BMI discrepancy and new-onset atrial fibrillation: findings from UK biobank

Background/Introduction Obesity is a well-recognized risk factor for atrial fibrillation (AF), but the role of discrepancies between genetically predicted and observed body mass index (BMI) in AF development is not well understood. Purpose We investigated the association between BMI discrepancy and the risk of new-onset AF. Methods In the UK Biobank, 118,945 participants free of AF at baseline and without extreme BMI values were included. BMI discrepancy was defined as genetically predicted BMI minus observed BMI. Incident AF was the primary outcome. Associations were assessed using Cox proportional hazards regression and restricted cubic spline models. Results The study population (mean age 56.8 ± 8.0 years; 45.6% men; average observed BMI 27.2 ± 4.3 kg/m²) was divided into quintiles of BMI discrepancy. Mean values ranged from -6.1 ± 2.6 kg/m² in the lowest quintile (more obese than predicted) to 5.5 ± 1.5 kg/m² in the highest quintile (leaner than predicted). Compared with the fourth quintile, the fifth quintile showed a significantly elevated likelihood of AF (adjusted HR 1.22, 95% CI 1.10–1.35, P<0.001). The spline analysis indicated a J-shaped curve, with the lowest AF risk observed when measured BMI was modestly leaner—around 1 kg/m²—than the genetic estimate. In contrast, participants whose BMI was substantially lower than predicted (greater than 5 kg/m²) demonstrated heightened AF susceptibility. Conclusion AF incidence exhibited a J-shaped association with BMI discrepancy. Individuals slightly leaner their genetically predicted BMI had the lowest risk, whereas those who were markedly leaner faced considerably increased risk. These findings suggest that both genetic predisposition and attained body weight are essential factors to consider in AF risk assessment.