Dynamic functional connectivity analysis reveals improved association between brain networks and eating behaviors compared to static analysis

Uncontrollable eating behavior is highly associated with dysfunction in neurocognitive systems. We aimed to quantitatively link brain networks and eating behaviors based on dynamic functional connectivity analysis, which reflects temporal dynamics of brain networks. We used 62 resting-state functional magnetic resonance imaging data sets representing 31 healthy weight (HW) and 31 non-HW participants based on body mass index (BMI). Brain networks were defined using a data-driven group-independent component analysis and a dynamic connectivity analysis with a sliding window technique was applied. The network centrality parameters of the dynamic brain networks were extracted from each brain network and they were correlated to eating behavior and BMI scores. The network parameters of the executive control network showed a strong correlation with eating behavior and BMI scores only when a dynamic (p < 0.05), not static (p > 0.05), connectivity analysis was adopted. We demonstrated that dynamic connectivity analysis was more effective at linking brain networks and eating behaviors than static approach. We also confirmed that the executive control network was highly associated with eating behaviors.