et al., 2013). In contrast, recent studies using resting-state fMRI and sleep EEG reported decreased path length and increased network efficiency in major depressive disorder (Leistedt et al., 2009; Zhang et al., 2011). These differences in brain network between mood disorders might be due to a variety of factors, including modality (e.g., EEG, MEG, fMRI, and DTI), selection of network matrix (binary or weighted), number of nodes, applied threshold, different populations, as well as the neurophyisological differences between major depression and BD. In the present study, the changes in topological network properties could be attributable to fragmentation or breakdown of the optimal balance between functional integration and segregation. In addition, these alterations were most apparent at frontal, central, and centro-parietal regions, which highly overlapped with the subnetwork from the NBS (Fig. 6).
