Transiently storing information and mentally manipulating it is known as working memory. These operations are implemented by a distributed, fronto-parietal cognitive control network in the brain. The neural mechanisms controlling interactions within this network are yet to be determined. Here, we show that during a working memory task the brain uses an oscillatory mechanism for regulating access to prefrontal cognitive resources, dynamically controlling interactions between prefrontal cortex and remote neocortical areas. Combining EEG with non-invasive brain stimulation we show that fast rhythmical brain activity at posterior sites are nested into prefrontal slow brain waves. Depending on cognitive demand this high frequency activity is nested into different phases of the slow wave enabling dynamic coupling or de-coupling of the fronto-parietal control network adjusted to cognitive effort. This mechanism constitutes a basic principle of coordinating higher cognitive functions in the human brain.

Working memory (WM) involves three cognitive events: information encoding, maintenance and retrieval; these are supported by brain activity in a network of frontal, parietal and temporal regions. Manipulation of WM load and duration of the maintenance period can modulate this activity. Although such modulations have been widely studied using the ERP technique, a precise description of the time course of brain activity during encoding and retrieval is still required. Here, we used this technique and principal component analysis to assess the time course of brain activity during encoding and retrieval in a delayed match to sample task. We also investigated the effects of memory load and duration of the maintenance period on ERP activity. Brain activity was similar during information encoding and retrieval and comprised six temporal factors, which closely matched the latency and scalp distribution of some ERP components: P1, N1, P2, N2, P300 and a slow wave. Changes in memory load modulated task performance and yielded variations in frontal lobe activation. Moreover, the P300 amplitude was smaller in the high than in the low load condition during encoding and retrieval. Conversely, the slow wave amplitude was higher in the high than in the low load condition during encoding, and the same was true for the N2 amplitude during retrieval. Thus, during encoding, memory load appears to modulate the processing resources for context updating and post-categorization processes, and during retrieval it modulates resources for stimulus classification and context updating. Besides, despite the lack of differences in task performance related to duration of the maintenance period, larger N2 amplitude and stronger activation of the left temporal lobe after long than after short maintenance periods were found during information retrieval. Thus, results regarding the duration of maintenance period were complex, and future work is required to test the time-based decay theory predictions.

Different cortical dynamics of alpha oscillations (8–13 Hz) have been associated with increased working memory load, which have been mostly interpreted as a neural correlate of functional inhibition. This study aims at determining whether different manifestations of load-dependent amplitude and phase dynamics in the alpha band can coexist over different cortical regions. To address this question, we increased information load by manipulating the number and spatial configuration of domino spots. Time–frequency analysis of EEG source activity revealed (i) load-independent increases of both alpha power and interregional alpha-phase synchrony within task-irrelevant, posterior cortical regions and (ii) load-dependent decreases of alpha power over areas of the left pFC and bilateral posterior parietal cortex (PPC) preceded in time by load-dependent decreases of alpha-phase synchrony between the left pFC and the left PPC. The former results support the role of alpha oscillations in inhibiting irrelevant sensorimotor processing, whereas the latter likely reflect release of parietal task-relevant areas from top–down inhibition with load increase. This interpretation found further support in a significant latency shift of 15 msec from pFC to the PPC. Together, these results suggest that amplitude and phase alpha dynamics in both local and long-range cortical networks reflect different neural mechanisms of top–down control that might be crucial in mediating the different working memory processes.

Working memory (WM) is a keystone of our cognitive abilities. Increasing load has been shown to dampen its performance and affect oscillatory neural activity in different frequency bands. Nevertheless, mixed results regarding fast frequencies activity and a lack of research on WM load modulations of cross-frequency phase-amplitude coupling mechanisms preclude a better understanding of the impact of increased WM load levels on brain activity as well as inter-regional communication and coordination supporting WM processes. Hence, we analyzed the EEG activity of 25 participants while performing a delayed-matching-to-sample (DMS) WM task with three WM load levels. Current density power and distribution at the source level for theta, beta, and gamma frequencies during the task's delay period were compared for each pair of WM load conditions. Results showed maximal increases of theta activity in frontal areas and of fast frequencies' activity in posterior regions with WM load, showing the involvement of frontal theta activity in WM maintenance and the control of attentional resources and visual processing by beta and gamma activity. To study whether WM load modulates communication between cortical areas, posterior beta and gamma amplitudes distribution across frontal theta phase was also analysed for those areas showing the largest significant WM load modulations. Higher beta activity amplitude at bilateral cuneus and right middle occipital gyrus, and higher gamma activity amplitude at bilateral posterior cingulate were observed during frontal theta phase peak in low than high memory load conditions. Moreover, greater fast beta amplitude at the right postcentral gyrus was observed during theta phase trough at right middle frontal gyrus in high than low memory load conditions. These results show that WM load modulates whether interregional communication occurs during theoretically optimal or non-optimal time windows, depending on the demands of frontal control of posterior areas required to perform the task successfully.

Memory capacity suffers an age-related decline, which is supposed to be due to a generalized slowing of processing speed and to a reduced availability of processing resources. Information encoding in memory has been demonstrated to be very sensitive to age-related changes, especially when carried out through self-initiated strategies or under high cognitive demands. However, most ERP research on age-related changes in working memory (WM) has used tasks that preclude distinction between age-related changes in encoding and retrieval processes. Here, we used ERP recording and a delayed match to sample (DMS) task with two levels of memory load to assess age-related changes in electrical brain activity in young and old adults during successful information encoding in WM. Age-related decline was reflected in lower accuracy rates and longer reaction times in the DMS task. Beside, only old adults presented lower accuracy rates under high than low memory load conditions. However, effects of memory load on brain activity were independent of age and may indicate an increased need of processing after stimulus classification as reflected in larger mean voltages in high than low load conditions between 550 and 1000 ms post-stimulus for young and old adults. Regarding age-related effects on brain activity, results also revealed smaller P2 and P300 amplitudes that may signal the existence of an age dependent reduction in the processing resources available for stimulus evaluation and categorization. Additionally, P2 and N2 latencies were longer in old than in young participants. Furthermore, longer N2 latencies were related to greater accuracy rates on the DMS task, especially in old adults. These results suggest that age-related slowing of processing speed may be specific for target stimulus analysis and evaluation processes. Thus, old adults seem to improve their performance the longer they take to evaluate the stimulus they encode in visual WM.