Twenty iliac crest cadaver specimens were obtained from 7 females and 13 males, ages 57–85 years, to empirically test the mathematically derived conversions of linear perimeter to three-dimensional area-volume ratios and to test the Delesse principle for volume. The scaling factor for linear perimeters was measured by coating the specimens with fast green stain. A weight per area coating constant was calculated and directly applied to the weight of coating on bone. External volume was measured by calipers and trabecular volume was measured by displacement. All data were repeated on the same specimens and correlated with standard histomorphometric technique. Our direct and histomorphometric measurements of volume on the same specimens showed almost identical means and were highly correlated (r=0.992,P<0.001). The direct and histomorphometric measures of surface density were also highly correlated (r=0.980,P<0.001). A comparison of values derived from 20 and 60 fields was run on the histomorphometric determinations of volume and surface density. The extremely high correlations for volume (r=0.995,P<0.001) and surface density (r=0.998,P<0.001) suggest that 20 fields are sufficient for these two measurements. The conclusions were that (a) Delesse' principle can be accurately applied to trabecular bone; (b) stereologic measures of bone volume are highly accurate; (c) inter- and intrapatient variation is due largely to sample site variation; (d) the direct measurement of trabecular surface was no different than the histomorphometric one; and (e) the conventional scaling factor of 4/π (1.273) is correct and is nearly identical to the empirically derived factor of 1.199.
Bone histomorphometryIsotropyStereologySurface densityVolume