• epidemiology;
  • population study;
  • vitamin D;
  • osteoporosis;
  • hip fracture


We used serum 25(OH)D data from NHANES III and incident hip fracture cases identified using linked mortality and Medicare records, and found that serum 25(OH)D was significantly related to reduced hip fracture risk in non-Hispanic white adults ≥65 yr of age.

Introduction: The role of vitamin D status in reducing fracture risk is unclear. We examined the relationship between serum 25 hydroxyvitamin D [25(OH)D] and incident hip fracture risk in older non-Hispanic white adults.

Materials and Methods: The study sample consisted of 1917 white men and women ≥65 yr of age who were examined in the third National Health and Nutrition Examination Survey (NHANES III, 1988-1994), a nationally representative survey. Incident hip fractures were ascertained using linked mortality and Medicare records that were obtained for NHANES III participants. Serum 25(OH)D values were measured with a radioimmunoassay kit. Cox proportional hazards models were used to estimate the relative risk (RR) of hip fracture by serum 25(OH)D level.

Results: There were 156 incident hip fracture cases in the sample. Cases were older, had lower BMD and body mass index, more prevalent spine or wrist fractures and weight loss before baseline, and ate fewer kilocalories and less calcium than noncases. After adjusting for these differences, serum 25(OH)D values exceeding 60 nM were significantly related to hip fracture risk. For example, the multivariate-adjusted RR was 0.64 (95% CI, 0.46-0.89) among individuals with serum 25(OH)D values ≥62.5 nM compared with those with values below this level. When grouped into quartiles, the multivariate-adjusted RR for the second, third, and fourth versus the first quartile of serum 25(OH)D were 0.50 (95% CI, 0.25-1.00), 0.41 (95% CI, 0.24-0.70), and 0.50 (95% CI, 0.29-0.86), respectively.

Conclusions: Serum 25(OH)D was related to a significantly lower hip fracture risk in this cohort of older white adults, even after adjusting for several relevant confounding variables. The relationship did not seem to be linear across all values. Our results support other studies suggesting that serum 25(OH)D values exceeding 60 nM are associated with health benefits.