Wearable data may inform mood prediction, but useful signals vary by device
A review identified movement, heart rate and sleep measures in prediction research, without establishing a device's ability to diagnose depression or anxiety.
Based on the published abstract. The full paper may contain additional methods, results and limitations.
The 30-second takeaway
Across studies combining phones with wearable devices, movement, step counts, heart rate and sleep emerged as recurring inputs for predicting depression or anxiety. Their reported usefulness varied by device. The review describes which data features researchers used and considered important; it does not establish that a wearable can accurately diagnose a mental health condition.
The abstract gives no pooled prediction accuracy, error rates or monitoring durations. Feature importance and frequency of use cannot establish whether a model works reliably for an individual or in routine care.
What the review measured
Researchers measured how often each feature appeared in studies and how often it was judged important when used. Those metrics summarize research patterns; they are not direct tests of how accurately a device identifies depression or anxiety.
Devices showed different patterns
In smartwatch studies, sleep and heart rate stood out, while movement and step measures were common but often not important. Actiwatch studies emphasized activity, and smart band studies also highlighted phone use.
The original publication
Key Features of Digital Phenotyping for Monitoring Mental Disorders: Systematic Review.
Jung HW, Kim DY, Lee I et al.
J Med Internet Res · 2025
- PubMed ID
- 41191793
- Record checked
AI-assisted research and writing. This explains one selected publication; it is not a complete review of everything known. Our approach.
One more question, understood.
Keep track of the research you’ve explored.