Abstract
Obstructive sleep apnea (OSA) has been widely associated with DN in observational studies; however, the causal nature and direction of this association remain uncertain. To clarify this, we performed a bidirectional two-sample Mendelian randomization (MR) analysis using large-scale genetic datasets. Genetic instruments for OSA were derived from a genome-wide association study (GWAS) comprising up to 476,853 individuals, while genetic associations for DN were obtained from another GWAS including 452,280 participants. We applied multiple MR methods to ensure robust causal inference, including inverse variance weighting, MR-Egger regression, weighted median, weighted mode, and simple mode. Sensitivity analyses were thoroughly conducted using Cochran's Q test for heterogeneity and the MR-Egger intercept test to assess potential pleiotropy. Furthermore, multivariable MR was employed to evaluate the independent effect of OSA on DN after adjusting for hyperlipidemia and hypertension. The results indicated a significant causal effect of OSA on DN risk, supported by IVW (OR = 1.41, 95% CI: 1.12-1.77, p = 0.003) and weighted median estimates (OR = 1.57, 95% CI: 1.16-2.13, p = 0.003). Reverse MR analysis revealed no evidence of a causal effect of DN on OSA. Importantly, after accounting for hyperlipidemia and hypertension, multivariable MR confirmed that OSA exerts an independent causal influence on DN (OR = 0.90, 95% CI: 0.14-1.67, p = 0.021). These findings suggest that OSA may contribute to the pathogenesis of diabetic nephropathy through specific mechanisms, independent of traditional metabolic risk factors.