Multi-omics landscape and machine learning predictors of acute and chronic coronary syndrome diagnosis in young patients

BACKGROUND: Acute coronary syndrome (ACS) is a leading global cause of death, and its incidence is increasingly rising in young adults, who exhibit distinct clinical characteristics from elderly patients. However, multi-omics studies focusing specifically on young coronary heart disease (CHD) patients remain scarce, hindering precise diagnosis and mechanism exploration. METHODS: Here, we enrolled 206 young chest pain patients (18-45 years old), including 122 ACS patients, 38 chronic coronary syndrome (CCS) patients, and 46 individuals with healthy coronary arteries (NC). We performed integrated analyses of peripheral blood mononuclear cell transcriptomics, serum metabolomics, stool metabolomics, and gut microbiome metagenomics to characterize CHD subtypes and develop targeted diagnostic tools. RESULTS: Our results showed that single omics layers had limited ability to distinguish CHD subtypes, while multi-omics integration significantly improved diagnostic efficacy. We identified unique molecular signatures for different subtypes: STEMI was associated with abnormal amino acid and carbohydrate metabolism, CCS was dominated by amino acid metabolism disturbances, and both STEMI and ACS showed enriched inflammation-related pathways. Novel biomarkers including p-chlorobenzene sulfonamide, cotinine, and the gut bacterium Streptococcus parasanguinis were identified, with Streptococcus parasanguinis validated as an atherogenic pathogen in a murine model. We constructed three multi-omics fusion diagnostic models (ACS vs. NACS, CCS vs. NC, STEMI vs. NSTE-ACS) with AUC values of 0.99, 0.95, and 0.96, respectively, and integrated them into a comprehensive diagnostic pipeline. Furthermore, multi-omics functional analysis unraveled a synergistic "microbiota-metabolism-immunity" regulatory network underlying CHD subtypes, linked to disordered amino acid and carbohydrate metabolism and aberrant inflammatory activation. CONCLUSION: This study provides a systematic molecular landscape of young CHD, a high-precision diagnostic strategy, and novel targets for mechanism research and targeted intervention, addressing the unmet clinical need for precise management of young CHD patients.