Urolithin A
Urolithin A is a postbiotic metabolite produced when gut bacteria convert ellagitannins — polyphenolic compounds found in pomegranates, walnuts, raspberries, and certain other plant foods — into an absorbable form. It does not come from food directly but is synthesised within the colon by specific microbial species, meaning its availability depends entirely on the composition of an individual's gut microbiome. Research interest in urolithin A has grown substantially over the past decade, centred primarily on its proposed ability to stimulate mitophagy — a selective cellular process by which damaged or dysfunctional mitochondria are identified and cleared — and thereby support mitochondrial quality over time.
What it is
Urolithin A belongs to a family of compounds called urolithins, characterised by a benzo-coumarin scaffold. It is the predominant urolithin isoform detected in human plasma and urine following consumption of ellagitannin-rich foods, and it circulates primarily as glucuronidated and sulfonated conjugates. The capacity to produce urolithin A varies substantially between individuals: one study in 100 healthy adults found that only 12 percent had detectable plasma levels at baseline, and that approximately 40 percent produced meaningful amounts following pomegranate juice consumption. This variability is attributable to differences in colonic microbiota composition and has led to the development of directly supplemented synthetic urolithin A, which bypasses microbiome-dependent conversion and achieves plasma levels approximately six times higher than those resulting from pomegranate juice intake.
The mechanism proposed to underlie urolithin A's health effects is activation of mitophagy via the PINK1-Parkin pathway and related mitochondrial quality-control systems. During ageing, mitophagy activity is thought to decline, leading to an accumulation of dysfunctional mitochondria in high-demand tissues including skeletal muscle, heart, and neurons. By hypothetically restoring mitophagic flux, urolithin A is proposed to improve mitochondrial efficiency, reduce associated oxidative stress and inflammation, and thereby support tissue function. This mechanistic story is well-elaborated in preclinical models and is supported by some molecular evidence from human muscle biopsies, but the translation into consistent functional outcomes in clinical trials remains incomplete.
Commercially, urolithin A is primarily sold under the brand name Mitopure, manufactured by the Swiss company Amazentis SA. This proprietary formulation has been used in all pivotal human trials published to date, which has important implications for interpreting the evidence: independent replication using non-proprietary preparations has not yet been published.
What the evidence shows
Human clinical evidence for urolithin A is currently limited to a small number of randomised controlled trials, all of short to moderate duration and all involving the proprietary Mitopure formulation. The most substantive findings come from two four-month trials in middle-aged and older adults respectively, plus a smaller eight-week trial in resistance-trained athletes.
The middle-aged trial (Singh et al., 2022, Cell Reports Medicine, n=88, ages 40 to 64) tested 500 mg and 1,000 mg daily doses against placebo in overweight, sedentary individuals over four months. Hamstring strength improved significantly in both UA groups compared with placebo, with an approximately 12 percent gain reported. Aerobic endurance, assessed by peak VO2, showed clinically meaningful improvements at the 1,000 mg dose. Physical performance on the 6-minute walk test improved directionally but not significantly versus placebo at group level. Notably, the study's primary endpoint — peak power output — was not significantly improved. Plasma acylcarnitine levels and C-reactive protein fell significantly in UA-treated groups, reflecting changes in mitochondrial metabolic markers and systemic inflammation. The majority of study authors were employees or scientific advisors of Amazentis SA, and the trial was company-sponsored.
The older-adult trial (Liu et al., 2022, JAMA Network Open, n=66, ages 65 to 90) administered 1,000 mg daily for four months. Muscle endurance — number of contractions to fatigue in hand and leg muscles — improved significantly compared with placebo. However, the co-primary endpoints of 6-minute walk distance and maximal ATP production in hand muscle assessed by phosphorus magnetic resonance spectroscopy did not differ significantly from placebo. Plasma biomarkers of mitochondrial health, including acylcarnitines and ceramides, fell with UA supplementation, as did C-reactive protein. The trial was funded by Amazentis, and several authors were company employees.
A smaller trial (Zhao et al., 2024, Journal of the International Society of Sports Nutrition, n=20) examined eight weeks of 1,000 mg daily in resistance-trained young male athletes. No significant differences were observed between UA and placebo on muscle strength, muscle endurance, oxidative stress markers, or protein metabolism measures. This null result in a trained, younger population is consistent with the view that effects may be more relevant in contexts of age-related mitochondrial decline.
A phase 1 randomised controlled trial in 50 healthy middle-aged adults (Denk et al., 2025, Nature Aging) examined immune cell populations after four weeks of 1,000 mg daily. UA supplementation expanded naive-like, less terminally exhausted CD8-positive T cells and increased their capacity for fatty acid oxidation. These findings are mechanistically interesting but represent immunological biomarker endpoints in a short exploratory trial; clinical immune outcomes were not assessed.
A 2025 systematic review preprint identified three trials meeting inclusion criteria for muscle outcomes, with total participation of 174 individuals and mean ages ranging from 24 to 72. Pooled analysis of 6-minute walk distance across two trials showed a non-significant improvement of 23 metres (95% CI: minus 6 to 52 metres). No significant effect on muscle mass was found. Four of twelve outcome measures showed statistically significant positive effects from urolithin A, with seven showing non-significant positive trends.
The overall picture is of a compound with a biologically plausible mechanism and some directionally consistent secondary findings, but one whose primary endpoints have consistently fallen short of significance in both pivotal trials. The positive results on muscle endurance and strength arise entirely from secondary endpoints in small, industry-funded studies, and should be interpreted accordingly. The entire published clinical evidence base is tied to a single manufacturer, trial sizes are too small for confident conclusions, and despite strong mechanistic rationale, clinical outcomes have not consistently validated this pathway.
Five questions
Does low status cause harm? This question is not directly applicable in the conventional sense, as urolithin A is not a dietary nutrient but a gut-derived metabolite. Inability to produce urolithin A from ellagitannin-rich foods does not constitute a deficiency state. However, there is a plausible biological argument that people who cannot convert dietary precursors — which represents the majority of the population — may not accumulate sufficient urolithin A to meaningfully activate mitophagy. Whether this translates into differential rates of age-related muscle decline or other outcomes has not been formally studied.
Does supplementation prevent disease? There is currently no human evidence that urolithin A supplementation prevents any disease. The compound has not been tested in populations with sarcopenia, heart failure, neurodegenerative conditions, or other clinical disease states. Preclinical models have generated hypotheses about relevance to ageing-associated conditions, but these have not been translated into clinical trial programmes examining disease incidence or progression.
Does it affect biomarkers? Yes, with some consistency. Plasma acylcarnitines and ceramides — markers of mitochondrial metabolic activity — fell with urolithin A supplementation across multiple trials. C-reactive protein declined similarly. In muscle biopsy samples, proteins associated with mitophagy and mitochondrial biogenesis were upregulated compared with placebo. In the immune trial, urolithin A altered T cell subset composition and oxidative metabolism phenotypes. These biomarker changes are not validated surrogates and do not establish clinical benefit. They are mechanistically plausible and directionally coherent, but whether acylcarnitine reduction or T cell phenotype shifts translate into functional outcomes over meaningful timescales is unknown.
Does it help clinical populations? This has not been tested. All human trials conducted to date have enrolled healthy adults — either sedentary middle-aged individuals, community-dwelling older adults, or resistance-trained athletes. No trials have been conducted in patients with sarcopenia, age-related muscle wasting, frailty, or mitochondrial disease. Extrapolation from healthy populations to clinical groups would be speculative.
Does it benefit healthy individuals? The existing evidence base is almost entirely in healthy individuals, so this is where the available data can be most directly applied. In healthy older adults and sedentary middle-aged individuals, urolithin A supplementation produced improvements in some secondary measures of muscle endurance and strength relative to placebo over four months, though primary endpoints were not met in either trial. It is unclear whether these changes translate into meaningful improvements in mobility, independence, or fall risk in real-world settings. In healthy young athletes, no benefit was observed. The balance of evidence suggests a suggestive but unconfirmed functional signal in older, untrained healthy adults, but the small total evidence base, reliance on secondary endpoints, and near-complete dependence on industry-funded research require substantial caution in interpretation.
Individual variation
The most consequential source of individual variation in urolithin A biology is gut microbiome composition. Conversion of ellagitannins to urolithin A requires a specific complement of colonic bacteria, including species within Gordonibacter and Ellagibacter genera. Studies indicate that only a minority of people produce urolithin A at levels likely to have pharmacological relevance when relying on dietary precursor intake alone. Directly supplemented synthetic urolithin A bypasses this limitation, delivering consistent plasma exposure regardless of microbiome status. Whether non-converters derive greater clinical benefit from supplementation than natural producers has not been tested in stratified trials.
Age is likely relevant, though not in the simple direction of older being better candidates. The mechanistic rationale for urolithin A benefit centres on restoring declining mitophagy activity — a process that appears to become impaired with advancing age. This would predict greater potential benefit in older adults, which is broadly consistent with the trial data showing null results in young athletes and modest signals in middle-aged and older populations. However, the evidence base is too small to make confident age-stratified predictions.
Sex-specific data are limited. The available trials have included mixed-sex populations without reporting sex-stratified outcomes. Hormonal changes associated with menopause and perimenopause affect mitochondrial function in skeletal and cardiac muscle, and there is theoretical interest in whether mitophagy activation might be particularly relevant in this context, but no sex-stratified analyses have been published from urolithin A trials. This is a meaningful gap in the evidence base.
Body composition and activity level appear to modify response. The middle-aged trial enrolled overweight, sedentary individuals and found strength and endurance signals. The athletic trial found nothing. This pattern is consistent with the view that interventions targeting age-related mitochondrial decline are more likely to show measurable effects against a background of low mitochondrial reserve than in populations with already-high mitochondrial fitness.
Testing and status assessment
There are no clinically available tests for urolithin A status, and no established reference ranges or status categories. Plasma urolithin A can be measured in research settings, but this is not routinely available and would not directly inform supplementation decisions.
Assessment of gut microbiome composition — relevant to natural production capacity — is available commercially but is not currently validated for predicting urolithin A conversion, and the clinical utility of microbiome profiling for supplement decisions is generally low.
Surrogate markers of mitochondrial health, including plasma acylcarnitines, may be measured in some clinical contexts, but their relevance to urolithin A supplementation decisions is not established. They are research endpoints rather than clinical monitoring tools.
Safety
Short-term trials suggest good tolerability for urolithin A at doses between 500 and 1,000 mg per day, but long-term safety remains uncharacterised. No drug-related serious adverse events were reported across the published trials, which ran for up to four months. Standard safety monitoring — including liver function tests, kidney markers, haematology, and electrocardiograms — showed no clinically significant changes compared with placebo. Preclinical toxicology identified no genotoxicity, and the no-observed-adverse-effect level in 90-day rat studies was determined to be considerably higher than the doses used in human trials.
Adverse events reported across trials were predominantly mild and not clearly attributable to urolithin A. In the middle-aged trial, adverse event rates were broadly similar between active and placebo groups, and a greater proportion of musculoskeletal events in UA-treated participants were attributed to the muscle biopsy procedure rather than to the supplement itself.
Drug interactions have not been formally studied. Urolithin A is metabolised via glucuronidation and sulfonation pathways, with some involvement of CYP450 enzymes. Theoretical interactions exist with drugs that induce or inhibit hepatic CYP enzymes — including rifampicin, carbamazepine, and clarithromycin — but these have not been characterised clinically. Whether urolithin A meaningfully inhibits or induces CYP enzymes at physiological concentrations is not established. People taking anticoagulants, immunosuppressants, or complex polypharmacy should seek advice before using urolithin A, given the absence of interaction data.
Safety data in pregnancy and breastfeeding are not available. Long-term safety beyond four months has not been studied in humans, and the safety profile in populations with significant organ disease — hepatic, renal, or cardiac — has not been characterised. All current safety data derive from trials in healthy adults.
What can reasonably be concluded
Urolithin A has a well-characterised biological mechanism and a plausible rationale for supporting mitochondrial health and muscle function in ageing. The human evidence base, while early-stage, is more substantive than for most longevity-oriented compounds, with randomised controlled trials in relevant populations published in respected journals. Some functional endpoints — particularly muscle endurance in older adults and muscle strength in sedentary middle-aged adults — have shown improvement over placebo in four-month trials.
However, several important qualifications apply. The positive findings come from secondary endpoints in trials where primary endpoints were not met, and it is unclear whether they translate into meaningful improvements in mobility, independence, or fall risk in real-world settings. Despite strong mechanistic rationale, clinical outcomes have not consistently validated this pathway. The entire published clinical evidence base is tied to a single proprietary formulation manufactured by the company that sponsored all trials. Independent academic replication is absent. Trial sizes are small, durations are short relative to the timescales of ageing, and disease populations have not been studied. The biomarker changes observed — in acylcarnitines, CRP, and mitochondrial protein expression — are not validated surrogates and do not establish clinical benefit. An Emerging rating reflects the state of the evidence: a genuine early signal with substantial uncertainty remaining.
Where evidence is limited or outcomes are uncertain, conclusions should be treated as provisional and subject to revision as the evidence base develops.