Oral-gut microbiome and plasma metabolome reveal incremental restratification of coronary artery disease from the clinical gray zone

ABSTRACT: Clinical prediction models have been widely applied for risk stratification of coronary artery disease (CAD), yet approximately 20% to 40% of screened individuals fall into the intermediate-risk interval and face a decision gap. Here, we adopted a two-stage validation design. In the first stage, a clinical prediction model was developed in a retrospective cohort of 1,500 patients, and predicted probabilities of 0.3 to 0.7 were defined as the clinical gray zone. In the second stage, tongue surface and fecal samples from a prospective cohort of 105 patients were analyzed by shotgun metagenomic sequencing, and fasting plasma was profiled by untargeted metabolomics. The oral and gut microbiomes maintained independent community structures in CAD, jointly explaining only 0.4% of the variation in plasma metabolites. The incremental effects of multi-omics were concentrated in gray-zone patients, with the gray-zone AUC increment being 1.63-fold and 1.93-fold the whole-population increment for CHD detection and ACS warning, respectively, and 88.9% and 75.9% of patients being correctly reclassified. Omics requirements differed across disease stages, with the metabolome alone providing incremental value in chronic coronary syndrome and combined microbiome-metabolome analysis being required for ACS. The nomogram constructed from selected variables showed robust discriminative performance. These findings validate that the incremental clinical value of multi-omics is concentrated within the uncertain interval of the clinical model and provide an evidence base for precise restratification of patients in the clinical gray zone.

INSTRUMENT(S): Liquid Chromatography MS - negative - reverse-phase, Liquid Chromatography MS - positive - reverse-phase

PROVIDER: MTBLS15425 | MetaboLights | 2026-08-31

REPOSITORIES: MetaboLights