Abnormal large-scale brain functional network dynamics in social anxiety disorder

AimsAlthough static abnormalities of functional brain networks have been observed in patients with social anxiety disorder (SAD), the brain connectome dynamics at the macroscale network level remain obscure. We therefore used a multivariate data-driven method to search for dynamic functional network connectivity (dFNC) alterations in SAD.MethodsWe conducted spatial independent component analysis, and used a sliding-window approach with a k-means clustering algorithm, to characterize the recurring states of brain resting-state networks; then state transition metrics and FNC strength in the different states were compared between SAD patients and healthy controls (HC), and the relationship to SAD clinical characteristics was explored.ResultsFour distinct recurring states were identified. Compared with HC, SAD patients demonstrated higher fractional windows and mean dwelling time in the highest-frequency State 3, representing "widely weaker" FNC, but lower in States 2 and 4, representing "locally stronger" and "widely stronger" FNC, respectively. In State 1, representing "widely moderate" FNC, SAD patients showed decreased FNC mainly between the default mode network and the attention and perceptual networks. Some aberrant dFNC signatures correlated with illness duration.ConclusionThese aberrant patterns of brain functional synchronization dynamics among large-scale resting-state networks may provide new insights into the neuro-functional underpinnings of SAD. Four distinct recurring states were identified, including (in order of decreasing frequency) state with "widely weak" FNC, state with "widely moderate" FNC, state with "locally strong" FNC (mainly involving perceptual networks), and state with "widely strong" FNC. SAD patients significantly spent more time in a "weakly connected" state and less time in "strongly connected" states and demonstrated decreased FNC mainly within/between the default mode network, the attention network, and perceptual networks.image