Do any peptides improve endurance in humans?

No human trial has demonstrated it. The two compounds the category is built on — AICAR and MOTS-c — have two and five ClinicalTrials.gov registrations respectively between them, and not one tests endurance performance in trained people. The animal results are real and were published in Cell and Nature Communications, but every one of them was obtained in rodents, and in the AICAR case in rodents that were not exercising.

What is AICAR and why does it keep coming up?

AICAR is a small molecule that activates AMPK, the enzyme that senses low cellular energy and switches muscle toward oxidative metabolism — the same switch endurance training flips. In 2008 a Cell paper reported that four weeks of it raised running endurance in sedentary adult mice by 44%, and that single sentence created a market. It is not a peptide, which is worth noting given where it is sold.

Is MOTS-c an exercise mimetic?

In mice it improved physical performance at three ages, including when started late in life. In humans the same research group showed the opposite direction of causation: exercise raises MOTS-c in muscle and blood. Both findings are in the same paper. A peptide that training produces is an interesting biomarker of adaptation; treating it as a replacement for the training that produces it is an inference the human data does not support.

Does SS-31 improve aerobic capacity?

It is the fairest test the category has had, because SS-31 acts directly on mitochondria and has been through real drug development. In 41 older adults the primary endpoint was maximal ATP synthetic rate in skeletal muscle, and the drug-placebo differences were smaller than their own standard deviations, with placebo moving further than drug on mitochondrial coupling at day seven. That is a short exposure in a small sample rather than a definitive negative, but it is what exists.

Are endurance peptides banned in sport?

The growth-hormone-axis compounds are, unambiguously: on the 2026 Prohibited List, growth hormone releasing factors sit in S2.2.4 and IGF-1 and thymosin beta-4 in S2.3, all prohibited at all times, which our drug-testing page sets out with the list wording. The metabolic modulators sold for endurance are governed by a separate section of the same list, and any athlete subject to testing should check the current published text rather than a secondary summary — including this one.

What actually improves endurance, on the evidence?

Training, and then a short list of legal supplements. The International Olympic Committee's consensus statement puts it in one clause: of the large number of products marketed to athletes, only a few — it names caffeine, creatine, specific buffering agents and nitrate — have good evidence of benefit. Each of those has been tested repeatedly in trained humans against placebo, which is precisely what no endurance peptide has.

Peptides for Endurance: the Market Descends From One Experiment on Sedentary Mice

Every compound sold as an endurance enhancer traces back to a 2008 mouse study in which untrained animals ran 44% longer. Eighteen years later, the human column is still empty.

Endurance is the hardest claim to fake, because it is measured with a clock. A compound that adds muscle can hide behind photographs and scale weight; one sold for aerobic capacity has to survive a time trial.

Which makes the state of the evidence here unusually clean to describe. The category exists. Its human column is empty.

This is research journalism. It reports what the sources state, including the amounts used in trials and on labels where those are on the record, each figure carrying its source and date — and it never tells anyone what to take.

One paper, eighteen years ago, in mice that did not train

In 2008 a group at the Salk Institute published a paper in Cell titled "AMPK and PPARdelta agonists are exercise mimetics" (PMID 18674809). Its aim was stated plainly and was not athletic: to find orally active drugs that reproduce the metabolic benefits of endurance exercise, for the treatment of metabolic disease.

The finding that escaped the paper was this. Mice given the AMPK activator AICAR for four weeks — with no exercise training of any kind — increased treadmill running endurance by 44%.

That sentence is the endurance-peptide market. Every "exercise in a bottle" pitch, every AMPK diagram on a vendor page, and a considerable amount of the anti-doping literature descend from it.

Eighteen years later, an intervention search of ClinicalTrials.gov returns two registrations for AICAR in total, and neither tests endurance performance in trained humans. The molecule that launched the category has been studied in people mainly for other reasons entirely, and — a detail the marketing rarely carries — AICAR is not a peptide.

MOTS-c, and a causal arrow pointing the wrong way

The second pillar is newer and better documented. MOTS-c is a peptide encoded in mitochondrial DNA rather than in the nuclear genome, and in 2021 a Nature Communications paper reported that it improved physical performance in mice at 2, 12 and 22 months of age, with benefits still appearing when treatment began at 23.5 months (PMID 33473109).

The same paper contains the human result, and it runs in the opposite direction to the sales pitch: exercise induced MOTS-c in human skeletal muscle and in circulation.

Read together, those two findings describe a peptide that training produces. That is genuinely interesting — it makes MOTS-c a candidate biomarker of exercise adaptation. It is not evidence that supplying the peptide reproduces the training, and the paper does not claim it is. The market's version of the story quietly reverses the arrow.

Five registrations exist for MOTS-c. Its 254 PubMed records are mostly mechanistic and mostly rodent or cell work.

The compound that got a fair test, and did not separate

There is one member of this class with a completed drug-development programme, a marketing approval and a mechanism aimed squarely at mitochondrial energy production: elamipretide, sold in the grey market as SS-31.

Its trial in ordinary older adults measured exactly the right thing — maximal ATP synthetic rate in skeletal muscle, a direct index of mitochondrial capacity. In 41 participants, the change was 0.17 mM/sec on drug against 0.12 on placebo two hours after infusion and 0.09 against 0.06 at day seven, with standard deviations two to three times the size of those differences. On the secondary measure of mitochondrial coupling, placebo moved further than drug by day seven.

The full account, including the narrow rare-disease approval it did earn, is on our SS-31 page. Its relevance here is that when the category's best-developed molecule was given a proper mitochondrial endpoint in humans, nothing separated.

Thymosin beta-4, sold as TB-500 and marketed for recovery and work capacity, has 18 registrations. None carries an endurance or aerobic-capacity endpoint; the compound is covered alongside BPC-157 on our Wolverine-stack page.

The registry pattern, stated once

Compound	ClinicalTrials.gov registrations (intervention search)	Registrations with an endurance endpoint in trained humans
AICAR	2	0
MOTS-c	5	0
Thymosin beta-4 (TB-500)	18	0
Elamipretide (SS-31)	21	0 (one 41-participant mitochondrial-capacity trial in older adults)

This is the same shape this site has found in every compound class it has audited: a real mechanism, a real animal result, a large detection literature, and an empty column where the human performance trial should be. The growth-hormone secretagogues produced it, the comparison with anabolic steroids turned on it, and it recurs here.

What the clock actually responds to

The honest comparison is not between peptides and nothing. It is between peptides and the interventions that have been tested in trained humans against placebo, repeatedly, with performance endpoints.

The International Olympic Committee's consensus statement on dietary supplements and the high-performance athlete settles that in a single clause: of the many products marketed to athletes, "only a few (including caffeine, creatine, specific buffering agents and nitrate) have good evidence of benefits" (PMID 29540367, British Journal of Sports Medicine, 2018). The same statement warns that supplement use carries a real risk of inadvertent ingestion of prohibited substances.

Creatine's place on that list is the bridge worth noticing, and it is the same conclusion our creatine comparison reached from the muscle-growth side: the compounds with the strongest human evidence in this whole field are the cheap legal ones, and the ones with the best stories are the ones with the empty registry.

For a tested athlete there is a second asymmetry. Caffeine and nitrate are unrestricted. The growth-hormone releasing factors and IGF-1 are prohibited at all times under S2.2.4 and S2.3 of the 2026 list, and the metabolic modulators sold for endurance fall under a separate section of the same document — the per-compound status this site has verified against the published text is on our drug-testing page.