[Development of a simple predicting tool for low bone mass of postmenopausal women: a study in Shanghai].

OBJECTIVE

To develop a simple screening tool for low bone mass of postmenopausal women.

METHODS

405 postmenopausal women in Shanghai who visited the department of osteoporosis consecutively, aged 62.8 +/- 8.0 (47 approximately 90), underwent questionnaire survey on the risk factors of osteoporosis and fracture. Dual energy X-ray absorptiometry (DXA) was conducted on the left or right femoral neck to measure the bone mineral density (BMD) to identify osteoporosis (T-score <or= -2.5). Univariate linear regression was conducted to identify the variables with significant association with the femoral neck BMD to be used in multiple variable regression analysis. The screening index was obtained by the formula: index = independent variable X corresponding weight. A receiver operating characteristic (ROC) curve was drawn with the sensitivity (true positive rate) as the vertical coordinate and 1-specificity (false positive rate) as the horizontal coordinate. The area under the ROC curve (AUC ROC) was calculated so as to establish a parsimonious model.

RESULTS

The final tool was based only on age and body weight. The formula of screening index was: index = 2 x weight (kg)/10 + [-1 x age (year)/10]. The risk index thus obtained had a sensitivity of 93.4%, a specificity of 52.6%, and an AUC ROC of 0.818 (95% CI for the mean: 0.766 approximately 0.870). 58% of the high-risk women had osteoporosis, compared with 26% and 2% of the intermediate and low-risk women respectively.

CONCLUSION

The established and verified screening tool can easily predict the bone mass status in the postmenopausal women, thus screening the high-risk population and saving the cost of BMD measurement unnecessary for the low-risk population.