Related Experiment Video
Updated: Oct 2, 2025
A Machine Learning Approach to Design an Efficient Selective Screening of Mild Cognitive Impairment
Published on: January 11, 2020
Developing and validating a Chinese multimorbidity-weighted index for middle-aged and older community-dwelling
Wei-Hua Hu1, Yu-Yang Liu1, Cong-Hui Yang1
1Department of Medical Statistics, School of Public Health, Sun Yat-sen University, Guangzhou, P.R. China.
Objective:
To develop and validate an index to quantify the multimorbidity burden in Chinese middle-aged and older community-dwelling individuals.
Methods:
We included 20,035 individuals aged 45 and older from the China Health and Retirement Longitudinal Study (CHARLS) and 19,297 individuals aged 65 and older from the Chinese Longitudinal Healthy Longevity Survey (CLHLS). Health outcomes of physical functioning (PF), basic and instrumental activities of daily living (ADL and IADL) and mortality were obtained. Based on self-reported disease status, we calculated five commonly used western multimorbidity indexes for CHARLS baseline participants. The one that predicted the health outcomes the best was selected and then modified through a linear mixed model using the repeated individual data in CHARLS. The performance of the modified index was internally and externally evaluated with CHARLS and CLHLS data.
Results:
The multimorbidity-weighted index (MWI) performed the best among the five indexes. In the modified Chinese multimorbidity-weighted index (CMWI), the weights of the diseases varied greatly (range 0.2-5.1). The top three diseases with the highest impact were stroke, memory-related diseases and cancer, corresponding to weights of 5.1, 4.3 and 3.4, respectively. Compared with the MWI, the CMWI showed better model fits for PF and IADL with larger R2 and smaller Akaike information criterion, and comparable prediction performances for ADL, IADL and mortality (e.g. the same predictive accuracy of 0.80 for ADL disability).
Conclusion:
The CMWI is an adequate index to quantify the multimorbidity burden for Chinese middle-aged and older community-dwelling individuals. It can be directly computed via disease status examined in regular community health check-ups to facilitate health management.
