Related Experiment Video
Updated: Dec 22, 2025
12:18
A Machine Learning Approach to Design an Efficient Selective Screening of Mild Cognitive Impairment
Published on: January 11, 2020
7.8K
Applying machine learning methods to develop a successful aging maintenance prediction model based on physical
TianPan Cai1, JingWen Long1, Jie Kuang1
1Jiangxi Province Key Laboratory of Preventive Medicine, Nanchang University, Nanchang, China.
Geriatrics & Gerontology International
|May 3, 2020
Summary
Machine learning models accurately predict successful aging (SA) using physical fitness tests. A deep learning model showed the best performance, highlighting the importance of fitness interventions for maintaining SA in older adults.
Area of Science:
- Gerontology and Artificial Intelligence
- Public Health and Aging Research
Background:
- Successful aging (SA) is a growing public health concern.
- Predicting SA maintenance is crucial for targeted interventions.
- Physical fitness is a key component of healthy aging.
Purpose of the Study:
- To develop and evaluate machine learning models for predicting successful aging (SA) based on physical fitness tests.
- To identify key physical fitness predictors of SA maintenance.
Main Methods:
- A cohort of 3657 community-dwelling adults aged ≥60 years was recruited.
- Physical fitness tests (balance, agility, speed, reactions, gait) and questionnaires were administered.
- Four machine learning models (logistic regression, deep learning, random forest, gradient boosting) were applied to predict SA over a 3-year follow-up.
Main Results:
- The baseline prevalence of SA was 26.99%, with an annual incidence rate of 11.04% transitioning to non-SA.
- All four machine learning models demonstrated high predictive performance (AUC >85%).
- The deep learning model achieved the highest accuracy (89.3%) and AUC (90.0%), while logistic regression excelled in sensitivity. Key predictors included age, arm curl, 30-s sit-to-stand, and reaction time.
Conclusions:
- Machine learning models, particularly deep learning, are effective for predicting successful aging (SA) maintenance.
- Physical fitness interventions are essential for promoting and sustaining SA in older adults.
- Identifying key fitness predictors can guide personalized interventions.
