- Design
- Bayesian meta-analysis of randomised placebo-controlled trials
- Population
- 10 trials, 6,515 adults with overweight or obesity
- Primary outcome
- Incident atrial fibrillation, atrial arrhythmia and any arrhythmia
- Effect
- AF OR 2.20 (95% CrI 0.81 to 6.75); any arrhythmia OR 1.84 (1.04 to 3.90)
A Bayesian meta-analysis in the Journal of the American Heart Association (September 2026) pooled 10 placebo-controlled trials of tirzepatide in adults with overweight or obesity, covering 6,515 people, 4,491 of them on tirzepatide.
The pooled odds ratio for atrial fibrillation was 2.20 (95% credible interval 0.81 to 6.75) and for atrial arrhythmia 2.16 (0.90 to 5.87) — both wide and both crossing 1. For any arrhythmia it was 1.84 (1.04 to 3.90).
Event numbers were low, arrhythmia was not a primary endpoint in any contributing trial, and 'any arrhythmia' groups quite different rhythms together, including benign ones. Tirzepatide also raises resting heart rate, which may contribute to reported events.
This is a signal to note, not a reason to withhold the drug. It is a reason to ask about palpitations at follow-up.
- Ask about palpitations at tirzepatide follow-up visits, particularly in patients with a cardiac history
- Record a baseline heart rate and ECG where arrhythmia risk is already raised
- Do not stop tirzepatide for an isolated sinus tachycardia without symptoms
- Investigate new palpitations on tirzepatide as you would in anyone else
- Report suspected arrhythmias to pharmacovigilance — the data are thin
Why it matters
Heart-rate effects of incretin drugs have usually been waved away as harmless; this raises the question without answering it.
Don't overread it
Arrhythmia was an adverse-event count, not a trial endpoint, and the atrial fibrillation estimate crosses 1.
The statistics, in plain English
An odds ratio of 2.20 sounds like a doubling, but the credible interval from 0.81 to 6.75 includes no effect, so the data are compatible with no increase at all. The 'any arrhythmia' interval (1.04 to 3.90) only just excludes 1, which with rare events and a composite outcome is fragile.
Read the rest in the app
You have read your two free briefings this month. The app carries all 27 specialties, every morning, free — and this finding is waiting in it.
Scan to keep reading on your phone. No account needed to start.
Tomorrow morning, before your first patient
One edition a day for diabetes & endocrinology, written by the desk, every claim tied to its paper. Six minutes.
Get the app — free