| Objective To construct a risk prediction tool for osteoporosis (OP) in postmenopausal women based on traditional Chinese medicine (TCM) syndromes.?Methods The general information and TCM symptoms of the subjects were collected with questionnaire survey. Bone mineral density was detected.??Logistic regression method was used to construct the model, and its predictive efficiency was analyzed by comparing it with osteoporosis self-assessment tool for Asians (OSTA) index, which is a frequently-used forecasting tool. External data were used to further verify the predictive efficiency.??Results Age, weight, hunchback, and bradypsychia were the risk factors for OP according to the analysis. The model was constructed by combining these factors with the typical TCM symptoms of kidney deficiency syndrome with OP, such as inability of waist and knee, lower limb cramps, nocturnal frequent urination, and tinnitus. The area under ROC curve (AUC) of the model was 0.765, which was higher than that of OSTA index.??The optimal cut-off threshold was > ?0.79, the sensitivity was 69.43, and the specificity was 72.60.?Validating?with the external data further, the sensitivity and specificity were 68.85 and 75.00, respectively.??Conclusion TCM syndromes play a role in the risk prediction of OP. In this study, the prediction performance of COPT model is better than that of OSTA index. The verification results of the external data further support that the model has good performance in OP prediction. |