Discordant normal ranges for proximal femur bone density in Australia

The aim of the study was to investigate the reference ranges currently in use for femoral neck bone mineral density (BMD) in Australia. Representative scans were obtained of the proximal femur from 78 dual-energy X-ray absorptiometry centers across Australia. Femoral neck BMD was standardized using published conversion equations and corresponding T-scores calculated using an Australian reference range (ARR) and using the NHANES-III reference range (NH3RR). Reported T-scores (TR) from the DXA centers were compared with T-scores derived from the NH3RR (TNH) and from the ARR (TA). Examination of the regression lines comparing TR with TNH, and TR with TA, demonstrated that the reference ranges used by centers with Lunar, and some of the Norland and Hologic instruments, differed only slightly from both the NH3RR and the ARR. A number of centers, however, with Norland or Hologic scanners, were using reference ranges which differed from both the NH3RR and ARR. The reference ranges used with these latter Hologic or Norland scanners could result in T-score differences of up to I standard deviation in the clinically relevant T-score range. The NH3RR was compared with the ARR using a set of BMD values covering the clinical range. There was close agreement between TNH and corresponding TA, confirming close agreement between the NH3RR and ARR. There are thus clinically important differences in the reference ranges currently used within Australia for interpretation of femoral neck BMD. For consistency, and in view of its validation, the NH3RR is recommended as the most appropriate Caucasian normal range in Australia. The ARR for femoral neck BMD does not differ significantly from the NH3RR and is also acceptable. Problems relating to the use of discordant BMD reference ranges potentially may exist in other countries.