OBJECTIVE: To investigate whether advanced glycation end products (AGEs) participate in the development of coronary artery disease (CAD) in nondiabetic and diabetic subjects. RESEARCH DESIGN AND METHODS: Serum concentrations of AGEs were measured using a newly established enzyme-linked immunosorbent assay in 48 nondiabetic patients (normal glucose tolerance, n = 20; impaired glucose tolerance, n = 28) who received coronary angiography for the study of chest pain or suspected CAD. Insulin sensitivity was examined by the euglycemic-hyperinsulinemic glucose clamp technique and was estimated as the mean glucose infusion rate during the last 30 min of clamp time (M value). RESULTS: Patients were classified into four groups based on the number of significantly stenosed vessels, defined as 0-, 1-, 2-, or 3-vessel disease. Serum concentrations of AGEs were significantly higher in nondiabetic subjects with CAD than in control subjects (2.42 +/- 0.65 vs. 1.96 +/- 0.40 mU/ml, P < 0.01) and significantly correlated with the number of significantly stenosed vessels (r = 0.678, P < 0.001). M values significantly inversely correlated with serum concentrations of AGEs (r = -0.490, P < 0.05). In multiple regression analysis, with the number of significantly stenosed vessels as the dependent variable, serum concentrations of AGEs, 2-h plasma glucose, and areas under the plasma glucose response curve were independently associated. CONCLUSIONS: This pilot study indicates the relation between AGEs and the severity of CAD in nondiabetic patients. The measurement of serum AGE concentrations may be predictive of vascular damage.