NUTRITION, BONE MASS, AND SUBSEQUENT RISK OF HIP FRACTURE IN WHITE WOMEN

The interrelationships between nutritional status, concurrent bone den sity, bone dimensions, and risk of subsequent hip fracture were invest igated using data from the first National Health and Nutrition Examina tion Survey and its three follow-up studies. A cohort of 890 White wom en aged 45 years or older who received nutritional and bone measuremen ts in the baseline survey in 1971-75 were recontacted in 1982-84, 1986 , and 1987. Height, weight, serum albumin, total energy intake, bone d ensity, and bone dimensions were measured at baseline. Thirty-three in cident hip fractures were identified during 12,190 person-years of fol low-up. Total energy intake and serum albumin were little correlated w ith concurrent bone density and bone dimensions. Serum albumin, total energy intake, and weight tended to be inversely associated with risk of hip fracture. The multivariable relative risks were 0.75 (95% confi dence interval [CI] = 0.57-0.98) for one standard deviation increment of serum albumin, 0.67 (95% CI = 0.42-1.07) for dietary energy intake, and 0.61 (95% CI = 0.38-0.97) for weight, respectively. Height was po sitively, but not significantly associated with risk of hip fracture. Nutritional effects on hip fracture were independent of concurrent bon e density and bone dimensions. Higher bone density was protective of s ubsequent hip fracture (relative risk = 0.55, 95% CI = 0.36-0.83, for one standard deviation increment of bone density) independent of effec ts related to concurrent nutritional status. Thus, poor nutritional st atus and reduced bone density appeared to increase the risks of subseq uent hip fracture independently. (C) 1998 Wiley-Liss, Inc.