Site‐matched calcaneal measurements of broad‐band ultrasound attenuation and single X‐ray absorptiometry: Do they measure different skeletal properties?

Because of the differences in the interactions of ultrasound and x‐ray waves with bone, quantitative ultrasound (QUS) techniques may yield information about skeletal status not accessible by regular bone densitometry (BD) techniques. However, relatively strong correlations have been reported between broad‐band ultrasound attenuation (BUA) and several x‐ray‐based BD methods. We assessed the precision and association of single x‐ray absorptiometry (SXA) and BUA of the calcaneus. We examined both BUA and SXA at the calcaneus using special software for matching the regions of interest. An algorithm was derived and applied to correct the observed correlation coefficients for the attenuation effect caused by the precision errors for BUA and SXA. In a group of 33 volunteers covering a wide range of age and calcaneal bone mineral densities, the site‐matched and precision‐adjusted correlation coefficient between BUA and SXA was r = 0.58, with a standard error of the estimate (SEE) of 14.41 dB/MHz, or 17.08%. For the subgroup of 25 women the correlation was stronger, with r = 0.72 and SEE = 11.53 dB/MHz, or 14.33%. SXA precision was 0.79% for the regular region of interest (ROI) and 1.22% for the site‐matched ROI. BUA precision was 2.76% for the entire subject group and 1.63% for women of age 40 or older. The observed correlation coefficient between ultrasound and x‐ray based techniques of the order of 0.7 is significant, but it leaves about 50% of the variability unexplained. It remains to be investigated to what extent the unexplained 50% is related to bone strength or structure or whether it reflects a variability unrelated to osteoporosis status.