Reduction in Left Frontal Alpha Oscillations by Transcranial Alternating Current Stimulation in Major Depressive Disorder Is Context Dependent in a Randomized Clinical Trial

Abstract

Background: Left frontal alpha oscillations are associated with decreased approach motivation and have been proposed as a target for noninvasive brain stimulation for the treatment of depression and anhedonia. Indeed, transcranial alternating current stimulation (tACS) at the alpha frequency reduced left frontal alpha power and was associated with a higher response rate than placebo stimulation in patients with major depressive disorder (MDD) in a recent double-blind, placebo-controlled clinical trial.

Methods: In this current study, we aimed to replicate successful target engagement by delineating the effects of a single session of bifrontal tACS at the individualized alpha frequency (IAF-tACS) on alpha oscillations in patients with MDD. Eighty-four participants were randomized to receive verum or sham IAF-tACS. Electrical brain activity was recorded during rest and while viewing emotionally salient images before and after stimulation to investigate whether the modulation of alpha oscillation by tACS exhibited specificity with regard to valence.

Results: In agreement with the previous study of tACS in MDD, we found that a single session of bifrontal IAF-tACS reduced left frontal alpha power during the resting state when compared with placebo. Furthermore, the reduction of left frontal alpha oscillation by tACS was specific for stimuli with positive valence. In contrast, these effects on left frontal alpha power were not found in healthy control participants.

Conclusions: Together, these results support an important role of tACS in reducing left frontal alpha oscillations as a future treatment for MDD.

Keywords: Alpha oscillations; Depression; Emotional images; Functional connectivity; Individual differences; Transcranial alternating current stimulation.

Figure 1.. IAF-tACS decreased left frontal IAF power in patients with MDD.

(A) Individual alpha frequency (IAF) was localized using an eyes-closed resting-state EEG. The power spectrum from occipital-parietal electrodes (outline of region of interest shown in insert) for each participant is depicted. The frequency with maximum power within the alpha band from 8–13 Hz (labeled grey box) is denoted with a black circle for all patients with MDD. (B) After localizing IAF for each participant, transcranial alternating current stimulation (tACS) was applied to the scalp at IAF. A hypothetical participant is depicted. Dotted line shows the IAF with maximum power in the grey rectangle of the canonical alpha band. IAF-tACS delivered identical current at 1 mA zero-to-peak with a split-wire to two 5×5cm stimulator pads on the left and right frontal cortices (over F3 and F4). The 5×7cm return stimulator pad over Cz acted as an electrical sink. (C) Electric field model shows the regions with maximum electric field strength (V/m) and depicts the electric force lines. On the left, an axial view at the +37 z-plane of Montreal Neurological Institute (MNI) space oriented anterior (A) to posterior (P). On the right, a coronal view at the +28 y-plane of MNI space oriented left (L) to right (R). (D) IAF-tACS produced a selective decrease in IAF power for verum versus sham (modulation index) over the left frontal region of interest (white circle, * p < 0.05). Electrodes with a significant change are depicted with a black dot, p < 0.05, that were in a cluster of at least three contiguous significant electrodes.

Figure 2.. IAF-tACS negated the increase in left frontal IAF power during the resting-state.

(A) In the absence of stimulation, a selective increase in IAF power over the left frontal region of interest (white circle, * p < 0.05) was observed. (B) Verum IAF-tACS did not produce a significant change in left frontal IAF power. Electrodes with a significant difference from post- to pre-stimulation are depicted with a black dot, p < 0.05, that were in a cluster of at least three contiguous significant electrodes.

Figure 3.. Individual difference in tACS effect by depression severity.

Individual differences in the impact of stimulation on IAF power across the scalp was correlated with baseline depression severity in patients with MDD quantified by self-report using the BDI-II (A) and clinician-report using the HAM-D (B). The difference in correlation between the verum and sham group found a focal difference in left frontal electrodes. Left frontal region of interest highlighted with a white circle. * p < 0.05. Dots represent electrodes with a significant difference and at least three contiguous significant electrodes.

Figure 4.. Analysis of the effect of IAF-tACS on left frontal IAF power by medication status.

Patients with MDD were categorized by use of an antidepressant within the past two weeks. A two-way ANOVA using between antidepressant-use and stimulation (verum or sham) found a main effect of stimulation (F(1,37)=4.791, p=0.035, η_p²=0.11), a trend-level effect of antidepressant-use (F(1,37)=3.277, p=0.078, η_p²=0.08), and no interaction (F(1,37)=0.801, p=0.377, η_p²=0.02). Post-hoc testing revealed a significant effect of stimulation with antidepressant-use (N=17, t(15)=−2.848, p=0.012, d=1.386) driven by a decrease in left frontal IAF power for verum (N=8, t(7)=2.683, p=0.031, d=0.949). Without antidepressants there was no significant effect of stimulation (N=24, t(22)=−1.049, p=0.305, d=0.428). Critically, there was no difference in depression severity (HAM-D) between patients groups by antidepressant-use (t(39)=0.124, p=0.902, d=0.0398). *p<0.05; error bars are SEM; units are modulation index.

Figure 5.. Elevated left frontal alpha power to images rated as positive is inhibited by IAF-tACS in patients with MDD.

(A) Patients with MDD passively viewed positive, neutral, and negative images from the IAPS (example images provided). Stimuli were presented for two seconds (s) separated by a two to three second inter-trial interval (ITI). (B) Across all valences at baseline in patients with MDD, time-frequency analysis revealed a significant decrease in alpha amplitude and increase in theta amplitude in left frontal electrodes (insert in the upper right). Black outline shows significant cluster at p < 0.05 after permutation-based cluster-correction by mass. Vertical line at zero denotes onset of the image. The dashed grey rectangle from 0.1 to 1.5 seconds and 8 to 12 Hz is the alpha power that was extracted. (C) At baseline, the contrast of positive versus neutral alpha power revealed a selective increase in left frontal electrodes and decrease in parietal-occipital electrodes. The white outline is the left frontal region of interest. * p < 0.05 for a priori analysis. Black dots represent electrodes with a significant difference that were in a cluster of at least three contiguous electrodes. (D) At baseline, the contrast of negative versus neutral alpha power revealed no modulation in the left frontal region of interest (white circle). “n.s.” is not significant. (E) For the contrast of verum versus sham, there was a selective reduction of left frontal alpha power for positive images. The tACS effect is the change in alpha amplitude for image viewing data post minus pre stimulation. (F) For the left frontal region of interest, the change in alpha power from stimulation is shown for all valence and stimulation conditions. The comparison of interest from the baseline analysis revealed a significant difference for verum versus sham. * p < 0.05. Error bars are within-participant SEM. Alpha power was z-transformed across scalp electrodes so units are z.

Figure 6.. Healthy controls without left frontal alpha engagement show no effect of stimulation.

(A) At baseline, there was no increase in left frontal alpha power for positive relative to neutral images. White circle depicts left frontal region of interest. N.s. means not significant. (B) Left frontal alpha power does not show modulation from IAF-tACS in healthy control participants. Error bars are within-participant SEM. (A,B) Units are z-transformed alpha power values across scalp channels. (C) Control participants do not exhibit an impact of IAF-tACS for verum versus sham in left frontal IAF power. (D) Unlike the patients with MDD, healthy controls do not show an increase in left frontal IAF power during resting-state after sham stimulation, as in patients with MDD. Black dots depict p < 0.05 with at least 3 contiguous electrodes. (C,D) Units are modulation index.

Figure 7.. Individual difference in baseline functional connectivity positively tracked with depression severity.

(A) Individual differences analysis revealed that depression severity (HAM-D) was positively correlated with functional connectivity strength (weighted phase lag index, wPLI) between the left and right frontal cortices in patients with MDD. The artifact zone surrounding the seed in F3 is depicted with a black box with rounded corners. A dot here denotes an electrode with a correlation of p < 0.05 and with at least 3 contiguous significant electrodes. A white outline was drawn around F4 and the surrounding electrodes. (B) Individual differences analysis for IAF power did not reveal any significant relationship with depression severity in patients with MDD. Functional connectivity and IAF power values were divided by the sum of scalp channels.