Body Mass Index and the Risk of Atrial Fibrillation: A Mendelian Randomization Study

Abstract

Although observational studies have shown positive associations between body mass index (BMI) and the risk of atrial fibrillation (AF), the causal relationship is still uncertain owing to the susceptibility to confounding and reverse causation. This study aimed to examine the potential causality of BMI on AF by conducting a two-sample Mendelian randomization (TSMR) study.

Methods: The independent genetic variants associated with BMI (n = 303) at the genome-wide significant level were derived as instrumental variables (IV) from the Genetic Investigation of Anthropometric Traits (GIANT) consortium consisting of 681,275 individuals of European ancestry. We then derived the outcome data from a GWAS meta-analysis comprised of 60,620 cases and 970,216 controls of European ancestry. The TSMR analyses were performed in five methods, namely inverse variance weighted (IVW) method, MR-Egger regression, the weighted median estimator (WME), the generalized summary data-based Mendelian randomization (GSMR), and the robust adjusted profile score (RAPS), to investigate whether BMI was causally associated with the risk of AF.

Results: We found a genetically determined 1-standard deviation (SD) increment of BMI causally increased a 42.5% risk of AF (OR = 1.425; 95% CI, 1.346 to 1.509) based on the IVW method, which was consistent with the results of MR-Egger regression, WME, GSMR, as well as RAPS. The Mendelian randomization assumptions did not seem to be violated.

Conclusion: This study provides evidence that higher BMI causally increased the risk of AF, suggesting control of BMI and obesity for prevention of AF.

Keywords: BMI; Mendelian randomization; atrial fibrillation; causal inference.

Figure 1

Forest plot of six Mendelian Randomization estimators of the effect of body mass index on atrial fibrillation. Abbreviations: MR, Mendelian randomization; OR, odds ratio; CI, confidence interval; WMS, weighted median estimator; RAPS, the robust adjusted profile score; GSMR, generalized summary data-based Mendelian randomization; SIMEX, simulation extrapolation.

Figure 2

Scatter plot of SNPs associated with BMI and the risk of AF. The plot related the effect sizes of the SNP−BMI association (x−axis, SD units) and the SNP−AF associations (y−axis, log (OR)) with 95% confidence intervals. The regression slopes of the lines correspond to causal estimates using three Mendelian randomization methods (the inverse variance weighted method, MR-Egger regression, and weighted median estimator). Abbreviations: BMI, body mass index; AF, atrial fibrillation; MR, Mendelian randomization; SNP, single-nucleotide polymorphism.

Figure 3

Funnel plot to assess the robustness. Scattering points represented the effect estimated using a single SNP as an instrumental variable. The vertical lines denoted the overall estimate obtained by the inverse variance weighted estimate and the MR-Egger regression.