Related Experiment Video
Updated: Sep 14, 2026
A Machine Learning Approach to Design an Efficient Selective Screening of Mild Cognitive Impairment
Published on: January 11, 2020
Nomogram for predicting the 12-year risk of ADL disability among older adults
Weifeng Qi1, Zhenhua Yin2, Yanping Sun3
1Department of Epidemiology and Health Statistics, Public Health College, Qingdao University, NO. 308 Ningxia Road, Qingdao, 266071, Shandong Province, China.
Objectives:
Previous studies have identified plenty of risk factors for activities of daily living (ADL). However, there are no reliable and widely available prediction models for ADL disability up to now. This study aimed to develop and validate a nomogram for predicting the 12-year risk of ADL disability in older adults.
Methods:
Data from 4,809 participants in the English Longitudinal Study of Ageing (ELSA) and 18,620 participants in the Survey of Health, Ageing and Retirement in Europe (SHARE) were used as training set and validation set, respectively. We used the least absolute shrinkage and selection operator (LASSO) and Cox regression to screen the predictors and develop the nomogram. The P value, concordance index (C-index), integrated area under the ROC (receiver operating characteristic) curve (AUC) and calibration curves were used to validate the nomogram.
Results:
During 12 years, 30.0% (n = 1,441) participants developed ADL disability in the training set, while the corresponding percentages were 18.5% in the validation set (n = 3,445). After screening, 13 variables were contained in the final prediction model. In ADL nomogram, the C-index and AUC were 0.744 ± 0.013 and 0.793 in internal valid ation, respectively, while in external validation, the C-index and AUC were 0.755 ± 0.009 and 0.796.
Conclusions:
This study developed and validated a nomogram that predicts functional disability. The application of the predictive model could have important implications for patient prognosis and health care.
