Key takeaways
- Every published trial we found that recorded retatrutide’s side effects tested Eli Lilly’s investigational product and was funded by Lilly. It is not approved anywhere.
- In every trial abstract we read that ranks adverse events, gastrointestinal events were the most common; heart rate rose in the phase 2 obesity trial, peaking at 24 weeks.
- In TRIUMPH-2, 4% to 12% of people in the three retatrutide groups stopped treatment for good because of adverse events or death, against 5% with placebo; in the knee osteoarthritis trials, serious adverse events were more common than with placebo.
- FDA’s public data held 83 reports naming retatrutide to 28 October 2026; in 74 it was a suspect or interacting product, 45 of those were serious and 2 recorded a death (our count). FDA says such reports cannot establish cause.
- Products sold online as retatrutide are not the trial drug: FDA has warned sellers of research-labelled retatrutide, and in Australia a product sold as retatrutide and linked to a hospital admission held semaglutide and no retatrutide.
01Retatrutide side effects at a glance
What trials, regulators and reports have recorded, dated 18 November 2026.
| Question | What is recorded |
|---|---|
| What retatrutide is | An investigational drug from Eli Lilly: one peptide that activates the GIP, GLP-1 and glucagon receptors. The TGA wrote on 5 November 2026 that it “has not been approved by the TGA or any comparable international regulator”. |
| Where its side effects were measured | In the published trials, all of them Lilly’s trials of its own product, funded by Lilly. The longest ran for 80 weeks. A phase 2 trial by another sponsor is registered, with no results yet. |
| Most common adverse events in trials | Gastrointestinal: nausea, diarrhoea, vomiting and constipation. Where the abstracts grade them, mostly mild to moderate. |
| Other events more frequent than with placebo | A rise in heart rate (phase 2 obesity trial, peaking at 24 weeks); low blood pressure and dysaesthesia, an altered skin sensation (TRIUMPH-2); serious adverse events (both knee osteoarthritis trials). |
| Stopped treatment because of adverse events | TRIUMPH-2: 4% to 12% across the three retatrutide groups, against 5% with placebo (adverse events or death). TRANSCEND-T2D-1: 2% to 5% with retatrutide, 0% with placebo. |
| Deaths in trials | TRIUMPH-2: two, three and one in the three retatrutide groups, one with placebo. TRANSCEND-T2D-1: two, both with retatrutide. TRIUMPH-2’s investigators judged all its deaths unrelated to the study intervention; the TRANSCEND-T2D-1 abstract describes its two as unrelated to the study drug. |
| FDA adverse-event reports | 83 reports name retatrutide (data to 28 October 2026). In 74 it is a suspect or interacting product; 45 of those are serious and 2 record a death (our count). FDA says a report cannot establish cause. |
| Regulator safety finding | FDA, October 2026: retatrutide has “not been found safe and effective for any condition”. |
| Products sold online | Not the trial drug. Regulators in the US, UK and Australia have acted on them. In Australia, one product sold as retatrutide and linked to a hospital admission held semaglutide and no retatrutide. |
02What side effects did retatrutide trials record?
Every published trial below tested Eli Lilly’s product, was funded by Lilly and compared it with placebo; two also used dulaglutide, an approved prescription GLP-1 medicine. We list what each abstract reports on adverse events. We give no efficacy results and no dose arms.
| Trial and paper | Who took part, and the comparison | Length | Adverse events, as the abstract reports them |
|---|---|---|---|
| Phase 1b · The Lancet, 2022 | 72 adults with type 2 diabetes; placebo and dulaglutide | 12 weeks | Treatment-emergent adverse events in 63% of people given retatrutide, 60% given dulaglutide and 54% given placebo; gastrointestinal events most frequent. 29 of the 72 left the study early; the abstract gives no reasons by group. |
| Phase 2 obesity · NEJM, 2023 | 338 adults with obesity, or overweight with a weight-related condition; placebo | 48 weeks | Gastrointestinal events most common, mostly mild to moderate. Heart rate rose, peaked at 24 weeks and declined after that. |
| Phase 2 type 2 diabetes · The Lancet, 2023 | 281 adults with type 2 diabetes in the US; placebo and dulaglutide | 36 weeks | Mild to moderate gastrointestinal events in 35% of the retatrutide groups combined, 13% with placebo and 35% with dulaglutide. No severe hypoglycaemia and no deaths. |
| TRANSCEND-T2D-1, phase 3 · The Lancet, June 2026 | 537 adults with type 2 diabetes not controlled by diet and exercise, in the US, Mexico and India; placebo | 40 weeks | Gastrointestinal events most frequent, generally mild to moderate, and they subsided over time. Stopped the study drug because of adverse events: 2% to 5% with retatrutide, 0% with placebo. No severe hypoglycaemia. Two deaths, both with retatrutide, described in the abstract as unrelated to the study drug. |
| TRIUMPH-1, phase 3 · NEJM, online 29 Sep 2026 | 2,339 adults with obesity and without diabetes; placebo | 80 weeks | Gastrointestinal events most common. The abstract gives no adverse-event rates. |
| TRIUMPH-2, phase 3 · The Lancet, online 29 Sep 2026 | 1,152 adults with obesity and type 2 diabetes, in eight countries; placebo | 80 weeks | Diarrhoea in 27% to 34% across the retatrutide groups against 13% with placebo; nausea 14% to 28% against 8%. Low blood pressure 1% to 6% against under 1%; dysaesthesia 4% to 7% against 1%. Stopped treatment for good because of adverse events or death: 4% to 12% against 5%. Deaths: two, three and one in the retatrutide groups, one with placebo, all judged unrelated by the investigators. |
| TRIUMPH-4 and TRIUMPH-1-OA, phase 3 · NEJM, online 8 Nov 2026 | 445 and 574 adults with painful knee osteoarthritis, without diabetes; placebo | 68 and 80 weeks | In both trials, serious adverse events were more common with retatrutide than with placebo, and gastrointestinal events were the most common. |
Sub-studies are not listed, and the table holds what the abstracts report (read 18 November 2026).
Both halves matter. In every abstract we read that ranks adverse events, gastrointestinal events came first, and where an abstract compares them with placebo they were more frequent with retatrutide. Some events that weigh more were also more frequent: low blood pressure and dysaesthesia in TRIUMPH-2, serious adverse events in the two knee trials. The TRIUMPH-2 authors, several of them Lilly employees, described its safety profile as “generally consistent with other molecules with GLP-1 receptor agonism”. That is the authors’ comparison, not a regulator’s finding.
These figures describe Lilly’s product, given to screened volunteers under trial monitoring, for 12 to 80 weeks. Trials of this size and length can show common, early problems. They cannot rule out uncommon or delayed ones (why: are peptides safe?). What the trials measured beyond safety: retatrutide research guide.
03What have regulators said about retatrutide’s safety?
Retatrutide is not approved anywhere. Here is what regulators in the US, the UK and Australia have said about it.
United States. FDA’s page on unapproved GLP-1 drugs (content current 1 October 2026) says retatrutide and cagrilintide “cannot be used in compounding under federal law” and “are not components of FDA-approved drugs and have not been found safe and effective for any condition”. It adds that FDA has warned companies that “illegally sold unapproved drugs containing semaglutide, tirzepatide, retatrutide, survodutide or mazdutide that are falsely labeled ‘for research purposes’ or ‘not for human consumption’”, and calls those products “of unknown quality” and possibly “harmful to their health”.
United Kingdom. The MHRA said on 15 October 2026 that “Retatrutide has not been authorised for use in the UK” and that products sold as retatrutide outside legitimate clinical trials “are being supplied illegally and could pose serious risks to public health.” More: are peptides legal in the UK?
Australia. On 14 September 2026 the TGA advised that “‘retatrutide’ products marketed for weight loss may pose a serious risk to your health”, after the serious adverse effect described below. It wrote on 5 November 2026 that retatrutide “is still undergoing clinical trials in humans” and that any product obtained outside a clinical trial “could be counterfeit”. The same advisory warns that peptide products, alone or with other medicines, “can cause serious side effects”. More: are peptides legal in Australia?
04What adverse-event reports exist for retatrutide?
FDA’s database. On 18 November 2026 we searched FDA’s public adverse-event data (openFDA, last updated 28 October 2026) for retatrutide. Our count: 83 reports name it. In 74 it is listed as a suspect or interacting product, and all 74 are spontaneous reports, not reports from a study. Of those 74, 45 are marked serious, 23 record a hospital admission and 2 record a death. Both list cardiac arrest; one also names kava as a suspect product. The most frequent reaction terms were vomiting (13 reports) and nausea (11). Some reports also carry terms for a product problem, such as “product quality issue”, “counterfeit product administered” or “product dispensed by unauthorised provider”. In the other 9 of the 83, retatrutide is listed only as a product taken alongside the suspect one.
None of this gives a rate or a cause. FDA says that in these reports “a causal relationship cannot be established between product and reactions”, that there is “no certainty that the reported event … was actually due to the product”, that FDA “does not receive reports for every adverse event”, and that the data “cannot be used to estimate the incidence of these events”. Which product is the suspect is the reporter’s view, and FDA notes “that information is not validated”.
A regulator’s case. On 14 September 2026 the TGA reported “a serious adverse effect linked to a product sold as retatrutide”: the person “experienced uncontrollable vomiting, resulting in a torn oesophagus, and required treatment in hospital.” TGA testing found that the product, in an unmarked vial, “contained undeclared semaglutide and no retatrutide”, at about eight times the level in the semaglutide products on the Australian register. Semaglutide is a prescription GLP-1 medicine.
Published case reports. We found two case reports of reactions in PubMed (18 November 2026), both about products obtained outside the trials. In Zurich, doctors admitted a woman in her mid-20s with persistent nausea and vomiting after a series of injections of an unverified product reported to contain retatrutide, given in a wellness setting by a person without medical qualification; her symptoms resolved within a week. The authors rated the reaction “probable”, but wrote that because the product “was not analytically verified”, it “cannot be determined whether retatrutide itself, an inaccurate dose, a contaminant or another active substance was responsible.” In Edinburgh, a man in his mid-30s with type 1 diabetes was admitted with severe vomiting, diarrhoea, raised ketones and acute kidney injury shortly after using a product bought online and marketed as retatrutide. A Shigella infection was confirmed later, the product was not analysed, and the author wrote that “causality cannot be established”.
Had a reaction to any product? Seek medical care first, then report it: how to report a reaction in the US, UK or Australia. How to weigh reports like these: peptide side effects, documented and unknown.
05Why are products sold online as retatrutide not the trial drug?
Every trial above tested Lilly’s product, made and checked for those trials. Material sold online under the same name has not been through them, and what it contains has to be shown sample by sample.
FDA. Warning letters dated 24 August 2026 to four online sellers listed their retatrutide products, labelled for research use and sold under names including “GLP-3 RT” and “GLP-3R”, among unapproved new drugs.
TGA. In November 2026 the TGA’s laboratory tested products from one brand sold as retatrutide and found they “contained less retatrutide than stated on the label, with significant variation between samples”.
Independent testing. In a 2026 letter, University of Queensland researchers reported laboratory results for three products sold as retatrutide in Australia. All three held retatrutide, identified by molecular weight. The measured amount ranged from about half to almost double the amount on the label: 51%, 165% and 190%. Sterility and endotoxin were not tested. One co-author leads the drug-checking service through which the samples came. Three samples cannot describe a market. The TGA, testing one brand, also found labels that did not match the contents.
In the TGA’s September case, the product held no retatrutide at all. Other documented cases: contamination and mislabeling. What a research-use label does and does not mean: research use only, explained.
06What we don’t know about retatrutide side effects
- Uncommon or delayed effects. The longest published trials ran 80 weeks. TRIUMPH-Outcomes, planned for about 10,000 adults with obesity, is listed as active, not recruiting, with primary completion expected in February 2029 (NCT06383390).
- The full safety results of TRIUMPH-3, in 1,946 adults with obesity and cardiovascular disease. ClinicalTrials.gov lists it as completed (NCT05882045), with no posted results, and we found no journal paper (PubMed, 18 November 2026).
- The detailed adverse-event tables of the 2026 papers. We read the abstracts; several full texts are paywalled.
- Results from other sponsors. A phase 2 trial of retatrutide sponsored by Hudson Biotech, in China (NCT07467447), is recruiting, with no results posted (ClinicalTrials.gov, 18 November 2026).
- How often reactions happen outside trials. Spontaneous reports cannot give a rate, and nobody knows how many people use products sold online.
- What any given vial sold online contains: retatrutide, how much of it, or something else.
Suspected reaction to any product? Seek medical care first, then report it. For anything about your health, a doctor who knows you is the right call.
This is a plain-English summary of public documents and studies, checked on 18 Nov 2026. It is not medical advice.