Relationships Between Bone Mineral Density and Incident Vertebral Fracture Risk With Raloxifene Therapy

Abstract

Low bone mineral density (BMD) in osteoporotic patients predicts an increased risk of bone fracture, but it is not clear whether increasing BMD in those given antiresorptive drugs predicts a reduced risk. The investigators explored this relationship after 1 or 3 years of treatment with raloxifene or placebo in 7705 postmenopausal women with osteoporosis who participated in the MORE (Multiple Outcomes of Raloxifene Evaluation) trial. Raloxifene was given in a daily dose of 60 or 120 mg. Logistic regression models were used to relate baseline BMD and changes in BMD to the risk of new vertebral fractures (Figs. 1 and 2). All women received supplemental calcium and vitamin D. BMD was estimated by dual-energy x-ray absorptiometry in the femoral neck and lumbar spine at yearly intervals. The risk of a new vertebral fracture was greatest in women having the lowest baseline BMD values at either measurement site. Raloxifene reduced the risk of incident vertebral fractures by 41% compared with the placebo group, regardless of baseline BMD at the femoral neck, but women with relatively low baseline lumbar spine BMD had the most marked reduction in fracture risk from raloxifene. After 3 years, raloxifene therapy was associated with increases of only 2% to 3% in BMD. Raloxifene-treated women had a significantly lower risk of spinal fracture than placebo patients regardless of the change in femoral neck BMD, and they also had a lower fracture risk at any percentage change in lumbar spine BMD. At the 75th percentile of the population, with a 6.1% increase in lumbar spinal BMD, the risk of incident vertebral fracture was reduced 52% by raloxifene therapy. With lesser increases in lumbar spine BMD at the 25th and 50th centiles, raloxifene lowered the risk of incident spinal fracture by 39% and 46%, respectively. These findings indicate that the percentage change in BMD in postmenopausal women treated with raloxifene account for only 4% of the observed reduction in risk of vertebral fracture. Not unexpectedly, changes in BMD during treatment are a poor predictor of risk reduction.