The Value Of FRAX (?) In Predicting The Risk Of Osteoporotic Fracture In The Main Part Of Menopausal Non - Osteoporosis In Beijing

Objective1. To assess the ability of the Chinese version of the WHO fracture risk assessment tool (FRAX(?)) to predict the incident major osteoporotic fractures of non-osteoporotic women in the Peking Vertebral Fracture Study cohort;2. To investigate whether the use of femoral bone mineral density influence the prediction results of FRAX(?) tools;3. To construct a new risk assessment model utilizing the lumbar spine BMD and FRAX(?) scores, and then to compare its prediction with original FRAX(?) tools;4. To analyze the effect of the rate of bone mineral density change of femoral neck in five years on the risk of incident major osteoporotic fracture.Subjects and methods1. Subjects:1100 postmenopausal women participated in our PK-VF follow-up study. After excluding 268 subjects according to the exclusion criteria,832 women, aged 40 or older, were enrolled in our prospective study.2. Methods:(1) Data about clinical risk factors was adopted by questionnaires, radiology and dual-energy x-ray absorptiometry(DXA). In the records of major osteoporotic fractures, vertebral fractures were confirmed by radiology, whereas the results of other non-vertebral fractures were acquired by questionnaires;(2) The estimated individual 10-year absolute risks of major osteoporotic fractures were calculated with the FRAX(?) algorithms for China both with and without the inclusion of the bone mineral density (BMD) measurements. Calibration was evaluated by comparing the five-year estimated (E) fractures predicted with FRAX(?) with the number of observed (O) fractures, using Hosmer-Lemeshow goodness-of-fit test to calculate a Chi square value;(3) Their discriminative ability for the probability of major osteoporotic fractures was assessed by the area under the receiving operating characteristic (ROC) curves. The AUC values of FRAX(?) with and without the inclusion of BMD were compared then; (4) The lumbar spine BMD and FRAX(?) tools are combined to construct new logistic regression equation, ROC curve was also obtained, and its AUC value was compared with the original FRAX(?) tools; (5) The rate of femoral neck BMD change in five years was calculated, and the relationship between the incident major osteoporotic fractures and the rate was assessed by logistic regression analysis.Results1. Fifty (6.0%) women sustained at least one incident major osteoporotic fracture (hip, vertebral, forearm or humerus), whereas no patients hits the standard of high risk;2. The E/O ratios for FRAX(?) tools calculated with and without femoral neck BMD were 0.30 and 0.38 respectively. The difference between E and O cases reached statistical significance (χ2, p<0.05);3. Areas under the ROC curves were 0.620 and 0.606 for FRAX(?) tools with and without femoral neck BMD respectively, with no statistical significance;4. AUC values of the new risk assessment model combined with lumbar spine BMD was 0.689, significantly higher than that of FRAX(?) tools with femoral neck BMD;5. No significant correlation was found between the rate of femoral neck BMD change in five years and incident major osteoporotic fractures.Conclusions1. The Chinese FRAX(?) tools underestimates the risk for major osteoporotic fractures in non-osteoporotic women, the algorithm had only modest calibration and discriminative ability;2. There is no significant difference between the prediction of FRAX(?) tools calculated with and without femoral neck BMD;3. Lumbar spine BMD predicts incident major osteoporotic fractures and adds moderate prediction information beyond FRAX(?) tools calculated with femoral neck BMD.4. The rate of change of femoral neck BMD in 5 years is not enough to predict the risk of major osteoporotic fracture.