Theta and alpha oscillations during working-memory maintenance predict successful long-term memory encoding

Abstract

To date, much is known about the neural mechanisms underlying working-memory (WM) maintenance and long-term-memory (LTM) encoding. However, these topics have typically been examined in isolation, and little is known about how these processes might interact. Here, we investigated whether EEG oscillations arising specifically during the delay of a delayed matching-to-sample task reflect successful LTM encoding. Given previous findings of increased alpha and theta power with increasing WM load, together with the assumption that successful memory encoding involves processes that are similar to those that are invoked by increasing WM load, alpha and theta power should be higher for subsequently remembered stimuli. Consistent with this assumption, we found stronger alpha power for subsequently remembered stimuli over occipital-to-parietal scalp sites. Furthermore, stronger theta power was found for subsequently remembered stimuli over parietal-to-central electrodes. These results support the idea that alpha and theta oscillations modulate successful LTM encoding.

Copyright 2009 Elsevier Ireland Ltd. All rights reserved.

Figures

Figure 1

Trial timing of the delayed matching-to-sample-task, in which either objects or letter strings had to be maintained throughout the delay period. The length of the delay was randomly varied between 5 and 7 seconds with an average of 6s. In addition, topographic maps of overall alpha and theta power during the WM-maintenance delay (2 – 7 s poststimulus) are shown separately for objects (upper maps) and letter strings (lower maps). The maps reveal comparable topographies for objects and letter strings across material types with a parietal-occipital maximum for alpha and a midfrontal maximum for theta. Therefore, the data were collapsed across material type in the subsequent analyses (see Methods for details).

Figure 2

Topographic maps of alpha (A) and theta (B) power for remembered and forgotten stimuli, as well as the “remembered-forgotten” difference maps (Dm effects). The individual alpha power maxima and the maximum effect were found over the parietal-occipital cortex. As usually found for theta, the individual theta power maxima were located over the midfrontal cortex between electrodes Cz and Fz. The maximum difference between later remembered vs. forgotten stimuli, however, was located more posterior over central-parietal electrodes (see Results for details).