Bone Abstracts

Recent meta-analyses of vitamin D without co-administration of calcium have not demonstrated fracture prevention, possibly through lack of power, inappropriate choice of doses, or failure to target the intervention to deficient populations. Bone mineral density (BMD) is able to detect biologically significant effects in much smaller cohorts, so is a relevant surrogate measure with which to re-assess the skeletal efficacy of these supplements. We searched Web of Science, Embase and the Cochrane Database for randomized trials comparing interventions that differed only in vitamin D content (D₃ or D₂, but not a vitamin D metabolite), and presented BMD results. Studies in groups with other metabolic conditions were not eligible. 23 studies, mean duration 23.5 months, comprising 4082 participants, 92% women, average age 59 years, met the inclusion criteria. Nineteen studies were in predominantly Caucasian populations. Mean baseline serum 25-hydroxyvitamin D was <50 nmol/l in 8 studies (1791 participants). Twelve studies administered calcium supplements to all trial participants. Ten studies (2294 participants) used vitamin D doses <800 IU/day. BMD was measured at 1–5 sites in each study, so 70 tests of statistical significance were carried out across the studies. There were 5 findings of significant benefit, 2 of significant detriment, and the rest were non-significant. Only one study found benefit at >1 site. Meta-analysis showed a small benefit at the femoral neck (0.8%, 95% CI 0.2, 1.4) with evidence of heterogeneity among trials. There was no effect at any other site, including the total hip. There was evidence of a bias toward positive results at the femoral neck and total hip. We conclude that vitamin D supplementation did not change BMD, except at the femoral neck where there were small increases of uncertain clinical significance. The widespread use of vitamin D supplements in the management of osteoporosis should, therefore, be re-examined.