Allergic diseases and T2DM: a bidirectional multivariable Mendelian randomization study and mediation analysis

Background Clinical studies involving observation have uncovered a mutual relationship between allergic disorders and diabetes, yet the precise causal link remains undetermined. Methods We conducted two-sample bidirectional Mendelian randomization analyses using single nucleotide polymorphisms (SNPs) associated with allergic conditions (asthma, allergic rhinitis, atopic dermatitis) from genome-wide studies and SNPs related to type 2 diabetes from FinnGen. Initially, we evaluated the causal link between allergic disorders and type 2 diabetes through a univariate Mendelian randomization study, incorporating inverse variance weighting, MR-Egger, and the weighted median estimator. To address potential confounding, we employed multivariate Mendelian randomization. Finally, we validated mediators influencing the correlation between asthma and type 2 diabetes. Results The Inverse variance weighted method showed that asthma genetically increased the risk of type 2 diabetes [Asthma-type 2 diabetes: beta(95%CI)=0.892 (0.152-1.632), p = 0.018]. Allergic rhinitis and type 2 diabetes exhibit a mutual protective effect: beta(95% CI)=-1.333 (-2.617 to -0.049), p = 0.042; type 2 diabetes-Allergic rhinitis: beta(95%CI)= -0.002 (-0.004 to -0.000), p = 0.018. The Multivariable Mendelian randomization study results showed that after excluding confounding factors, asthma still demonstrates statistical significance in relation to type 2 diabetes. Through mediation analysis, it was discovered that lung function and the percentage of monocytes in leukocytes exert an inhibitory effect on the mediation between asthma and type 2 diabetes. Conclusion The Multivariable Mendelian randomization study indicates asthma as a risk factor for type 2 diabetes. Lung function, and the percentage of monocytes in leukocytes, play an inhibitory role in asthma and type 2 diabetes mediating effects.