Predicting and explaining social isolation: Insights from an interpretable machine learning model in ageing populations
Li, Sicheng Lam, Kyle Qiu, Jianing Fang, Ya
Li, Sicheng
Lam, Kyle
Qiu, Jianing
Fang, Ya
Author
Supervisor
Department
Personalized Medicine
Embargo End Date
Type
Journal article
Date
2025
License
Language
English
Collections
Research Projects
Organizational Units
Journal Issue
Abstract
Background and Objectives
Social isolation affects one in four people and is associated with adverse health outcomes, yet accurate social isolation risk prediction models remain lacking. This study develops and validates an interpretable machine learning (ML) approach to predict social isolation and identify key predictors among middle-aged and older adults in China.
Research Design and Methods
Training data came from the China Health and Retirement Longitudinal Study. Baseline data from 2011 included 12,785, 12,323, and 11,590 participants for the 2-, 4-, and 7-year prediction models. External validation used data from the China Family Panel Studies 2010, 2012, and 2016. Five ML algorithms were used to construct prediction models with 283 candidate predictors. SHapley Additive exPlanations explained the feature importance. Classic logistic regression and restricted cubic spline (RCS) explored potential causal associations.
Results
In the development phase, the Gradient Boosting Machine (GBM) achieved the best performance across 2-, 4-, and 7-year models (AUC-ROC = 0.767, 0.729, and 0.749). In the external validation, the GBM had AUC-ROC with 0.649 and 0.678 for the 2- and 7-year prediction models. Age, monthly non-food consumption, and net primary residence value were consistently identified as the top predictors. Environmental exposures (greenness exposure, rainy days) and community environment (convenience stores, out-migrants) also emerged as important predictors. RCS analysis revealed non-linear associations between these external factors and social isolation.
Discussion and Implications
With multimodal data, our best-performing model GBM outperformed existing models for identifying social isolation risk. The interpretability of our model highlights actionable and potentially reversible targets, especially at a community and environmental level.
Citation
S. Li, K. Lam, J. Qiu, and Y. Fang, “Predicting and explaining social isolation: Insights from an interpretable machine learning model in ageing populations,” Gerontologist, Dec. 2025, doi: 10.1093/GERONT/GNAF297
Source
Gerontologist
Conference
Keywords
Social isolation, Interpretable Machine learning, Intervention target, Older adults, China
Subjects
Source
Publisher
Oxford University Press