Causal relationship between cardiac arrhythmias and sepsis: a two-sample bidirectional mendelian randomization study

OBJECTIVE: This study used a bidirectional two-sample Mendelian randomization (MR) framework to investigate whether genetic liability to cardiac arrhythmias increases the risk of sepsis and whether genetic liability to sepsis increases the risk of cardiac arrhythmias. METHODS: Summary statistics were obtained from publicly available genome-wide association studies in individuals of European ancestry. Cardiac arrhythmia data were derived from FinnGen (32,416 cases and 139,739 controls; IEU ID: finn-b-CARDIAC_ARRHYTM), and sepsis data were derived from UK Biobank (11,643 cases and 474,841 controls; IEU ID: ieu-b-4980). Independent single-nucleotide polymorphisms (SNPs) associated with each exposure at P < 5 × 10^-6 were selected, clumped at r^2 < 0.001 within a 10,000-kb window, harmonized, and filtered to retain instruments with F-statistics > 10. The primary analysis used random-effects inverse-variance weighting (IVW), with weighted median, MR-Egger, simple mode, and weighted mode analyses as complementary methods. Horizontal pleiotropy and outliers were assessed using MR-PRESSO and the MR-Egger intercept. Heterogeneity was evaluated with Cochran's Q test. Because the two causal directions were prespecified, false-positive control was strengthened by combining concordant directional evidence across methods with Benjamini-Hochberg false-discovery-rate assessment for the two primary IVW tests. RESULTS: Genetic liability to cardiac arrhythmias was associated with a higher risk of sepsis (IVW OR = 1.066, 95% CI: 1.007-1.128, P = 0.0268). Genetic liability to sepsis was also associated with a higher risk of cardiac arrhythmias (IVW OR = 1.078, 95% CI: 1.005-1.156, P = 0.0347), although the reverse-direction analysis was based on only three SNPs and should be interpreted cautiously. Effect directions were concordant across complementary MR models, Cochran's Q test did not indicate substantial heterogeneity, MR-Egger intercepts did not support directional pleiotropy, and the two primary IVW tests showed an FDR-adjusted q value of 0.0536. CONCLUSION: The present bidirectional MR analysis supports a robust association from cardiac arrhythmia liability to sepsis risk and suggests a possible reverse association from sepsis liability to cardiac arrhythmias. These findings support closer rhythm and infection surveillance in high-risk patients while highlighting the need for larger and biologically resolved validation studies.