Abstract
PURPOSE: Osteoporosis significantly increases fracture risk and mortality, yet robust tools for predicting long-term mortality in this population are lacking.This study aimed to develop and validate a nomogram for predicting 5-year all-cause mortality among patients with osteoporosis. METHODS: A retrospective cohort study was conducted using data from 2,165 osteoporosis patients sourced from the NHANES database (2007-2023; training set) and 304 patients from Dandong Central Hospital (2017-2024; validation set). Potential risk factors were analyzed through LASSO regression, followed by multivariate logistic regression, to identify independent predictors.A nomogram was constructed employing significant predictors. Finally, the C-index, ROC curve, calibration curve, and decision curve analysis were utilized to validate the model in both the training and validation sets. RESULTS: In the study population, 192 patients died in the training set and 36 patients died in the experimental set. At the same time, we collected detailed baseline demographic data. Specifically, the age distribution of the training set was 56.07 ± 17.62, and that of the experimental set was 57.11 ± 18.34. Among them, 49.52% of the training set were male, and 50.99% of the experimental set were male. During the study period, we recorded 228 deaths. Seven independent predictors of 5-year all-cause mortality were identified: increased Age (OR=1.090,95%Cl: 1.115-2.313), Male gender (OR=1.606,95%Cl: 1.071-1.109), Smoking (OR=1.945,95%Cl: 1.289-2.933), higher FBG (OR=1.006,95%Cl: 1.002-1.010), higher Uric acid (OR=1.177,95%Cl: 1.039-1.332); Alcohol use (OR=0.583,95%Cl: 0.410-0.827) and higher BMI (OR=0.946,95%Cl: 0.909-0.985) were protective. The resulting nomogram demonstrated strong discriminatory ability in both the training set (AUC = 0.834) and validation set (AUC = 0.862). In the validation set, the precision rate was 0.514, the recall rate was 0.5, and the F1-score was 0.507. Calibration plots and the Hosmer-Lemeshow test indicated good agreement between predicted and observed outcomes (p > 0.05). Decision curve analysis confirmed significant clinical utility across a wide range of risk thresholds. CONCLUSION: This study developed and validated a novel nomogram incorporating seven common clinical factors, which can predict the 5-year all-cause mortality risk in patients with osteoporosis. Although the tool demonstrated good performance and has the potential to assist in clinical risk stratification and personalized management, there are still some limitations in the study design. Therefore, its clinical applicability should be interpreted with caution until further external validation.