The aim of the present study was to investigate the association between resting-state EEG functional connectivity and spatial working memory (SWM) in adolescents, using a graph-theoretical approach. Sixty-three healthy adolescents (38 females; M = 16.8, SD = 0.47) participated in the study. Resting-state EEG was recorded by a wireless dry-sensor 32-channel headset (Cognionics Quick-32r) configured in a standard 10-20 montage. Graph metrics were calculated in alpha, theta, beta-low and beta-high frequency bands, using connection-strength thresholds of 50 and 80%. SWM performance was measured with the computerized Corsi Block-Tapping Task (CBT). Robust bootstrapped linear regressions revealed significant associations between graph metrics and CBT accuracy only for theta and beta-low frequency bands. Specifically, characteristic path length and modularity correlated negatively with CBT accuracy, while participation coefficient correlated positively, indicating that participants with more integrated resting-state networks performed better on the SWM task. A k-means clustering analysis divided the participants into two groups characterized by either integrated (low CPL and modularity, high participation coefficient) or segregated (high CPL and modularity, low participation coefficient) network organization. Participants with integrated theta-band networks demonstrated significantly higher CBT accuracy than those with segregated networks, whereas no such effect was observed in the beta band. Overall, theta-band networks were more integrated than beta-band networks. These findings highlight the importance of global integration of resting-state functional brain networks, particularly within the theta frequency range, for spatial working memory performance in adolescence.
Pavlova et al. (Fri,) studied this question.