Daily-life dynamics of general and cognitive fatigue in childhood cancer survivors: An experience sampling and wearable study

Background Fatigue is a common late effect in childhood cancer survivors (CCS), yet its moment-to-moment fluctuations and time-varying covariates in daily life remain poorly understood. This study examined within-person dynamics of general and cognitive fatigue, focusing on prior fatigue, emotional tension, sleep duration, physical activity, and cardiorespiratory fitness. Methods Seventy-eight CCS (median age = 18.6 years) participated in five 7-day experience sampling method (ESM) bursts over one year, completing seven daily smartphone assessments on fatigue and emotional states. Fitbit devices provided continuous data on nightly sleep and step counts in the two hours preceding each assessment. Multilevel mixed-effects models examined lagged within-person associations and stable between-person differences in fatigue, including moderation by cardiorespiratory fitness (VO 2 peak). Results Participants completed 12345 assessments in total. Fatigue was predominantly explained by within-person variability (∼70%), with smaller between-person differences (∼30%). Prior fatigue strongly predicted subsequent fatigue, indicating temporal persistence (general: b=0.44,p<.001; cognitive: b=0.19,p<.001). Higher tension predicted higher subsequent fatigue (general: b=0.08,p<.001; cognitive: b=0.07,p<.001). Sleeping longer than one’s usual duration was associated with lower next-day fatigue (general: b=–0.10,p<.001; cognitive: b=–0.05,p<.001). Recent physical activity showed domain-specific associations, relating to higher general fatigue (b=0.07,p=.001) but lower cognitive fatigue (b=–0.02,p=.042). General fatigue increased nonlinearly across the day, accelerating later in the afternoon (b=0.22,p=.028). VO 2 peak was not significantly associated with fatigue, and did not significantly moderate the relationship between recent physical activity and fatigue. Conclusion Fatigue in CCS is a dynamic, within-person process shaped by prior fatigue, emotional tension, sleep duration, recent activity, and time of day. Intensive longitudinal monitoring offers valuable insight into real-time fatigue dynamics and may support personalized management strategies, including pacing, optimized scheduling, and tailored movement guidance.