Why Urolithin A Trials Keep Measuring Muscle

Four of the nine human urolithin A trials I could verify measured muscle, so muscle is the largest group but not a majority. The papers point to three reasons for the muscle focus: the animal work measured muscle function, the first human trial measured muscle gene activity, and strength and endurance can be scored with functional tests. Three of the four muscle trials list Amazentis SA among the authors' affiliations.

The short answer Four of the nine human urolithin A trials I could verify measured muscle. Muscle is the largest group, but not a majority. The papers point to three reasons for the muscle focus: the animal work measured muscle function, the first human trial measured muscle gene activity, and strength and endurance can be scored with functional tests. Blood, immune cells, skin (topical only), and blood-vessel function have also been measured in people. The human heart and brain evidence is one 10-person heart trial and no oral trial of cognition or brain structure that I could find.

The ingredient guide covers the compound itself, and the full urolithin A guide covers the mechanism. This page covers one question: why the human trials looked at muscle, and what else they looked at. The idea that most human trials measured muscle is only partly right. Four of nine is close to half, and muscle is still the largest single group.

Why do urolithin A trials focus on muscle?

The starting point was animals. In 2016, Ryu and colleagues reported that urolithin A improved exercise capacity in two mouse models of age-related muscle decline and in young rats (Ryu et al., 2016, Nature Medicine). The first human study came later. Andreux et al., 2019, Nature Metabolism was a first-in-human safety trial in healthy, sedentary older adults. Its secondary outcomes included plasma acylcarnitines and mitochondrial gene expression in skeletal muscle after 500 mg and 1,000 mg a day for four weeks.

The papers point to three reasons for the muscle focus:

• The animal data pointed at muscle. The mouse work that the human trials built on measured muscle function.
• The human trials could sample muscle. Andreux measured muscle gene expression, and Singh et al., 2022, Cell Reports Medicine measured mitophagy and mitochondrial proteins in skeletal muscle.
• Muscle function has functional tests. Singh 2022 used peak oxygen uptake and a six-minute walk. Liu et al., 2022, JAMA Network Open counted contractions to fatigue in a hand muscle and a leg muscle, and used magnetic resonance spectroscopy to measure ATP production. Zhao et al., 2024 measured maximal isometric contraction and repetitions to failure in resistance-trained men.

My inference, which none of these papers states: a needle biopsy and a walk test are both quick and inexpensive, so muscle is an easy first tissue to measure. Treat that as a guess about practicality, not a finding.

The papers do not say why the authors chose muscle, and affiliations are part of the picture. Three of the four muscle trials list Amazentis SA among the authors' affiliations: Andreux 2019, Singh 2022, and Liu 2022. The fourth, Zhao 2024, lists none. Read the muscle results with that in mind.

Muscle was not the only tissue studied. The animal and cell work covers the brain, retina, cartilage, arteries, and immune cells. So muscle is where people looked first. It is not proof that urolithin A acts only there.

What did the human trials measure?

Nine human trials gave urolithin A itself. Eight gave it by mouth, and one applied it to the skin.

The grades use the site's vocabulary. Muscle, blood, immune, and topical skin are Preliminary, because each rests on a few small or mixed trials. In the two muscle trials with performance primary endpoints, the primary endpoint did not reach significance. Heart, vessels, brain, retina, and cartilage are Insufficient, because each rests on one small trial, one diet correlation, or animal and cell work.

The 100-person study by Singh et al., 2021 was not about function. It randomized healthy adults to a glass of pomegranate juice or a 500 mg urolithin A product. About 40% of people turned the juice into urolithin A, and the product gave more than six times the plasma exposure of the juice. The food side of this is covered in why most adults make little urolithin A from pomegranate.

Has urolithin A been studied outside muscle in people?

Yes, in four areas, each with a small or mixed record.

Immune cells. Denk et al., 2025, Nature Aging gave 1,000 mg a day for four weeks to 50 healthy middle-aged adults, compared with placebo. The main shift was small. A naive-like CD8 T-cell subset rose by 0.50 percentage points (95% CI 0.16 to 0.83), and CD8 fatty-acid oxidation capacity rose by 14.72 percentage points.

Skin, applied topically. D'Amico et al., 2026, iScience tested urolithin A applied to the skin in randomized, placebo-controlled studies. Periorbital wrinkle depth fell after eight weeks, and UVB-induced redness was reduced. A repeat-insult patch test supported safety. The paper is topical, not oral, so it says nothing about what a capsule does for skin. Several co-authors are from L'Oréal and Lancôme, and readers should know that.

Blood vessels and brain blood flow. da Costa Pinaffi-Langley et al., 2026, JAMA Network Open gave 1,000 mg a day for four weeks to 50 middle-aged adults with obesity. The primary measure, brachial artery flow-mediated dilation, did not differ from placebo. Brain blood-flow responses during a task were larger in the urolithin A group, and task performance did not change. The authors call that effect worth further study.

Blood markers. Plasma acylcarnitines and C-reactive protein were lower with urolithin A in two of the muscle trials (Singh 2022; Liu 2022).

Has urolithin A been tested in the brain or the heart?

Brain. Not in a dosed human trial of cognition or brain structure that I could find. Most of the brain evidence comes from mice and worms. Hou et al., 2024 reported better learning, memory, and smell after long-term treatment in three Alzheimer's mouse models. Fang et al., 2019 reported that mitophagy stimulation, with urolithin A among the compounds tested, reversed memory deficits in worm models of Alzheimer's. In people, a diet trial found that higher urinary urolithin A went with less decline in hippocampal occupancy score (Kaplan et al., 2022). That is a correlation among people following a dietary pattern, not a dosing result.

Heart. Once, in a very small trial. Jamialahmadi et al., 2024 gave 500 mg twice daily to 10 people with heart failure with reduced ejection fraction. Each person had one 4-week period on urolithin A and one on placebo, separated by a 2-week washout, in a crossover design. Echocardiogram measures did not change significantly, and pro-BNP and CRP were unchanged. HDL cholesterol rose by about 6.5 mg/dL against placebo. The authors concluded that the results do not support a benefit and suggested higher doses and longer use. A 2024 systematic review of five human studies in 250 healthy people reported no effect on cardiovascular outcomes. The mouse work on artery plaque (Xu et al., 2024) is preclinical.

How should you read these trials?

Four checks sort out most of the confusion.

• Find the primary endpoint first. In Singh 2022, the primary endpoint was peak power output, and it did not improve significantly. In Liu 2022, the primary endpoints were the six-minute walk and ATP production in a hand muscle, and neither differed significantly from placebo. The gains that were reported came from other measures. The interactive clinical data charts lay out each endpoint.
• Match the dose and length. The human trials used 500 to 1,000 mg a day, for four weeks to four months. None of the nine trials went above 1,000 mg a day. The dosage page explains what that does and does not show.
• Check the population. Liu 2022 enrolled adults aged 65 to 90. Denk 2025 enrolled healthy middle-aged adults. Zhao 2024 enrolled resistance-trained men, and the heart trial enrolled patients with heart failure. A result in one group does not transfer to another.
• Check the affiliations. Six of the nine human trials list Amazentis SA among the authors' affiliations. The three that do not are the heart trial, the athlete trial, and the obesity trial. Affiliation does not make a result wrong, but it is the first thing a careful reader should check.

The decision by situation:

• If you want a muscle claim: the trials show small gains on some endurance and strength measures, with mixed primary endpoints. The evidence is preliminary.
• If you want a brain claim: there is no oral human trial of cognition or brain structure. The mouse and worm data are not human results.
• If you have heart failure: the one heart trial found no echocardiogram benefit. Talk with your cardiologist before any change.

When to stop or ask someone

Ask a clinician before you start if you take prescription medicines, have heart failure, or are pregnant or trying to become pregnant. None of the abstracts I checked report interactions with prescription drugs or outcomes in pregnancy, so that gap is real. The safety data come from trials lasting four weeks to four months. Liu et al., 2022 found no significant difference in adverse events from placebo. The 2024 review described the adverse events it saw as mild or moderate and unrelated to treatment. If new symptoms appear after you start, stop and call a clinician.

Next reads

• The urolithin A complete guide sets out the mechanism and the wider evidence that this question sits inside.
• Urolithin A in trained runners explains how the athlete question differs from the adult trials.
• How to compare urolithin A labels before you buy, including dose per day and what to check on a label.

More on Urolithin A: The Complete Guide