Abstract
BACKGROUND AND AIMS: The comorbid state of obstructive sleep apnea (OSA) and coronary artery disease (CAD) has emerged as a significant public health challenge. However, the underlying metabolic regulation and molecular biomarkers linking to the CAD and OSA comorbidity (CADOSA) remain poorly understood. This study aimed to identify non-invasive biomarkers for the CADOSA comorbidity. METHODS: We randomly recruited a total of 143 participants from a hospital-based setting, comprising four groups: healthy controls (n=38), patients with OSA-only (n=37), CAD-only (n=36), and comorbid CADOSA (n=32). Participants were predominantly male (69.2%), with a mean age of 59.44 years (SD 12.25) and 24.77±3.94 BMI index. Comprehensive clinical evaluation and blood metabolomic profiling were conducted. We integrated differential expression analysis, functional enrichment, machine learning, receiver operating characteristic (ROC) curve, and matched transcriptomes association analysesto uncover key metabolites and potential molecular regulatory mechanisms associated with CADOSA. RESULTS: Baseline clinical data revealed that CADOSA comorbidity significantly exacerbates more severe cardiac dysfunction and sleep disordered breathing than two unique disease stages, despite no significant differences in age or BMI among the groups. Common disordered metabolites in patients with OSA-only or CAD-only were primarily associated with fatty acid, steroid, amino acid, and energy metabolism. While OSA-specific metabolic pathways included ascorbate and aldarate metabolism, alpha-linolenic acid metabolism, and primary bile acid biosynthesis. In contract, CAD specific pathways consisted of arginine biosynthesis, sphingolipid metabolism, pyruvate metabolism, and glycerophospholipid metabolism. Specifically, CADOSA patients exhibited unique metabolic pathways involving unsaturated fatty acid biosynthesis. Machine learning analysis identified the newly discovered and top-ranking potential metabolomic biomarkers for each condition: upregulated D-glucuronolactone for OSA, increased 12-hydroxyoctadecanoic acid for CAD, and downregulated tryptophan for CADOSA. Matched transcriptome association analysis revealed highly expressed and strongly correlated metabolite-gene pairs in the CADOSA comorbidity, including tryptophan-DMXL2, (C12-AE1S)-ZDHHC11, and IBMX-FCAR, which acting as potential up- and down-stream regulatory relationships underlying progress of CADOSA. CONCLUSION: This study elucidates the distinct clinical features and metabolic profiles of OSA, CAD, and comorbid CADOSA. These newly identified metabolic biomarkers and associative gene-metabolite pairs highlight a distinct comorbidity-specific metabolic and transcriptional architectures, providing novel insights into the development of molecular biomarkers and therapeutic targets for CADOSA.