Assessment of the strength of the proximal femur in vitro: relationship with ultrasonic measurements of the calcaneus

Abstract

Matched pairs of the right proximal femur and right calcaneus were obtained from 64 cadavers (28 female, 36 male). Ultrasonic velocity and broadband ultrasonic attenuation were measured in the calcaneus using a laboratory ultrasound system. Bone mineral density (BMD) was measured at the calcaneus and at the trochanteric and neck regions of the femur using dual-energy X-ray absorptiometry. Femoral strength was determined in a mechanical test simulating a fall onto the greater trochanter. Femoral BMD was more strongly correlated with femoral strength (r2 = 0.71, 0.88 for neck BMD and trochanteric BMD, respectively) than were any of the other predictive variables investigated (p < 0.05). Calcaneal ultrasonic measurements alone produced correlations with femoral strength of r2 = 0.40-0.47, with no significant differences observed in predictive ability between the various ultrasonic parameters. In multiple regression analysis, ultrasound was, in general, not a significant additional independent predictor of femoral strength when combined with either femoral or calcaneal BMD, and combining ultrasonic parameters did not improve the ability to predict femoral strength. Calcaneal width was found to be significantly correlated with both femoral strength and femoral BMD, and this explained the slightly better correlations with femoral strength found for those ultrasonic parameters which were not effectively normalized for calcaneal width. In summary, calcaneal ultrasound did not significantly enhance the prediction of femoral strength compared to femoral BMD measurements alone. Given the substantial differences between the in vitro and in vivo situations, this finding does not necessarily contradict emerging clinical data indicating that ultrasound and BMD have comparable and independent predictive ability for hip fracture risk. Reasons for the apparent discrepancy are discussed, including the enhanced accuracy of DXA in vitro. Nevertheless, it is suggested that further fundamental investigations into the efficacy of current ultrasonic techniques are warranted.