Biomarkers of endothelial dysfunction and risk of type 2 diabetes mellitus.

Endothelial dysfunction occurs in diagnosed type 2 diabetes mellitus but may also precede development of diabetes. To determine whether elevated plasma levels of biomarkers reflecting endothelial dysfunction (E-selectin; intercellular adhesion molecule 1 [ICAM-1]; and vascular cell adhesion molecule 1 [VCAM-1]) predict development of type 2 diabetes in initially nondiabetic women. Prospective, nested case-control study within the Nurses' Health Study, an ongoing US study initiated in 1976. Of 121 700 women initially enrolled, 32 826 provided blood samples in 1989-1990; of those free of diabetes, cardiovascular disease, or cancer at baseline, 737 developed incident diabetes by 2000. Controls (n = 785) were selected according to matched age, fasting status, and race. Risk of confirmed clinically diagnosed type 2 diabetes by baseline levels of E-selectin, ICAM-1, and VCAM-1. Baseline median levels of the biomarkers were significantly higher among cases than among controls (E-selectin, 61.2 vs 45.4 ng/mL; ICAM-1, 264.9 vs 247.0 ng/mL; VCAM-1, 545.4 vs 526.0 ng/mL [all P values < or =.004]). Elevated E-selectin and ICAM-1 levels predicted incident diabetes in logistic regression models conditioned on matching criteria and adjusted for body mass index (BMI), family history of diabetes, smoking, diet score, alcohol intake, activity index, and postmenopausal hormone use. The adjusted relative risks for incident diabetes in the top quintile vs the bottom quintiles were 5.43 for E-selectin (95% confidence interval [CI], 3.47-8.50), 3.56 for ICAM-1 (95% CI, 2.28-5.58), and 1.12 for VCAM-1 (95% CI, 0.76-1.66). Adjustment for waist circumference instead of BMI or further adjustment for baseline levels of C-reactive protein, fasting insulin, and hemoglobin A(1c) or exclusion of cases diagnosed during the first 4 years of follow-up did not alter these associations. Endothelial dysfunction predicts type 2 diabetes in women independent of other known risk factors, including obesity and subclinical inflammation.