Matched health-record study associates bariatric surgery with fewer new atrial arrhythmias

What happened

A large health-record analysis found fewer newly documented cases of atrial fibrillation or atrial flutter after metabolic and bariatric surgery than in a matched group that did not have surgery. The same analysis found lower all-cause mortality in the surgical group, but it was observational and cannot establish cause and effect.

Researchers used the TriNetX Global Collaborative Network, which draws de-identified electronic records from 165 health care organizations. They identified adults with severe obesity and excluded people with a previous arrhythmia. Before matching, the database included about 39,000 surgical patients and more than 3.2 million nonsurgical patients. The investigators then matched 38,953 people in each group using demographic and known arrhythmia-risk variables.

During follow-up, records showed atrial fibrillation or flutter for 2.1% after surgery, versus 3.3% among controls. Researchers reported a hazard ratio of 0.73 (95% confidence interval, 0.67 to 0.79). This is a relative time-to-event estimate, not a statement that surgery prevents 27 of every 100 cases.

Recorded deaths from any cause occurred in 0.5% of surgical patients and 2.8% of controls. The hazard ratio was 0.24, with a 95% confidence interval of 0.20 to 0.28. The study did not find statistically clear differences in ventricular tachycardia, ventricular fibrillation, or cardioversion. Those events were less common, and the ventricular-fibrillation estimate was especially imprecise.

What it means

The study adds a large, international electronic-record analysis to evidence connecting bariatric surgery with cardiovascular outcomes. Its atrial-arrhythmia result remained after the researchers balanced the groups on measured factors included in their matching model.

Absolute numbers matter alongside relative estimates. Newly recorded atrial events differed by 1.2 percentage points between the matched cohorts. The mortality gap was larger, but such a strong observational association also calls for careful attention to differences that a health-record database may not capture.

What it does not mean

This was not a randomized trial. People who undergo surgery differ from people receiving nonsurgical care in ways that can include health engagement, access to specialists, operative eligibility, socioeconomic circumstances, and follow-up. Propensity matching addresses recorded variables, not every possible source of confounding.

Electronic records can also miss diagnoses, outside care, medication use, and deaths. The study does not tell readers whether one operation is better than another for rhythm outcomes, and it does not justify surgery solely to prevent atrial fibrillation.

The neutral findings for ventricular rhythms should not be read as proof of no effect. Their confidence intervals allow both benefit and harm, especially for ventricular fibrillation. The appropriate treatment decision still depends on overall surgical indications, expected benefits, operative risk, and patient priorities.