Wearable- and Mobile App–Based Activity Pacing and Fatigue Management in Post–COVID-19 Condition: Exploratory Observational Study

Background Post–COVID-19 fatigue affects millions worldwide; yet, evidence-based management strategies remain limited. Activity pacing, regulating activity to match available energy and minimize symptom exacerbation, may support symptom management, although optimal pacing approaches remain unclear. Objective This study aimed to explore associations between activity pacing strategies delivered through a mobile app and daily fatigue levels in individuals with post–COVID-19 fatigue. Methods In this exploratory observational study, 19 adults with post–COVID-19 fatigue used wearable devices (Fitbit Inspire 3) for objective activity monitoring and a mobile app (FatigueSense) to self-report daily symptoms (fatigue and energy levels) and optionally select activity pacing goals (light, balanced, or active) over a period of 3 months. We examined associations between pacing strategies and symptom outcomes by using mixed-effects linear models with random intercepts. Same-day outcomes (fatigue and energy reported on the day of goal selection) were analyzed controlling for age and sex. Next-day outcomes (fatigue and energy reported the day following goal selection) were analyzed controlling for age, sex, and prior-day symptoms. Results Across 2182 observation days, participants self-selected pacing goals on 816 (37.4%) days, demonstrating symptom-responsive behavior with higher baseline fatigue on pacing days (mean score 1.74, SD 0.59 vs 1.58, SD 0.54 on nonpacing days on a 0-3 scale where 0=“none” and 3=“severe”; P=.004). On pacing days (584/816, 71.6% with complete data from 18 participants), active pacing was associated with reduced same-day fatigue (β=−0.34, 95% CI −0.49 to −0.19; P<.001) and increased same-day energy (β=8.8, 95% CI 4.7-12.9; P<.001) compared to light pacing. Balanced pacing also showed significant reductions in fatigue (β=−0.15; P=.008) and increases in energy (β=5.8; P<.001) compared to light pacing. Next-day effects were attenuated and nonsignificant (fatigue: β=−0.05, P=.53; energy: β=1.1, P=.61). Individual heterogeneity was substantial, with an intraclass correlation coefficient of 0.32, indicating that 32% of the variance was attributable to between-person differences. Among participants trying multiple strategies, 58.3% (7/12) showed meaningful responses (≥0.3-point fatigue reduction) to structured pacing strategies. Conclusions Structured activity pacing strategies (active and balanced) were associated with improved same-day symptom management in individuals with post–COVID-19 fatigue. However, substantial confounding by indication (self-selection of pacing based on symptom state) and individual heterogeneity limit causal interpretation. These exploratory findings warrant testing in randomized controlled trials to establish efficacy and identify responder characteristics.