Passive Detection of Pre-PEM Physiological Destabilization in ME/CFS and Long COVID: A prospective wearable study protocol using nocturnal HRV

Post-exertional malaise (PEM) is a defining and disabling feature of myalgic encephalomyelitis/chronic fatigue syndrome (ME/CFS) and a substantial subset of Long COVID. Despite its clinical importance, PEM remains difficult to predict and lacks objective early markers. This study protocol introduces a focused and minimalistic approach to test a single hypothesis: that PEM onset is preceded by a measurable physiological destabilization detectable in nocturnal heart rate variability (HRV). The study employs a fully passive, naturalistic design using continuous wearable monitoring (Oura Ring), without behavioral instructions, activity tracking, or intervention. Participants prospectively record only PEM onset, severity, and duration, enabling event-aligned analysis of discrete state transitions. By deliberately minimizing measurement burden and conceptual complexity, the protocol isolates the temporal relationship between physiological state and symptom onset. The primary objective is to determine whether nocturnal HRV declines systematically in the days preceding PEM onset within individuals. Secondary analyses explore whether a broader multivariate physiological signature precedes PEM. This protocol is designed as a proof-of-concept study in a carefully selected cohort with clearly identifiable and recurrent PEM episodes, allowing high-resolution temporal analysis under real-world conditions. The study is conceptually grounded in the Persistent Systemic Threat-Signaling State (PSTS) framework, a systems-level model of post-acute infection syndromes (PAIS) in which persistent activation and reduced recovery capacity define a state of impaired stability. Within this framework, post-exertional malaise (PEM) is hypothesized to represent a state transition preceded by progressive physiological destabilization. In this context, nocturnal heart rate variability (HRV) is evaluated as a candidate leading indicator of pre-onset system instability rather than a direct mechanistic driver. The underlying conceptual model is described in a separate preprint. If confirmed, the findings would provide empirical support for the existence of a detectable pre-PEM window and establish a foundation for future work on prediction, monitoring, and intervention.