THE USE OF QUANTITATIVE COMPUTED-TOMOGRAPHY TO ESTIMATE RISK OF FRACTURE OF THE HIP FROM FALLS

We conducted an in vitro investigation of the loads and energies needed to fracture the proximal part of the femur in twelve fresh cadavera under loading conditions simulating one particular type of fall. The fracture loads ranged from 778 to 4,040 newtons and the work to fracture, from five to fifty-one joules. We also investigated the relationship between the fracture loads and several potential indices of bone strength, which were measured non-invasively at the subcapital, basicervical, and intertrochanteric regions with quantitative computed tomography. A very high positive correlation with the fracture load resulted from use of an intertrochanteric index - the product of the average trabecular computed-tomography number and the total cross-sectional area of the bone (R2 = 0.93, standard error of estimate = 295 newtons, and p < 0.00001). We expect the use of this parameter to result in improved assessments of the degree of osteoporosis and of the component of risk of fracture of the hip that is associated with bone strength. However, the measured work to fracture for the isolated femur was an order of magnitude smaller than estimates of the energy available during a typical fall (about 450 joules), suggesting that energy absorbed during falling and impact, rather than bone strength, may be the dominant factors in the biomechanics of fracture of the hip. Clinical relevance: The very high correlation between the intertrochanteric computed-tomography index and the femoral fracture loads supports the assertion that quantitative computed tomography provides an effective predictor of the component of risk of fracture of the hip in vivo that is associated with bone strength. However, the fact that the average work to fracture for these isolated femora represented only about one-twentieth of the total energy available during a typical fall indicates that factors associated with the biomechanics of falling are of far greater importance than is currently assumed.