Prevotella stercorea defines dual ecological pathways linking the gut microbiome to infection susceptibility
Description
Research hypothesis We hypothesised that Prevotella stercorea, previously associated with reduced infection risk in Gambian children, exerts protection through two mechanistically distinct ecological pathways: a community-mediated pathway operating through microbiome richness and colonisation resistance, and a species-autonomous pathway independent of community structure. We further hypothesised that expression of the species-autonomous pathway is conditioned by host immune-metabolic reserve, operationalised using weight-for-age z-score (WAZ), and reflected in host inflammatory phenotypes. What the data shows This dataset contains species-level gut microbiome taxonomic abundance data, clinical metadata, anthropometric measures, inflammatory biomarkers, and adverse event records from 633 children aged 6–35 months enrolled in the IHAT-GUT randomised controlled trial (NCT02941081) in rural Gambia. Stool samples were collected at Days 1, 15, and 85 of follow-up. Adverse events, including acute respiratory infection (ARI), diarrhoea, fever, and infection-related illness, were recorded prospectively over 113 days. Systemic inflammatory biomarkers include C-reactive protein (CRP), alpha-1-acid glycoprotein (AGP), and faecal calprotectin. Notable findings P. stercorea positively predicts gut microbiome richness across all timepoints, consistent with a role in community assembly and colonisation resistance. However, its association with reduced ARI frequency (IRR = 0.946, p = 0.002) persists unchanged after richness adjustment, while richness itself is not associated with ARI, supporting a richness-independent, species-autonomous pathway. In contrast, diarrhoeal outcomes show directionally consistent evidence of partial richness mediation. A co-occurring congener, P. copri, does not retain an independent ARI association in joint models despite comparable associations with richness, indicating species specificity. Host context modifies these relationships: in lower-WAZ children, P. stercorea associates with reduced CRP and reduced ARI/fever burden, whereas in higher-WAZ children, baseline inflammatory tone predicts subsequent colonisation, consistent with host-context-dependent immune coupling. Data interpretation and use Species-level microbiome abundances are expressed as relative abundances and should be log-transformed (log₁+x) prior to regression modelling. Species richness is defined as the number of taxa with non-zero abundance per sample. Adverse event frequencies and durations were truncated at 5 episodes and 30 days, respectively, to reduce outlier influence. The complete reproducible analysis pipeline, including R scripts, model specifications, and mediation analyses, is available at: https://github.com/ofordile-star/IHAT_Pstercorea