Meta-analysis of the effects of proton pump inhibitors on the human gut microbiota

Abstract

Abstract Mounting evidence has linked changes in human gut microbiota to proton pump inhibitor (PPI) use. Accordingly, multiple studies have analyzed the gut microbiomes of PPI users, but PPI–microbe interactions are still understudied. Here, we performed a meta-analysis of four studies with available 16S rRNA gene amplicon sequencing data to uncover the potential changes in human gut microbes among PPI users. Despite some differences, we found common features of the PPI-specific microbiota, including a decrease in the Shannon diversity index and the depletion of bacteria from the Ruminococcaceae and Lachnospiraceae families, which are crucial short-chain fatty acid-producers. Through training based on multiple studies, using a random forest classification model, we further verified the representativeness of the six screened gut microbial genera and 20 functional genes as PPI-related biomarkers, with AUC values of 0.748 and 0.879, respectively. Functional analysis of the PPI-associated 16S rRNA microbiome revealed enriched carbohydrate- and energy-associated genes, mostly encoding fructose-1,6-bisphosphatase and pyruvate dehydrogenase, among others. In this study, we have demonstrated alterations in bacterial abundance and functional metabolic potential related to PPI use, as a basis for future studies on PPI-induced adverse effects.

Keywords

• Gut microbiota
• Proton pump inhibitors
• Meta-analysis
• Supervised learning technique
• Biomarkers