Gut microbiota and metabolome signatures in preterm infants with high versus low risk for neurodevelopmental impairment: a prospective, matched, longitudinal multi-omics study

Preterm birth is a leading global cause of neurodevelopmental impairment (NDI), yet early predictive biomarkers remain elusive. The gut microbiome, developing in parallel with the brain and communicating via the microbiota-gut-brain axis, holds potential as a source of such biomarkers. However, specific longitudinal multi-omics signatures predictive of NDI risk in preterm infants are poorly defined. We conducted a prospective, matched, longitudinal study of 60 preterm infants, classified at 3 months corrected age (CA) into high-risk (HR, n=30) or low-risk (LR, n=30) groups for NDI based on combined motor (TIMP) and neurological (GMs) assessments. Fecal samples from birth (meconium) and 3 months CA underwent shotgun metagenomic sequencing and untargeted metabolomics. Groups were rigorously matched for gestational age, birth weight, sex, and clinical exposures. While α- and β-diversity did not differ between groups, profound taxonomic and functional divergence emerged. At 3 months CA, the LR gut was enriched with Akkermansia muciniphila , whereas the HR gut was dominated by Klebsiella variicola . Functional metagenomics revealed a dysbiotic HR trajectory, enriching pathways for bacterial virulence, stress response, and—notably—multiple pathways annotated for human neurodegenerative diseases, contrasting with LR expansion of core biosynthesis. Metabolomics confirmed a dysfunctional HR state, showing impaired amino acid metabolism and aberrant neuroactive pathway enrichment. Critically, meconium features correlated with 3-month neurobehavioral scores, demonstrating ultra-early predictive potential. Integrated networks at 3 months directly linked Akkermansia muciniphila and co-varying glycerophospholipids to superior neurodevelopmental scores, forming a beneficial “ Akkermansia -lipid” axis, while Klebsiella variicola and triterpenoids formed a dysbiotic hub. Our study defines a high-risk gut ecosystem trajectory in preterm infants, characterized by early commensal depletion, pathobiont expansion, and a functional shift towards inflammation and neuroinflammation. These signatures offer novel targets for early risk prediction and microbiome-targeted interventions.