Objectives
To examine whether (1) serum 25-hydroxy-vitamin D level (25[OH]D) is a risk factor for hyperglycemia, as assessed by glycated hemoglobin (HbA1c), in African American men (AAM) and (2) 25(OH)D is a predictor of HbA1c in AAM and Caucasian American men (CAM).
Methods
We prospectively assessed 25(OH)D and HbA1c in 1,074 men, outpatients with and without diabetes, at an urban Veteran Administration Medical Center (66.8% AAM, 26.4% CAM, 6% Hispanic, 0.4% Asian, and 0.4% Native American men). Multivariate regression analyzed the determinants of HbA1c after accounting for potential confounders.
Results
We found high prevalence of low (< 30 ng/mL) 25(OH)D (81%) and elevated (≥5.7%) HbA1c (53.5%). The 25(OH)D was inversely associated with HbA1c in all men (r = −0.12, P<.001), in AAM (r = −0.11, P = .003), and in CAM (r = −0.15, P = .01). In the entire group the independent determinants of HbA1c included body mass index (BMI), age, 25(OH)D levels, systolic blood pressure (BP), triglycerides, high-density lipoprotein (HDL), and current alcohol use (P<.0001, .013, .009, .01, .008, .034, and .048, respectively) while glomerular filtration rate (GFR) and marital status showed borderline significance (P = .08 and .09, respectively). In AAM these determinants included BMI, 25(OH)D levels, systolic BP, and current alcohol use (P<.0001, .01, .02, and .03, respectively), while age had borderline significance (P = .06). In CAM, these included BMI, age, and triglycerides (P = .01, .03, and .004, respectively) but not 25(OH)D levels (P = .50).
Conclusion
Circulating low 25(OH)D is a risk factor for hyperglycemia, as assessed by HbA1c, in AAM. The 25(OH)D level is an independent determinant of HbA1c in AAM, but not in CAM, including men with and without diabetes.