Ten-year probabilities of clinical vertebral fractures according to phalangeal quantitative ultrasonography.

Abstract

The objectives of the present study were to estimate 10-year probabilities of clinical vertebral fractures in women, according to age and bone mineral assessment using phalangeal quantitative ultrasound (QUS). Risks were computed from UK derived data on the incidence of a first symptomatic vertebral fracture and mortality rates for each year of age using Poisson models. The 10-year probability of vertebral fracture was determined as the proportion of individuals fracture-free at that site from the age of 45 years. We assumed that the risk of fracture increased with decreasing QUS as assessed by an independent re-analysis of a previously published, multicenter cross-sectional study. For amplitude-dependent speed of sound (AD-SoS) information was available from 8,502 women, and vertebral fracture risk increased 1.7-fold for each SD decrease in measurement. For fast wave amplitude (FWA), available in 6,573 women, the risk gradient was 2.4/SD. In a subset of the population ( n =1,572) in whom bone mineral density was measured at the lumbar spine, the gradient of risk was 2.3/SD, with similar gradients of risk noted for AD-SoS (1.8/SD) and FWA (2.6/SD). Ten-year probabilities increased with age and decreasing Z -score. The use of absolute risk permits information from different types of bone mineral measurements to be applied for the assessment of patients, either alone or in combination with other independent risk factors.