Independent · Evidence-led · We don't sell peptides
EU / NordicsUpdated weeklyEN

Explore this article's sources with AI

ChatGPTClaudePerplexityGeminiGrokGoogle AI

Follow PeptideMethods on Google

First published

Semaglutide vs Tirzepatide Heart Risk

A claims-based target-trial emulation found no significant cardiovascular difference, but it cannot prove semaglutide and tirzepatide are equal.

Why we wrote this. A target-trial emulation can look randomized in a headline. We explain the observed estimate without treating it as proof of equivalence.

In this article (5 sections)
  1. What the target-trial emulation compared
  2. The estimate was close to statistical uncertainty
  3. Why emulating a trial is still observational
  4. Randomized evidence answers different questions
  5. What remains unresolved

A 2026 analysis of US insurance claims found no statistically significant difference in a broad cardiovascular composite between semaglutide and tirzepatide users with overweight or obesity, atherosclerotic cardiovascular disease, and no diabetes. The hazard ratio comparing semaglutide with tirzepatide was 1.21, with a 95% confidence interval from 0.99 to 1.51[1]. This was an observational target-trial emulation, not a randomized head-to-head trial, and the result does not prove equal cardiovascular effects.

The researchers used Komodo Health national claims data in an active-comparator, new-user design. They applied inverse-probability weighting and Cox models to make the treatment groups more comparable on recorded characteristics[1]. Those methods can reduce measured imbalance. They cannot remove differences that claims data do not capture, such as clinical reasoning, adherence, achieved dose, or changes in weight and risk factors over time.

What the target-trial emulation compared

Eligible adults had documented atherosclerotic cardiovascular disease and a body mass index of at least 27 with an overweight or obesity diagnosis during the previous year. They also needed continuous insurance coverage. The study excluded people with diabetes or contraindications, as well as prior users of a GLP-1 receptor agonist or another obesity medicine[1]. The comparison therefore concerns new users in a defined high-risk group, not everyone who receives either medicine.

The primary outcome was a revised five-part composite of myocardial infarction, stroke, hospitalization for heart failure or unstable angina, coronary revascularization, and mortality. Secondary analyses examined a revised three-part composite and the individual endpoints[1]. Combining events can increase statistical power, but each component has a different clinical meaning. A composite result should not be read as proof that every component moved in the same way.

The PubMed abstract does not state the treatment-group denominators, event counts, follow-up duration, or an absolute risk difference. Those omissions matter for interpretation. A hazard ratio describes relative rates over observed follow-up, while absolute event counts and follow-up show the scale and precision of the result. We do not infer those missing values from the title or from another database study.

The estimate was close to statistical uncertainty

For the revised five-part outcome, the estimated hazard was 21% higher with semaglutide than with tirzepatide, but the confidence interval ranged from 1% lower to 51% higher. Because the interval included 1.00, the prespecified comparison was not statistically significant[1]. The abstract also reports no significant difference in the secondary outcomes. A nonsignificant finding means the analysis did not clearly separate the groups. It is not evidence that the medicines are interchangeable.

Sensitivity analyses kept the five-part estimates pointed in the same general direction, yet the result changed with decisions about early events, off-label use, analytic method, and outcome definition[1]. That sensitivity is useful information about robustness. It also limits any confident ranking of the two medicines from this study alone.

Why emulating a trial is still observational

A target-trial emulation starts by specifying the randomized trial researchers would like to run, then applies those eligibility rules, treatment definitions, follow-up choices, and outcomes to existing data. This can prevent common design errors, including mixing established users with new users or measuring outcomes before treatment begins. It does not randomize treatment. Clinicians, patients, insurance rules, product availability, and calendar time still influence which medicine appears in a claim.

Inverse-probability weighting gives more influence to participants who help balance measured baseline differences. Its validity depends on the recorded variables, the treatment model, and adequate overlap between groups. Unmeasured confounding can remain even when weighted tables look balanced. Claims also identify events through billing records rather than blinded clinical adjudication, which can introduce outcome misclassification.

Randomized evidence answers different questions

SELECT randomized 17,604 adults with established cardiovascular disease, overweight or obesity, and no diabetes to semaglutide 2.4 mg or placebo. Its primary cardiovascular outcome occurred in 6.5% of the semaglutide group and 8.0% of the placebo group over a mean follow-up of 39.8 months, for a hazard ratio of 0.80[2]. That establishes a semaglutide benefit against placebo in the SELECT population. It does not compare semaglutide directly with tirzepatide.

SURMOUNT-5 directly randomized 751 adults with obesity and no diabetes to tirzepatide or semaglutide, but it studied weight change rather than cardiovascular events. Mean weight loss at 72 weeks was 20.2% with tirzepatide and 13.7% with semaglutide[3]. Greater weight loss cannot be converted into a cardiovascular-event advantage without a trial designed and powered for cardiovascular outcomes.

These studies belong beside the new claims analysis, not inside it. SELECT provides randomized cardiovascular evidence for semaglutide versus placebo. SURMOUNT-5 provides randomized weight evidence for tirzepatide versus semaglutide. The new study offers an observational comparison of coded cardiovascular outcomes between the two medicines. The semaglutide evidence page and tirzepatide evidence page keep those questions separate.

What remains unresolved

The target-trial emulation does not establish equivalence or noninferiority. It also does not show that either medicine should be selected for a particular person. A stronger comparison would randomly assign semaglutide and tirzepatide, use a prespecified cardiovascular endpoint with blinded adjudication, report absolute as well as relative risks, and follow enough participants long enough to produce a precise estimate.

For now, the study supports a narrow statement: this weighted claims analysis did not detect a statistically significant cardiovascular difference between the treatment groups, and its estimate was sensitive to several analytic choices. Prescribing decisions still require the approved indications, contraindications, adverse effects, access, and the patient's clinical history to be considered by a qualified professional.

Medical disclaimer: This article is for educational and journalistic purposes only and does not constitute medical advice. Peptides discussed may be classified as prescription medicines or research chemicals depending on your jurisdiction. Always consult a qualified healthcare professional before using any peptide product. PeptideMethods.com does not sell, distribute, or facilitate the sale of any peptide product.

Frequently asked

Did semaglutide and tirzepatide have the same cardiovascular risk?

The weighted claims analysis found no statistically significant difference, but that does not prove equivalence. The confidence interval allowed effects in either direction, and treatment was not randomized.

What is a target-trial emulation?

It applies the planned structure of a hypothetical randomized trial to observational data. It can improve study design, but it cannot create random treatment assignment or remove all confounding.

What cardiovascular outcome did the study use?

The primary composite included myocardial infarction, stroke, hospitalization for heart failure or unstable angina, coronary revascularization, and mortality. Secondary analyses used a narrower composite and individual endpoints.

Does greater weight loss prove better heart protection?

No. SURMOUNT-5 compared weight loss, not cardiovascular events. A cardiovascular advantage requires a study designed around adjudicated cardiovascular outcomes.

Sources

  1. [1]Cardiovascular outcomes in semaglutide versus tirzepatide users: target trial emulation (PMID 43010688)Tier 1 · primary↩
  2. [2]Semaglutide and cardiovascular outcomes in obesity without diabetes: SELECT (PMID 37952131)Tier 1 · primary↩
  3. [3]Tirzepatide compared with semaglutide for obesity: SURMOUNT-5 (PMID 40353578)Tier 1 · primary↩

No revisions yet. First published .

About the editorial team

PeptideMethods is written and edited by the PeptideMethods Editorial Team and published by Digital Compass Group Ltd. The team is not made up of medical professionals; every health, regulatory or dosage claim on the site is tied to a primary source and is not a substitute for advice from a qualified clinician.

See our editorial policy and methodology for how we research, source and verify.

Read the pillars