Is the NHANES III femoral neck database discordant with the total hip and trochanteric region databases?

The results of dual-energy X-ray absorptiometry (DXA) testing frequently leads to a significant clinical treatment decision, based on T-score or Z-score results. Using the nearly universally accepted World Health Organization (WHO) criteria, which in turn are based on the validity of the population database, a flawed database may lead to an incorrect and deleterious clinical categorization. We have observed that the Hologic National Health and Nutrition Examination Survey III (NHANES III) femoral neck scores are frequently disproportionately low compared with scores in the other major hip regions and that the discordance is statistically significant. Although the WHO now uses the femoral neck T-score (in postmenopausal women and men 50 yr or older), many clinicians still use the lowest (worst) T-score in any 1 of 3 major scan regions--femoral neck, total hip, or lumbar spine--suggesting that patient misclassification may occur, leading to the potential for unnecessary pharmacologic intervention.