Related Experiment Video
Updated: Aug 15, 2026
05:19
Closed-Loop Neurostimulation for Biomarker-Driven, Personalized Treatment of Major Depressive Disorder
Published on: July 7, 2023
Digital Markers for Passive Remote Monitoring of Bipolar Disorder: Systematic Review
Thomas P Kutcher1, Isha Chakraborty1, Kristin Kostick-Quenet2
1Department of Electrical & Computer Engineering, Rice University, Houston, TX, United States.
JMIR Mental Health
|August 13, 2026
Summary
Passive digital markers from wearables show promise for detecting bipolar disorder (BD) mood shifts by monitoring activity, sleep, and speech. Further research is needed to validate these digital health tools for clinical use.
Area of Science:
- Digital Health
- Computational Psychiatry
- Wearable Technology
Background:
- Bipolar disorder (BD) is characterized by mood episodes (mania, depression) that are difficult to detect early due to infrequent clinical contact.
- Wearable devices and smartphones offer continuous, real-world monitoring of behavior and physiology for potential insights into BD mood dynamics.
Purpose of the Study:
- To systematically review passively collected digital markers for bipolar disorder (BD) mood states.
- To characterize devices, modalities, and analytical approaches used in BD digital marker research.
- To identify gaps and priorities for translating digital marker research into clinical practice.
Main Methods:
- A systematic review following PRISMA guidelines was conducted across multiple databases (MEDLINE, PsycINFO, Scopus, IEEE Xplore, ACM Digital Library).
- Studies included adults with Bipolar I or II disorder measuring passively collected digital markers correlated with mood states (depressive, manic, hypomanic, mixed, euthymic).
- Exclusion criteria included studies relying solely on active measures; risk of bias was assessed.
Main Results:
- Fifty-seven studies were included, with most having small sample sizes and short monitoring durations.
- Eight digital marker domains were identified: activity, heart rate (HR), electrodermal activity (EDA), geolocation, smartphone use, light exposure, sleep, and speech.
- Consistent patterns linked depression to reduced mobility/socialization and mania/hypomania to increased activity/communication; circadian features and speech/keyboard data showed predictive potential, though external validation was often lacking.
Conclusions:
- Passively collected digital markers show promise for detecting BD mood states, particularly those related to sleep-wake patterns, activity, socialization, geolocation, and speech.
- Clinical translation requires longer monitoring periods, standardized features, external validation, privacy-preserving data collection, and expanded physiological measurements beyond HR and EDA.
- Developing reliable digital tools is crucial for earlier detection and improved management of bipolar disorder episodes.