Skip to main content
ARS Home » Northeast Area » Boston, Massachusetts » Jean Mayer Human Nutrition Research Center On Aging » Research » Publications at this Location » Publication #284833

Title: A pooled analysis to define vitamin D dose requirements for fracture prevention in seniors

Author
item BISCHOFF-FERRARI, HEIKE - University Of Zurich
item WILLETT, WALTER - Harvard School Of Public Health
item ORAV, E - Harvard School Of Public Health
item LIPS, PAUL - Vrije University
item LYONS, RONAN - Swansea University
item FLICKER, LEON - University Of Western Australia
item WARK, JOHN - University Of Melbourne
item JACKSON, REBECCA - The Ohio State University
item CAULEY, JANE - University Of Pittsburgh
item MEYER, HAAKON - University Of Oslo
item PFEIFER, MICHAEL - Fuerstenhof Hospital
item SANDERS, KERRIE - University Of Melbourne
item STAEHELIN, HANNES - University Of Basel
item THEILER, ROBERT - Triemli Hospital
item MEUNIER, PIERRE - University Of Lyon
item DAWSON-HUGHES, BESS - Jean Mayer Human Nutrition Research Center On Aging At Tufts University

Submitted to: New England Journal of Medicine
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 12/29/2011
Publication Date: 7/5/2012
Citation: Bischoff-Ferrari, H., Willett, W., Orav, E.J., Lips, P., Lyons, R., Flicker, L., Wark, J., Jackson, R., Cauley, J., Meyer, H., Pfeifer, M., Sanders, K., Staehelin, H., Theiler, R., Meunier, P., Dawson-Hughes, B. 2012. A pooled analysis to define vitamin D dose requirements for fracture prevention in seniors. New England Journal of Medicine. 367(1):40-49.

Interpretive Summary: One strategy to prevent fractures in the elderly may be broad vitamin D supplementation. However, data from several study-level meta-analyses are conflicting. The discordant findings may in part be explained by different inclusion criteria of trials or different adjustments made for adherence. We have conducted an individual-patient-level meta-analysis of 11 available double-blind vitamin D intervention trials, taking into account the impact of vitamin D dose, use of additional vitamin D supplements, and adherence. The objective was to assess the effect of vitamin D on fracture risk. The analysis included data in 31,022 participants of whom 91% were women and mean age was 76 years. There were 1,111 incident hip and 3,770 non-vertebral fractures. Comparing all participants randomized to vitamin D versus control, there was a 10% non-significant reduction of hip (HR = 0.90; 95% CI: 0.80-1.01) and a 7% reduction of non-vertebral fractures (HR = 0.93; 0.87-0.99). When considering actual dose of vitamin D taken, fracture reduction was demonstrated only at the highest quartile (range 792-2000 IU; median 800 IU/day) compared to control, with a 30% reduction at the hip and a 14% reduction at any non-vertebral site. We conclude that vitamin D lowered fracture risk at the hip and any non-vertebral site; however, this was significant only at the highest actual dose level (median 800 IU /day).

Technical Abstract: Meta-analyses reached conflicting results regarding vitamin D and fracture reduction. We pooled individual participant-level data from 11 double-blind RCTs of oral vitamin D supplementation (daily, weekly, 4-monthly) with or without calcium compared with placebo or calcium in seniors age 65 and older. Primary endpoints were incidence of hip and any non-vertebral fractures using Cox regression analyses; adjusting for age, gender, type of dwelling, and study. Beyond the intent-to-treat analyses, our primary aim was to refine this comparison by including quartiles of actual dose of vitamin D by incorporating adherence and supplement use outside the study protocol at the participant level. 31,022 seniors (mean age 76 years, 91% women) with 1111 incident hip and 3770 nonvertebral fractures. Comparing participants randomized to vitamin D versus control, there was a 10% non-significant reduction of hip (HR = 0.90; 95% CI: 0.80-1.01) and a 7% reduction of non-vertebral fractures (HR = 0.93; 0.87-0.99). By quartiles of actual dose, fracture reduction was demonstrated only at the highest actual intake level (median 800 IU; range 792-2000 IU/day) compared to control, with a 30% reduction at the hip (HR=0.70; 0.58-0.86) and a 14% reduction at any non-vertebral site (HR = 0.86; 0.76-0.96). The highest actual dose reduced hip fractures in community-dwelling (HR = 0.68; 0.48-0.96) and institutionalized (HR = 0.70; 0.55-0.89) seniors. Vitamin D lowered fracture risk at the hip and any non-vertebral site; however, this was only at the highest actual dose level (median 800 IU; range 792-2000 IU/day).