Meta-analysis shows reduced oral microbial Shannon diversity in Crohn's disease versus controls

Background: Emerging evidence suggests that the oral microbiome may contribute to aberrant gut immune responses in Inflammatory Bowel Disease (IBD). Methods: We performed a comprehensive, harmonised analysis of aggregated oral microbiome 16S rRNA datasets across multiple cohorts. Data were processed using a unified bioinformatics pipeline including DADA2 for taxonomic assignment, PICRUSt2 for functional prediction, MaAsLin2 for multivariable modelling, and machine learning. Results: Across 25 studies (n = 1,136 IBD; n = 759 controls), meta-analysis showed significantly reduced oral microbial Shannon diversity in IBD (standardised mean difference -0.31, p = 0.007). Secondary bioinformatics analysis of six datasets plus in-house data confirmed this reduction (Shannon diversity; Hedges' SMD = -0.372, p < 0.001), driven primarily by Crohn's disease (CD). Beta diversity demonstrated global compositional shifts, with CD demonstrating greater divergence from controls than ulcerative colitis (UC). Multivariable modelling identified reproducible taxa enriched in IBD, including Corynebacterium, Serratia and Streptoccocus, while Porphyromonas and Ruminococcaceae.G1 were enriched in controls. Functional pathway prediction indicated reduced butyrate metabolism in IBD sub-types and increased aromatic amino acid and related metabolite degradation pathways. Machine learning classifiers achieved modest discrimination (mean AUC ~0.67), supporting the potential of saliva-based microbiome profiling to study dysbiosis in IBD. Conclusions: These findings demonstrate that the oral microbiome in IBD is characterised by reduced diversity and reproducible structural community reorganisation. Together, these data support a contributory role for the oral-gut axis in CD pathogenesis and provide a rationale for targeted mechanistic and longitudinal studies to define causal links between oral dysbiosis and intestinal inflammation.