Neurobehavioral Association Between Macroscopic Network Decoupling and Reduced Eating Impulsivity: A Longitudinal fMRI Study in Bulimia Nervosa

Abstract

ABSTRACT Background Fluoxetine is an established pharmacological intervention for bulimia nervosa (BN); however, the macroscopic functional network remodeling underlying its efficacy in suppressing eating impulsivity remains poorly understood. Methods Fifty‐eight patients with BN and 58 healthy controls assessed at both baseline and follow‐up underwent resting‐state functional magnetic resonance imaging (rs‐fMRI). Network‐Based Statistic (NBS) analysis with edgewise paired comparisons was used to identify whole‐brain connectivity changes in patients with BN. The resulting BN‐derived component was subsequently examined in both groups using a Group × Time analysis. Exploratory brain‐behavior associations within BN were examined using Spearman and partial Spearman correlations. Results During follow‐up, patients with BN showed significant improvements in depressive symptoms and externally driven eating behavior. Within‐group NBS analysis identified a 23‐edge component showing reduced functional connectivity in BN at follow‐up, with network‐level family‐wise error correction (p < 0.05). This component primarily involved dorsal attention, ventral attention/salience, and somatomotor networks. Whole‐brain NBS analysis in healthy controls did not identify a significant longitudinal component. When mean connectivity of the BN‐derived component was extracted in both groups, connectivity decreased more strongly in BN than in healthy controls (Group × Time β = −0.053, p = 0.027). Conclusions The follow‐up period in patients with BN was accompanied by clinical improvement and reduced functional connectivity among attention‐, salience‐, and sensorimotor‐related systems. These findings provide preliminary evidence of treatment‐period‐associated functional network reconfiguration in BN. Trial Registration ChiCTR2200066885