Clinical Subtype Stratification Reveals Gut Microbial Alterations, Candidate Markers, and Co-Occurrence Network Remodeling in Coronary Artery Disease

Although gut microbiota dysbiosis is associated with coronary artery disease (CAD), its dynamic alterations across distinct clinical stages remain largely unexplored. This study aimed to identify stage-specific microbial biomarkers across the CAD spectrum. We analyzed 16S rRNA sequencing data from 306 subjects, including non-CAD controls and patients with mild coronary stenosis (MCS), stable angina (SA), unstable angina (UA), and acute myocardial infarction (AMI). Diversity analyses, machine learning, and SPIEC-EASI network modeling were applied to evaluate microbial community shifts and identify core diagnostic markers. Significant structural remodeling of the gut microbiota was observed, particularly during the AMI stage. Veillonella, Intestinimonas, and Streptococcus were identified as the core biomarkers for the MCS, SA, and UA/AMI stages, respectively. From the MCS to the UA stage, ecological network connectivity continuously declined, highlighting critical barrier vulnerability, before transitioning into an antagonism-dominated network in the AMI stage. Machine learning models effectively discriminated among the various disease stages, and key microbial markers remained statistically robust after adjusting for major clinical confounders. The gut microbiota exhibits dynamic, stage-specific dysbiosis throughout CAD progression, providing promising non-invasive biomarkers for precise disease staging and diagnosis.