Not legally, and not as a medicine. It is not approved in any country. No regulator has been asked to approve it: Eli Lilly said on 23 July 2026 that it intends to file an application with the American regulator in the first quarter of 2027, and nothing has been filed in Europe or Britain either. The only lawful way to take it is to be enrolled in one of the company's clinical trials. Everything else — the websites, the clinics, the shops — is selling an unapproved drug, which the American regulator has repeatedly said is illegal regardless of a 'research use only' label on the vial.

What does the third receptor actually add?

Honestly: nobody has shown, in a person, what it adds. The theory is clean. GLP-1 and GIP make you eat less; glucagon makes the body burn more and pushes the liver to break down its stored fat. In obese mice that is exactly what happened — the extra weight loss beyond what the two gut receptors produced was traced to the glucagon receptor raising energy burn. In people, the only place its fingerprints are clearly visible is the liver, where fat fell by more than 80% and a blood marker of fat being burnt for fuel rose with the amount given. Whether it raises whole-body energy burn in a human being has been measured, in a study of 85 adults that finished in August 2025, and nothing has been published. Glucagon is also the most likely explanation for the faster heart rate and for the liver's tendency to push sugar into the blood, which in the trials was outweighed by the other two receptors.

How much weight does it actually take off?

In the published trial, 24.2% over 48 weeks on the highest amount against 2.1% on dummy injections. In the much larger unpublished phase 3 trial, 28.3% over 80 weeks against 2.2% — but that figure counts only the people who stayed on it, and counting everyone randomised gives 25.0% against 3.9%. In people who also have type 2 diabetes it is smaller: 20.8% against 4.0%. Read those alongside the number of people who quit: at the highest amount, between 11% and 18% of people left the trials because of side effects, against about 5% on dummy injections.

Is it better than tirzepatide or semaglutide?

On the headline weight numbers it looks better, and nobody has run the trial that would settle it. There is no published head-to-head comparison against either drug, so every claim rests on comparing across separate trials with different people in them. When an independent group did that across 262 trials and 99,791 people, it put tirzepatide at 14.9% below lifestyle change alone at one year on moderate to high certainty evidence, and retatrutide — grouped with two other unapproved drugs — at 13.1% to 14.6% on very low to low certainty evidence. Two head-to-head trials are running, one against semaglutide and one against tirzepatide, and neither has reported. What is settled is the other half: retatrutide's own trials show more people abandoning it than the trials of the approved drugs show for them.

Is the burning-skin thing real?

Yes, and it is the most distinctive thing about this drug. In the phase 3 knee trial one in five people on the highest amount reported it, against 1 in 140 on dummy injections. It means an ordinary touch registering as burning or as pain. The company describes it as mostly mild to moderate and mostly clearing up while treatment continued, with no timings published. It is not unique to retatrutide — a French analysis of official side-effect reports found the same thing with semaglutide and tirzepatide, more often at higher amounts — but it is much more common here. And it was there in the phase 2 trial too, in 4 of 62 people on the highest amount and none of 70 on dummy injections, under a different name and unremarked.

Does it protect the heart?

Unknown, and the one set of numbers that exists says nothing either way. In the trial of people who already had heart disease, a wider count of heart and blood-vessel events came 44 times on the drug against 52 on dummy injections, and a narrower count of heart attack, stroke and heart death came 27 times against 23. Both ranges cross the point of no difference. That trial was designed to measure weight, not events. The trial designed to measure events enrolled 10,000 people, is still running and is not expected to finish until February 2029. In the meantime the drug clearly improves the things that predict heart disease — blood pressure down, triglycerides down, an inflammation marker halved — which is not the same as preventing anything.

Somebody told me a man died after taking it. Is that true?

The British Medical Journal published a fact check on 9 July 2026 under exactly that question, following press coverage of a death. This register has not been able to read that article in full and so cannot report what it concluded, which is stated here rather than glossed over. What can be verified independently is thinner and more useful: the American regulator's public side-effect file holds 27 reports naming retatrutide as of 28 April 2026, most of them from 2025 and 2026, of which 22 are flagged serious, and among the reported terms are counterfeit product and product misuse. A report is not a cause. But the underlying problem is not in dispute: what is in an unapproved vial bought online or over a shop counter is unknown, and at least one certificate of analysis supplied with such a product has been confirmed fake by the laboratory whose name was on it.

Metabolic

Retatrutide

Q: What happens if I stop?

Nobody has published an answer for this drug. Not one trial has followed anyone after stopping. The extension trial went in the other direction: it kept people on it longer, and they kept losing. Across the approved drugs in the same family, pooled figures put about 60% of the lost weight back within a year of stopping — but that is a finding about other drugs, applied here by assumption. A trial specifically about holding on to weight after retatrutide is running and is not due to report until 2028.

Also known as LY3437943

Still being tested in people

7 of 51 peptides sit at this level

Category: Metabolic
Doping status: Banned in sport
Sources: 32
Trials: 2 trials · 2,677 people

Retatrutide is an experimental injection that switches on three hormone receptors at once — GIP and GLP-1, which are the two gut hormones the body releases after a meal, and glucagon, which is the hormone the body uses to raise blood sugar and burn stored fuel. The first two cut appetite and are already the basis of approved medicines; the third is what makes this one different, and it is also the least understood part of it. In the largest completed trial, people on the highest amount lost about 28% of their body weight over 80 weeks against about 2% on dummy injections — but that figure comes from a company press release, not a published paper, and no regulator anywhere has approved the drug or even been asked to. It is legally available only inside a clinical trial, and the version being sold online and in shops is of unknown content.

RetatrutideTrials

Trials
What it is
Sold and legal
In your body
What changed
What can go wrong
Amounts studied
What people say
Unknowns
Questions
Sources

Weight-loss trials

2 trials · 2023–2026

FAS 2 · 2023 · NEJM
Treatment-8.7 % … -24.2 %
placebo-2.1 %
N
338
WEEKS
48
[source] — Fas 2, NEJM 2023
TRIUMPH-1 · 2026 · ELI LILLY
Treatment-19 % … -28.3 %
placebo-2.2 %
N
2 339
WEEKS
80
Topline figures from a press release. Not published in a peer-reviewed journal, and therefore not comparable with the rows above.
[source] — TRIUMPH-1, Eli Lilly 2026

What it is

An experimental peptide that switches on three receptors at once: GIP, GLP-1 and glucagon. It is not approved as a medicine in any country.

How it is sold, and whether it is legal

Sold as: Injection under the skin (only within clinical trials).
Legal status in the EU: No marketing authorisation in the EU or the US. Eli Lilly said in July 2026 that it plans to file with the FDA in the first quarter of 2027, pushed back from an earlier plan for 2026. The substance is legally available only within clinical trials; anything else falls outside medicines law.

What it does in your body

9 parts of the body · 7 measured in people, 1 only seen in animals, 1 only seen in a dish

Hitting three receptors means the GLP-1 and GIP effects on insulin release and appetite are combined with switching on the glucagon receptor. How much each receptor contributes to the effect in people is not established. The sequence has not been verified against a citable source and is left empty.

Appetite, in the brain: Hunger drops and the urge to keep eating past the point of fullness drops with it. When adults with type 2 diabetes were asked to rate their own appetite, the ones on the middle and higher amounts reported less overall appetite, less hunger and less of a sense that they could keep eating. They also scored lower on a measure of the tendency to overeat once eating has started — the thing that turns one biscuit into the packet. How much of this comes from GLP-1, how much from GIP and how much from the glucagon receptor has not been separated out in people.; Self-rated appetite and eating behaviour measured in 275 adults with type 2 diabetes at 24 and 36 weeks, comparing retatrutide against both dummy injections and dulaglutide, an existing GLP-1 medicine. Everyone on 4 mg or more rated their appetite, hunger and expected food intake lower than the dummy group at 24 weeks. The comparison against dulaglutide was less consistent. The falls in hunger and in overeating tracked the weight loss, but loosely — the correlations were 0.28 and 0.36, which is a real relationship and a long way from a tight one.; Measured in people; Source [05]Source [06]
Stomach and gut: Feeling sick is the commonest thing that happens, followed by loose stools, constipation and vomiting. It arrives while the amount is being built up rather than at the end, and it is strongly tied to how fast that build-up goes. Two groups in the phase 2 trial finished on exactly the same amount but were worked up to it at different speeds, and the slower group had roughly a third as much sickness as the faster one. Most of it is mild or moderate and settles while treatment carries on, but it is the main reason people leave the trials.; In the phase 2 obesity trial's posted registry results, nausea was recorded in 28 of 62 people on the highest amount against 8 of 70 on dummy injections; vomiting in 12 of 62 against 1 of 70; constipation in 10 of 62 against 2 of 70. Loose stools barely moved: 9 of 62 against 8 of 70. Of the two groups that both ended on 8 mg, the one built up more slowly had sickness in 6 of 35 people and the one built up faster in 21 of 35. In the phase 3 obesity trial, nausea ran 28.6% to 42.4% across the three amounts against 14.8% on dummy injections, and loose stools 25.2% to 34.1% against 13.5% — figures from a company press release, not a published paper.; Measured in people; Source [01]Source [02]Source [07]
Blood sugar, and the pancreas: This is where the three receptors pull against each other, and it matters. GLP-1 and GIP tell the pancreas to release insulin when blood sugar is high, which brings sugar down. Glucagon does the opposite: it tells the liver to release stored sugar into the blood, which pushes sugar up. In people the first two win. In the phase 3 diabetes trial, long-term blood sugar fell by up to 1.94 percentage points against 0.81 on dummy injections, and 82% to 89% of people got below the usual diabetes threshold. In the heart-disease trial, high blood sugar was actually reported as a problem less often on the drug than on dummy injections — 3.1% against 13.4%. No severe low-blood-sugar episode has been reported in the published trials.; Long-term blood sugar measured in 537 adults with type 2 diabetes over 40 weeks at 48 centres in the United States, Mexico and India, published in full in the Lancet. Counting everyone randomised, it fell 1.69, 1.86 and 1.94 percentage points on the three amounts against 0.81 on dummy injections; the gaps were 0.88 (95% CI 1.18 to 0.59), 1.04 (1.32 to 0.76) and 1.12 (1.39 to 0.85). Starting level was 7.9%. The figure for high blood sugar being reported less often comes from the TRIUMPH-3 press release, not a paper.; Measured in people; Source [08]Source [09]Source [10]
The liver: Fat stored inside the liver falls further and faster here than with any approved drug in this class, and this is the one place where the glucagon receptor's fingerprints are visible in people. Switching on the glucagon receptor tells the liver to burn fat rather than store it, and the blood shows exactly that: a marker of fat being burnt for fuel rose in step with the amount given. In people who started with a fatty liver, most ended the study with a normal one.; Liver fat measured by scan in 98 adults who had at least 10% of their liver made of fat at the start, inside the phase 2 obesity trial, published in Nature Medicine. At 24 weeks liver fat had fallen 42.9%, 57.0%, 81.4% and 82.4% across the four amounts against a 0.3% rise on dummy injections. By 24 weeks liver fat was back in the normal range in 27%, 52%, 79% and 86% of people against none of the dummy group. Liver enzymes did not show a pattern of harm.; Measured in people; Source [11]
Heart and blood vessels: Two things happen at once and they point in opposite directions, exactly as they do with the approved drugs in this class — but the resting heart rate rise is bigger here. Blood pressure falls, blood fats improve, and a marker of inflammation in the blood drops by about half. Against that, the heart beats faster at rest, by an amount that grows with the amount given. In the phase 2 trial the rise peaked at 24 weeks and then came down again while treatment continued. Nobody knows what a persistent rise of this size does over years.; Resting heart rate pooled in a 2026 systematic review: 3.46 beats a minute above dummy injections (95% CI 1.74 to 5.18) across 330 people on retatrutide against 70 on dummy injections. By amount, the rise was 1.26 beats (95% CI −3.01 to 5.67) on 1 mg, 2.63 (−1.35 to 6.56) on 4 mg, 4.35 (0.09 to 8.39) on 8 mg and 5.52 (1.23 to 9.89) on 12 mg — the two lower figures answer nothing, the two higher ones do. For comparison the review put the whole GLP-1 class at 3.47 beats (2.65 to 4.29) across 12 trials and 15,313 people. Blood pressure pooled across the trials fell 6.79 mmHg on the top number (95% CI 8.36 to 5.23) and 2.46 on the bottom (3.25 to 1.67).; Measured in people; Source [01]Source [12]Source [13]
Skin and nerves: Some people's skin starts to hurt when it should not. The medical word is dysaesthesia — an ordinary touch, a shirt, a shower, registering as burning or as pain. It is the one effect on this drug that is not a familiar feature of the approved drugs in this class, and it is much commoner at the top amount than below it. In the trials it was mostly mild or moderate and mostly went away while treatment carried on. It was already there in phase 2 under a different name, in a handful of people, and nobody flagged it.; In the phase 3 knee trial, dysaesthesia was reported by 20.9% of people on the top amount and 8.8% on the one below, against 0.7% on dummy injections. In the phase 3 obesity trial it was 5.1%, 12.3% and 12.5% across the three amounts against 0.9%; in the diabetes-and-obesity trial 4.5%, 5.6% and 7.3% against 0.7%. All four figures are from company press releases, not published papers. In the phase 2 obesity trial's posted registry results, allodynia — pain from a touch that should not hurt — was recorded in 4 of 62 people on the top amount and in none of the 70 on dummy injections, with oversensitive skin in a further 2 of 62.; Measured in people; Source [07]Source [14]Source [15]
Body fat, and muscle: The weight that comes off is mostly fat, and the waist shrinks by more than the scales alone would suggest. Lean tissue — muscle and everything else that is not fat — comes off too, but the trial that measured it found the share of the loss that was lean tissue was about the same as with other weight drugs, not worse. That is a comparison of proportions, not of amounts: a bigger total loss at the same proportion still means more muscle gone.; Body composition scanned in a substudy of the phase 2 diabetes trial: 189 people enrolled, 103 had both a starting and a 36-week scan. Total body fat fell 15.2% on the 4 mg groups pooled, 26.1% on the 8 mg groups pooled and 23.2% on 12 mg, against 4.5% on dummy injections and 2.6% on dulaglutide. Against dummy injections the differences were 21.6 points (95% CI 27.1 to 16.1) for 8 mg and 18.7 points (25.1 to 12.3) for 12 mg. Waist measured in the phase 3 obesity trial fell 16.3, 21.8 and 24.1 cm across the three amounts against 3.6 cm — a press release figure, from a starting waist of 118.3 cm.; Measured in people; Source [02]Source [16]
How much energy the body burns: This is the claim the whole drug is built on and the one with the thinnest human evidence. The argument runs: the two gut receptors make you eat less, the glucagon receptor makes you burn more, and that is why the weight loss is bigger than with drugs that only do the first. The burning-more half has been demonstrated in obese mice, where blocking the glucagon receptor removed the extra weight loss. In people it has not been shown. A study designed to answer exactly this — 85 adults with obesity, measuring both how much they ate and how much energy they burnt over 24 hours — finished on 26 August 2025 and has published nothing.; In obese mice, the extra weight loss over what the two gut receptors produced was attributed to the glucagon receptor raising energy burn. In the same paper, tests in a dish found roughly balanced activity at the glucagon and GLP-1 receptors and more at the GIP receptor — a fact about cells, not about people. The human study, registered as NCT06313528, measured 24-hour energy burn, sleeping metabolic rate and calories eaten at supervised meals; it is marked completed with no results posted.; Only seen in animals; Source [17]Source [18]
Heart muscle, tested outside the body: Small pieces of heart muscle taken from people having heart surgery squeezed harder and relaxed faster when retatrutide was added to the bath they were sitting in. Blocking any one of the three receptors weakened the effect; blocking the receptors that adrenaline works through did not. That is a direct effect on heart muscle rather than a knock-on effect of adrenaline, and it happened at concentrations a person on this drug actually reaches in the blood. What it means for a heart inside a living body over years is not known.; Strips of right atrial tissue from adults undergoing open heart surgery for severe coronary artery disease, tested at concentrations from 10 to 100 nanomolar. The paper notes that 8 mg of retatrutide produces a peak blood concentration of about 182 nanomolar in people, so the tissue was tested inside the range people reach. This is tissue in a bath, not a person.; Only seen in a dish, not in a body; Source [19]

What changed when it was measured

13 findings · 12 measured in people, 1 where studies in people disagree

In the phase 2 study (NEJM 2023, 338 participants, 48 weeks) weight change ranged from −8.7% in the lowest dose group to −24.2% in the highest, against −2.1% for placebo. According to Eli Lilly's headline report from the phase 3 study TRIUMPH-1 (21 May 2026, 2,339 participants, 80 weeks) weight loss ranged from −19.0% to −28.3% across dose groups against −2.2% for placebo. Those phase 3 results come from the company's press release and have not yet been published in a peer-reviewed journal.

Slow on, and nobody has published what happens coming off. Changes in eating turn up early: in the exit interviews after the phase 2 trial, 31 of the 36 people who had been on the drug described their eating having changed within the first eight weeks. Sickness and the other gut effects cluster in the same early stretch, while the amount is being built up, and mostly settle while treatment carries on. Resting heart rate climbed to a peak at about 24 weeks in the phase 2 trial and then came down again. The burning or oversensitive skin was described by the company as mostly resolving during treatment, with no timing published. Weight comes off far more slowly than the headline numbers suggest and does not stop when the trials do. In the phase 3 diabetes trial the investigators noted that weight loss had not reached a plateau when the 40-week trial ended. In the phase 3 obesity trial people were still losing at 80 weeks, and a group that carried on to 104 weeks went from about 28% down to about 30% down. What happens when someone stops is the largest hole in the record: not one published study has followed anyone off this drug. Across the approved drugs in the same family, pooled figures put about 60% of the lost weight back within a year of stopping, but that is a class finding from other drugs, not a measurement of this one.

Body weight in adults with obesity, over 48 weeks: Down 8.7% on the lowest amount, 17.1% on 4 mg, 22.8% on 8 mg and 24.2% on 12 mg, against 2.1% on dummy injections. At the top amount, 100% of people lost at least 5% of their weight, 93% lost at least 10% and 83% lost at least 15% — against 27%, 9% and 2% on dummy injections. This is the published trial, and it is the smallest of the weight trials.; 338 adults with obesity, or overweight plus at least one weight-related condition, randomised across six drug groups and a dummy group, 48 weeks; 51.8% were men; Measured in people; Source [01]
Body weight in adults with obesity, over 80 weeks — the largest trial, unpublished: Counting only people who stayed on the drug: down 19.0%, 25.9% and 28.3% across the three amounts against 2.2% on dummy injections. Counting everyone randomised, including those who stopped: down 17.6%, 23.7% and 25.0% against 3.9%. At the top amount 62.5% lost at least a quarter of their body weight against 2.2%, and 45.3% lost at least 30% against 0.5%. Two thirds — 65.3% — were no longer in the obesity range at the end. Every figure in this row comes from an Eli Lilly press release of 21 May 2026. The trial was presented at a conference on 6 June 2026 and had still not been published in any journal as of 2 August 2026.; 2,339 adults with obesity or overweight randomised four ways, 80 weeks, with groups inside it who also had knee osteoarthritis or sleep apnoea; Measured in people; Source [02]Source [20]Source [21]
Whether the weight loss had stopped by the end — it had not: A group who carried on for a further 24 weeks kept losing. At 104 weeks they were down 27.9%, 29.5% and 30.3% depending on which amount they had started on. The comparison here is not a dummy group in any useful sense: the people who had been on dummy injections were switched onto the drug for the extension and lost 19.2% themselves. So the honest reading is that two years in, the curve was still going down, not that the gap had widened. Press release figures.; 532 people from the same trial who had a body mass index of 35 or above at the start, followed to 104 weeks; Measured in people; Source [02]
Blood sugar and weight in adults with type 2 diabetes — the one phase 3 trial that has been published: Long-term blood sugar fell 1.69, 1.86 and 1.94 percentage points across the three amounts against 0.81 on dummy injections; the gaps were 0.88 (95% CI 1.18 to 0.59), 1.04 (1.32 to 0.76) and 1.12 (1.39 to 0.85), all p<0.0001. Weight fell 11.5%, 13.9% and 15.3% against 2.6%. 91% of people completed the treatment period still on the drug. No severe low-blood-sugar episode occurred. Two people died, both in the lowest-amount group, and the investigators judged neither death related to the drug.; 537 adults with type 2 diabetes not controlled by diet and exercise alone, average 2.5 years since diagnosis, at 48 centres in the United States, Mexico and India, 40 weeks; Measured in people; Source [08]
Liver fat in people who started with a fatty liver: Down 42.9%, 57.0%, 81.4% and 82.4% across the four amounts at 24 weeks, against a 0.3% rise on dummy injections. By 48 weeks the top amount was down 86.0% against a 4.6% fall in the dummy group. Liver fat was back inside the normal range in 86% of people on the top amount at 24 weeks against none of the dummy group. Weight in this subgroup fell between 8.6% and 25.9% against 0.1%.; 98 adults inside the phase 2 obesity trial who had at least 10% of their liver made of fat at the start, 48 weeks; Measured in people; Source [11]
Body fat measured by scan, in people with type 2 diabetes: Total body fat down 15.2% on the 4 mg groups pooled, 26.1% on the 8 mg groups pooled and 23.2% on 12 mg, against 4.5% on dummy injections and 2.6% on dulaglutide, an existing weekly GLP-1 medicine. Against dummy injections that is a gap of 21.6 points (95% CI 27.1 to 16.1) for 8 mg and 18.7 points (25.1 to 12.3) for 12 mg. The authors report that the share of the total loss that was lean tissue was similar to other weight drugs.; 189 adults enrolled in the body-composition substudy of the phase 2 diabetes trial, of whom 103 had both a starting and a 36-week scan; Measured in people; Source [16]
Blood sugar and weight in adults who had both type 2 diabetes and obesity, over 80 weeks: Weight down 12.7%, 19.1% and 20.8% across the three amounts against 4.0% on dummy injections. Long-term blood sugar down 1.4, 1.6 and 1.5 percentage points against 0.2, from a starting level of 7.7%. The weight loss is markedly smaller than in people without diabetes on the same amounts in the same programme — 20.8% against 28.3% — and that gap is a consistent feature of this whole class of drugs. Press release figures.; 1,152 adults with type 2 diabetes and obesity or overweight, randomised four ways, 80 weeks; Measured in people; Source [10]
Weight and heart risk factors in people with severe obesity who already had heart disease: Weight down 21.6% and 22.6% on the two amounts against 3.2% on dummy injections. On the top amount, blood triglycerides down 37.0%, non-HDL cholesterol down 16.5%, the top blood pressure number down 9.3 mmHg, waist down 19.0 cm and a marker of inflammation in the blood down 51.2%. Press release figures.; 1,949 adults with a body mass index of 35 or above and established heart or artery disease, with or without type 2 diabetes, 80 weeks; Measured in people; Source [10]
Heart attacks and strokes — the only such figures that exist, and they answer nothing: In the trial above, a wider count of heart and blood-vessel events happened 44 times on the drug and 52 times on dummy injections — a hazard ratio of 0.82, meaning events came at 82% of the rate, with a range of possible true values running from 0.55 to 1.22. A narrower count of heart attack, stroke and heart death happened 27 times against 23, a hazard ratio of 1.12 with a range from 0.64 to 1.96. Both ranges cross 1, which is the point of no difference, in both directions, so neither shows benefit and neither shows harm. This was a weight trial that counted events as they came, not a trial built to answer the question. The trial built to answer it, in 10,000 people, is not due to report until 2029.; The same 1,949 adults with severe obesity and established heart disease, 80 weeks; Measured in people; Source [10]Source [22]
Knee pain in people with obesity and knee osteoarthritis: Pain scores fell 4.5 and 4.4 points on the two amounts — 75.8% and 74.3% — against 2.4 points, or 40.3%, on dummy injections, from a starting score of 6.0. About one in eight people ended the trial with no knee pain at all: 14.1% and 12.0% against 4.2%. Weight fell 26.4% and 28.7% against 2.1%. Note how much the dummy group improved: losing 40% of your knee pain on a dummy injection is a large effect, and the drug's own effect is what sits on top of that. Press release figures.; 445 adults with obesity or overweight and knee osteoarthritis, 84% of them with a body mass index of 35 or above, randomised three ways, 68 weeks; Measured in people; Source [14]Source [23]
Breathing interruptions during sleep: In the group inside the phase 3 obesity trial who also had moderate or severe sleep apnoea, the number of times breathing stopped or went shallow each hour fell by up to 60.6%. Knee pain in the parallel group inside the same trial fell by up to 73.1%. No comparison figure for the dummy group has been released for either, and neither has been published — these numbers come from a conference presentation reported in a medical journal's meeting coverage.; Subgroups within the 2,339 adults of the phase 3 obesity trial who also had knee osteoarthritis or obstructive sleep apnoea, 80 weeks; Measured in people; Source [09]Source [21]
Blood pressure and blood fats, pooled across the trials: Top blood pressure number down 6.79 mmHg (95% CI 8.36 to 5.23), bottom number down 2.46 (3.25 to 1.67). Total cholesterol down 21.88 mg/dL (27.79 to 15.97), LDL cholesterol down 13.10 (16.10 to 10.10), triglycerides down 40.90 (48.50 to 33.30). HDL cholesterol did not move at all: 0.01 mg/dL, with a range from 1.30 down to 1.27 up.; Randomised trials of retatrutide pooled in a 2026 meta-analysis; the underlying trials are the phase 1 and phase 2 studies; Measured in people; Source [13]
How it stands against the approved drugs, when someone else does the comparing: Nobody has run retatrutide head to head against tirzepatide or semaglutide in a published trial, so every comparison is indirect. A 2026 review of 262 trials in 99,791 people put tirzepatide at 14.9% below lifestyle change alone at one year (95% CI 16.0 to 13.9) and weekly semaglutide at 9.8% (10.6 to 9.1), both on moderate to high certainty evidence. It put retatrutide, alongside two other unapproved drugs, at 13.1% to 14.6% — on very low to low certainty evidence, which is the review's way of saying the number could move a lot. A separate review of glucagon-receptor drugs put retatrutide first on weight, at 13.44 kg below dummy injections (95% CI 18.38 to 8.51), and among the worst of that group on how well people tolerated it.; 262 randomised trials of 19 weight drugs, 99,791 participants, follow-up 12 to 172 weeks; and separately 14 randomised trials of glucagon-receptor drugs; Studies in people disagree; Source [24]Source [25]Source [26]

What can go wrong

11 effects, 3 serious

In TRIUMPH-1, nausea was reported by 28.6–42.4%, diarrhoea by 25.2–34.1% and altered sensation by 5.1–12.5% across dose groups, against 14.8%, 13.5% and 0.9% for placebo. Up to 11.3% in the highest dose group stopped because of side effects, against 4.9% for placebo. Long-term safety is unknown.

Inflammation of the pancreasSerious: Severe pain high in the belly that bores through to the back, often with vomiting. It is a known risk of every approved drug in this family and there is no reason to expect this one to be different, but the trials so far are far too small to say how often it happens here. Nobody has published a pooled count across the phase 3 programme.; One person in the phase 2 obesity trial, out of 62 on the highest amount, against none of the 70 on dummy injections. That is a single case, and a single case sets no rate.; Source [07]
Gallbladder inflammation and kidney injurySerious: Both are recognised knock-on effects in this class rather than direct ones. Rapid weight loss causes gallstones by itself; kidneys fail when days of vomiting or loose stools have emptied the body of fluid. Kidney injury also appears in the small number of side-effect reports the American regulator has received about people using retatrutide obtained outside a trial.; One case each in the phase 2 obesity trial, both in a group of 35 people on 8 mg, against none of the 70 on dummy injections. Single cases, no rate.; Source [07]Source [27]
Whatever is in the vial, when it did not come from a trialSerious: This is not a side effect of retatrutide. It is a side effect of buying something labelled retatrutide. A news investigation published on 2 July 2026 bought a vial over the counter at a Brooklyn convenience store for $95 with no questions asked, and found that the certificate of analysis supplied with a competing product was fake — the laboratory named on it confirmed the document was not theirs, and the accompanying graph identified a different drug altogether. The regulator's own position is that these products are of unknown quality and may be harmful.; Not measured, and unmeasurable. The American regulator's public side-effect file holds 27 reports mentioning retatrutide as of 28 April 2026, of which 22 are flagged serious. Among the reported terms are suspected counterfeit product, counterfeit product administered and intentional product misuse.; Source [27]Source [28]Source [29]
Skin that burns or hurts when touched: An ordinary touch — clothing, bedding, water in the shower — registering as burning or as pain. The company describes it as generally mild to moderate with most cases clearing up while treatment continued. It is not new to this family of drugs: a French pharmacovigilance analysis found the same pattern with semaglutide and tirzepatide, more often at higher amounts and with the stronger drugs, usually settling after the drug was stopped. What is new is how common it is here. It was present in phase 2 too, under the name allodynia, in 4 of 62 people on the top amount and none of 70 on dummy injections, and nobody named it until phase 3.; The signature effect of this drug, and strongly tied to the amount. 20.9% of people on the top amount in the phase 3 knee trial against 0.7% on dummy injections; 12.5% against 0.9% in the phase 3 obesity trial; 7.3% against 0.7% in the diabetes-and-obesity trial; 6.4% against 1.3% in the heart-disease trial. All four are press release figures.; Source [02]Source [07]Source [14]Source [15]
Feeling sick: It clusters in the weeks while the amount is being increased and mostly settles afterwards. How the amount is built up matters more than almost anything else: of two phase 2 groups that both ended on the same 8 mg, the one taken up more slowly had sickness in 6 of 35 people and the one taken up faster in 21 of 35. That is roughly a threefold difference in how many people felt ill, for the same drug at the same final amount.; The commonest effect throughout. 28 of 62 people on the top amount in the published phase 2 trial against 8 of 70 on dummy injections. In the phase 3 trials, between 28.6% and 43.2% depending on the trial and the amount, against 5.8% to 14.8% on dummy injections — press release figures.; Source [01]Source [02]Source [07]
Loose stools: The phase 2 numbers are worth pausing on. Loose stools were about as common on dummy injections as on the drug in that trial, which is what it looks like when a symptom is mostly not the drug. The phase 3 trials, which are ten times larger, found a clear gap. Beyond the discomfort, this is the route by which people get dried out, which is how the kidney problems in this class happen.; 25.2% to 34.1% across the amounts in the phase 3 obesity trial against 13.5% on dummy injections; 30.1% against 8.7% in the heart-disease trial. Press release figures. In the published phase 2 trial the gap was much smaller: 9 of 62 on the top amount against 8 of 70 on dummy injections.; Source [07]Source [10]
Constipation: The mirror image of the loose stools, from the same slowing of the gut, and the same person can have both in different weeks.; 23.8% to 26.1% across the amounts in the phase 3 obesity trial against 10.9% on dummy injections. In phase 2, 10 of 62 on the top amount against 2 of 70.; Source [02]Source [07]
Vomiting: Usually early and usually short-lived, and it tracks the sickness. In the published phase 2 trial one person in the whole 70-person dummy group vomited, against roughly one in five on the top amount — a much wider gap than for the loose stools.; 10.6% to 25.3% across the amounts in the phase 3 obesity trial against 4.8% on dummy injections. In phase 2, 12 of 62 on the top amount against 1 of 70.; Source [02]Source [07]
A faster resting heart: It peaked at about 24 weeks in the phase 2 trial and came down again while treatment carried on. Most people never notice it. Set against the approved drugs, the class average is about 3.5 beats a minute, so the lower amounts of retatrutide sit inside the normal range for this family and the top amount sits above it. Whether a persistent rise of five beats a minute matters over years has not been established for any drug in this class, and there is no long-term trial here at all.; An average rise of 3.46 beats a minute above dummy injections (95% CI 1.74 to 5.18), rising with the amount: 4.35 beats on 8 mg and 5.52 on 12 mg, against 1.26 on the lowest amount, where the range crossed no difference and answered nothing.; Source [01]Source [12]
Urinary infections: A small and inconsistent excess that only became visible when the trials got large — in one of the three the dummy group had more than the lower amount did. It has drawn enough attention to prompt a published commentary asking whether the timing of the cases explains it. Whether it is a real effect of the drug is not settled.; 7.5% to 8.8% across the amounts in the phase 3 obesity trial against 5.3% on dummy injections; 3.8% to 8.0% against 6.6% in the diabetes-and-obesity trial; 6.1% and 7.0% against 5.3% in the heart-disease trial. Press release figures.; Source [02]Source [10]Source [30]
Giving up because of the side effects: The single most useful number on this page, because it is what the benefit costs in practice. At the lowest amount, as many people quit for side effects as quit on a dummy injection. At the highest amount in the knee trial, nearly one in five did. The independent review of 262 weight-drug trials makes the same point about the whole field: the drugs that take off the most weight are also the ones people abandon most.; Between 4.1% and 11.3% across the amounts in the phase 3 obesity trial against 4.9% on dummy injections; 12.2% and 18.2% against 4.0% in the knee trial, which used only the two higher amounts; 2% to 5% against 0% in the published diabetes trial.; Source [02]Source [14]Source [24]

Who it is known to be dangerous for

Nobody has established this, and that is the honest answer rather than a gap in the research for this page. Every other record in this register can point at a product information document where a regulator has written down who must not take the drug. Retatrutide has no such document anywhere in the world, because no regulator has approved it and none has yet been asked to — Eli Lilly said on 23 July 2026 that it intends to file an application with the American regulator in the first quarter of 2027. Until that happens and is assessed, there is no authorised list of people it is dangerous for. What can be said is what the trials did. They enrolled adults; nobody under 18 has been studied. The type 2 diabetes trials enrolled people whose diabetes was recent — an average of 2.5 years since diagnosis in the published phase 3 trial — and studied the drug on its own, not alongside insulin or the older sulfonylurea tablets, so the combination that causes dangerously low blood sugar with every other drug in this family has not been tested here. Pregnancy and breastfeeding were not studied. The heart-disease trial required a body mass index of 35 or above, so it says nothing about people with heart disease and a lower weight. A separate trial in people with reduced kidney function is running, and its results are not out. The risks the whole family of drugs carries — inflammation of the pancreas, gallstones, kidneys failing after days of vomiting or diarrhoea, food still sitting in the stomach when someone is put under anaesthetic — have no reason not to apply here, and there is no reason to assume they are smaller. And separately from all of that: it is on the World Anti-Doping Agency's prohibited list at all times, as a substance in clinical development without regulatory approval, so an athlete subject to testing must not take it.

The amounts the studies used

8 amounts

These are the amounts the studies below gave their participants, and they are here as facts about those studies. They are not a recommendation, not a starting point and not a range to pick from. What a person should take, if anything, is a question for a doctor who knows them.

The phase 2 obesity trial, 338 adults with obesity or overweight — the trial the published weight figures come from: 1 mg, 4 mg, 8 mg or 12 mg once a week, injected under the skin. Two of the amounts were tested twice over, once reached gradually and once reached faster, to see whether that changed how ill people felt; 48 weeks; Source [01]
The phase 2 type 2 diabetes trial, 281 adults, which also compared the drug against dulaglutide, an existing weekly GLP-1 medicine: 0.5 mg, 4 mg, 8 mg or 12 mg once a week, injected under the skin, against dulaglutide 1.5 mg once a week and against dummy injections; 36 weeks, with the main measurement at 24 weeks; Source [06]
The liver-fat substudy inside the phase 2 obesity trial, 98 adults who had at least 10% of their liver made of fat: 1 mg, 4 mg, 8 mg or 12 mg once a week, injected under the skin; 48 weeks, with the main measurement at 24 weeks; Source [11]
TRANSCEND-T2D-1, the phase 3 trial in 537 adults with type 2 diabetes — the only phase 3 trial published in full: 4 mg, 9 mg or 12 mg once a week, injected under the skin, reached by stepwise increases over the earlier part of the trial; 40 weeks; Source [08]
TRIUMPH-1, the phase 3 obesity trial in 2,339 adults, whose results exist only as a company press release: 4 mg, 9 mg or 12 mg once a week, injected under the skin, reached by stepwise increases over the earlier part of the trial; 80 weeks, with 532 people carrying on to 104 weeks in an extension; Source [02]Source [20]
TRIUMPH-2, the phase 3 trial in 1,152 adults with type 2 diabetes and obesity or overweight, results reported only as a company press release: 4 mg, 9 mg or 12 mg once a week, injected under the skin; 80 weeks; Source [10]
TRIUMPH-3, the phase 3 trial in 1,949 adults with severe obesity and established heart disease, results reported only as a company press release: 9 mg or 12 mg once a week, injected under the skin; 80 weeks; Source [10]
TRIUMPH-4, the phase 3 trial in 445 adults with obesity or overweight and knee osteoarthritis, results reported only as a company press release: 9 mg or 12 mg once a week, injected under the skin; 68 weeks; Source [14]Source [23]

What people report

What follows is what people say online. It is not evidence, it is not graded, and it is not checked by anyone. People who had a bad time and people who are selling something both post more than people for whom nothing happened. It is here because you would go and read it anyway, and knowing who is talking is better than not.

What people describe online is not what the trials describe — the top complaint is being hungry

In an analysis of 7,823 people posting about their own retatrutide use, the single most reported symptom was increased appetite, at 17.1%, followed by fatigue at 15.2% and increased energy at 12.4%. Nausea came fourth at 11.6%. Food craving was reported by 7.3% and insomnia by 6.1%. The authors' own reading is that this could be the glucagon receptor, or product quality and unknown strength, or people taking several unapproved peptides at once, or simply that the people posting are different from the people in trials — and that they cannot tell which.

Read in A cross-sectional analysis of 148,640 posts and comments from 6 retatrutide-specific and 21 broader peptide or weight-management communities on Reddit between May 2021 and December 2025, posted to medRxiv on 3 June 2026 by researchers at the University of Pennsylvania and Boston Children's Hospital. It is a preprint and has not been through peer review.