Related Experiment Video
Updated: Jan 15, 2026

Author Spotlight: Automated Lifespan Monitoring – Discovering Aging Dynamics with the Lifespan Machine
Published on: January 26, 2024
Identifying subjective life expectancy risk factors in physically active and inactive middle-aged and older adults
Jian Yang1, Zhihui Li2, Ming Wu3
1College of Physical Education and Health, East China Normal University, Shanghai, China.
Machine learning models identified key risk factors for subjective life expectancy in older adults. Age and perceived health were crucial, guiding differentiated health strategies for active and inactive individuals.
Area of Science:
- Public Health
- Gerontology
- Computational Biology
Background:
- Physical activity significantly impacts public health and is linked to well-being and subjective life expectancy.
- Understanding risk factors for subjective life expectancy is crucial for developing targeted health interventions.
- This study focuses on differentiating these factors based on physical activity levels in middle-aged and older adults.
Purpose of the Study:
- To identify risk factors for subjective life expectancy in middle-aged and older adults.
- To compare these factors between individuals with active and inactive physical activity levels.
- To provide an evidence base for differentiated health intervention strategies.
Main Methods:
- Utilized data from 10,945 participants in the China Health and Retirement Longitudinal Study (CHARLS) 2018.
- Developed and compared five machine learning models (RF, LR, SVM, XGBoost, LightGBM) for active and inactive groups.
- Employed Synthetic Minority Over-sampling Technique (SMOTE), 70/30 train/test split, and ten-fold cross-validation with grid search for model optimization and evaluation (AUC, accuracy, sensitivity, specificity, F1-score).
Main Results:
- Support Vector Machine (SVM) performed best in the inactive group (AUC: 0.797), identifying 'age' as the key factor.
- Light Gradient Boosting Machine (LightGBM) performed best in the active group (AUC: 0.775), identifying 'perceived health' as the key factor.
- Significant differences in model performance and key risk factors were observed between active and inactive groups.
Conclusions:
- Machine learning models effectively identify critical risk factors for subjective life expectancy in older adults.
- Findings highlight the importance of 'age' for inactive individuals and 'perceived health' for active individuals.
- Results support the development of tailored health management strategies based on physical activity levels.
More Related Videos
12:18A Machine Learning Approach to Design an Efficient Selective Screening of Mild Cognitive Impairment
Published on: January 11, 2020
11:21Methodology for Establishing a Community-Wide Life Laboratory for Capturing Unobtrusive and Continuous Remote Activity and Health Data
Published on: July 27, 2018