Abstract
INTRODUCTION: Hyperlipidemia, impaired glucose tolerance, and inflammatory status are critical contributors to the progression of coronary artery disease (CAD). Biomarkers associated with these pathways may predict clinical outcomes in dialysis-dependent CAD patients. This study aims to compare conventional lipid profiles, inflammatory markers, and insulin resistance-related indicators to evaluate their associations with mortality and prognostic performance in this high-risk population. METHODS: This multicenter retrospective cohort study included 1249 patients from January 2015 to June 2021. 864 patients were finally included in the analysis. Baseline lipid profiles, inflammatory markers, and insulin resistance-related indicator (the triglyceride-glucose [TyG] index) were analyzed. RESULTS: Among 864 patients, 252 deaths (29.2%) occurred over three years. Both the TyG index (hazard ratio [95% confidence interval]: 1.22 [1.03-1.44], p=0.024) and inflammatory marker (neutrophil-to-lymphocyte ratio [NLR]: 1.02 [1.00-1.04], p=0.038) were independently associated with mortality, whereas lipid profiles showed no significant association. Adding the TyG index or NLR to the Global Registry of Acute Coronary Events (GRACE) score (TyG c-index: 0.644 [0.609-0.679]; NLR c-index: 0.642 [0.607-0.677]) or baseline model (TyG c-index: 0.707 [0.676-0.738]; NLR c-index: 0.705 [0.672-0.738]) improved predictive performance. The combined model integrating TyG index and NLR demonstrated the highest discriminative ability for mortality prediction (GRACE score c-index: 0.660 [0.625-0.695]; Baseline model c-index: 0.713 [0.682-0.744]). CONCLUSION: The findings indicate that systemic inflammation and insulin resistance are more significant risk factors for three-year mortality in dialysis-dependent CAD patients than dyslipidemia. This suggests that targeted anti-inflammatory therapies and regulation of glucose-lipid metabolism may offer greater benefits compared to conventional lipid-lowering strategies in this high-risk cohort.