Retatrutide produced substantial average weight loss in adults with type 2 diabetes in TRIUMPH-2, a phase 3 trial published on 29 September 2026. But readers may encounter two different headline results: 18.8% in the paper’s primary analysis and 20.8% in Lilly’s announcement. Both refer to the highest studied dose at 80 weeks. They answer different questions.
The distinction matters because a treatment result depends on how researchers handle stopping treatment, other medicines and missing measurements. Understanding those choices makes the news more useful than choosing the larger percentage.
Who took part in TRIUMPH-2?
The published trial abstract describes 1,152 adults with type 2 diabetes and a body mass index of at least 27. Participants had stable diabetes treatment before entering the study; most already used oral glucose-lowering medicines. This was a diabetes population, an important distinction from TRIUMPH-1, which studied adults without diabetes.
Participants were randomly assigned to placebo or weekly retatrutide injections in three research groups. The main weight comparison assessed the 9 mg and 12 mg groups after 80 weeks; the 4 mg comparison was a key secondary endpoint. These are studied regimens, not instructions for using an investigational drug. The NCT05929079 registry identifies the study as completed.
Two analyses, with their own placebo results
An estimand specifies the treatment effect a trial intends to estimate. The paper reports a treatment-regimen analysis including all randomized participants, using multiple imputation to handle missing data. Its 12 mg result was an average 18.8% loss, alongside 5.1% with placebo.
Lilly’s release instead highlights the efficacy estimand: an estimate under a hypothetical scenario in which everyone remained on the assigned intervention, allowing interruptions or dose adjustments, without prohibited weight-management treatment. That analysis gives 20.8% for the 12 mg group and 4.0% for placebo. It is not simply an average of participants who finished treatment.
| Randomized research group | Treatment-regimen analysis: paper | Efficacy analysis: sponsor release |
|---|---|---|
| Retatrutide 4 mg | 11.9% | 12.7% |
| Retatrutide 9 mg | 16.8% | 19.1% |
| Retatrutide 12 mg | 18.8% | 20.8% |
| Placebo | 5.1% | 4.0% |
Read each column as a complete comparison. Pairing 20.8% with 5.1%, or 18.8% with 4.0%, mixes analyses and produces a comparison neither source reports.
What those percentages can tell you
The trial provides evidence that retatrutide reduced weight more than placebo in this population. The size of the reported average depends on the question being asked and the assumptions used to answer it. Neither percentage predicts an individual’s outcome.
The ICH guidance on estimands distinguishes a treatment-policy question, which retains outcomes regardless of specified events such as treatment discontinuation, from a hypothetical question about what would happen under a defined alternative scenario. Neither approach makes missing measurements disappear. Readers should still ask what was collected and how unavailable outcomes were estimated; our guide to weight-loss trial estimands explains these choices.
TRIUMPH-2 also compared retatrutide with placebo, rather than another active weight-management drug. Comparing its headline with a separate trial cannot establish which medicine works better. Population, follow-up and analysis choices can differ. Those are core questions when reading weight-loss trials.
Tolerability belongs beside the weight result
The abstract reports that gastrointestinal events, particularly diarrhea and nausea, were more common with retatrutide than placebo. Low blood pressure and dysesthesia, an altered or unpleasant sensation, were also more frequent. These findings deserve attention alongside the average weight reduction.
Treatment discontinuation needs careful wording. The paper’s abstract reports permanent discontinuation because of adverse events or death; the sponsor release describes discontinuation because of adverse events. Those labels should not be treated as interchangeable. We have not reconstructed a detailed safety table from differently defined summaries.
The accessible abstract cannot settle every safety question. We could not access the full paper and have not reviewed its supplements, detailed event adjudication or subgroup analyses. The weight findings therefore should not be presented as proof of long-term safety for everyone who might seek treatment.
A trial result does not create an approved product
As checked on 4 October 2026, retatrutide remains investigational in the United States. The FDA’s current warning says retatrutide is not a component of an FDA-approved drug and has not been found safe and effective for any condition. A completed phase 3 study is a research milestone, separate from a regulatory decision.
The trial results do not establish the safety or effectiveness of products offered online as “research peptides”. Our research-peptide safety guide covers that distinction. For the broader development picture, see the retatrutide evidence guide.
