The Tromsø study: determinants of precision in bone densitometry.

Studies of precision determinants in bone densitometry are scarce. A total of 111 subjects recruited from the population-based multipurpose Tromsø Study (Norway), 27-75 years of age, had repeated forearm bone single X-ray absorptiometry (SXA) measurements. Measurement conditions were systematically varied in series up to eight scans. Median coefficients of variation (CV) for two scans performed 1 week apart, by two different operators were 0.79% and 0.98% at distal and ultradistal sites, respectively. The CV distribution was skewed: 5% of the subjects had individual CVs above 2.2% (distal) and 3.4% (ultradistal). Age (P = 0.0097) and repositioning were important determinants of precision. The SXA bone mineral density (BMD)-measurement method is sufficiently precise to establish BMD level. The minimal individual percentage BMD change that can be detected with 95% certainty was 2% and 3% at distal and ultradistal sites, respectively. Detection of BMD changes less than this should rely on multiple repeat measurements at each point in time.