Abstract
Background We investigated the associations of glycemic markers (HbA(1C) [hemoglobin A(1C)], fasting plasma glucose, and insulin resistance-homeostasis model assessment of insulin resistance) with subclinical cardiovascular disease (CVD) among blacks. Methods We included 4303 community-dwelling blacks (64% women; mean age, 54.5 years) without prevalent CVD. Subclinical CVD was defined as ≥1 of the following: any coronary artery calcification (CAC), elevated carotid intima-media thickness (cIMT), left ventricular (LV) hypertrophy, LV ejection fraction <50%, and peripheral artery disease (ankle-brachial index, <0.90). Estimates of cross-sectional associations of glycemic markers (fasting plasma glucose, HbA(1C), and homeostasis model assessment of insulin resistance) with subclinical CVD measures were adjusted for traditional CVD risk factors. Results Each 1% increment in HbA(1C) was associated with higher odds of CAC, abnormal cIMT, and subclinical CVD (all P <0.001). Adjusted mean values of LV mass (LVM), LVM index, relative wall thickness, CAC, and cIMT were increasingly abnormal with worsening HbA(1C) categories (all P<0.05). Each 10-mg/dL increase in fasting plasma glucose was associated with higher odds of LV hypertrophy, CAC, abnormal cIMT, and subclinical CVD (all P <0.005). Adjusted mean values of LVM, LVM index, relative wall thickness, CAC, ankle-brachial index, and cIMT were more abnormal across categories of worsening fasting plasma glucose (all P <0.05). Each unit increment in log-transformed homeostasis model assessment of insulin resistance conferred a higher odd of having LV hypertrophy ( P<0.01). Across quartiles of homeostasis model assessment of insulin resistance, we observed progressively abnormal adjusted mean values of LVM, LVM index, relative wall thickness, and ankle-brachial index (all P <0.01). Conclusions Among blacks, glycemic markers were differentially associated with various measures of subclinical CVD.