The most striking finding in obesity pharmacology, and an unapproved drug still in phase 3. Both of those are true at the same time.
Retatrutide, developed by Eli Lilly under the code LY3437943, is the compound that broke the pattern in obesity pharmacology. Where earlier drugs act on one hormone receptor or two, this one acts on three at once.
It is not an approved medicine. It is an investigational drug in phase 3 trials, and everything below comes from studies where participants were screened, supervised, and monitored throughout. That context is part of the data, not a footnote to it.
An agonist is a molecule that switches a receptor on. Retatrutide switches on three:
The progression across those three is the clearest way to see what each addition buys. In their respective trials, single-receptor semaglutide produced roughly 15% average weight loss, dual-receptor tirzepatide roughly 20%, and retatrutide 24.2% at 48 weeks.
GLP-1 is released in the gut after eating. It does four things. It prompts insulin release when blood sugar rises. It slows how fast the stomach empties. It cuts the liver's sugar output. And it signals fullness through the hypothalamus, the part of the brain that sets the background drive to seek food. That last one is why these drugs reduce appetite.
GIP is also released in the gut after eating. It supports insulin release alongside GLP-1. It also shapes how nutrients are shared out — how much of what you eat gets taken up and used by tissue, rather than left in the blood. Retatrutide is about nine times more potent here than the body's own GIP.
That may matter for what gets lost. In tirzepatide trials, about 25% of the weight lost was lean mass. In semaglutide trials it was 39 to 45%. Some of that gap is put down to GIP activity.
Glucagon is the one that sets retatrutide apart. It works in the opposite direction to insulin, prompting the liver to put stored sugar back into circulation and to break fat down. Switching this receptor on has three effects. The liver burns more fat. The body uses more energy at rest. And stored fat releases fatty acids to be used as fuel.
So the three together reduce intake, steady blood sugar, and raise expenditure. Earlier drugs mostly did the first.
Here is the counterintuitive part, and it is the most interesting thing about the molecule's design.
Retatrutide is the weakest of the three at suppressing appetite. That was deliberate. One molecule cannot be strongest at three receptors at once, so something had to give. The designers settled for about 40% of semaglutide's GLP-1 strength. In exchange they got roughly nine times natural potency at GIP, plus glucagon activity that neither rival has.
So hunger is more noticeable on retatrutide than on semaglutide, while the measured weight loss is higher. Appetite is something you feel. Fat burning in the liver is not. Judging this compound by hunger measures the weakest of its three mechanisms.
Each receptor has its own sensitivity, described by a value called EC50 — the concentration producing half the maximum possible response. For retatrutide, GIP is the most sensitive, GLP-1 sits in the middle, and glucagon is the least sensitive.
That means glucagon activity is recruited gradually as concentration rises, along a smooth S-shaped curve, rather than switching on at a threshold. The liver-fat substudy tracked beta-hydroxybutyrate, a marker the liver makes when it is actively burning fat. It rose to two or three times baseline in the 4 mg arm and above, alongside the biggest changes in body composition. Below that dose, glucagon activity was present but too small to show as a clear signal.
Phase 2, obesity (Jastreboff and colleagues, New England Journal of Medicine, 2023). 338 adults, 48 weeks:
| Weekly dose | Average weight loss |
|---|---|
| 1 mg | 8.7% |
| 4 mg | 17.1% |
| 8 mg | 22.8% |
| 12 mg | 24.2% |
| Placebo | 2.1% |
Two details matter as much as the headline. Weight was still falling when the trial ended, so 48 weeks is not a plateau. And the gains shrink as the dose climbs. Going from 4 mg to 8 mg added about five percentage points. Going from 8 mg to 12 mg added roughly one.
Phase 2, type 2 diabetes (Rosenstock and colleagues, The Lancet, 2023). 281 adults, 36 weeks. HbA1c fell by up to about 2 percentage points at the highest dose. No severe hypoglycaemia was reported.
Phase 2a, liver fat (Sanyal and colleagues, Nature Medicine, 2024). 98 participants with metabolic dysfunction-associated steatotic liver disease. At 24 weeks, liver fat fell 81.4% at 8 mg and 82.4% at 12 mg, against a slight increase on placebo. Normal liver fat — under 5% — was reached by 86% of participants at 12 mg and by none on placebo. These are among the largest liver fat reductions recorded for any drug.
Phase 2 body composition (Coskun and colleagues, Lancet Diabetes & Endocrinology, 2025). Total fat mass fell 23.2% from baseline at 12 mg. The authors' conclusion is the part that matters. The share of the loss that came from lean tissue was much like other obesity treatments. The bigger total loss did not come disproportionately from lean mass.
Structural work (Li and colleagues, Cell Discovery, 2024). Cryo-electron microscopy showed how one molecule engages all three receptors. The first nine residues are shared across all three. The middle section differs, giving each receptor its own fit.
Phase 3. The TRIUMPH programme runs across roughly 5,800 participants. The sponsor's report on TRIUMPH-4 describes average weight loss of 28.7% at 68 weeks at the top dose, plus improvement in knee osteoarthritis pain. Those figures come from a company announcement, not a peer-reviewed paper. Hold them more loosely than the phase 2 results above until the full publication appears.
Gut effects dominate, and they rise with dose. Nausea, vomiting, diarrhoea and constipation are the common ones. They cluster around dose increases and generally ease after the first weeks at a given level.
The distinctive one is dysesthesia, sometimes described as allodynia — a skin sensitivity in which ordinary touch registers as uncomfortable. Clothing against skin becomes noticeable; some report tingling or heightened sensitivity to pressure and temperature. It is dose-related, reported in roughly 7 to 21% of participants depending on dose, against under 1% on placebo. The suspected mechanism involves GLP-1 receptors in the nervous system.
Feeling cold is also widely reported, plausibly related to shifts in energy metabolism. Resting heart rate rose in trials, peaking around week 24 before declining.
One case of pancreatitis was reported in phase 2 and resolved. Discontinuation for adverse effects ran at roughly 12 to 18% at the higher doses in phase 3.
Trial eligibility is informative in itself. As with the rest of this drug class, a personal or family history of medullary thyroid carcinoma rules someone out. So does multiple endocrine neoplasia type 2, and so do pregnancy and breastfeeding.
Extra caution applies to several groups: a history of pancreatitis, gallbladder disease (rapid weight loss raises gallstone risk on its own), type 1 diabetes, diabetic retinopathy, gastroparesis, and active eating disorders.
Interactions matter too. Combined with insulin or sulfonylureas, there is a risk of blood sugar dropping too low. Slowed stomach emptying can change how oral medicines are absorbed, including contraceptives. Combined with SGLT2 inhibitors, the risk of ketoacidosis rises. Phase 3 added a ketoacidosis warning, which follows from glucagon activity raising ketone production.
Retatrutide has the strongest weight-loss data of any investigational obesity drug, and the liver-fat results are remarkable on their own terms.
It is also unapproved, with no marketing authorisation anywhere as of early 2026, a submission not yet filed, and no long-term safety record. Everything above was measured under supervision. Participants were screened, and anyone judged to be at too much risk was excluded. Monitoring was in place to catch problems early.
That last point is not a formality. The trial results describe what happened under those conditions. They do not describe what happens without them.
Filing is expected in late 2026 or 2027. We will cover the phase 3 publications as they appear.
This article explains published research and the rules around it. It carries no dose, no protocol, and no claim that any compound treats, cures or prevents a condition in anyone. Where the evidence is thin we say so. See our editorial standards.
Research-grade material, sold for laboratory research use. We link to a single supplier rather than a list, so you know exactly where the recommendation goes. Check the certificate of analysis against what a COA can and cannot show before you buy anything.
Commercial link. Nothing on this page is a dose, a protocol, or a claim that this compound treats or prevents any condition.
KLOW is GLOW plus 10 mg of KPV — that part is simple. The more useful fact is that both are, by mass, mostly copper peptide, and that is the ingredient whose injected form has no human trials.
An open study found sleep normalised in six of seven patients. The controlled study by the same investigator found weak effects and said so. The difference between them is a placebo arm.
It does not raise dopamine — it magnifies what other things do to it. That one property explains why caffeine feels stronger on it, and why MAO inhibitors are a hard contraindication.
