WHEN SHOULD BONE-DENSITY MEASUREMENTS BE REPEATED

We calculated how long to wait before repeating bone mineral density ( BMD) measurements to reassess fracture risk. Correlation results from serial measurements of 495 postmenopausal Japanese-American women were used to estimate 95% confidence intervals (CI) for future BMD. After 7 years of follow-up, BMD correlations with the initial measurement ra nged between 0.81 and 0.94, depending on age group and measurement sit e. In this analysis, the period between measurements was defined as th e time required for the lower 95% CI to fall below the BMD value corre sponding to doubling of fracture risk. Progressive bone loss causes fr acture risk to double after 10 years, on average. However, the 95% CIs indicate that a second BMD measurement will detect risk doubling afte r only 2 or 3 years for some women. For untreated, early postmenopausa l women, the period between measurements was approximately 2-5 years f or the radius and 4-6 years for the calcaneus, depending on the initia l BMD level. The period was approximately 1 year longer for women age 60 and older. Treatments that halve the bone loss rate would increase the period by 1-3 years. In the absence of a second measurement of BMD , the CI will continue to expand with time, corresponding to a wider r ange in risk between individuals, and a greater proportion of women wi ll be at increased fracture risk. Obtaining a second BMD measurement p inpoints the patient's status within the precision of the measurement. We conclude that repeated BMD measurements will provide a more accura te estimate of fracture risk than a single, baseline measurement.