SLIT2 and ANGPTL3 as putative mediators linking obesity to atrial fibrillation: A Mendelian randomization study

The observational association between obesity and atrial fibrillation has been well-established. However, its causal mediators underlying this association remain unclear. This study aimed to investigate the causal relationship between body mass index (BMI), as well as potential mediating factors underlying this association, using Mendelian randomization (MR). Genetically predicted BMI was significantly associated with an increased risk of atrial fibrillation (inverse-variance weighted odds ratio [OR] = 1.350, 95% confidence interval [CI] = 1.268-1.436, P < .001). Sensitivity analyses using multiple methods yielded consistent results, and no evidence of directional pleiotropy was detected after outlier correction. A total of 91 inflammatory cytokines and 4907 plasma proteins were included as potential mediators. Among the screened mediators, SLIT2 exhibited a significant inverse mediating effect on the association between BMI and atrial fibrillation (OR = 1.019, 95% CI = 1.005-1.034), accounting for 7% of the total effect. These results suggest that higher BMI is associated with reduced SLIT2 expression, which in turn increases the risk of atrial fibrillation. In contrast, ANGPTL3 showed a significant positive mediating effect (OR = 1.010, 95% CI = 1.003-1.017), accounting for 3.64% of the total effect, indicating that elevated BMI increases ANGPTL3 levels, thereby further increasing the risk of atrial fibrillation. This MR study provides genetic evidence supporting a causal association between obesity and atrial fibrillation and suggests that SLIT2 and ANGPTL3 may be involved as mediators linking obesity-related inflammation and lipid dysregulation to atrial fibrillation susceptibility. First, a 2-sample MR analysis was conducted using genome-wide association studies meta-analysis of European ancestry to estimate the causal effect of genetically proxied BMI on atrial fibrillation. Subsequently, inflammatory cytokines and plasma proteins were considered as candidate mediators, with their cis-protein quantitative trait locus used as instrumental variables. The inverse-variance weighted method was applied as the primary MR approach, while MR-Egger, weighted median, weighted mode, MR-Pleiotropy Residual Sum and Outlier, and leave-one-out methods were used for sensitivity analyses. Shared causal variants were further evaluated using false discovery rate correction and colocalization analysis within a ± 500 kb window (PPH4 > 0.75).