Back to 2026 Display Posters
Neuroactive Functional Dysbiosis of Urinary Microbiome in Pediatric Neurogenic Bladder
Deepthi Ramya Ravindran, M.sc, Sachit Anand, MBBS, MCh, Ajay Verma, MBBS, MCh, Jitendra Kumar Meena, MD, DNB, ACME, Himalaya Kumar, Msc, Pankaj Hari, MD, FIAP, FAMS, Vineet Ahuja, MBBS, MD, DM.
All India Institute of Medical Sciences, New Delhi, India.
BACKGROUND: Pediatric Neurogenic bladder (NB) is characterized by impaired bladder function, urinary stasis and recurrent urinary infections. Increasing evidence suggests that urinary microbiome dysbiosis contributes to disease progression beyond traditional clinical factors. However, microbial dysbiosis that extends beyond taxonomic disruption to functional neuro-metabolic mechanisms remains poorly characterised in NB. We investigated microbiome-derived neuroactive pathways and their role in disease stratification using multivariate analysis and machine learning approaches.
METHODS: A total of 83 children, including NB (n = 39) and age-and gender-matched controls (n = 44), were recruited in this cross-sectional study. NB was further stratified into subgroup A (renal deterioration, n = 15) and subgroup B (no renal deterioration, n = 23). Urinary microbiome profiling was performed using 16S rRNA sequencing. Based on PiCRUST2 functional inference, a panel of 25 neuroactive pathways was identified. Subsequently, the neuroactive pathways were categorised into functional pathway classes. A composite neurodysbiosis index (NDI) was computed. Supervised machine learning models were analysed for Multivariate discrimination using Partial Least Squares Discriminant Analysis (PLS-DA), and feature importance calibration by Random Forest (RF) models.
RESULTS: As compared to controls, heatmap clustering revealed consistent enrichment and upregulation of neuro-metabolic pathways in NB groups. This correlated with enrichment of pathobiont genera and depletion of commensals, indicating coordinated microbial and functional dysbiosis. The associated pathways include aromatic amino acid metabolism (e.g., ARO-PWY, COMPLETE-ARO-PWY), Coenzyme A and lipid metabolism (COA-PWY, PHOSLIPSYN-PWY), Folate and one-carbon metabolism (FOLSYN-PWY, MET-SAM-PWY), Energy and redox pathways (GLYCOLYSIS, fermentation-related pathways), Neurotransmitter-associated and cofactor biosynthesis pathways (TRNA-CHARGING-PWY, NAD biosynthesis). The neurodysbiosis index followed a gradient of C > B > A, with significant differences between A vs C (p = 0.004) and A vs B (p = 0.002), indicating progressive neuro-functional disruption. In supervised machine learning models, PLS-DA demonstrated clear separation between control and NB groups, indicating strong multivariate discrimination. A clear stepwise enrichment of neuroactive pathways was observed (C → B → A). SHAP-like directional analysis confirmed that these pathways contributed positively to disease classification, with stronger effect sizes in A vs C than B vs C, supporting a severity-dependent functional shift. RF models identified key neuroactive pathways driving disease discrimination. Classification performance was high in pathway-based model AUC ≈ 0.88, Combined features AUC ≈ 0.98, Subgroup discrimination at A vs Rest (AUC ≈ 0.91), B vs Rest (≈ 0.85), C vs Rest (≈ 0.89) respectively.
CONCLUSIONS: Along with microbial dysbiosis, pediatric NB is characterised by coordinated enrichment and upregulation of neuroactive pathways spanning amino acid metabolism, lipid biosynthesis, and one-carbon metabolism. These pathways demonstrate a severity-dependent gradient and provide strong discriminatory power in machine learning models. PLS-DA and RF models showed a robust neuro-metabolic signature linked to underlying disease progression, positioning neuroactive functional characteristics as key microbial biomarkers for stratification and mechanistic insight.
Back to 2026 Display Posters
