Peptides for weight loss: which ones have human trial evidence, which were tested and failed, and which have never been given to a person

Peptides for weight loss: which ones have human trial evidence, which were tested and failed, and which have never been given to a person

The version of this page that stood here from March 2026 was called "Top 5 fat loss peptides that actually work". It ranked on page six, got one click in three months out of five thousand impressions, and I understand why: it read like every other page for the same search. Five categories, mechanisms for each, a comparison table with no numbers in it, and a line about dosing protocols in the excerpt. I wrote it, and it was not good enough. This version starts from a different question: which of these has ever been given to a person in a trial with a weight endpoint, and what happened? Asked that way, the list sorts itself. Three compounds have general-obesity approvals and tens of thousands of trial participants between them. Four more have finished phase 3. One failed a 502-patient trial nearly twenty years ago and is still sold on the strength of a paper that reports no efficacy data. And most of the rest, including several I sell, have never been tested for weight in a human at all. The numbers are below, with the trials they came from.

Michael Phelps

Founder & Peptide Research Specialist, PrymaLab

Research reference · First published March 4, 2026 · Rewritten from primary sources October 10, 2026 · ~40 min read

What are fat loss peptides, and why does the phrase cover two different things?

A "fat loss peptide" is any short chain of amino acids sold with a weight-loss claim. The phrase lumps together approved drugs with the largest trial programmes in modern pharmacology and research compounds with no human data at all, because both are peptides and both are sold online. The word "peptide" tells you about the chemistry. It tells you nothing about the evidence.

Semaglutide is a 31-residue analogue of the gut hormone GLP-1, and it has been given to tens of thousands of people in randomised trials, including 17,604 in a single cardiovascular outcomes study.^[1] AOD-9604 is a 16-residue analogue of the C-terminal fragment of human growth hormone, and the only peer-reviewed human publication about it reports safety and tolerability with no weight-loss results.^[2] MOTS-c is a 16-residue peptide encoded in mitochondrial DNA, and in its briefing document for a July 2026 advisory committee meeting the FDA could identify no "clinical studies or human exposure data for MOTS-c via any route of administration".^[3] All three are peptides. All three appear on lists of "best peptides for weight loss". One of them has an answer to the question of whether it works.

So this article sorts by evidence rather than by chemistry. Tier 1 is drugs approved for general obesity with multiple phase 3 trials and cardiovascular outcome data. Tier 2 is completed phase 3 weight-loss trials without US or EU approval. Tier 3 is phase 2 only. Tier 4 is compounds that were tested for weight and failed, or whose programmes were stopped. Tier 5 is compounds approved for something else and shown not to reduce weight. Tier 6 is compounds that have never been given to a person with a weight or fat-mass endpoint. The last tier is the longest, and it contains most of what is sold under this phrase.

How do peptides for weight loss work?

The compounds with evidence work on the gut-brain hormones that regulate appetite: GLP-1, GIP, glucagon and amylin. They slow gastric emptying and reduce food intake, and in trials people eat less and lose weight. The compounds without evidence are mostly claimed to work on growth hormone or mitochondrial pathways, and the claim has not been tested against a scale.

GLP-1 receptor agonists (liraglutide, semaglutide) mimic the incretin hormone released after meals; GIP/GLP-1 dual agonists (tirzepatide) add the second incretin receptor; GLP-1/glucagon dual agonists (survodutide, mazdutide) add a receptor that raises energy expenditure and liver fat oxidation; the triple agonist retatrutide hits all three; and amylin analogues (cagrilintide, the cagrilintide half of CagriSema) act on a separate satiety hormone from the pancreas.^{[4][5][6][7][8]} The mechanism is not the point of this article, because mechanism is where most "fat loss peptide" marketing lives and it is the part that cannot be checked by a reader. What can be checked is the trial. A compound with a plausible mechanism and no trial is a hypothesis. GHRP-2 has a plausible mechanism for fat loss through growth hormone release, and when seven lean men were infused with it before a buffet, they ate 35.9% more.^[9]

The growth-hormone-axis story deserves one more sentence because it carries most of the non-incretin market. The idea is that raising GH or IGF-1 shifts the body toward fat oxidation. The one well-designed body-composition trial of a GH secretagogue, the oral non-peptide MK-677 given to 65 older adults in a two-year trial with one-year primary endpoints, increased fat-free mass by 1.1 kg and found that "no significant differences were observed in abdominal visceral fat or total fat mass", with body weight up 2.7 kg against 0.8 kg on placebo and increased appetite as the commonest side effect.^[10] That is the best human result the GH-secretagogue mechanism has produced, and it is a weight gain.

Which peptides have the strongest human evidence for weight loss?

Tirzepatide, semaglutide and liraglutide, in that order of effect size, are the only peptides with approvals, multiple phase 3 trials and cardiovascular outcome trials. Retatrutide, CagriSema, mazdutide and survodutide have completed phase 3 with large effects but no US or EU approval. Everything else sold as a fat-loss peptide is phase 2 at best, and most of it has no human weight data.

Tier 1. Semaglutide: STEP 1 randomised 1,961 adults without diabetes two-to-one against placebo for 68 weeks and found −14.9% against −2.4%, a treatment difference of −12.4 percentage points, with 50.5% of semaglutide participants losing at least 15% of their weight against 4.9% on placebo.^[4] SELECT then randomised 17,604 people with cardiovascular disease and overweight, without diabetes, and found a 20% relative reduction in major adverse cardiovascular events (6.5% against 8.0%, hazard ratio 0.80).^[1] Tirzepatide: SURMOUNT-1 randomised 2,539 adults across three doses and placebo for 72 weeks and found −15.0%, −19.5% and −20.9% in the three dose arms against −3.1%, with 57% of the highest-dose group losing at least 20%.^[5] Liraglutide: SCALE randomised 3,731 adults for 56 weeks and found −8.4 against −2.8 kg, a difference of 5.6 kg, with 63.2% against 27.1% losing at least 5%.^[11]

Tier 2. Retatrutide: TRIUMPH-1 randomised 2,339 adults with obesity for 80 weeks and reported −17.6%, −23.7% and −25.0% in the three dose arms against −3.9%, published in September 2026; Lilly says it plans a US filing in the first quarter of 2027.^[7][12] CagriSema: REDEFINE 1 randomised 3,417 adults for 68 weeks and found −20.4% against −3.0%, a difference of 17.3 points; Novo Nordisk filed an NDA on 18 December 2025 and no decision had been announced by 9 October 2026.^[13][14] Mazdutide: GLORY-2 randomised 461 Chinese adults two-to-one to mazdutide or placebo for 60 weeks and found −16.65% against −1.50%; China's regulator approved it for weight management in June 2025.^[15][16] Survodutide: SYNCHRONIZE-1 randomised 725 adults for 76 weeks and found −12.2% and −13.0% in the two dose arms against −5.4%, a smaller margin than the others, with gastrointestinal events in 89.7% of the high-dose group.^[6]

Tier 3. Cagrilintide alone, the long-acting amylin analogue, produced 10.8% loss at its top dose against 3.0% on placebo over 26 weeks in a 706-person phase 2 trial, and beat liraglutide by 1.8 points; phase 3 began in November 2025.^[8][17] Maridebart cafraglutide, an antibody-peptide conjugate, produced 12.3% to 16.2% loss against 2.5% in a 592-person phase 2 trial.^[18] Amycretin's phase 1b/2a reported up to 24.3% at 36 weeks, but at a single site with heavy dropout, and its phase 3 programme only started in February 2026.^[19][20] These are real signals from real trials. They are not approvals, and the gap between a phase 2 effect and an approved label is where a lot of compounds have died.

Tiers 4 to 6 are the subject of their own section below, because the compounds in them are what most people searching this phrase will actually be offered.

What did the phase 3 trials measure?

Weight change against placebo over 56 to 80 weeks in adults with obesity, plus the proportions reaching 5%, 10%, 15% and 20% thresholds, discontinuation and adverse events. The table reports each trial's own figures and estimand by trial and agent, without dose arms, and nothing in it translates to anything outside those trials.

Two cautions on reading the table. First, estimands differ. REDEFINE 1's 20.4% is a treatment-policy figure that counts people who stopped the drug; Novo Nordisk's trial-product figure for the same trial is 22.7%.^[13][14] TRIUMPH-1's 25.0% is the treatment-regimen figure; Lilly's efficacy estimand is 28.3%.^[7][21] A page that quotes the larger number without saying which it is has made a choice for you. Second, the populations differ: GLORY-2 enrolled Chinese adults with a mean BMI of 34, SURMOUNT-1 a mean of 38, and the glucagon-containing agents were tested at different durations. The bars in Figure 2 are each trial against its own placebo, not against each other. The head-to-head trials are the next section.

How do the approved agents compare head to head?

In the only randomised comparison of the two market leaders at weight-management doses in people without diabetes, tirzepatide beat semaglutide 2.4 mg by 6.5 percentage points over 72 weeks. Semaglutide beat liraglutide by 9.4 points in STEP 8. CagriSema failed to show non-inferiority to tirzepatide in REDEFINE 4. No randomised trial has compared retatrutide with semaglutide or tirzepatide (its only active comparator so far is dulaglutide, in a phase 2 diabetes trial), so its apparent lead rests on cross-trial comparison.^[23]

SURMOUNT-5 randomised 751 adults with obesity and without diabetes to maximum-tolerated-dose tirzepatide or semaglutide for 72 weeks, open-label: least-squares mean change −20.2% against −13.7%, waist circumference −18.4 against −13.0 cm, both p<0.001.^[24] STEP 8 randomised 338 adults to semaglutide, liraglutide or matching placebo for 68 weeks (126 and 127 in the active arms), with the active comparison open-label: −15.8% against −6.4%, a difference of 9.4 points, with 55.6% against 12.0% losing at least 15%, and 27.6% of liraglutide participants discontinuing treatment against 13.5%.^[25] REDEFINE 4 compared CagriSema with tirzepatide over 84 weeks in 809 people and, by Novo Nordisk's own filing, "did not achieve its primary endpoint of demonstrating non-inferiority on weight loss", with 20.2% against 23.6% on the treatment-regimen estimand.^[26]

Three things follow. The order among the approved peptides is tirzepatide, then semaglutide, then liraglutide, and it comes from randomised trials rather than from comparing press releases. CagriSema sits between semaglutide and tirzepatide: it beat semaglutide 2.4 mg by 6.5 points in REDEFINE 5 (−18.4% against −11.9% over 68 weeks in 331 East Asian adults) and lost to tirzepatide in REDEFINE 4.^[27][26] And every obesity head-to-head trial so far has been open-label, which is a limitation worth knowing even if it is unlikely to reverse a 6.5-point difference.

What happens when treatment stops, and what happens to muscle?

Most of the weight comes back. One year after semaglutide was withdrawn in the STEP 1 extension, participants had regained two-thirds of their loss. After tirzepatide was switched to placebo in SURMOUNT-4, weight rose 14.0% in 52 weeks. About a quarter of the weight lost is lean mass, the same proportion as with placebo, and no trial has measured whether that matters for function.

The withdrawal data are consistent across agents. In the STEP 1 extension, 327 participants were followed for a year after stopping semaglutide and the lifestyle programme: loss at week 68 was 17.3%, regain by week 120 was 11.6 points, net loss 5.6%, and "cardiometabolic improvements seen from week 0 to week 68 with semaglutide reverted towards baseline".^[28] In STEP 4, after a 20-week run-in, 803 people were randomised to continue semaglutide or switch to placebo for 48 weeks: −7.9% against +6.9%, a 14.8-point gap.^[29] In SURMOUNT-4, after 36 weeks of tirzepatide (−20.9%), 670 people were randomised to continue or switch to placebo for 52 weeks: −5.5% against +14.0%, and 89.5% of those who continued kept at least 80% of their loss against 16.6% of those who stopped.^[30] A 2026 meta-analysis of 17 cessation studies put pooled regain at 7.2% of body weight: 7.19% for semaglutide, 4.83% for liraglutide, 13.04% for tirzepatide.^[31] Regain scales with how much was lost, which means the most potent agents carry the largest rebound, and none of the tier 2 agents has published withdrawal data yet.

On muscle, the SURMOUNT-1 DXA substudy of 160 participants found body weight −21.3%, fat mass −33.9% and lean mass −10.9% with tirzepatide at 72 weeks, against −5.3%, −8.2% and −2.6% on placebo; "of the body weight lost, approximately 75% was fat mass and 25% was lean mass for both tirzepatide and placebo".^[32] The STEP 1 DXA substudy, reported as a congress abstract rather than a paper, gave lean body mass −9.7% with semaglutide and a rising lean-to-fat ratio.^[33] A network meta-analysis of 22 trials estimated lean mass loss at 0.86 kg, "approximately 25% of the total weight loss".^[34] The 2026 Annals systematic review of 35 trials found the proportion of loss from muscle-related measures exceeded the 25% benchmark in about two-thirds of incretin trials and that "no study reported objective physical function outcomes".^[35] So in the SURMOUNT-1 substudy the lean fraction matched placebo, although Batsis's review found incretin arms above the 25% benchmark in about 65% of cases; the absolute lean loss is larger because the total loss is larger, and whether any of it changes what a person can do has not been measured.

What are the risks?

For the incretin class: gastrointestinal events in most participants, gallbladder disease at roughly double the placebo rate, a small pancreatitis signal, and the regain described above. For products bought outside a pharmacy, the one published assay found purity of 7.7% to 14.37% against a 99% claim and endotoxin in every vial. The non-incretin compounds have their own, mostly unmeasured, safety records.

The approved labels carry the pooled numbers. Wegovy: 6.8% of semaglutide 2.4 mg patients discontinued for adverse reactions against 3.2% on placebo; cholelithiasis 1.6% against 0.7%; cholecystitis 0.6% against 0.2%; adjudicated acute pancreatitis in 4 semaglutide patients against 1.^[36] Zepbound: gastrointestinal adverse events in 56% at every dose against 30%; cholecystitis 0.7% against 0.2%; acute pancreatitis 0.2% in both arms.^[37] Saxenda: gastrointestinal disorders in 68% against 39%; cholelithiasis 2.2% against 0.8%; acute pancreatitis 0.3% against 0.1%.^[38] A meta-analysis of 76 randomised trials put the relative risk of gallbladder or biliary disease at 1.37 overall and 2.29 in weight-loss trials.^[39] A 2025 meta-analysis of 21 cardiovascular outcome trials in 99,599 patients found gallbladder events up 26% and gastrointestinal events up 63%, with no difference in pancreatitis.^[40] The higher-potency agents are less tolerable: discontinuation for adverse events reached 11.3% at the highest retatrutide dose in TRIUMPH-1 by Lilly's figures, and dysaesthesia, an unusual skin-sensation side effect, affected 12.5% at that dose and 22.9% in the higher-dose semaglutide arm of STEP UP.^[21][41]

The risk that is specific to the market this page is read from is the product itself. Ashraf and colleagues bought semaglutide from online sellers without a prescription. Three of six orders never arrived. The three vials that did contained 28.56% to 38.69% more semaglutide than the label stated, had purity of 7.7% to 14.37% against a claimed 99%, and all three contained endotoxin at 2.1645 to 8.9511 EU/mg; the authors classed every vial as "probable substandard and falsified".^[42] In FAERS, reports involving compounded GLP-1 products carried adjusted reporting odds of 48.92 for preparation errors and 19.00 for contamination relative to the approved products.^[43] The FDA's own page, current to 1 October 2026, logs 990 adverse-event reports for compounded semaglutide and more than 730 for compounded tirzepatide, including dosing errors "some requiring hospitalization".^[44] None of that is a property of the molecule. All of it is a property of where it came from.

What about the compounds sold as fat-loss peptides outside the incretin class?

They sort into three groups: compounds that were tested for weight and failed or were stopped, compounds that were tested for something else and turned out weight-neutral, and compounds that have never been given to a person with a weight endpoint. The last group is the largest and includes most of what the phrase "fat loss peptides" is used to sell.

AOD-9604: the largest weight-loss programme, and the clearest failure

AOD-9604 is a 16-residue synthetic analogue of residues 177 to 191 of human growth hormone with an added N-terminal tyrosine, developed by Metabolic Pharmaceuticals in Melbourne on the premise that the lipolytic region of GH could be separated from its effects on growth and glucose.^[2][45] It is not the same molecule as hGH fragment 176-191, which WADA lists separately and for which no human trial of any kind exists.^[46] Between 2001 and 2006 the sponsor ran six randomised trials in 893 people, the two largest registered in Australia rather than on ClinicalTrials.gov.^[2]

What the record shows is best read from the FDA's 2024 review, because the efficacy data were never published in a journal. The only peer-reviewed human paper, Stier and colleagues 2013, reports safety and tolerability across the six trials and no weight results at all; its line that the profile was "indistinguishable from placebo" is a safety statement that gets quoted as if it were efficacy.^[2][22] The decisive trial, METAOD006, a phase 2b study, randomised 502 adults with BMI 30 to 45 to three oral doses or placebo for 24 weeks alongside a supervised diet and exercise programme, powered for an 80% chance of detecting a 1.8 kg placebo-subtracted difference. It did not find one. On 21 February 2007 the company announced that the results "do not support the commercial viability of the drug as a treatment for obesity. Development of the drug for this condition is terminated."^[22] FDA's summary: "In most of the studies we identified, AOD-9604 failed to show benefit for weight reduction when compared to placebo."^[22] The agency also found no human data for the injectable or transdermal routes that are actually sold, flagged bone-health signals in rats, possible liver toxicity in monkeys and equivocal genotoxicity, and found no US or foreign regulatory recognition of the compound; the "GRAS" status that vendors repeat traces to a food-ingredient expert panel described in Stier's paper, not to any FDA action.^[22][2] On 4 December 2024 the Pharmacy Compounding Advisory Committee voted 0 to 12 against adding it to the 503A list.^[47] The AOD-9604 reference covers the molecule in detail.

The growth hormone secretagogues

CJC-1295 DAC, a GHRH analogue that binds albumin, raised GH and IGF-1 for days in healthy volunteers.^[48] That and two further hormone-response studies in healthy adults are the whole of its published human pharmacology. Its phase 2 in HIV lipodystrophy was terminated in 2006 after a participant died of a myocardial infarction, and "the data from that study has not been published"; FDA counts 63 healthy adults as the total human exposure, most dosed once.^[49][50] The committee voted 0 to 13 against listing four of the five CJC-1295 substances nominated, and 1 to 12 against the fifth.^[47] Ipamorelin's only published randomised trial was in post-operative ileus after bowel resection, where it missed its endpoint (time to first tolerated meal 25.3 against 32.6 hours, p=0.15), with two deaths in the ipamorelin arm; a second 320-patient ileus trial was completed in 2014 and has never reported results; FDA found no human study that measured weight or body composition.^[51][52][53] GHRP-2, in the one human feeding study, increased food intake by 35.9%.^[9] GHRP-6 and hexarelin have no registered trials.^[54] The only "body composition" report for a GHRP combination is a chart review of 14 compliant men on testosterone therapy whose IGF-1 rose; nobody measured their weight or fat.^[55] Tesamorelin, the one currently approved GHRH analogue, reduced visceral fat by 15.4% against placebo in 806 patients with HIV lipodystrophy while leaving subcutaneous fat unchanged, and its 2025 label states that it "is not indicated for weight loss management as it has a weight-neutral effect".^[56][57] MK-677, the oral non-peptide secretagogue, is the contrast case already described: lean mass up, fat unchanged, weight up, and a hip-fracture trial stopped for a heart-failure signal.^[10][58]

MOTS-c, 5-amino-1MQ and Lipo-C

MOTS-c reduced diet-induced obesity in mice. In humans, the FDA wrote in its briefing document for the July 2026 compounding committee meeting, "the nomination did not include, and FDA has not identified, any clinical studies or human exposure data for MOTS-c via any route of administration", and it noted that MOTS-c is rapidly hydrolysed in human blood in vitro, which raises the question of whether it can reach an active concentration in the body at all.^[3] The compound sometimes cited as MOTS-c's human trial is CB4211, a modified analogue made by CohBar, whose phase 1b in 20 people with fatty liver found liver fat falling about equally on drug and placebo (−5.03% against −4.88%); CohBar then merged into an oncology company.^[59][60][61] The first registered trial of MOTS-c itself started in China in February 2026 with insulin sensitivity, not weight, as its endpoint.^[62] 5-amino-1MQ is a small-molecule inhibitor of the enzyme NNMT, not a peptide; it reduced fat mass in mice, has rat-only pharmacokinetics, and has zero registered human studies.^[63][64][54] Lipotropic injections (methionine, inositol, choline, B12, sometimes carnitine) are compounded preparations; a PubMed search for controlled human weight-loss trials of injectable lipotropics returned none.^[65]

Two non-peptides that get listed anyway

Tesofensine is a triple monoamine reuptake inhibitor, not a peptide. Its phase 2 produced weight loss 9.2 and 10.6 percentage points greater than diet and placebo in the two higher-dose arms over 24 weeks in 203 patients, in a paper that now carries a 2013 expression of concern.^[66] The Mexican phase 3 that is often cited was never published or registered, and on 6 November 2024 the licensor announced that the Mexican regulator had not approved it; the February 2023 event sometimes described as an approval was a committee opinion.^[67] It is not approved anywhere this review could find. The tesofensine reference has the rest. MK-677 is covered above.

Which weight-loss peptides are approved, and what is the status of the rest?

Liraglutide, semaglutide and tirzepatide are approved by the FDA and EMA for chronic weight management in general obesity. Retatrutide, survodutide, CagriSema, cagrilintide, amycretin and MariTide are investigational in the US and EU; mazdutide is approved in China. Every non-incretin compound above is unapproved for weight loss, and the FDA has stated in writing that "research use only" labels do not change what a product sold with weight-loss claims is.

The FDA's position on research-labelled GLP-1 products is now stated in three warning letters that use nearly identical language. To Summit Research Peptides, 10 December 2024: "Despite statements on your product labeling marketing your products as 'RESEARCH USE ONLY' and 'INTENDED AS A RESEARCH CHEMICAL ONLY,' evidence obtained from your websites establish that your products are intended to be drugs for human use."^[68] To Pink Pony Peptides, 31 March 2026, the same finding, with the claims "Reduce appetite and increase satiety" and "three times the fat-fighting horsepower" quoted back as evidence of intent.^[69] To Tex Peptides, 24 August 2026, the same again, adding that selling bacteriostatic water alongside the vials provides "the means to prepare an injectable drug for human administration".^[70] The agency's standing page states that "FDA has warned companies that have illegally sold unapproved drugs containing semaglutide, tirzepatide, retatrutide, survodutide or mazdutide that are falsely labeled 'for research purposes' or 'not for human consumption'".^[44] The 2026 compounding picture for the non-incretin peptides is different and more confusing: in April 2026 the FDA removed twelve peptides, including MOTS-c, from its Category 2 "significant safety risk" list, which the market has read as rehabilitation; the agency's position is that removal "does not make these substances eligible for 503A compounding", and that they remain unapproved new drugs when marketed for a therapeutic use.^[71][72]

On sport: no GLP-1 drug appears on the WADA 2026 Prohibited List, semaglutide and tirzepatide markers are in the 2026 Monitoring Program, and the 2027 List approved in September 2026 again leaves them permitted.^[73][74][75] Every growth-hormone-axis peptide on this page is prohibited at all times under S2.2.3, which names AOD-9604 and hGH 176-191 as examples of GH fragments, CJC-1295, sermorelin and tesamorelin as GHRH analogues, ipamorelin and ibutamoren as secretagogues, and hexarelin, GHRP-2 and GHRP-6 as GHRPs; MOTS-c is named under S4.4.1 as an AMPK activator.^[46] The efficacy question and the eligibility question are independent, and for an athlete the second one is settled.

Where PrymaLab fits

PrymaLab supplies research-grade peptides, including semaglutide, tirzepatide, retatrutide and several of the non-incretin compounds discussed above, for laboratory use only, through the metabolic peptide catalogue, alongside identity and purity testing listings for laboratories verifying the material they hold. None of it is Wegovy, Zepbound or Saxenda, and this page exists so that anyone comparing the compounds can see which of them has a human record and which does not, including the ones in the catalogue that do not.

One analytical point follows from the Ashraf data. The approved drugs are characterised to pharmaceutical standards: identity by mass spectrometry against a reference standard, purity by HPLC with impurities identified, endotoxin by LAL assay, content by validated assay. A vial from an unknown source has none of that behind it, and a "99%" on a label is a claim, not a measurement, until a certificate of analysis for that lot shows it. PrymaLab tests every lot in its own laboratory (RP-HPLC for purity, LC-MS for identity); check that the certificate matches the lot number on your vial and can be verified at https://prymalab.net/testing/verify/. Under PrymaLab's third-party testing guarantee, if any independent laboratory finds the wrong compound or less than the stated purity or amount, the order is refunded and the test reimbursed. The purity testing guide covers what HPLC, mass spectrometry and endotoxin testing can each establish, and what a certificate has to show before a purity figure means anything.

What this article does not settle

• Whether any tier 2 agent will be approved, or when. CagriSema's decision is pending; retatrutide's filing is a plan; survodutide has no announced filing date; REDEFINE 4 missed non-inferiority against tirzepatide.^[14][12][26]
• Whether retatrutide beats tirzepatide. No randomised comparison exists; the apparent 3- to 4-point lead is cross-trial.^[7][5]
• Whether the lean-mass loss matters. The proportion matches placebo; no trial has measured function.^[32][35]
• Regain after the tier 2 agents. No withdrawal data have been published for retatrutide, survodutide, mazdutide or CagriSema.^[31]
• AOD-9604's exact per-arm numbers. METAOD006 was never published; the figures above come from FDA's reconstruction of company filings.^[22]
• The CJC-1295 death. The attending physician attributed it to undiagnosed coronary disease; the data were never published and the registry and secondary reports disagree on enrolment (120 against 192).^[49][50]
• The July 2026 compounding committee votes. Secondary reports say six peptides were recommended for the 503A list; FDA minutes were not posted at the time of writing.^[72]

GLP-1, GIP, glucagon, amylin and the rest: what the names mean

GLP-1 (glucagon-like peptide-1) is a gut incretin hormone; GLP-1 receptor agonists are liraglutide, semaglutide and the non-peptide orforglipron. GIP is the other incretin; tirzepatide is a GIP/GLP-1 dual agonist. Glucagon receptor agonism is the second half of survodutide and mazdutide and one third of the triple agonist retatrutide. Amylin is a pancreatic satiety hormone; cagrilintide is its long-acting analogue, and CagriSema is cagrilintide plus semaglutide in one injection. Maridebart cafraglutide (MariTide) is an antibody carrying a GLP-1 agonist peptide and blocking GIP. Wegovy, Ozempic and Rybelsus are semaglutide; Zepbound and Mounjaro are tirzepatide; Saxenda and Victoza are liraglutide; Foundayo is orforglipron. Vendor codes such as "GLP-2" for tirzepatide and "GLP-3" for retatrutide are not scientific names; GLP-2 is a real and different gut hormone and there is no GLP-3. AOD-9604 is a 16-residue hGH 177-191 analogue and is not hGH fragment 176-191. Tesamorelin, sermorelin and CJC-1295 are GHRH analogues; ipamorelin, GHRP-2, GHRP-6 and hexarelin are growth hormone secretagogues; MK-677 is a non-peptide secretagogue. MOTS-c is a mitochondrial-derived peptide. 5-amino-1MQ and tesofensine are small molecules, not peptides, however they are listed.

Frequently asked questions

References