Urolithin A side effects: a clean record, but short trials

Most of our side-effect write-ups are cautionary. This one is different: urolithin A is one of the few longevity supplements whose safety data actually looks clean. The interesting question isn’t “what goes wrong” — it’s “how confident can we be, given how short the trials were?” For the efficacy side of the ledger, see our urolithin A evidence monograph.

What the human trials found: well tolerated

The safety story starts with a proper first-in-human study. In a randomized, double-blind, placebo-controlled trial in sedentary elderly adults, urolithin A was given as single doses up to 2,000 mg and as 500–1,000 mg per day for 28 days. It was safe and well tolerated, with no serious adverse events and a pharmacokinetic and adverse-event profile that did not differ meaningfully from placebo.^[1]That is a stronger safety foundation than almost anything else on the longevity shelf.

The two later muscle RCTs reinforced it. In middle-aged adults, four months of supplementation was reported as safe with no significant excess of adverse events.^[2] In a separate trial in older adults, urolithin A up to 1,000 mg per day for four months was again safe and well tolerated, with adverse events comparable to placebo.^[3] Across the program, the tolerability finding replicated even where the efficacy findings were modest.

The mild, rare stuff

When adverse events did turn up, they were minor and did not cluster into a pattern. The most commonly mentioned complaints in this class are occasional, mild gastrointestinal effects — the sort of thing you would expect from any oral supplement — and importantly they did not occur at a higher rate than placebo in the controlled trials.^[3] No dose-limiting toxicity, no signal on liver or kidney panels, and no serious adverse events attributable to the compound were reported across the human program.^[1] The preclinical toxicology backs this up: a formal safety assessment in animals found no adverse effects and supported urolithin A’s recognition as generally safe, the basis for its GRAS (Generally Recognized As Safe) status.^[4]

The honest limits

Here is where balance matters. The trials that established this clean profile ran weeks to a few months, not years.^[1][3] Mitophagy induction is a biologically active process, and no one has characterized what daily use looks like over the multi-year horizons people actually take longevity supplements for — so long-term safety is unestablished, not proven benign. The trial populations were also relatively healthy adults; effects in people with significant chronic disease are less well studied.

Two more honest caveats. First, drug–supplement interactions are largely unstudied — urolithin A has not been systematically tested against common medications, so “no known interactions” mostly reflects a lack of investigation rather than a clean bill of health. If you take prescription drugs, that gap is worth a conversation with your clinician. Second, it is a supplement, not an approved drug: it hasn’t been through the long-term outcome trials and post-marketing surveillance that would let anyone speak confidently about rare or delayed harms. Staying within studied doses is the sensible move — see our urolithin A dosage guide for what the trials actually used.

The honest bottom line

Urolithin A is genuinely one of the better-tolerated compounds we cover: across a first-in-human safety study and two placebo-controlled RCTs, it was well tolerated with adverse events comparable to placebo, no serious events, and a supporting toxicology package behind its GRAS status.^[1][4]The mild reports that exist are occasional and gastrointestinal, not a dose-limiting pattern. The single real caveat is duration: this is a clean short-term record, and multi-year safety plus interactions remain genuinely open questions. Treat it as a low-risk supplement whose safety story is reassuring but incomplete — which is a very different, and more defensible, profile than a peptide like MOTS-c, whose human side-effect data barely exists.