A new perspective for evaluating the efficacy of tACS and tDCS in improving executive functions: A combined tES and fNIRS study

Abstract

Background

Executive function enhancement is considered necessary for improving the quality of life of patients with neurological or psychiatric disorders, such as attention‐deficit/hyperactivity disorder, obsessive‐compulsive disorder and Alzheimer's disease. Transcranial electrical stimulation (tES) has been shown to have some beneficial effects on executive functioning, but the quantification of these improvements remains controversial. We aimed to explore the potential beneficial effects on executive functioning induced by the use of transcranial alternating current stimulation (tACS)/transcranial direct current stimulation (tDCS) on the right inferior frontal gyrus (IFG) and the accompanying brain function variations in the resting state.

Methods

We recruited 229 healthy adults to participate in Experiments 1 (105 participants) and 2 (124 participants). The participants in each experiment were randomly divided into tACS, tDCS, and sham groups. The participants completed cognitive tasks to assess behavior related to three core components of executive functions. Functional near‐infrared spectroscopy (fNIRS) was used to monitor the hemodynamic changes in crucial cortical regions in the resting state.

Results

Inhibition and cognitive flexibility (excluding working memory) were significantly increased after tACS/tDCS, but there were no significant behavioral differences between the tACS and tDCS groups. fNIRS revealed that tDCS induced decreases in the functional connectivity (increased neural efficiency) of the relevant cortices.

Conclusions

Enhancement of executive function was observed after tES, and the beneficial effects of tACS/tDCS may need to be precisely evaluated via brain imaging indicators at rest. tDCS revealed better neural benefits than tACS during the stimulation phase. These findings might provide new insights for selecting intervention methods in future studies and for evaluating the clinical efficacy of tES.

This study found that both 20 Hz tACS and tDCS on the right IFG could effectively enhance the behavioral performance related to executive functions, and variations in cortical activity are more sensitive for assessing benefits than behavior changes by tES intervention.

1. INTRODUCTION

As top‐down active mental processes, executive functions are necessary for maintaining mental health and promoting personal growth; they mainly include inhibition (self‐control and control of informational interference to ensure concentration), working memory (a crucial structure for thought processes that is closely related to reasoning ability), and cognitive flexibility (a prerequisite of innovation and changes in cognitive state) (Diamond, 2012). Good executive functioning can promote self‐regulation, which is key to ensuring an individual's adaptability to their environment and a prerequisite for success in school or work (Diamond & Lee, 2011a; Hofmann et al., 2012). However, patients with a variety of clinical diseases, such as obsessive‐compulsive disorder, obesity, and attention‐deficit/hyperactivity disorder, have varying degrees of cognitive symptoms related to impaired executive functions, which in turn becomes an important factor affecting their quality of life (Barkley, 1997; Dohle et al., 2018; Pauls et al., 2014). Therefore, improving executive functioning is an effective tool for improving mental health and reducing cognitive deficits. Previous studies have found that regular computer cognitive training, physical exercise or music training can effectively improve an individual's executive functioning (Diamond & Lee, 2011b; Siponkoski et al., 2020; Westwood et al., 2023). However, the training effect of these intervention methods might be reduced by high time costs (repeated training is required) and a low degree of involvement (occupying more cognitive resources). Transcranial electrical stimulation (tES) is a safe, noninvasive technique that alters neuronal activity by delivering a low‐intensity current to a target brain area; this technique has been shown to improve cognitive performance (Lin et al., 2021; Yavari et al., 2018). Transcranial direct current stimulation (tDCS) and transcranial alternating current stimulation (tACS) are two widely used noninvasive stimulation paradigms (Hong et al., 2022). tDCS mainly changes neuronal membrane potentials by delivering a stable direct current, while tACS uses an external oscillation source to induce neuronal oscillations, thereby promoting cognitive improvement through nerve entrainment (Dedoncker et al., 2016; McLaughlin et al., 2022). tDCS has been shown to improve global executive functions, but this improvement is limited, appearing to have more effect on refresh (working memory) than inhibition and cognitive flexibility, and this enhancement depends on the state of cognitive load of the individual (Figeys et al., 2023; Imburgio & Orr, 2018). A study found that tACS could ameliorate declining brain function and improve working memory and long‐term memory in older adults (Grover et al., 2022). However, the effect of tACS on executive functions is more complex because, unlike tDCS, it involves frequency parameters. It has been confirmed that active tACS can effectively promote the performance of executive functions, but this enhancement effect varies depending on the frequencies that are used (Reinhart, 2017). tACS with θ band frequencies has a limited positive effect on executive functions, but the internal neural mechanism is unclear (Mosbacher et al., 2021), while tACS with γ band frequencies only promotes perception and has no significant effect on executive functions (Klink et al., 2020). The β frequency band is widely believed to be closely related to most cognitive functions, and 20 Hz stimulation of the functional cortex could have a positive influence on executive functions (Emmons et al., 2019; Yaple et al., 2017). However, the cognitive enhancement effects of tACS with this frequency have not been fully explored. Therefore, the effects of these two tES paradigms on executive function enhancement are inconsistent, which leads to uncertainty regarding the optimal stimulus mode and decreases the robustness of the behavioral quantification of the intervention effect. In addition, the lack of horizontal comparison between tACS and tDCS—both of which are potential clinical therapies for cognitive impairments—also poses a challenge for the formulation of therapeutic programs. Therefore, in the current study, Experiment 1 aimed to clarify the differences in the effects of the two tES paradigms on executive function performance to determine the most effective stimulus type.

The effects of current stimulation on the cortex mainly depend on the electrode montage in tES (Wörsching et al., 2018). Thus, choosing the most efficient brain area to be modulated by tES is crucial for obtaining beneficial behavior results and neural activity. Previous studies have found that damage to the right inferior frontal gyrus (IFG) leads to increased switching costs (low cognitive flexibility), ultimately impairing executive function performance (Aron et al., 2004a). Studies using functional magnetic resonance imaging (fMRI) have proven that the right IFG is activated during the process of inhibition control (Omata et al., 2018; Simmonds et al., 2008). Furthermore, the activation state of the right IFG could be positively changed in the context of working memory (Breitling et al., 2020). A drug study emphasized the importance of the right IFG in executive functions; the right IFG, as a seed point, was associated with other brain regions by enhanced functional connectivity accompanied by high behavioral performance (Borchert et al., 2016). Therefore, activity in the right IFG appears to be a potentially important neural mechanism underlying executive functions. Here, the right IFG was selected as the target brain region for stimulation with the aim of comparing the promoting effects of tACS and tDCS on executive functions. However, simple cognitive assessment is susceptible to the influence of ceiling or floor effects, which may affect the validity of the measurement. The prefrontal cortex (PFC) is a crucial brain region involved in many human core cognitions, including executive functions (Funahashi, 2001; Koechlin, 2003). Therefore, we planned to monitor the activity of brain regions involved in executive functions at rest during the interventions. Functional near‐infrared spectroscopy (fNIRS) is a noninvasive optical imaging technique that relies on the difference in the absorption of NIR light between oxyhemoglobin (HbO) and deoxyhemoglobin (HbR) (Scholkmann et al., 2014). This brain imaging technique can quantify the concentration changes of HbO and HbR by variations in the light intensity measured by emitting continuous‐wave light through the skull into the brain (Pinti et al., 2020a). fNIRS has a higher temporal resolution and a lower spatial resolution than fMRI but a lower temporal resolution and a higher spatial resolution than electroencephalography (EEG). Moreover, fNIRS has better motion tolerance than both fMRI and EEG. The continuous recording of hemodynamic variation through fNIRS has been considered an advantageous research model to measure brain activity during tES modulation (Di Rosa et al., 2019). In Experiment 2, fNIRS was used to continuously detect the activity of functional brain regions in the resting state at different stimulation phases (prestimulation, stimulation, and poststimulation). These imaging data could help quantify the benefits of stimulation and explain the neural variation features of cortices related to executive functions in the resting state more accurately.

In this study, we aimed to separately use tACS and tDCS to modulate the right IFG to improve executive functions; we further aimed to obtain a more precise assessment of the benefits by combining tACS or tDCS with fNIRS to detect brain activity at rest during the intervention. To deepen our understanding of the potential of different noninvasive electrical stimulation methods to improve executive functions, we compared the effects and mechanisms of two tES paradigms by evaluating cognitive performance and neural activity at rest.

2. MATERIALS AND METHODS

2.1. Participants and ethics

A total of 229 healthy male adults (mean + SD = 22.17 ± 2.52 years) were recruited for this study, with 105 participants in Experiment 1 and 124 participants in Experiment 2. For each experiment, participants were randomly assigned to the tACS group, the tDCS group or the sham group. All participants were right‐handed (Oldfield, 1971) and had normal or corrected‐to‐normal vision and hearing. No participants had a history of neurological or psychiatric disorders or head injuries, and no participants took antipsychotic, hypnotic, or antiseizure medications that could influence neural activity. This study was approved by the Ethics Committee of Xijing Hospital (Ethics Committee approval number: KY20202063‐F‐2). The study was performed in accordance with the ethical standards established in the 1964 Declaration of Helsinki and its later amendments or comparable ethical standards. All participants provided informed consent after the experimental procedures had been fully explained and acknowledged their right to withdraw from the study at any time. The sample size needed was estimated by G*Power 3.1.9.6 as at least 54 participants in each experiment given the following parameters: medium effect size (0.25), 1 − β = 0.90, and α = 0.05 (Cohen, 1992; Faul et al., 2007).

2.2. Cognitive task

The score of the N‐back task is considered a robust indicator to represent the level of working memory (Foraster et al., 2022). The color‐word Stroop (CW‐Stroop) task was used to assess inhibition ability by requiring individuals to recognize the color of words with conflicts of semantics and color (Schulte et al., 2012). Cognitive flexibility was measured by a shifting number task (Monsell, 2003). In the 3‐back task (Figure 1a), the screen displayed a 3 by 3 grid, with one square of the grid shaded in blue. Each time the grid flashed, the position of the blue square changed, and the subjects were asked to remember the current position of the blue square. Starting from the fourth flash, the participants were instructed to determine whether the position of the current blue square was the same as that of the blue square in the first flash. If so, they were asked to press the “J” key; if not, they pressed the “F” key. This task consisted of 9 practice trials followed by 63 formal trials. For the CW‐Stroop task (Figure 1b), a symbol (“XX”) or Chinese character (“红” “黄” “蓝” “绿”) in a random color (red, yellow, blue, green) appeared in the middle of the screen. The “D,” “F,” “J,” or “K” keys were pressed when the color of the symbol or Chinese character was red, yellow, blue, or green. The three conditions (neutral, congruent and incongruent) involved in this task randomly occurred at a ratio of 1:1:1, and there was a total of 20 practice trials followed by 108 formal trials for each condition. For the shifting number task (Figure 1c), a number from “0” through “9” (except “5”) appeared in the center of the screen in colored text (red or blue). When the number presented was red, the participant was asked to judge the parity of the number, pressing the “F” key if it was odd or the “J” key if it was even. When the number presented was blue, they compared the size of the number with 5. If the number was smaller than 5, they pressed the “F” key. If the number was larger than 5, they pressed the “J” key. The ratio of odd numbers, even numbers, numbers larger than “5” and numbers smaller than “5” was 1:1:1:1. There were 8 practice trials followed by 80 formal trials. The sequence of these three tasks was counterbalanced.

FIGURE 1.

The task designs assessing three core components of executive functions. (a) 3‐back task; (b) CW‐Stroop task; (c) shifting number task.

2.3. tES protocol

A Soterix Medical MXN‐33 High‐Definition Transcranial Electrical Stimulator (Soterix Medical, Inc., New York) was used to modulate the tES in the present study. Four returned electrodes and one central electrode were used to stimulate the target cortical areas of each participant, and each electrode had a diameter of 12 mm. According to the international 10‐10 EEG montage, the central electrode was placed at FC6, as in previous studies (Holland et al., 2011), and the four return electrodes were placed at F6, FT8, C6, and FC4 to cover the right IFG (Figure 2b). Additionally, we used HD‐explore software (Soterix Medical, Inc., New York) to simulate the theoretical electric field and current intensity to ensure the efficiency of the electrode montage. For the tACS group, a 1.5‐mA current (according to the tolerance limits of participants) with a frequency in the β band (20 Hz) was applied for 10 min in all participants. For the tDCS group, the current intensity and duration were the same as those in the tACS group; however, the central electrode was anodal, and another four cathodal return electrodes were included. The sham group received sham stimulation at 1.5 mA for 60 s, comprising 30 s for ramping up to 1.5 mA and then 30 s for ramping down. To ensure blinding efficacy, two colleagues who were not involved in setting up the tES parameters administered the intervention under a blinded operation schema, in which they only saw code names for each of the three tES paradigms, and the detailed parameters were not displayed on the operation interface. Each participant was asked to complete a blinding efficacy questionnaire, indicating whether they thought that they received active or sham stimulation and how confident they were in this judgment (on a scale from 1 = not confident to 3 = very confident). Additionally, participants were asked to complete a side effects questionnaire to assess whether they experienced any abnormal sensations (e.g., itching) potentially related to stimulation.

FIGURE 2.

(a) The arrangement of optodes and electrodes. Four channels and five current electrodes around FC6 were placed to modulate or measure activity in the right IFG, and the other 10 channels were placed adjacent to the Fpz to measure activity in the frontal pole region (pictured above); an actual photograph is also shown (below). (b) The details of the electrode layout are shown, and the theoretical current intensity on the right IFG is predicted with Soterix HD‐Explore software. (c) The designs of Experiment 1 and Experiment 2. Each experiment was divided into three stages (prestimulation, stimulation, and poststimulation).

2.4. Raw fNIRS data acquisition

We used a 14‐channel LABNIRS fNIRS system (Shimadzu Co., Kyoto, Japan) emitting light at 780, 805, and 830 nm to record the concentration changes of oxygenated hemoglobin (oxy‐Hb) and deoxygenated hemoglobin (deoxy‐Hb) and to simultaneously assess neural imaging variations in each participant (sample rate of 27.78 Hz). According to the international 10‐10 EEG montage, Fpz and FC6 were used as the frontal pole region and right IFG, respectively. For each participant, a 2 × 2 optode patch (4 channels with 2 emitters and 2 detectors) with FC6 set as the reference point was placed on the right side of the head to record changes in the right IFG, and a 4 × 2 optode patch (8 channels with 4 emitters and 4 detectors) with Fpz as the reference point was symmetrically placed on the forehead to record changes in the frontal pole region (Lu et al., 2020), as shown in Figure 2a. The distance between the source and the adjacent detector was ~30 mm. Notably, the optode patch on the right IFG was designed to match the tES electrode patch, in which the anodal electrode (FC6) was surrounded by four optodes. To ensure the quality of the acquired brain signal and prevent weak electrode contact caused by movement of the optode in Experiment 2, a cap for multibrain stimulation combined with real‐time signal acquisition was designed and used in this study. We used nondeforming plastic tape to hold the optode patch in place and maintain the relative distance between the adjacent optodes or electrodes. The participants were instructed to open their eyes as much as possible during the rest state and note “+” in the center of the screen, and they were asked to relax and remain motionless as much as possible.

2.5. Procedure

This study consisted of two experiments (Experiment 1 and Experiment 2). Experiment 1 focused on the promotion effect of tES on the behavior of the three core components of executive function. Three behavioral tasks (CW‐Stroop, 3‐back, and switch number tasks) were used to evaluate the cognitive variation in the prestimulation and poststimulation phases. According to the behavioral results of Experiment 1, active tES modulation produced a behavioral benefit without a significant difference between tACS and tDCS. Thus, fNIRS was used in Experiment 2 to continuously monitor the hemodynamic variation in the resting state in the three intervention phases to explore the different influences of tACS and tDCS on brain function. Three phases (prestimulation, stimulation, and poststimulation) were included in each of the two experiments. All the details can be found in Figure 2c.

2.6. Data analysis

2.6.1. Behavioral data

To ensure the quality of participant responses, trials with reaction times ≤150 ms or ≥1500 ms were excluded. The reaction time and accuracy rate of the 3‐back task were used as the evaluation index of working memory. The Stroop effect in the CW‐Stroop task was calculated as the reaction time (correct answers) of the incongruent condition minus the reaction time (correct answers) of the congruent condition and used to evaluate inhibition. Switching cost in the shifting number task was calculated as the reaction time of switching trials minus the reaction time of repeated trials and was used to evaluate cognitive flexibility; lower switching costs indicated higher cognitive flexibility.

2.6.2. fNIRS data

Data format conversion

LABNIRS collection and analysis software was used to convert the original “omm” files into “txt” format without performing any data preprocessing. Then, the format conversion function (shimaduz2nirs) in the Homer_2 toolbox was applied to convert the “txt” data to the “nirs” format supported by Homer_2. Then, the coefficient of variation (CV) method was employed to eliminate the channels with poor signal quality, and the threshold for the CV was set to 7.5% (Hocke et al., 2018).

Data preprocessing

First, the original optical intensity was converted to the optical density (OD) by the hmrIntensity2OD function. The enPCA function (nSV = 0.8) was used to reduce physiological noise (Hu et al., 2021). Subsequently, the motion artifact reduction algorithm (MARA) was utilized. MARA is a correction method that detects artifacts by evaluating whether the amplitude variation or standard deviation within short time windows exceeds user‐specified thresholds. Time windows containing artifacts were widened, and then artifacts were approximated with smoothing splines and subtracted to obtain motion‐corrected signals. Then, the hmrMotionArtifactByChannel function was applied to identify motion artifacts, and the MotionCorrectSpline function was applied to correct motion artifacts (tMotion = 0.5, tMask = 2, STDEVthresh = 20, AMPthresh = 0.5, pSpline = 0.99) (Cooper et al., 2012; Reindl et al., 2022). Then, the OD was converted to oxy‐Hb and deoxy‐Hb concentration changes by the hmrOD2Conc function (ppf = 6.0 6.0 6.0) based on the modified Beer–Lambert law (Baker et al., 2014). Previous studies have demonstrated that oxy‐Hb concentration changes are more sensitive than deoxy‐Hb concentration changes in reflecting changes in regional cerebral blood flow (Hoshi, 2007; Sato et al., 2013); thus, oxy‐Hb concentration change data were used as the main indicator in further analyses with deoxy‐Hb concentration change data as an auxiliary indicator added to Data S1. Notably, the hmrBandpassFilt function (hpf = 0.01, lpf = 0.1 Hz) was applied to eliminate physiological noise and baseline drift during functional connectivity processing. To ensure the consistency of data processing, the same length of data was used for functional connectivity analysis, that is, 3 min for prestimulation, stimulation (3 min in the middle of the 10 min stimulation phase) and poststimulation. Finally, Pearson correlation analysis was performed on the oxy‐Hb concentration change data of the 14 channels at the three stimulation phases in each group, and Fisher's z‐transformed Pearson correlation coefficient between channels was indicated as an evaluation index of functional connection strength.

2.6.3. Statistical analysis

Most indices were analyzed by mixed‐design analysis of variance (ANOVA) (the between‐subjects factor was group: tDCS, tACS and sham; the within‐subjects factor was time: phases of stimulation) or one‐way ANOVA (the factor was group: tDCS, tACS and sham). The significance threshold was set to P < .05. Significant interaction effects were further explored with simple effect analysis. Multiple comparisons were adjusted with the Bonferroni correction. In addition, effect sizes are reported as partial eta‐squared (η _p ²) values. All analyses were performed using IBM SPSS software v25.0.

3. RESULTS

3.1. Experiment 1: Improvement in executive function behaviors by tACS/tDCS

3.1.1. Characteristics, blinding efficacy, and side effects of participation

As shown in Table 1, there were no differences in age or educational level among the three groups at baseline. There was no significant difference among the groups in the number of guesses (χ ² = 1.02, P = .907). In addition, the confidence in guesses did not significantly differ (P = .812) among the three groups according to a Kruskal–Wallis test. All participants had good tolerance for the discomfort of receiving tES, and the side effects reported were slight and transient. All details can be found in Table 1.

TABLE 1.

Basic characteristics, blinding efficacy, and self‐reported side effects of participants in Experiment 1.

	tACS (n = 35)	tDCS (n = 35)	Sham (n = 35)	P value
	Mean ± SD	Mean ± SD	Mean ± SD
Age (years)	22.20 ± 2.01	22.91 ± 3.04	21.80 ± 1.71	.133
Education (years)	16.29 ± 1.96	16.86 ± 2.23	15.83 ± 1.79	.103
Number of guesses (active)	34 (97.14%)	33 (94.29%)	33 (94.29%)
Confidence in guesses (active)	2.97 ± 0.17	2.97 ± 0.17	2.97 ± 0.17
Number of guesses (sham)	1 (2.86%)	1 (2.86%)	1 (2.86%)
Confidence in guesses (sham)	3.00	2.00	2.00
Number of guesses (uncertain)	0	1 (2.86%)	1 (2.86%)
Confidence in guesses (uncertain)	0	3.00	3.00
Itching	3 (8.57%)	4 (11.43%)	5 (14.29%)
Skin redness	0	1 (2.86%)	0
Headache	1 (2.86%)	1 (2.86%)	0
Cervical pain	0	0	0
Distraction	1 (2.86%)	2 (5.71%)	2 (5.71%)
Acute mood variation	0	0	0
Sleepiness	1 (2.86%)	2 (5.71%)	1 (2.86%)

Note: The P values were calculated with one‐way ANOVA; P < .05 was considered significant.

3.1.2. Performance variation of three core components in executive functions by tACS/tDCS

The 3 (group: tACS, tDCS and sham) × 2 (prestimulation and poststimulation) mixed‐design ANOVAs were separately conducted in cognitive tasks. As shown in Figure 3a,b, the results of the 3‐back task indicated that the main effects of time were significant for both reaction time (F _(1,102) = 125.24, P < .001, η ² _p  = 0.551) and accuracy rate (F _(1,102) = 45.85, P < .001, η ² _p  = 0.310), but the interaction effects of reaction time and accuracy rate were not significant (Ps > .05). A mixed‐design ANOVA showed a significant interaction effect of the Stroop effect (F _(2,102) = 4.144, P = .019, η ² _p  = .075). Further simple effect analysis showed that the Stroop effect was significantly decreased after tES intervention (Ps < .01) (Figure 3c). The variation in the switching cost of the shifting number task had a significant interaction effect between time and group (F _(2,102) = 3.148, P = .047, η ² _p  = .058). Further simple effect analysis showed that the switching cost was significantly reduced in both intervention groups (Ps < .001) (Figure 3d). Behavioral results did not provide evidence that distinguished the effects of the two interventions, so the cortical activity data in Experiment 2 required further analysis.

FIGURE 3.

Behavioral performance in the 3‐back task, CW‐Stroop task, and shifting number task. (a,b) Variations in reaction time and accuracy rate for the 3‐back task. (c) Stroop effect in the CW‐Stroop task. (d) Shift cost in the shifting number task. Boxes extend from the 25th to 75th percentiles, with the horizontal line representing the median. Whiskers show the min and max values. Colored dots represent individual data points. Bonferroni‐corrected contrast: *P < .05, ***P < .001.

3.2. Experiment 2: Positive effect on hemodynamic activity of the PFC by tACS/tDCS

3.2.1. Characteristics, blinding efficacy, and side effects of participation

As shown in Table 2, there were no differences in age or educational level among the three groups at baseline. There was also no significant difference among the groups in the number of guesses (χ ² = 3.85, P = .427). In addition, the confidence in guesses did not significantly differ (P = .808) among the three groups according to a Kruskal–Wallis test. All participants had good tolerance for the discomfort of receiving tES, and the side effects reported were slight and transient. All details can be found in Table 2.

TABLE 2.

Basic characteristics, blinding efficacy, and self‐reported side effects of participants in Experiment 2.

	tACS (n = 42)	tDCS (n = 42)	Sham (n = 40)	P value
	Mean ± SD	Mean ± SD	Mean ± SD
Age (years)	22.71 ± 3.10	21.69 ± 2.28	21.78 ± 2.47	.149
Education (years)	16.86 ± 2.16	15.98 ± 2.16	16.25 ± 1.98	.150
Number of guesses (active)	40 (95.24%)	41 (97.62%)	36 (90.00%)
Confidence in guesses (active)	2.85 ± 0.36	2.76 ± 0.54	2.86 ± 0.42
Number of guesses (sham)	1 (2.38%)	1 (2.38%)	1 (2.50%)
Confidence in guesses (sham)	3	2	2
Number of guesses (uncertain)	1 (2.38%)	0 (0.00%)	3 (7.50%)
Confidence in guesses (uncertain)	2	0	2
Itching	8 (19.05%)	6 (14.29%)	6 (15.00%)
Skin redness	0	0	0
Headache	1 (2.38%)	1 (2.38%)	1 (2.50%)
Cervical pain	0	0	1 (2.50%)
Distraction	1 (2.38%)	1 (2.38%)	1 (2.50%)
Acute mood variation	0	0	0
Sleepiness	2 (4.76%)	3 (7.14%)	3 (7.50%)

Note: The P values were calculated with one‐way ANOVA; P < .05 was considered significant.

3.2.2. Effect on cortical functional connectivity in the resting state by tACS/tDCS

The Pearson correlation coefficient was used as the index to plot the functional connectivity strength of all channels in the three groups at three phases (Figure 4a). The F values of time, group main effect, and interaction effect matrix derived from the 3 (group: tACS, tDCS and sham) × 3 (time: prestimulation, stimulation and post stimulation) mixed‐design ANOVA were plotted, with P < .05 (Figure 4b). The pairs of channels with significant interaction effects were channels 3 and 10 as well as channels 7 and 13. Further simple effect analysis showed that the functional connectivity strength of channels 3–10 and channels 7–13 significantly decreased during stimulation compared with that in the prestimulation phase in the tDCS group, while the other two groups did not show similar changes in these two pairs of channels (Figure 5a,c). Especially in the stimulation phase, the functional connectivity of channels 3–10 and channels 7–13 during tDCS intervention was significantly lower than that in the tACS group and the sham group (Figure 5b,d), which might indicate that the stimulation phase is the key phase in which the neural enhancement characteristics of the two tES paradigms in the resting state can be distinguished.

FIGURE 4.

The functional connectivity between pairs of channels and their statistical significance according to mixed‐design ANOVAs. (a) Mean correlation matrices of fNIRS signals for the three groups during the three phases. (b) The main effects of group and time and the interaction effect between group and time. The white arrows indicate the connectivity of channels 3–10, channels 7–13, which both showed significant interaction effects.

FIGURE 5.

The variation in the functional connectivity of channels 3–10 (a,b) and channels 7–13 (c,d) among the three groups from the prestimulation phase to the poststimulation phase. Boxes extend from the 25th to 75th percentiles with the horizontal line representing the median, and whiskers show the min and max values. Colored dots indicate individual data points. All values are presented as the mean ± SEM in broken line graphs. Bonferroni‐corrected contrast: *P < .05, **P < .01, ***P < .001.

4. DISCUSSION

As the basis for normal psychological function and individual success, executive functions play an important role in social interaction, work, and education. tES is an effective method to improve the cognitive ability of healthy people or patients with cognitive impairments, but there is a lack of comparative studies on the improvement of executive functions by different tES paradigms. To our knowledge, this is the first study that assesses and demonstrates the enhancement of executive functions by using two kinds of tES paradigms at the right IFG. We found that the behavioral performance related to executive functions was significantly improved after both tACS and tDCS. However, only the significant decrease in neural activity (improved neural efficiency) in the right IFG and frontal pole region during the stimulation phases was affected by the stimulation type. The details of the results are discussed below.

Both tES paradigms positively affected executive function performance in the present study. tDCS or tACS can benefit inhibition in healthy people or individuals with attention‐deficit/hyperactivity disorder (Klírová et al., 2021; Schroeder et al., 2020). Previous studies have found that tDCS can effectively improve individuals' cognitive flexibility by increasing dopamine secretion (Borwick et al., 2020). A study on the decline in cognitive flexibility caused by substance use disorders found that tACS could positively affect behavioral deficits (West et al., 2021). The results of this study were consistent with these previous studies in that a single tDCS or tACS session could improve the level of cognitive flexibility and inhibition of individuals. However, this study was also different from previous studies in that the stimulation site was focused on the right IFG, which is considered one of the important brain areas in executive functions (Aron et al., 2004b). Our results emphasized the importance of the right IFG as a targeted brain region in noninvasive interventions to improve executive functions. However, when compared with the sham group, there were no significant improvements in working memory after a single tDCS or tACS session, indicating that enhancements in working memory performance were observed among all three groups. This may be related to the ceiling effect caused by the high cognitive level of the subjects and the relatively simple 3‐back task used in the present study. A similar phenomenon was also found in our previous study under the condition of repeated stimulation (Lu et al., 2021). More importantly, several studies have demonstrated that tES combined with behavioral training can promote better working memory performance, especially for subjects with high cognitive levels, and tES could magnify the effect of behavioral training on working memory (Ke et al., 2019; Pergher et al., 2022). Notably, the 20 Hz tACS of the right IFG in this study was proven to be effective in promoting executive functions. A study that aimed to increase α oscillations in functional brain areas to promote speech decision making found that α‐band tACS did not have a significant effect on executive function decision performance (Werchowski et al., 2022). Although the improvement in nerve entrainment on executive functions caused by β‐band tACS is controversial, it has been proven that β‐band tACS can have a positive effect on performance in the signal‐stop task, while γ‐band tACS leads to a negative effect (Leunissen et al., 2022). Therefore, this study provides evidence for the positive effects of β‐band tACS on executive functions. The results of Experiment 1 demonstrated the effectiveness of the two tES paradigms in improving executive control function, but the behavioral effects of tACS/tDCS cannot explain the potential difference in enhancement benefits. Therefore, the stimulation effects of the two intervention techniques on brain function can be compared from the perspective of the underlying neural mechanisms by evaluating the imaging data of Experiment 2.

Separate enhancements of synaptic plasticity and oscillations in tES via endogenous voltage‐dependent Hebbian plasticity or nerve entrainment are thought to be important mechanisms that promote behavior (Pariz, 2023; Korai, 2021), and variation in hemodynamics reflects the degree of functional cortical activity according to the theory of neurovascular coupling (Pinti et al., 2020b). In the present study, fNIRS was used to measure the signal of the cortex during electrical stimulation without current interference due to its optical basis, and the activity on the right IFG and the frontal pole region in the resting state was continuously observed by fNIRS during the entire tES modulation process. Here, it should be noted that the basis for maintaining normal cognitive function is integrity of the collaborative interaction between near and far neurons across different brain regions, to which nerve fibers (structural or material connections) and rhythmic activations (functional or statistical connections) contribute (Edison, 2020; Sporns, 2022). Therefore, it is not possible to establish a causal relationship between activity changes in stimulation‐targeted brain regions and variations in cognitive performance. In other words, although the right IFG was stimulated in this study, the frontal pole region was observed since the PFC, which involves the IFG, is considered the crucial brain area for the development of most cognitive functions of human beings (Funahashi, 2001; Koechlin, 2003). We found that the strength of functional connectivity between the right IFG and the frontal pole region and within the frontal pole region significantly decreased after tDCS but not in the other groups. The reduction in functional connectivity strength might indicate an improvement in connectivity efficiency; that is, communication between neurons could be maintained with less connectivity strength to obtain the same or an even higher level of cognitive performance (Lu et al., 2022). Previous studies have also found that active tDCS can promote language learning by reducing functional connectivity between the left and right brain regions (Fiori et al., 2018). However, in this study, tACS increased the functional connectivity strength between the right IFG and the frontal pole region after stimulation, which may support the fact that the neural oscillation induced by β‐band tACS did not effectively promote neural efficiency. In addition, another study previously showed that rhythmic noninvasive stimulation (such as tACS) is more dependent on cortex state changes (Kasten & Herrmann, 2022). Therefore, a possible reason for the reduction in connectivity after stimulation observed in this study is that the brain state changes from the task state to the resting state after stimulation withdrawal, which leads to the weakening of the effect of tACS modulation. Moreover, the different states of the task would change the phase of endogenous oscillations, which can induce unstable phase differences and reduce the benefits of tACS on brain activity (Haslacher et al., 2023).

There are several limitations in the current study. Although we recruited many participants and there were no significant differences in the demographic information of both samples (Table S3), it was still difficult to fully interpret the results through the nonidentical and male‐only samples of the two experiments. In addition, only the assessment of brain function in the resting state was used as the indicator, not the evaluation of the brain function in the task state; some neural information related to the task would be lacking, which should be explored in future studies. tACS showed a similar improvement in the performance of executive function to that of tDCS, but it did not show effective enhancement of neural efficiency. This is related to the fact that tACS can elicit internal neural oscillations rather than direct neuronal excitability changes to influence cognitive performance. Future studies might be needed to evaluate the amplitude and latency of specific frequency band oscillations during the stimulation process of tACS by using EEG, transcranial magnetic stimulation (TMS) and other tools. The tasks used in this study were all classical tests with low difficulty for subjects with high cognitive abilities, which resulted in the lack of evaluation validity for executive functions. Therefore, there were no significant differences between tACS and tDCS in the elicited performance changes for the three core elements of executive functions. Due to the limited number of fNIRS channels in the study, the channels were set around the right IFG but did not fully pass through this brain area, and it is also impossible to measure all brain regions related to executive functions. Although principal component analysis (PCA) is used in fNIRS data processing, the lack of short channels makes it difficult to minimize physiological noise contamination, which could possibly impact the results. Future studies may require more comprehensive and pure neuroimaging evidence and monitoring a larger range of brain regions.

In summary, the present study demonstrated the behavioral improvements of tDCS/tACS on inhibition and cognitive flexibility in the core components of executive functions in healthy adults while exploring the potential neural mechanism at rest during tACS/tDCS modulation. The stimulation phase is proposed as a key state for distinguishing the benefits of two tES paradigms. tDCS can promote the neural efficiency of the right IFG and frontal pole region more effectively than tACS, and neural variation can be regarded as an effective indicator to quantify the intervention benefits of these two tES paradigms. Although neurological or psychiatric patients were not involved in the present study, these findings also suggest a new perspective of a potential clinical diagnostic and treatment method, which contributes to the improvement of executive function deficits in patients with attention‐deficit/hyperactivity disorder, obsessive‐compulsive disorder or Alzheimer's disease.

AUTHOR CONTRIBUTIONS

Hongliang Lu, Yajuan Zhang, Danmin Miao and Xia Zhu designed the research; Hongliang Lu, Xuanyi Tan and Zhilong Zhang performed the research; Hongliang Lu, Huake Qiu and Peng Huang analyzed the data; Hongliang Lu and Yajuan Zhang wrote the article; Danmin Miao and Xia Zhu provided editorial support for this article.

FUNDING INFORMATION

This work was supported by the Major Project of Medicine Science and Technology of PLA (AWS17J012), Major Project of AFMU (2022ZZXM017), Air Force Command Project (KJZL2022‐2) and The General Project of Key R&D Plan of Shaanxi Province‐Social Development Field (2022SF‐023).

CONFLICT OF INTEREST STATEMENT

The authors declare no competing interests.

Supporting information

Data S1. Supporting Information.

Lu, H. , Zhang, Y. , Qiu, H. , Zhang, Z. , Tan, X. , Huang, P. , Zhang, M. , Miao, D. , & Zhu, X. (2024). A new perspective for evaluating the efficacy of tACS and tDCS in improving executive functions: A combined tES and fNIRS study. Human Brain Mapping, 45(1), e26559. 10.1002/hbm.26559

DATA AVAILABILITY STATEMENT

The data that support the findings of this study are available from the corresponding author upon reasonable request.