Related Experiment Video
Updated: Jan 17, 2026
08:05
Design and Analysis for Fall Detection System Simplification
Published on: April 6, 2020
11.1K
Predicting fall risk among older adults in Chinese communities with advanced machine learning techniques: a
Aihong Liu1, Lingling Zhang1, Debin Huang2
1Department of Nursing, Union Hospital, Tongji Medical College, Huazhong University of Science and Technology, Wuhan, China.
Frontiers in Public Health
|September 17, 2025
Summary
A new machine learning model using CatBoost effectively predicts fall risk in older adults. This tool can help prevent falls through early clinical assessment and intervention strategies.
Area of Science:
- Gerontology
- Artificial Intelligence
- Public Health
Background:
- Community-dwelling older adults are at high risk of falls, leading to significant morbidity and mortality.
- Early identification of individuals at risk is crucial for implementing preventive measures.
- Existing fall risk assessment methods may lack the precision needed for effective intervention.
Purpose of the Study:
- To develop and evaluate an advanced machine learning model for predicting fall risk in community-dwelling elders.
- To identify key factors contributing to fall risk in this population.
- To provide actionable insights for early fall prevention strategies.
Main Methods:
- A cohort of 977 community-dwelling older adults was recruited, with data collected via structured questionnaires.
- Participants were randomly assigned to training (732) and testing (245) sets.
- Five machine learning models (Random Forest, GBDT, LGBM, XGBoost, CatBoost) and Logistic Regression were compared using AUC, accuracy, precision, sensitivity, specificity, and F1 score.
Main Results:
- The CatBoost model achieved the highest Area Under the Curve (AUC) of 0.8719, demonstrating superior predictive performance compared to other models.
- Key predictors identified by SHapley Additive exPlanations (SHAP) include history of falls, comorbidities, polypharmacy, sleep disorders, ADL, TUG results, frailty, and assistive device use.
- The overall fall incidence in the study population was 20.0% (195 out of 977 participants).
Conclusions:
- The CatBoost-based machine learning model shows excellent performance in predicting fall risk among community-dwelling older adults.
- This model can serve as a valuable tool for early clinical assessment and the development of targeted fall prevention programs.
- Identifying high-risk individuals enables timely interventions to reduce fall incidence and improve quality of life.
