...

Double-blind, randomized, placebo-controlled pilot clinical trial with gamma-band transcranial alternating current stimulation for the treatment of schizophrenia refractory auditory hallucinations

Date:


Study participants

This study was a double-blind, randomized, placebo-controlled pilot clinical trial conducted at the Tianjin Anding Hospital from October 2020 to May 2022 and was approved by the Ethics Committee of the Tianjin Anding Hospital. Fifty schizophrenic participants with refractory auditory hallucinations were recruited from the inpatient department of Tianjin Anding Hospital. Of these, 32 participants who completed all study sessions were included in the analysis (see Fig. S1). The study protocol was approved by the Ethics Committee of Tianjin Anding Hospital [Approval Number: 2020(2020–30)]. All methods were performed in accordance with the relevant guidelines and regulations. All participants gave written informed consent after being informed about the study.

Inclusion and exclusion criteria

All participants were diagnosed with schizophrenia or schizoaffective disorder (confirmed by the Structured Clinical Interview for Diagnostic and Statistical Manual of Mental Disorders 5th edition [41] (DSM-V)) with symptoms during this treatment present for greater than 12 weeks and who also met the following criteria: 18–70 years old (male and female are not limited); junior high school or above; treated with ≥ 2 different antipsychotic medications (chlorpromazine equivalent dose (The drug equivalent dosage instructions see Table S1) > 600 mg), last medication measure lasting ≥ 6 weeks with poor outcome (CGI ≥ 4 points or PANSS reduction rate < 20% since this treatment) [PANSS reduction rate = pre/post difference/(baseline score – 30)]; Consistent PANSS hallucinatory behavior (P3) item score of ≥ 4 or PSYRATS score change of ≤ 20% in the last two weeks prior to patient enrollment as assessed by the supervising physician, exhibiting stable and significant hallucinatory symptoms; no transcranial magnetic stimulation, transcranial direct current stimulation, or electroconvulsive therapy in the last four weeks.

Participants were excluded from the study for the following criteria: have a mental illness other than the study illness (according to DSM-V); severe physical or neurological illness; any brain device or implant, including cochlear implants and aneurysm clips; previous autoimmune disease or family history of immune disease; antibiotics and immunosuppressive drugs in the last four weeks.

Demographics and baseline scale statistical analysis

For demographic analysis, chi-square tests were conducted for categorical variables, including gender, family history, marital status, clozapine use, and the use of antipsychotic drugs (APD). Data normality was confirmed with the Shapiro-Wilk test (P > 0.05). Independent samples t-tests were employed to assess continuous variables such as years of education, duration of illness, and equivalent drug dosage. Regarding baseline scores on the PANSS and PSYRATS scales for both groups, we also performed independent samples t-tests.

Study design and blinding procedures

Participants who completed the study attended a total of four weeks of the treatment period and two weeks of the non-treatment period (Fig. 1A). In the treatment period, participants in the active tACS group received 20 min of 40Hz-tACS per day for five days (Monday to Friday) per week for four consecutive weeks in hospital settings, whereas participants in the sham tACS group received 20 sessions of sham stimulation for the same duration (see “Electrode Montage and Stimulation Protocol” below for details). Participants in the active tACS group received 40 Hz stimulation (3 s ramp up, 3 s ramp down) for 20 min. Participants in the sham tACS group only received 3 s ramp-up and a 3 s ramp-down to simulate the sensation of active tACS on the skin for the purpose of blinding. Clinical assessments were conducted at baseline, week2, week 4 and the follow-up (week6) of the treatment period. Resting-state EEG and task-related EEG recordings were conducted at baseline (before the stimulation), week2, week4 (after the stimulation) and follow-up (week6). (labeled in Fig. 1A). Participants were randomly assigned following simple randomization procedures (computerized random numbers) to one of treatment groups.

Fig. 1: Transcranial alternating current stimulation (tACS) protocol.
figure 1

A Experimental design consisted of two stages: the treatment period (left) and the non-treatment period (right). During the treatment period, participants were randomly assigned to the tACS group (Bottom right, red frame) or the sham group (Bottom right, blue frame). Ramp-in and ramp-out is 3 s for all conditions, with 0 s of active stimulation for sham stimulation, 20 min of active stimulation for 40Hz-tACS. Participants received 40Hz-tACS or sham stimulation five times a week (Monday through Friday) for four consecutive weeks. Clinical assessments and EEG recordings were conducted on baseline (before the stimulation), week2 and week4 (after the stimulation). After the end of treatment, Clinical assessments and EEG recordings were conducted on week6. Our analyses focused on the clinical assessment and task-related EEG of the four key times. B Gamma-band (40 Hz) auditory steady-state response paradigm. C Transcranial alternating current stimulation (tACS) configuration for all participants. The first figure (from left to right) is shown that two stimulators were used; one connected to the electrode over AF3 (red pad), one connected to the electrode over CP5 (red pad), and both connected to the electrode over Cz (blue pad). tACS at AF3 and CP5 has an amplitude of 1 mA (zero-to-peak), while the return current at Cz has an amplitude of 2 mA (zero-to-peak). The second and third figures show the stimulated voltage distribution over the brain from two different views (left and top) when currents at AF3 and CP5 reach their positive peak.

The randomization of all participants, group assignment, and preparation of the stimulation equipment (40Hz-tACS or sham) were conducted by a single researcher. Once the stimulation equipment was set up, the experimenter performed electrical stimulation on the patient. The stimulation interface was oriented only towards the experimenter, ensuring that the patient was unaware of the specific type of stimulation being performed. In addition, the rater who assessed the patient’s symptoms was unaware of the group assignment.

Electrode montage and stimulation protocol

Transcranial alternating stimulation was delivered via Starstim 32 (Neuroelectrics, Spain). Three carbon-silicone electrodes were applied to the scalp with a saline-soaked round sponge pad for all participants. One electrode was placed at AF3 (in the International 10–20 system) in the area of the dorsolateral prefrontal cortex (dlPFC); another electrode of the same dimensions was placed at CP5 in the area of the left TPJ. These two electrodes are shown in red in Fig. 1C. A third return electrode, the same as two additional electrodes, was placed over Cz (the blue pad in Fig. 1C). For participants who received tACS, 40 Hz alternating currents were applied to the dlPFC and TPJ electrodes in phase with each other (red waves in Fig. 1C), with a zero-to-peak amplitude of 1 mA; 40 Hz alternating currents at the return electrode were antiphase to those of the dlPFC and TPJ electrodes (blue waves in Fig. 1C). The placement of electrodes was based on previous work using tACS for the treatment of schizophrenia [33, 42]. The simulated voltage distribution and the electric field strength shown in Fig. 1C were simulated using ROAST3.0 [43] on the MNI 152 Head. During the stimulation, participants sat comfortably in a quiet room and were asked to keep their eyes open, with the aim of maintaining a constant brain state. At the same time, the operator in the room avoided interaction with the patient as much as possible. After each stimulation, the participant was asked about the sensation of the stimulation. All participants enrolled in the experiment reported no discomfort throughout the stimulation process.

EEG tasks and acquisition

Stimulus procedure

Auditory steady-state stimuli were conducted binaurally through noise-cancelling headphones (WH-1000XM3, SONY) at a sound pressure level (SPL) of 90 dB. After fitting the headphones, the operator asked participants for readiness. Upon confirmation, the operator initiated the program, and the sound playback commenced. The response to auditory steady-state stimuli (click-trains) was recorded (Fig. 1B). Before the formal stimulation, a 0.5 s beep signaled the start of the experiment. Considering environmental factors and the specificity of the participants, during the click stimulation, participants were required to look at a picture of the headphones on a computer monitor while listening to a 3 s long click train at a rate of 40 Hz, with two click trains separated by a 1.5 s interval, for a total of 56 click trains. Participants underwent EEG acquisition while receiving the auditory stimuli in order to record the response to the auditory click-trains.

EEG acquisition

EEGs were continuously digitized at a rate of 1000 Hz using a 64-channel SynAmps2 system (Neuroscan, USA), and the interelectrode impedance was kept below 10 kΩ. The electrode montage was based on standard positions in the International 10–20 electrode system [44]. The system acquisition bandpass was 0.1–200 Hz. Data recording was referenced to a linked left mastoid electrode (“M1”) with the ground electrode located at “AFz”.

Clinical outcome measures and statistics

The primary outcome measure was the change in the auditory hallucination symptom scores measured by Psychotic Symptom Rating Scales (PSYRATS) from baseline to week2, week4, and follow-up. In addition, the outcome measures also measured by the positive and negative syndrome scale (PANSS) were assessed at baseline, week 2, week4, and follow-up. For statistical analysis, custom-built scripts in R (R Foundation for Statistical Computing, Vienna, Austria) were used and were available by request. Data normality was confirmed with the Shapiro-Wilk test (P > 0.05) (see Table S3). To assess equality of variance, we used Levene’s test (P > 0.05). We adopted a two-way repeated-measure ANOVA (rm-ANOVA), with the intergroup factor being “condition” (40Hz-tACS and sham), and the intragroup factor being “time” (baseline, week2, week4, follow-up). This study mainly focused only on scale scores, for which there was a significant interaction. Further analysis was performed using the Students’ test to compare the severity of symptoms between the two groups; a one-way ANOVA was used to compare the change in symptoms between the two groups over the treatment and non-treatment periods, and a paired Student’s test (Bonferroni corrected) was chosen for the post-hoc test. For scores where there was a significant interaction, we further calculated the symptom score reduction rates for the two groups with the formula:

$${\rm{Reduction\; rates}}=\frac{{\rm{baseline\; score}}-{\rm{week}}* {\rm{score}}}{{\rm{baseline\; score}}}\times 100 \%$$

Two-samples t-tests were used to compare the rate of score reduction between the two groups at different time points (baseline, week2, week4, follow-up). To control for baseline differences in AH symptom scores, baseline scores were included as a covariate in the analysis of reduction rates at week2, week4, and follow-up. An analysis of covariance (ANCOVA) was performed to compare the rate of symptom reduction between the two groups across these time points.

EEG data analysis and statistic

Preprocessing

EEG data analyses were performed in Matlab using the EEGLAB toolbox and a combination of custom MATLAB scripts (MathWorks, Natick, MA). First, electrodes that were not used, such as CB1, CB2, HEO, VEO, EKG, and EMG, were removed from the experiment. All data were down-sampled to 250 Hz. Offline, EEG data recorded immediately after stimulation were imported into EEGLAB 2022.1 running under MATLAB 2020a and re-referenced to the average of the left and right mastoids. The function pop_eegfiltnew.m was applied for a 0.5–120 Hz bandpass filter (finite impulse response filter, cutoff frequency (−6 dB): [0.25 Hz 120.25 Hz], zero-phase, non-causal). Notch filtered at 50 and 100 Hz was utilized to remove industrial frequency interference. Artifact subspace reconstruction in the plugin clean_rawdata() of EEGLAB was used to automatically reject high-amplitude artifacts [45]. The parameters applied were: flat line removal, 10 s; electrode correlation, 0.7; ASR, 100 (this value was chosen to consider the actual data quality); window rejection, 0.5. The rejected channels are spatially interpolated by spherical. Further, independent component analysis (ICA) was used to remove eye blinks, eye movement, muscular artifacts, and heartbeats. To select brain ICs among all types of ICs, the EEGLAB plugin ICLabel() was used [46].

Continuous EEG data were segmented into epochs that started at –1 s and ended at 4 s relative to sound stimulus onset. Bad epochs were rejected by thresholding the magnitude (±100 μV) of each epoch. We found no significant difference in the epochs and components between all conditions (two-way rm-ANOVA, α = 0.05, see Table S2). Moreover, we set eight regions of interest (i.e., ROIs) according to the anatomical location (Fig. 3A) [47], namely the left frontal region (green area: FP1, AF3, F7, F5, F3, F1, FC5, FC3, FC1, C1), right frontal region (green area: FP2, AF4, F8, F6, F4, F2, FC6, FC4, FC2, C2), left temporal region (yellow area: FT7, T7, TP7), right temporal region (yellow area: FT8, T8, TP8), left parietal region (blue area: C5, C3, CP5, CP3, CP1, P7, P5, P3, P1), right parietal region (blue area: C6, C4, CP6, CP4, CP2, P8, P6, P4, P2), left occipital region (purple area: PO7, PO5, PO3, O1), and right occipital region (purple area:PO8, PO6, PO4, O2).

Functional connectivity

Functional connectivity (phase synchrony) between two-channel locations was measured by the phase locking value (PLV), the absolute value of the mean phase difference between the two signals expressed as a complex unit-length vector [48, 49]. When analyzing bio-signals, especially electrical brain activities, PLV represented a crucial factor to consider in terms of synchronization. It gauged the frequency-specific synchronization, which was not directional and denotes long-range integrations. It evaluated how the phase difference between two signals varies over time [48]. We computed PLV as:

$${\rm{PLV}}=\frac{1}{N}{{|}}\mathop{\sum }\limits_{n=1}^{N}\exp (j\theta (t,n)){{|}}$$

where θ (t, n) was the phase difference \({\emptyset}_{1}({\rm{t}},{\rm{n}})-{\emptyset}_{2}({\rm{t}},{\rm{n}})\). PLV measured the intertrial variability of this phase difference at t: If the phase difference varies little across the trials, PLV was close to 1; otherwise, it was close to 0 [50]. We selected 3 s post-stimulation EEG recordings (30–50 Hz bandpass filtering) from each epoch of 40 Hz click stimulation for further analysis. The mean connectivity between every electrode in a specific region and all the electrodes in the other region was computed for every ROI. The connectivity between both regions was determined by averaging all the electrode pairs between them.

Functional brain network controllability

The functional brain network controllability analysis was based on brain connectivity networks. Brain network controllability reflected the possibility of driving a current network state to other desired target states with external energy input. Average controllability and modal controllability were the most frequently employed metrics to measure network controllability [39]. Average controllability was a measure of the ability of a node to drive the brain to all possible, easily reachable states considering the average input energy cost [51]. Average controllability could be measured by the trace of the Gramian matrix, which was inversely proportional to the control energy required to drive shifts in brain states [52]. Brain regions with high average controllability could switch the brain to many easily achievable states with less input energy. Modal controllability quantified the ease with which a single control node could drive the brain into difficult-to-reach states [53]. Areas with high modal controllability were not hubs of the network but instead had a low degree of control, which imposed high energy costs for completing complex goal-specific operations [54]. For a more detailed derivation of the formula and the code, please refer to the study by Gu et al. [39]. For times with significant differences, the Pearson correlation between average controllability and modal controllability was explored for all participants.

Statistic

The statistical analysis of EEG data was performed using custom R scripts. The neural data met the assumptions of the statistical tests used. Data normality was confirmed with the Shapiro-Wilk test (P > 0.05). To assess equality of variance, we used Levene’s test (P > 0.05). Similar to clinical scale score statistics, two-way rm-ANOVA (α = 0.05) was selected to statistically analyze the PLV, average controllability and modal controllability within the eight ROIs obtained, and the results focused only on the regions with significant interactions. Further analysis was performed using the Students’ test to compare the values of PLV, average controllability and modal controllability between the two groups. A one-way ANOVA was used to compare the change in EEG features between the two groups over time of treatment and non-treatment period. Post-hoc comparisons within two-way rm-ANOVAs were Bonferroni-corrected.

LEAVE A REPLY

Please enter your comment!
Please enter your name here

Subscribe

spot_imgspot_img

Popular

More like this
Related

‘Good Cop/Bad Cop’ Star Luke Cook Balances Fitness With Fun

Luke Cook may be known for playing disturbing...

The 9-Week Workout Plan for Moms To Spring Back into Shape

Spring is here, and that means more carpool...

Article One – Shaping the Course of Spring

This spring, Article One redefines the...

Canadian government offers student loan forgiveness to dentists, hygienists in rural areas

Dentists and dental hygienists who move to underserved...