Free entry from The Complete Supplement Guide
Urolithin A
Graded Moderate · Longevity & Immune Resilience · free to read
Four placebo-controlled human trials exist, not twenty-five. The largest missed its pre-specified primary endpoint, a second had its endpoints replaced after the study began, and an independent 2025 review concluded the evidence is insufficient in any population. Mitophagy has never been measured in a human.
The full entry, the nine categories and how all 147 entries were graded are in the guide.
Get The Complete Supplement Guide →Last reviewed October 2026
The verdict
Is Urolithin A worth taking?
Urolithin A is not something you eat. It is what your gut bacteria make from what you eat — the ellagitannins in pomegranate, walnuts and berries are broken down to ellagic acid, and certain gut bacteria convert that to urolithin A. Supplements sell the finished metabolite so you do not have to rely on your microbiome producing it.
That premise is sound, and it is the most interesting thing about the compound. The problem is everything built on top of it.
Four placebo-controlled trials exist in humans. The largest missed its pre-specified primary endpoint. A second had its endpoints replaced after the study had started. A third, in trained runners, found no improvement in the 3,000 meter time trial in either group. An independent systematic review published in 2025 — funded by the UK's National Institute for Health and Care Research, not by anyone selling the compound — pooled the human evidence and concluded that insufficient evidence exists to support using urolithin A to improve muscle function in any population at present.
The bottom line. The conversion biology is real and worth understanding, and the compound looks safe at the doses tested. But every functional benefit it is sold for has either failed its own primary endpoint or has never been tested against placebo at all. If you want the substrate, eat pomegranate and walnuts. If you want the finished metabolite because you suspect you are a non-converter, understand that you are buying a mechanism, not a measured outcome.
What it does
What Urolithin A does in the body
The proposed mechanism is mitophagy — the process by which a cell identifies damaged mitochondria, tags them, and recycles them. Mitochondrial quality control declines with age, and in worms and mice urolithin A improves it, extends lifespan in C. elegans and improves muscle function in aged rodents. That preclinical work is genuinely good.
In humans, mitophagy has never been measured. Not once, in any trial. What has been measured are proxies: mitochondrial proteins in muscle biopsies, plasma acylcarnitines, and in one trial the phosphorylation of Parkin, a protein involved in tagging damaged mitochondria. In that trial — the 2025 study in trained distance runners — the Parkin marker did not reach significance (p=0.098), and mitochondrial respiration did not change.
This matters because the marketing language collapses the distinction. “Activates mitophagy in humans” is a claim about a measurement nobody has taken. The careful version, which is what the papers themselves say when you read the prose rather than the comparison tables, is that urolithin A induces a signature consistent with improved mitochondrial health. That is a different and much weaker statement, and it is the honest one.
The evidence
What the research actually shows
Moderate, and that grade is generous. Here is every placebo-controlled human trial, in full, because the number of them is itself the story — you will see claims of twenty-five human trials, and there are four.
Singh 2022, Cell Reports Medicine. n=88, 500 mg or 1,000 mg daily, four months, middle-aged overweight or obese adults. The pre-specified primary endpoint was peak power output. It was not met — the paper states there was no significant change in peak power output against placebo. Lean body mass by DEXA was unchanged across groups. The six-minute walk distance improved by 33.4 meters, but that is a within-group change from baseline; against placebo it did not reach significance (p=0.098). Funded by Amazentis SA, which sells the supplement; six authors were company employees.
Liu 2022, JAMA Network Open. Older adults, 1,000 mg daily, four months. The trial's endpoints were replaced after the study had begun — the paper says so. Six-minute walk performance was not a significant improvement over placebo. Muscle ATP showed no significant improvement. At four months there was no significant effect of treatment on muscle endurance. The published paper records that the funder had a role in the decision to submit the manuscript for publication.
Whitfield 2025, Sports Medicine. n=42, 1,000 mg daily, four weeks, highly trained male distance runners at altitude. The paper's own subtitle is the finding: urolithin A enhances recovery but not running performance. The 3,000 meter time trial did not significantly improve in either group. VO₂max showed no group-by-time interaction (p=0.138). What did reach significance was lower perceived exertion (p=0.02) and a reduction in creatine kinase after the time trial. Funded by Amazentis; three authors are company employees.
Zhao 2024. Young male strength-training athletes, eight weeks. The independent review that pooled it rated this one as having some concerns over randomization, allocation concealment and statistical reporting.
What the independent pooling found. The 2025 systematic review — three of these trials, 174 participants, funded by the UK's National Institute for Health and Care Research — found four of twelve outcomes significant. Pooled six-minute walk distance was +23 meters with a 95 percent confidence interval of −6 to 52 meters, p=0.12. Its conclusion, verbatim in substance: insufficient evidence exists to support the use of urolithin A to improve muscle function in any population at present.
And the one that gets presented as a new trial but is not. A 2025 paper in iScience reported reductions in plasma ceramides, lipid markers associated with cardiovascular risk, after four months at 1,000 mg. It is widely cited as fresh evidence. It is a lipidomics analysis of banked plasma from the same older-adult trial above — the one whose endpoints were replaced and whose functional outcomes were null. Sixty-six participants, Amazentis-funded, and the between-group p-values are in a supplementary table rather than the text. A biomarker re-analysis of a trial that missed its functional endpoints is not independent confirmation of anything.
Can you make it?
Whether your gut can make it is the real question
This is the part of the story that deserves the attention, and the marketing gets it almost right before overstating it.
People fall into urolithin metabotypes based on which gut bacteria they carry. UM-A produce urolithin A. UM-B produce urolithin B and isourolithin A instead. UM-0 produce no urolithins at all — you can feed them pomegranate indefinitely and nothing comes out the other side.
The best population data comes from a Caucasian cohort of 839 people aged 5 to 90 (Cortés-Martín and colleagues, Food & Function, 2018). Its central finding is not the one usually quoted: aging is the main factor determining how a person metabolizes ellagic acid. The UM-0 group — the true non-converters — runs at approximately 10 percent in Spanish and Chinese healthy volunteers, and researchers in the same journal have noted it could be higher in the United States.
So the honest framing is: roughly one person in ten makes none of it, a larger group makes the related urolithin B instead, and metabotype shifts with age. You will also see the line that only one in three people can get enough urolithin A from diet alone. I could not find a citation for that figure anywhere on the site that publishes it, and it does not match the metabotype literature.
There is no consumer test for your metabotype. The practical read: if you eat pomegranate or walnuts regularly and have never noticed anything, that tells you nothing either way, because urolithin production is not something you can feel.
The claims
How to read a urolithin A sales page
The compound is sold almost entirely on the strength of claims that are structured to survive scrutiny while reading as proof. Three patterns are worth recognizing, because they are not unique to this compound.
A study count with no study list. You will see figures like twenty-five human clinical trials and five hundred studies. Four placebo-controlled human trials exist. The larger numbers are reached by counting cell culture, rodent work, reviews, and separate analyses of the same trial's samples as distinct studies.
The strong claim in the table, the hedged claim in the prose. “Directly activates mitophagy” appears as a checkmark in product comparison charts. In the body text of the same pages, and in the papers themselves, it becomes “induces a signature of improved mitochondrial function.” When a claim is only ever made in a format that cannot carry a citation, that is the tell.
The within-group result presented as the effect. A group improving from its own baseline is not evidence the supplement worked; that is what a placebo arm is for. The 33 meter walking improvement that circulates widely is a within-group change. Against placebo it was not significant.
None of this makes the compound worthless. It makes the marketing an unreliable guide to it.
Food first
Getting Urolithin A from food
You cannot eat urolithin A. You eat the precursor and your gut bacteria do the rest, which makes this one of the few compounds where the food version and the supplement version are doing genuinely different things.
The precursors are ellagitannins, and the food map is short. Pomegranate is the headline source; the specific ellagitannins are punicalagins. Walnuts are the richest nut source and the most practical, because a handful is a normal thing to eat. Then raspberries, blackberries, strawberries and cloudberries, and chestnuts and pecans in smaller amounts. Oak-aged wine and spirits pick up ellagitannins from the barrel.
Two things a cook should know, because both are places the compound gets thrown away.
In pomegranate, the punicalagins are concentrated in the peel and the white pith, not the arils. Juice pressed from the whole fruit carries substantially more than juice squeezed from hand-picked seeds. If you are buying bottled pomegranate juice, whole-fruit pressed is the one that matters, and the faint astringency at the back of the palate is what you are tasting.
In walnuts, much of it sits in the bitter papery brown skin. Blanching walnuts to slip that skin off — a standard move when you want a sweeter, cleaner nut in a dessert or a salad — takes the ellagitannins with it. Toasting dry, skin on, keeps them.
The quantities are not trivial either. A cup of pomegranate juice or a small handful of walnuts delivers ellagitannins in the hundreds of milligrams, which is the same order of magnitude the trials dosed as finished urolithin A. The variable is not how much you eat. It is whether your gut converts it, which is the section above.
What the guide adds
The part that is not on this page
Nothing has been measured beyond four months. The longest human exposure in any trial is four months. This is marketed as a longevity compound; there is no human data on a year of use, let alone a decade.
No trial has been run in anyone who is ill. All four placebo-controlled trials recruited healthy middle-aged adults, healthy older adults with reduced walking performance, or trained athletes. There is no trial in diagnosed sarcopenia, no trial in cardiovascular disease, no trial in any clinical population — which is exactly where a mitochondrial quality-control drug should be most useful, and where the independent review called for the work to be done.
Every trial that found anything was funded by the manufacturer. All four were supported by the company that holds the patents, with company employees among the authors in three of them. That is normal in this industry and it does not make the results false. It does mean the only pooled analysis conducted by people with nothing to sell concluded the evidence is insufficient.
The population data is narrow. The metabotype work is in Caucasian, Spanish and Chinese cohorts. The older-adult trial that produced the ceramide finding was entirely White. There is no reason to expect the biology differs, but the gut microbiome is the mechanism here, and gut microbiomes differ considerably across diets and regions.
Nothing about lifespan or healthspan has been measured in a human. The lifespan extension is in C. elegans. The muscle work is in aged rodents. Those are the findings that built the compound's reputation and they remain the strongest findings about it.
Safety looks fine. At 500 to 1,000 mg daily for up to four months it has been well tolerated with no serious adverse events reported. The honest risk here is not harm; it is cost. At typical pricing this is one of the most expensive supplements on the market per month, bought for a mechanism rather than a demonstrated outcome.
About the author
Who graded this
Daniel Edlund is a Le Cordon Bleu-trained private chef in Los Angeles, a certified personal trainer and strength coach, and an internationally licensed massage therapist. Thirty years of strength training and fifteen years cooking in private estate kitchens. He is the author of The Complete Supplement Guide, which grades 147 entries against the clinical literature, and of The EASY Way to Organic Cooking, first published in 2013.
This page is information, not medical advice. Supplements interact with prescription medication and with each other, and the right answer depends on your own bloodwork and history. Talk to your doctor before starting anything, particularly if you are pregnant, nursing, managing a health condition, or already taking medication.