per 500 mg list, and $0.29 with the code.
Liposomal Resveratrol is sold as LIPO RESVERATROL, 125 mg of trans-resveratrol per capsule, 90 capsules, $41.95 list ($37.76 with the code). That is $0.47 per capsule at list. Scaling the label to 500 mg, the same size as one powder scoop, is $1.86 list and $1.68 with the code. Renue markets this as a liposomal form. The four trials cited here did not test whether a liposomal preparation changes absorption in people, and this note does not claim that it does.
Set next to the trials, without turning a label into a schedule: Timmers used 150 mg per day, which is less than one 500 mg scoop and a little more than one 125 mg capsule [1]Citation 1. Witte used 200 mg per day, still under one scoop and under two 125 mg capsules [4]Citation 4. Kennedy’s single doses were 250 mg and 500 mg, so one scoop matches the higher of those single doses and two 125 mg capsules match the lower one [3]Citation 3. Poulsen used 500 mg three times a day, which is three scoops a day, or twelve of the 125 mg capsules [2]Citation 2. The trial materials were not these shop products. A matching milligram is not a matching preparation.
References
- Timmers et al. (2011). Calorie restriction-like effects of 30 days of resveratrol supplementation on energy metabolism and metabolic profile in obese humans. Cell Metabolism. Back to text
- Poulsen et al. (2013). High-dose resveratrol supplementation in obese men: a randomized, placebo-controlled clinical trial of substrate metabolism, insulin sensitivity, and body composition. Diabetes. Back to text
- Kennedy et al. (2010). Effects of resveratrol on cerebral blood flow variables and cognitive performance in humans: a double-blind, placebo-controlled, crossover investigation. American Journal of Clinical Nutrition. Back to text
- Witte et al. (2014). Effects of resveratrol on memory performance, hippocampal functional connectivity, and glucose metabolism in healthy older adults. Journal of Neuroscience. Back to text