Causal relationship between ulcerative colitis and male infertility: A two-sample Mendelian randomization study

To explore the causal relationship between ulcerative colitis (UC) and male infertility using Mendelian randomization method with single nucleotide polymorphism (SNP) as the instrumental variables.

Genetic loci closely associated with UC were extracted as instrumental variables and male infertility was the outcome variable in pooled data from the gene-wide association study (GWAS),which was derived from European ethnic groups. The UC data(ebi-a-GCST003045) contained a total sample size of 27432 individuals and 110944 SNPs, and the male infertility data(finn-b-N14_MALEINFERT) contained a total sample size of 73479 individuals and 16377329 SNPs. The SNPs highly correlated with UC were screened from ebi-a-GCST003045(P<5×10-8 as the screening condition, the linkage disequilibrium coefficient was 0.001,and the width of the linkage disequilibrium area was 10000 kb).SNPs related to male infertility from finn-b-N14_MALEINFERT (the minimum r2>0.8,replacing the missing SNPs with SNPs with high linkage, and deleting SNPs without substitution sites) were extracted. MR analysis was performed using MR-Egger regression, the weighted median and the inverse-variance weighted (IVW) respectively, and the causal relationship between UC and male infertility was evaluated by OR and 95% CI, and the Egger-intercept method was used to test for horizontal multiplicity, and the sensitivity analysis was performed using "leave-one-out method". Finally, we used Bayesian Weighted Mendelian Randomization (BWMR) approach to test the results of MR study.

A total of 86 SNPs were included as IVs, with OR and 95% CI of 1.095(0.820~1.462)、1.059(0.899~1.248)、1.125(1.002~1.264) for MR-Egger, the weighted median and IVW results respectively, and P value of less than 0.05 for IVW, indicating that a causal relationship between UC and male infertility was causally related. The results of MR analysis combined with BWMR analysis also showed positive genetic causal relationship between UC and male infertility.MR-Egger regression showed an intercept of -2.21×10-3 with a standard error of 0.006 and P = 0.751, there was no horizontal pleiotropy for the IVs of exposure factors. Heterogeneity tests showed no heterogeneity and the results of the "leave-one-out" sensitivity analysis were stable.

There is a causal association between UC and male infertility, which increases the risk of developing male infertility.