PubMed2005
The femoral neck (FN) is a cantilever with external and internal dimensions determining its size, shape, the spatial distribution of the mineralized cortical and trabecular bone tissue mass, and its strength. Geometric indices of FN strength are often derived using FN dimensions estimated in vivo from dual X-ray absorptiometry (DXA) assuming that the FN cross section approximates a circle or a square. As DXA does not measure FN depth, we examined whether circular, square, or elliptical models of FN cross sections predict FN depth, and so its external volume, shape, volumetric bone mineral density (vBMD), and geometric indices of strength. We studied paired FN specimens from 13 Caucasian female cadavers (mean age 69 years, range 29 to 85) using DXA, micro-computed tomography (mu-CT), and direct calliper measurements. DXA accurately measured FN width (supero-inferior diameter) but models assuming a circular and a square cross section overestimated FN depth (antero-posterior diameter) and volume, and so underestimated vBMD by 15.0 +/- 5.8% (circular cross section) and by 33.2 +/- 4.6% (square cross section) (both P < 0.05). As depth was less than the width, an elliptical model with a constant depth/width ratio of 0.75 reduced the accuracy error in vBMD to 14.0 +/- 8.5% (P = 0.10). However, as FN width increased, FN depth increased relatively less. An elliptical model using a quadratic equation to mimic this changing in shape with increasing size reduced the error in vBMD to 4.4 +/- 7.7% (NS). Circular cross-section models overestimated section modulus at the mid-FN by about 51%. The elliptical models reduced the error two- to three fold. Images from micro-CT scanning show that the FN cross-sectional shape resembles an ellipse with the long axis and the maximum moment of inertia (I(max)) oriented in the supero-inferior direction, and the cortical mass concentrated inferiorly. The larger the cross section, the more elliptical the shape, and the greater the I(max) supero-inferiorly, while I(min) (in the antero-posterior direction) remains relatively constant. The shape, spatial distribution of bone, and moments of inertia are likely to be adaptations to bending moments during bipedalism. Assuming the FN cross section approximates a circle or square produces errors in FN depth, volume, vBMD, and geometric indices of bone strength. Studies are needed to determine the effects of age, sex, and race on FN size and shape in health and disease.