Tirzepatide and HFpEF: What 3,308 Matched Patients Show
A retrospective cohort study published in Therapeutic Advances in Cardiovascular Disease on 17 August 2026 asked whether men and women with obesity and heart failure with preserved ejection fraction fare differently while taking tirzepatide. Across 3,308 matched patients, the answer on cardiovascular outcomes is largely no.
The more interesting part is the sentence the authors put at the end of their own conclusion, which tells readers what the study cannot be used to claim.
A framing note: tirzepatide is an approved medicine and this is a study of patients. Peptra Labs supplies it as reference material for laboratory research only, and nothing here is guidance about human use.
How the tirzepatide study was built
The authors used TriNetX, a federated database drawing on 110 healthcare organisations. They identified adults with obesity, heart failure with preserved ejection fraction, and a tirzepatide prescription between 1 June 2022 and 1 January 2025.
Males and females were then balanced using 1 to 1 propensity score matching, giving 1,654 patients per group. Follow-up ran one year.
The primary outcome was all-cause hospitalisation. Secondary outcomes were all-cause mortality, heart failure exacerbation, major adverse cardiovascular events, incident atrial fibrillation, nausea and vomiting, and diarrhoea.
Propensity score matching is worth a sentence of its own. It builds two groups that resemble each other on the variables the researchers can see, so that a comparison between them is less distorted by obvious differences. It is the standard tool for database work of this kind, and its weakness is stated in the name: it matches on scores derived from recorded data.
The falsification endpoints
The authors included two prespecified falsification endpoints, and this detail deserves more attention than it usually gets.
A falsification endpoint is an outcome that the exposure should have no plausible effect on. If the analysis finds an association there, something is wrong with the matching or with residual confounding, because the study has detected an effect that cannot exist. It is a built-in alarm for a study design that cannot randomise.
The authors report that the falsification endpoints showed no significant associations, which is the result you want and a reason to take the rest of the numbers more seriously than an unchecked database analysis.
What the analysis found
| Outcome | Hazard ratio, males vs females | 95% CI |
|---|---|---|
| All-cause hospitalisation | 0.87 | 0.77 to 0.98 |
| All-cause mortality | 1.73 | 0.95 to 3.18 |
| Heart failure exacerbation | 1.14 | 0.90 to 1.44 |
| Major adverse cardiovascular events | 1.28 | 0.99 to 1.65 |
| Incident atrial fibrillation | 1.21 | 0.61 to 2.43 |
| Nausea and vomiting | 0.50 | 0.37 to 0.67 |
| Diarrhoea | 0.84 | 0.59 to 1.19 |
Two results reach significance. Men had a lower hazard of all-cause hospitalisation, and a substantially lower rate of coded nausea and vomiting. Everything else has a confidence interval crossing 1.
Two of those non-significant results are worth a second look rather than a dismissal. All-cause mortality at 1.73 and major adverse cardiovascular events at 1.28 both point the same direction, with intervals that only just include 1. That is not evidence of a difference. It is also not evidence of no difference, and in a study of this size the distinction is mostly about statistical power.
What the authors say it does not show
Their conclusion is unusually disciplined, and the final sentence is the one that matters.
They state that among tirzepatide-treated patients with obesity and heart failure with preserved ejection fraction, no statistically significant sex-associated differences were detected in major cardiovascular outcomes. Men had a modestly lower observed hazard of hospitalisation and fewer coded nausea and vomiting events.
Then: because this was a within-treatment observational comparison without an external comparator, the findings should not be interpreted as evidence of differential tirzepatide effectiveness by sex.
That sentence rules out the headline most coverage would reach for. Everyone in this study was taking the drug. There is no untreated group. So the comparison is between men and women who are all on treatment, which can describe how the two groups did, and cannot separate the drug from everything else that differs between men and women with this condition.
Two more limits worth naming
Coded events are not observed events. Nausea and vomiting here are diagnosis codes entered into records across 110 organisations. A hazard ratio of 0.50 could reflect a real difference in symptoms, a difference in how readily men and women report them, a difference in how clinicians code them, or all three. Database studies measure what was written down.
Propensity score matching balances what is measured. It cannot balance what is not in the database. The falsification endpoints give some reassurance that unmeasured confounding is not overwhelming, which is not the same as absent.
Where this sits
Tirzepatide in this condition is not a fringe question. A trial of tirzepatide for heart failure with preserved ejection fraction and obesity was published in the New England Journal of Medicine in 2025, and a comparison of semaglutide and tirzepatide in the same population appeared in JAMA later that year. Those are the studies that can address effectiveness. This one addresses a narrower question about how outcomes distribute within a treated population.
We reported earlier on a meta-analysis of blood pressure related adverse events with the same compound, where the endpoint was also recorded adverse events rather than measured physiology. The two papers share a methodological shape: both are informative about what shows up in records, and both need care before being read as biology.
For laboratories working with the compound, our tirzepatide reference material documentation covers identity and purity fields, and the European research buyer guide covers procurement.
References
- Mortada I, Lee AW, Buckingham D, et al. Sex-based differences in clinical outcomes among patients with obesity and heart failure with preserved ejection fraction treated with tirzepatide. Therapeutic Advances in Cardiovascular Disease, 17 August 2026. doi 10.1177/17539447261478928
- Tirzepatide for Heart Failure with Preserved Ejection Fraction and Obesity. New England Journal of Medicine, 30 January 2025
- Semaglutide and Tirzepatide in Patients With Heart Failure With Preserved Ejection Fraction. JAMA, 14 October 2025
The products referenced on this site are supplied for laboratory research use only. They are not medicines and are not intended for human or veterinary use. This article summarises published research for informational purposes and is not medical advice. Statements about third-party studies belong to their authors.
