Bone mineral density, ultrasound velocity, and broadband attenuation predict mechanical properties of trabecular bone differently

Abstract

Measurement of areal bone mineral density (BMD(areal)), broadband ultrasound (US) attenuation (BUA), and speed-of-sound (SOS) are widely used ways to perform clinical assessment of bone quality. In this study, bovine (n = 37) and human (n = 32) trabecular bone was investigated in vitro and in vivo to reveal relationships between mechanical properties, mineral density, and US parameters BUA and SOS. To fulfill these aims, clinical US and dual-energy X-ray absorptiometry (DXA) techniques, as well as dynamic and destructive mechanical testing, were utilized. BUA correlated positively and linearly with BMD(areal) only in human calcaneus of low or moderate density (r = 0.849, n = 32, p < 0.01). When calcaneal areas with high BMD(areal) were included in the analysis, however, the in vivo study revealed that the BUA-BMD(areal) relationship could be described by a second-order polynomial fit (r(2) = 0.618, n = 408). In high-density human or bovine bone, the BUA-bone density relationship was negative. In the in vitro assessment, BUA correlated linearly and negatively with volumetric BMD (BMD(vol)) (r = -0.540, n = 29, p < 0.01) and with storage modulus, as measured at 1 Hz (r = -0.505, n = 28, p < 0.01). A weak positive correlation was found between BUA and mechanical loss tangent (r = 0.322, n = 28, p < 0.1). SOS correlated strongly positively with BMD(vol) (r = 0.888, n = 29, p < 0.01), as well with storage modulus (r = 0.649, n = 28, p < 0.01). In contrast, SOS correlated negatively with loss tangent (r = -0.417, n = 28, p < 0.05). When tested dynamically in the frequency range of 0.01-22.7 Hz, bovine trabecular bone was only slightly viscoelastic. In summary, the most accurate parameters for measuring storage modulus and strength of bovine trabecular bone were SOS and BMD(vol), respectively. BUA failed to predict the mechanical properties of high-density trabecular bone. In vivo mapping of the calcaneus revealed the importance of standardized and reproducible localization of the measurement site for the validity of BUA values.