Fat loss.

This is the most price-aware query in the peptide category, and it happens to be the one with the best evidence base attached to it. All four molecules below have randomised controlled trials behind them, and three of them have their headline numbers printed in abstracts of the New England Journal of Medicine. That is genuinely rare in this field. It is also exactly where the risk sits, because the numbers are large enough that a careless sentence turns a trial into a promise. None of these trials tested a product bought for research. Every one of them tested pharmaceutical-grade, protocol-managed product in a clinical setting, and the distance between those two things is the whole subject of this page.

What the literature shows

PMID 33567185 — A double-blind trial in 1 961 adults with a body-mass index of 30 or higher, or 27 or higher with at least one weight-related condition, none of whom had diabetes. Both arms received lifestyle intervention. Funded by the manufacturer. Sixty-eight weeks of treatment. The coprimary endpoints were percentage change in body weight and weight reduction of at least 5%.

Mean change in body weight from baseline to week 68 was −14.9% on semaglutide against −2.4% on placebo, a treatment difference of −12.4 percentage points (95% CI −13.4 to −11.5, P < 0.001), or −15.3 kg against −2.6 kg. Weight reduction of 5% or more: 86.4% against 31.5%. Of 10% or more: 69.1% against 12.0%. Of 15% or more: 50.5% against 4.9%. Nausea and diarrhoea were the most common adverse events, typically transient and mild to moderate. Discontinuation owing to gastrointestinal events was 4.5% against 0.8%.

PMID 35658024 — A phase 3 double-blind trial in 2 539 adults with obesity, excluding diabetes, over 72 weeks including a 20-week dose-escalation period. Baseline mean weight was 104.8 kg and mean BMI 38.0, with 94.5% of participants at a BMI of 30 or above. Read the population size and the mean BMI together — this is a trial cohort, not a shopper.

Mean percentage change in weight at week 72 was −15.0% (95% CI −15.9 to −14.2), −19.5% (−20.4 to −18.5) and −20.9% (−21.8 to −19.9) across the three active dose levels, against −3.1% (−4.3 to −1.9) on placebo, P < 0.001 for all comparisons. Weight reduction of 20% or more was reached by 50% (CI 46 to 54) and 57% (CI 53 to 61) in the two higher levels, against 3% (CI 1 to 5) on placebo. Adverse events caused discontinuation in 4.3%, 7.1% and 6.2% across the active levels against 2.6% on placebo. Most adverse events were gastrointestinal, occurring primarily during dose escalation.

PMID 37366315 — A phase 2 — not phase 3 — double-blind trial in 338 adults, 51.8% of them men, over 48 weeks, covering six active dose levels plus placebo. This is the largest number in the file and the smallest dataset behind it. The distinction is not a formality: phase 2 exists to find out whether a signal holds, not to confirm an outcome.

At 48 weeks the least-squares mean percentage change in body weight was −8.7% in the lowest dose level, −17.1% in the combined next level, −22.8% in the combined next and −24.2% in the highest, against −2.1% on placebo. At 48 weeks, weight reduction of 15% or more had occurred in 60% of participants in the lower of the two mid dose levels, 75% in the upper of them and 83% in the highest level, against 2% on placebo. The most common adverse events were gastrointestinal, dose-related and mostly mild to moderate, partially mitigated by a lower starting dose. Dose-dependent increases in heart rate peaked at 24 weeks and declined thereafter.

PMID 22495074 — Two phase 3 randomised double-blind studies in men and women with HIV-associated abdominal fat accumulation, 26 weeks followed by a randomised 26-week continuation, with 402 subjects in the per-protocol analysis. This is the only study on the page whose primary endpoint is visceral fat rather than body weight, and the only molecule here that is an approved medicine rather than an investigational one. The analysis compares responders with non-responders — that structure matters when reading the numbers.

Responders were defined a priori as those with at least an 8% reduction in visceral adipose tissue. Compared with non-responders at 26 weeks, responders showed a greater mean reduction in triglycerides, −0.6 ± 1.7 against −0.1 ± 1.2 mmol/L (P = .005), attenuated change in fasting glucose, +1 ± 16 against +5 ± 14 mg/dL (P = .01), and attenuated change in haemoglobin A1c, 0.1 ± 0.3% against 0.3 ± 0.4% (P < .001). The paper’s background records that the compound decreases visceral adipose tissue by 15% to 20% over 6 to 12 months in this population.

Where these compounds stand

The two most recognised drugs on this page are the two the Prohibited List never mentions, and the compound with the least regulatory paperwork is the one the list reaches through a catch-all clause. Both halves of that are checkable, so each row below carries the list section and the database entry rather than a verdict.

The list is the 2026 Prohibited List, in force since 1 January 2026. ‘Medicine status’ is the US position, read from FDA Drugs@FDA on 1 October 2026: an application number means a marketing authorisation exists in that database, and ‘no application’ means the database holds none under that name — a statement about the database, not a clearance for anything.

Compound	Status	Source
Tirzepatide	Not prohibited in sport: the list never names tirzepatide, and the S0 catch-all only reaches substances with no current approval. Approved medicine — NDA215866 (Mounjaro) and NDA217806 (Zepbound).	WADA 2026 · Drugs@FDA
Semaglutide	Not prohibited in sport, on the same reading. Approved medicine — NDA213051 (Ozempic, Rybelsus) and NDA215256 (Wegovy).	WADA 2026 · Drugs@FDA
Retatrutide	Not named in the list and no application in Drugs@FDA, so S0’s catch-all covers it: a substance with no current approval for human therapeutic use.	WADA 2026, S0 · Drugs@FDA
Tesamorelin	Prohibited — S2.2.4 names tesamorelin among the GHRH analogues. Approved medicine — BLA022505 (Egrifta). One of the two rows here that is both listed and approved.	WADA 2026, S2.2.4 · Drugs@FDA
AOD9604	Prohibited — S2.2.3 names it as a growth hormone fragment, “e.g. AOD-9604 and hGH 176-191”. No application in Drugs@FDA.	WADA 2026, S2.2.3 · Drugs@FDA

S0 is the clause behind the third row, and it is worth reading in full rather than as a footnote: it covers “any pharmacological substance which is not addressed by any of the subsequent sections of the List and with no current approval by any governmental regulatory health authority for human therapeutic use”, with drugs in clinical development, discontinued drugs and veterinary-only drugs given as its own examples. Unnamed plus unapproved is the combination that puts a substance inside it, which is the entire difference between retatrutide and tirzepatide.

Two things this table is not saying. It is not a ranking: the list turns on authorisation, not on how a compound acts or how well it is studied, so an approved drug can be on it and an unlisted one can be off it. And a marketing authorisation is not an endorsement of the compound beyond the indication the application covers, or a statement about what a person should do. What these compounds do in a person is the evidence table above, and for most of them the trial literature is thinner than the marketing around them.

What the data do not show

Six things this evidence base does not contain, in order of how often they get skipped.

None of these trials tested a peptide bought for research. They tested pharmaceutical-grade product, administered under protocol, with a 20-week escalation period where one was specified. The results are evidence about the molecule. They are not evidence that a research pen reproduces the trial, and this page does not say it does.

The populations are not the people browsing. Mean BMI 38.0 in SURMOUNT-1, with 94.5% of participants at 30 or above (35658024). A BMI of 30 or 27-plus-a-comorbidity in STEP 1 (33567185). These are clinical cohorts selected on body composition and comorbidity.

Retatrutide’s headline is phase 2. −24.2% at 48 weeks in 338 people (37366315). There is no phase 3 and no long-term outcome data behind it. The number is the largest in the file and its evidence base is the thinnest, and those two facts belong in the same sentence.

Weight is a surrogate endpoint, not a health outcome. None of these four papers reports cardiovascular events, mortality, or any clinical outcome as a primary result. The evidence is about kilograms. That is a real and useful thing to know and it is not the same as knowing something about a person’s health.

What happens when the pen stops is not addressed by any of the four. No trial here reports sustained weight after withdrawal as a result. Treat it as unknown rather than as permanent.

And the discontinuation numbers are real numbers, printed in the same abstracts as the headline percentages: 4.5% against 0.8% for gastrointestinal events in STEP 1 (33567185), and 4.3% to 7.1% across active levels in SURMOUNT-1 (35658024). They deserve the same prominence as the percentages they sit next to.

Four molecules lead this protocol and a fifth is included on purpose. Tirzepatide carries the largest human dataset of anything on our shelf — 2 539 adults. Semaglutide carries the best-known name and the trial people have heard of. Retatrutide carries the strongest phase 2 numbers in the file and a card that says phase 2. Tesamorelin is the only approved molecule we sell and the only one with a visceral-fat endpoint, which is an honest frame for someone who does not want the GLP-1 comparison at all. AOD9604 is fifth because the query asks about it by name, and the card’s job there is to say what was actually measured.