Exploiting large-scale genetic data to elucidate potential mechanisms between type 2 diabetes and atrial fibrillation

Background Observational studies suggest an association between type 2 diabetes mellitus (T2DM) and atrial fibrillation (AF), but the underlying mechanisms remain unclear. Identifying pathways that elucidate the relationship between T2DM and AF could enhance risk stratification and guide therapeutic strategies. Purpose We aimed to use genetic epidemiological approaches to determine the relevance of distinct pathways underlying the relationship between T2DM and AF. Methods We used clusters of genetic variants associated with T2DM representing different mechanisms from the T2DGGI consortium obtained through k-means clustering to explore the pathways between T2DM to AF. Two-sample MR using the inverse variance weighted method (IVW) assessed the associations between T2DM genetic clusters and AF, using summary data for T2DM from T2DGGI, and for AF from Nielsen (60,620 cases and 970,216 controls) and FinnGen (55,853 cases and 231,952 controls). Sensitivity analyses considering potential pleiotropy included weighted median, MR-Egger regression, and MR-PRESSO approaches. A data-driven method (MRClust) was also used to identify novel clusters of T2DM-associated variants representing distinct causal mechanisms. These clusters were further explored through PheWeb based phenome-wide association studies. To account for multiple testing, significance thresholds of p< 0.007 = (0.05/7) for T2DGGI cluster analyses and p<5×10⁻⁵ for phenome-wide associations. Results Genetically-predicted T2DM was associated with a higher risk of AF consistently across datasets [OR per log odds higher genetically-predicted T2DM =1.06, 95% CI: 1.04-1.08], P<0.001]. Clusters chiefly representing obesity (OR=1.33, 95%CI: 1.28-1.38, P<0.001), body fat (OR=1.07, 95%CI: 1.03-1.13, P=0.003) and residual glycaemia (OR=1.07, 95%CI: 1.03-1.12, P<0.001) were associated with a higher risk of AF. Other clusters showed no significant associations after multiple testing corrections (Figure). Using MRClust, we identified three informative clusters of T2DM variants with heterogeneous effects on AF. One of which, supported a strongly positive causal association with AF and was linked to obesity-related traits. Sensitivity analyses confirmed the robustness of the findings, with consistent estimates across multiple methods. Conclusion Genetically-predicted T2DM was associated with a higher risk of AF, with heterogeneous effects across clusters. T2DM pathways most strongly associated with obesity are likely to be the most influential causal mechanism between T2DM and AF. These findings emphasise the importance of managing obesity to mitigate the risk of AF and highlight the value of clustering methods in improving our understanding of disease mechanisms.