Serum 25-hydroxyvitamin D levels and incident falls in older women

Three hundred eighty-seven home-dwelling older women were divided into quartiles based on mean serum 25-hydroxyvitamin D (S-25(OH)D) levels. The rates of falls and fallers were about 40% lower in the highest S-25(OH)D quartile compared to the lowest despite no differences in physical functioning, suggesting that S-25(OH)D levels may modulate individual fall risk.

Vitamin D supplementation of 800 IU did not reduce falls in our previous 2-year vitamin D and exercise RCT in 70-80 year old women. Given large individual variation in individual responses, we assessed here effects of S-25(OH)D levels on fall incidence.

Irrespective of original group allocation, data from 387 women were explored in quartiles by mean S-25(OH)D levels over 6-24 months; means (SD) were 59.3 (7.2), 74.5 (3.3), 85.7 (3.5), and 105.3 (10.9) nmol/L. Falls were recorded monthly with diaries. Physical functioning and bone density were assessed annually. Negative binomial regression was used to assess incidence rate ratios (IRRs) for falls and Cox-regression to assess hazard ratios (HR) for fallers. Generalized linear models were used to test between-quartile differences in physical functioning and bone density with the lowest quartile as reference.

There were 37% fewer falls in the highest quartile, while the two middle quartiles did not differ from reference. The respective IRRs (95% CI) for falls were 0.63 (0.44 to 0.90), 0.78 (0.55 to 1.10), and 0.87 (0.62 to 1.22), indicating lower falls incidence with increasing mean S-25(OH)D levels. There were 42% fewer fallers (HR 0.58; 040 to 0.83) in the highest quartile compared to reference. Physical functioning did not differ between quartiles.

Falls and faller rates were about 40% lower in the highest S-25(OH)D quartile despite similar physical functioning in all quartiles. Prevalent S-25(OH)D levels may influence individual fall risk. Individual responses to vitamin D treatment should be considered in falls prevention.