Abstract
PURPOSE: This study aimed to identify functional brain connectivity patterns associated with cognitive performance in Beta-thalassemia major (β-TM) children and to determine whether hematological factors influence cognition indirectly through alterations in connectivity. METHOD: We recruited 25 children with β-TM and 35 age-matched healthy controls. Cognitive performance was assessed using the Wechsler Intelligence Scale (WIS). Resting-state functional MRI data were processed to construct whole-brain functional connectivity matrices. We applied network-based statistics (NBS) to compare connectivity differences between groups and connectome-based predictive modeling (CPM) with cross-validation to predict cognitive scores. Mediation analyses were further conducted to test whether hematological metrics (hemoglobin level, red blood cell distribution width) impacted cognition through functional connectivity. FINDING: Compared to controls, β-TM children showed significantly reduced WIS scores and widespread disruptions in functional connectivity, particularly in cerebellar, motor, and temporal networks. The CPM approach identified a predictive network that largely overlapped with the NBS-derived network and robustly predicted WIS scores. Mediation analysis revealed that hemoglobin and red blood cell distribution width influenced cognitive scores indirectly through altered connectivity, indicating a full mediation effect. CONCLUSION: This study provides evidence that hematological abnormalities in β-TM children impair cognitive performance via their impact on functional brain networks. Functional connectivity signatures derived from CPM may serve as early neuromarkers of cognitive vulnerability and could inform future monitoring and intervention strategies in this population.