For years, comparisons between semaglutide and tirzepatide relied on cross-trial analysis — SURMOUNT-1 showed tirzepatide produced 20.9% weight loss; STEP trials showed semaglutide produced roughly 14.9%. These numbers came from separate trials, run in different periods, with different populations and protocols, making direct comparison methodologically shaky. SURMOUNT-5 changed that: it's the first published head-to-head, active-comparator trial directly testing the two medications in the same patient population at the same time.[1]
Trial design
SURMOUNT-5 was a Phase 3b, multicenter, open-label, randomized, active-controlled trial conducted at 32 sites in the U.S. and Puerto Rico between April 2023 and November 2024.[1] It enrolled 751 adults with obesity or overweight (BMI ≥30, or ≥27 with at least one weight-related complication such as hypertension, dyslipidemia, obstructive sleep apnea, or cardiovascular disease) without type 2 diabetes. Mean age was approximately 45 years; 64.7% of participants were women. The trial specifically enrolled a diverse population — 19% Black or African American, 26% Hispanic or Latino.
Participants were randomized 1:1 to subcutaneous tirzepatide or semaglutide, once weekly, titrated to the maximum tolerated dose (10mg or 15mg for tirzepatide; 1.7mg or 2.4mg for semaglutide), alongside a behavioral support program, for 72 weeks. The primary endpoint was percentage weight change from baseline to week 72.
Results
In absolute terms, that translated to a mean loss of 22.8 kg with tirzepatide versus 15.0 kg with semaglutide. Waist circumference reduction was similarly larger with tirzepatide (-18.4 cm vs. -13.0 cm). Participants on tirzepatide were more likely to reach every weight-loss threshold tested — ≥10%, ≥15%, ≥20%, and ≥25% — than those on semaglutide.[2]
The results are consistent with — in fact, almost identical to — what we've seen in trials in which these drugs were evaluated independently.
That comment from principal investigator Louis Aronne, MD, is a meaningful methodological note: the head-to-head numbers closely tracked what the earlier, separate SURMOUNT-1 and STEP trials had already suggested indirectly. SURMOUNT-5 didn't overturn the prior indirect comparison — it confirmed it under a more rigorous, controlled design.[3]
A notable subgroup finding: sex differences
Weight loss in female participants was greater than in male participants in this trial — a reversal of the pattern typically observed with lifestyle interventions alone, where men often lose weight somewhat faster than women. The trial data does not fully resolve why this reversal occurs with pharmacotherapy specifically, though it is a specific, testable finding worth further mechanistic investigation.[1]
Cardiometabolic secondary endpoints
Beyond weight, tirzepatide showed significantly greater improvements than semaglutide at 72 weeks across several cardiometabolic risk parameters: systolic and diastolic blood pressure, HbA1c, fasting insulin, triglycerides, and HDL cholesterol.[4] Consistent with prior GLP-1 literature generally, both treatment arms showed a dose-response relationship between magnitude of weight loss and magnitude of cardiometabolic improvement — greater weight loss correlated with greater risk-factor improvement in both groups, not just in the tirzepatide arm specifically.
A modeled, not observed, cardiovascular projection
A post-hoc analysis using SURMOUNT-5's patient-level data and NHANES population data projected 10-year cardiovascular disease risk reduction for tirzepatide versus semaglutide.[5] This is a modeled estimate extrapolating from risk-factor changes, not a direct observation of cardiovascular events — SURMOUNT-5 was not designed or powered to capture actual CV event rates over 72 weeks. A genuine head-to-head cardiovascular outcomes trial comparing the two drugs directly does not yet exist.
SURMOUNT-5 excluded patients with type 2 diabetes. Whether the same magnitude of difference holds in a diabetic population is a separate, unanswered question — the trials establishing efficacy in diabetic populations for each drug were conducted separately, not head-to-head.
The trial was open-label, meaning both participants and investigators knew which medication each person was receiving — a standard limitation for trials comparing two already-approved, distinctly-dosed injectable medications, where blinding is logistically difficult, but one that can introduce reporting or behavioral bias relative to a double-blind design.
Real-world confirmation
A separate 6-month retrospective cohort study using Truveta de-identified U.S. electronic health record data, examining adults who initiated tirzepatide or semaglutide between December 2023 and June 2024, found real-world weight reduction and cardiometabolic improvements directionally consistent with SURMOUNT-5's randomized trial findings — an encouraging sign that the controlled-trial results are translating into comparable real-world effectiveness.[6]
Compare GLP-1 telehealth providers
Source GLP-1 tracks pricing, licensing, and provider structure across the market.