Metagenomic characterization of gut microbiota in patients with Disorders of Gut–Brain Interaction

Abstract

Patients with Disorders of Gut–Brain Interaction (DGBI) experience recurrent gastrointestinal symptoms without identifiable organic abnormalities. While earlier studies focused on neural and hormonal regulation, recent evidence suggests that gut microbiota dysbiosis and altered microbial metabolism may play key roles in DGBI. A total of 86 DGBI patients meeting the Rome V criteria were enrolled. Demographic data, disease duration, and symptom scores (Wexner Constipation Score, Self-Rating Anxiety Scale, and Abdominal Bloating Score) were collected. Fecal samples were subjected to metagenomic sequencing, followed by diversity, taxonomic, and functional analyses to characterize the gut microbiota profile of DGBI patients. DGBI patients and healthy controls (HC) showed comparable α-diversity and no significant overall separation in β-diversity. At the genus level, Prevotella was more abundant in HC, whereas other dominant genera showed similar patterns between groups. Co-occurrence network analysis revealed altered microbial associations in DGBI. A species-level random forest model moderately discriminated DGBI from HC (AUC = 0.769). Among DGBI subtypes, FD showed lower species richness, while IBS-M exhibited higher microbial evenness and diversity. Subtype-specific microbial signatures were identified, including Ruminococcus gnavus in IBS-D, Prevotella stercorea and Asaccharobacter celatus in IBS-C, and Victivallis vadensis in FC. Moreover, exploratory analysis showed a negative correlation between Asaccharobacter celatus abundance and Wexner constipation scores in the IBS-C subtype. In conclusion, this study showed that patients with DGBI exhibited limited changes in overall gut microbial diversity compared with healthy individuals, but certain differences in microbial community patterns were observed across clinical subtypes, suggesting potential microbial ecological heterogeneity among different DGBI phenotypes.