Publications

Abstract

BACKGROUND: The aim of this study was to examine the effect of different reference ranges in bone mineral density on the diagnosis of osteoporosis. METHODS: This cross-sectional study involved 357 men and 870 women aged between 18 and 89 years, who were randomly sampled from various districts within Ho Chi Minh City, Vietnam. BMD at the femoral neck, lumbar spine and whole body was measured by DXA (Hologic QDR4500). Polynomial regression models and bootstraps method were used to determine peak BMD and standard deviation (SD). Based on the two parameters, we computed T-scores (denoted by TVN) for each individual in the study. A similar diagnosis was also done based on T-scores provided by the densitometer (TDXA), which is based on the US White population (NHANES III). We then compared the concordance between TVN and TDXA in the classification of osteoporosis. Osteoporosis was defined according to the World Health Organization criteria. RESULTS: In post-menopausal women, the prevalence of osteoporosis based on femoral neck TVN was 29%, but when the diagnosis was based on TDXA, the prevalence was 44%. In men aged 50+ years, the TVN-based prevalence of osteoporosis was 10%, which was lower than TDXA-based prevalence (30%). Among 177 women who were diagnosed with osteoporosis by TDXA, 35% were actually osteopenia by TVN. The kappa-statistic was 0.54 for women and 0.41 for men. CONCLUSION: These data suggest that the T-scores provided by the Hologic QDR4500 over-diagnosed osteoporosis in Vietnamese men and women. This over-diagnosis could lead to over-treatment and influence the decision of recruitment of participants in clinical trials.