Retatrutide: what the studies say, 2026 trial results

What it is

Retatrutide is an experimental peptide developed by Eli Lilly, coded LY3437943 in early papers. A peptide is a short chain of amino acids, the building blocks of proteins. Retatrutide was designed to act on three receptors at once. A receptor is a docking point on the surface of a cell, and when a matching molecule attaches to it the cell changes its behaviour. These three receptors normally answer to the hormones GLP-1, GIP and glucagon, which help regulate appetite, insulin and energy use. A molecule that switches a receptor on is called an agonist, so retatrutide is described as a triple agonist (Jastreboff 2023). In the trials it was injected under the skin once a week. It remains investigational. No regulator in the EU, the UK or the United States has approved it as a medicine, and the company has said it plans a US application in the first quarter of 2027 (Lilly, July 2026).

What the studies looked at

All the trials below were randomised, double-blind and placebo-controlled. Participants were assigned by chance to retatrutide or to a placebo, an injection with no active ingredient, and neither they nor the study staff knew who received which. A phase 2 trial tests several doses in a few hundred people. A phase 3 trial tests the chosen doses in thousands and supports a regulatory application.

• Phase 2, obesity (Jastreboff 2023): 338 adults with a body-mass index (BMI) of 30 or more, or 27 to 30 with a weight-related condition. 51.8 % were men. Groups received 1 mg, 4 mg, 8 mg or 12 mg once weekly for 48 weeks, some starting at 2 mg or 4 mg and stepping up. A substudy followed 98 of them who had at least 10 % liver fat on an MRI scan (Sanyal 2024).
• Phase 2, type 2 diabetes (Rosenstock 2023): 281 adults with type 2 diabetes, HbA1c of 7.0 to 10.5 % and BMI of 25 to 50, at 42 centres in the USA, for 36 weeks. HbA1c is a blood test that reflects average blood sugar over about three months. Groups received 0.5 mg, 4 mg, 8 mg or 12 mg weekly, placebo, or dulaglutide 1.5 mg, an approved weekly diabetes medicine used as a comparison.
• TRIUMPH phase 3 programme (Giblin 2026): four trials with over 5,800 participants, all using 4 mg, 9 mg or 12 mg weekly. TRIUMPH-1 followed 2,339 adults with obesity and no diabetes for 80 weeks, including 574 with knee osteoarthritis (Jastreboff 2026). TRIUMPH-2 followed 1,152 adults with obesity and type 2 diabetes in eight countries for 80 weeks (Bellido 2026). TRIUMPH-3 enrolled 1,946 adults with obesity and cardiovascular disease (NCT05882045) and TRIUMPH-4 enrolled 445 adults with obesity and knee osteoarthritis (NCT05931367). A further trial, TRIUMPH-Outcomes, plans to follow about 10,000 adults for heart and kidney events until 2029 (NCT06383390).

What the studies say

Weight, phase 2. At 48 weeks the mean change in body weight was −8.7 % with 1 mg, −17.1 % with 4 mg, −22.8 % with 8 mg and −24.2 % with 12 mg, against −2.1 % with placebo. By then every participant on 8 mg or 12 mg had lost at least 5 % of their starting weight, and 83 % of the 12 mg group had lost 15 % or more, against 2 % on placebo (Jastreboff 2023).

Type 2 diabetes, phase 2. At 24 weeks HbA1c fell by 0.43 percentage points with 0.5 mg and by up to 2.02 points with 12 mg, against 0.01 with placebo and 1.41 with dulaglutide. At 36 weeks weight had fallen by 3.19 % with 0.5 mg and by up to 16.94 % with 12 mg (Rosenstock 2023).

Liver fat. At 24 weeks liver fat had fallen by 42.9 % (1 mg), 57.0 % (4 mg), 81.4 % (8 mg) and 82.4 % (12 mg), while it rose by 0.3 % with placebo, and normal liver fat, under 5 %, was reached by 27 %, 52 %, 79 % and 86 % of those groups and by nobody on placebo (Sanyal 2024).

TRIUMPH-1. At 80 weeks the mean weight change was −17.6 % with 4 mg, −23.7 % with 9 mg and −25.0 % with 12 mg, against −3.9 % with placebo. These figures count everyone randomised, including people who stopped injecting. Knee pain scores on a 1-to-10 scale fell by 3.2, 3.5 and 3.6 points against 1.9 (Jastreboff 2026). The company's announcement gave larger weight figures, −19.0 %, −25.9 % and −28.3 % against −2.2 %, calculated only for people who stayed on treatment (Lilly, May 2026).

TRIUMPH-2. In adults with type 2 diabetes the mean weight change at 80 weeks was −11.9 % with 4 mg, −16.8 % with 9 mg and −18.8 % with 12 mg, against −5.1 % with placebo (Bellido 2026). The company reported HbA1c reductions of 1.4, 1.6 and 1.5 points from a starting value of 7.7 %, against 0.2 (Lilly, July 2026).

TRIUMPH-3 and TRIUMPH-4. These two trials exist so far only as company announcements, not peer-reviewed papers. The company reported −22.6 % with 12 mg at 80 weeks in adults with cardiovascular disease, against −3.2 % (Lilly, July 2026), and −28.7 % with 12 mg at 68 weeks in adults with knee osteoarthritis, against −2.1 % (Lilly, December 2025).

What the researchers flagged

• Digestive effects. In every trial the most common adverse events were gastrointestinal: nausea, diarrhoea, constipation and vomiting. In phase 2 they were dose-related, mostly mild to moderate, and partly reduced by starting at 2 mg rather than 4 mg (Jastreboff 2023). In TRIUMPH-2 diarrhoea affected 27 %, 34 % and 34 % of the 4 mg, 9 mg and 12 mg groups against 13 % on placebo, and nausea 14 %, 21 % and 28 % against 8 % (Bellido 2026).
• Heart rate. In phase 2 heart rate rose with the dose, peaked at 24 weeks and then declined (Jastreboff 2023).
• Blood pressure and skin sensations. In TRIUMPH-2 low blood pressure occurred in 1 %, 5 % and 6 % of the retatrutide groups against under 1 % on placebo, and dysesthesia, an abnormal skin sensation such as tingling or burning, in 4 %, 6 % and 7 % against 1 % (Bellido 2026). In TRIUMPH-4 the company reported dysesthesia in 8.8 % and 20.9 % of the 9 mg and 12 mg groups against 0.7 %, and discontinuation for adverse events in 12.2 % and 18.2 % against 4.0 % (Lilly, December 2025).
• Stopping. In TRIUMPH-2, 4 %, 12 % and 8 % of the retatrutide groups stopped permanently because of adverse events or death, against 5 % on placebo. Seven participants died, six on retatrutide and one on placebo, and the investigators judged none of the deaths related to the treatment (Bellido 2026).
• Limits. The phase 2 diabetes trial was run in the USA only (Rosenstock 2023). The longest published follow-up is 80 weeks, so effects after stopping, and over many years, are not yet known. The outcomes trial is not due before 2029 (NCT06383390), and the full TRIUMPH-3 and TRIUMPH-4 papers have not appeared.