High-definition transcranial direct current stimulation of the parietal cortices modulates the neural dynamics underlying verbal working memory

Abstract

Verbal working memory (vWM) is an essential limited-capacity cognitive system that spans the fronto-parietal network and utilizes the subprocesses of encoding, maintenance, and retrieval. With the recent widespread use of noninvasive brain stimulation techniques, multiple recent studies have examined whether such stimulation may enhance cognitive abilities such as vWM, but the findings to date remain unclear in terms of both behavior and critical brain regions. In the current study, we applied high-definition direct current stimulation to the left and right parietal cortices of 39 healthy adults in three separate sessions (left anodal, right anodal, and sham). Following stimulation, participants completed a vWM task during high-density magnetoencephalography (MEG). Significant neural responses at the sensor-level were imaged using a beamformer and whole-brain ANOVAs were used to identify the specific neuromodulatory effects of the stimulation conditions on neural responses serving distinct phases of vWM. We found that right stimulation had a faciliatory effect relative to left stimulation and sham on theta oscillations during encoding in the right inferior frontal, while the opposite pattern was observed for left supramarginal regions. Stimulation also had a faciliatory effect on theta in occipital regions and alpha in temporal regions regardless of the laterality of stimulation. In summary, our data suggest that parietal HD-tDCS both facilitates and interferes with neural responses underlying both the encoding and maintenance phases of vWM. Future studies are warranted to determine whether specific tDCS parameters can be tuned to accentuate the facilitation responses and attenuate the interfering aspects.

Keywords: alpha; encoding; magnetoencephalography; maintenance; oscillations; theta.

FIGURE 1

Visit timeline, current flow model, and MEG task paradigm. Using HD‐tDCS, participants received 20 min of anodal and sham stimulation over the left and right parietal lobes. Stimulation montages were pseudorandomized across the three visits, each separated by at least 1 week. Current distribution modeling using our HD‐tDCS montage revealed focal stimulation of the left and right parietal lobes (left). Following HD‐tDCS, participants completed a modified Sternberg verbal working memory task during MEG recording (right). The total visit time from the beginning of stimulation to the end of the MEG task was approximately 75 min.

FIGURE 2

(Top) Grand‐averaged time‐frequency spectrogram from a sensor in parieto‐occipital areas (i.e., MEG 2313) with time (ms) shown on the x‐axis and frequency (Hz) denoted on the y‐axis. All signal power data are expressed as percent difference from the baseline, with the percent bar shown on the top of the spectrogram. (Bottom): Significant time‐frequency responses are plotted topographically to show the mean spatial distribution across the time‐frequency window.

FIGURE 3

Grand‐averaged beamformer images (pseudo‐t) across all participants, and HD‐tDCS montages for each time‐frequency component. Separate color scale bars are shown on the sides for each.

FIGURE 4

Repeated measure whole‐brain ANOVAs on the theta encoding data identified several brain regions where the strength of theta oscillatory activity differed by the parietal HD‐tDCS montage, including the (a) right inferior frontal (b) right occipitotemporal cortices, (c) left cuneus, (d) left supramarginal regions, and right cerebellum (not shown). Each violin plot shows individual data points with the middle bars representing the means of each condition. Dotted circles in the images to the right of each plot indicate the relevant region. *p < .05, **p < .005.

FIGURE 5

Repeated measure whole‐brain ANOVAs on the (a) alpha encoding and (b, c) maintenance responses indicated several brain regions where alpha oscillatory strength differed by the parietal HD‐tDCS montage. These regions included the (a) right inferior temporal during encoding and the (b) left medial frontal and (c) right inferior temporal regions during maintenance. Each violin plot shows individual data points with the middle bars representing the means of each condition. Dotted circles in the images to the right of each plot indicate the relevant region. *p < .05, **p < .005.