Related Experiment Video
Updated: Jan 6, 2026
12:18
A Machine Learning Approach to Design an Efficient Selective Screening of Mild Cognitive Impairment
Published on: January 11, 2020
7.9K
Functional disability screening in the elderly: a machine learning approach with ELSI-Brazil data
Dalton Breno Costa1, Carmen Moret-Tatay2, João Carlos Néto3
1Pontifícia Universidade Católica Do Rio Grande Do Sul, PUCRS, Porto Alegre, Brazil.
Geroscience
|November 1, 2025
Summary
Machine learning models effectively predict functional disability in older adults using health and sociodemographic data. Key indicators include depressive symptoms and self-rated health, enabling early risk identification for interventions.
Area of Science:
- Gerontology
- Artificial Intelligence
- Public Health
Background:
- Functional disability in the elderly is a growing public health concern.
- Early identification and intervention are crucial for maintaining quality of life and reducing healthcare costs.
Purpose of the Study:
- To investigate, validate, and apply Machine Learning (ML) algorithms for predicting functional disability in elderly individuals.
- To identify multidimensional risk indicators for functional disability.
- To understand factors influencing functional disability screening.
Main Methods:
- Analysis of data from 4502 participants in the ELSI-Brazil study (2015-2016).
- Development of classification models using 49 predictor variables (sociodemographic, health, behavioral).
- Application of SMOTE, tenfold cross-validation, Bayesian optimization, and SHAP analysis for model interpretation.
Main Results:
- The Ridge Classifier model demonstrated robust performance (ROC-AUC: 0.785).
- Key predictors for functional disability included depressive symptoms, concern about mobility, and self-rated health.
- High negative predictive value (84.5%) indicates effective screening.
Conclusions:
- ML techniques integrated with multidimensional health data offer a promising tool for early screening and intervention of functional disability in the elderly.
- This approach can support clinical decision-making and health policies for active and healthy aging.
- Identifying key risk factors like mental health and self-perceived health is vital for targeted interventions.
