Copyright © 2005 American Medical Association. All rights reserved. From: Fracture Prevention With Vitamin D SupplementationA Meta-analysis of Randomized Controlled Trials JAMA. 2005;293(18):2257-2264. doi:10.1001/jama.293.18.2257 Figure Legend: QUOROM indicates Quality of Reporting of Meta-analyses; RCTs, randomizedcontrolled trials. *Vitamin D or active vitamin D compared with treatmentsother than calcium or placebo. Date of download: 11/1/2017 Copyright © 2005 American Medical Association. All rights reserved.

Copyright © 2005 American Medical Association. All rights reserved. From: Fracture Prevention With Vitamin D SupplementationA Meta-analysis of Randomized Controlled Trials JAMA. 2005;293(18):2257-2264. doi:10.1001/jama.293.18.2257 Figure Legend: Squares represent relative risks (RRs) and size of squares is proportionalto the size of the trials. Error bars represent 95% confidence intervals (CIs).Trials are sorted by trial duration ranging from 24 to 60 months for hip fractureand 12 to 60 months for nonvertebral fracture. For 3 trials with hip fractures, which included 5572individuals with a vitamin D dose of 700 to 800 IU/d, the pooled RR was 0.74(95% CI, 0.61-0.88; Q test P = .74). For5 trials with nonvertebral fractures, whichincluded 6098 individuals with a vitamin D dose of 700 to 800 IU/d, the pooledRR was 0.77 (95% CI, 0.68-0.87; Q test P = .41).For the 2 trials, with avitamin D dose of 400 IU/d, trial duration ranged from 24 months to 36 to41 months. Date of download: 11/1/2017 Copyright © 2005 American Medical Association. All rights reserved.

Copyright © 2005 American Medical Association. All rights reserved. From: Fracture Prevention With Vitamin D SupplementationA Meta-analysis of Randomized Controlled Trials JAMA. 2005;293(18):2257-2264. doi:10.1001/jama.293.18.2257 Figure Legend: Circles and squares represent relative risks (RRs) and error bars represent95% confidence intervals. Trendline is based on series of effect sizes (opencircles and squares). All trials identified for the primary analyses for bothfractures are shown as a reference number outside each circle or square. Ameta-regression, which included 9294 individuals, indicated a significantinverse relationship between higher achieved 25-hydroxyvitamin D levels inthe treatment group and hip fracture risk (β = –0.009; P = .02; log RR of hip fracture is estimatedto decrease by 0.009 per 1-nmol/L increase in 25-hydroxyvitamin D). A meta-regression,which included 9820 individuals, indicated a significant inverse relationshipbetween higher achieved 25-hydroxyvitamin D levels in the treatment groupand nonvertebral fracture risk (β = −0.006; P = .03; log RR of nonvertebral fracture is estimated todecrease by 0.006 per 1-nmol/L of 25-hydroxyvitamin D achieved in the treatmentgroup). To convert 25-hydroxyvitamin D to ng/mL, divide values by 2.496. Date of download: 11/1/2017 Copyright © 2005 American Medical Association. All rights reserved.