Methane and fatty acid metabolism pathways are predictive of Low-FODMAP diet efficacy for patients with irritable bowel syndrome

Abstract

Identification of microbiota-based biomarkers as predictors of low-FODMAP diet response and design of a diet recommendation strategy for IBS patients. We created a compendium of gut microbiome and disease severity data before and after a low-FODMAP diet treatment from published studies followed by unified data processing, statistical analysis and predictive modeling. We employed data-driven methods that solely rely on the compendium data, as well as hypothesis-driven methods that focus on methane and short chain fatty acid (SCFA) metabolism pathways that were implicated in the disease etiology. The patient's response to a low-FODMAP diet was predictable using their pre-diet fecal samples with F1 accuracy scores of 0.750 and 0.875 achieved through data-driven and hypothesis-driven predictors, respectively. The fecal microbiome of patients with high response had higher abundance of methane and SCFA metabolism pathways compared to patients with no response (p-values<6×10-3). The genera Ruminococcus 1, Ruminococcaceae UCG-002 and Anaerostipes can be used as predictive biomarkers of diet response. Furthermore, the low-FODMAP diet followers were identifiable given their microbiome data (F1-score of 0.656). Our integrated data analysis results argue that there are two types of patients, those with high colonic methane and SCFA production, who will respond well on a low-FODMAP diet, and all others, who would benefit a dietary supplementation containing butyrate and propionate, as well as probiotics with SCFA-producing bacteria, such as lactobacillus. This work demonstrates how data integration can lead to novel discoveries and paves the way towards personalized diet recommendations for IBS.