Lampe et al. (2025)
- Authors: Viktoria Lampe, Tim Riester, Lena-Marie Theil, Priyanka Dey, Philipp Wieder, Frank Müller, Dominik Schröder, Tim Schmachtenberg, Eva Hummers, Eva Maria Noack, Sigrun May, Frank Klawonn, Georg Martin Behrens, Rebecca Rubin, Alexandra Dopfer-Jablonka, Sandra Steffens, Christine Happle
- Institutes: Hannover Medical School, Hannover, Germany, University Medical Center Göttingen, Göttingen, Germany, Ostfalia University of Applied Sciences, Wolfenbüttel, Germany, Helmholtz Centre for Infection Research, Braunschweig, Germany
- Publisher: Research Square (Preprint)
- Link: DOI
Summary
This protocol introduces the ‘U-WaTCH’ study, which aims to transform how patients manage ‘crashes’ by moving from subjective tracking to objective, AI-driven prediction. By comparing Post-COVID Syndrome with rheumatic conditions, the research seeks to identify universal physiological signatures of Post-Exertional Malaise (PEM). If successful, this could lead to wearable tools that warn patients of an impending crash before it occurs, significantly improving the safety of daily activity and pacing.
What was researched?
This research outlines the protocol for a prospective study designed to develop machine learning algorithms that can predict Post-Exertional Malaise (PEM) ‘crashes’ using continuous data from wearable devices.
Why was it researched?
While PEM is a debilitating hallmark of Post-COVID Syndrome and various inflammatory conditions, patients currently lack objective tools to predict or prevent symptom exacerbations. Researchers aim to bridge this gap by identifying physiological biomarkers in wearable sensor data that precede a clinical crash.
How was it researched?
The ‘U-WaTCH’ study will recruit 300 participants divided into three groups: patients with Post-COVID Syndrome and PEM, patients with rheumatic conditions and PEM, and healthy controls. Participants will wear an Apple Watch SE for up to 180 days to collect continuous data on heart rate, activity, and sleep, while simultaneously logging symptoms via a mobile app.
What has been found?
As a study protocol, the paper details the intended methodology rather than final clinical results. It establishes a framework for using non-parametric machine learning to analyze high-dimensional, longitudinal sensor data to detect individual-specific patterns of physiological strain.
Discussion
The study’s strengths include its long-term (180-day) observation period and the inclusion of a non-COVID inflammatory control group to ensure findings are specific to PEM rather than a single disease. A potential limitation is the reliance on participants’ self-reported ‘crash’ events to train the predictive algorithms.
Conclusion & Future Work
The U-WaTCH study represents a critical step toward personalized, real-time PEM monitoring. Future results are expected to validate whether wearable technology can provide actionable warnings to help patients stay within their ‘energy envelope’ and avoid relapses.