Dual-photon absorptiometry (DPA) was performed on both hips of 40 patients to determine if the calculated bone-mineral density (BMD) of one hip could be used to predict the BMD of the opposite hip. For the Ward triangle, femoral neck, and greater trochanter the correlation coefficients between the BMD of the two hips was .920, .917, and .843, and the standard errors (SE) of the estimate for the linear regression of the left hip on to the right were 0.067, 0.063, and 0.077 g cm-2. The absolute error of predicting one hip from the other was not a function of BMD and thus the relative error increases with lower BMD values. The relative errors were 17%, 8%, and 7% for BMDs of 0.4, 0.8, and 1.0 g cm-2, respectively. The interobserver variability was small, with an r value of .96 and an SE of the estimate value of 0.036 g cm-2. The relative error in the mild-to-moderate osteoporosis categories was 2.5 times the precision of the instrument, indicating that the asymmetry of BMD is due to real differences between hips. Therefore the BMD of one hip cannot be used to predict that of the other with sufficient accuracy to discriminate clinically relevant trends in BMD.

No abstract

No abstract