Urolithin A and Mitophagy

Urolithin A is a metabolite gut bacteria make from ellagitannins in pomegranates, walnuts and some berries. It is one of the few longevity-adjacent compounds with randomized, placebo-controlled human trials behind it, which makes it worth reading closely rather than dismissing. The question this article answers is narrow and specific: what endpoints did those trials measure, and did they hit them?

What the trials actually tested

The first-in-human study, published by Andreux and colleagues in Nature Metabolism in 2019, was a double-blind, randomized, placebo-controlled trial in healthy sedentary elderly volunteers. Its primary outcome was safety, and urolithin A passed. Four weeks of dosing at 500 mg and 1,000 mg also changed plasma acylcarnitines and skeletal muscle mitochondrial gene expression — but those were secondary outcomes, and the authors described the result as “a molecular signature of improved mitochondrial and cellular health,” not as a functional benefit.

Two efficacy trials followed, and both missed their primary endpoints.

Trial	Population	Dose / duration	Primary endpoint	Result
Andreux 2019, Nature Metabolism	Healthy sedentary elderly	500 / 1,000 mg, 4 weeks	Safety	Met — favourable safety profile
Liu 2022, JAMA Network Open	66 sedentary adults aged 65–90	1,000 mg, 4 months	6-minute walk distance; maximal ATP production in hand muscle	Neither significant vs placebo
Singh 2022, Cell Reports Medicine	Middle-aged adults	Two doses, 4 months	Peak power output	Not significant
Denk 2025, Nature Aging	50 adults aged 45–70	1,000 mg, 4 weeks	CD8+ T-cell phenotype and fatty-acid oxidation	Met, with small differences

In the JAMA Network Open trial, six-minute walk distance rose 60.8 m on urolithin A and 42.5 m on placebo — a real improvement in both arms, and no statistically significant separation between them. Maximal ATP production showed no treatment effect either. What did separate was a secondary measure: at two months, contractions to fatigue increased more on urolithin A in both the first dorsal interosseous (+95.3 vs +11.6) and tibialis anterior (+41.4 vs +5.7) muscles. By four months both groups had improved and the difference was gone.

Across four randomized trials, urolithin A’s positive findings sit almost entirely in secondary and exploratory endpoints — muscle endurance, plasma acylcarnitines, gene and protein expression, immune-cell markers. The pre-specified primary endpoints in the two efficacy trials were not met.

Two things worth knowing before you read a headline

• All of this research is industry-funded. Andreux 2019, Liu 2022, Singh 2022 and Denk 2025 were funded by Amazentis SA, which commercialises urolithin A, and company employees and advisors appear on the author lists of all four. Industry funding does not invalidate a peer-reviewed randomized trial, but it does mean there is no independent replication yet.
• Most people do not make much of it themselves. In a crossover study of 100 healthy adults aged 18–80, only 12% had detectable urolithin A in plasma on their normal diet, and roughly 40% converted pomegranate precursors into it at 24 hours — the other 60% converted poorly or not at all. Direct supplementation with 500 mg produced more than six-fold higher exposure than pomegranate juice and levelled out that variability. (This study, too, was funded by Amazentis.)

What has not been measured

No trial has run long enough, or in enough people, to say anything about healthspan, disability, falls, independence, or survival. The 2025 Nature Aging trial in 50 adults is the most recent, and its endpoints were immune-cell composition and metabolic capacity — a treatment difference of 0.50 percentage points in naive-like CD8+ cells and 14.72 percentage points in CD8+ fatty-acid oxidation over four weeks. These are mechanistic readouts. They are not outcomes anyone experiences.

That pattern — a coherent mechanism, real trials, surrogate endpoints — is the norm rather than the exception in this field, and we cover why in why most longevity supplements disappoint and in our primer on autophagy.

The takeaway

Urolithin A is better evidenced than most of the longevity shelf, and that is a low bar. It has an established safety profile in short trials, a reproducible effect on mitochondrial and metabolic biomarkers, and one secondary muscle-endurance signal that did not persist to four months. Both efficacy trials failed their primary endpoints, all the published trials come from the company selling it, and nothing has been demonstrated about how long or how well anyone lives. If you were waiting for evidence that it changes an outcome you would notice, it does not exist yet.

Sources

• Andreux et al., “The mitophagy activator urolithin A is safe and induces a molecular signature of improved mitochondrial and cellular health in humans” (Nature Metabolism, 2019)
• Liu et al., “Effect of Urolithin A Supplementation on Muscle Endurance and Mitochondrial Health in Older Adults” (JAMA Network Open, 2022)
• Singh et al., “Urolithin A improves muscle strength, exercise performance, and biomarkers of mitochondrial health in a randomized trial in middle-aged adults” (Cell Reports Medicine, 2022)
• Denk et al., “Effect of the mitophagy inducer urolithin A on age-related immune decline” (Nature Aging, 2025)
• Singh et al., “Direct supplementation with Urolithin A overcomes limitations of dietary exposure and gut microbiome variability” (European Journal of Clinical Nutrition, 2021)