Introduction

Inflammatory bowel disease (IBD) shows significant variations in clinical presentation, prevalence, and epidemiology across the world. We hypothesized that these differences might be, in part, due to regional variations in gut microbiomes. To test this, we examined gut microbiota of IBD patients and healthy controls in Northen part of India, and compared our findings to Dutch patients and population.

Aims & Methods

We collected stool samples and clinical metadata from 226 IBD patients (198 ulcerative colitis, 28 Crohn’s disease) and 66 controls in Northern India. Metagenomic shotgun sequencing was employed to assess gut microbiota diversity, composition, and function. Associations between microbiome profiles and IBD subtypes, along with clinical parameters like medication usage, drug response, and disease activity, were investigated. We compared our findings to a Dutch cohort comprising 436 IBD patients and 903 controls, using a consistent analytical approach. Finally, we used machine learning approaches to evaluate the generalizability of Indian IBD signals to the Dutch population.

Results

Indian IBD patients exhibited reduced microbiome diversity and alterations in beta diversity (FDR < 0.05, replicated in the Dutch cohort). We identified 319 taxa and 71 biochemical pathways significantly altered in Indian patients (FDR < 0.05), with 39% of these signals replicated in the Dutch cohort (at FDR < 0.05). Shared dysbiosis patterns across Indian and Dutch IBD patients included expansion of opportunistic pathogens (e.g., Clostridium, Streptococcus, Lactobacilli genera) and oral bacteria (e.g., Streptococcus oralis, Bifidobacterium dentium), alongside reduction in butyrate producers (e.g., Faecalibacteria, Roseburia hominis, Blautia massiliensis). Furthermore, we identified unique pathobionts specific to the Indian cohort, including Scardovia and Oribacterium genera, Actinomyces dentalis and Klebsiella pneumoniae. Machine learning models trained on the Indian cohort exhibited high predictive accuracy within the Indian test set (Sensitivity 0.84, Specificity 0.95) and demonstrated generalizability to the Dutch cohort (Sensitivity 0.77, Specificity 0.69). Additionally, we found associations between medication use and gut microbiota independent of disease activity: Tofacitinib usage (by 26 UC patients) correlated with decreased Lactobacilliales and increased Clostridia, while Prednisolone and Azathioprine usage associated with reduced Roseburia and increased Streptococcus and Granulicatella.

Conclusion

IBD patients from Northern India exhibit gut dysbiosis patterns similar to European patients, alongside enrichment in India-specific pathobionts. Microbiome-based diagnostic models show generalizability across populations, demonstrating the potential utility of microbiota-based approaches in IBD diagnosis and management across diverse populations.