Objective: The relative strength of risk factors for cancer‐related fatigue (CRF) among breast cancer survivors (BCSs) remains unclear. This study aims to systematically evaluate and compare the strength of different risk factors for CRF using a network meta‐analysis (NMA) approach. Methods: This NMA included observational studies on female BCSs with extractable data related to risk factors for CRF. The PubMed, Cochrane Library, and Embase databases were systematically searched, and the study protocol was registered in PROSPERO (reference no. CRD42025642021). A random‐effects meta‐analysis was performed to estimate pooled effect sizes, and an NMA with P‐scores was used to rank the relative strength of risk factors. Subgroup analyses, sensitivity analyses, and meta‐regression were conducted to assess methodological quality and explore potential sources of heterogeneity. Results: Thirty observational studies ( n = 36,302 female BCSs) that were published between 2004 and 2024 were included in this NMA. Depression exhibited the strongest association with CRF (OR = 3.34, 95% CI 2.50–4.46, P‐score = 0.9727), followed by insomnia (OR = 2.35, 95% CI 1.45–3.81, P‐score = 0.6549), pain (OR = 1.94, 95% CI 1.33–2.84, P‐score = 0.4587), and anxiety (OR = 1.85, 95% CI 1.23–2.79, P‐score = 0.4132). Subgroup analysis revealed that the associations of the four risk factors with CRF remained significant at the three posttreatment time points ( 5 years), with the exception of anxiety and insomnia at 5 years. Meta‐regression demonstrated that higher study quality (measured via the Newcastle–Ottawa scale NOS) was significantly correlated with stronger associations of anxiety and insomnia with CRF ( β = 0.305 and 0.221, p < 0.05, respectively). Sensitivity analysis confirmed the robustness of the main findings. Conclusion: Depression plays a central role in CRF development and should be prioritized in survivorship care. Integrating multimodal interventions for depression, sleep disturbances, and pain management may improve fatigue outcomes.
Huang et al. (Wed,) studied this question.