Aberrant large- and mesoscale network segregation and integration in bulimia nervosa

Growing evidence indicates that bulimia nervosa (BN) is more likely driven by network-level dysfunctions rather than abnormalities in isolated brain regions. However, previous studies focusing on specific regions or connectivity features have led to a limited understanding of the complexity and integrality of network in BN. This study aimed to investigate large- and mesoscale network alterations in BN from the perspective of network segregation and integration. Using resting-state functional magnetic resonance imaging data from 85 BN patients and 71 healthy controls (HCs) matched for age, sex, and education, we applied a graph-theoretic framework to analyze functional network architecture in BN. Specifically, we examined group differences in within- and between-system functional connectivity (FC) as well as system segregation at both the whole-brain and system levels. Associations between network measures and clinical features were further explored. Compared with HCs, BN patients exhibited whole-brain reduced within- and between-system FC accompanied by increased system segregation. At the system level, patients showed increased between-system FC of the default mode network (DMN), control network and ventral attention network, along with decreased DMN system segregation. The increased integration of DMN with other systems was associated with higher Beck Depression Inventory (BDI-21) scores in BN patients. The neural mechanisms of BN involve an imbalance between network segregation and integration at both global and system levels, most pronounced within the DMN, whose alterations are associated with depression and disordered eating behavior. These findings may provide potential neural substrates for some core behavioral deficits in BN. Bulimia nervosa (BN) is a serious eating disorder, yet most neuroimaging studies have focused on isolated regions or specific networks, which limits our understanding of how large-scale brain network systems are organized and interact in BN. Although some evidence suggests abnormalities in network organization, it remains unclear how within-network connectivity, between-network connectivity, and overall network system level organization are jointly altered, and how these changes relate to clinical symptoms. Given that cognition and behavior depend on coordinated activity across distributed brain network systems, we examined BN using a whole-brain network approach. In this study, we used resting-state fMRI to scan 85 BN patients and 71 healthy controls matched for age, sex, and education. We analyzed the brain as a set of interacting networks and measured how strongly regions communicated within the same network and between different networks, as well as how “segregated” or “integrated” each network was. Overall, patients with BN showed a disrupted balance in brain network organization. Across the whole brain, there was generally weaker connectivity both within and between networks, alongside a tendency toward increased segregation between networks for BN. At the same time, one key network system, the default mode network (DMN), which is involved in self-referential thinking and emotional processing, displayed relatively increased integration with other brain networks. Importantly, stronger DMN integration was linked to more severe depressive symptoms and eating-disorder related psychopathology in individuals with BN. Our findings suggest that BN is associated with an imbalance in how brain networks are organized and communicate, particularly involving the DMN. This imbalance may contribute to difficulties in regulating self-focused thoughts, shifting attention, and controlling impulsive behavior. Exploring BN from the perspective of large-scale brain networkorganization may help improve our understanding of its complex pathophysiological mechanisms.