Why vitamin D clinical trials should be based on 25-hydroxyvitamin D concentrations

Many health benefits are attributed to vitamin D, with those findings supported mostly by observational outcome studies of relationships to serum 25-hydroxyvitamin D [25(OH)D]. However, many randomized controlled trials (RCTs) aiming to confirm those findings have failed, perhaps because serum 25(OH)D is an index of UVB exposure and non-vitamin D mechanisms or because disease reduces serum 25(OH)D content. But the most likely reason for that failure is inappropriate design, conduct, analysis, and interpretation of RCTs. Most RCTs used principles designed to test pharmaceutical drugs; that design incorporates the assumptions that the RCT is the sole source of the agent and that dose-response relationships are linear. However, neither assumption is true for vitamin D, since neither vitamin D dose-responses or health outcome-serum 25(OH)D concentration relationships are linear-larger changes being induced with low rather than high baseline 25(OH)D values. Here, we propose a hybrid observational approach to vitamin D RCT design, based primarily on serum 25(OH)D concentration, requiring an understanding of serum 25(OH)D concentration-health outcome relationships, measuring baseline 25(OH)D values, recruiting non-replete subjects, measuring serum 25(OH)D during the trial for adjustment of supplemental doses for achievement of pretrial selection of target 25(OH)D values, where possible, and analyzing health outcomes in relation to those data rather than solely to vitamin D dosages.