Evidence boundary: The human findings below belong to Lilly-sponsored trials of investigated LY3437943 in defined participant populations. They are study results, not claims about commercially available research material.
Trials of Lilly’s investigated LY3437943 have reported some of the largest mean body-weight changes in randomized metabolic research. These results are a major reason for scientific interest in the triple-agonist strategy, and understanding the trial designs helps place their magnitude in context.
Phase 2: the strongest peer-reviewed obesity evidence
The Phase 2 obesity trial randomized adults with obesity or overweight plus a weight-related condition, without type 2 diabetes. At 24 weeks, every tested dose met the prespecified primary weight endpoint.
By week 48, mean weight changes were:
| Group | Mean weight change |
|---|---|
| Lower dose | −8.7% |
| Intermediate dose | −17.1% |
| Higher dose | −22.8% |
| Highest dose | −24.2% |
| Placebo | −2.1% |
These peer-reviewed randomized findings established a clear dose-response pattern. Mean weight was still declining at week 48, leaving the eventual plateau and longer-term trajectory as open questions for later trials.
Phase 3: larger results, less complete publication
TRIUMPH-1 was the largest completed obesity trial available at the cutoff. A 2026 conference presentation reported week-80 efficacy-estimand mean changes of −19.0%, −25.9%, and −28.3%, compared with −2.2% for placebo.
Those figures came from a large randomized trial whose primary endpoint was reported met. They substantially strengthened the Phase 2 signal, although the available conference presentation did not yet provide the completeness of a full peer-reviewed report.
Where does “30%” come from?
“Thirty percent weight loss” can refer to different things:
- The proportion of participants who crossed a 30% threshold
- A selected extension group’s mean change
- An efficacy estimand that models outcomes as if participants remained on treatment
- An individual result rather than the trial average
These meanings are not interchangeable. The principal week-80 TRIUMPH-1 mean at the highest reported dose was below 30%, even though some participants lost 30% or more.
Why estimands change the number
An efficacy estimand estimates what might occur if participants remained on assigned treatment without certain intercurrent events. A treatment-regimen estimand includes outcomes regardless of discontinuation or prohibited additional treatment under its specified missing-data strategy.
The efficacy estimate is generally larger. Neither is automatically deceptive; they answer different questions. Problems arise when a report presents the larger number without explaining which question it answers.
Does this prove superiority?
Retatrutide’s reported numbers are numerically larger than some results from separate tirzepatide and semaglutide trials, making the comparison scientifically compelling. They do not yet establish superiority because differences in trial design, population, duration, adherence, estimand, and missing-data handling can materially change the reported mean.
The direct retatrutide-versus-tirzepatide trial had not reported by the cutoff.
Bottom line
Investigators reported mean body-weight changes up to −24.2% in a peer-reviewed Phase 2 obesity trial and up to −28.3% in Phase 3 conference data. These are notable research findings under the cited study conditions. The “30%” description is most accurate when tied to responder thresholds or selected analyses, while direct comparative interpretation awaits the head-to-head program.