Association between Vitamin D and Adiponectin and Its Relationship with Body Mass Index: The META-Health Study.

BACKGROUND: Low vitamin D and adiponectin levels are both associated with obesity and cardiovascular disease. Previous studies have indicated that vitamin D levels are directly associated with adiponectin, and that this association varies across body mass index (BMI) categories; stronger with increasing BMI. Few studies examined this association in African-Americans (AA), known to have lower levels of vitamin D and adiponectin, and in whites. METHODS: We assessed whether serum vitamin D is associated with serum adiponectin in a biracial population-based sample. Cross-sectional analyses were performed on 426 non-diabetic participants (218 whites and 208 AA) from the META-Health Study, a random sample from the metro Atlanta. Age-adjusted correlations and multivariable linear regression were used for analyses. We investigated the effect modification of the BMI categories of lean, overweight, and obese as defined by standard cut-points (25 and 30 kg/m(2)). RESULTS: The mean (SD) age of our study sample was 50.5 (9) years. The mean (SD) levels of vitamin D were 27.4 (9.8) ng/mL in white women, 25.5 (9.3) ng/mL in white men, 16.9 (7.3) ng/mL in AA women, and 18.8 (7.3) ng/mL in AA men. The mean (SD) levels of adiponectin were 17.0 (17.1) μg/mL in white women, 9.9 (11.3) μg/mL in white men, 6.6 (4.8) μg/mL in AA women, and 9.4 (11.6) μg/mL in AA men. Among lean white women (n = 63), there was a significant direct association between vitamin D and adiponectin (β = 0.02, p = 0.04) after adjustment for age, systolic blood pressure, HDL-cholesterol, triglycerides, income, and season of blood drawing. On the contrary, in lean AA women (n = 23), there was a significant inverse association (β = -0.06, p = 0.01). CONCLUSION: The association of vitamin D and adiponectin is dependent on race, gender, and BMI category. Among lean white women, there was a significant direct association, whereas in lean AA women the association was inverse. No association was present among obese individuals.