Machine learning applications across living environments, activity behaviors, and mental health among older adults: a systematic review
- 1School of Human Settlements Science, University of the Chinese Academy of Sciences
Abstract
Background: Machine learning is increasingly used to examine how living environments, activity behaviors, and mental health are interconnected in later life. However, evidence across these domains remains dispersed, and the contribution of machine learning to understanding their relationships has not been systematically synthesized.
Methods: Following PRISMA 2020-informed procedures, we systematically reviewed empirical studies published between 1 January 2016 and 30 August 2026. Eligible studies included participants aged 60 years or older or reported separately extractable findings for this age group, examined relationships involving at least two of living environment, activity behaviors, and mental health, and used machine learning as a substantive component of the empirical analysis.
Results: Of 4,337 records identified, 27 studies met the eligibility criteria. Evidence was concentrated on individual relationships between living environment, activity behaviors, and mental health, while relatively few studies considered all three domains together. Across this literature, machine learning was used to construct environmental and behavioral measures, predict behavioral and mental-health outcomes, and characterize nonlinear relationships, thresholds, interactions, and population or spatial heterogeneity. Collectively, these studies point to considerable variation in environment–behavior–health relationships across populations, places, and analytical contexts. Nevertheless, the evidence remains distributed across partially connected relationships, with limited integration of activity behaviors into analyses linking living environments with mental health.
Conclusion: Current research provides a growing but fragmented evidence base on the interconnections among living environments, activity behaviors, and mental health in later life. Machine learning offers new opportunities to characterize the complexity and heterogeneity of these relationships. Greater integration of environmental, behavioral, and mental-health evidence across populations, places, and time could provide a stronger basis for understanding how living environments relate to mental health in later life.
Related
- Fairness in multimodal machine learning applications in clinical decision support: a systematic review
- Bridging the Gap Between Research and Practice: A Systematic Literature Review of IoT and Machine Learning Applications in Mental Healthcare Systems
- Machine Learning Applications for Wearable Sensor Data in Tennis and Padel: A Systematic Review of Movement Recognition, Performance Monitoring, and Emerging Injury-Related Applications.
- Antidepressant medication use and objectively measured physical activity and sedentary behaviors in adults: A cross-sectional analysis of a nationally representative sample of Canadian adults
- A systematic review to characterize functional tic-like behaviors
BibTeX
@article{wang2026machine,
title = {{Machine learning applications across living environments, activity behaviors, and mental health among older adults: a systematic review}},
author = {Yuxuan Wang and Jun Lan},
journal = {Frontiers in Public Health},
year = {2026},
month = oct,
doi = {10.3389/fpubh.2026.1942765},
url = {https://www.frontiersin.org/articles/10.3389/fpubh.2026.1942765/full},
note = {CC BY 4.0},
}
RIS
TY - JOUR
TI - Machine learning applications across living environments, activity behaviors, and mental health among older adults: a systematic review
AU - Yuxuan Wang
AU - Jun Lan
PY - 2026
DA - 2026/10/26
JO - Frontiers in Public Health
DO - 10.3389/fpubh.2026.1942765
UR - https://www.frontiersin.org/articles/10.3389/fpubh.2026.1942765/full
AB - Background: Machine learning is increasingly used to examine how living environments, activity behaviors, and mental health are interconnected in later life. However, evidence across these domains remains dispersed, and the contribution of machine learning to understanding their relationships has not been systematically synthesized. Methods: Following PRISMA 2020-informed procedures, we systematically reviewed empirical studies published between 1 January 2016 and 30 August 2026. Eligible studies included participants aged 60 years or older or reported separately extractable findings for this age group, examined relationships involving at least two of living environment, activity behaviors, and mental health, and used machine learning as a substantive component of the empirical analysis. Results: Of 4,337 records identified, 27 studies met the eligibility criteria. Evidence was concentrated on individual relationships between living environment, activity behaviors, and mental health, while relatively few studies considered all three domains together. Across this literature, machine learning was used to construct environmental and behavioral measures, predict behavioral and mental-health outcomes, and characterize nonlinear relationships, thresholds, interactions, and population or spatial heterogeneity. Collectively, these studies point to considerable variation in environment–behavior–health relationships across populations, places, and analytical contexts. Nevertheless, the evidence remains distributed across partially connected relationships, with limited integration of activity behaviors into analyses linking living environments with mental health. Conclusion: Current research provides a growing but fragmented evidence base on the interconnections among living environments, activity behaviors, and mental health in later life. Machine learning offers new opportunities to characterize the complexity and heterogeneity of these relationships. Greater integration of environmental, behavioral, and mental-health evidence across populations, places, and time could provide a stronger basis for understanding how living environments relate to mental health in later life.
SN - 2296-2565
ER -