Studies have indicated that quantitative ultrasound (QUS) variables may be
influenced by the mechanical properties of bone which in turn are determine
d by bone's material and structural properties. However. from these studies
it is unclear what rule density, elasticity, and structure play in determi
ning velocity. Eighteen defatted, 12-mm cubic trabecular bone specimens wer
e cut from cadaveric specimens. Amplitude-dependent speed of sound (SOS) us
ing a single point QUS system was assessed in three orthogonal axes. Magnet
ic resonance images were obtained, from which measures of apparent trabeucl
ar structure were derived. The specimens were nondestructively tested in co
mpression along three orthogonal axes defined by the sides of the cubes. Th
e elastic modulus tin the three directions) and the strength tin one direct
ion) were determined. Trabecular BMD was measured by quantitative computed
tomography. SOS varied significantly with direction of measurement, with th
e highest value in the axial direction (axial:1715 m/s, sagittal: 1662 m/se
cond, and coronal: 1676 m/s). SOS of each of the three axes was generally a
ssociated with the various mechanical (r = 0.300.87), density (r = 0.81-0.9
3), and bone structural variables (0.3-0.8). However, after adjusting the S
OS correlations by density, only the correlation with elasticity remained s
ignificant in the coronal direction. BMD alone explained 88-93% of variance
in SOS whereas in the multivariate model, BMD plus elasticity and/or aniso
tropic variables explained 96-98% of the variance in SOS. Variability of SO
S is explained mostly by density and to a small extent by elasticity or ani
sotropy. Since only 2-6% of the variance of the QUS measurement is not expl
ained by density and elasticity, one could conclude that the remaining vari
ance reflects other properties of bone or perhaps simply measurement error.
Evidence that these other properties may be structure related is only foun
d in the anisotropy of QUS parameter.