Section 1 of 9Overview
Overview
Urolithin A (UA) is a small molecule that some gut microbial communities produce from ellagic acid and ellagitannins in foods such as pomegranate, berries and walnuts. It is not a peptide. A chemically synthesized version has also been studied as an oral food or supplement ingredient.
Human evidence now extends beyond a catalog lead, but it remains limited. Small, short randomized trials report changes in selected mitochondrial biomarkers, muscle-endurance measures and immune-cell phenotypes. Those findings must be read together with the negative primary results: two central performance trials did not show a statistically significant benefit on their prespecified primary efficacy endpoints.
Biomarker change is not an anti-ageing outcome
No reviewed trial showed that urolithin A slows human ageing, prevents disease, extends life or treats a diagnosed condition. Study-arm quantities below are reported only to make the evidence interpretable; they are not instructions.
Section 2 of 9Verified identity and origin
Verified identity and origin
PubChem CID 5488186 and FDA GRAS Notice 791 identify urolithin A as 3,8-dihydroxybenzo[c]chromen-6-one, formula C13H8O4, CAS 1143-70-0, with molecular weight 228.20 g/mol. This structure distinguishes it from other urolithins and conjugated metabolites.
The molecule can be formed by gut microbes after exposure to ellagitannin-rich foods, but people differ in whether and how much they produce. Direct synthetic urolithin A therefore creates a different and more standardized exposure than eating pomegranate. Evidence for a defined synthetic study product does not automatically transfer to every supplement, extract or food.
Section 3 of 9Mechanistic basis
Mechanistic basis
The foundational 2016 work found mitophagy and functional effects in worms and rodents. Later human studies measured plasma acylcarnitines, muscle gene or protein expression and immune-cell metabolism. These are biologically relevant signals, but they remain surrogate or mechanistic outcomes. They do not by themselves demonstrate improved health, prevention of age-related disease or longer survival.
Section 4 of 9Source-integrity correction
Source-integrity correction
The imported PeptideGuide record labels its single citation as the foundational urolithin A mitophagy study but links to PMID 27400127. That PMID is an unrelated paper about truncated netrin-1 and vascular permeability in diabetic retinopathy. The actual Ryu et al. urolithin A study is PMID 27400265. The title alone was therefore not accepted as provenance, and the wrong PMID must not be reused as urolithin A evidence.
Section 5 of 9What randomized human studies show
What randomized human studies show
| Study | Design | Main result | Boundary |
|---|---|---|---|
| Andreux 2019 / NCT02655393 | Phase 1, 60 healthy older adults; single doses and 4-week repeated exposure | The primary short-term safety/tolerability objective was met; exposure and mitochondrial biomarker signals were reported | Not an efficacy trial; too short and selected to establish long-term safety |
| Liu 2022 / NCT03283462 | 66 adults aged 65–90; 1,000 mg/day or placebo for 4 months | Neither co-primary comparison—6-minute walk distance or maximal hand-muscle ATP production—was significantly improved versus placebo; selected muscle-endurance and plasma markers changed | Secondary signals do not overturn the negative co-primary results |
| Singh 2022 / NCT03464500 | 88 middle-aged adults; 500 mg/day, 1,000 mg/day or placebo for 4 months | The prespecified primary peak-power endpoint was not significantly different from placebo; selected leg-strength and biomarker outcomes favoured UA | Proof-of-concept secondary outcomes need confirmatory trials |
| Runners / NCT04783207 | 42 trained male distance runners; 1,000 mg/day or placebo for 4 weeks | The 3,000 m time-trial result was not significantly improved; perceived exertion and creatine-kinase measures differed | Small, short and population-specific; no general performance claim |
| Immune trial / NCT05735886 | 50 healthy middle-aged adults; 1,000 mg/day or placebo for 4 weeks | Primary cellular phenotype and immune-metabolism measures changed | No infection, cancer, vaccine-response or clinical disease endpoint was tested |
Several authors across this evidence base were employees, shareholders or patent holders connected to the ingredient developer. That does not invalidate the studies, but it increases the importance of independent replication, prespecified endpoints and the distinction between primary and secondary results.
ClinicalTrials.gov shows continuing research in glucose metabolism, frailty, muscle immobilization, oncology and cognition. A registry record proves that a study exists; it does not prove benefit, completion or peer-reviewed results.
Section 6 of 9Safety boundary
Safety boundary
Across the reviewed trials, adverse events were generally similar to placebo and no short-term safety signal dominated. The evidence is nevertheless based on small, selected groups and follow-up of at most four months in the central published performance trials. It does not establish long-term safety, pregnancy safety, use in children, interactions, or suitability for people with complex disease or medication regimens.
Product identity matters. The FDA notice describes a specified synthesized ingredient with purity and contaminant limits. It cannot validate the identity, purity or composition of an unrelated retail product.
Section 7 of 9Regulatory and sport status
Regulatory and sport status
In GRN 791, FDA said it had no questions at that time about the notifier's GRAS conclusion for specified food uses. The same letter explicitly says this is not an FDA GRAS affirmation, did not evaluate health-benefit labelling claims and did not address every applicable legal provision. It is therefore not a drug approval and not proof that urolithin A treats anything.
The European Commission hosts a non-confidential summary of a 2018 novel-food application. An application summary records what an applicant requested; it is not, by itself, an authorisation. Food and supplement status must be checked for the current jurisdiction and product.
An exact-name review of WADA's 2026 Prohibited List found neither “urolithin A” nor “Mitopure”. WADA categories can extend beyond named examples, so absence of the name must not be turned into a guarantee for athletes.
Section 8 of 9Russian-language and regional evidence review
Russian-language and regional evidence review
A targeted PubMed search for urolithin A with Russian affiliations found Russian-origin reviews and laboratory/ex-vivo work, including a 2022 cancer review, but no direct-administration Russian human trial. These sources expand the mechanistic map; they do not establish a clinical treatment effect.
Section 9 of 9Bottom line
Bottom line
Urolithin A has credible early human research and a clearer evidence base than many “longevity” compounds. The most defensible conclusion is still modest: short trials show biological activity and selected secondary signals, while important primary performance outcomes were negative and no anti-ageing or therapeutic benefit has been established.