Context: Hyperhomocysteinemia is an independent risk factor for premature atherosclerosis and thromboembolism. 25-Hydroxyvitamin D [25(OH)D] may modulate the expression of genes involved in homocysteine metabolism.
Objective: Little is known about the relationship between homocysteine and 25(OH)D. We hypothesized an inverse and nonlinear association between 25(OH)D and homocysteine.
Design: We analyzed data from the continuous National Health and Nutrition Examination Survey 2001-2006 for asymptomatic adults (≥18 y).
Setting: Linear regression models with spline adjusted for cardiovascular disease risk factors were used to explore nonlinearity.
Main outcomes measure: Mean change (β-coefficients with 95% confidence intervals) in homocysteine was reported per 10 ng/mL change in 25(OH)D.
Results: Mean (SD) age and homocysteine levels of 14 630 participants were 47.2 (20) years and 8.8 (4.7) μmol/L, respectively, whereas the median (interquartile range) of 25(OH)D was 21 (15-27) ng/mL. Without using spline, we observed an inverse relation between homocysteine and 25(OH)D both in simple [-0.25 (-0.34 to -0.02) μmol/L] and multivariable [-0.13 (-0.18 to -0.01) μmol/L] regression. With spline, in a univariate model, an increase in 25(OH)D was associated with a significant decrease in homocysteine [-0.56 (-0.75 to -0.37) μmol/L] until 25(OH)D reaches but not if above its median (21 ng/mL). Similarly, in multivariable spline models, the inverse relationship between homocysteine and 25(OH)D remain significant [-0.49 (-0.67 to -0.31) μmol/L] only below the population median of 25(OH)D.
Conclusions: From a large community-based cohort of asymptomatic adults, we found an inverse relation between 25(OH)D and homocysteine among those with 25(OH)D concentration of 21 ng/mL or less. We did not observe any statistical decrease in homocysteine once 25(OH)D concentration rose above 21 ng/mL.