Causal relationship between eosinophilic esophagitis and inflammatory bowel disease: a bidirectional two-sample Mendelian randomization study.

BACKGROUND

Eosinophilic esophagitis (EoE) and inflammatory bowel diseases (IBDs), including Crohn's disease (CD) and ulcerative colitis (UC), are immune-mediated gastrointestinal diseases with overlapped pathogenesis and are sometimes concurrently diagnosed, but their causal relationship remains unclear. We investigated the causal relationship between EoE and IBD and its subtypes via a two-sample bidirectional Mendelian randomization (MR) approach.

METHODS

MR analyses were performed using summary data of a genome-wide association study (GWAS) on individuals of European ancestry. Independent single-nucleotide polymorphisms correlated with EoE (from a GWAS meta-analysis containing 1,930 cases and 13,634 controls) and IBD (from FinnGen GWASs containing 9,083 IBD, 2,033 CD, and 5,931 UC cases, and GWASs of IBD genetic consortium containing 12,882 IBD, 6,968 UC, and 5,956 CD cases) were selected as instruments. We applied the inverse variance weighted (IVW) method as the primary analysis followed by several sensitivity analyses. For the forward MR study, estimates from IVW methods were subsequently meta-analyzed using a random-effect model.

RESULTS

Our results suggested a causal effect of EoE on IBD [pooled odds ratio (OR), 1.07; 95% confidence interval (CI), 1.02-1.13] and EoE on UC (pooled OR, 1.09, 95% CI, 1.04-1.14). No causal link between EoE and CD was observed (pooled OR, 1.05; 95% CI, 0.96-1.16). The reverse MR analyses revealed no causal effect of IBD (and its subtypes) on EoE. Sensitivity analyses confirmed the robustness of primary results.

CONCLUSIONS

Our findings provided evidence of a suggestive causal effect of EoE on IBD (specifically on UC) in the European population. Increased awareness of concurrent or subsequent IBD in patients with EoE is called for. Still, the present evidence is not adequate enough and ought to be validated by further investigations.