[[DP - Sleep]]
[[Depression]]
# Wearables as relapse biomarkers (depression): actigraphy markers predict relapse over 1–2 years https://pubmed.ncbi.nlm.nih.gov/41670991/
A longitudinal JAMA Psychiatry study operationalized relapse using clinical escalation endpoints (MADRS 22 for 2 weeks, hospitalization, suicidal intent/behavior, or antidepressant escalation) and linked relapse risk to sleep/activity regularity features from continuous actigraphy. [PubMed abstract](https://pubmed.ncbi.nlm.nih.gov/41670991/) reports that lower sleep regularity, lower relative amplitude (day–night activity contrast), lower sleep efficiency, and higher wake after sleep onset / nighttime activity were associated with relapse.
Why it matters for HOPES: Reinforces sleep/activity rhythm features as plausible “early warning” signals and supports building dashboard-facing metrics (e.g., relative amplitude, regularity) that can trigger clinician review.
Source: [https://pubmed.ncbi.nlm.nih.gov/41670991/](https://pubmed.ncbi.nlm.nih.gov/41670991/)
---
# What the study found:
A large, long-term study (JAMA Psychiatry) tracked people with depression using wearables (actigraphy = movement/sleep sensors) over 1–2 years. It found that certain sleep and activity patterns could predict whether someone would relapse.
Specifically, people who relapsed tended to show:
- Irregular sleep — inconsistent sleep/wake timing
- Low relative amplitude — their days and nights looked similar (less active by day, more restless at night — blurred contrast)
- Poor sleep efficiency — more time in bed, less actual sleep
- High WASO — waking up a lot during the night
How they defined "relapse": Not just self-report — they used hard clinical endpoints: MADRS score ≥22 for 2 weeks, hospitalization, suicidal behavior, or medication escalation.