Bone is a unique self-repairing structural material that is capable of adapting to a variety of environmental stimuli Generally, healthy bone can respond to a wide-range of mechanical skeletal loading by either increasing bone mass or modifying bone architecture. However, frequently a loading stimulus exhausts the remodeling capacity of bone and the result is a stress fracture or bone stress injury (BSI). In particular, in both civilian and military training environments, young women have an alarming incidence of bone stress injury. Ideally, in order to minimize the rate of bone stress injuries, a measurement tool that could objectively quantify a subject’s fracture resistance prior to onset of increased physical activity is desirable. Procedures that measure either bone mass or bone geometry have been proposed as methods to estimate bone strength. However, these methods do not provide a mechanical measure of bone competence. Therefore, a search for reliable and informative alternative methods continues. One such alternative investigated in this study was the measurement of tibial flexural wave propagation in vivo. An advantage of this method is that it can provide information on the structural and mechanical properties of bone. However, a number of methodological issues remain to be answered before this technology can be successfully implemented in a screening environment.

The purpose of the present study was to determine the most reliable method, for measuring flexural wave propagation, and subsequently, to demonstrate the effect of soft-tissue, shank mass, and subject characteristics on the measured velocity. Following completion of informed consent, the right tibiae of twenty-five young women were sampled on four separate days, two setups per day, and ten trials per setup. The sampling rate was 400 KHz with an accelerometer separation distance of 10cm and the hammer impact on the tibial tubercle. Two velocity determination methods, peak (time domain) and phase (frequency domain), as well as the attenuation coefficient were analyzed. The mean peak velocity was 219 m.s⁻¹ (range 142 to 355 m.s⁻¹) and the mean phase velocity was 210 m.s⁻¹ (range 144 to 409 m.s⁻¹). The peak method was more reliable (ICCs ranged from 0.811 to 0.963) than the phase velocity method (ICCs ranged from 0.589 to 0.888). A step-wise regression analysis was used to find the best factors to predict peak velocity. Two models were developed:​ 1) subject age and history of late menarche (p<0.001, R² = 46.2%) and 2) subject age and width of the proximal tibia (p<0.003, R² = 41.6%). Additionally, a simple regression model demonstrated that the level of subject risk factors for BSI explained a moderate (R² = 23%) amount of the variance in peak velocity (p<0.015). Peak velocity was not associated with shank mass or soft-tissue characteristics, suggesting that peak velocity measured underlying tibial bone characteristics. The attenuation coefficient was highly inversely correlated (-0.944) with peak velocity and therefore, should not be used as an independent measure of bone quality. In conclusion, the peak method can be used to discriminate between subjects with varying levels of flexural wave velocity, however, individual change would need to be fairly substantial in order to not be masked by the measurement variability.