SURMOUNT-5: tirzepatide as compared with semaglutide for the treatment of obesity

Aronne LJ, Horn DB, le Roux CW, et al. · New England Journal of Medicine · May 2025 · NCT05822830

Study record

Results

Tirzepatide −20.2% (95% CI −21.4 to −19.1) versus semaglutide −13.7% (95% CI −14.9 to −12.6), p<0.001. Absolute: 22.8 kg versus 15.0 kg. At least 25% body-weight loss: 31.6% versus 16.1%. Tirzepatide was superior on the primary and all five key secondary endpoints.

Limitations

Open-label — participants and investigators knew which drug they were receiving, which can bias adherence and reporting. Funded by Eli Lilly, which manufactures the winning drug. Both caveats belong with the number every time it is quoted. Applies to approved subcutaneous injections.

What this study does not prove

This section exists on every study page we publish, because the most common way clinical evidence is misused is not by misquoting the result — it is by stretching the result past the population, dose, duration or dosage form that was actually tested.

Absolute versus relative: reading the number correctly

Trial results are usually reported as relative figures, because relative figures are larger and therefore more persuasive. A "20% reduction in cardiovascular events" sounds transformative. The absolute reduction in SELECT was from 8.0% to 6.5% — about 1.5 percentage points over roughly three years. Both statements describe the same result honestly; only one of them tells you what to expect for yourself.

The same applies to weight-loss figures. A mean reduction of 20.9% is a mean. Individual results in these trials ranged from substantial loss to none at all, and a mean tells you nothing about where you personally would land. Anyone quoting a trial average as a promise is misusing it.

Funding and conflicts of interest

Every pivotal trial in this field was funded by the company that manufactures the drug it tested. That is normal in pharmaceutical research and it does not make the results false — these are large, well-conducted, peer-reviewed studies. It does mean the funding belongs in the citation every time, particularly for head-to-head trials where the funder makes the winning drug. SURMOUNT-5 was funded by Eli Lilly and found Lilly's drug superior. The result is plausible and consistent with the separate trial programmes; the disclosure still belongs beside it.

Where this sits against the other evidence

No single trial should be read alone. The strength of the GLP-1 evidence base is that multiple independent trial programmes — SURMOUNT for tirzepatide, STEP for semaglutide, SCALE for liraglutide, SELECT for cardiovascular outcomes — point in a consistent direction across tens of thousands of participants. That consistency is what makes the class credible.

What that consistency does not do is extend to products the trials never tested. Every one of those programmes studied an FDA-approved subcutaneous injection. None studied a compounded preparation, a microdose regimen, or an orally disintegrating tablet. The evidence is strong exactly where it was collected and silent everywhere else, and the gap between those two things is where most of the marketing in this industry operates.

How to read a trial result without being misled

Three habits will protect you from most of the misuse of this literature.

Ask what was actually administered. Not the molecule — the product. Every trial on this page tested an FDA-approved subcutaneous injection. If you are being sold a compounded preparation, a microdose, or an orally disintegrating tablet, this trial is evidence about a related product, not about yours.

Ask who was studied. SURMOUNT-1 and STEP 1 enrolled people without type 2 diabetes and gave every arm a lifestyle intervention. SELECT enrolled people who already had cardiovascular disease. A result in one population is not automatically a result in another, and the further you are from the enrolled population, the less the number tells you.

Ask for the absolute figure. Relative reductions are larger and therefore more persuasive. The absolute figure is the one that tells you what to expect.

Relevance to patients

The trial tested an FDA-approved subcutaneous injection. If you are considering a compounded product, a microdose programme, or an orally disintegrating tablet, this trial is not evidence for what you are buying. It is evidence about a related but different product. That distinction is the single most important thing to carry away from any GLP-1 trial page, including this one.

Limitations of this analysis

Every page on this site should tell you where it stops being reliable. This one stops here.

Prices decay quickly. This is the fastest-moving data we publish. Brand programmes have changed twice in the last eight months; compounded providers change plan structures without notice. Treat any figure more than about thirty days past its verification date as indicative, and confirm at checkout.

Competitor pricing is reported, not captured by us. We hold dated captures for brand pricing and for NexLife. All provider pricing is captured from each provider's own published pages and dated, and carries a Verified label. Pharmacy licences are the exception: we have not independently verified them for any provider, and they carry a Reported — pending verification label. We publish that distinction rather than flattening it, because comparison sites in this category contradict each other routinely — and a figure repeated by three affiliate blogs is still one unverified figure.

We have not audited pharmacy licences. Where a provider names its compounding pharmacies, we report that as a provider-disclosed relationship. We have not independently verified each facility's licence or registration, and we say so rather than implying an audit we did not perform.

Advertised availability is not your availability. Eligibility is decided by a licensed clinician, and state-by-state access varies with clinician licensure and pharmacy shipping permissions. No page can promise you a price you will actually be offered.

We are commercially funded. The publisher and certain principals have financial relationships with some of the providers listed here, and we may earn a commission from provider links. That is disclosed in the footer of every page. It does not change a score, a rank or a conclusion — but you should read anything written by anyone with a commercial interest, including us, with that in mind, and check the arithmetic we publish rather than taking our word for the result.