LAUSR

In Response: We thank Dr Kwak and Dr Choi for their interest and suggestions regarding our recent publication on the association between free and bioavailable 25-hydroxyvitamin D and mortality in patients with coronary artery disease.1 With respect to the suggestion that the column heading of Table 3 may be erroneous, we have carefully checked the previous versions and confirmed this is a typo. It might occur during copyediting or printing phase, but we did not notice it in the last print version unfortunately. We apologize for this error and appreciate Dr Kwak and Dr Choi for their careful reading and this kind comment. Based on the concern of the calculation method for bioavailable 25-hydroxyvitamin D due to using of the formula involving other variables, and the convenience of commercial ELISA kit for free 25-hydroxyvitamin D, we concur with the authors that free 25-hydroxyvitamin D may be more feasible than bioavailable 25-hydroxyvitamin D for clinical situation. However, we also note that only <1% vitamin D metabolites are free. Albumin-bound vitamin D metabolites, which are also biologically active forms in vivo, account for the majority of bioavailable vitamin D, with their circulating levels being almost 1000× higher than free vitamin D levels.2,3 In our study, the extreme-quartile adjusted hazard ratios for cardiovascular mortality were 2.58 for bioavailable 25-hydroxyvitamin D but only 1.64 for free 25-hydroxyvitamin D. Additionally, as we reported in the Online Table IV, incorporation of combination of bioavailable 25-hydroxyvitamin D, but not total D or free D, into the conventional cardiovascular risk model improved the prediction capability of cardiovascular death. These data, therefore, implied that bioavailable D should not be underestimated despite its labor-intensive requirements. For other analogous endocrine hormones, for example, testosterone, even after decades' research, there remain controversies in the implication of total, free, and bioavailable testosterone. As regard to determinations for free/bioavailable testosterone, calculated methods have been recommended as the most useful assessments for determining androgen status.4 Direct measurement of free testosterone using equilibrium dialysis, but not immunoassay, is also recommended by the Endocrine Society,5 but this method is too complex for routine clinical use.6 As for vitamin D, it remains to be determined whether free 25-hydroxyvitamin D using ELISA methods is better biomarker than calculated bioavailable 25-hydroxyvitamin D. Likewise, a concern also exists in term of an ELISA immunoassay for free vitamin D since its levels in circulation are extremely low (the unit of measurement, pg/mL). Laboratory harmonization of measurements of vitamin D metabolites are required to allow the pooling of research data,7 and future largescale studies based on standardized measurements are needed to better assess the true vitamin D status for physiology and clinical situation.