JMIR Publications

ABSTRACT

Background:

Persistent post-COVID fatigue symptoms are frequently accompanied by sleep disturbance and altered autonomic regulation. Previous studies have mainly examined how sleep and heart rate variability (HRV) relate to next-day symptoms, while reverse-direction temporal associations are less studied.

Objective:

To examine temporal associations between objectively measured sleep quality, nocturnal heart rate variability (HRV), and daily fatigue and energy levels in adults reporting post-COVID fatigue symptoms.

Objective:

To examine temporal associations between objectively measured sleep quality, nocturnal heart rate variability (HRV), and daily fatigue and energy levels in adults reporting post-COVID fatigue symptoms.

Methods:

: We conducted an exploratory longitudinal observational study using continuous wearable monitoring (Fitbit Inspire 3) in 14 adults reporting post-COVID fatigue symptoms. Participants wore devices 24/7 for a median of 28 days (range: 12-73 days), yielding 678,057 minute-level observations. Sleep metrics (duration, efficiency, sleep stages) and nocturnal HRV parameters (root mean square of successive differences [RMSSD], high-frequency power, low-frequency power) were automatically derived. Participants self-reported fatigue on a 4-level scale (None/Mild/Moderate/Severe; observed responses spanned 3 levels) and energy on a 5-level scale (0,25,50,75,100) throughout each day via smartphone app (Fatiguesense). We analyzed temporal associations using two linking strategies: (1) prospective—predicting morning/afternoon symptoms from the previous night’s sleep and HRV; (2) reverse-direction—predicting subsequent-night sleep and HRV from evening symptoms. Mixed-effects models accounted for within-person repeated measurements. Physical activity (daily steps) was examined as a potential moderator.

Results:

In prospective analyses (previous night → next-day symptoms), the strongest correlations were observed between increased high-frequency HRV and lower fatigue (ρ=-0.359, p<0.001) and increased energy (ρ=+0.498, p<0.001), exceeding sleep duration (ρ=-0.145, p=0.033). REM sleep duration negatively correlated with fatigue (ρ=-0.271, p<0.001). Higher low-frequency/high-frequency ratio (an exploratory autonomic index) was associated with worse fatigue (ρ=+0.204, p=0.002) and lower energy (ρ=-0.422, p<0.001). In ordinal mixed-effects models, we tested both parsimonious (sleep duration + HRV-RMSSD + low-frequency/high-frequency ratio) and comprehensive (all sleep metrics + HRV metrics) specifications. HRV-RMSSD remained the only significant predictor in comprehensive models (fatigue: p=0.026, OR=0.49; energy: p=0.016, OR=1.73), demonstrating effects independent of sleep duration, efficiency, and architecture. Sleep duration effects observed in parsimonious models were mediated through sleep quality. In retrospective analyses (evening symptoms → subsequent night’s sleep/HRV), fatigue showed no associations with subsequent sleep or HRV. However, energy exhibited significant retrospective associations: higher evening energy predicted higher nocturnal HRV-RMSSD (ρ=+0.267, p=0.024) and high-frequency HRV (ρ=+0.365, p=0.002), confirmed in mixed-effects modeling (β=0.920, p=0.018). Physical activity did not independently predict symptoms but showed a positive interaction with sleep duration (p=0.014).

Conclusions:

In this exploratory pilot sample of adults reporting post-COVID fatigue symptoms, nocturnal HRV and sleep measures showed temporally specific associations with daily fatigue and energy ratings. Prospective associations were generally stronger than reverse-direction associations, which were mainly observed for evening energy and subsequent nocturnal HRV. These findings are hypothesis-generating and should be interpreted cautiously, given the small sample, lack of a control group, and consumer wearable measurement constraints. Clinical Trial: N/A