Efficacy of Transcranial Alternating Current Stimulation (tACS) for Treating Hallucinations: A Systematic Review and Meta-analysis

Abstract

Purpose of the Review:

Transcranial alternating current stimulation (tACS) is a non-invasive technique that modulates brain oscillations and may help treat hallucinations when conventional treatments fail. This systematic review evaluated the efficacy of tACS in reducing hallucinations in psychiatric disorders.

Collection and Analysis of Data:

A systematic search of PubMed, Scopus, EMBASE, and clinical trial registries was conducted. Studies included adult psychiatric populations receiving tACS with hallucinations measured using standardized scales. Three randomized controlled trials (RCTs) were included in the meta-analysis, and five additional studies were synthesized qualitatively. Risk of bias was assessed using the Cochrane RoB 2 tool for RCTs, and the Joanna Briggs Institute tools for other designs, and certainty of evidence was assessed using GRADE.

Meta-analysis of three RCTs (72 participants) showed no significant difference between tACS and sham treatment in reducing hallucinations post-intervention (SMD = –0.14, 95% CI = –0.61 to 0.32) or at study endpoint (SMD = 0.10, 95% CI = –0.41 to 0.61). Qualitative findings suggested potential benefits with gamma-frequency (40 Hz) tACS for auditory hallucinations in schizophrenia. Side effects were mild and infrequent. Overall risk of bias was moderate to high, and evidence certainty was low.

Conclusions:

Gamma-frequency tACS shows potential, but evidence remains inconclusive. Larger, well-designed trials are needed to determine its effectiveness for hallucinations in psychiatric disorders. Persistent auditory hallucinations pose treatment challenges in schizophrenia. Non-invasive brain stimulation such as TMS and tDCS has been found to offer some benefit for auditory hallucinations. However, the literature on the efficacy of tACS for auditory hallucinations has not been reviewed. This systematic review has synthesized all published studies and reported the pooled effect for efficacy and safety.

Key Messages:

• Question: What is the evidence for the efficacy of tACS in reducing hallucinations across psychiatric disorders?

• Findings: There is no significant difference between tACS and sham treatment in reducing hallucinations. Gamma-frequency (40 Hz) tACS may be effective for the treatment of hallucinations in schizophrenia.

• Meaning: There is preliminary but inconclusive evidence supporting the efficacy of tACS for auditory hallucinations in schizophrenia.

Transcranial alternating current stimulation (tACS) is a non-invasive brain-stimulation technique that delivers alternating electrical currents through scalp electrodes, modulating brain oscillations. ¹ Because these oscillations occur across different brain regions, stimulating one area can produce widespread effects. Parameters such as electrode placement, frequency, duration, amplitude, phase, and timing can be adjusted. tACS has been applied to neurocognitive disorders (e.g., Alzheimer’s, Parkinson’s, Huntington’s), psychotic disorders (schizophrenia, schizoaffective, psychotic bipolar disorder), depressive disorders, and other conditions like substance use, neurodevelopmental, and sleep–wake disorders.^2–4

Hallucinations—perceptions without external stimuli—are a core feature of psychosis, both primary and secondary, and occur across sensory modalities. Treatments include medications, psychotherapy, neurofeedback, and virtual reality, either alone or in combination.^5–8 However, the effect sizes are small to moderate in long-term follow-ups.^9,10 Additionally, brain-stimulation techniques such as repetitive transcranial magnetic stimulation (rTMS), tDCS, and tACS have been studied for managing hallucinations in psychotic disorders.^11–13

tACS may improve cognition and reduce positive and negative symptoms in schizophrenia and other disorders.^2,14 A scoping review found promising effects on hallucinations and delusions, though results are mixed, especially in clozapine-resistant cases.^13,15,16 Neuromodulation is generally a third-line approach for treatment-resistant hallucinations, affecting 25%–40% of schizophrenia patients.^17,18 To address this gap, a systematic review and meta-analysis were conducted to evaluate the efficacy of tACS specifically for hallucinations.

Methods

The review was conducted following the Cochrane Handbook for Systematic Reviews of Interventions and reported in accordance with the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines (Supplementary Table S3).^19,20 The protocol was prepared and registered with PROSPERO (Ref: CRD420250656267, registered on February 23, 2025). No ethics committee approval was required for systematic review and meta-analysis.

Eligibility Criteria

Studies were included if they met the following criteria: (a) any study design, including case reports and case series, that used tACS as intervention; (b) adult participants with neuropsychiatric conditions in whom hallucination change was an outcome; and (c) for the meta-analysis, randomized controlled trials (RCTs) that evaluated hallucination outcomes using standardized rating scales such as the Auditory Hallucination Rating Scale (AHRS), ²¹ Positive and Negative Syndrome Scale (PANSS), ²² or Psychotic Symptom Rating Scales (PSYRATS). ²³ Exclusion criteria were: studies not involving human participants; studies that did not assess hallucinations using validated measures; studies involving children or adolescents; and studies lacking sufficient detail on the tACS protocol or stimulation parameters. Research that did not report changes in hallucination outcomes following tACS, did not define the study population using established diagnostic criteria, or did not clearly specify the indication for tACS was also excluded.

Search Strategy and Selection of Studies

The literature was searched across several databases, including MEDLINE (PubMed), Scopus, EMBASE, and clinical trial registries. All types of articles published up to February 25, 2025, were included, provided they met the following criteria based on population (psychiatric conditions associated with hallucinations in any modality, such as psychotic disorders, bipolar disorder, major depressive disorder, post-traumatic stress disorder [PTSD], and dementia); intervention (tACS); comparison (sham, treatment as usual, or any other modality of non-invasive brain stimulation); and outcome (reduction in hallucination frequency and intensity measured by standardized rating scales). The search string used in PubMed retrieved 14 articles (without any filters), which was also applied to EMBASE (retrieving 21 articles using the “query translator” feature) and Scopus (retrieving 27 articles) (see Supplementary Table S1).

Study Screening and Selection

The articles retrieved from the databases were screened using Rayyan software. ²⁴ Two authors (DD and VV) independently removed duplicates manually. Titles and abstracts were screened independently by two authors (DS and GP), with conflicts resolved by another author (RKD). Full-text screening was conducted independently by two authors (DD and VV), with conflicts resolved by another author (RKD) (list of excluded studies with reasons in Supplementary Table S2). Data were then extracted from the included articles and cross-checked for errors. The PRISMA flowchart (Figure 1) outlines the study selection process.

Figure 1. PRISMA Flowchart.

A total of 62 records were identified from PubMed (14), EMBASE (21), and Scopus (27), after removal of 42 duplicates. Of the remaining 20 records screened, 8 were excluded, and 12 were assessed for retrieval. All 12 were retrieved, and 8 full-text articles were assessed for eligibility. Four were excluded due to duplication, irrelevance to hallucinations or tACS, or unrelated outcomes. Among the eight included studies, three were RCTs^25–27 and included in the meta-analysis. These studies were conducted in the USA (n = 2)^25,26 and China (n = 1) ²⁷ between 2018 and 2025, with a total of 72 participants across both intervention and control groups. The remaining five studies were included solely for qualitative synthesis.^16,28–31

Outcome Measures

The main outcome was the reduction in hallucination scores (across any type) as reported by standardized rating scales, among participants with psychiatric conditions such as psychotic disorders, bipolar disorder, major depressive disorder, PTSD, and dementia, who were administered either tACS or a sham treatment.

Data Extraction

The primary outcome measures were extracted at both the post-intervention and study endpoint from all three RCTs,^25–27 for meta-analysis, and from the remaining five studies^16,28–31 for qualitative synthesis. Hallucination scores were reported as mean, standard deviations (SD), and the number of participants (n) in both the tACS and sham groups. Additional data extracted from the studies included: first author’s name, year of publication, country of study, study design, details regarding randomization, blinding, and allocation, participant descriptions, inclusion and exclusion criteria, the number of participants in each group (intervention vs. sham), details related to tACS (dose, duration, frequency, number of sessions, and stimulation days), and outcomes (reduction in standardized rating scale scores for hallucinations across all reported sessions).

Assessment of Bias and Certainty of Evidence

Risk of bias was assessed independently by two authors (DS and GP), with conflicts resolved by another author (RKD). For the RCTs, the Cochrane Risk of Bias Assessment Tool version 2 (RoB2) ³² was used (Supplementary Figure S4). A traffic light plot was generated using the ROBVIS online tool (www.mcguinlu.shinyapps.io/robvis/). For the other included studies, the Joanna Briggs Critical Appraisal Tools (https://jbi.global/critical-appraisal-tools) were used (Supplementary Tables S5, S6, and S7). The certainty of the evidence was assessed using the GRADE approach, with GRADEpro software (https://www.gradepro.org/).

Statistical Analysis

The meta-analysis of the included RCTs was conducted using Review Manager (RevMan) version 5.4.1 (https://revman.cochrane.org/info). Data were pooled using a random-effects model, and the analysis was conducted using the inverse-variance method. Although two different types of tACS were used in the studies, the effect sizes were pooled because both shared a common mechanism of action. The I² statistic was used to assess heterogeneity. The main outcome, the reduction in hallucination scores at both post-intervention and study endpoints, was reported using the standardized mean difference (SMD) and 95% confidence intervals (CI). Due to the small number of included studies, publication bias was assessed using a Doi plot with the Luis Furuya-Kanamori (LFK) index, ³³ plotted using effect size (Cohen’s d), which was calculated via an online calculator (https://lbecker.uccs.edu/#means%20and%20standard%20deviations), and standard error (SE) with the formula: Sqrt (SD1 ² /n1) + (SD2 ² /n2). The analysis was performed in R using the “metasens,” “meta,” and “ggplot2” packages. ³⁴

Baseline Characteristics of the Included Studies for Meta-analysis (Table 1)

Table 1.

Characteristics of the Studies Included in Meta-analysis.

Author (Year), Country, Study Design	Participants (Age, Sex), Groups	tACS Protocol	Outcomes	Comments
Zhang et al. (2022) 27 USA Double-blind, randomized controlled pilot trial, crossover design	21.8 years, M/F = 16/25. Four groups (tACS + tACS, tACS + sham, Sham + tACS, Sham + sham) tACS (n = 14) Sham (n = 11).	Active: 10 Hz AC applied to the DLPFC and TPJ electrodes in phase with each other with a zero-to-peak amplitude of 1 mA. Sham: 20 s tACS flanked by a 10 s ramp-up and a 10 s ramp-down to simulate the sensation of tACS on the skin for the purpose of blinding. Five consecutive days of 10 Hz tACS for 40 min each day.	Primary outcome: Change in AHRS Group 1: tACS + tACS group (n = 8) Baseline: 25.0 ± 7.4 5 days: 25.1 ± 8.0 2 months: 23.9 ± 9.6 Group 2: tACS + Sham group (n = 6) Baseline: 19.0 ± 6.8 5 days: 19.8 ± 5.1 2 months: 15.7 ± 5.0 Group 3: Sham + tACS group (n = 5) Baseline: 23.4 ± 6.5 5 days: 25.6 ± 6.8 2 months: 26.4 ± 6.8 Group 4: Sham + Sham group (n = 6) Baseline: 23.2 ± 7.0 5 days: 21.0 ± 5.9 2 months: 17.6 ± 7.1 AE: Not mentioned.	The mean difference was taken for post-intervention (after 5 days) and post follow-up (after 2 months) only for Groups 1 and Groups 4 for comparison. There were no significant changes observed.
Wang et al. (2025) 26 China Double-blind, randomized, placebo-controlled pilot clinical trial	46.72 years, M/F = 14/18. Two groups (tACS and sham) tACS (n = 25) Sham (n = 25).	Active: 20 min of 40 Hz AC applied to DLPFC and TPJ electrodes in phase with each other, with a zero-to-peak amplitude of 1 mA; 40 Hz AC at the return electrode were antiphase to those of the DLPFC and TPJ electrodes. Sham: Received a 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. Daily sessions for 5 days (Monday to Friday) per week for 4 consecutive weeks in hospital settings.	Primary outcome: Change in AHRS component of PSYRATS tACS group: Baseline: 33.94 ± 4.49 2 weeks: 28.88 ± 5.95 4 weeks: 25.38 ± 7.68 6 weeks: 24.94 ± 7.47 Sham group: Baseline: 31.44 ± 6.51 2 weeks: 28.06 ± 6.89 4 weeks: 27.44 ± 6.45 6 weeks: 26.81 ± 7.26 AE: Not mentioned	The tACS group had a significant reduction of baseline AHRS scores at post-intervention and also at the last follow-up compared to the sham group.
Mellin et al. (2018) 25 USA Double-blind, randomized, sham-controlled pilot clinical trial	38.86 years M/F = 15/7 3 (tACS, tDCS, Sham) tACS (n = 8) tDCS (n = 7) Sham (n = 7)	tACS: 10 Hz tACS stimulation waveform was a sinewave with a peak-to-peak amplitude of 2 mA. tDCS: Stimulation paradigm was +2 mA at the frontal site (F3/Fp1) and −2 mA at the temporo-parietal site (T3/P3). Sham: 10 seconds of ramp-in to 20 seconds of 10 Hz tACS, with a ramp-out of 10 seconds for a total of 40 seconds of stimulation. Twice-daily 20-minute stimulation sessions, separated by 3 hours, over 5 consecutive days.	Primary outcome: Change in AHRS scores from baseline. Group 1 tACS: Baseline: 25.88 ± 3.8 5 days: 22.13 ± 5.82 1 week: 21.88 ± 6.22 1 month: 22.25 ± 6.09 Group 2 tDCS: Baseline: 23.00 ± 6.9 5 days: 21.86 ± 9.58 1 week: 21.43 ± 6.55 1 month: 23.29 ± 7.63 Group 3 Sham: Baseline: 24.71 ± 5.91 5 days: 22.43 ± 5.56 1 week: 22.29 ± 4.19 1 month: 20.86 ± 4.85 AE: Headache, neck pain, scalp pain, tingling, itching, ringing/buzzing noise, burning sensation, local redness, sleepiness, trouble concentrating, improved mood, worsening of mood, dizziness, flickering lights.	The tACS group had a significant reduction at post-intervention (Day 5) compared to the sham and tDCS groups. At the last follow-up, the tACS group had a sustained reduction in AHRS compared to the tDCS group.

tACS: Transcranial alternating current stimulation; tDCS: Transcranial direct current stimulation; dlPFC: dorsolateral prefrontal cortex; TPJ: Temporo-parietal junction; AHRS: Auditory Hallucination Rating Scale; PSYRATS: Psychotic Symptom Rating Scale.

In three RCTs, the mean ages of the populations were 21.8, ²⁶ 38.8, ²⁵ and 46.7 ²⁷ years. Stimulation parameters varied between 10 Hz and 40 Hz, administered either as single or double sessions per day, lasting from 20 to 40 minutes, over a period ranging from 5 consecutive days to up to 4 weeks. All studies included sham as a control, with one study, ²⁶ also having a tDCS arm for comparison in a crossover design. The diagnosis of all subjects was schizophrenia with auditory hallucinations resistant to pharmacological treatment.

Each study provided baseline (pre-intervention) data, but the follow-up durations differed:

Mellin et al. ²⁵ reported follow-ups at 5 days, 1 week, and 1 month; Wang et al. ²⁷ at 2 weeks and 1 month; and Zhang et al. ²⁶ at 5 days and 2 months. As the authors did not respond to requests for additional data, the meta-analysis focused on post-intervention and last follow-up scores. Additionally, the study by Zhang et al. ²⁶ employed a crossover design with two arms for the first 5 days (tACS vs. Sham) and four arms for the following 2 months (tACS + tACS vs. tACS + Sham vs. Sham + tACS vs. Sham + Sham). To account for potential carryover effects, data from the first 5 days were used for the post-intervention analysis, and for the last follow-up analysis, data from the two arms (tACS + tACS and Sham + Sham) were included.

Intervention Protocols

The tACS protocols varied across the included studies. Mellin et al. ²⁵ applied 10 Hz tACS with a peak-to-peak amplitude of 2 mA, delivered in 20-minute sessions twice daily (separated by 3 hours) over 5 consecutive days. Zhang et al. ²⁶ used 10 Hz tACS at a zero-to-peak amplitude of 1 mA, targeting the dorsolateral prefrontal cortex (dlPFC) and temporo-parietal junction (TPJ), administered for 40 minutes daily over 5 consecutive days. In contrast, Wang et al. ²⁷ administered 40 Hz tACS with a zero-to-peak amplitude of 1 mA for 20 minutes daily, 5 days per week, over a period of 4 consecutive weeks. All three studies incorporated appropriate sham conditions, typically involving brief periods of stimulation (ranging from 3 to 20 seconds) with ramp-up and ramp-down phases to replicate the sensation of active stimulation and ensure effective blinding of participants.

Primary Outcome: Reduction in Hallucinations

The outcome measures varied across the studies. Mellin et al. ²⁵ and Zhang et al. ²⁶ used the AHRS, a subscale of the PSYRATS, while Wang et al. ²⁷ used the full PSYRATS scale to assess hallucinations. In the meta-analysis of post-intervention outcomes, the SMD was –0.14 (95% CI = –0.61, 0.32; I² = 0%; n = 72; 3 RCTs; p = .55; Figure 2), indicating no significant advantage of tACS over sham in reducing hallucinations. Similarly, at the study endpoint, the SMD was 0.10 (95% CI = –0.41, 0.61; I² = 0%; n = 61; p = .70; Figure 3), again showing no superiority of tACS over sham. Across both analyses, heterogeneity among the studies was minimal.

Figure 2. Reduction in Hallucination at Post-intervention Between tACS and Sham Treatment.

Figure 3. Reduction in Hallucination at the Study Endpoint Between tACS and Sham Treatment.

Individually, Mellin et al. ²⁵ reported a reduction in AHRS scores from baseline (mean ± SD: 25.88 ± 3.80) to post-treatment (22.13 ± 5.82), with effects maintained at 1-week (21.88 ± 6.22) and 1-month (22.25 ± 6.09) follow-ups. Zhang et al. ²⁶ reported mixed findings across their four groups: the tACS + tACS group showed minimal change from baseline (25.0 ± 7.40) to 5 days (25.1 ± 8.00), with modest improvement at 2 months (23.9 ± 9.60), while the tACS + sham group worsened slightly at 5 days but improved by 2 months (from 19.0 ± 6.80 to 15.7 ± 5.00). Wang et al. ²⁷ demonstrated a progressive reduction in PSYRATS scores with tACS, from baseline (33.94 ± 4.49) to 2 weeks (28.88 ± 5.95), 4 weeks (25.38 ± 7.68), and 6 weeks (24.94 ± 7.47), with the sham group also improving, though to a lesser extent.

Assessment of Publication Bias, Risk of Bias, and Certainty of Evidence

Publication bias for post-intervention outcomes (Supplementary Figure 8) was assessed using a Doi plot, which showed a relatively symmetrical distribution with an LFK index of 0.27, indicating minimal publication bias. However, the Doi plot for study endpoint outcomes (Supplementary Figure 9) revealed significant asymmetry, with an LFK index of 4.26, suggesting substantial publication bias. This suggests caution when interpreting the long-term effects.

Risk of bias, assessed using the Cochrane ROB2 tool, revealed some concerns in domain 5 (bias in the selection of the reported results) across all three studies. In domain 1 (bias arising from the randomization process), Zhang et al. ²⁶ raised some concerns, while Mellin et al. ²⁵ showed a high risk of bias. The overall risk of bias was rated as high for Mellin et al. ²⁵ and as having some concerns for Zhang et al. ²⁶ and Wang et al. ²⁷ (Figure S2). The sensitivity analysis, which removed the study with a high risk of bias, did not change the pooled estimate. According to the GRADEpro assessment (Supplementary Table S10), the certainty of the evidence for the reduction in hallucinations at post-intervention and at the study endpoint was rated as low.

Adverse Effects

Only Mellin et al. ²⁵ reported adverse effects associated with tACS, including headache, neck pain, scalp pain, tingling, itching, ringing or buzzing noise, burning sensation, local redness, sleepiness, trouble concentrating, mood changes, dizziness, and flickering lights. These side effects were assessed using a Likert scale ranging from 1 (absent) to 4 (severe). The mean scores for each side effect were all below 2, indicating that the reported side effects were generally mild. Notably, the effect of tACS on sleepiness was the least pronounced compared to tDCS and sham, with statistical significance (p = .03). The other two studies, Zhang et al. ²⁶ and Wang et al., ²⁷ did not report any adverse effects.

Baseline Characteristics of the Included Studies for Qualitative Synthesis Only

A total of five studies investigating tACS for hallucinations were included in the qualitative synthesis (Table 2). These studies comprised three case reports,^16,30,31 one retrospective cohort study, ²⁸ and one non-RCT with a crossover design. ²⁹ The publications were from four countries: USA (n = 2),^29,30 India (n = 2),^16,31 and China (n = 1). ²⁸ Sample sizes varied, with three single-case reports,^16,30,31 one study with 6 participants, ²⁹ and one larger study with 17 participants. ²⁸ The majority of participants had a diagnosis of schizophrenia with persistent auditory hallucinations (AH) that were resistant to conventional treatments.

Table 2.

Characteristics of Studies Included in Qualitative Synthesis.

Author (Year), Country, Study Design	Participants (Age, Sex), Diagnosis, Groups	tACS Protocol	Outcomes	Comments
Force et al. (2021) 30 USA Case report	1 (35-year-old female) Schizophrenia with persistent AH	Once weekly for 20 weeks, tACS, for 40 mins at 2 mA and 10 Hz. Location: Left dlPFC (between F3 and Fp1) and the temporal parietal junction (between T3 and P3). Reference electrode over the central region (Cz). Zero phase difference.	AHRS, HPSVQ	Improvement with respect to duration and controllability of auditory hallucinations. Side effects: Transient scalp pain (60% of sessions), tingling (60%), and burning sensation (55%) at the electrode site.
Pathak et al. (2024) 16 India Case report	1 (42-year-old female) Clozapine-resistant schizophrenia.	Two sessions per day for 10 days with an inter-session interval of 3 hours for a total of 20 sessions, with each one for 20 mins at 2 mA and 10 Hz. Location: 35 cm2 electrodes applied over AFz and Pz using a Soterix 1×1 CT device.	SAPS, AHRS, PSYRATS	Improvement in the scores, which lasted for 2 weeks, followed by a worsening noticed. Due to logistical reasons, patients could not come for 10 days, so they were offered 5 tACS sessions per day for 2 days with an inter-session interval of 20 minutes (accelerated tACS), but only mild improvement was reported with it. Side effects: No side effects noticed with accelerated tACS.
Dudi et al. (2024) 31 India Case report	1 (20-year-old male) Schizophrenia with persistent AH not responding to medications, ECT or CBT	Two sessions per day for 5 days with an inter-session interval of 3 hours- for a total of 10 sessions with each one for 20 mins at 2 mA and 40 Hz. Location: 2 electrodes over the scalp: one placed at a point midway between FP1 and F3 (left dlPFC) and another midway between P3 and T3 (left TPJ).	AHRS (baseline score 31) PSYRATS-AH (baseline score 36).	Patients started reporting improvement in AH by the 6th session, and after the 10th session, scores in AHRS (score-9) and PSYRATS-AH (score-16). Four weeks after intervention, while he was on Olanzapine 30 mg/d along with Flupentixol decanoate 20 mg every 21 days, his scores were: AHRS (15) and PSYRATS-AH (23). Side effects: Mild itching at the site of electrodes at the beginning of stimulation, which would disappear over time during the session.
Wang et al. (2023) 28 China Retrospective cohort study	17 (8 males/9 females) between 18 and 70 years Schizophrenia with refractory AH.	One session per day for 5 days a week, given for 4 weeks for a total of 20 sessions, with each one for 20 mins at 1–2 mA and 40 Hz. Location: tACS at AF3 and CP5 (in phase) at 1 mA (zero-to-peak) and return current at Cz (reversed-phase) at 2 mA (zero-to-peak).	PSYRATS	Nine responders (>20% reduction in their auditory hallucination PSYRATS scores after 4 weeks of tACS). Eight non-responders (≤20% reduction in PSYRATS scores over the same period). Week 6 follow-up was conducted to monitor sustained effects, but no detailed analysis or results from the follow-up are reported in the study. Side effects: Not mentioned.
Raymond et al. (2023) 29 USA Non-randomized, single blind controlled trial with crossover (conducted in two 4-week phases)	6 (18–55 years) Schizophrenia, schizoaffective, or psychotic bipolar disorder (confirmed by the Structured Clinical Interview for DSM-5), without an antipsychotic medication change for at least 4 weeks.	Initially, two sessions of 20 min of HD-tDCS (six patients –3 M/3 F) were administered for 5 consecutive days, a 4-week washout period, following which participants received HD-tACS (twice-daily for 5 days) (four patients-2 M/2 F). HD-tACS was delivered in phase for 20 mins at 2 Hz and current as per the location given below. Location (goal of delivering focalized current to the bilateral extra-striate visual cortex): AF7, 0.333 mA AF8, 0.35 mA P6, –0.325 mA P9, 0.333 mA P10, 0.65 mA PO7, –1.00 mA PO8, –0.675 mA O1, 0.334 mA	PANSS	HD-tDCS led to significant short-term reductions in general psychopathology on PANSS and event-related potentials (SSVEP P1), which correlated with clinical improvements. HD-tACS resulted in long-term improvements in general psychopathology on PANSS and global functioning. Side effects: Not mentioned.

dlPFC: dorsolateral prefrontal cortex; AHRS: Auditory Hallucination Rating Scale; HPSVQ: Hamilton Program for Schizophrenia Voices Questionnaire; tACS: Transcranial alternating current stimulation; SAPS: Scale for the Assessment of Positive Symptoms; PSYRATS: Psychotic Symptom Rating Scales: AH: Auditory Hallucination; ECT: Electroconvulsive therapy; CBT: Cognitive behavior therapy; TPJ: Temporo-parietal junction; HD-tACS: High-definition tACS; HD-tDCS: High-definition transcranial direct current stimulation.

Intervention Parameters

The tACS protocols varied across studies in terms of frequency, electrode placement, and stimulation parameters. The studies utilized different frequency bands: two studies used alpha-band stimulation at 10 Hz,^16,30 two studies gamma-band stimulation at 40 Hz,^28,31 and one used 2 Hz stimulation. ²⁹ Regarding current amplitudes, most studies (n = 3) applied a 2 mA current,^16,30,31 while the other two studies^28,29 used varying amplitudes depending on specific electrode sites. In terms of electrode placement, all studies targeted brain regions implicated in hallucinations, with a primary focus on the left DLPFC and TPJ. Raymond et al. ²⁹ uniquely targeted the bilateral extra-striate visual cortex. Treatment durations ranged from 20 to 40 minutes per session, and the number of sessions varied widely, from 10 sessions (5 days with twice-daily sessions) to 20 weekly sessions over a period of 20 weeks.

Primary Outcome: Reduction in Hallucinations

The studies used various standardized assessment tools to measure hallucination severity. The AHRS was employed in three studies,^25,26,30 while the PSYRATS were used in four studies.^16,27,28,31 PANSS was used in only one study. ²⁹ Additionally, other measures included the Hamilton Program for Schizophrenia Voices Questionnaire (HPSVQ) in two studies,^26,30 the Global Assessment of Functioning (GAF) in one study, ²⁹ and the Montgomery-Åsberg Depression Rating Scale (MADRS) in one study. ²⁹

All five studies reported positive outcomes with tACS intervention for hallucinations. In the case reports, Force et al. ³⁰ found improvements in the duration and controllability of AH after 20 weekly sessions of 10 Hz tACS. Pathak et al. ¹⁶ observed an improvement in hallucination scores that lasted for 2 weeks following 10 days of twice-daily 10 Hz tACS sessions, although subsequent worsening was noted. Dudi et al. ³¹ documented significant improvement in a 20-year-old male with treatment-resistant hallucinations, with AHRS scores decreasing from 31 to 9 and PSYRATS-AH scores reducing from 36 to 16 after 10 sessions of 40 Hz tACS.

In the cohort study, Wang et al. ²⁸ reported that 9 out of 17 participants (53%) were classified as responders, showing more than a 20% reduction in PSYRATS scores after 4 weeks of gamma-band (40 Hz) tACS intervention. In the controlled trial, Raymond et al. ²⁹ found that high-definition tACS (HD-tACS) led to long-term improvements in general psychopathology, as measured by PANSS, and in GAF, whereas HD-tDCS resulted in significant short-term reductions in symptoms.

Assessment of Risk of Bias

Joanna Briggs critical appraisal tools were used to assess the quality of studies included in the qualitative analysis, based on their respective study designs. For the case reports,^16,30,31 all eight domains of the appraisal tool were reported (Supplementary Table S5). However, for the quasi-experimental study, ²⁹ details regarding the reliability of outcome measures were not provided (Supplementary Table S6). Additionally, for the cohort study, ²⁸ there were no details regarding the identification of confounding factors or the strategies used to address these confounders (Supplementary Table S7).

Follow-up and Side Effects

Follow-up data were limited across the included studies. Only Dudi et al. ³¹ provided detailed follow-up results at 4 weeks post-intervention, showing some sustainability of therapeutic effects, with AHRS scores reduced to 15 and PSYRATS-AH scores to 23, compared to baseline scores of 31 and 36, respectively. Wang et al. ²⁸ mentioned conducting a 6-week follow-up but did not report detailed outcome data.

In terms of adverse effects, Dudi et al. ³¹ reported mild itching at the electrode site, which subsided during the session. Force et al. ³⁰ mentioned transient scalp pain (60% of sessions), tingling (60%), and burning sensation (55%) at the electrode site. The remaining studies^16,28,29 did not report adverse effects.

Discussion

The application of tACS as a non-invasive brain-stimulation technique has gained increasing attention for its potential to modulate neural oscillations in individuals experiencing AH. ³⁵ Hallucinations, which are a cross-cutting symptom across various psychiatric disorders, can often become treatment-resistant and significantly impair daily functioning. This meta-analysis incorporates findings from three RCTs involving a total of 72 participants to evaluate the efficacy of tACS in reducing hallucination severity, as measured by standardized rating scales. Additionally, the qualitative analysis of five supplementary studies further enriches our understanding of the clinical potential and application of tACS in managing hallucinations.

Summary of Main Findings

The meta-analysis of post-intervention outcomes demonstrated a non-significant trend toward tACS efficacy over sham treatment, with an SMD of –0.14 (95% CI = –0.61, 0.32). Similarly, the analysis at study endpoints also revealed a non-significant trend, with no clear advantage of tACS over sham treatment. In both analyses, heterogeneity was minimal, and although the direction of findings generally favored tACS, the pooled effects did not reach statistical significance. This study aimed to assess the preliminary evidence supporting the use of tACS as a potential intervention for hallucinations across psychiatric disorders. However, only three studies met the criteria for inclusion in the meta-analysis, and all focused exclusively on patients with schizophrenia.

Based on post-intervention data, there was no significant evidence of publication bias (LFK index = 0.27), whereas follow-up outcomes indicated potential bias, as suggested by the Doi plot (LFK index = 4.26). While individual studies reported reductions in hallucination scores using standardized rating scales, the aggregated effect was not statistically significant. These findings underscore the need for further research, particularly in patients with treatment-resistant schizophrenia. Although the current evidence is insufficient to draw firm conclusions, the observed trends suggest that tACS remains a promising avenue for future clinical trials targeting hallucinations.

Qualitative Synthesis

The qualitative synthesis of five additional studies in this review adds important context to the findings. Two case reports^16,30 used 10 Hz tACS and reported initial improvements in AH; however, these effects were not consistently sustained over time. This suggests the possibility of habituation to the stimulation and a subsequent decline in clinical effectiveness, a phenomenon commonly observed with neuromodulation interventions. ³⁶ Another case report ³¹ demonstrated significant improvement in a patient with treatment-resistant hallucinations using 40 Hz tACS. Supporting this, the cohort study ²⁸ using the same 40 Hz frequency reported that approximately half of the participants responded to treatment.

While the majority of studies focused on schizophrenia and AH, one comparative study, ²⁹ explored the effects of 2 Hz HD-tACS and HD-tDCS targeting the visual cortex. However, this study found improvements limited to general psychopathology. Overall, there is limited but emerging evidence for the application of tACS in psychiatric disorders beyond schizophrenia, and to date, other modalities of hallucinations—such as visual, tactile, or olfactory—have not been systematically evaluated. This highlights the need for broader research into tACS across different types of hallucinations and psychiatric conditions.

Stimulation Frequency and Parameters

The overall data suggest that gamma-frequency (40 Hz) tACS stimulation, which targets gamma oscillations essential for inter-regional brain communication, may hold particular promise in schizophrenia, where these oscillations are often disrupted.^27,28,31 In the case report by Dudi et al., ³¹ the application of 40 Hz tACS to the left DLPFC and left TPJ resulted in rapid and sustained improvement in persistent, treatment-resistant AH. Similarly, Wang et al. ²⁷ in their RCT demonstrated that 40 Hz tACS applied to the fronto-temporo-parietal region significantly reduced AH compared to a sham control and, when combined with pharmacological treatment, further reduced AH. These results provide early but encouraging evidence that 40 Hz tACS, by targeting disrupted fronto-temporal connectivity, may be an effective intervention for AH in schizophrenia.

Several studies have also explored the use of alpha-frequency (10 Hz) tACS, particularly targeting the left frontal and temporal regions, which are associated with enhanced alpha oscillatory activity. Two RCTs—by Mellin et al. ²⁵ and Zhang et al. ²⁶ —examined the effects of repeated 10 Hz tACS on AH. Both studies reported short-term improvements, with some reduction in hallucination severity at 1 week compared to sham treatment. However, these effects were not sustained at follow-ups. Thus, while alpha-frequency tACS may offer some transient therapeutic potential, its long-term efficacy and reliability in reducing AH remain unclear and warrant further investigation.

Recently, there has been emerging evidence supporting the use of theta-frequency (4–8 Hz) tACS, primarily aimed at alleviating negative and cognitive symptoms in schizophrenia, as demonstrated in some case series.^37,38 However, its potential application for reducing hallucinations has not yet been explored.

Across most of the reviewed studies^16,25,30,31 a current intensity of 2 mA was used. However, the duration and number of tACS sessions varied considerably—ranging from short-term protocols involving 5 days of stimulation with single or double daily sessions, to longer-term interventions spanning several weeks. This variability in protocol design poses a challenge for drawing robust, generalizable conclusions. Furthermore, the targeted brain regions also differed across studies, with some focusing on the fronto-temporo-parietal areas and others on the visual cortex, adding to the heterogeneity in findings and complicating comparisons across trials.

Safety and Tolerability

No serious adverse events leading to dropouts were reported in any of the studies included in this review. Only one study ²⁵ provided detailed documentation of side effects, including scalp pain, itching, tingling, and phosphenes. These adverse effects were described as mild and transient, with no significant impact on treatment adherence. To improve reporting and enhance the comparability of future studies, adverse effects should be systematically measured and consistently reported across all trials. This includes using standardized side effect rating scales, documenting the frequency, severity, and duration of each adverse event, and clearly stating whether any effects led to treatment discontinuation.

Limitations

However, several limitations must be acknowledged. The meta-analysis was based on a small number of RCTs (n = 3) with only 72 participants, which falls short of the optimal information size and lacks sufficient statistical power to draw conclusions. Additionally, the included studies differed significantly in their tACS protocols, including stimulation parameters, session duration, and frequency, introducing methodological heterogeneity. Despite this, a strength of the analysis lies in the consistent use of standardized rating scales—specifically the PSYRATS and AHRS—across the studies, allowing for more valid comparisons of hallucination-related outcomes.

It is also important to note that most participants were on concurrent antipsychotic medications, which could have influenced the therapeutic effects attributed to tACS. Finally, the limited reporting of safety and tolerability in two of the three RCTs restricts the ability to draw firm conclusions about the overall safety profile of tACS in this population.

Future research should prioritize larger, rigorously designed, double-blind, sham-controlled RCTs to more definitively evaluate the efficacy of tACS for treating hallucinations. Standardization of tACS protocols—including frequency, intensity, session duration, and electrode placement—is critical to ensure consistency and replicability across studies. Comparative investigations exploring different stimulation frequencies, particularly alpha (10 Hz) versus gamma (40 Hz), and targeting specific brain regions implicated in hallucinations, such as the fronto-temporo-parietal network, will be essential in identifying optimal stimulation parameters. In addition, the systematic and transparent reporting of adverse effects in all future trials is necessary to establish a comprehensive safety and tolerability profile for tACS, which will inform clinical application and patient care.

Conclusions

While our meta-analysis shows a non-significant trend toward improvement in AH with tACS in schizophrenia, the evidence remains inconclusive. Gamma-frequency tACS may offer potential benefits, but the small number of studies and variability in study designs highlight the need for further research to confirm its efficacy and establish tACS as a viable treatment for hallucinations.