Serum 25 hydroxy vitamin D 25(OH) and the risk of hip fracture: the Women's Health Initiative (WHI)

Vitamin D deficiency is common in older persons, especially in home based populations, in general medical inpatients and in women with acute hip fracture. To test the hypothesis that serum measures of 25(OH)D predict the likelihood of hip fracture in community dwelling women, we preformed a nested case control study in the WHI Observational Study Cohort. We excluded women who had a history (hx) of hip fracture or used hormones or other osteoporosis medications at baseline. A total of 39,793 women met these criteria. We randomly chose 400 women who had an incident hip fracture, confirmed by medical record over a median of 7. I y. A control was selected for each case, matched on baseline age (+/- one year), race/ethnicity and date of blood draw (within 75 days). 25(OH)D was measured in fasting baseline serum using DiaSorin kits. The sensitivity of the assay was ISng/ml; interassay coefficients of variation ranged 8.6% to 12.5%. Conditional logistic regression models were conducted to evaluate the association between hip fractures and Vitamin D. Odds ratio (OR) with 95% confidence intervals (CI) were estimated per SD and across quartiles of 25(OH)D (defined by the distribution in controls) with p values for tests of linear trend. We adjusted for potential confounders. The mean age of cases and controls was 70.8 years; 95% were White. ZS(OH)D levels ranged from 3.7 to 48.6 ng/ml. The mean (SD) Vitamin D level (ng/ml) in the cases was 22.4 (8.1) and 23.8 (7.2) in the controls, p=0.007. The risk of hip fractures was 77% higher among women whose 25(OH)D was <19ng/ml (Table). Additional adjustment for other risk factors for fracture had no effect. In conclusion, low serum 25(OH)D levels are an independent prognostic risk factor for hip facture in community dwelling postmenopausal women.