Two large claims studies disagree on tirzepatide versus semaglutide for heart events

Two big observational studies of US health records have reached opposite conclusions about which drug is better at preventing heart attacks, strokes and deaths: semaglutide (sold as Ozempic, Wegovy and Rybelsus) or tirzepatide (Mounjaro and Zepbound). One found a 29% lower rate of major cardiovascular events with semaglutide ^[1]. The other found lower event rates and lower death rates with tirzepatide ^[2].

What each study found

The STEER analysis used Komodo Research Data, a US insurance claims database. It included patients age 45 and older with overweight or obesity and at least one claim for heart attack, ischemic stroke or peripheral artery disease, who started semaglutide or tirzepatide between May 13, 2022 and January 31, 2025 ^[1]. Patients with diabetes were excluded. After propensity score matching, 10,625 patients were in each group ^[1].

In STEER, semaglutide was linked to a 29% lower risk of a revised 3-point MACE outcome — heart attack, stroke or death from any cause — with a hazard ratio of 0.71 and a p-value of 0.046 ^[1]. A broader 5-point outcome that added coronary revascularization and heart failure hospitalization was 22% lower (HR 0.78, p = 0.040) ^[1]. In a secondary analysis that stopped counting once patients went more than 30 days without refilling, the gaps were larger: HR 0.43 for the 3-point outcome and 0.57 for the 5-point outcome ^[1]. The listed authors were employees and shareholders of Novo Nordisk, which makes semaglutide, at the time of the study ^[1].

The second study used the TriNetX Research Network and looked at a different population: people with obesity (BMI 30 or above) and type 2 diabetes, with qualifying events from May 1, 2022 through November 21, 2024 ^[2]. After 1:1 propensity matching, there were 47,804 patients in each group, with a mean age of 75, 45% male and 74% identified as white ^[2]. Outcomes were measured over one year ^[2].

There, tirzepatide came out ahead. MACE occurred in 3.7% of tirzepatide patients versus 4.1% on semaglutide (risk ratio 0.918, 95% CI 0.862–0.978) ^[2]. All-cause death was 0.2% versus 0.4% (risk ratio 0.436, 95% CI 0.338–0.562) ^[2]. Average HbA1c was lower with tirzepatide (6.565% vs. 6.848%, p < 0.001), and gastrointestinal side effects were slightly less common (9.8% vs. 10.2%) ^[2]. There was no significant difference in heart failure exacerbation or urinary tract infections, and hospital or emergency department use was similar ^[2]. The authors reported no conflicts of interest ^[2].

Why the two disagree

The studies are not measuring the same thing in the same people. STEER looked at people with established atherosclerotic cardiovascular disease and no diabetes ^[1]. The TriNetX study looked at an older group with type 2 diabetes and obesity ^[2]. The outcome definitions differ too: STEER's "revised" MACE counted death from any cause alongside heart attack and stroke and followed patients through January 2025 ^[1], while the TriNetX analysis used a fixed one-year window ^[2].

Both are retrospective and observational. Propensity score matching balances the characteristics researchers can see in the data, such as age and diagnoses, but it cannot balance things the records do not capture — for example, how sick someone looked in clinic, why a doctor picked one drug, or whether a patient could afford to stay on it. Neither study randomly assigned treatment, so neither can prove one drug caused fewer events than the other.

Why it matters for patients

Headlines drawn from either study can make one drug sound clearly better for the heart. The two papers together show how fragile that conclusion is when it comes from claims and electronic health record data rather than a randomized trial. Funding and authorship are also worth noting: the semaglutide-favoring analysis was conducted by employees of semaglutide's manufacturer ^[1].

These sources do not report results from any randomized head-to-head trial comparing semaglutide and tirzepatide on cardiovascular events, so which drug is better for preventing heart attacks and strokes is not yet settled. Whether such a trial is underway is not addressed in these sources.

People weighing the two drugs may find it more useful to discuss their own history — prior heart attack or stroke, diabetes status, side effects, cost and coverage — with a clinician than to lean on either database result alone.