Research information only. PrecisePep is an educational library. It is not medical advice and is not intended to assist, encourage or enable self-administration. The compounds documented here are research chemicals — most are not approved for human use. Read the full disclaimer.
PrecisePep/Peptide Library/Retatrutide
Metabolic & Adipose Regulation
LY3437943 · Reta · "Triple-G" · GGG tri-agonist
An investigational once-weekly triple agonist at the GLP-1, GIP and glucagon receptors — currently the most weight-effective single molecule in published obesity trials, and still years from any approval.
Retatrutide is the first molecule to combine agonism at all three of the principal nutrient-sensing incretin receptors — GLP-1, GIP and the glucagon receptor — in a single once-weekly peptide. It sits one generation beyond tirzepatide (GLP-1/GIP dual) and two beyond semaglutide (GLP-1 alone), and the added glucagon arm is what makes it categorically different: it raises energy expenditure and drives hepatic fat oxidation rather than only suppressing intake.
In the 48-week phase 2 obesity trial published in The New England Journal of Medicine, the 12 mg arm produced a mean 24.2% reduction in body weight, and the weight curves had not plateaued when dosing stopped.[1] That is a magnitude of effect not previously seen from a pharmacological agent outside bariatric surgery. A separate phase 2a trial in MASLD reported near-total normalisation of liver fat in a large fraction of participants at higher doses.[2]
The TRIUMPH phase 3 programme has now reported, and it did not disappoint. Four trials — in obesity, in obesity with type 2 diabetes, in severe obesity with established cardiovascular disease, and in obesity with knee osteoarthritis — have read out across late 2025 and 2026, covering roughly 5,900 randomised participants in total. The headline figure is a mean 28.3% body weight reduction at 80 weeks on 12 mg, rising to 30.3% at 104 weeks in a higher-BMI extension cohort.[8] A biologics licence application is planned for the first quarter of 2027.
Phase 3 also did what phase 3 is for: it surfaced things phase 2 did not. A dysesthesia signal appeared at rates up to fourteen times placebo, discontinuation for adverse events reached 13.5% at the top dose in one trial, and the cardiovascular event numbers from TRIUMPH-3 — which was not designed as an outcome trial — do not establish benefit in the way the weight numbers establish efficacy. Those are on this page alongside the results.
None of it makes retatrutide an approved drug today. Everything circulating in the research-chemical market remains unapproved material of unverified provenance, used far outside any trial protocol — and now at doses and durations the trials have characterised, which makes the deviation easier to see rather than harder.
Retatrutide's glucagon-receptor arm increases resting energy expenditure. That is also why heart rate rises on it — a consistent, dose-dependent trial finding, not an anecdote. In an unmonitored setting there is no ECG, no lipase panel, no HbA1c trend and nobody watching lean mass. The trials that produced the impressive numbers had all of those.
GLP-1 receptor agonism slows gastric emptying, amplifies glucose-dependent insulin secretion, and acts on hypothalamic and hindbrain circuits to reduce appetite and the reward salience of food — the effect users describe as the disappearance of "food noise."
GIP receptor agonism contributes additional glucose-dependent insulinotropic signalling and improved postprandial lipid handling. Clinically the more interesting property is that GIP agonism appears to blunt the nausea and emetic signalling driven by pure GLP-1 agonism, which is part of why multi-agonists tolerate up-titration better than their potency alone predicts.
Glucagon receptor agonism is the differentiator. Hepatic GCGR activation increases fatty-acid oxidation and thermogenesis, raising total energy expenditure rather than only reducing intake. It is directly responsible for the profound liver-fat clearance seen in the MASLD trial. It is also the arm that carries the most metabolic risk: unopposed glucagon signalling raises hepatic glucose output, which is why the GLP-1 and GIP arms must be present to hold glycaemia in check.
The three activities are deliberately imbalanced within the molecule rather than equipotent, and the C20 fatty-diacid chain drives albumin binding, giving a half-life long enough for weekly dosing.
Effects that have been reported. A study-tier entry means a result was published — not that the effect is established, reproducible, or transferable from a rodent to a person.
Highest confidenceReported in peer-reviewed literature, registered clinical trials, or regulatory filings. Note the model: much peptide literature is in vitro or rodent work, not human trial data.
Phase 3, randomised, double-blind, placebo-controlled, 2,339 participants randomised 1:1:1:1 to 4, 9 or 12 mg or placebo, from a baseline mean weight of 248.5 lbs (BMI 40.0). Mean reductions were dose-ordered: 19.0% at 4 mg, 25.9% at 9 mg and 28.3% (70.3 lbs) at 12 mg, against 2.2% on placebo. 65.3% of the 12 mg arm reached a BMI below 30.
Key secondary endpoints at 80 weeks in TRIUMPH-1. At 12 mg: 62.5% achieved ≥25% weight loss, 45.3% achieved ≥30%, and 27.2% achieved ≥35% — against 2.2%, 0.5% and 0.3% on placebo. Waist circumference fell 9.5 inches at 12 mg. These are magnitudes previously associated with bariatric surgery rather than with pharmacology.
A pre-specified extension enrolled 532 participants with baseline BMI ≥35 who completed the main study, continuing to 104 weeks. Mean weight loss on 12 mg reached 30.3% (85.0 lbs) from a baseline of 268.3 lbs. Weight was still falling at two years, which addresses the plateau question phase 2 could not.
1,152 adults with type 2 diabetes and obesity or overweight, 80 weeks, 4/9/12 mg versus placebo. Weight fell 12.7%, 19.1% and 20.8% respectively against 4.0% on placebo, with HbA1c reductions up to 1.6 percentage points. Weight loss is consistently smaller in type 2 diabetes than in obesity alone — a well-described class pattern, not a failure of this agent.
1,949 adults with BMI ≥35 and established cardiovascular disease, randomised 1:1:2 to 9 or 12 mg or placebo over 80 weeks. Weight fell 21.6% and 22.6% against 3.2% on placebo, from a baseline of 245.6 lbs. This is the population where the risk-benefit calculation matters most and where the compound had not previously been tested.
445 participants with obesity and knee osteoarthritis randomised 1:1:1 to 9 or 12 mg or placebo over 68 weeks. WOMAC pain scores fell 4.5 points (−75.8%) on 9 mg and 4.4 points on 12 mg from a baseline of 6.0, against 2.4 points (−40.3%) on placebo. Roughly one in eight treated participants reported being completely free of knee pain at week 68, against just over 4% on placebo. Weight loss reached 28.7% at 12 mg.
TRIUMPH-1 reported significant improvements in waist circumference, non-HDL cholesterol, triglycerides, systolic blood pressure and high-sensitivity C-reactive protein — the expected downstream consequences of weight loss at this magnitude, now measured in a phase 3 population.
The trial that made this compound famous, retained here for continuity. Randomised placebo-controlled, dose-ordered across 1, 4, 8 and 12 mg, with the weight curve still falling at week 48. Phase 3 subsequently confirmed and exceeded it at 80 and 104 weeks.
A randomised phase 2a trial in metabolic dysfunction-associated steatotic liver disease reported large relative reductions in liver fat content by MRI-PDFF, with a substantial proportion of participants reaching normal liver fat at higher doses.
Phase 2 work in T2D reported clinically meaningful HbA1c reductions alongside weight loss, consistent with combined incretin agonism.
Trial data reported reductions in systolic blood pressure, triglycerides and non-HDL cholesterol — expected downstream consequences of the degree of weight and visceral-fat loss achieved.
The glucagon-receptor component distinguishes retatrutide from GLP-1-only agents by increasing resting energy expenditure — the mechanistic basis for the greater weight effect at comparable appetite suppression.
Unproven — inference onlyExtrapolated from a known mechanism, receptor profile, or adjacent compound class. Biologically plausible but not directly demonstrated for this compound in this context.
TRIUMPH-4 demonstrated the pain reduction but was not designed to separate mechanical unloading from any systemic anti-inflammatory effect of fat loss, or from a direct joint action. The placebo arm also improved substantially (−40.3%), which is characteristic of osteoarthritis pain trials and is worth holding in view alongside the treatment figure.
GLP-1 receptor signalling modulates mesolimbic dopamine reward. The literature for this effect is strongest for semaglutide (alcohol, nicotine, compulsive behaviours); extrapolation to retatrutide is mechanistically reasonable and undemonstrated for this molecule.
Incretin receptors are expressed in the CNS, and the GLP-1 class has an active neurodegeneration literature. No retatrutide-specific human neuro outcome data exists. Treat as a class hypothesis only.
Anticipated by analogy to the semaglutide and tirzepatide outcome trials and by the magnitude of risk-factor improvement. A retatrutide cardiovascular outcomes trial is required before this can be claimed; blood-pressure and lipid changes are surrogates, not events.
Retatrutide reduces hyperinsulinaemia, which is the driver of PCOS, so the mechanism plainly applies. The randomised PCOS evidence belongs to approved GLP-1 receptor agonists — liraglutide, exenatide, increasingly semaglutide — not to an unapproved triple agonist that adds a glucagon arm and has no PCOS study of any kind.
Acts independently of thyroid status. Sequencing is the caveat — hypothyroid weight gain is modest and largely resolves with adequate replacement, so confirming TSH is in range comes first.
Follows plausibly from the degree of visceral and total fat loss, by analogy with tirzepatide OSA trial data. Not established for retatrutide.
Uncontrolled reportsAggregated from research-community logs, forums, vendor literature and self-reported experience. Uncontrolled, unverified, subject to placebo, polypharmacy and product-purity confounds. Lowest evidentiary weight.
The most consistent community report is that 0.5–2 mg weekly — well below trial doses — produces near-complete loss of intrusive food thoughts. This is the basis of the widespread microdosing practice and it is entirely uncontrolled data.
A resting-HR rise of roughly 5–7 bpm is reported so consistently that the community treats it as confirmation the glucagon arm is active. Framing a cardiovascular effect as a desirable marker is exactly the kind of reasoning uncontrolled use produces — the trials recorded the same signal as an adverse finding.
Widely reported and attributed to increased thermogenesis via GCGR. Plausible, unquantified.
Frequently reported at low doses before meaningful weight change, which users interpret as a direct anti-inflammatory action. Confounded by placebo, by early water-weight change, and by concurrent compounds in almost every reported stack.
Self-ordered fasting insulin, HOMA-IR and HbA1c improvements are commonly posted. Directionally consistent with trial data; individually unverifiable.
Adverse effects, contraindications and unknowns. The grey-market tier is where under-reporting is worst: uncontrolled use produces no systematic safety surveillance at all, so absence of a reported harm is not evidence of safety.
Highest confidenceReported in peer-reviewed literature, registered clinical trials, or regulatory filings. Note the model: much peptide literature is in vitro or rodent work, not human trial data.
Abnormal or unpleasant sensation, most commonly described as tingling, burning or altered skin sensation. Reported in 5.1–12.5% of retatrutide participants against 0.9% on placebo in TRIUMPH-1, with similar patterns in TRIUMPH-2 (4.5–7.3% vs 0.7%) and TRIUMPH-3 (6.4% vs 1.3%). Described as generally mild to moderate, with most participants continuing treatment. This is the clearest example of why phase 3 exists: at up to fourteen times the placebo rate it is unambiguous, and it was not a headline finding of the phase 2 programme.
In TRIUMPH-1, against placebo: nausea 28.6–42.4% (vs 14.8%), diarrhoea 25.2–34.1% (vs 13.5%), constipation 23.8–26.1% (vs 10.9%), vomiting 10.6–25.3% (vs 4.8%). Dose-dependent and concentrated in the titration period. These remain the most frequent adverse events and the main driver of discontinuation.
Dose-ordered across the programme. TRIUMPH-1: 4.1% at 4 mg, 6.9% at 9 mg, 11.3% at 12 mg, against 4.9% on placebo. TRIUMPH-3: 9.8% at 9 mg and 13.5% at 12 mg against 4.8%. Roughly one in seven people at the highest dose in the cardiovascular-disease population stopped because of side effects — which is the number to weigh against the 28% figure, not instead of it.
Reported at 7.5–8.8% against 5.3% on placebo in TRIUMPH-1, with comparable patterns in the other trials. A modest but consistent excess across the programme.
In the cardiovascular-disease population, MACE-5 events numbered 44 on retatrutide against 52 on placebo (hazard ratio 0.82), while MACE-3 events numbered 27 against 23 (hazard ratio 1.12). TRIUMPH-3 was not designed or powered as a cardiovascular outcome trial, the event counts are small, and the two composites point in opposite directions. This does not demonstrate cardiovascular benefit, and it does not demonstrate harm. It is the absence of an answer, in a class where semaglutide, liraglutide and dulaglutide all have one.
Mean resting heart rate rose in a dose-ordered fashion, peaking around the mid-point of the trial before partially attenuating. Attributed to glucagon-receptor-mediated increases in energy expenditure and sympathetic tone.
Glucagon agonism raises hepatic glucose output. Trials monitored for this; it is the reason the incretin arms must remain in balance and the reason unsupervised dosing in anyone with impaired glucose tolerance is a genuine hazard.
A property of all rapid-weight-loss pharmacology, not unique to retatrutide. Body-composition sub-studies across the incretin class consistently show a meaningful fraction of total loss coming from fat-free mass.
Approved agents in this class carry warnings for pancreatitis, gallbladder disease (cholelithiasis, cholecystitis), acute kidney injury secondary to dehydration, diabetic retinopathy complications with rapid glycaemic change, and — from rodent data — a boxed contraindication in medullary thyroid carcinoma and MEN2. These apply as class considerations pending retatrutide-specific characterisation.
Local erythema and nodules reported at typical rates for subcutaneous peptide therapeutics.
Unproven — inference onlyExtrapolated from a known mechanism, receptor profile, or adjacent compound class. Biologically plausible but not directly demonstrated for this compound in this context.
If lean-mass loss is proportionally sustained over multi-year use without resistance training and elevated protein intake, the theoretical endpoint is clinically relevant sarcopenia — particularly in older users. Not yet demonstrated over long follow-up because that follow-up does not exist.
Persistent delayed gastric emptying has been an emerging concern across the GLP-1 class, including anaesthesia aspiration risk from retained gastric contents. Anyone with a surgical procedure planned should assume this applies.
Rapid, large weight reduction reduces bone mineral density through mechanical unloading. Untested for retatrutide specifically.
Regain after withdrawal is well characterised across the incretin class. There is no reason to expect retatrutide behaves differently, and the larger the loss the larger the potential rebound.
Improving insulin sensitivity and losing weight restores ovulation in a substantial proportion of people with PCOS. Approved agents in this class are contraindicated in pregnancy with defined washouts; retatrutide has no label, no pregnancy safety data and a ~6-day half-life, so exposure persists for weeks after the last dose.
Delayed gastric emptying alters absorption of a narrow-therapeutic-index, fasting-dependent drug, and direction of change is not predictable. Compounded here by the glucagon-driven 5–7 bpm heart rate rise, which overlaps precisely with the symptoms of thyroid over-replacement.
Delayed gastric emptying alters the absorption kinetics of orally administered drugs. In community stacks combining five to fifteen compounds, no interaction data exists at all.
Uncontrolled reportsAggregated from research-community logs, forums, vendor literature and self-reported experience. Uncontrolled, unverified, subject to placebo, polypharmacy and product-purity confounds. Lowest evidentiary weight.
Reported far more prominently in community logs than trial adverse-event tables suggest, likely because trial reporting categories aggregate them. Frequently the reason for dose reduction.
Attributed variously to caloric deficit, blunted dopaminergic reward signalling, and inadequate electrolyte or protein intake. Common enough that it is a routine topic in community protocol discussion.
Frequently reported at 3–5 months. Consistent with rapid weight loss and caloric restriction generally rather than a specific drug toxicity.
Reported alongside the documented heart-rate rise. Community response is usually to reduce dose; nobody in that setting is being screened for arrhythmia.
Cholecystectomy after rapid loss appears periodically in community reports, matching the known class association between rapid weight loss and gallstone formation.
Under-dosed, mislabelled and non-sterile material is a recurring theme in community third-party testing threads. Two vials labelled identically are not necessarily the same product.
Figures below are documented, not recommended. Study ranges come from published protocols in the stated model; grey-market ranges are what the research community reports using and carry no safety validation of any kind.
Highest confidenceReported in peer-reviewed literature, registered clinical trials, or regulatory filings. Note the model: much peptide literature is in vitro or rodent work, not human trial data.
Across the TRIUMPH programme, participants started at 2 mg once weekly and escalated stepwise every four weeks to their assigned target: one step to 4 mg, three steps to 9 mg, four steps to 12 mg. This is the first properly characterised escalation schedule for this compound, replacing inference from phase 2. It is slower than most community protocols, and the adverse-event and discontinuation figures on this page were produced with it rather than despite it.
The phase 3 programme settled on three maintenance doses. Weight loss is dose-ordered across all of them, and so are adverse events and discontinuation — 11.3% discontinued for adverse events at 12 mg in TRIUMPH-1 against 4.1% at 4 mg. The 4 mg dose still produced 19.0% weight loss, which exceeds what approved semaglutide achieves in its obesity trials, at roughly a third of the discontinuation rate of the top dose. More is not automatically the right answer.
The trial used slow, fixed, protocol-driven escalation over months to reach maintenance doses of 1, 4, 8 or 12 mg once weekly, with escalation steps designed specifically to manage gastrointestinal tolerability rather than to chase weight loss.
The ~6-day half-life supports fixed weekly subcutaneous dosing; trials used a consistent weekly day with subcutaneous abdominal, thigh or upper-arm administration.
Unproven — inference onlyExtrapolated from a known mechanism, receptor profile, or adjacent compound class. Biologically plausible but not directly demonstrated for this compound in this context.
Because adverse events are dose-ordered while appetite suppression saturates earlier than weight loss does, the mechanistically defensible position is that the smallest dose producing appetite control is the one with the best risk profile — even though it is not the one that produced the headline trial numbers.
Community and some clinical practice extend titration intervals for women on the basis of lower average body mass and reported greater GI sensitivity. There is no trial evidence supporting a sex-specific schedule.
Escalating only when weight loss stalls — rather than on a fixed calendar — is a widely-proposed approach. It is a reasonable inference from dose-ordered adverse events but was not how the trials were run, and it changes the exposure profile the safety data was generated under.
Uncontrolled reportsAggregated from research-community logs, forums, vendor literature and self-reported experience. Uncontrolled, unverified, subject to placebo, polypharmacy and product-purity confounds. Lowest evidentiary weight.
The dominant grey-market practice sits far below trial doses. Reported rationale: near-complete appetite suppression with minimal GI burden. No safety or efficacy data exists at these doses for this molecule.
The schedule reproduced in the Gold Standard protocol: begin at 1 mg weekly, raise by 0.5 mg only after a genuine ≥2-week plateau (4 weeks for women) and only in the absence of side effects.
Some users halve the weekly dose across two injections to flatten peak concentration and reduce nausea. Pharmacokinetically coherent given the long half-life; entirely undocumented.
Continued low-dose administration to defend against regain is common practice in the community. It is also the scenario with the longest cumulative exposure and the least data.
| Community tier | Weekly dose | Reported intent |
|---|---|---|
| Micro | 0.5 – 1 mg | Appetite/food-noise control, minimal GI burden |
| Low | 1 – 2 mg | Steady loss, tolerability-first |
| Moderate | 2 – 4 mg | Accelerated loss; GI effects commonly reported |
| High (trial-equivalent) | 8 – 12 mg | Rarely used outside trials; heaviest AE burden |
The study table is the only one attached to monitored safety data. The community table describes what people report doing — at doses the trials never characterised, with material the trials never used. The gap between them is the single most important thing on this page.
Combination effects. Almost no peptide combination has been studied as a combination in humans; stack logic is overwhelmingly mechanistic inference and community practice. Each linked compound has its own page.
Highest confidenceReported in peer-reviewed literature, registered clinical trials, or regulatory filings. Note the model: much peptide literature is in vitro or rodent work, not human trial data.
The combination of a long-acting amylin analogue with an incretin agonist has been formally studied as cagrilintide + semaglutide, showing additive weight effect with tolerable safety. This is the trial-grade evidence that the amylin + incretin concept works — it is not evidence about retatrutide specifically.
Unproven — inference onlyExtrapolated from a known mechanism, receptor profile, or adjacent compound class. Biologically plausible but not directly demonstrated for this compound in this context.
Amylin agonism acts on hindbrain satiety circuits distinct from incretin pathways, so the pairing is expected to be additive on satiety while allowing a lower retatrutide dose — and therefore fewer dose-driven adverse events. Mechanistically sound, never trialled as this pair.
The stated rationale is preservation of lean mass during aggressive fat loss. GH axis stimulation is nitrogen-sparing in principle; whether it meaningfully offsets incretin-driven lean-mass loss in humans has not been tested.
Community logic pairs aggressive fat loss with BPC-157/TB-500 blends for connective-tissue support and gut tolerance. No interaction data; the gut-healing rationale for BPC-157 against GLP-1-driven GI effects is entirely mechanistic speculation.
GH is counter-regulatory to insulin. Stacking a strong GH secretagogue with a glucagon-receptor agonist plausibly pushes hepatic glucose output in the same direction. This is a reason for caution that community protocols rarely mention.
Uncontrolled reportsAggregated from research-community logs, forums, vendor literature and self-reported experience. Uncontrolled, unverified, subject to placebo, polypharmacy and product-purity confounds. Lowest evidentiary weight.
The most widely circulated metabolic pairing. Reported rationale is that cagrilintide lets retatrutide stay at a lower dose, cutting side effects while maintaining loss.
The most defensible thing in the community playbook, and the only "synergy" here with a solid independent evidence base — resistance training and adequate protein preserve lean mass during any caloric deficit.
Reported as targeting visceral adipose from two directions. Both compounds independently reduce visceral fat; the combination is untested.
Routine community practice for managing constipation and fatigue. Low-risk supportive measures rather than pharmacological synergy.
Whole schedules — dosing, titration, cycle length, and off-time — as documented in trials, as proposed from mechanism, and as circulated in the research community. Reproduced for study and comparison only.
Highest confidenceReported in peer-reviewed literature, registered clinical trials, or regulatory filings. Note the model: much peptide literature is in vitro or rodent work, not human trial data.
The registered phase 3 programme spans obesity, obesity with type 2 diabetes, cardiovascular risk and knee osteoarthritis. Protocol documents, eligibility criteria, exclusions and endpoints are public on ClinicalTrials.gov and are the only place a genuinely complete retatrutide protocol exists.
Fixed-interval escalation with defined step sizes and stopping rules, run against protocol-mandated monitoring: heart rate, glycaemia, lipase/amylase, gallbladder symptoms and body composition.
Unproven — inference onlyExtrapolated from a known mechanism, receptor profile, or adjacent compound class. Biologically plausible but not directly demonstrated for this compound in this context.
Escalate only on genuine plateau and only in the absence of adverse effects; hold at the lowest dose that maintains appetite control. A coherent risk-minimising inference from the dose-ordered adverse-event data, not a validated protocol.
Given documented lean-mass loss, the mechanistically obvious accompaniment to any incretin protocol is elevated protein intake and progressive resistance training. This part is well supported by independent literature even though the peptide protocol around it is not.
Baseline and periodic: resting heart rate and blood pressure, HbA1c and fasting glucose, lipase, liver panel, lipids, and body composition. This is what the trials measured; it is the gap between trial use and unmonitored use.
Uncontrolled reportsAggregated from research-community logs, forums, vendor literature and self-reported experience. Uncontrolled, unverified, subject to placebo, polypharmacy and product-purity confounds. Lowest evidentiary weight.
1 mg retatrutide weekly with 0.5 mg cagrilintide weekly; increase retatrutide by 0.5 mg every 4 weeks only if weight loss has plateaued for ≥2 weeks (≥4 weeks for women) and only if there are zero side effects. Reproduced here as circulated.
Places retatrutide/cagrilintide titration in the opening quarter of a year-long phased matrix, before regenerative and neuro-cognitive compounds are layered in, on the stated principle of never introducing multiple modalities simultaneously.
After reaching goal weight, continuing 0.5–1 mg weekly to defend against regain. Common practice, longest cumulative exposure, zero data.
Both conditions bring people to the incretin class, and the trial evidence in PCOS is real — but it belongs to the approved agents, not to this one. That distinction is the whole of this section.
Retatrutide is investigational. It is not approved for anything, anywhere — not for obesity, not for diabetes, and certainly not for PCOS or hypothyroidism. The randomised PCOS evidence that exists is for approved GLP-1 receptor agonists, and it does not transfer to an unapproved triple agonist with no PCOS data of its own.
The fertility and contraception item below applies regardless, and matters more here because nobody is supervising it.
Pathophysiology
Nicotine dependence is maintained by mesolimbic dopamine signalling. Quitting also produces an average weight gain of several kilograms, and fear of that gain is a measured barrier to attempting cessation and a measured cause of relapse.
Mechanistic rationale
The randomised evidence in this class is two small trials. Dulaglutide alongside varenicline did not improve abstinence — 63% against 65% — but prevented post-cessation weight gain, −4.6 kg against +0.5 kg. Exenatide alongside a nicotine patch did improve abstinence, 46.3% against 26.8%. Semaglutide has a large observational signal whose own investigators said it does not justify off-label use. Nothing has been tested for this agent, and in every trial the GLP-1 was added to a treatment that already works rather than replacing it.
Community reports
The consistent finding across the randomised literature is the weight, not the quitting. Post-cessation weight gain is real, specific and worth addressing; improved abstinence is not established, and the whole randomised evidence base is roughly 410 participants.
Components carrying the argument: GLP-1 receptor — class inference only
Pathophysiology
Type 2 diabetes is insulin resistance plus beta-cell failure. The glucagon receptor arm of this molecule is the one that made its use here interesting and the one that made it uncertain.
Mechanistic rationale
TRIUMPH-2 settled the question phase 2 could only gesture at. In 1,152 adults with type 2 diabetes and obesity or overweight over 80 weeks, retatrutide produced HbA1c reductions of up to 1.6 percentage points alongside weight loss of 12.7%, 19.1% and 20.8% at 4, 9 and 12 mg. The theoretical concern was that glucagon receptor agonism raises hepatic glucose output — exactly the wrong direction in diabetes. At phase 3 scale and duration the GLP-1 and GIP arms evidently compensate. It is still not approved, with a licence application planned for the first quarter of 2027.
Community reports
Retatrutide is the most-discussed grey-market compound in this space and people with type 2 diabetes are using it. There is now a characterised titration schedule and a characterised adverse-event profile to deviate from, which is an improvement on deviating from nothing — but there is still no approved product, no label, and no cardiovascular outcome trial, in a condition where two approved alternatives have one.
Components carrying the argument: GLP-1 + GIP + glucagon triple agonism
Pathophysiology
Insulin resistance drives hyperinsulinaemia, which stimulates ovarian androgen production and suppresses SHBG, disrupting ovulation. Weight gain worsens insulin resistance, closing the loop.
Mechanistic rationale
Retatrutide reduces hyperinsulinaemia and produces the largest weight reductions published for any single agent, so the mechanism plainly applies. What does not transfer is the evidence. The randomised PCOS trials are of liraglutide, exenatide and increasingly semaglutide — approved drugs with defined labels. Retatrutide adds a glucagon-receptor arm those agents do not have, has no PCOS study of any kind, and is not approved for any indication. Anyone reasoning "GLP-1s work in PCOS, so this will" is skipping the step where that gets demonstrated.
Community reports
Community reports mirror the approved agents — weight loss, returning cycles, improved skin. Uncontrolled, and in a population using an unapproved compound without monitoring.
Components carrying the argument: GLP-1 arm mainly; the glucagon arm is the unstudied variable
Pathophysiology
Restoring insulin sensitivity and losing 5%—10% of body weight restores ovulation in a substantial proportion of people with PCOS — often in people who had assumed conception was difficult.
Mechanistic rationale
The approved agents in this class are contraindicated in pregnancy and require discontinuation before conception. Retatrutide has no label to state a washout, no pregnancy safety data, and no prescriber tracking who is taking it. Its half-life is roughly six days, so meaningful exposure persists for weeks after the last dose. An unplanned pregnancy on an investigational triple agonist is the scenario this section exists to flag.
Community reports
Rarely discussed in community protocols, which focus almost entirely on weight and dose titration.
Components carrying the argument: Applies to any effective incretin agonist
Pathophysiology
Low thyroid hormone lowers basal metabolic rate and promotes fluid retention. Weight gain attributable to hypothyroidism itself is modest and largely resolves with adequate replacement.
Mechanistic rationale
The mechanism does not depend on thyroid status. The interaction does: delayed gastric emptying alters levothyroxine absorption, and levothyroxine is narrow-therapeutic-index, fasting-dependent and absorption-sensitive. Retatrutide also raises resting heart rate by 5—7 bpm via the glucagon arm — a symptom that overlaps precisely with over-replacement, making the two difficult to tell apart without a blood test.
Community reports
Elevated resting heart rate is reported so consistently on retatrutide that the community treats it as confirmation the drug is working. In a replaced hypothyroid person that same finding has a second possible cause.
Components carrying the argument: Gastric emptying + the glucagon-driven heart rate effect
| Question | Position |
|---|---|
| Approved for type 2 diabetes? | Not yet — BLA planned Q1 2027 |
| Phase 3 HbA1c reduction | Up to 1.6 points — TRIUMPH-2, 80 weeks |
| Glucagon arm | Raises hepatic glucose output — net effect favourable through phase 3 |
| Approved for anything? | No — investigational, phase 3 |
| Does the PCOS trial evidence apply? | No — that evidence is for approved agents |
| Does the mechanism apply? | Yes — hyperinsulinaemia is the shared target |
| Pregnancy washout guidance? | None — no label exists; half-life ~6 days |
| Thyroid interaction | Gastric emptying vs levothyroxine |
| Confounder unique to retatrutide | Glucagon arm raises heart rate 5—7 bpm |
| Better-evidenced alternative | An approved GLP-1 RA, with a prescriber |
PCOS has treatments that work. Metformin is first-line for the metabolic phenotype; combined hormonal contraceptives address hyperandrogenism and cycle regulation; letrozole is first-line for ovulation induction; inositol has reasonable evidence and an excellent safety profile; approved GLP-1 receptor agonists now have randomised evidence in this population. Weight loss of 5–10% alone restores ovulation in a meaningful proportion.
Levothyroxine is inexpensive, once daily, titrated against a reliable blood test, and restores euthyroid status in most people. Nothing here treats the thyroid. The unglamorous first move in someone still gaining weight is confirming replacement is adequate and correctly taken — fasting, away from calcium, iron, coffee and proton pump inhibitors.
The comparison worth making here is not retatrutide versus nothing. It is retatrutide versus semaglutide or tirzepatide — approved agents, obtainable lawfully, with randomised PCOS evidence, pregnancy guidance, defined contraindications and a clinician accountable for monitoring. Retatrutide may eventually surpass them. It has not been shown to in this population, and it currently offers none of that infrastructure.
Physical-chemistry reference for laboratory handling of the lyophilised material. Reproduced so that stability and concentration mathematics can be studied — see the calculator for the arithmetic and the reconstitution guide for sterile technique.
Grey-market retatrutide is supplied lyophilised, commonly in 5, 10, 15, 20 or 30 mg vials. Reconstituted with bacteriostatic water; because a single vial covers many weeks of a weekly dose, multi-draw preservative-containing diluent is the norm. Direct the stream down the vial wall, swirl gently, never shake — agitation shears peptide chains.
Lyophilised: −20 °C long-term, protected from light. Reconstituted: 2–8 °C. Community reports of 4–8 weeks refrigerated stability are unverified for this molecule; approved analogues in the class carry defined in-use periods measured in weeks.
Most commonly 5 mg, 10 mg, 15 mg and 30 mg lyophilised vials.
Cloudiness, particulate or discolouration indicates degradation or contamination. A weekly-dose compound in a 10 mg vial can sit reconstituted for two to three months at typical microdoses — considerably beyond any characterised in-use stability window.
Regulatory status changes and differs by country. This is a summary for orientation, not legal advice — check your own jurisdiction.
Investigational. Not approved in any jurisdiction. The TRIUMPH phase 3 programme has reported across late 2025 and 2026 — TRIUMPH-1 in obesity, TRIUMPH-2 in obesity with type 2 diabetes, TRIUMPH-3 in severe obesity with established cardiovascular disease, and TRIUMPH-4 in obesity with knee osteoarthritis — and Eli Lilly has stated an intention to submit a biologics licence application to the FDA in the first quarter of 2027.
A completed phase 3 programme is not an approval. Review takes the better part of a year after submission and examines the full dataset rather than the press releases. Until then there is no approved product, no label, no approved dose and no legitimate supply. Every vial in circulation is unapproved material of unverified provenance.
Note also what the programme did not deliver: a cardiovascular outcome trial. TRIUMPH-3 enrolled a cardiovascular-disease population but was not powered for cardiovascular endpoints, and its two MACE composites point in opposite directions. Semaglutide, liraglutide and dulaglutide all carry that evidence. This does not.
Regulatory and trial claims re-checked against primary sources on 16 August 2026. Checked: TRIUMPH-1 to -4 phase 3 readouts and BLA timing. Approvals, trial readouts and compounding decisions move faster than anything else on this page — a date here means someone looked, not that nothing has changed since.
Peptide hormones and metabolic modulators are addressed under the WADA Prohibited List. Any athlete subject to testing should treat an unapproved investigational metabolic agent as prohibited until they have specifically verified otherwise.
Links resolve to PubMed records, trial registries, publisher pages or primary documents. Where a body of work rather than a single paper is cited, the link opens a PubMed query so the full result set — including newer papers than this page — can be reviewed. Verify every claim against the primary source before relying on it.
For educational and research reference only. Nothing on this site is medical advice, a recommendation, or an instruction to administer any substance to a human being. Every figure on this page is a record of what has been reported, not a recommendation. Full disclaimer.
