Is a calculated total hip BMD of clinical use?

Abstract

The diagnosis of osteoporosis is based on bone mass measurement. To avoid the errors associated with the measurement of spinal bone density the total hip has been accepted as the standard measurement site. This information is not available for many early measurements. We have assessed whether it is possible to derive clinically useful information about total hip bone mineral density (BMD) from measurements at other hip sites. The bone mass measurements of 46 patients participating in a current trial of therapy for osteoporosis were reviewed. The total hip BMD as directly measured was compared with that obtained from the formula: Total hip BMD = 0.48 x Neck BMD + 0.62 x Trochanteric BMD + 0.03. In 30 patients with follow-up data the rate of change in hip BMD over a year was also determined by both methods. In the pretreatment state there was good agreement between the two measures (r2 = 0.96, SEE 0.012 g/cm2). If the formula was used to compute a change in total hip BMD, the agreement between both methods remained good. However, the standard error of the estimate of the change represented 59% of the observed change. This indicates that the error associated with this estimate is too great to allow clinically meaningful conclusions to be drawn from calculated total hip BMD. We conclude that, whilst it may be possible to obtain reasonable point estimates of total hip BMD from other measures in the hip, these estimates are too imprecise to allow conclusions about change in BMD to be made.