Investigating the causal relationship of gut microbiota with GERD and BE: a bidirectional mendelian randomization

Abstract

Background Gut microbiota(GM) have been proven associated with lots of gastrointestinal diseases, but its causal relationship with Gastroesophageal reflux disease(GERD) and Barrett’s esophagus(BE) hasn’t been explored. We aimed to uncover the causal relation between GM and GERD/BE and potential mediators by utilizing Mendelian Randomization(MR) analysis. Methods Summary statistics of GM(comprising 301 bacteria taxa and 205 metabolism pathways) were extracted from MiBioGen Consortium(N = 18,340) and Dutch Microbiome Project(N = 7,738), GERD and BE from a multitrait meta-analysis(NGERD=602,604, NBE=56,429). Bidirectional two-sample MR analysis and linkage disequilibrium score regression(LDSC) were used to explore the genetic correlation between GM and GERD/BE. Mediation MR analysis was performed for the risk factors of GERD/BE, including Body mass index(BMI), weight, type 2 diabetes, major depressive disorder(MDD), smoking initiation, alcohol consumption, and dietary intake(including carbohydrate, sugar, fat, protein intake), to detect the potential mediators between GM and GERD/BE. Results 11 bacterial taxa and 13 metabolism pathways were found associated with GERD, and 18 taxa and 5 pathways exhibited causal relationship with BE. Mediation MR analysis suggested weight and BMI played a crucial role in these relationships. LDSC identified 1 taxon and 4 metabolism pathways related to GERD, and 1 taxon related to BE. Specie Faecalibacterium prausnitzii had a suggestive impact on both GERD(OR = 1.087, 95%CI = 1.01–1.17) and BE(OR = 1.388, 95%CI = 1.03–1.86) and LDSC had determined their correlation. Reverse MR indicated that BE impacted 10 taxa and 4 pathways. Conclusions This study established a causal link between gut microbiota and GERD/BE, and identified the probable mediators. It offers new insights into the role of gut microbiota in the development and progression of GERD and BE in the host.