Biomechanics of fracture risk prediction of the hip and spine by quantitative computed tomography

In this review, engineering concepts provide a framework for reviewing current efforts relating quantitative CT (QCT) measures and bone density and strength. A factor of risk for a particular set of in vivo loading conditions is defined as the ratio of the in vivo loads on the hip or spine to the known in vitro failure loads for these regions. Experiments designed to relate QCT data to failure loads for the proximal femur and vertebrae are remarkably accurate, with coefficients of determination (R²) as high as 0.93 and relative errors as low as 14%. However, when these strength data are compared against estimated in vivo loads, it is apparent, especially for the spine, that loads are very close to those that cause fracture. The findings thus emphasize the importance of incorporating in vivo load estimates in all attempts at fracture risk prediction. They also help explain the relatively poor predictive power of previous approaches to fracture risk prediction, which do not account for the confounding influences of these loads.