SURMOUNT-5 supplies a direct comparison that is more informative than placing results from separate trials side by side. In the report published online on 11 May 2025, average weight reduction favored tirzepatide over semaglutide. The value of that finding depends on understanding the question the trial actually asked. This review uses the original published abstract and the public ClinicalTrials.gov record, checked on 3 October 2026. It does not claim a review of the journal’s full supplementary analyses.
The core comparison in one place
The original SURMOUNT-5 report describes 751 participants randomized to a 72-week, open-label comparison. The study used maximum-tolerated weekly injection regimens. Reported mean weight change was −20.2% with tirzepatide and −13.7% with semaglutide. Gastrointestinal events were the most common adverse events in both groups.
That result supports a comparison between the studied treatment strategies. It should be quoted with the time point and population, rather than reduced to a universal ranking of every possible way to use either molecule. Treatment choice requires a wider assessment than one average weight endpoint.
Who was eligible shapes the application
The registry describes adults with obesity, or overweight and a qualifying weight-related condition. Diabetes was excluded. Other eligibility restrictions also selected the study population.
If your question concerns people with diabetes, children, a different clinical history or another formulation, the match is incomplete. That is a reason to look for relevant evidence, not a reason to discard this trial. Ask how closely the people studied resemble the population you are asking about. An important result can be directly useful for one question and indirect for another.
Randomization and open labeling are separate features
The registry documents randomized allocation with no masking. Randomization helps create comparable treatment groups at the start. Open labeling means the assigned treatment is known, leaving potential influences from expectations and behavior.
A measured body-weight outcome is different from a subjective impression, but an objective measurement does not remove every possible influence on a trial. Conversely, lack of masking does not erase randomization. A balanced appraisal identifies each design feature and the problem it addresses. The useful question is how that feature affects the particular outcome and interpretation.
Compare percentage points with relative language carefully
Subtracting the reported average reductions gives a difference of 6.5 percentage points. That is an arithmetic comparison of the two percentages, not a prediction that a person will lose another 6.5 kilograms. Kilograms depend on starting weight and the person’s actual response.
A relative comparison uses a different denominator and can sound much larger. When a headline says one treatment produced a certain percentage more weight loss, find the underlying averages. Keep both the unit and reference quantity visible. This makes it possible to compare descriptions without mistaking different presentations of the same result for different findings.
An average is different from a responder threshold
The registry lists weight-reduction thresholds among the secondary outcomes. An average asks about the group’s mean change; a threshold asks what proportion reached a defined amount. They describe different aspects of response.
Do not turn a group mean into the proportion who will obtain that mean. Likewise, crossing a threshold does not describe every other outcome or adverse experience. Our trial-reading guide explains how population, comparator, time and outcome work together. For personal expectations, ask the clinician how uncertainty and individual response will be reviewed.
Check the analysis question before combining numbers
The public registry also lists an outcome based on the assumption that participants remained on treatment and reached the highest dose. A result with that assumption answers a different question from an analysis intended to reflect a broader treatment strategy.
Before combining estimates from a table, identify how treatment discontinuation, switching and missing measurements were handled. This review does not supply a detailed audit of the paper’s methods or identify every numerical result by estimand. Our estimand explainer provides background. The practical lesson is to keep each number attached to its analysis question.
Weight and waist changes leave other questions open
The comparison assessed weight and waist outcomes at the stated time point. Those endpoints should not silently become claims about direct superiority for cardiovascular events, lifespan or every obesity-related condition.
When a discussion moves to a different health outcome, ask for its evidence separately. A biological rationale or improvement in an intermediate measure can motivate further research without establishing that clinical endpoint. Similarly, a trial ending at a particular week cannot tell you the whole course of treatment over many years. Time and outcome remain part of the claim.
Safety requires more detail than the most common event
An abstract’s summary of common adverse events is a starting point. Assessing personal suitability also requires the current prescribing information, relevant medical history and the care team’s judgment. This review does not use the abstract to declare one medicine universally safer.
If a comparison matters to a decision, ask about the events that led to stopping treatment, serious events, symptom management and follow-up. Those questions need their own denominators and definitions. A larger average benefit cannot by itself settle whether a particular person can tolerate or appropriately receive a treatment.
Funding is context for appraisal
The publication identifies Eli Lilly as the funder, and the registry identifies the company as sponsor. This commercial relationship should be visible when discussing the evidence.
Sponsorship does not prove a result is false, and publication does not remove the need for appraisal. Look for transparent methods, a clearly specified comparison and appropriately limited conclusions. Distinguish the original report from a sponsor’s promotional interpretation. This review cites the publication and trial record rather than treating a press release as an independent evaluation.
Bring a focused question to a treatment discussion
SURMOUNT-5 can inform a conversation about the average weight effects of the studied regimens. It did not test a personalized switching strategy, and it does not prescribe which treatment you should start or how to change a dose.
Ask how the trial fits your goals, eligibility, current treatment, tolerance and ability to continue care. Our medication consultation guide can help organize that discussion. The strongest use of a direct comparison is a more precise decision conversation: which part of the evidence applies, which remains uncertain and how your response will be assessed.
