Structural equation modeling to explore putative causal factors for chronic fatigue in childhood cancer survivors: a DCCSS LATER study

Purpose: To explore putative causal relations for chronic fatigue (CF) in childhood cancer survivors (CCS) using structural equation modeling (SEM). Methods: Interrelationships between factors that were previously associated with CF and their causal relation with CF were investigated using SEM and causal discovery methodology. A bootstrap method was used to ascertain how robust each finding was, presenting the percentage of times that each discovered edge was found in 1000 bootstrap samples as a measure of confidence (with > 50% needed to be confident in a found edge). Results: 1927 CCS (51.7% male) with a mean age of 35 years (SD 9) participated in the study (23.6% reported CF). Results indicated that sex had a putative causal effect on CF (bootstrap confidence 81%), while CF was causally linked to helplessness, physical activity, pain, BMI, and sleep problems (bootstrap confidence 82%, 78%, 78%, 51%, and 51%, respectively). The relation between CF and depression was found to be two-way (bootstrap confidence 68%), indicating a reciprocal relation or the presence of a latent confounder. The same applied to the relations between CF and anxiety, self-esteem, and social functioning, but this could not be confirmed with high confidence (bootstrap confidence < 50%). Conclusion: This study provides insight into the complex etiology of CF and could give guidance in the development of appropriate prevention and/or intervention strategies for CF in CCS. Implications for Cancer Survivors: Results show the consequences of having CF and could help to understand the impact it has on daily life.