Reassessing the association between celiac disease and cardiovascular events: A large propensity-matched analysis...

Celiac disease is an immune-mediated enteropathy increasingly recognized as a systemic disorder with potential cardiovascular involvement. Prior studies suggested a modestly higher risk of atrial fibrillation and cardiovascular disease, though traditional risk factors such as hypertension, hyperlipidemia, and obesity are less common in this population. We aimed to evaluate the independent association between celiac disease and cardiovascular outcomes using a large real-world database.

We used the TriNetX Global Collaborative Network, comprising anonymized electronic health records (EHR) from over 150 million individuals worldwide. Adults (≥18 years) with celiac disease (ICD-10 K90.0) were compared with controls without celiac disease. Propensity-score matching (1:1) was applied on demographics, comorbidities, laboratory parameters, and cardiovascular medications. Outcomes included all-cause death, acute myocardial infarction, ischemic stroke or thromboembolism, incident atrial fibrillation, ventricular tachycardia, ventricular fibrillation, or cardiac arrest, incident heart failure, acute pulmonary edema or cardiogenic shock, myocarditis, and pericarditis.

After propensity-score matching, 41,071 patients per cohort were followed for a mean of 4.2 ± 1.5 years. Compared with matched controls, patients with celiac disease had lower risks of all-cause death (HR 0.75, 95% CI 0.71-0.80), acute myocardial infarction (HR 0.76, 95% CI 0.67-0.88), and incident heart failure (HR 0.91, 95% CI 0.85-0.98). There were no significant differences between groups for ischemic stroke or thromboembolism, atrial fibrillation, tachycardia/ventricular fibrillation or cardiac arrest, acute pulmonary edema or cardiogenic shock, myocarditis or pericarditis.

In this large, contemporary EHR-based analysis, celiac disease was not associated with excess cardiovascular risk after adjustment for comorbidities and risk factors, suggesting that prior associations likely reflected residual confounding rather than a direct pathogenic link.