GLP-1 Medicines and Atrial Fibrillation: What the Trials Show About Heart Rhythm, Ablation and AF Risk

Do GLP-1 medicines raise or lower the risk of atrial fibrillation? Across the randomised trials, the class as a whole does not raise it, and semaglutide in particular has been linked to fewer new cases. The best-supported pathway is the one cardiologists have studied for over a decade: sustained weight loss reduces how often atrial fibrillation comes back and how bad it feels.

Atrial fibrillation, or AF, is an irregular and often fast rhythm that starts in the upper chambers of the heart. It can come and go (paroxysmal), last for days or weeks (persistent), or become the heart's settled rhythm (permanent). Obesity is one of the strongest drivers of it, which is why GLP-1 drugs drew attention from rhythm specialists.

There was an early worry too. GLP-1 receptor stimulation raises heart rate, and one older drug in the class looked like it nudged AF upward. This article goes through what the meta-analyses found, what the STEP-HFpEF program showed in people who already have AF, the ablation data, the weight-loss trials that came first, and what has not been studied, including low doses.

Do GLP-1 medicines cause atrial fibrillation?

Monami 2017 and Nreu 2020

The pooled trial data say the GLP-1 class does not increase atrial fibrillation. The concern was reasonable: a 2017 meta-analysis by Matteo Monami and colleagues in the Journal of Endocrinological Investigation opened by noting that GLP-1 receptor stimulation is associated with a rise in heart rate, and that pooled phase III data for albiglutide had shown a significant increase in AF risk.

That 2017 review gathered 113 randomised trials of at least 12 weeks in people with type 2 diabetes. In the trials that recorded AF events, covering 17,966 people on a GLP-1 drug and 15,305 on comparators, the odds ratio for AF was 0.87, and the difference from comparators was not statistically significant. The authors concluded the class was not associated with AF, with the possible exception of albiglutide.

A larger update in 2020 by Besmir Nreu and colleagues, in Nutrition, Metabolism and Cardiovascular Diseases, pooled 43 trials lasting 52 weeks or more with 63,134 patients. GLP-1 receptor agonists reduced major cardiovascular events (odds ratio 0.87) and did not increase AF (odds ratio 0.94). The authors wrote that the AF data seemed to exclude any major safety issue. If the faster pulse on a GLP-1 is what worries you, our article on resting heart rate on GLP-1 protocols explains where that rise comes from.

Does semaglutide lower the risk of atrial fibrillation?

Saglietto and colleagues, European Journal of Clinical Investigation, 2024

In the trial data so far, semaglutide is linked to fewer new cases of atrial fibrillation. Andrea Saglietto and colleagues pooled 10 randomised trials with 12,651 participants (7285 on semaglutide and 5366 on placebo) and found a 42% lower risk of incident AF, a relative risk of 0.58, with no heterogeneity between studies.

Oral and injected semaglutide gave similar results (relative risks of 0.53 and 0.59). Meta-regression found no clear influence of the share of people with diabetes or of baseline BMI. The authors described the population as people at high cardiovascular risk, mainly with type 2 diabetes.

A narrower 2025 meta-analysis in the European Journal of Heart Failure, by Raja Ravikulan and colleagues, looked only at people with heart failure with preserved or mildly reduced ejection fraction. Pooling four trials (SELECT, FLOW, STEP-HFpEF and STEP-HFpEF DM) with 3743 participants, GLP-1 therapy reduced incident AF with a risk ratio of 0.54. Heterogeneity was moderate (I² 51%), and the authors called for heart failure trials with adjudicated AF outcomes. That caveat applies to all of these numbers: AF was mostly counted as an adverse event, not a planned endpoint.

What did STEP-HFpEF show in people who already have AF?

Verma and colleagues, Journal of the American College of Cardiology, 2024

In people with obesity-related heart failure, semaglutide helped those with atrial fibrillation at least as much as those without it, and on symptoms more. The STEP-HFpEF and STEP-HFpEF DM trials randomised adults with a left ventricular ejection fraction of 45% or more and a BMI of 30 or more to semaglutide 2.4 mg once weekly or placebo for 52 weeks.

A pooled analysis led by Subodh Verma found that 518 of the 1,145 participants (45%) had a history of AF. Of those, 40% had paroxysmal AF, 24% persistent and 35% permanent. People with AF were older, more often men, had higher NT-proBNP and more often had NYHA class III symptoms.

On the Kansas City Cardiomyopathy Questionnaire clinical summary score (KCCQ-CSS), a measure of heart failure symptoms and physical limits, semaglutide improved scores by 11.5 points over placebo in people with AF and 4.3 points in people without. The win ratio for the hierarchical composite endpoint was 2.25 with AF and 1.30 without. Weight, CRP and NT-proBNP fell whatever the AF status, and serious adverse events were less frequent on semaglutide. The trials were funded by Novo Nordisk.

Why does weight loss matter so much for atrial fibrillation?

Abed 2013 and LEGACY 2015

Losing weight reduces how often AF happens and how much it disrupts daily life, and that was shown years before GLP-1 drugs were used for obesity. In a randomised trial published in JAMA in 2013, Hany Abed and colleagues in Adelaide assigned 150 overweight or obese adults with symptomatic AF to structured weight management or general lifestyle advice, with both groups getting intensive risk factor care.

Over a median of 15 months, the weight management group lost 14.3 kg against 3.6 kg in controls. Their AF symptom burden scores fell by 11.8 points versus 2.6, and symptom severity scores by 8.4 versus 1.7. On 7-day Holter monitoring, cumulative time in AF fell by 692 minutes in the intervention group and rose by 419 minutes in controls.

The LEGACY cohort from the same team followed 355 people with AF and a BMI of 27 or more. Those who lost 10% or more of their weight had a 6-fold greater probability of arrhythmia-free survival than those who lost less. Weight fluctuation of more than 5% offset part of that benefit, with a 2-fold higher risk of recurrence. For a GLP-1 user, that second finding is the practical one: steady loss that holds appears to matter more than a fast drop that rebounds. Our guide to GLP-1 maintenance dosing covers how people try to keep weight stable after the loss phase.

Can a GLP-1 medicine reduce AF coming back after ablation?

Karakasis 2024 and Guo 2025

Observational studies suggest GLP-1 drug use goes with fewer AF recurrences after catheter ablation, but no randomised trial has confirmed it. A 2024 meta-analysis in Advances in Therapy by Paschalis Karakasis and colleagues pooled three propensity score-matched studies with 6031 people. Over 12 months, GLP-1 use was associated with lower AF recurrence than no GLP-1 use, a hazard ratio of 0.549.

A 2025 prospective study in Circulation: Arrhythmia and Electrophysiology by Jun Guo and colleagues in Shanghai enrolled 437 patients with type 2 diabetes and a BMI of 24 or more after their first AF ablation. The 158 who chose semaglutide had fewer atrial arrhythmia recurrences over 12 months after the 3-month blanking period (hazard ratio 0.68) than the 279 who declined it.

Patients picked their own group in that study, and the semaglutide group started with a slightly higher BMI and HbA1c. Weight change also differed: 8.2% lost on semaglutide versus 4.6% without. That makes it hard to say whether the drug, the weight loss or the type of patient who opts in explains the difference. The authors and the meta-analysis both asked for randomised trials.

Is there data on tirzepatide and atrial fibrillation?

Wu and colleagues, Diabetes Research and Clinical Practice, 2025

Tirzepatide has only observational AF data so far. Jheng-Yan Wu and colleagues used the TriNetX health record network to compare adults with both AF and type 2 diabetes who did or did not use tirzepatide. After propensity matching, 11,194 patients were analysed.

Tirzepatide use was associated with a lower risk of the combined outcome of cardioversion, intravenous antiarrhythmic drugs or AF ablation, a hazard ratio of 0.65. The authors call this a reduced AF burden and ask for randomised trials. Records-based studies can match on many variables, but people prescribed a newer and costlier drug can differ in ways no database captures. For how the two main drugs compare more broadly, see our semaglutide vs tirzepatide breakdown.

Does microdosing a GLP-1 change atrial fibrillation risk?

What the evidence can and cannot support

That has not been studied. STEP-HFpEF used semaglutide 2.4 mg weekly, and the meta-analyses pooled trials run at standard treatment doses without reporting AF results by low dose. There is no data on whether a microdose protocol changes the chance of developing AF, the number of episodes, or recurrence after ablation.

The weight-loss studies hint at a reasonable question, not an answer. Abed and LEGACY showed that sustained weight loss reduces AF burden, and LEGACY found that weight swings worked against it. Whether a smaller dose that produces slower, steadier loss would help rhythm in the same way is unknown.

The practical summary is narrow. GLP-1 medicines do not appear to raise AF risk in trials, semaglutide has been linked to fewer new cases, and losing weight and keeping it off is one of the better-supported ways to reduce AF symptoms. Palpitations, a fluttering or irregular pulse, breathlessness or dizziness deserve an ECG and a clinician's review, whether or not someone is on a GLP-1. Sleep apnea is another common AF driver; our article on GLP-1s and sleep apnea covers that link.

Frequently Asked Questions