Advances in digital phenotyping for relapse risk monitoring and community-based treatment and rehabilitation in methamphetamine use disorder
- 1School of Forensic Medicine, National Health Commission Key Laboratory of Drug Addiction Medicine, Kunming Medical University
Abstract
Methamphetamine use disorder (MUD) lacks widely approved pharmacological treatments with consistent efficacy, and community-based care relies largely on psychosocial interventions. Earlier identification of relapse risk in daily life is therefore a practical priority. Digital phenotyping uses smartphones, wearable devices, ecological momentary assessment (EMA), and passive sensing to collect repeated information on behavior, activity rhythms, subjective states, and environmental exposure in naturalistic settings. This narrative review maps evidence on digital and contextual signals for dynamic relapse-risk monitoring in MUD and examines how these signals may support human-reviewed community responses. It focuses on EMA, passive sensing, wearable devices, multimodal dynamic risk-state modeling, explainable risk prompts, just-in-time adaptive intervention (JITAI), and human-in-the-loop workflows. Direct MUD evidence primarily supports feasibility and concurrent or post-use correlates, while broader digital-intervention studies suggest potential benefits for treatment engagement and self-monitoring. These findings may support contextual understanding, but they do not establish prospectively validated relapse warning or demonstrated benefit from monitoring-triggered support. Long-term relapse outcomes, external validation, data standards, ethical governance, and real-world implementation remain limited. Digital phenotyping should therefore be viewed not as a mature automated relapse prediction system, but as a supportive framework for dynamic risk monitoring and stratified community response. Future work should prioritize multicenter validation and comparative evaluation of population-level, individualized, and hybrid risk models, including cold-start performance and recalibration, alongside transparent and explainable modeling, privacy-preserving data governance, and integration with existing community service workflows.
Related
- Sex-Differences in the Relationship between Methamphetamine-Use Experience and Spatial Memory in Non-Treatment Seeking Adults
- Continuous Glucose Monitoring for Metabolic Phenotyping and Precision Nutrition: Current Evidence, Clinical Applications and Future Perspectives
- Atrial Cardiomyopathy and Ischemic Stroke: A Monitoring-Integrated Staging and Phenotyping Framework Beyond Atrial Fibrillation
- Evaluation of treatment preferences among orthopedic surgeons and physical medicine and rehabilitation specialists for lateral epicondylitis : A cross-sectional survey study
- Co-production using Theory of Change in the development of the SUCCEED Africa community-based rehabilitation intervention for people living with psychosis in West and Southeast Africa
BibTeX
@article{wang2026advances,
title = {{Advances in digital phenotyping for relapse risk monitoring and community-based treatment and rehabilitation in methamphetamine use disorder}},
author = {Zhiwen Wang and Genmeng Yang and Ziyan Zhu and Ruike Xu and Yanxia Peng},
journal = {Frontiers in Psychiatry},
year = {2026},
month = nov,
doi = {10.3389/fpsyt.2026.1952829},
url = {https://www.frontiersin.org/articles/10.3389/fpsyt.2026.1952829/full},
note = {CC BY 4.0},
}
RIS
TY - JOUR
TI - Advances in digital phenotyping for relapse risk monitoring and community-based treatment and rehabilitation in methamphetamine use disorder
AU - Zhiwen Wang
AU - Genmeng Yang
AU - Ziyan Zhu
AU - Ruike Xu
AU - Yanxia Peng
PY - 2026
DA - 2026/11/05
JO - Frontiers in Psychiatry
DO - 10.3389/fpsyt.2026.1952829
UR - https://www.frontiersin.org/articles/10.3389/fpsyt.2026.1952829/full
AB - Methamphetamine use disorder (MUD) lacks widely approved pharmacological treatments with consistent efficacy, and community-based care relies largely on psychosocial interventions. Earlier identification of relapse risk in daily life is therefore a practical priority. Digital phenotyping uses smartphones, wearable devices, ecological momentary assessment (EMA), and passive sensing to collect repeated information on behavior, activity rhythms, subjective states, and environmental exposure in naturalistic settings. This narrative review maps evidence on digital and contextual signals for dynamic relapse-risk monitoring in MUD and examines how these signals may support human-reviewed community responses. It focuses on EMA, passive sensing, wearable devices, multimodal dynamic risk-state modeling, explainable risk prompts, just-in-time adaptive intervention (JITAI), and human-in-the-loop workflows. Direct MUD evidence primarily supports feasibility and concurrent or post-use correlates, while broader digital-intervention studies suggest potential benefits for treatment engagement and self-monitoring. These findings may support contextual understanding, but they do not establish prospectively validated relapse warning or demonstrated benefit from monitoring-triggered support. Long-term relapse outcomes, external validation, data standards, ethical governance, and real-world implementation remain limited. Digital phenotyping should therefore be viewed not as a mature automated relapse prediction system, but as a supportive framework for dynamic risk monitoring and stratified community response. Future work should prioritize multicenter validation and comparative evaluation of population-level, individualized, and hybrid risk models, including cold-start performance and recalibration, alongside transparent and explainable modeling, privacy-preserving data governance, and integration with existing community service workflows.
SN - 1664-0640
ER -