Start with the study behind the headline
The published abstract identifies SURMOUNT-5 as a randomized, open-label phase 3b comparison involving 751 participants. Its primary outcome was the percentage change in body weight at 72 weeks. This is a direct comparison within one trial, which addresses a different question from placing results from unrelated studies side by side.
The trial registry specifies adults with obesity, or overweight and a qualifying weight-related condition, and excludes diabetes. It names Eli Lilly and Company as the sponsor. Those details belong alongside the result: readers need to know who was studied, how the trial was organized and who funded it.
20.2% versus 13.7%: read both kinds of difference
The abstract reports mean weight changes of −20.2% with tirzepatide and −13.7% with semaglutide. Subtracting the magnitudes gives 6.5 percentage points. Dividing that gap by 13.7 gives approximately 47%, the relative comparison used in the sponsor announcement. It does not mean that participants lost 47% of their starting weight.
These calculations describe the reported averages; they are not a forecast for a reader. An individual might have a different response or be unable to continue treatment. Our guide to weight-loss estimates explains why the time point, population and analysis approach need to travel with the number.
A strong comparison still has a defined scope
The registry records random assignment and no masking. Randomization helps make the groups comparable at the start; open-label means treatment assignment was not concealed from participants and investigators. That design feature is worth knowing without treating it as a reason to dismiss every result.
The sponsor’s trial description says both groups received diet and physical-activity counseling and used maximum tolerated study doses. It identifies the headline analysis as the treatment-regimen estimand. We would not reinterpret those averages as results from the starting dose, from unsupervised treatment or from buying medicine without the trial’s care arrangements.
A weight result is not a complete safety ranking
The abstract reports that gastrointestinal events were the most common adverse events in both groups; most were mild to moderate and occurred mainly during dose escalation. The sponsor announcement explicitly says the study was not powered to compare the treatments’ safety and tolerability.
That matters when a headline turns “greater average loss” into “better for everyone.” Treatment choice also involves the individual’s medical history, other medicines, preferences, access and ability to continue. This article does not replace current prescribing information or a clinical discussion, and it does not provide a dose or switching plan.
What this result cannot establish
The eligibility criteria do not support automatically extending the finding to children or people with diabetes. Nor does a trial of these study regimens measure the quality of a particular telehealth clinic. The FDA explains that compounded GLP-1 products are not FDA approved and are not reviewed for safety, effectiveness and quality before marketing.
A recommendation displayed on this website is therefore separate from the study finding. A shared ingredient name does not demonstrate that another preparation or service has reproduced the trial. When reading a provider’s claim, look for evidence about the actual product and arrangement being offered.
How we reviewed the evidence
We checked the journal abstract through Europe PMC, the registry, the sponsor’s May 11, 2025 announcement and FDA information on September 26, 2026. We could not access the full journal article and supplementary appendix, so this is a review of those accessible records, not a reanalysis of participant data or a full methods appraisal.
Use our clinical-trial registry review to follow the study record and the news-headline checklist when evaluating a new claim. The useful conclusion is specific: this trial favored tirzepatide for its measured weight outcome at 72 weeks in the studied population.
Common questions
Did participants lose 47% of their body weight?
No. The reported mean reductions were 20.2% and 13.7%. About 47% describes the relative difference between those averages.
Does this review establish which medicine is right for me?
No. It explains a study result and its limits. Individual treatment decisions require clinical assessment and current product information.
Sources and fact-checking
Sources checked 2026-09-26. Company pages document their own published terms; government and research sources provide context. This is a review of published information, not a hands-on service test or a clinical evaluation. Prices and availability can change.