In a database study of more than 75,000 people with chronic kidney disease, the ones taking a GLP-1 drug developed a common irregular heart rhythm far less often than the ones who were not. Atrial fibrillation showed up in 6.9 percent of the GLP-1 users over three years, against 12.4 percent of the non-users. That works out to a hazard ratio of 0.59, roughly a 41 percent lower risk.
The finding was published July 3 in Cardiovascular Drugs and Therapy ↗. The authors used TriNetX, a large US network of hospital and clinic records. They pulled patients with chronic kidney disease (kidneys filtering below normal) who had never had atrial fibrillation or kidney failure. Atrial fibrillation is a fast, disorganized heartbeat that raises the risk of stroke and heart failure. It is one of the most common rhythm problems in people with weak kidneys.
The GLP-1 drugs ↗ here are the class sold as Ozempic, Wegovy, and Mounjaro, which act on the GLP-1 receptor ↗ and are prescribed mainly for type 2 diabetes and obesity. To compare like with like, the team used propensity score matching, pairing each of 37,768 GLP-1 users with a non-user of similar age, sex, race, and baseline heart and kidney risk. The average patient was 67, about 63 percent were diabetic, and the average body mass index was 34.
The benefit did not need diabetes or extra weight
The reason this is more than one more line in the GLP-1 benefit column is what happened when the authors split the patients up. Because these drugs are given for diabetes and obesity, the easy explanation for any heart effect is that it simply follows from better blood sugar or lost weight. But the atrial fibrillation reduction held in people who had neither.
Among patients without diabetes, the hazard ratio was 0.42, a 58 percent lower risk. Among patients who were not obese, it was 0.63. Among those who were neither diabetic nor obese, it was 0.59. The benefit showed up whether or not the two conditions the drugs are marketed for were present.
The other outcomes moved the same way. Heart failure flare-ups came with a hazard ratio of 0.48. Major adverse cardiovascular events, 0.58. Procedures used to manage atrial fibrillation, like ablation and cardioversion, 0.58. Progression to kidney failure, 0.70. Death from any cause, 0.52. Every endpoint pointed in the same direction and most of the confidence intervals were tight, which is what a cohort this size buys.
What the number can and cannot say
This is an observational study, not a randomized trial, so it shows an association and not proof. Matching narrows the gap between the two groups but cannot close it. Patients who get prescribed a newer, more expensive drug tend to be more engaged with their care, and some of the difference could ride on that rather than on the molecule.
The sentence worth arguing with is the non-diabetic figure. A hazard ratio of 0.42 in people without diabetes says something specific. If a GLP-1 effect on the heart's electrical stability is real and independent of weight and glucose, then the mechanism is not the obvious metabolic one. No large trial has yet been built around atrial fibrillation as its main endpoint in kidney disease. A 75,000-patient real-world cohort is the kind of thing that raises that question faster than a trial can answer it.
It fits a pattern our own coverage has been tracking. In June a separate cohort found GLP-1 heart protection held across every level of kidney function ↗, with the worst kidneys gaining the most. The rhythm data adds a specific electrical endpoint to that cardiorenal picture, and it does it in the group where the metabolic explanation runs out.