Construction and validation of a model for predicting disability in older adults with chronic lung disease

INTRODUCTION: Chronic lung disease and disability pose a serious threat to the physical and mental health of the older adults, and there are relatively few studies on the prediction of disability in this group. AIMS: This study was to explore the risk factors for disability in older adults with chronic lung disease and develop a validated risk prediction model. METHOD: This study involved 845 older adults with chronic lung diseases from the 2015 China Health and Retirement Longitudinal Study (CHARLS) survey. They were split into a training group (n = 592) and a validation set (n = 253) at a 7:3 ratio. Univariate and LASSO multiple logistic regression screened variables affecting disability occurrence. A multivariate logistic regression built the prediction model and created nomograms. A calibration curve assessed the model's accuracy, while the receiver operating characteristic curve's area and decision curve analysis evaluated its prediction performance. RESULTS: The final model included age, gender, falls, physical pain, depression, health satisfaction, child satisfaction, intellectual disability, SBP, respiratory function.The model demonstrated good discrimination (AUC 76.5%, 95% confidence interval CI 0.70-0.80). The calibration curve closely aligned with the ideal curve, and the Brier score was 0.193. For the validation set, the model demonstrated comparable discrimination (AUC 76%, 95% CI 0.70-0.82), Decision curve analysis demonstrated the high clinical utility of the model. CONCLUSION: The predictive model of disability in older adults with chronic lung disease constructed on the basis of the CHARLS with logistic regression selection of factors is good and has a certain predictive value for the risk of disability in older adults with chronic lung disease.