This randomized, double-blind, controlled clinical trial evaluates the efficacy of combining HD-tDCS with AE to improve cognitive function in individuals with CIAS. Eligible participants will be randomly assigned to one of four groups in equal proportions (1:1:1:1): MRI-guided HD-tDCS +AE (Group 1), MRI-guided HD-tDCS alone (Group 2), AE alone (Group 3), and a control group (Group 4). Assessments will be conducted at baseline (T0), at 2 weeks (T1), at 4 weeks (T2), and at 6 months post-intervention (T3). The specific intervention measures for each group are delineated as follows: the MRI-guided HD-tDCS + AE group first undertakes a 15-minute warm-up exercise. Subsequently, HD-tDCS and AE treatments are administered concurrently for a duration of 30 minutes. The HD-tDCS group receives 30 minutes of electrical stimulation per session. The AE group’s intervention comprises a 15-minute warm-up followed by a 30-minute formal AE stage, amounting to a total of 45 minutes. The control group does not receive HD-tDCS or AE but instead is provided with routine psychiatric care, including regular psychiatric follow-ups, medication management, and general health guidance, but no experimental interventions. Each intervention for the aforementioned groups is conducted once daily, five days a week, over a period of four weeks, culminating in a total of 20 sessions. The study design is illustrated in Figure 1.

Participants will be recruited from the Department of Psychiatry at the Second Affiliated Hospital of Xinxiang Medical University. The trial will be publicized via the hospital’s official website and various media outlets. Information leaflets will be distributed within the Department of Psychiatry, where psychiatrists will provide an overview of the study. Potential participants will receive both oral and written information regarding the research procedures, along with their potential benefits and risks. All participants and their legal guardians, will sign informed consent forms. In accordance with the Helsinki Declaration on Ethical Principles for Medical Research Involving Human Subjects, this study received approval from the Research Ethics Committee of the Second Affiliated Hospital of Xinxiang Medical University (Approval Number: XYEFYLL-2025-16, Approval Date: February 17, 2025) and has been registered with the Chinese Clinical Trial Center (Registration Number: ChiCTR250010698). The trial will take place at the Second Affiliated Hospital of Xinxiang Medical University. This research protocol adheres to the 2013 Standard Protocol Items: Recommendations for Intervention Trials (SPIRIT) Statement Guidelines (44).

Eligible individuals must meet the following inclusion criteria (1): The current episode of the patient aligns with the diagnostic criteria for schizophrenia according to the Diagnostic and Statistical Manual of Mental Disorders-5 (DSM-5) (2). patients who achieved the stable period through oral antipsychotic drug treatment, as judged by the following criteria: delusion, hallucinatory behaviors, exaggeration, and suspicion/victimization items in the positive and negative symptom scale (PANSS), abnormal thought content scores of ≤ 5 in the general psychopathology scale, and PANSS conceptual disorganization scores of ≤ 4, if all the aforementioned criteria were met, the patient was considered to be in a stable period of schizophrenia (45) (3). Patient must be undergoing treatment with atypical antipsychotic medications, with equivalent doses calculated using the defined daily dose method (4). To minimize study population heterogeneity only Han ethnic participants will be enrolled (5). Aged between 18 and 55 years (6). Written informed consent must be obtained from the patient. Exclusion criteria (1): Individuals with organic brain lesions, intellectual disabilities, or other physical ailments (2). Those experiencing intracranial hypertension (3). Frequent or persistent migraines (4). Individuals with a personal history of epilepsy or a familial history of epilepsy (5). Patients possessing metallic implants within their bodies (6). Severe substance uses disorder (7). Pregnant women or those who are breastfeeding (8). Subjects currently displaying significant abnormalities in laboratory tests (i.e. blood routine, liver and kidney function, electrolytes, and thyroid function. Results that exceed the normal range and are recognized by the attending psychiatrist as having clinical significance, such as severe anemia, acute liver/kidney function disorders) (9). Patients undergoing modified electroconvulsive therapy (10). Individuals with a history of limb disability or leg injury. Participants will be evaluated based on the inclusion and exclusion criteria outlined in Table 1.

The randomization process will be conducted by an independent researcher who is not involved in the experiment. This researcher will utilize a random number table to randomly allocate participants into four distinct groups, labeled as “1, 2, 3, and 4.”. Specifically, the group coding is as follows: Group 1 corresponds to the HD-tDCS combined with AE group, Group 2 to the HD-tDCS group, Group 3 to the AE group, and Group 4 to the control group. The random allocation process will be implemented using Microsoft Excel 2019. Upon the recruitment of all 48 eligible participants, each individual will be assigned a unique identifier and paired with a randomly generated value using Excel’s “=RAND ()” function. The “Paste Special” function will then be employed to convert these random numbers into fixed values. Subsequently, the list of participants will be sorted in descending order based on the random values, with participants allocated to groups 1 to 4, each comprising 12 individuals, according to their ranking positions. Once the randomization scheme is established, it will be sealed within an opaque envelope, which will be opened sequentially in accordance with the order of subject enrolment. The allocation scheme contained within each envelope will dictate the group assignment for the corresponding participant. This approach ensures that researchers are unable to access the randomization scheme in advance, thereby safeguarding the integrity of the randomization process.

All participants and researchers will remain blinded throughout the research process. During the entire research process, the designated stimulus conditions will be kept strictly confidential and will only be disclosed to the principal researcher. Four distinct groups will be tasked with administering the treatment, with strict measures in place to prevent inter-group communication regarding the particulars of each treatment. Each group will be informed only of its own treatment protocol and will remain unaware of the grouping details or the treatment plans of the other groups. The principal investigator will assign specific measures to each group, which will then implement the experimental treatment for their respective patient cohorts. Patients will remain unaware of their group allocation and will be instructed not to discuss the grouping, the treatment received, the completion of questionnaires, or any other aspects of the experimental protocol. During the phase of result measurement, personnel involved in data collection and analysis will also be blinded to the grouping information and specific measures. Subsequently, the principal investigator will conduct two unblinding procedures: the first will occur before data locking, and the second will follow data analysis. The initial unblinding will occur after data locking, organizing the data into groups 1–4 without disclosing the actual correspondence between these groups and their respective identities. The second unblinding will take place after the data analysis, explicitly revealing the specific identities of groups 1-4. Furthermore, to ensure that participants are unaware of their group assignments, the intervention activities will be conducted in designated rooms. Although the researchers will be aware of the group assignments, they must not disclose this information to the participants and will not be involved in the evaluation or analysis of the study. Additionally, the evaluators and data analysts will remain unaware of the intervention group’s situation throughout the entire research process.

All researchers involved in data management and statistical analysis will remain blinded to the grouping information throughout the entire process. Specifically, data will be analyzed using coded group identifiers (e.g., Group A, B, and C), and the blinding code will remain undisclosed until the completion of the statistical analysis and the locking of the database. This procedure aligns with the principle of the blind method for minimizing assessment bias, as outlined in clinical trial protocols (46).

A Blinding Integrity (BI) assessment will be conducted following the final intervention session. Participants will be required to specify their perceived group allocation via a forced-choice questionnaire, asking: “Which intervention do you believe you received?” The response options will include (1): Combined HD-tDCS and AE (2), HD-tDCS alone (3), AE alone (4), Control, or (5) Unsure. To rigorously assess the blinding effect, Bang’s Blinding Index (BI) will be calculated for each group. This index allows for the distinction between genuine unblinding and random guessing. The BI is the primary tool for quantitatively evaluating the effectiveness of blinding implementation in clinical trials. By calculating the BI separately for each group, we can accurately assess the maintenance of blinding across different intervention groups (47).

Unlike drug trials, these trials—such as those involving surgery, rehabilitation, psychological, and physical therapy—cannot physically blind the “treatment itself” (48). During the informed consent process, we informed participants that the study was designed to investigate the effects of non-pharmacological interventions on cognitive function of individuals with schizophrenia. In the informed consent form, all intervention measures were described as “interventions with potential benefits,” without disclosing which group was the experimental group and which was the control group.The language used in the informed consent form was intentionally general to avoid revealing specific differences between groups, thereby minimizing the expectancy effect. Upon completion of the study, it is imperative to fully disclose the actual research design and the group assignments to all participants.

To construct precise and anatomically accurate head models, T1- and T2-weighted MRI data will be collected to optimize electrode placement and account for individual anatomical variations. These models will be developed and simulated using SimNIBS software to estimate the electric field distributions resulting from HD-tDCS stimulation. Eight distinct 4*1 montages centered on the mPFC will be used to simulate 280 the E-field for each brain. The finite element method will be utilized to compute the normal component of the induced electric field. The head models, derived from individual MRI data will differentiate among five tissue types: skin, skull, cerebrospinal fluid, gray matter, and white matter. The mPFC will be delineated in each participant’s brain according to the Ranta atlas. The study will utilize an MRI-guided HD-tDCS device (MxN-9-9002A, Soterix Medical, New York, USA). We used five small circular electrodes (1×1 cm each) with a small area for high-precision electrode arrangement, concentrating the stimulation current on the target brain region. Specifically, the central electrode (anode) will be placed in the mPFC (fz) according to international 10–20 system, while the four surrounding cathodes will be directly located in front of Fp1, Fp2, F7, and F8, thus forming a circular current circuit. The current intensity will be maintained at 2 mA. Each HD-tDCS session will involve the delivery of a 2mA direct current for a duration of 30 minutes, with ramp-up and ramp-down periods of 30 seconds each. HD-tDCS will be administered once daily for 30 minutes, five times a week over a total duration of 4 weeks, culminating in 20 treatment sessions.

AE training will be administered using a stationary bicycle. The training sessions will be characterized by moderate to low intensity and will include both a warm-up phase and a formal AE phase. Each session is customized according to each individual’s maximum heart rate (HRmax), which is calculated using the formula: HRmax = (220 - Age) × 0.7. (34). Each session will last for 45 minutes, comprising a 15-minute warm-up and a 30-minute formal AE phase. During the warm-up, participants’ heart rates are expected to remain between 55% and 60% of HRmax. In the formal AE phase, heart rates should be maintained between 60% and 70% of HRmax (34). A portable monitor will be utilized to record heart rates throughout the training, ensuring participants maintain an appropriate level of physical exertion. Blood pressure measurements will be taken both before and after the sessions. The intervention will occur once daily, 5 days per week, over a period of 4 weeks, culminating in 20 treatment sessions.

Structural T1-weighted MRI images will be acquired using a three-dimensional magnetization-prepared rapid gradient-echo (3D MPRAGE) sequence with the following parameters: repetition time (TR)/echo time (TE)/inversion time (TI) = 240/2.14/100 ms, flip angle = 8 degrees, field of view (FOV) = 224 x 224 mm, voxel size = 0.7 mm isotropic, bandwidth = 210 Hz/pixel, integrated parallel acquisition techniques (iPAT) = 2, and an acquisition time of 7 minutes and 40 seconds. T2-weighted images will be acquired using a 3D T2 sampling perfection with application-optimized contrasts using different flip angle evolutions (T2-SPACE) sequence, with parameters: TR/TE = 320/565 ms, variable flip angle, FOV = 224 x 224 mm, voxel size = 0.7 mm isotropic, bandwidth = 755 Hz/pixel, iPAT = 2, and an acquisition time of 8 minutes and 24 seconds. The total duration for structural imaging acquisition will be 16 minutes and 4 seconds. Resting-state functional MRI (fMRI) images will be acquired with the following specifications: TR/TE = 720/33.1 ms, flip angle = 52 degrees, FOV = 208 x 180 mm, matrix size = 104 x 90, slice thickness = 2.0 mm, 72 slices, 2.0 mm isotropic voxels, multiband factor = 8, echo spacing = 0.58 ms, and bandwidth = 229 Hz/pixel. The acquisition of resting-state functional images will require 14 minutes and 33 seconds.

A 48-channel functional near-infrared spectroscopy (fNIRS) device (NirScan model from Danyang Huichuang Medical Equipment Co. Ltd in China) will be employed in this study. The fNIRS data will be collected during the administration of The Chinese version of the Verbal Fluency Test (VFT) serves to evaluate verbal fluency, working memory, verbal recall, attention, and retrieval capabilities. The VFT task is structured into three distinct phases: a 30-second pre-task baseline period, a 60-second task period, and a 70-second post-task period. The baseline phase is characterized by the absence of VFT task performance. During the 30-second pre-task baseline phase, participants are instructed to count repeatedly from one to five until the commencement of the task period. In the subsequent 60-second VFT task period, participants are required to generate as many phrases as possible using simple words such as “white,” “north,” and “big.” Following the completion of the phrase generation task, participants are instructed to resume counting from one to five repeatedly throughout the 70-second post-task period. Fifteen light source probes and sixteen light detector probes will be positioned on the bilateral frontotemporal cortex, with a distance of 3 cm maintained between each light source and detector probe. In accordance with the 10/20 electrode placement system, the central probe will be positioned at FPz, while the lower boundary of the probe array will extend from Fp1 to Fp2. Hemodynamic changes will be assessed through measurements of oxyhemoglobin, deoxyhemoglobin, and total hemoglobin. Resting-state and task-state fNIRS data acquisitions will be performed for all participants in this study. The raw fNIRS data will be preprocessed utilizing the MATLAB Home 3 toolkit. Five statistical metrics—mean, variance, skewness, kurtosis, and peak value—will be calculated for changes in oxy-Hb signals, based on the spatial average across all 48 channels. This methodology will generate a total of 240 independent features for each subject.

The resting-state EEG data were recorded utilizing a PN-NET multichannel electroencephalogram device with 64 scalp electrodes, 1 reference electrode on the nose tip, and 2 electro-ocular electrodes positioned at the right eye corner and beneath the left eye. Data collection transpired in a tranquil examination room by a consistent researcher. Subjects were instructed to maintain stillness, wakefulness, and relaxation, minimize bodily and ocular movements, and abstain from specific tasks. Electrode impedance was maintained below 5 kΩ to ensure data fidelity during collection. A sampling rate of 1,024 Hz was employed, capturing EEG data with closed eyes over a 5-minute interval.

Participants will undergo a fasting anterior elbow vein puncture to collect a 5 ml blood sample, which will be transferred to an anticoagulant tube and centrifuged at 300 rpm for 10 minutes. The supernatant and sedimented blood cells will be separately transferred into two Eppendorf tubes. Subsequently, the centrifuge tube will be placed in a freezing centrifuge for further centrifugation. The supernatant will be aspirated, and its protein concentration will be quantified using the BCA kit. Subsequently, equal volumes of samples and molecular weight standards will be subjected to Tris-SDS-PAGE electrophoresis, followed by transfer onto a PVDF membrane. The membrane will be incubated in a 5% skimmed milk solution for blocking for a duration of 2 hours. After a washing step, primary antibodies specific to BDNF and β-actin, diluted at a ratio of 1:500, will be applied and incubated overnight at 4°C. Thereafter, secondary antibodies, also diluted at a ratio of 1:500, will be introduced and incubated at room temperature for 2 hours. The PVDF membrane containing the target protein will then be immersed in a developing solution for analysis using a chemiluminescence imaging analyzer. The exposure time will be adjusted based on the protein’s abundance, and the expression levels of BDNF will be quantitatively assessed using ImageJ software.

The primary outcome measure will be the change in scores on the Chinese version of the Food and Drug Administration utilizes the MATRICS Consensus Cognitive Battery (MCCB). Assessments will be conducted at baseline (T0), at 2 weeks (T1), at 4 weeks (T2), and at 6 months post-intervention (T3). The MCCB consists of 10 cognitive tests (3), it covers a total of seven different cognitive domains, including speed of processing, attention/vigilance, working memory, verbal learning, social cognition, reasoning and problem-solving (49). MCCB provides a comprehensive evaluation of specific cognitive domains associated with schizophrenia. Assessments will be conducted at baseline (T0), at 2 weeks (T1), at 4 weeks (T2), and at 6 months post-intervention (T3).

For the assessment of secondary outcomes, this study employed the Repeatable Battery for the Assessment of Neuropsychological Status (RBANS) and the Wisconsin Card Sorting Test (WCST) to evaluate global and executive functions. Simultaneously, the Facial Emotion Perception Test (FEPT) and the Voice Emotion Perception Test (VEPT) were utilized to assess emotion perception abilities within the domain of social cognition. The severity of clinical symptoms and overall changes were quantified using the Positive and Negative Syndrome Scale (PANSS) and the Brief Psychiatric Rating Scale (BPRS). Furthermore, improvements in social functioning and quality of life were evaluated using the Social Disability Screening Schedule (SDSS) and the Schizophrenia Quality of Life Scale (SQLS). The RBANS comprises five components: immediate memory, visuospatial skills, language, attention, and delayed memory. Its notable features include ease of administration, brief administration time, and robust reliability and validity. The RBANS is effective for detecting cognitive impairment and can also be used to assess changes in cognitive function following treatment or intervention (50). The WCST is a widely utilized standardized tool in neuropsychological research and clinical cognitive assessment, primarily crafted to assess cognitive flexibility and abstract thinking within executive functions. Participants must discern classification rules based on evolving feedback (e.g., color, shape, or number) and adapt their strategies accordingly. This assessment effectively mirrors the operation of the prefrontal cortex and is frequently employed to pinpoint executive function impairments linked to psychiatric disorders, brain injuries, and ageing (51). The FEPT is predominantly employed to gauge an individual’s capacity to perceive and comprehend the emotions conveyed in the facial expressions of others. This evaluation presents a sequence of standardized facial expression images (e.g., happiness, sadness, anger, fear, etc.), requiring participants to recognize the type or intensity of emotion (52). The VEPT is employed to assess an individual’s ability to perceive emotional nuances in others’ voices. Participants listen to voice samples (neutral or expressing emotions like joy, sadness, anger, etc.) and must identify the specific emotion conveyed (53). The BPRS is utilized to gauge the intensity of psychiatric symptoms, with higher scores indicating more pronounced symptom severity (54). The PANSS is a widely used standardized tool in psychiatric clinical and research settings, primarily crafted to evaluate symptom severity in individuals with schizophrenia (55). It consists of the Positive Symptom Scale and the Negative Symptom Scale, each comprising 7 items, alongside a general psychopathology scale with 16 items, totaling 30 items. The SDSS is employed to quantitatively evaluate social functioning impairment in patients, encompassing crucial aspects such as work and social interactions. It is user-friendly and suitable for rehabilitation assessments (56). The SQLS assesses patients’ quality of life using various indicators that mirror their subjective experiences, offering a benchmark for evaluating effectiveness (57).

To explore the underlying mechanisms, the study adopted a multimodal assessment framework. MRI and fNIRS were utilized to investigate changes in brain activity, structure, and functional connectivity. EEG and P300 were employed to capture the neurophysiological dynamics of electrical brain activity. EM served as an objective behavioral-physiological marker of cognitive processing. Additionally, plasma BDNF levels were measured to evaluate the molecular correlates of neurotrophic support and plasticity. This integrated approach aims to elucidate the therapeutic effects of MRI-guided HD-tDCS combined with AE for CIAS and to uncover the neurobiological mechanisms underlying these effects.The procedure for the site visit is outlined in Table 2.

We performed the necessary calculations using G*Power software (version 3.1.9.7). Under the following command, test family: F tests; Statistical test: ANOVA: Repeated measures, within-between interaction (58).The research parameters were defined as follows: a Type I error rate of 0.05, a statistical power of 80%, and an effect size of 0.25. Utilizing a repeated measures ANOVA model and considering a potential participant dropout rate of 20%, we determined that the required sample size is 48 patients, with 12 patients assigned to each group. In preliminary studies where prior sample size data are unavailable, a group size of 12 subjects is deemed appropriate. We will justify the rationale for selecting a sample size of 12 per group from three perspectives: feasibility, improved precision in estimating means and variances, and adherence to regulatory requirements (59). This sample size is expected to provide adequate statistical power to address the study objectives.

The evaluation of cognitive function and the severity of psychiatric symptoms will be conducted by two psychiatrists, who will remain blinded to the group assignments. All demographic information and scale-related data will be recorded in electronic Case Report Forms (eCRFs) and stored on a designated website. Access to the securely stored eCRFs will be restricted to the project leader and the principal investigator. To maintain confidentiality, anonymization will be implemented during data entry by substituting patients’ actual names with unique identification numbers. The independent data monitoring committee will be established to monitor safety, the occurrence of adverse events, and advise on trial design decisions.

The Shapiro-Wilk test will be employed to determine whether the quantitative data conform to a normal distribution. When the quantitative data are normally distributed, they will be described using the mean and standard deviation. If the data deviate from a normal distribution, the median and interquartile range will be utilized for presentation. Categorical data will be presented as counts and percentages. To evaluate differences in baseline characteristics among groups, appropriate statistical tests such as t-tests, nonparametric tests, chi-square tests, Mann-Whitney U tests, or ANOVA will be selected based on the type and distribution of the data. Additionally, Fisher’s exact test or chi-square test will be used to compare adverse reactions between groups.

This study will adhere to the CONSORT guidelines and perform data analysis based on the intention-to-treat principle, ensuring that all participants are evaluated according to their initial randomization assignments. To assess differences in primary and secondary outcomes both between and within groups over time, generalized linear mixed models and repeated-measures ANOVA will be employed as appropriate. A stepwise model selection approach will be implemented to derive a succinct multivariate regression model. Age, gender, and the Defined Daily Dose (DDD) of antipsychotic medications will be included as covariates in all models. The treatment effect will be evaluated using the likelihood ratio test to ascertain whether the coefficients for treatment and the interaction between time and treatment are both zero. To correct for multiple comparisons across time points, the Bonferroni correction method will be applied to adjust the p-values. All statistical analyses will be conducted using SPSS version 20.0, with a significance threshold set at P < 0.05 to denote statistical significance. Research framework is illustrated in Figure 2.

Despite being deemed low-risk for participants, the Institutional Review Board of the Second Affiliated Hospital of Xinxiang Medical University opted to form a data monitoring committee for the ensuing experiments to safeguard data quality and participant well-being.

To evaluate the safety and potential adverse effects of HD-tDCS, this study will thoroughly investigate both severe and mild adverse events. All participants will be required to complete the HD-tDCS Adverse Reaction Scale (ARS), which encompasses common side effects such as tingling, mild redness, itching, and discomfort at the stimulation site. Participants will rate their adverse reactions on a scale from 0 to 5, with all occurrences documented in the CRFs. The HD-tDCS equipment will be operated by trained therapists proficient in its use, aiming to minimize the risk of participants encountering any significant health hazards or adverse effects. In the event of a severe adverse reaction, the participant will be withdrawn from the study, and the incident will be promptly reported to the ethics committee.

To ensure data quality, a dual verification system will be implemented throughout the study, covering both the treatment phase and the 6-month follow-up period. The lead researcher will conduct weekly evaluations of CRFs and estimation forms. Additionally, a bi-weekly comparison between hard copy and electronic records will be performed to verify consistency. Any discrepancies will be documented and discussed in team meetings to facilitate prompt corrective actions.

The procedure for implementing substantial protocol modifications requires submitting all changes to the Ethics Review Committee of the Second Affiliated Hospital of Xinxiang Medical University for approval. Upon receiving approval, the research team will formally notify all relevant stakeholders in writing. Subsequently, the approved amendments will be promptly recorded on the Chinese Clinical Trial Registry platform.

Prior to enrolling any participants, the research team will register and disclose the trial details on the Chinese Clinical Trial Registry. Furthermore, there is a commitment to publicly release the complete study outcomes. The dissemination of results will be achieved through presentations at international conferences and submissions for peer-reviewed publication in scientific journals.

Regarding confidentiality, all study personnel have received training and certification in human subjects’ research protections through the Second Affiliated Hospital of Xinxiang Medical University Training Program. The Principal Investigator provides ongoing training and oversight to the evaluation coordinator to maintain the confidentiality and privacy of all participants and their data.

Raw data will be generated at the Second Affiliated Hospital of Xinxiang Medical University. Derived data supporting the findings of this study will be will be accessible upon request from the corresponding author.