Wastewater-derived gut microbiome: global patterns and associations with population health

Urban wastewater systems aggregate gut microbiome signals from large populations, enabling exploration of their associations with population-level health patterns. However, it remains unclear whether wastewater-derived gut microbiome (WGM) simply represents a composite of individual gut microbes or whether it forms structured ecological patterns with potential relevance to population-level health-burden gradients. We analyzed 661 untreated wastewater metagenomes from 220 cities across 97 countries spanning all inhabited continents to characterize the global WGM. WGM exhibited broad phylogenetic diversity, with more than 95% of detected genera occurring in over 80% of samples worldwide. Co-occurrence network analysis revealed distinct, densely connected modules and structurally heterogeneous hub and bridging taxa, indicating organized network architecture that was not explained by abundance alone. Under the iCAMP framework, WGM turnover was predominantly classified as dispersal limitation across climate zones, continents, development levels, and sampling years. We then explored associations between WGM and population-level health indicators using a multi-model machine-learning framework. Across 21 health-burden indicators covering non-communicable diseases, infectious diseases, injuries, and mental health conditions, WGM-based classification of low- versus high-burden countries exceeded shuffled-label expectations for 19 indicators. Recurrently selected WGM taxa were identified across multiple feature-selection strategies. Together, these results show that operationally defined WGM profiles show reproducible structure and broad associations with country-level health-burden indicators.