LAUSR

To the Editor: We were interested to read the article by Yu et al that was published recently in Circulation Research.1 Many researchers generally use serum total 25-hydroxyvitamin D (25(OH)D) to analyze associations between vitamin D nutritional status and chronic diseases.2–4 The authors’ objective here, however, was to examine associations between bioavailable and free 25(OH)D levels and mortality in patients with coronary artery disease. They suggest that use of free or bioavailable vitamin D metabolite rather than total 25(OH)D may be better to predict cardiovascular diseases. Although their results are very interesting; we consider the article to have one minor error. The title of table 3 refers to “Hazard Ratios for Mortality According to Serum Bioavailable and free 25-hydroxyvitamin D in Coronary Artery Disease Patients of Various Subpopulations,” yet hazard ratios are provided only for quartiles of “Free 25-hydroxyvitamin D,” with none provided for “serum bioavailable.” We guess that the values provided are hazard ratios for quartiles of “bioavailable” 25-hydroxyvitamin D and that the column heading as “free” 25-hydroxyvitamin D may be erroneous. This study suggests a new perspective in measures of vitamin D status: bioavailable and free 25(OH)D concentrations. In this study, online Figure 1 showed a strong correlation between serum bioavailable 25(OH)D and free 25(OH)D; 10% to 15% of 25(OH)D exists in a loose albumin-bound state [bioavailable 25(OH)D], and <0.01% of 25(OH)D is free in the circulation.1 Yu et al suggested the incorporation of a combination of bioavailable 25(OH) D into a model with conventional cardiovascular risk factors improved the prediction capability of cardiovascular death (shown Table IV in the Online Data Supplement). Based on these results, bioavailable 25(OH)D may better assess the vitamin D status for physiology. The computation of bioavailable 25(OH)D, however, requires measures of various other biomarkers, including total 25(OH)D, DBP (vitamin D-binding protein), albumin, DBP isoform. Also, DBP or albumin-bound vitamin D status depends on liver function, kidney disease, sex steroids, and genetic background, whereas free 25(OH)D is less dependent and remains constant.5,6 We are, therefore, cautious that the use of bioavailable 25(OH)D is plausible in large-scale studies. And, the question still remains as to the role of bioavailable 25(OH)D on various adverse health outcomes, including coronary artery disease. For these reasons, and because free 25(OH)D can be measured easily by ELISA kit, we suggest that free 25(OH)D may be more feasible than bioavailable 25(OH)D to estimate vitamin D status in large-scale studies. In fact, many studies suggested that free 25(OH)D is a good measure to estimate vitamin D status.5,7 In summary, we suggest that bioavailable 25(OH)D may better assess the true vitamin D status for physiology, while free 25(OH)D may be more feasible for clinical situation. Nevertheless, the laboratory harmonization of measurements of vitamin D metabolites is required to allow the pooling of research data,8 and further studies are needed to verify whether bioavailable or free 25(OH) D is a potential sensitive predictor of chronic diseases including coronary artery disease.