Where the category stands in May 2026 — what's available, who to trust, what costs what. 22-min read, medically reviewed.
Read the guideAn independent overview of the compounded-GLP-1 telehealth market and the providers we cover — market size, what the FDA changed, and an honest read on each.
View the provider overviewOverview & reviewsPrimers for newcomers and deep dives for the curious. The foundation articles are published; more are in the pipeline.
View all Learn articlesWhat compounded and branded semaglutide and tirzepatide actually cost, from each provider's own published pricing.
In 2015, a research team at USC went looking for proteins encoded inside an unlikely stretch of DNA: the 16,569 base pairs of the mitochondrial genome [1], long assumed to code only for the machinery mitochondria need to run themselves. They found one. Sixteen amino acids long, hiding inside the 12S ribosomal RNA gene [2], with measurable effects on whole-body glucose metabolism when injected into mice. They called it MOTS-c.
That discovery did something larger than identify a new peptide. It reframed mitochondria as endocrine organs, structures that not only burn fuel but also talk back to the rest of the cell, and to other tissues, through their own peptide signals. MOTS-c is the most studied member of a small family of these mitochondrial-derived peptides (MDPs). It's also the one most aggressively promoted in the gray market of research peptides, with claims about metabolism, aging, and exercise capacity running well ahead of the human data.
This guide walks through what the evidence actually shows. Most of it's rodent work. The human story is thinner than the wellness marketing implies.
MOTS-c stands for Mitochondrial Open Reading Frame of the 12S rRNA Type-C. It's a 16-amino-acid peptide with the sequence MRWQEMGYIFYPRKLR and a molecular weight of about 2,175 daltons [3]. Small, by peptide standards.
What makes it unusual isn't the size. It's the address. Nearly every peptide in the human body, from insulin to glucagon to GLP-1, is encoded by nuclear DNA, translated in the cytoplasm, and shipped out from there. MOTS-c is encoded inside the mitochondrion, in a short open reading frame nested within a ribosomal RNA gene that biologists spent decades assuming did nothing except make ribosomes. The implication is that mitochondria, which most textbooks still describe as cellular power plants, also operate as signaling organs with their own hormonal vocabulary.
A handful of other MDPs have since been characterized, including humanin and SHLP1 through SHLP6, but MOTS-c has accumulated the most preclinical data. It circulates in human plasma. It rises during exercise. It falls with age. And in rodents, it does things to glucose handling that look genuinely interesting.
The lab was Pinchas Cohen's at USC, and the first author on the 2015 Cell Metabolism paper was Changhan Lee [4]. They were using computational tools to scan mitochondrial DNA for short open reading frames, the kind of small sequences that conventional gene-finding software tends to skip. The 12S rRNA gene flagged a candidate. They synthesized the predicted peptide, injected it into mice, and watched what happened to insulin sensitivity.
It improved. Substantially. In high-fat-diet animals, daily MOTS-c treatment over several weeks prevented diet-induced obesity and restored glucose tolerance. In 12-month-old mice, the peptide pushed insulin sensitivity back toward what the same strain looked like at 3 months. That's the kind of result that gets attention.
It also helped, narratively, that MOTS-c rose sharply in human plasma after exercise. Here was a molecule made by mitochondria, dumped into circulation when muscles worked hard, with metabolic effects that looked like a long-term training adaptation in pill form. Or so the framing went.
Worth flagging early: Lee is a consultant and shareholder of CohBar, the biotech that developed a MOTS-c analog into clinical trials. That conflict is disclosed in subsequent papers. It doesn't invalidate the work, but it's part of the context.
The mechanism is reasonably well worked out in cell and animal models, less clear in humans.
The headline effect is activation of AMPK, the cellular energy sensor. AMPK turns on when ATP runs low and ADP/AMP accumulates, and it shifts the cell into fuel-burning mode, pulling glucose in, oxidizing fat, suppressing anabolic pathways. Many of the drugs and lifestyle interventions associated with metabolic improvement (metformin, exercise, caloric restriction) hit AMPK somewhere along the way.
MOTS-c gets there indirectly. It inhibits the folate cycle and the linked de novo purine biosynthesis pathway, which causes AICAR (5-aminoimidazole-4-carboxamide ribonucleotide) to accumulate. AICAR is a known endogenous AMPK activator. Once AMPK is on, the downstream cascade looks familiar to anyone who has read about exercise physiology: GLUT4 transporters move to the muscle cell surface and pull glucose in, acetyl-CoA carboxylase gets phosphorylated and inhibited, fat oxidation rises, PGC-1α turns on mitochondrial biogenesis, and the liver throttles back gluconeogenesis.
There's a second mechanism that's genuinely novel. Under metabolic stress (glucose restriction, oxidative challenge), MOTS-c translocates into the nucleus and binds antioxidant response elements directly, regulating nuclear gene expression. A mitochondrial-encoded peptide reaching into the nucleus to set transcriptional programs is, as far as we know, the only documented example of this kind of bidirectional organelle communication. Whether it matters clinically is a separate question. But it's not nothing.
A third pathway shows up in muscle: MOTS-c suppresses myostatin via the CK2/PTEN/AKT/FOXO1 axis, which in principle should protect against muscle wasting. Rodent data supports this. Human data is correlational.
This is where the preclinical evidence is strongest, and where the human signal is most suggestive.
In rodents, MOTS-c does what an AMPK activator should do. Mice on high-fat diets given 0.5 mg/kg/day for eight weeks didn't gain the weight their controls did and kept their glucose tolerance intact (Lee 2015, PMID 25738459) [5]. Aged mice given short courses regained insulin sensitivity that approached young-animal baseline. In type 2 diabetic rats, the peptide lowered fasting glucose and improved glucose handling; at least one recent report described beneficial effects on diabetic cardiac remodeling, with improvements in cardiac structure and mitochondrial function. Gestational diabetes mouse models showed reduced hyperglycemia and better offspring outcomes.
The human side is observational. Studies examining serum MOTS-c across the spectrum from normal glucose to prediabetes to T2DM have found that levels tend to be lower in poorly controlled diabetics, those with elevated HbA1c, though the effect sizes and study designs vary across the literature. At least one pediatric obesity study has reported meaningfully lower MOTS-c levels in obese boys compared with lean controls, though the precise magnitude differs depending on cohort and methodology. The pattern is consistent: metabolic disease correlates with lower circulating MOTS-c.
Correlation is not causation. Lower MOTS-c in people with diabetes could mean the peptide is protective and people with less of it slide into disease. It could also mean diabetes damages the mitochondria that make MOTS-c, in which case the low levels are a downstream marker, not a lever.
Nobody has yet run the interventional trial that would distinguish these possibilities in humans.
If MOTS-c has a signature finding in humans, this is it. In a study of a small group of healthy young men, a single bout of stationary cycling drove skeletal muscle MOTS-c expression up by several-fold. Circulating plasma levels rose more modestly, by roughly 1.6-fold, and stayed elevated four hours after the session. Muscle levels were still markedly elevated at the four-hour mark.
That's the kind of dynamic range you'd expect from a hormone, not a structural protein.
A separate cross-sectional study of young adults found serum MOTS-c positively correlated with jumping power, peak force, and leg muscle mass. It did not correlate with VO2max. Read carefully, this suggests an association with anaerobic or strength capacity rather than aerobic capacity. The exercise mimetic label that circulates online tends to elide that distinction.
Rodent work fills in the rest. In mice, weeks of voluntary wheel running raised skeletal muscle MOTS-c by several-fold. A single MOTS-c injection in sedentary mice increased running time by about 12% and distance by 15% [6]. In one striking study of aged mice (22 months old, geriatric in mouse terms), two weeks of MOTS-c roughly doubled running time [7]. Lifespan extension in the same protocol trended upward but didn't reach statistical significance (p=0.23) [8]. Small sample. Real signal in performance, ambiguous signal in survival.
The honest read: MOTS-c clearly behaves like an exercise-induced factor in humans, and clearly improves exercise capacity when injected into rodents. The leap from those two facts to "exercise in a syringe" is a leap the data doesn't yet support.
Plasma MOTS-c declines with age. Research comparing younger and older adults has found levels drop significantly across age groups, though the precise study designs vary. The young-versus-old comparison appears robust. Whether the decline causes age-related metabolic dysfunction or merely tracks it's, again, not resolved.
The most-cited piece of human longevity evidence is a polymorphism study by Fuku and colleagues in an Asian centenarian cohort [9]. A specific variant in the MOTS-c open reading frame was overrepresented in people who reached extreme old age. Genetic association studies of this kind are suggestive, not conclusive. Populations are stratified, variants travel in linkage blocks, and replication across ethnic groups has been limited. But it's at least consistent with the broader hypothesis that MOTS-c signaling matters for healthy aging.
Mechanistically, the AMPK story does most of the heavy lifting. AMPK activation overlaps with the cellular changes seen in caloric restriction and in exercise, two of the few interventions with reasonably solid evidence for extending healthspan in mammals. If MOTS-c is one of the endogenous routes into that signaling node, the longevity framing has a defensible biological basis. What it doesn't have is a clinical trial in humans showing that exogenous MOTS-c does anything to age-related endpoints.
Several preclinical models report anti-inflammatory effects. The proposed mechanism runs through AMPK activation and downstream suppression of MAP kinase signaling, with JNK, p38, and c-Fos all dampened, which reduces pro-inflammatory cytokine output. In mouse models, intraperitoneal MOTS-c produced both antinociceptive and anti-inflammatory effects.
Cross-sectional human data show lower circulating MOTS-c in stable COPD patients, in obstructive sleep apnea, and in chronic kidney disease, where plasma levels in CKD patients have been reported as meaningfully depressed compared with healthy controls. These are diseases of chronic inflammation and mitochondrial dysfunction. The MOTS-c deficit could be cause or consequence; the studies can't distinguish.
Preclinical work has also looked at vascular calcification, pulmonary fibrosis in cultured human lung cells, and bone loss in ovariectomized mice. The bone finding (MOTS-c suppressed ovariectomy-induced bone loss via AMPK) is interesting given the overlap with metabolic and endocrine aging. None of this has reached human trials.
Stated plainly: the human evidence for MOTS-c is almost entirely observational.
What exists:
Cross-sectional studies showing lower MOTS-c in people with metabolic disease (T2DM, pediatric obesity, COPD, OSA, CKD), and in older adults compared to younger ones. A genetic study linking a MOTS-c polymorphism to longevity in an Asian centenarian cohort. A small exercise physiology study showing acute exercise-induced rises in muscle and plasma MOTS-c. A correlation between serum MOTS-c and anaerobic muscle performance in young adults. A correlation between MOTS-c and plasma myostatin levels.
What doesn't exist:
A completed Phase 2 or Phase 3 efficacy trial of MOTS-c in humans for any indication. Not for diabetes, not for obesity, not for sarcopenia, not for aging, not for exercise performance.
The closest thing to interventional human data is a Phase 1a/1b trial of CB4211, a MOTS-c analog developed by CohBar, in patients with nonalcoholic fatty liver disease [10]. The trial enrolled dozens of subjects. Full peer-reviewed results have not been published in the sources we reviewed; CohBar has since wound down operations. That's essentially the entire human interventional record on MOTS-c-class compounds.
For a peptide marketed as a metabolic and longevity intervention, this is a thin file.
The unknowns aren't just "more research needed" filler. They're specific and they matter.
Half-life in humans is not established. Pharmacokinetic parameters (absorption, distribution, clearance) have not been published for any human administration route. Without that, there's no rational basis for translating rodent dosing into human protocols. The 0.5 to 15 mg/kg ranges used in mice were chosen by researchers based on what produced measurable effects in mice [11]. Whether the human-equivalent dose lands at 5 mg or 50 mg per injection, and how often, is a matter of guesswork.
The peptide doesn't cross the blood-brain barrier via systemic administration. CNS effects reported in mouse models required intracerebroventricular or intranasal delivery with a cell-penetrating carrier. Marketing claims that systemic MOTS-c will improve cognition don't square with the route-of-administration data.
The direction of causality in human observational studies is unresolved. Low MOTS-c in disease might be a driver or a downstream marker. Until someone runs an intervention in humans and measures endpoints, this stays unanswered.
Sex-dependent regulation has been underexplored. Male animals show greater disruption of MOTS-c in metabolic disease; premenopausal women may have estrogen-driven mitochondrial biogenesis that blunts the decline. Most rodent studies were done in male mice. What that means for dosing or response prediction in women has not been worked out.
MOTS-c is not approved by the FDA for any indication. It's classified as investigational.
The compounding pathway is where things get interesting, and where most consumer-facing MOTS-c access currently sits. The FDA has placed MOTS-c on its list of substances in compounding that may present real safety risks [12], citing three concerns: possible immunogenicity for some routes of administration, complexities around peptide-related impurities and active ingredient characterization, and the absence of human exposure data through any route. That last point, the FDA explicitly stating it lacks information about whether MOTS-c would cause harm if administered to humans, is unusual language for a compound that some clinics already dispense.
That review has now happened. MOTS-c was one of twelve peptides FDA removed from Category 2 on April 15, 2026, and on July 23 the agency's Pharmacy Compounding Advisory Committee voted to recommend it for the 503A bulks list — 7 in favour, 5 against, 2 abstentions, with the free base and the acetate voted separately and landing the same way [13]. The uses FDA had evaluated were obesity and osteoporosis. The agency's own reviewers had proposed rejecting it, on the same grounds set out above: no human exposure data, unresolved characterisation and impurity questions.
The vote does not change what a pharmacy may do. Committee recommendations are non-binding, FDA has issued no final determination, and MOTS-c is still not on the 503A Bulks List — which is what would make it legally compoundable, and which requires rulemaking that has not started and has no timeline. Our report on the vote has the record. The peptide remains in the gray zone, now with a favourable advisory opinion attached to it.
WADA, the World Anti-Doping Agency, has listed MOTS-c on its Prohibited List as a banned substance at all times under the AMPK activator category. Athletes subject to drug testing should treat this as a hard line.
No major medical society (ADA, AHA, Endocrine Society) references MOTS-c in clinical practice guidelines.
The honest answer on safety: we don't have one, at least not for humans.
The FDA's own position is that it has no human exposure data on MOTS-c products through any route and lacks sufficient information to characterize the safety profile. A FAERS database query returned no adverse event signal specific to MOTS-c. That isn't reassuring. It reflects the absence of formal pharmacovigilance, not the presence of a clean record.
Preclinical safety in rodents has been generally unremarkable across the published dose ranges studied. No overt toxicity signals, no meaningful effects on food intake or body weight in normal-diet animals at therapeutic doses. Animals tolerate it. Whether that translates to a small synthetic peptide injected into humans over months or years is genuinely not known.
The theoretical concerns the FDA has flagged are reasonable. Immunogenicity is a known issue with peptide therapeutics: the immune system can build antibodies that neutralize the drug, or worse, cross-react with endogenous human MOTS-c. Impurity profiles in compounded peptides vary by source and have been documented as inconsistent in independent testing of other research peptides. None of this is MOTS-c-specific data. It's the standard risk surface for an unregulated injectable peptide with no human safety database behind it.
Clinics that dispense MOTS-c in a research-informed context generally advise against use in active or recent cancer, in pregnancy or breastfeeding, and in patients unwilling to undergo baseline labs. These are reasonable defaults given how much we don't know.
No, but they're cousins. All three are mitochondrial-derived peptides, encoded inside mitochondrial DNA rather than nuclear DNA. Humanin was the first MDP discovered, in 2003 [13], and has been studied primarily in neurodegeneration and cytoprotection. The SHLPs (small humanin-like peptides) are six related peptides with overlapping metabolic and apoptotic effects [14]. MOTS-c is structurally distinct, encoded in a different mitochondrial gene, and signals through AMPK in a way the others don't.
This is the marketing claim. The honest version is narrower: MOTS-c rises sharply during exercise in humans, and exogenous MOTS-c improves exercise performance in rodents. It activates AMPK, which exercise also activates. But exercise produces dozens of adaptive signals across multiple organ systems, and reducing all of that to a single peptide injection is a leap the human data doesn't support. No published trial has tested whether MOTS-c administration improves fitness, body composition, or metabolic markers in humans.
It depends on what "buy" means. MOTS-c is not FDA-approved for any indication. An FDA advisory committee voted 7–5 with two abstentions in July 2026 to recommend adding it to the 503A Bulks List, but that advice is non-binding and FDA has not acted, so it is still not on the list as a permitted compoundable substance. Some compounding pharmacies dispense it under physician supervision in a regulatory gray area. Research-only suppliers sell it labeled "not for human consumption," which transfers legal risk to the buyer. Athletes subject to WADA testing shouldn't use it under any circumstance.
Plasma levels in healthy young adults run around 125 pg/mL at rest, rising to about 190 pg/mL after acute exercise. Levels decline with age and are lower in metabolic disease. Serum median in controls is about 3.89 ng/mL. These are reference points, not targets. No one has established what a "therapeutic" level would be in someone receiving exogenous peptide.
Two reasons. The most advanced human program, CohBar's CB4211, got through Phase 1 and then the company wound down operations before larger trials could launch. Beyond that, MOTS-c is hard to develop as a conventional drug: it's a peptide (oral bioavailability is poor), the patent landscape around a naturally occurring human peptide is messy, and the indication space (general metabolic improvement, aging) doesn't map cleanly onto FDA approval pathways. The peptide may have more commercial life as an analog or modified version than as MOTS-c itself.
MOTS-c is a real discovery with a thin clinical file. The biology, a peptide encoded in mitochondrial DNA, secreted under metabolic stress, activating AMPK, translocating to the nucleus to regulate gene expression, is genuinely novel and worth understanding. The animal data on metabolism and exercise capacity is consistent and reasonably reproduced. The human observational data is suggestive in the same direction.
What's missing is the trial that would tell anyone whether injecting MOTS-c into a person does what the marketing claims. We don't have safety data through formal channels. We don't have established dosing. We don't have human efficacy outcomes for any condition. The FDA has explicitly said it lacks the information to characterize risk.
That gap is the story. The peptide is interesting. The clinic ahead of the evidence is the part to be cautious about.
Editorial note: Informational only — not medical advice. Decisions about peptide therapy should be made with a licensed healthcare provider familiar with your medical history. See our methodology. Last reviewed July 2026.