Abstract
BACKGROUND: Bronchial asthma is a heterogeneous inflammatory airway disease with complex etiology. While the respiratory microbiome and host metabolism are implicated, their integrated roles in defining asthma subtypes remain underexplored. METHODS: This multi-omics study utilized 16S rRNA sequencing and untargeted metabolomics on sputum samples from asthma patients and healthy controls. We characterized airway microbial and metabolic profiles, identified asthma subgroups through unsupervised clustering, and investigated microbe-metabolite interactions. RESULTS: Asthma patients exhibited typical clinical hallmarks including elevated IgE and impaired lung function. Microbiome analysis revealed significant enrichment of Streptococcus, Veillonella, and Prevotella as key asthma-associated taxa, alongside dysregulation of crucial lipid metabolic pathways (eg, alpha-linolenic and arachidonic acid). Four highly discriminative diagnostic biomarkers (AUC >0.9), including cis-aconitic acid and LPE 20:1, were identified, indicating specific metabolic perturbations. Unsupervised clustering stratified asthma patients into 2 distinct endotypes: a severe Streptococcus-dominant cluster (Cluster 1) with profound systemic and local metabolic disturbances, and a milder cluster (Cluster 2) with microbial and metabolic profiles resembling healthy controls. These endotypes displayed differential lipid and amino acid metabolism, suggesting unique underlying mechanisms. CONCLUSION: Our findings precisely delineate asthma endotypes driven by distinct microbiome-metabolome interactions, providing novel diagnostic biomarkers and pathway-specific therapeutic targets. This study critically advances our understanding of asthma heterogeneity, highlighting the importance of integrated multi-omics for personalized precision medicine strategies. CLINICAL TRIAL NUMBER: Not applicable.