Constructing the prediction model based on DXA between sarcopenia and BMD in middle-aged and elderly men with T2DM.

OBJECTIVE

To explore the relationship between sarcopenia and bone mineral density (BMD) in middle-aged and elderly male patients with type 2 diabetes mellitus (T2DM), construct a prediction model for sarcopenia based on dual-energy X-ray absorptiometry (DXA), and evaluate its clinical value.

METHODS

A total of 523 middle-aged and elderly male patients with T2DM in the hospital from January 2021 to December 2024 were selected and divided into the training set (366 cases) and the validation set (157 cases) at a ratio of 7:3. The BMD -value was measured by DXA, and clinical data were collected. A prediction model was constructed using multivariate logistic regression in the training set, and the model efficacy was evaluated by receiver operating characteristic (ROC) curve, calibration curve, and decision curve analysis (DCA).

RESULTS

The incidence of sarcopenia was 27.05% (99/366) in the training set and 28.02% (44/157) in the validation set. Multivariate analysis showed that age, HbA1c, and HOMA-IR were independent risk factors for sarcopenia, while the lumbar L1-L4 -value, and femoral neck -value were independent protective factors for sarcopenia ( < 0.05). The C-index of the nomogram model were 0.773 (in the training set) and 0.750 (in the validation set) respectively. The calibration curve showed good agreement between predicted and actual values, and the Hosmer-Lemeshow test were significant (all  > 0.05). The ROC curve showed the area under the curve (AUC) of the nomogram model for predicting the risk of sarcopenia was 0.773 (95% CI: 0.652-0.895) and 0.750 (95% CI, 0.686-0.814) in the training set and the validation set, respectively. The sensitivity and specificity were 0.714, 0.887 and 0.688, 0.796, respectively.

CONCLUSION

The prediction model constructed based on DXA can effectively predict the risk of sarcopenia in middle-aged and elderly male patients with T2DM, providing a basis for clinical early screening and intervention.