The prediction of bone mineral density (BMD) response to osteoporosis treatment remains challenging as current prediction models rely on complex clinical data or sequential BMD measurements. This study aims to research the association between baseline 24-hour urinary creatinine excretion (24 h Ucr) and the change of BMD after treatment with zoledronic acid (ZOL) and develop an effective model for BMD improvement prediction in older postmenopausal women with osteoporosis. A total of 135 postmenopausal women (aged ≥ 60 years) with osteoporosis receiving ZOL treatment were included. Correlation analysis was performed to assess the association of baseline 24 h Ucr with BMD change after one year of ZOL treatment. The prediction model for the changes in BMD was established by multivariable binary logistic regression analysis with backward selection. Bootstrap resampling was used to internally validate the final model. Higher baseline 24 h Ucr was negatively correlated with lumbar BMD changes (r =－0.230, p = 0.007). After adjusting for confounding factors, baseline 24 h Ucr was found to be an independent predictor of changes in lumbar BMD after ZOL treatment (odds ratio 0.694, 95% confidence interval 0.530–0.908, p = 0.008). The new model logit (P) developed achieved a strong area under the curve (AUC) of 0.831 for discriminating treatment responders. The model was internally validated using the bootstrap re-sampling procedure that calculated small AUC optimisms of 0.018 (95% CI－0.060–0.085). Our study identifies baseline 24 h Ucr as a promising, readily available and independent predictor for the response to ZOL therapy. This finding offers a potential tool for personalising geriatric osteoporosis management by identifying patients most likely to achieve significant BMD improvement.
Zeng et al. (Thu,) studied this question.