Comparative evaluation of local and international reference databases for forearm densitometry: different impacts on diagnostic decisions.

Abstract

Reference databases play a key role in the management of osteoporosis. The aim of this preliminary study was to compare the diagnostic consequences of using either an international or a local reference database in peripheral densitometry. For this purpose, standard curves for bone mineral density (measured by dual-energy X-ray absorptiometry at the distal and proximal forearm) were generated for healthy Hungarian men and women. In total, 303 healthy volunteers of both sexes (age range: 20-94 yr) were recruited from four osteoporosis centers. Subjects with medical conditions or taking medication affecting the bone metabolism were excluded. Bone densitometry was performed with pDEXA (Norland-Stratec, Fort Atkinson, WI) devices in each center after cross-calibration of the machines. The precision error of the forearm measurement was also determined (<1% in vitro, and 1.2-2.5% in vivo). In females, the peak forearm density was detected in the 30-39-yr group. The density decreased by 8% per 5 yr in early postmenopausal females, and by 10% per 10 yr in late postmenopausal females. In males, the highest bone mineral density was found in the 30-39-yr group for the distal forearm, but 1 decade later for the proximal site. Subsequently, a 5% decrease in density occurred per 10 yr, except in the 8th decade, in which a 20% decrease was demonstrated. One thousand four hundred thirty-four patients with suspected osteoporosis were classified according to the forearm density T-scores using both the new Hungarian reference database and the international database provided by the manufacturer. Comparison of the results measured at the distal forearm with the two different databases led to similar outcomes. However, at the proximal site, one fifth of the female patients were reclassified from the low-density group to the normal group using the domestic normative database. An opposite difference was observed for the males: use of the Hungarian reference data resulted in 40% more men being categorized in the low-density group than when the international normal database was applied. Our results suggest that not only geographic differences, but also the reference database used, can influence the prevalence of the diagnosis of osteoporosis. Further data are currently being collected to increase the statistical power of the study.