Objective: Obesity is associated with elevated levels of biomarkers of inflammation and endothelial dysfunction [including C-reactive protein (CRP), E-selectin, and intercellular adhesion molecule-1], as well as insulin resistance (IR) and type 2 diabetes. We tested the hypothesis that these biomarkers mediate associations among obesity, IR, and risk of diabetes.
Research methods and procedures: We stratified 510 initially non-diabetic women in the Nurses' Health Study cohort into four phenotypes above/below median BMI (27 kg/m2) and waist circumference (81 cm): low BMI-low waist (LBLW; N = 190), low BMI-high waist (LBHW; N = 74), high BMI-low waist (HBLW; N = 27), and high BMI-high waist (HBHW; N = 219).
Results: In models assessing associations of weight phenotype with IR [fasting insulin (FI)], adjusted for age and diabetes risk factors, mean FI was higher comparing HBHW women (13.6 microU/mL, p < 0.0001) and LBHW (11.5 microU/mL, p = 0.02) with LBLW women (8.6 microU/mL); HBLW and LBLW women were not significantly different. Differences in FI levels were most strongly attenuated after adjustment for E-selectin comparing LBHW with LBLW women (11.7 vs. 9.7 microU/mL, p = 0.2).
Discussion: In logistic regression models, LBHW predicted diabetes (risk factor-adjusted relative risk 2.06, 1.05 to 6.40), compared with LBLW, but was no longer significant after adjustment for E-selectin or CRP. After adjusting for CRP and E-selectin, only HBHW and E-selectin were significantly associated with risk of diabetes. In women with central adiposity and low BMI, endothelial dysfunction and inflammation may mediate the relationship among central fat, IR, and incident diabetes.