Association between vitamin D status and long-term falls-related hospitalization risk in older women.

Abstract

The dose-response relationship of vitamin D status and the risk of serious falls requiring hospitalization in older women is unclear. We examined the association between plasma 25-hydroxyvitamin D (25OHD) with falls-related hospitalizations over 14.5 years in a large cohort of older women. In 1348 community-dwelling Australian women aged ≥70 years, plasma 25OHD concentrations were assessed at baseline (1998) using LC-MS/MS. Fall-related hospitalizations were obtained from linked data systems. Baseline grip strength and timed-up-and-go (TUG) were assessed as measures of muscle strength and physical function, respectively. Mean plasma 25OHD was 66.9 ± 28.2 nmol/L. The number of women in the low (LOW; <50 nmol/L), medium (MED; 50 to <75 nmol/L), and higher 25OHD (HIGH; ≥75 nmol/L) categories were 384 (28.5%), 491 (36.4%), and 473 (35.1%), respectively. In the multivariable-adjusted analysis, compared to LOW, women in HIGH had significantly lower hazards for a falls-related hospitalization (HR 0.76 95%CI 0.61-0.95). Restricted cubic spline regression models highlight increasing gradient of risk for a falls-related hospitalization with decreasing 25OHD levels. Generalized additive modeling highlighted higher 25OHD to be associated with better TUG performance. Including TUG into the multivariable-adjusted models did not alter the relationship between 25OHD and injurious falls (HIGH vs. LOW HR 0.76 95%CI 0.60-0.95). In community-dwelling older Australian women, maintaining plasma 25OHD at 75 nmol/L or above may confer benefits to muscle function and long-term prevention of injurious falls requiring hospitalization. This relationship appears to be independent of better physical function observed in women with higher 25OHD levels.