correlation matrix) (rij, N×N) can be obtained for each subject. Each functional connectivity matrix can be converted to a binary, undirected network G using a cost threshold (t, 0<t<1), which is equivalent to the ratio between the number of edges and all possible edges [35].In this study, we first applied a range of cost threshold (0.05≤t≤0.5, step = 0.01) to investigate the network properties. Such a thresholding approach can normalize all networks to have the same number of edges or wiring cost and thus provide an avenue to detect age-related changes in the topological organization [36], [37]. Finally, we adopted the following complementary approaches to select the small-world regime as a range of cost threshold (0.2≤t≤0.35): (1) the average of the number of connections over all nodes is larger than the log of the number of nodes (N = 90) ensuring that the small-world properties are estimable [38] and (2) the resulting brain networks are sparse but fully connected and have distinguishable properties in comparison with the degree-matched random and regular networks, respectively [39], [40], [41]. To confirm our results, we also repeated all analyses using weighted, undirected network (see Text S1) and found similar results in both global network
