Impact of reference point selection on DXA-based measurement of forearm bone mineral density

Few studies have systematically evaluated the technical aspects of forearm bone mineral density (BMD) measurement. We found that BMD remained stable regardless of the reference point; however, the ROI identified was not always consistent. Our study highlights the importance of using the same reference point for serial measurements of forearm BMD.

Forearm fractures are clinically important outcomes from the perspective of morbidity, health care costs, and interruption of work. BMD of the forearm, as derived by dual-energy x-ray absorptiometry (DXA), is a better predictor of fracture at the forearm compared with BMD measured at other sites. However, very few studies have evaluated the technical aspects of selecting the ROI for forearm BMD measurement. This study aimed to compare the BMD values measured at the 1/3 radius site using three different reference points: the ulnar styloid process, the radial endplate, and the bifurcation of the ulna and radius.

Healthy Chinese patients participating in the control group of an ongoing study at Zhejiang Provincial People's Hospital were recruited for this study. For each patient, a DXA scan (GE Lunar Prodigy) of the forearm was performed and BMD values were separately calculated using each of the three reference points to identify the ROI. Pearson correlation coefficients were calculated to examine the correlation between the BMD measures derived from each reference point. The F test and independent t test were applied for more robust analysis of the differences in the variances and means.

Sixty-eight healthy Chinese volunteers agreed to participate in this study. The root mean square standard deviation (RMS-SD) percentages of BMD values measured at the 1/3 radius site were 2.19%, 2.23%, and 2.20% when using the ulnar styloid process, radial endplate, and the bifurcation of the ulna and radius as the reference points, respectively. Pearson's correlation coefficients for all pairwise comparisons among these three groups were greater than 0.99. F tests and independent t tests showed p values ranging from 0.92 to 0.99. However, we observed that among 10% of patients, choosing an ROI at the ulnar styloid process led to an inability to accurately determine the BMD at the ultra-distal radius.

Given equal ability to determine BMD at the 1/3 radius accurately, the radial endplate or the bifurcation of the ulna and radius should be preferentially selected as the reference point for routine forearm BMD measurements in order to avoid situations in which the ultra-distal radius BMD cannot be determined.