Retatrutide TRIUMPH-1: the phase 3 results

What TRIUMPH-1 found, in one paragraph

TRIUMPH-1 randomized 2,339 adults with obesity or overweight, without diabetes, to 80 weeks of retatrutide 4 mg, 9 mg, or 12 mg, or placebo. On the efficacy estimand, mean weight loss was 19.0%, 25.9%, and 28.3% against 2.2% on placebo. It was not the program's first readout: TRIUMPH-4 reported in December 2025.

Retatrutide is the most-watched obesity compound in late-stage development, and on May 21, 2026 Eli Lilly released topline results from TRIUMPH-1, a pivotal phase 3 obesity trial [1]. It was not the program's first readout. TRIUMPH-4, in 445 adults with obesity or overweight and knee osteoarthritis, completed on November 14, 2025 [10]. It reported on December 11, 2025 [9]. The TRIUMPH-1 headline number (up to 28.3% mean weight loss, on the efficacy estimand) is real and it is large [1]. The trial has since been published in the New England Journal of Medicine, which leads with the stricter treatment-regimen figure, 25.0% at 12 mg against 3.9% on placebo [15], and there is important context between any headline and what the data actually mean for a real person [8]. This post walks the numbers, the dose-by-dose detail, the side-effect profile, and the regulatory reality.

The short version: retatrutide's 12 milligram arms reached 28.3% in TRIUMPH-1 and 28.7% in TRIUMPH-4, both on the efficacy estimand [1] [9], and the published TRIUMPH-1 paper reports 25.0% on the stricter treatment-regimen estimand [15]. The safety signals look like a steeper version of the familiar GLP-1 class profile. And the drug is still investigational, with Eli Lilly saying in July 2026 that it plans a United States filing in the first quarter of 2027 [11]. It cannot be prescribed as an approved medicine today; outside clinical trials, Lilly runs a pre-approval expanded access programme open to a narrow group of patients with severe obesity [14].

The dose-by-dose weight loss numbers

At 80 weeks, on the efficacy estimand, retatrutide produced mean weight reductions of 19.0% (47.2 lb) at 4 mg, 25.9% (64.4 lb) at 9 mg, and 28.3% (70.3 lb) at 12 mg, against 2.2% for placebo. On the treatment-regimen estimand the same arms lost 17.6%, 23.7%, and 25.0%, against 3.9% for placebo.

TRIUMPH-1 (trial identifier NCT05929066) is a randomized, double-blind, parallel-arm phase 3 trial with 2,335 participants enrolled according to its registry entry [2]; Lilly's release describes 2,339 participants randomized 1:1:1:1 across three retatrutide doses and placebo [1]. Dosing started at 2 mg once weekly and escalated every four weeks to the assigned target [3]. The primary analysis ran to 80 weeks, and the per-arm figures in the table below are the release's efficacy estimand figures, which describe the effect had everyone stayed on the drug [1]. A subset of participants with a baseline body mass index of 35 or higher continued into an extension to 104 weeks [3].

The same release also reports the treatment-regimen estimand, which counts every randomized participant whether or not they stayed on the drug, and those figures are lower: 17.6% at 4 mg, 23.7% at 9 mg, and 25.0% at 12 mg, against 3.9% on placebo [1] [3]. Neither number is the true one. They answer two different questions, and which one belongs in a given sentence depends on whether you are asking what the drug does or what happens to a group of people prescribed it.

Two secondary thresholds from the same release are worth pulling out [1]. At the 12 mg dose, 45.3% of participants lost 30% or more of their body weight, a magnitude historically associated with bariatric surgery rather than medication [1]. And 65.3% of the 12 mg group reached a body mass index below 30, meaning they were no longer in the obesity category by the end of the trial [3]. The extension data are notable too: the higher-BMI group at 12 mg was still losing weight past 80 weeks and reached a mean of 30.3% (about 85.0 lb) at 104 weeks, suggesting the curve had not fully flattened for everyone at the standard endpoint [3].

For scale, this is the comparison the field has been waiting on, and it only holds up if the estimands match. The GLP-1 comparison tool puts retatrutide next to semaglutide and tirzepatide; semaglutide produced roughly 15% at 68 weeks in the STEP program [4], and tirzepatide 20.9% at the 15 mg dose at 72 weeks in SURMOUNT-1, under that trial's treatment-regimen estimand [5]. The retatrutide figure that sits on the same footing as SURMOUNT-1's is therefore not the headline 28.3% but the treatment-regimen 25.0% at 12 mg [1]. That is still the largest of the three, and it is the honest one to line up against them. Stacking 28.3% against 20.9% compares one drug under ideal adherence [1] with another measured across everyone randomized to it [5].

The cardiometabolic and secondary results

Beyond weight, the 12 mg group reduced waist circumference by 24.1 cm (9.5 inches) and improved non-HDL cholesterol, triglycerides, systolic blood pressure, and high-sensitivity C-reactive protein, on the efficacy estimand. Eli Lilly presented the detail at the American Diabetes Association meeting in June 2026. TRIUMPH-2 in diabetes and TRIUMPH-3 in cardiovascular disease reported in July 2026.

Weight is the headline, and TRIUMPH-1 also reported the downstream metabolic markers [3]. The company described improvements from baseline in waist circumference, non-HDL cholesterol, triglycerides, systolic blood pressure, and high-sensitivity C-reactive protein [1] [3]. The waist reduction in the 12 mg arm, reported as roughly 24.1 cm, is consistent with the large total-weight numbers [3].

A few caveats apply. The glucagon receptor component of retatrutide is the part that most differentiates it from tirzepatide, and glucagon agonism can raise hepatic glucose output and heart rate, so the full glycemic profile and any cardiovascular signal matter a great deal. The dedicated TRIUMPH-2 trial in type 2 diabetes and TRIUMPH-3 in severe obesity with established cardiovascular disease were designed to answer those questions, and both reported topline results on July 23, 2026: mean weight loss at 80 weeks of 20.8% at 12 mg in TRIUMPH-2, alongside an A1C reduction of up to 1.6%, and 22.6% at 12 mg in TRIUMPH-3, both on the efficacy estimand [11]. The TRIUMPH-2 paper, published in The Lancet in September 2026, reports 18.8% at 12 mg against 5.1% on placebo on the stricter treatment-regimen estimand [16].

TRIUMPH-3 is the one people reach for as cardiovascular evidence, and it is not that. Eli Lilly reported that major adverse cardiovascular events happened less often than the trial anticipated in both arms: 44 events on pooled retatrutide against 52 on placebo, a hazard ratio of 0.82 with a confidence interval running from 0.55 to 1.22 [11]. That interval crosses 1, which means the result is compatible with benefit, with no effect, and with modest harm. Lilly said the same release's on-treatment analysis was not pre-specified [11]. A dedicated cardiovascular outcomes trial is what would settle the question, and none has reported for retatrutide.

The side-effect profile

Adverse events were gastrointestinal and dose-related. At 12 mg, approximate rates were 42.4% nausea, 32.0% diarrhea, 26.1% constipation, and 25.3% vomiting, mostly mild to moderate. Dysesthesia (skin tingling) occurred in roughly 5% to 12.5% of retatrutide users. Discontinuation due to adverse events rose with dose to 11.3% at 12 mg, versus 4.9% on placebo.

The tolerability story is the familiar incretin pattern, scaled up. Nausea, diarrhea, constipation, and vomiting were the dominant adverse events, they were dose-related, and most cases were mild to moderate. At the top 12 mg dose the reported rates were about 42.4% nausea, 32.0% diarrhea, 26.1% constipation, and 25.3% vomiting, against single-digit-to-low-teens rates on placebo [3].

One effect worth knowing by name is dysesthesia, an altered or tingling skin sensation, reported in roughly 5% to 12.5% of retatrutide participants compared with under 1% on placebo [3]. It was generally mild to moderate [3]. The practical readout from the safety data is the discontinuation rate: about 4.1% at 4 mg, 6.9% at 9 mg, and 11.3% at 12 mg stopped because of adverse events, versus 4.9% on placebo [3]. That gradient is the tradeoff in plain terms: the dose that produces the most weight loss is also the dose more people cannot stay on [3].

Why retatrutide is still investigational

Retatrutide is not approved anywhere, and outside trials it reaches patients only through a narrow expanded access programme. TRIUMPH-1 and TRIUMPH-2 are now published, in the NEJM and The Lancet, as is TRANSCEND-T2D-1, but publication is not approval. On July 23, 2026 Eli Lilly said it plans to submit a Biologics License Application to the FDA in Q1 2027.

This is the part the excitement tends to skip. As of October 2026, retatrutide is an investigational molecule. Eli Lilly's own release describes it as investigational [1]: it holds no marketing approval for any indication, so it cannot be prescribed as a marketed medicine and there is no quality-assured product on the market. Outside trials, access is narrow: since June 2026 Eli Lilly has run a registered pre-approval expanded access programme for adults with a BMI of 35 or higher despite the highest available dose of an approved weight-management drug, two or more serious obesity-related complications, and no way into a trial [14]. Because it is investigational [1], nothing sold online as research retatrutide has passed any regulator's review, so its identity, purity, and actual dose are whatever the seller claims, and the trial safety data above describe a product a buyer cannot confirm they are getting.

There is also a difference between a topline announcement and the evidence base. Most numbers in this post first came from Eli Lilly's May 21, 2026 press release [1] [3], and the peer-reviewed paper (with full statistics, prespecified analyses, subgroup breakdowns, and independent review) is the document that the field ultimately judges a drug on. Detailed TRIUMPH-1 results reached the American Diabetes Association's 86th Scientific Sessions on June 6, 2026, alongside TRANSCEND-T2D-1, a phase 3 diabetes trial published in The Lancet the same day [12] [13]. The TRIUMPH-1 paper followed in the New England Journal of Medicine on September 29, 2026. It leads with the treatment-regimen estimand, 25.0% at 12 mg against 3.9% on placebo, and its abstract does not report the 28.3% efficacy-estimand figure from the press release [15].

The regulatory picture has also moved, and this is where the earlier version of this post is out of date. When TRIUMPH-1 read out, Eli Lilly had named no filing date. On July 23, 2026, announcing TRIUMPH-2 and TRIUMPH-3, it did: Lilly said it is completing the manufacturing and controls package required for a Biologics License Application and plans to submit retatrutide to the FDA in the first quarter of 2027 [11]. A stated filing plan is not a filing, a filing is not an approval, and approval, if it comes, follows regulatory review of the entire TRIUMPH program, not a single trial. For context on how the FDA handles approval timelines for this drug class, see the FDA peptide reclassification guide and the overview of FDA-approved peptides to understand which compounds in this space have already cleared regulatory review.

If you want the mechanism behind the numbers, how simultaneous GIP, GLP-1, and glucagon receptor activation produces this profile, and why the glucagon arm is the variable to watch, our retatrutide course walks through the receptor pharmacology and the clinical evidence in structured units. For how this slots in against the approved options and the off-ramp question that applies to every drug in this class, the tirzepatide muscle, plateau, and off-ramp guide covers the principles that transfer across compounds.

How to read trial headlines like this one

Mean trial figures hide the spread: some participants far exceed the average and some respond little or stop early. Headline percentages also come from the highest-dose arm under structured lifestyle support, on a selected trial population. Topline press releases precede peer review. Treat 28.3% as the top of a distribution under ideal conditions, not a number any individual should expect.

A 28.3% mean [1] is an average, and an average hides the spread. In any obesity trial, some participants lose far more than the mean, some lose much less, and some discontinue because of side effects and are handled by the statistical analysis rather than walking out with that result. The headline is also the top-dose arm, which carries the highest adverse-event burden, and trial participants receive structured diet and lifestyle support that real-world users may not. The CagriSema REDEFINE-1 trial analysis illustrates the same issue in a different compound: large headline numbers that require careful reading of the full trial design.

None of that diminishes the result. It contextualizes it. The first four TRIUMPH obesity trials have now reported: TRIUMPH-4 in December 2025 [9], TRIUMPH-1 in May 2026 [1], and TRIUMPH-2 and TRIUMPH-3 in July 2026 [11]. TRIUMPH-5, the head-to-head trial against tirzepatide, is still running, with primary completion estimated for November 2026 [17]. In TRIUMPH-1 the 12 mg arm lost 25.0% of body weight on the stricter treatment-regimen estimand [15], above the 20.9% tirzepatide reached at 15 mg in SURMOUNT-1 [5] and the 14.9% semaglutide reached in STEP 1 [4], though those are separate trials. The side effects are dose-related, and stopping because of them rises with dose [3]. Retatrutide remains an investigational compound with no marketing approval; outside trials, Lilly runs a narrow pre-approval expanded access programme [14]. That is the research, and the research is genuinely impressive on its own without the hype.