Claim check · Study interpretation, Effectiveness

When weight-loss results can be compared

Claim (representative wording, not a quotation): Tirzepatide caused more weight loss than semaglutide, because its trial showed a bigger percentage.

Plain-English answer

Percentages can be compared directly only when both drugs were randomized in the same trial, and WPF found that for one trial in this set (SURMOUNT-5). Two separate placebo-controlled trials differ in people, length and design, so subtracting them proves nothing.

Verdict: Overstated. Part of the claim is supported, but the claim says more than the sources show: a wider population, a stronger outcome, a human result from animal work, or a certainty the design cannot give.

What changed?

One change to this record since it was first published; this page shows version 1.6.0. Older versions still resolve.

  • · Source attribution corrected · Verdict unchanged

    SURMOUNT-5 confidence intervals: arm intervals come from the abstract, the between-group interval from the registry

    The correction changed source attribution only. Every estimate and every verdict stayed the same. The intervals were already correct; the records now say that the arm intervals come from the publication abstract and the between-group interval from the registry results, so a reader who goes to either source finds the figure where WPF says it is.

    Read this change and its sources · Earlier record (version 1.4.0) · Updated record (version 1.5.0) · Correction Log entry

    AI-assisted source verification. Documented human review: none recorded.

Scope

Population or organism
Adults with obesity in the five results of Flagship Weight Management 1.3.0
Study design
One head-to-head randomized trial, three separate placebo-controlled trials and one observational cohort
Boundary
Weight change as reported. Not safety, cost, tolerability or any individual's expected result.

Strongest evidence

  • SURMOUNT-5, result surmount-5-weight-72w (Flagship Weight Management 1.3.0) (WPF record)

    Population or organism: 751 adults with obesity, without type 2 diabetes

    Design: Phase 3b, open-label, randomized head-to-head, 72 weeks

    Where in the source: Point estimates: registry NCT05822830 results, primary outcome measure 1, posts -20.2 (tirzepatide) and -13.7 (semaglutide) as least-squares means, each with a standard error of 0.59, and a between-group least-squares mean difference of -6.5 (95% CI -8.1 to -4.9). The intervals -21.4 to -19.1 (tirzepatide's own mean) and -14.9 to -12.6 (semaglutide's own mean) and P<0.001 are from the NEJM abstract (PMID 40353578), not the registry; each belongs to one arm's mean, not to the difference between arms, and the abstract does not report the between-group difference.

  • STEP 1 and SURMOUNT-1 (separate placebo-controlled trials), results step-1-weight-68w and surmount-1-weight-72w (WPF record)

    Population or organism: 1,961 adults (STEP 1, 68 weeks) and 2,539 adults (SURMOUNT-1, 72 weeks)

    Design: Two separate double-blind placebo-controlled trials

    Where in the source: Each compares one drug with placebo; neither contains the other drug, so the two results are an indirect comparison.

  • Observational cohort, result jama-intern-med-2024-realworld (WPF record)

    Population or organism: 41,222 US adults; 18,386 after propensity-score matching

    Design: Observational cohort, not randomized, on-treatment follow-up to 12 months

    Where in the source: Hazard ratios 1.76, 2.54 and 3.24 for reaching 5%, 10% and 15% weight loss, checked against the open-access full text.

What remains uncertain

  • SURMOUNT-5 was open-label and its publication abstract does not name the estimand, so the match to the analysis plan was made by method.
  • Doses, populations and follow-up differ between the placebo-controlled trials, and WPF did not model those differences.
  • The cohort's differences may reflect factors that were not measured, because people were not randomized.

What would change the answer

  • A second randomized head-to-head trial in a different population, which would show whether the SURMOUNT-5 difference holds.
  • Network meta-analysis methods that adjust for trial differences, if applied and reported transparently.

Exact gaps

  • The four journal papers behind the trial rows returned HTTP 403 to automated requests, so their tables and figures were not located; registry results and analysis plans were read instead (see the Verify panel for each result).

Date checked and review status

Checked: · version 1.6.0

Review status: AI-assisted source verification. Human review: none recorded. Human review is pending and has not happened.

Profiles: /directory/semaglutide · /directory/tirzepatide · /directory/retatrutide

Films: /films/metabolic/06-semaglutide

Flagship results: /research/flagship/compare/weight-management#surmount-5-weight-72w · /research/flagship/compare/weight-management#step-1-weight-68w · /research/flagship/compare/weight-management#surmount-1-weight-72w

Research journeys: /research/journey/semaglutide

Teaching: /research/teaching/weight-management-comparison

Cite this check

The citation names the version and an address that keeps returning what was checked, even after a newer version is published.

When weight-loss results can be compared: Percentages can be compared directly only when both drugs were randomized in the same trial, and WPF found that for one trial in this set (SURMOUNT-5). Two separate placebo-controlled trials differ in people, length and design, so subtracting them proves nothing. Verdict: Overstated. Cited from World Peptide Foundation, Peptide Claims & Evidence, version 1.6.0, claim weight-loss-results-comparable, checked 2026-11-16. AI-assisted source verification; human review none recorded.

Page:
https://www.worldpeptidefoundation.org/research/check/weight-loss-results-comparable

Versioned record:
https://www.worldpeptidefoundation.org/api/research/claims/weight-loss-results-comparable?version=1.6.0

Versioned record: /api/research/claims/weight-loss-results-comparable?version=1.6.0