Ex vivo bone mineral density of the wrist: influence of medullar fat

The cone beam technology was recently proposed in a third generation of densitometers (dual photon X-ray absorptiometry, DXA such as the Lexxos densitometer). Because fat is a well-known problem with DXA, we have designed a cadaver study to compare the influence of medullary lipids on the measures performed with the Hologic QDR4500 and the Lexxos. Twenty-three human distal radii were obtained and analyzed in parallel on both densitometers; bone mineral density (BMD) was measured at the distal radius with standard softwares and on a standardized square regions of interest (ROI). Bones were then defatted and a new series of measurement was performed. Bones were then thoroughly dried and a cube was prepared at the distal radius with a banding saw. Trabecular and total bone volumes were measured by microcomputerized tomography. Ash eight was obtained after calcination of the blocks. BMD could be measured on the Lexxos before and after delipidation but this was not possible with the QDR4500. The X-ray image quality was better with the Lexxos. Delipidation had a very significant effect on measurements: after defatting, BMD values were considerably reduced (-49.8 +/- 19.4%). BMD before/after defatting were significantly correlated (r = 0.81, P < 0.0001) but bone mass appeared to reflect 66% of the variance. BMD was significantly correlated with BV/TV after defatting (r = 0.44, P < 0.03) but the correlation improved when cortices were taken into account (r = 0.70, P < 0.0001). Ash weight was significantly correlated with BMD and total bone volume (respectively, r = 0.84, P < 0.0001; r = 0.53, P < 0.03), but not with BV/TV. BMD at the distal radius is influenced by marrow fat and cortical density.