Association of standardized serum 25-hydroxyvitamin D with falls in post-menopausal women

Vitamin D status has long been related to falls risk. In this planned secondary analysis of a vitamin supplementation trial in postmenopausal women, standardized 25-hydroxyvitamin D concentration up to 60 ng/mL was not associated with increased falls. Women with 25(OH)D >= 60 ng/mL had higher odds of >= 2 falls. Purpose Falls are common and cause fractures. High circulating 25(OH)D may increase falls risk; thus, recent guidance recommends 25(OH)D not exceed 50 ng/mL. Prior falls studies have not reported standardized 25(OH)D (s25D) data. The purpose of this planned secondary analysis of a 4-year calcium/vitamin D supplementation trial was to evaluate the association of s25D with falls. Methods This study recruited 2,303 postmenopausal women. The analytic dataset consisted of pooled concatenated data from years 2-4 (N-Total = 5,732). Serum 25(OH)D was measured annually and subsequently retrospectively standardized using Vitamin D Standardization Program methods. Falls were recorded by diary. Incidence for >= 1 fall and >= 2 falls was assessed by s25D group (<= 20, 20- < 30, 30- < 40, 40- < 50, 50- < 60 and >= 60 ng/mL) using multivariable logistic regression. Results Mean (SD) baseline s25D was 32.6 ng/mL (8.3) with no difference between supplement and placebo groups. s25D increased to 41.3 ng/mL at year 2 in the supplement group then remained stable. By s25D group, incidence for >= 1 fall varied from 22-32% (p = 0.19). For >= 2 falls incidence varied (p = 0.03) from 6% (< 20 ng/mL) to 17% (>= 60 ng/mL.) There was no significant association between s25D and >= 1 fall. Those with s25D >= 60 ng/mL had a higher adjusted odds of >= 2 falls (OR = 1.99 +/- 1.2-3.3) compared to women with s25D of 30- < 40 ng/mL. Conclusions25D up to 60 ng/mL was not associated with greater risk for >= 1 or >= 2 falls. Women with a s25D >= 60 ng/mL were at higher odds for >= 2 falls, however this group included only similar to 2% of study observations; therefore, confirmation in other cohorts is necessary.