Tirzepatide linked to lower stroke risk than other GLP-1 drugs in atrial fibrillation study

A retrospective analysis published August 22 in the journal Nature Communications found that adults with both atrial fibrillation and type 2 diabetes had lower risks of stroke and death after starting tirzepatide than similar patients who started another GLP-1 receptor agonist. The authors wrote that randomized trials are required before treatment recommendations can change ^[1].

The researchers compared 14,118 matched pairs of adults who began taking tirzepatide or another GLP-1 receptor agonist, drawn from a large international federated electronic health record network, the TriNetX Global Collaborative Network, which is credited in the paper's acknowledgements ^[1].

Tirzepatide use was associated with lower risks of any stroke, ischemic stroke (blocked blood flow to the brain), hemorrhagic stroke (bleeding in the brain) and death from any cause ^[1].

The absolute difference in stroke risk was 2.1 percentage points at one year, the paper states. The absolute difference in mortality was 6.2 percentage points; the article does not attach a time frame to that mortality figure ^[1].

What the study can and cannot show

This was an observational, retrospective study based on existing health records, not a randomized trial in which patients were assigned by chance to one drug or the other. That design can show an association between a drug and an outcome, but it cannot prove that tirzepatide itself caused the difference ^[1]. The paper's own conclusion is that the findings support further assessment of tirzepatide as a potential option for people with both conditions, while adding that randomized trials are needed before treatment recommendations change ^[1].

Several specifics are not given in the article. It does not report which individual GLP-1 receptor agonists were used as the comparison, the doses patients took, how long they took them, or the relative risk estimates and confidence intervals behind the percentage-point differences. Those details have not been disclosed in the published text. Whether the pattern holds beyond one year also has not been reported ^[1].

Why it matters for patients

People who have both atrial fibrillation and type 2 diabetes remain at high risk of stroke even when they are taking blood thinners and drugs that lower blood sugar, according to the paper's background ^[1]. That makes the choice between tirzepatide and another GLP-1 medication a live question for a large group of patients.

For someone already taking one of these drugs, the practical weight of this finding is limited by its design. The study compared people starting tirzepatide with people starting a different GLP-1 receptor agonist. It did not test whether switching from one drug to another would change anyone's stroke risk, and no such switching evidence is reported here ^[1]. The authors' stated position is that this analysis alone should not change how these drugs are prescribed ^[1].

It is also worth noting that the two groups were assembled from medical records rather than assigned at random, so differences between the patients — in age, other illnesses, blood thinner use or how closely they were followed — could partly explain the gap in outcomes. The paper describes matching the pairs but does not detail how fully those differences were balanced ^[1].

What happens next

The authors call for randomized trials to test whether the association holds up when patients are assigned to treatments by chance. No specific trial, sponsor or timeline is named in the paper, and no date for a follow-up analysis has been reported ^[1].

The article was received October 21, 2025, accepted July 24, 2026, and published August 22, 2026 ^[1]. The work was funded by the Lo-Hsu Medical Foundation, Lotung Poh-Ai Hospital, through grants 11403 and 11404, and the authors declared no competing interests ^[1].