BONE DENSITOMETRY OF THE FOREARM - COMPARISON OF SINGLE-PHOTON AND DUAL-ENERGY X-RAY ABSORPTIOMETRY

Forearm bone mineral densitometry was performed initially by single-ph oton absorptiometry (SPA), but is now achievable by dual-energy X-ray absorptiometry (DXA) as well, with a good correlation between both mea surements. However, it is still unknown whether: (1) short-term precis ion of DXA is superior to SPA and (2) identical regions of interest (R OI) are mandatory to correlate SPA with DXA, The aim of this study was to answer these questions using a commercial system for DXA (DXA-FAS) and to test an in-house system using spine DXA and a soft-tissue comp ensator (DXA-STC). In ten subjects, four measurements on the same day showed significantly lower (p< 0.05) coefficients of variation (CV) fo r bone mineral density (BMD) by DXA-FAS (proximal site: 0.74%; ultradi stal site: 1.20%) than by SPA (1.26% and 2.25%). However, the CV for b one mineral content (BMC) were similar for DXA-FAS (0.73% and 1.58%) a nd SPA (0.79% and 1.34%). The significant difference (p <0.05) for sur face calculation by DXA-FAS (1.24% and 0.93%) compared with SPA (2.36% and 1.28%) explains all the advantages of DXA-FAS for short-term prec ision. The measurements taken on the same day on the ulna and the radi us or on the radius alone by SPA, DXA-FAS, and DXA-STC on 108 subjects aged 18-80 years were highly correlated [r ranging from 0.925 to 0.99 5 (p < 0.0001) and standard error of the estimate from 3.15% to 8.89%] . The need for a manual adjustment of the ROI was found to be mandator y for BMC but not BMD assessment, The use of DXA-STC is a fast method for forearm bone densitometry and its correlation with SPA is very hig h, However, its short-term precision for BMC (3.00% and 1.54%), BMD (2 .15% and 1.12%), and surfaces (1.99% and 1.12%) is significantly highe r (p < 0.05) than that of DXA-FAS, We conclude that short-term precisi on of DXA is better than that of SPA only for BMD and surface measurem ent but not for BMC. ROI should be adjusted manually for the assessmen t of BMC but not for that of BMD.