A direct comparison can answer a question that separate placebo-controlled studies cannot settle cleanly: how do two treatments perform when tested in the same trial? SURMOUNT-5 randomized 751 adults with obesity but without type 2 diabetes to injectable tirzepatide or injectable semaglutide. At 72 weeks, the estimated average weight reduction was greater with tirzepatide. That is a meaningful comparative finding, but it is not a universal ranking of medicines for every person or every health outcome.
The useful reading is both precise and practical. What regimens were compared? What does a least-squares mean represent? How much uncertainty surrounds the estimates, and what safety information is actually reported? This analysis uses the indexed primary report abstracts for SURMOUNT-5, STEP 1, SURMOUNT-1, and SELECT, together with available official FDA material. It does not represent a review of all full articles, protocols, or supplements.
At a glance
- Direct randomized comparison
- SURMOUNT-5 assigned 751 adults with obesity but without type 2 diabetes to tirzepatide or semaglutide in a 1:1, open-label trial lasting 72 weeks.Sources: [1]
- Studied injectable regimens
- The trial compared maximum tolerated once-weekly subcutaneous tirzepatide, 10 or 15 mg, with maximum tolerated once-weekly subcutaneous semaglutide, 1.7 or 2.4 mg.Sources: [1]
- Primary weight result
- Least-squares mean weight change was −20.2% with tirzepatide and −13.7% with semaglutide, a 6.5-percentage-point separation calculated from the reported rounded estimates.Sources: [1]
- Waist circumference
- Estimated mean waist change was −18.4 cm with tirzepatide and −13.0 cm with semaglutide; the reported 95% confidence intervals were −19.6 to −17.2 cm and −14.3 to −11.7 cm, respectively.Sources: [1]
- Safety finding, not safety supremacy
- Gastrointestinal events were the most common adverse events in both groups. Most were mild to moderate and occurred primarily during dose escalation; this does not establish universal safety superiority for either treatment.Sources: [1]
The question a head-to-head trial actually answers
SURMOUNT-5 was a phase 3b controlled trial with an active comparator: participants received one of two medicines rather than a medicine or placebo. Random assignment makes the groups more comparable at the outset and supports interpreting differences as treatment effects within the trial’s population and procedures. It is stronger evidence for this particular comparison than placing results from unrelated trials side by side.
The population matters. Participants were adults with obesity and without type 2 diabetes. The result therefore directly addresses that setting, not every person using either medicine. The abstract does not provide enough detail to judge all eligibility restrictions or how closely the sample resembles a particular clinical population.
The study was funded by Eli Lilly and registered as NCT05822830. Sponsor involvement is relevant context, not a reason by itself to discard randomized evidence. The important task is to examine the design and reported outcomes while recognizing that an abstract alone does not permit a complete assessment of analysis decisions, reporting completeness, or sponsor roles.
Maximum tolerated regimens are not interchangeable dose labels
Participants were assigned to once-weekly injections under the skin. The tirzepatide arm used a maximum tolerated dose of 10 or 15 mg; the semaglutide arm used 1.7 or 2.4 mg. These numbers identify the study arms. They are not instructions for starting, increasing, or selecting a dose.
This was a comparison of treatment strategies allowing the specified tolerated doses, not a single fixed-dose contest in which everyone necessarily received the highest listed dose. Tolerability is therefore part of what the trial tested. The abstract does not report how many participants finished at each dose, so it cannot tell us how much each dose level contributed to the overall difference.
Milligram amounts also cannot be compared as if they measured equivalent potency across different molecules. Nor does the study isolate which biological mechanism explains the observed separation. For an actual offer, the first documentary question remains what formulation and route are being supplied; the medication-identification guide helps organize that distinction.
STUDY DATA · GLP-1 Care Insight
SURMOUNT-5: mean weight change at 72 weeks
751 adults with obesity without type 2 diabetes, randomized in an open-label trial. Weekly subcutaneous maximum tolerated tirzepatide 10/15 mg versus semaglutide 1.7/2.4 mg.
| Study group | Least-squares mean weight change (%) |
|---|---|
| Tirzepatide | -20.2 |
| Semaglutide | -13.7 |
95% CIs: tirzepatide −21.4 to −19.1%; semaglutide −14.9 to −12.6%. Trial doses identify arms, not instructions. Results do not establish universal safety superiority or provider outcomes. Funded by Eli Lilly.
Original visualization of published aggregate data. No individual prediction or provider performance is represented. Original research ↗
Open-label does not erase randomization—but it changes interpretation
Open-label means participants and investigators knew the assigned treatment. Randomization and blinding solve different problems: randomization helps balance starting characteristics, whereas blinding can reduce changes in behavior, symptom reporting, or assessment that arise from knowing the assignment. SURMOUNT-5 had the former protection, not the latter.
Body weight is a more objective measurement than a participant’s impression of improvement. Still, knowledge of treatment could influence adherence, expectations, or other behaviors that affect weight. This is a potential source of influence, not evidence that such influence explains the result. The abstract does not quantify it.
The primary endpoint was percent weight change from baseline to week 72. That timing anchors the conclusion: greater average loss by 72 weeks under the tested regimens. It does not establish what would happen after stopping treatment or many years later. Those are different study questions, which should not be silently folded into this result.
Reading −20.2% versus −13.7% without losing the units
The primary report gives least-squares mean weight changes of −20.2% for tirzepatide and −13.7% for semaglutide. A least-squares mean is a model-based group estimate, not simply a promise about what a typical individual will lose. The abstract does not specify all model terms or missing-data assumptions, which limits a deeper audit of how the estimates were constructed.
Subtracting the reported rounded values gives a 6.5-percentage-point difference in weight change favoring tirzepatide. It is not a difference of 6.5% of starting weight loss, and it is not a statement that every participant lost an extra 6.5%. Percent and percentage points answer different questions.
As an arithmetic illustration only, for a starting weight of 100 kg, reductions of 20.2% and 13.7% correspond to 20.2 kg and 13.7 kg. The numerical gap would be 6.5 kg. This is not a prediction for a 100-kg reader; it shows how percentage change translates into weight units when the same starting weight is used.
| Measure | Tirzepatide | Semaglutide |
|---|---|---|
| Least-squares mean weight change | −20.2% | −13.7% |
| 95% CI for weight change | −21.4% to −19.1% | −14.9% to −12.6% |
| Least-squares mean waist change | −18.4 cm | −13.0 cm |
| 95% CI for waist change | −19.6 to −17.2 cm | −14.3 to −11.7 cm |
| Study regimen, once weekly under the skin | Maximum tolerated 10 or 15 mg | Maximum tolerated 1.7 or 2.4 mg |
Confidence intervals describe estimation, not individual outcomes
The 95% confidence interval for mean percent weight change was −21.4 to −19.1 with tirzepatide and −14.9 to −12.6 with semaglutide. These intervals describe statistical uncertainty around the modeled group estimates. They are not ranges containing 95% of participants’ weight changes, and they do not show the full spread of individual responses.
The report gives P<0.001 for the weight comparison. Under the statistical test’s assumptions, that supports a difference rather than random sampling variation alone. Statistical significance does not answer whether the difference matters equally to everyone, nor does it resolve tolerability, treatment burden, or outcomes other than weight.
The abstract supplies confidence intervals for each arm but not the confidence interval for the between-group weight difference. It would be inappropriate to manufacture that interval by subtracting endpoints. A guide to questions about advertised weight-loss percentages can help separate a reported group estimate from an individualized claim.
Waist reduction and responder thresholds add different information
Waist circumference fell by an estimated 18.4 cm with tirzepatide and 13.0 cm with semaglutide. The difference calculated from those rounded estimates is 5.4 cm. The report also gives P<0.001 for this comparison. Waist change adds a second body-size measurement, but it is not itself a count of cardiovascular events or proof of a particular health benefit.
Participants assigned tirzepatide were also more likely to reach weight reductions of at least 10%, 15%, 20%, and 25%. These threshold outcomes complement the average: they ask how often participants crossed specified milestones rather than how much the group lost on average.
However, the available abstract does not give the proportions reaching those thresholds in SURMOUNT-5. It therefore supports the direction of the comparison, not a precise statement about someone’s chance of reaching 20% loss. Nor can an average determine whether a particular participant benefited, had little response, or stopped treatment.
Why STEP 1 and SURMOUNT-1 are context, not substitutes
STEP 1 randomized 1,961 adults without diabetes to semaglutide 2.4 mg or placebo, alongside lifestyle intervention, for 68 weeks. Mean weight change was −14.9% versus −2.4%; the reported estimated treatment difference was −12.4 percentage points, with a 95% confidence interval of −13.4 to −11.5. The reported modeled difference need not equal subtraction of rounded group values exactly.
SURMOUNT-1 randomized 2,539 adults without diabetes to tirzepatide 5, 10, or 15 mg or placebo for 72 weeks, including 20 weeks of dose escalation. Mean changes were −15.0%, −19.5%, and −20.9% across the tirzepatide arms, versus −3.1% with placebo. Unlike SURMOUNT-5’s tolerated-dose strategy, this study reported separate assigned-dose arms.
Both were double-blind placebo-controlled studies. STEP 1’s primary estimand—the precise treatment effect being targeted—addressed effects regardless of discontinuation or rescue interventions; SURMOUNT-1’s treatment-regimen estimand addressed discontinuation in the randomized population. Differences in duration, comparator, dosing, participants, and analysis mean their headline numbers should not replace the direct SURMOUNT-5 comparison.
| Trial and population size | Comparison and duration | What the design directly addresses |
|---|---|---|
| SURMOUNT-5: 751 participants | Maximum tolerated injectable tirzepatide versus semaglutide; 72 weeks | Relative weight and waist outcomes between two active regimens |
| STEP 1: 1,961 participants | Injectable semaglutide 2.4 mg versus placebo, plus lifestyle intervention; 68 weeks | Semaglutide’s effect compared with placebo in that setting |
| SURMOUNT-1: 2,539 participants | Injectable tirzepatide 5, 10, or 15 mg versus placebo; 72 weeks | Results for separate assigned tirzepatide doses compared with placebo |
The safety evidence does not produce a universal winner
In SURMOUNT-5, gastrointestinal events were the most common adverse events in both groups. Most were mild to moderate and occurred primarily during dose escalation. That describes the dominant reported pattern; it does not mean all events were mild or that either treatment was risk-free.
The head-to-head abstract does not provide detailed event rates or discontinuation percentages. Earlier studies demonstrate why tolerability deserves separate attention. In STEP 1, gastrointestinal events led to treatment discontinuation in 4.5% of semaglutide participants versus 0.8% with placebo. In SURMOUNT-1, adverse events led to discontinuation in 4.3%, 7.1%, and 6.2% across the tirzepatide dose arms, versus 2.6% with placebo.
Those discontinuation measures cannot be used as a substitute safety contest: one concerns gastrointestinal events specifically, the other adverse events overall, in different trials. Greater average weight reduction also does not prove superior tolerability. The scientifically useful question is which adverse outcomes were measured, how often they occurred, and how they affected continued participation.
More weight loss does not settle cardiovascular benefit or durability
SELECT illustrates why outcomes cannot be ranked by weight results alone. It enrolled 17,604 people aged 45 or older with established cardiovascular disease, BMI of at least 27, and no diabetes. Once-weekly injectable semaglutide 2.4 mg was compared with placebo, not tirzepatide, over a mean follow-up of 39.8 months.
Its primary composite—cardiovascular death, nonfatal heart attack, or nonfatal stroke—occurred in 6.5% versus 8.0%. The hazard ratio was 0.80, with a 95% confidence interval of 0.72 to 0.90. That represents a lower hazard over follow-up, not a 20-percentage-point reduction in event probability. A lower composite also does not establish that every component improved individually.
Adverse events caused permanent trial-product discontinuation in 16.6% versus 8.2% in SELECT. The SELECT cardiovascular analysis addresses that different benefit–harm question. Likewise, questions about what happens after treatment ends belong to follow-up or withdrawal research, discussed separately in the STEP 1 extension analysis, not to SURMOUNT-5’s 72-week comparison.
Translating evidence into questions—not provider rankings
SURMOUNT-5 tested specified injectable regimens, not oral products, compounded preparations, or commercial care programs. FDA states that compounded drugs are not FDA-approved or reviewed before marketing for safety, effectiveness, or quality. Results from studied products do not establish equivalence for compounded versions.
FDA also identifies dosing errors and concerns about semaglutide salt ingredients. Such reports identify safety concerns but cannot, by themselves, establish event rates or causation. Product identity, formulation, and route therefore remain essential distinctions rather than minor details beneath a shared medicine name.
Editorial interpretation: a trial result should be one column in a decision record, not a score assigned to a provider. A comparison worksheet can keep clinical evidence separate from contract terms. The documentary CoreAge Rx review and Mochi Health review are places to check enrollment and payment questions; neither link implies access to the studied regimens or trial-like outcomes.
Questions & answers
Does SURMOUNT-5 prove tirzepatide is the better medicine for everyone?
No. It establishes greater average weight and waist reduction at 72 weeks with the tested tirzepatide strategy in adults with obesity without type 2 diabetes. Individual response, adverse effects, other health goals, and evidence for different outcomes remain separate questions. A weight-loss comparison is not a universal benefit–harm ranking.
Why was semaglutide’s weight result different from STEP 1?
STEP 1 reported −14.9% at 68 weeks, while SURMOUNT-5 reported a least-squares mean of −13.7% at 72 weeks. They used different comparators, blinding, and dosing strategies. Ordinary sampling variation and analytic choices can also matter. The abstracts do not isolate a single explanation, and the difference alone does not show that one result is wrong.
Can the confidence intervals tell me my likely weight loss?
Not directly. They quantify uncertainty around estimated group means, not the distribution of individual outcomes. Even a precisely estimated average can coexist with substantial variation between participants. The SURMOUNT-5 abstract does not supply enough individual-response information to build a personalized prediction.
Did the study show which treatment causes fewer serious adverse effects?
The available abstract does not establish a universal serious-adverse-effect advantage. It reports gastrointestinal events as most common in both groups, mostly mild to moderate and concentrated during dose escalation. Detailed comparative safety conclusions require event-specific counts, definitions, and uncertainty estimates that are not presented in that abstract.
Can a program apply the 20.2% result to a compounded or oral regimen?
Not on the basis of this trial. SURMOUNT-5 studied specified once-weekly subcutaneous regimens. A different route or formulation requires its own supporting evidence. FDA specifically notes that compounded products lack premarket approval review; a familiar ingredient name does not establish equivalent quality, exposure, effectiveness, or safety.
Our assessment
SURMOUNT-5 provides direct randomized evidence that its maximum tolerated injectable tirzepatide regimen produced greater average weight and waist reductions than its semaglutide regimen over 72 weeks. Its strongest contribution is replacing an indirect comparison with a shared trial framework. The most useful interpretation preserves that finding without expanding it into claims about every patient, long-term cardiovascular superiority, universal safety, or commercial programs. Questions still unresolved by the accessible abstract include detailed comparative safety, dose-specific contributions, individual response variation, and durability beyond the reported follow-up.
Sources & documentation
Scientific and official sources consulted October 20, 2026. Study findings apply to the populations, products and time periods described; they are not provider outcome data.
- Tirzepatide as Compared with Semaglutide for the Treatment of Obesity. ↗
- Once-Weekly Semaglutide in Adults with Overweight or Obesity. ↗
- Tirzepatide Once Weekly for the Treatment of Obesity. ↗
- Semaglutide and Cardiovascular Outcomes in Obesity without Diabetes. ↗
- FDA concerns about unapproved GLP-1 drugs ↗