Simultaneous densitometry and quantitative bone sonography in the estimation of osteoporotic fracture risk.

759 post-menopausal women (41-80 years old), 175 with and 584 without vertebral fracture, were studied by quantitative bone sonography and densitometry of the distal radius in order to evaluate the ability of ultrasound transmission velocity (UTV) to separate fractured from healthy women independently of bone mineral density (BMD) and to test the possibility of improving the discriminant ability of BMD by the simultaneous use of UTV. A second BMD measurement was made at the mid radial shaft. Both BMD and UTV were higher in healthy subjects than in fractured women; the latter being older, shorter and having a longer postmenopausal status. On logistic regression, standardized for 1 SD from the mean value of healthy women, UTV differentiated between healthy and fractured subjects after correction for BMD. UTV was also a significant predictor of fracture in a selected subgroup of healthy and fractured women paired for BMD (144 pairs). In this same subgroup, the difference in UTV between fractured and healthy women regression lines was related to elasticity (E) variation between pairs. Simultaneously evaluating BMD and UTV as fracture predictors, logistic regression showed an odds ratio that was twice that of each predictor alone and 1.2 times higher than that derived from the simultaneous evaluation of two different BMD predictors. These data confirmed that UTV differentiates between healthy and fractured women both as well as and independent of BMD. In addition, UTV separated fractured from healthy BMD matched women by measuring non mass related differences. The simultaneous use of BMD and UTV improved the discriminant ability of each of them alone and of two simultaneous measurements of BMD detected at different skeletal sites.