For education only · Not medical advice
← All articles
September 2026 Weight & Metabolic 6 min read

What Does Retatrutide Do? Phase 2, Phase 3, and What Is Still Unknown

Retatrutide has some of the largest published trial effects in metabolic medicine and no regulatory approval in any country, which makes it one of the few compounds where the honest summary is genuinely split between strong data and complete absence of a product.

The short answer

  • Retatrutide (development code LY3437943) is a single lipidated peptide that activates three receptors at once: GIP, GLP-1 and glucagon. The glucagon arm is what distinguishes it from tirzepatide.
  • Evidence tier: Human trials. Large phase 2 and phase 3 randomised trials exist for obesity, type 2 diabetes, knee osteoarthritis pain with obesity, and liver fat.
  • It is not approved anywhere as of September 2026. The manufacturer has said it plans a US submission in the first quarter of 2027.
  • In Canada it is an unauthorized drug with no record in the Drug Product Database, and it was named in a Health Canada advisory of unauthorized injectable peptides seized from an online seller (1 August 2025).
  • Every human number below comes from the manufacturer’s product under trial conditions. None of it describes a vial bought online, and one published case report documents serious harm after an online-sourced product was self-injected.

What does retatrutide do at the receptor level

Three receptors, three jobs.

The GLP-1 and GIP components behave the way they do in tirzepatide: reduced food intake through central pathways, slowed gastric emptying, and glucose-dependent insulin secretion. In vitro the molecule’s agonism is roughly balanced at the glucagon and GLP-1 receptors and stronger at GIP (Coskun 2022).

The glucagon receptor component is the addition. Glucagon raises hepatic glucose output, but it also increases hepatic fat oxidation and raises energy expenditure, mostly shown in animal work. In obese mice, the glucagon component added weight loss on top of the reduction in intake (Coskun 2022). Glucagon agonism alone would push glucose up; the incretin arms offset that, and glycated haemoglobin falls substantially in the diabetes trials rather than rising.

A fatty-acid moiety binds the peptide to albumin, giving a half-life of about 6 days, which is what makes once-weekly injection workable (Urva 2022). Exact sequence and molecular weight are not stated in the verified primary sources, and supplier-listed values are not a substitute.

For the general mechanism shared across this class, see how peptides work in the body.

The phase 2 numbers

The 48-week phase 2 obesity trial is the peer-reviewed anchor. Mean weight change was -8.7% at 1 mg, -17.1% at 4 mg, -22.8% at 8 mg and -24.2% at 12 mg, against -2.1% on placebo (Jastreboff 2023).

In type 2 diabetes, phase 2 in the USA reported an A1c change of -2.02% at 24 weeks on 12 mg (Rosenstock 2023).

The liver-fat substudy is the most striking single result in the file. Relative liver fat fell 42.9% at 1 mg and 82.4% at 12 mg by 24 weeks, and 86% of participants on 12 mg reached normal liver fat compared with 0% on placebo (Sanyal 2024). Reductions of that size are consistent with a direct hepatic glucagon effect but do not prove one.

A DXA substudy found fat mass fell by up to roughly 26%, with the lean-mass share of total weight lost similar to other drugs in the category (Coskun 2025).

The phase 3 numbers, and why they need an asterisk

The TRIUMPH programme has reported topline results by press release. These figures have not been peer-reviewed as of September 2026, which is a real limitation and not a formality.

TRIUMPH-1 (n = 2,339, 80 weeks, topline 21 May 2026) reported -19.0%, -25.9% and -28.3% at 4, 9 and 12 mg against -2.2% on placebo using the efficacy estimand; the same trial reported -17.6%, -23.7% and -25.0% against -3.9% using the treatment-regimen estimand. An extension subset reached -30.3% at 104 weeks on 12 mg.

That gap between estimands matters. The efficacy estimand answers “what happens in people who stay on the drug as intended”, the treatment-regimen estimand answers “what happens to everyone randomised, including those who stop”. Quoting only the larger number is the most common way these results get distorted.

TRIUMPH-2 (n = 1,152, obesity with type 2 diabetes) reported -12.7% to -20.8% and A1c change up to -1.6 points. TRIUMPH-3 (n = 1,949, severe obesity with established cardiovascular disease) reported up to -22.6%. TRIUMPH-4 (n = 445, knee osteoarthritis pain with obesity) reported up to -28.7% weight change and a significant reduction in WOMAC pain score.

One phase 3 result is published rather than topline: TRANSCEND-T2D-1 (40 weeks, n = 537) reported A1c changes of -1.69%, -1.86% and -1.94%, with placebo-adjusted differences of -0.88 to -1.12, and weight changes of -11.5%, -13.9% and -15.3% against -2.6% (Bajaj 2026).

Doses studied, and the numbers that are not doses

Retatrutide has no label dose, because it has no label anywhere.

What the trials used: 0.5 to 12 mg subcutaneously once weekly in phase 1b over 12 weeks, with stepwise escalation (Urva 2022). Phase 2 obesity used 1, 4, 8 and 12 mg weekly over 48 weeks, with the higher arms beginning at 2 or 4 mg; the lower 2 mg start partially reduced gastrointestinal events (Jastreboff 2023). Phase 3 maintenance doses were 4 mg, 9 mg and 12 mg weekly, reached by escalation schedules that were not published in the sources reviewed.

Grey-market regimens circulating online loosely imitate the phase 2 arms, or describe very small doses for “recomposition”. These are unvalidated community conventions with no evidence behind them, and they are applied to products of unknown identity, purity and content. This site does not publish them.

Content uncertainty here is documented rather than hypothetical. An Australian analysis examined the composition and labelling accuracy of products sold as retatrutide (Piatkowski 2026), and a 2026 case report describes a man with type 1 diabetes who developed severe vomiting, ketonaemia at 4.3 mmol/L and acute kidney injury after injecting an online product, followed by recurrent hypoglycaemia (Branine 2026). No combination of retatrutide with semaglutide, tirzepatide, cagrilintide or a growth hormone secretagogue has any human safety data at all. For the approved end of this class, see GLP-1 and semaglutide.

What the side effect profile looks like

From the TRIUMPH-1 topline, at 4, 9 and 12 mg against placebo: nausea 28.6/38.4/42.4% versus 14.8%; diarrhoea 25.2/34.1/32.0% versus 13.5%; constipation 23.8/25.9/26.1% versus 10.9%; vomiting 10.6/22.8/25.3% versus 4.8%. Discontinuation for adverse events ran 4.1/6.9/11.3% versus 4.9%, and reached 18.2% on 12 mg in TRIUMPH-4.

Two signals are less familiar. Dysaesthesia, meaning abnormal skin sensation, occurred in 5.1/12.3/12.5% versus 0.9%, a signal also seen with high-dose semaglutide. Urinary tract infections ran 7.5 to 8.8% versus 5.3%, with the timing and mechanism still under discussion (Koufakis 2026). Heart rate also rises in a dose-dependent way, peaking around week 24 and then declining (Jastreboff 2023).

Because there is no label, there are no formally established contraindications. Trials typically excluded personal or family history of medullary thyroid carcinoma or MEN 2, prior pancreatitis, pregnancy and type 1 diabetes.

What is still unknown

  • Long-term outcomes. The cardiovascular and kidney outcome trial (TRIUMPH-Outcomes) is ongoing, as are a chronic kidney disease trial and sleep apnoea baskets.
  • Chronic glucagon receptor agonism. Years of glucagon signalling, and what it does to amino-acid metabolism, lean mass, heart rate and arrhythmia risk, is uncharacterized.
  • Head-to-head position. A phase 3 comparison against tirzepatide is ongoing. Until it reports, claims that retatrutide is simply stronger rest on cross-trial comparison.
  • Basic pharmacokinetics beyond half-life. Bioavailability, volume of distribution and clearance are not reported in the verified abstracts.
  • Anything about grey-market product. Identity, purity and content are unverified by definition.

For how the approved members of this class compare with each other, see semaglutide vs tirzepatide vs retatrutide.

The bottom line

Retatrutide sits at Human trials on the strength of large, well-designed randomised studies, and it is unapproved in every jurisdiction as of September 2026. Its phase 2 obesity and liver-fat results are peer-reviewed; most of its phase 3 numbers are company toplines that have not yet been through review, and the two estimands in TRIUMPH-1 differ by more than three percentage points. Anything sold online under this name is a different object from the compound in those trials, and the published harm report is about the former, not the latter.

Education only. This page is educational. Most compounds referenced are not approved for human use, and a “research use only” label carries no legal status of its own. Nothing here constitutes medical advice, diagnosis, or treatment.

Sources

Peptides 101RetatrutideIncretins
For education only · Not medical advice