Causal effects of inflammatory bowel disease on risk of type 2 diabetes: a two-sample multivariable Mendelian randomization study.

ABSTRACT

AIM: This study aimed to explore the causal association between inflammatory bowel disease (IBD) and the risk of type 2 diabetes (T2D) based on a two-sample Mendelian randomization (MR) study. METHODS: Summary single nucleotide polymorphism (SNP)-phenotype association data were obtained from published two genome-wide association studies (GWAS) including SNPs related to IBD, UC, or CD in European participants (n = 71,997) and East Asian participants (n = 16,805). Two GWAS including SNPs associated with T2D included 655,666 Europeans and 433,540 East Asians. A series of screening processes were performed to select qualified instrumental SNPs strongly related to exposure. We applied the inverse variance weighted (IVW), the MR-Egger regression, and the weighted median to estimate the causal effects of IBD, ulcerative colitis (UC) or Crohn' disease (CD) on T2D. Cochran's Q test was conducted to evaluate the statistical heterogeneity between SNPs in the IVW method. The leave-one-out analysis was employed to assess whether the results were caused by any single SNP associated with IBD, UC, or CD. Odds ratio (OR) and 95% confidence interval (CI) were calculated. RESULTS: The IVW results demonstrated that IBD could increase the risk of T2D in the European population (OR = 1.0230, 95%CI 1.0073-1.0390). UC was positively associated with the risk of T2D according to the weighted median (OR = 1.0274, 95%CI 1.0009-1.0546) and IVW (OR = 1.0244, 95%CI 1.0071-1.0421) results in the European population. The IVW results indicated that the CD was positively associated with the risk of T2D in the European population (OR = 1.0187, 95%CI 1.0045-1.0330). In the East Asian population, there are no associations between the IBD, UC, or CD and the risk of T2D (all P > 0.05). MVMR results revealed that the causal effect UC on T2D was still statistically significant after including body mass index (BMI) or low-density lipoprotein (LDL). CONCLUSION: IBD, UC, or CD had causal effects on the risk of T2D in the European population, which might provide evidence for the prevention of T2D in patients with IBD, UC, or CD.