Revisiting alpha-theta cross-frequency dynamics during working memory

Abstract

Prior EEG research has shown that during working memory, alpha (8 to 14 Hz) and theta (4 to 8 Hz) oscillations tend to form a 2:1 frequency ratio. According to the Binary Hierarchy Brain Body Oscillation Theory (BHBBOT), a recent model grounded in mathematical analysis, this cross-frequency configuration reflects enhanced connectivity between brain regions generating these rhythms. However, this prediction has not yet been empirically tested. In this study, we leveraged high density EEG, source localization and connectivity metrics derived from Information Theory (IT) and the Theory of Weakly Coupled Oscillators (TWCO) to examine whether the previously observed alpha-theta cross-frequency dynamics during working memory are accompanied by changes in connectivity. Our results show that a significant increase in the proportion of 2:1 ratios between regions generating frontal theta and parietal alpha rhythms was accompanied by relative decreases in connectivity, as revealed by both IT and TWCO metrics. Furthermore, phase synchrony between these two regions was significantly reduced during working memory and correlated negatively with behavioral performance. In conclusion, our results show that the increased occurrence of 2:1 alpha:theta cross-frequency ratios during working memory reflects functional segregation (rather than integration) and therefore directly challenges some of the predictions of the BHBBOT.