Comparative Medical Ecology of Gut Microbiomes in Major Neurodegenerative, Neurodevelopmental, and Psychiatric (NNP) Disorders

This study provides a comprehensive medical ecology analysis of gut microbiome alterations in four neuropsychiatric disorders: Alzheimer’s disease (AD), Parkinson’s disease (PD), autism spectrum disorder (ASD), and the mood disorders (MDs) including major depressive disorder (MDD) (depression) and bipolar disorder (BD). Using diversity, heterogeneity, specificity, and AI/machine learning approaches, we examined microbiota and species-level dysbiosis by reanalyzing approximately 4000 gut microbiome samples from 23 independent studies on the four neuropsychiatric disorders. Key findings include: (i) Alpha-diversity increases across all disorders compared with healthy control, influenced by disease type and health status, while beta-diversity follows the Anna Karenina Principle (AKP), showing healthy individuals are more alike, while diseased individuals are not. Gamma and network diversity vary by disease type and species rarity­dominance spectrum. (ii) Unlike diversity, a common but limited approach ignoring species interactions, heterogeneity—considering both species abundance and interactions—remains understudied. TPLE (Taylor’s power law extension) ranked heterogeneity scaling as MD > ASD > PD ≈ H > AD, while TPLoN (TPL of network) showed an inverse order (H ≈ PD > ASD > MD > AD), except for AD. (iii) The SSD (species specificity and specificity diversity) framework not only identified disease-specific unique/enriched species (US/ES), which are promising candidates for probiotic development, but also identified key biomarker species crucial for the diagnosis of NNP disorders. In terms of species specificity, MD showed either negative correlations (with ASD, PD, and healthy controls) or insignificant correlation (with AD). In contrast, PD and AD were positively correlated but showed no correlation with ASD, potentially due to differences in patient age. These patterns align with intuitive perceptions of NNP disorders, though such perceptions have lacked microbiome-based evidence in existing literature. (iv) AI-Machine learning models achieved 80%-97% precision in diagnosing disorders with as few as top 25 most specific species. These results reveal shared and distinct microbiome features across neurodegenerative, neurodevelopmental, and psychiatric disorders, advancing understanding of their etiological mechanisms. The study highlights the potential of microbiome-based diagnostics and targeted therapies, such as probiotics and diagnostic biomarkers, to inform personalized interventions.