Abstract
The association between vitamin D deficiency (VDD) and incident atrial fibrillation/flutter (AF) remains controversial due to conflicting evidence from cross-sectional studies. This study assessed the association between VDD and incident AF using a large-scale, multi-institutional cohort with propensity score matching. This longitudinal cohort study utilized the TriNetX platform, analyzing individuals aged ≥40 years with 2 serum 25-hydroxyvitamin D measurements between 2010 and 2023. Patients with sustained VDD (<20 ng/mL) were compared to controls (≥30 ng/mL) using 1:1 propensity score matching. The primary outcome was incident AF within 3 years. Dose-response analysis across vitamin D categories and subgroup analyses were performed. After matching, 137,407 patients were analyzed in each group. VDD was significantly associated with higher AF risk (hazard ratio [HR]: 1.83, 95% confidence interval: 1.68-1.99, P < .001). A clear dose-response relationship emerged, with HRs of 1.90 for severe deficiency (<10 ng/mL), 1.94 for moderate deficiency (10-20 ng/mL), and 1.60 for insufficiency (20-30 ng/mL). The association persisted during 3 to 6 year follow-up and was consistent across all AF subtypes. VDD was also associated with increased ischemic stroke and pneumonia risk. Subgroup analyses revealed that dyslipidemia significantly modified the VDD-AF association (HR: 1.84 vs 1.50 without dyslipidemia, P = .017). VDD was significantly associated with increased incident AF risk in a dose-dependent manner. These findings support VDD as a potentially modifiable risk factor and highlight the importance of vitamin D status in comprehensive cardiovascular risk assessment. Further prospective and interventional studies are warranted to validate these associations and clarify the underlying mechanisms.