In this randomized double-blind trial, we investigated whether externally induced left-hemispheric frontoparietal theta synchronization by multi-electrode online theta (6Hz) transcranial alternating current stimulation (tACS) would enhance the influence of a working memory training on negative symptoms of schizophrenia.

Negative symptoms have a negative impact on the prognosis of schizophrenia, but effective treatment for this symptom dimension is still under investigation. Identifying a treatment target that has a close link to negative symptoms or highly impacts negative symptoms may help to develop an effective therapy to counteract negative symptoms of schizophrenia. Recent theoretical and empirical work linking negative symptoms and cognitive impairment in schizophrenia has identified a potential treatment target: cognitive deficits. Evidence has indicated that cognitive remediation (CR) has a positive effect on improving negative symptoms of schizophrenia, in particular behavioural negative symptoms.

Although it is still debated which active components of CR contribute to the improvements in negative symptoms. The framework proposed by Gold and colleagues may be a candidate to explain how CR improves negative symptoms (i.e., possible through improving working memory). Anhedonia (i.e., the diminished ability to experience pleasure and reduced reactivity to pleasurable stimuli) represents a challenging negative symptom in schizophrenia.The impairment in hedonic processing has been associated with reduced motivation to engage in potentially rewarding events. Working memory (WM) plays an important role in the formation, maintenance, and retrieval of affective and value representations, all of which are essential for anticipatory pleasure. Individuals represent events and forecast pleasure by using WM in order to recruit motivational resources. It has been suggested that there is a hedonic detector system within the WM model and that WM serves as a potential underlying cognitive mechanism for anticipatory pleasure and goal-related behaviours. Problems in WM may reduce the ability to retrieve and manipulate information to motivate and guide future behaviour, thereby contributing to diminished motivation and the pleasure experience. It is known that the recruitment of the prefrontal-striatal system (including dorsolateral prefrontal cortex, cingulate cortex, insula, and ventral striatum) implicates in hedonic processing. It is also known that the same brain regions can be viewed as core brain regions of the WM network because they are activated during WM load. The overlap in activation of the prefrontal-striatal system during hedonic processing and WM provides robust evidence supporting the relationship between hedonic capacity and WM. Evidence indicated a correlation between the activity in WM brain networks and the improvement in negative symptoms following antipsychotic treatment, suggesting a mediating effect of WM on negative symptoms improvement in the context of a pharmacological intervention. More recently, studies indicated that 20 sessions of WM training (dual n-back task training) showed neural transfer effect to enhance hedonic processing in individuals with high social anhedonia and ameliorate hedonic dysfunction in schizophrenia patients with prominent negative symptoms.

In addition to WM training, non-invasive brain stimulation (NIBS) is also a non-pharmacological method to improve brain neural plasticity. For example, repetitive transcranial magnetic stimulation (rTMS) can change the activity of cortical nerve temporarily or continuously and enhance neural plasticity. However, the ability of rTMS alone to improve cognitive function in schizophrenia is frequently limited to some extent, necessitating a combination with WM training to boost cognitive functions.

Transcranial alternating current stimulation (tACS), a safe NIBS technique that applies low-intensity alternating current, is also a potential therapeutic option in treating the cognitive impairment in schizophrenia. The stimulation frequency of tACS is usually set to coincide with the targeted brain endogenous rhythms. It would synchronize the neural oscillations in the stimulated cortical regions to the applied stimulation frequency. Different tACS current intensity (0.5 to 4 mA), stimulation frequency (0.1 to 80 Hz), electrode montages, phase difference across the stimulation site, with/without DC offset, and the states (e.g., at rest or concurrently under the tasks) during stimulation contribute to its different effects. If both the target electrode and the reference electrode are in the same phase of the cycle of the current at any given time, the phase difference will be 0 degree (i.e., in-phase). The phase difference will be 180 degrees (i.e., anti-phase) if the electrodes are in the opposite phase. In-phase and anti-phase tACS over brain regions elicits synchronization and desynchronization of neuronal activity across the brain regions, respectively.

The state-dependent tACS effects indicate that the effects of tACS are enhanced when the state of the targeted brain regions is active. Specifically, synchronization of frontoparietal regions at theta frequencies dominates during an executive task. tACS at the frequency close to the theta oscillation activated by an executive task would elicit more resonance and may, in turn, help to enhance executive function. tACS applied when an individual is concurrently engaged in a specific cognitive task is defined as online tACS. In healthy subjects, online theta (6Hz) in-phase tACS facilitated frontoparietal phase coupling (synchronization) resulting in improved WM performance, whereas online theta (6Hz) anti-phase tACS disrupted theta phase-coupling (desynchonization) resulting in impaired WM performance.

The prolonged after-effects of tACS may be related to a phenomenon called spike-timing-dependent plasticity (STDP), a process that modifies the connection strengths based on the relative timing of the input and output spikes (or action potentials) of a neuron. STDP means that synapse will be strengthened if input action potentials occur immediately before the output action potentials, and synapse will be made weaker if an input action potentials occur immediately after an output action potentials. The STDP process is known to explain in part long-term depression (LTD) and long-term potentiation (LTP) of nervous systems. To sum up, tACS at a frequency close to that of resonance frequency during tACS may intensify the synapses across the stimulated regions through the mechanism of STDP. Repetitive (multiple-session) tACS during specific time intervals may consolidate the neuroplasticity effects, and may, in turn, elicit a long-lasting effect for further clinical application in treating neuropsychiatric disorders. In schizophrenia patients, a few case reports indicated 1-5 sessions of online theta in-phase tACS over left frontoparietal regions improved the performance of WM and several other cognitive domains.

So far, there is no relevant research report exploring whether and how the joint intervention (i.e., online theta (6Hz) in-phase tACS over the left frontoparietal regions plus WM training) will have a combined effect on improving the performance of WM and other cognitive domains in schizophrenic patients. Neither is any research investigating whether online theta (6Hz) in-phase tACS over the left frontoparietal regions can potentiate the effects of WM training on improving negative symptoms of schizophrenia. To sum up, we hypothesize that the combination training mode of active online theta tACS plus WM training would have a greater ability to reduce negative symptoms of schizophrenia compared to the training mode of sham online theta tACS plus WM training. The study aims to test the efficacy of "active online theta tACS plus WM" versus "sham online theta tACS plus WM training" in improving negative symptoms, WM performance, and other cognitive domains performance of schizophrenia using a double-blinded, randomized sham-controlled trial design. In order to evaluate the efficacy of the interventions, we will use behavioral outcomes (e.g., negative symptoms assessment and neurocognitive assessment) and neurophysiological outcomes (e.g., EEG and heart rate variability).

Methods

Dual N back training

The study will use a dual n-back task for WM training. In this task, squares at 8 different locations will show up sequentially (stimulus length, 500 ms; inter-stimulus interval, 2,500 ms) on a computer screen every 3 seconds. Simultaneously with the presentation of the squares, one of eight consonants will show up sequentially through a speaker. Participants will have to judge whether the location of a square and the consonant they heard matches the one n-back before (the same n value for both visual and auditory targets). Each training session has 20 blocks consisting of 20 + n trials and takes around 25 min. Each block includes six auditory and six visual targets (four appearing in only one modality, and two appearing in both modalities simultaneously) whose locations are random. Participants will have to make responses manually by pressing the mouse left-click button for visual targets and the right-click button for auditory targets. No responses were required for non-targets. If a target is correctly detected, a green flash will be given, which constitutes a positive feedback. If a target is falsely detected, a blue flash will be given, which constitutes a negative feedback. The dual N back training is designed to continuously vary its difficulty by modifying the WM load (i.e., the level of n) and thereby track the participants' performance. Each training session begins at n = 1. Participants' performance will be analyzed after each block and the level of n for the next block will be adapted according to the following principle. The level of n increases by 1 in the next block if the mistakes per modality made by the participant are<3. Conversely, n would decrease by 1 if the mistakes per modality made by the participant are>5. In all other cases, the n is kept unchanged. Participants will come to the laboratory and take part in the WM training sessions twice daily for 5 weekdays (total 10 sessions), with each session lasting for approximately 25 minutes. The time interval between twice-daily sessions will be >3 hours.

Online left-hemispheric frontoparietal theta (6Hz) in-phase tACS

θ tACS will be administered during the dual n-back task, starting at the beginning of each task and lasting for 20 min. In the active θ tACS condition, sinusoidal tACS will be delivered by two battery-operated devices connected with two 4 × 1 wire adaptors (Equalizer Box, NeuroConn, Ilmenau, Germany). The electrode montages used for θ tACS will be positioned at left frontoparietal locations. The stimulation electrodes of the 1st stimulator will be placed at the International 10-10 electrode position F1, F5, AF3, and FC3 (stimulation electrodes) and CPz (return electrode) to cover the left frontal cortex. For the 2nd stimulator, the stimulation electrodes will be placed at P1, P5, CP3, and PO3 (stimulation electrodes) and FCz (return electrode) to cover the left parietal cortex. A NeuroConn digital to analog converter (DAQ) will control the two stimulators and create an in-phase (synchronous) setup (0° relative phase difference between the output signals of the two tACS-stimulators).

Inclusion Criteria:

1. Eligible participants aged 20-65 with DSM-V-defined schizophrenia or schizoaffective disorder.
2. Duration of illness > 2 years.
3. Being clinically stable and on an adequate therapeutic dose of antipsychotics for at least 8 weeks prior to enrolment.
4. Agreement to participate in the study and provide the written informed consent.

Exclusion Criteria:

1. Having unstable medical conditions, current psychiatric comorbidity or active substance use disorder (in exception to caffeine and/or tobacco).
2. Having a history of seizures, meningitis or encephalitis.
3. Having contraindications for transcranial electrical stimulation or transcranial magnetic stimulation, e.g., pacemakers, metallic or magnetic pieces in the head/brain, ear implants and other implantible brain medical devices.
4. Having a history of intracranial neoplasms or surgery, or a history of severe head injuries or cerebrovascular diseases.
5. Pregnancy or breastfeeding at enrollment.
6. Skin lesions on scalp at the area of electrode application

The allocation concealment was further ensured by the administration of tACS using "study mode of the device" in which a five-digit numerical code specific to individual participant was entered into the device (Eldith DC stimulator Plus, NeuroConn, Ilmenau, Germany) that resulted in either active or sham stimulation, i.e., the researcher got the randomization code and a unique five-digit numerical code for an individual participant from the study coordinator while tACS administrator entered the code for study mode into the device. The study coordinator had continuous access to the randomization list and unblinded the study after the final visit of the last participant. Not until the unblinding of the trial did the participants, tACS administrators, researchers and clinical raters know the actual stimulation types.