INTRODUCTION
Mild cognitive impairment (MCI) is a condition characterized by a noticeable cognitive decline that goes beyond what is expected for normal aging but does not meet the criteria for dementia. It is often seen as an intermediate stage between normal cognitive aging and more severe forms of cognitive impairment, such as Alzheimer’s disease (AD) [1]. The prevalence of MCI is estimated to range from 10% to 20% among individuals aged 65 years and older, and individuals with MCI are at an increased risk of progressing to dementia [2]. Dementia, particularly AD, is a growing global health concern that affects millions of people worldwide and imposes significant social and economic burdens on patients, caregivers, and healthcare systems [3]. Current treatments for MCI, including pharmacological interventions, provide limited benefits in preventing progression to dementia and often require complementary therapeutic approaches to manage cognitive decline effectively. This highlights the need for additional interventions such as cognitive training, physical exercise, and neuromodulation techniques to attenuate cognitive impairment.
Transcranial alternating current stimulation (tACS) is a noninvasive brain stimulation technique that has shown promise in modulating neural activity and enhancing cognitive function [4], garnering attention as a potential alternative therapeutic approach for patients with MCI or AD. tACS delivers weak electrical currents at specific frequencies to synchronize or modulate brain oscillations that are believed to play a crucial role in cognitive processes [5]. In particular, theta (4-7 Hz) and gamma (30-50 Hz) tACS have been shown to influence cognitive functions, including memory and attention [6]. This effect is hypothesized to result from the modulation of theta-gamma cross-frequency coupling by theta or gamma frequency tACS, which facilitates the dynamic interaction between distinct neural oscillatory activities critical for cognitive processing. Given this underlying mechanism, the modulation of theta-gamma coupling by tACS has the potential to improve cognitive function and memory, particularly because patients with MCI or AD are known to exhibit abnormal brainwave activity compared to the general population [7]. This suggests that tACS could serve as a promising therapeutic alternative for addressing these abnormalities and enhancing the cognitive function in these patients.
Recent studies have increasingly applied tACS to MCI or early AD to examine its effects on cognitive function and memory, with many of these studies primarily focusing on gamma-frequency tACS [8]. Preclinical research by Iaccarino et al. [9] demonstrated that applying 40 Hz light flickering to an AD mouse model reduced amyloid-beta peptides, suggesting the potential applicability of gamma-frequency stimulation for MCI and AD. Subsequent clinical studies have shown that applying 40 Hz gamma tACS to patients with MCI and AD can reduce dementia-related markers, such as tau burden [10,11] and amyloid-beta [10]. Additionally, improvements have been observed in various dementia-related clinical measures, including episodic memory, gamma spectral power [12], verbal learning and recall, cholinergic transmission [13], and overall cognitive function [14]. Collectively, these findings indicated the potential of gamma tACS as a therapeutic intervention for cognitive decline in patients with MCI or AD.
Despite the growing interest in gamma tACS for MCI and AD, theta frequency tACS has not received comparable attention. As mentioned earlier, improvements in cognitive function and memory observed with tACS are believed to result from its influence on theta-gamma cross-frequency coupling. Interestingly, several studies involving healthy control participants without cognitive impairment reported significant effects of theta tACS on cognitive function. A recent meta-analysis by Lee et al. [6] suggested that the cognitive enhancement effects of tACS are more critical for theta frequency than for gamma frequency in healthy populations. This highlights the necessity of examining the effects of theta tACS in patients with MCI or AD despite the intuitive challenges posed by patients with MCI, who tend to have a higher proportion of slow-wave activity and a lower proportion of fast-wave activity than healthy individuals [7]. Nevertheless, the cognitive benefits of theta tACS observed in healthy individuals, along with its proposed role in theta-gamma cross-frequency coupling relevant to working memory and executive function, suggest that theta tACS could plausibly exert beneficial effects even in MCI populations. Comparing its efficacy with gamma tACS could thus clarify frequency-specific therapeutic potential in this clinical group.
This study aimed to directly compare the effects of thetaand gamma-frequency tACS on the cognitive function of patients with MCI. Specifically, we will assess the impact of a single 30-minute session of theta or gamma tACS on attention, working memory, and executive function to explore which frequency shows more potential for cognitive enhancement. This study focused on the immediate effects of these stimulations to determine whether there was a differential response between theta and gamma tACS, providing initial evidence that could guide future interventions to optimize cognitive outcomes in patients with MCI.
METHODS
Ethics statement
All methods were conducted per relevant guidelines and regulations. This study was conducted in psychiatric units at Chuncheon Sacred Heart Hospital, a teaching hospital affiliated with College of Medicine, Hallym University, Republic of Korea. The study was registered in the Clinical Trials Registry of Korea (registration: KCT0010158/date: 25/01/2025). This study was approved by the Institutional Review Board of Chuncheon Sacred Heart Hospital, Republic of Korea (approval no. 2023-12-011). All the participants provided written informed consent.
Participants
Clinically diagnosed patients with MCI aged >55 years were recruited through advertisements at local hospitals with psychiatric units and community mental health centers. To be included in the study, participants had to meet the following criteria: 1) fulfill the DSM-5 criteria for mild neurocognitive disorder [15]; 2) meet Petersen’s criteria for MCI [16], which include memory complaints, normal activities of daily living, intact general cognitive function, abnormal memory performance for age, and no diagnosis of dementia; and 3) have a Clinical Dementia Rating (CDR) score between 0.5 and 2 [17] or a Global Deterioration Scale (GDS) score between 2 and 5 [18]. To ensure inclusiveness at the initial screening stage while minimizing the risk of excluding clinically borderline MCI cases, we employed a slightly extended range for CDR (0.5-2) and GDS (GDS 2-5).
Participants were excluded if they met any of the following conditions: 1) current use of cognitive enhancers (e.g., donepezil, rivastigmine, galantamine, memantine); 2) history of neurological conditions such as head trauma, epilepsy, or Parkinson’s disease; 3) history of psychiatric disorders as per DSM-5 criteria, including bipolar disorder, substance use disorder, alcohol use disorder, or personality disorders; 4) clinically significant medical conditions affecting brain function, such as cardiovascular or endocrine diseases, or severe asthma; 5) sensory impairments, including olfactory or visual deficits (e.g., color blindness); 6) inability to read Korean; and 7) participation in other clinical studies within the past month. The use of psychiatric or general medications, except cognitive enhancers, was permitted during the study.
Diagnostic interviews were conducted face-to-face by board-certified clinical psychologists and psychiatrists who reviewed each participant’s history, symptoms, and psychosocial functioning using all available information.
Sample size estimation was based on a previous study reporting a large effect size (Cohen’s f=0.983) for gamma-frequency tACS on Trail-Making Test (TMT) performance in MCI [19]. A priori power analysis with G*Power 3.1. (α=0.05, power=0.80, 3 groups, 2 time points; Heinrich Heine University) indicated a minimum of 9 participants. However, Considering the likelihood of attrition and the exploratory nature of the pilot study, we aimed to recruit at least 15 participants per group, substantially exceeding the minimum requirement, thereby ensuring sufficient statistical power for exploratory analysis. Of the 60 individuals initially screened, 47 (16 males) right-handed participants aged 63-89 years met the inclusion criteria and were enrolled in the study.
Study procedure
This study used a randomized, sham-controlled, pre-post, double-blinded, single-intervention design. The participants were randomly assigned to one of the three groups: sham (n=15), theta tACS at 5 Hz (n=16), or gamma tACS at 40 Hz (n=16). Each participant performed a set of cognitive tasks immediately before and after the tACS intervention to evaluate cognitive changes (Figure 1).
The cognitive tasks included the Digit Span (Forward/Backward) and Trail-Making Test A (TMT-A). The Digit Span test measures attention, short-term memory, and working memory, in which participants are asked to repeat a sequence of numbers as presented (forward) and in reverse order (backward) with increasing sequence length until incorrect recall [20]. The TMT-A assesses working memory, visual attention and task-switching abilities, requiring participants to connect 25 numbered circles as quickly and accurately as possible [21]. Performance was evaluated based on the time needed to complete the task and the number of errors made, with shorter times and fewer errors indicating better cognitive function.
Intervention protocol
The tACS intervention lasted for 30 min, during which the participants were seated and relaxed. The stimulation device was the MIND-D (YBRAIN, Republic of Korea; http://www.ybrain.com/). Identical electrodes (28.26 cm2; current density, 0.07 mA/cm2) were used for all conditions, with round-sponge patches soaked in 0.9% sodium chloride solution (approximately 7 mL of saline). The sham group received placebo stimulation, the theta group received 5 Hz stimulation, and the gamma group received 40 Hz stimulation. Stimulation was administered at an intensity of 2 mA for 30 min, targeting the dorsolateral prefrontal cortex (DLPFC), precisely positioned at F3 (left) and F4 (right) based on the extended international 10-20 system. The tolerability of the tACS intervention was assessed at the end of each session using a self-report measure, followed by a questionnaire by a research assistant who administered transcranial electrical stimulation (tES).
Statistical analysis
To assess the impact of the tACS intervention on cognitive function, repeated-measures analysis of variance (ANOVA) was conducted separately for each of the four outcome measures: Digit Span Forward, Digit Span Backward, TMT-A errors, and TMT-A completion time. Group (gamma tACS, theta tACS, sham) was included as a between-subjects factor, and time (pre-intervention, post-intervention) was included as a within-subjects factor. Post-hoc analyses were conducted for any statistically significant main effects or interactions observed in the ANOVA using Tukey’s correction for multiple comparisons. In addition, as supplementary analyses to assess group-level differences at each time point, independent-sample t-tests were conducted to compare cognitive scores between groups (gamma vs. sham, gamma vs. theta, theta vs. sham) separately at the pre-intervention and post-intervention stages. Statistical significance was set at p<0.05. All statistical analyses were conducted using the IBM SPSS Statistics 23 (IBM Corp.).
RESULTS
Demographic and clinical information
The demographic and clinical information of the participants are summarized in Table 1. The analysis revealed no significant differences among the sham, theta, and gamma groups regarding demographic characteristics or clinical conditions, indicating that the groups were comparable at baseline.
Digit Span
Separate repeated-measures ANOVA (RMANOVA) were conducted for the forward and backward digit spans (Table 2). No significant main effects or interactions were observed for the Digit Span Backward task, indicating no significant changes across groups. For the Digit Span Forward task, there was a significant main effect of time, F(1, 44)=11.79, p=0.001, indicating an overall improvement in performance following the intervention. Post-hoc analysis with Tukey’s correction revealed that the gamma tACS group showed a statistically significant increase in performance from pre- to post-intervention (pre=4.94, post=5.56), t(44)=2.996, p_Tukey=0.048.
To further examine group-level differences at each time point, additional between-group comparisons were performed using independent-sample t-tests. For Digit Span Forward, no significant differences were found between gamma and sham (t(29)=0.416, p=0.679), gamma and theta (t(29)=0.445, p=0.658), or theta and sham (t(30)=0.021, p=0.983) at the post-intervention stage. Similarly, at the pre-intervention stage, all comparisons were non-significant (gamma vs. sham: t(29)=1.613, p=0.114; gamma vs. theta: t(29)=0.644, p=0.523; theta vs. sham: t(30)=0.979, p=0.332).
For Digit Span Backward, the results were likewise nonsignificant at both the pre- (gamma vs. sham: t(29)=1.089, p=0.282; gamma vs. theta: t(29)=1.276, p=0.209; theta vs. sham: t(30)=0.204, p=0.839) and post-intervention stages (gamma vs. sham: t(29)=1.061, p=0.294; gamma vs. theta: t(29)=0.165, p=0.870; theta vs. sham: t(30)=1.224, p=0.228).
Trail-Making Test A
Separate RMANOVA analyses were conducted for TMT-A completion time and number of errors (Table 3). Five participants (sham=2; theta=2; gamma=1) were excluded from the TMT-A analysis due to performance failure during the pre-intervention assessment, defined as exceeding the 300-second time limit or committing five or more errors. For the TMT-A completion time, a significant main effect of time was found, F(1, 38)=4.61, p=0.038, indicating an overall reduction in completion time following the intervention. However, post-hoc analyses did not reveal any significant differences between the groups. For the number of errors in the TMT-A, there was a significant main effect of time, F(1, 38)=5.88, p=0.020, and a significant interaction between group and time (F(2, 38)=3.91, p=0.029). Post-hoc analysis with Tukey’s correction indicated that the gamma tACS group showed a statistically significant decrease in the number of errors from pre- to post-intervention (pre=1.133, post=0.200, t(38)=3.787, p_Tukey=0.007).
To further examine group-level differences at each time point, additional independent-sample t-tests were performed. For TMT-A completion time, no significant between-group differences were observed at either time point. At the pre-intervention stage, comparisons yielded the following: gamma vs. sham (t(26)=0.363, p=0.719), gamma vs. theta (t(26)=1.754, p=0.087), and theta vs. sham (t(26)=1.183, p=0.244). At the post-intervention stage, the differences remained nonsignificant (gamma vs. sham: t(26)=0.239, p=0.813; gamma vs. theta: t(26)=1.184, p=0.244; theta vs. sham: t(26)=0.913, p=0.367).
Similarly, for TMT-A error, no significant between-group differences were observed at either the pre- or post-intervention stage. At pre-test, t-tests showed: gamma vs. sham (t(26)=1.768, p=0.085), gamma vs. theta (t(26)=1.768, p=0.085), and theta vs. sham (t(26)=1.57×10-15, p>0.999). At post-test, the comparisons remained non-significant (gamma vs. sham: t(26)=0.942, p=0.352; gamma vs. theta: t(26)=0.388, p=0.700; theta vs. sham: t(26)=0.535, p=0.596).
DISCUSSION
This study aimed to compare the effects of theta and gamma tACS on cognitive function and memory improvement in individuals with MCI. Participants were assigned to one of three conditions: sham, theta (5 Hz), or gamma (40 Hz) stimulation and received tACS over the DLPFC (F3, F4) for approximately 30 min. Cognitive tasks, including the Digit Span and TMT-A, were administered immediately before and after stimulation to assess changes in performance. The results showed no significant differences in cognitive task performance before and after stimulation between the sham and theta groups. However, a significant reduction in TMT-A errors was observed in the gamma group, and while Digit Span Forward lacked a group×time interaction, only the gamma group showed a significant within-group improvement (Figure 2).
The TMT is a widely used neuropsychological measure of working memory, processing speed and cognitive flexibility [22], domains often impaired in individuals with MCI. Prior studies have consistently shown that patients with MCI exhibit prolonged completion times and higher error rates on TMT compared to cognitively healthy controls, reflecting early disruption in attention, executive function and working memory [23]. In the present study, the gamma tACS group demonstrated a significant reduction in TMT-A errors following stimulation, suggesting a facilitation of task-relevant cognitive processes. Given the established sensitivity of TMT to cognitive decline, particularly in populations at risk for MCI, this improvement may serve as a meaningful indicator of enhanced cognitive efficiency. Although the Digit Span Forward task also showed a significant pre-to-post improvement in the gamma group, the absence of a group×time interaction limits the strength of this finding. Nonetheless, the result may tentatively suggest a supplementary benefit of gamma-frequency stimulation on working memory and attention. These results indicate that gamma tACS enhances multiple facets of cognitive function, including working memory, processing efficiency, and executive control, thereby providing a potential intervention to mitigate cognitive impairment in patients with MCI.
The TMT-A task is known to be closely related to the functions of the prefrontal cortex, particularly the DLPFC, which is crucial for working memory and executive function [24]. The prefrontal cortex plays a central role in processes involving temporary storage and manipulation of information, including regulating complex cognitive tasks requiring flexibility and control. In patients with MCI, atrophy of the medial temporal lobe is often accompanied by reduced prefrontal cortex function, particularly affecting working memory and executive processes [25,26]. This reduction in prefrontal activity is a key contributor to the cognitive deficits observed in these populations, including poor performance on tasks such as TMT.
Given the known involvement of the DLPFC in these cognitive domains, it is plausible that applying gamma frequency tACS over this region temporarily enhances neural activity and connectivity within the prefrontal cortex, leading to improved task performance. Gamma oscillations are associated with enhanced synaptic plasticity and improved neural synchrony, essential for adequate working memory and executive control [27]. Therefore, the observed improvement in TMT-A task in the gamma tACS group may be attributed to the temporary modulation of DLPFC activity, which facilitates better cognitive performance in individuals with MCI. This suggests that gamma tACS could serve as a promising intervention to target specific deficits in the prefrontal cortex function, providing meaningful cognitive benefits for individuals at risk of progressing to AD.
Given that the primary focus of this study was to compare the effects of theta and gamma tACS, the absence of significant effects in the theta tACS group warrants the consideration of several factors that may have influenced the outcomes. Minorly, but worth considering, the lack of significant effects in the theta tACS group may be related to differences in the energy levels associated with the applied stimulation frequencies. In tES, the ability of the current to penetrate the scalp, skull, and cerebrospinal fluid (CSF) and ultimately reach the brain can vary significantly depending on the stimulation frequency [28,29]. Lower frequency stimulations, such as theta (5 Hz), may deliver less energy to the brain than higher frequencies, such as gamma (40 Hz), owing to increased resistance from these multiple barriers. This aligns with well-known Planck’s principle, which describes how energy is directly proportional to frequency [30]. Higher frequency electrical stimulation inherently carries more energy per cycle than lower frequency stimulation. Additionally, in biological systems, lower frequency currents encounter greater impedance from the scalp, skull, and CSF, further limiting energy transmission to the brain. As a result, lower frequency stimulation not only contains less intrinsic energy but also faces greater biological resistance, making it less effective at delivering energy to neural tissue compared to higher frequencies. This could indicate that theta tACS, at its lower energy level, is less effective in reaching and modulating neural activity in the targeted prefrontal cortex. Consequently, reduced energy delivery might explain why significant cognitive improvements were not observed in the theta group compared with the gamma group. Suppose we hypothesize that the neural potential targeted by tACS is independent of the stimulation frequency. In that case, the overall energy level of the electrical stimulation is crucial for the effective modulation of neural activity. The higher energy associated with gamma tACS likely significantly impacted synaptic plasticity and neural synchrony, improving cognitive performances. It is also important to note that this explanation is somewhat speculative, given that previous research has shown that theta tACS can be effective in healthy populations (e.g., studies indicating improvements in working memory and attention) [6]. It should be acknowledged that energy levels may not fully account for the observed lack of effects in the theta group, particularly considering that theta stimulation has demonstrated efficacy in non-clinical populations. This suggests that using energy-level differences as a logical basis for the lack of theta effects is inconclusive. However, the current findings suggest that the efficacy of theta tACS may not extend as effectively to populations with MCI, potentially because of reduced neural responsiveness or more significant structural and functional deficits. Therefore, considering the technical aspects of transcranial stimulation, the differential effects of theta and gamma tACS observed in this study highlight the importance of optimizing the stimulation parameters to achieve meaningful cognitive outcomes in clinical populations.
A significant factor to consider is the abnormal brain wave activity patterns commonly observed in individuals with MCI. It is well documented that MCI patients have an increased presence of slow-wave activity, including delta and theta waves, and a reduced presence of faster waves, such as beta and gamma oscillations [31,32]. This imbalance suggests that neural networks in patients with MCI are dominated by slow-wave dynamics, which can significantly influence the effects of neuromodulation interventions such as tACS.
While not directly assessed in the present study, theta-gamma coupling has been proposed as a theoretical mechanism relevant to cognitive enhancement. Theta and gamma oscillations are thought to play complementary roles—theta coordinating large-scale neural communication, and gamma supporting localized information processing [33]. Cognitive processes such as working memory may depend on the optimal balance between these rhythms. In healthy individuals, where baseline slow-wave activity is relatively low, theta stimulation may facilitate this coupling and thereby support cognitive performance. However, in individuals with MCI, who often exhibit elevated theta activity and reduced gamma power, this balance may already be disrupted. From this perspective, additional theta input may not enhance functional synchrony and could instead reinforce slow-wave dominance. While speculative, this framework may help explain the relatively limited cognitive response to theta tACS observed in the current study compared to gamma stimulation.
This study aimed to directly compare the effects of theta and gamma tACS on cognitive function in individuals with MCI. It revealed that gamma tACS is promising to enhance working memory and cognitive performance. While gamma stimulation resulted in significant improvements in TMT-A task, suggesting enhanced neural synchrony and cognitive flexibility, theta stimulation did not yield comparable gains. This outcome highlights the differential impacts of theta and gamma tACS in MCI populations, potentially attributable to the altered oscillatory dynamics observed in these individuals, such as increased slow-wave activity and impaired theta-gamma coupling. These findings underscore the potential of gamma tACS as a targeted intervention to mitigate cognitive decline in MCI and highlight the importance of individualized approaches to neuromodulation treatments that consider the unique neurophysiological characteristics of each patient.
However, this study had several limitations. First, the sample size was relatively small, which may have limited the generalizability of the findings. Further studies with larger sample sizes are needed to validate these findings. Second, the study did not include a comprehensive assessment of long-term effects, as the outcomes were measured immediately after stimulation. The persistence of the cognitive improvements observed with gamma tACS remains unknown, and future studies should investigate the long-term efficacy and potential cumulative effects of repeated tACS sessions. Third, this study focused on a specific frequency range for theta and gamma tACS, and possibly other stimulation parameters or individualized frequency adjustments could yield different outcomes. Fourth, while major psychiatric and neurological conditions were excluded and cognitive enhancers were prohibited, the use of psychotropic medications (e.g., benzodiazepines) and mild psychiatric symptoms such as subthreshold depression were not explicitly restricted to preserve ecological validity in this exploratory study. Nonetheless, these factors may have introduced confounding effects, and future confirmatory trials should consider more stringent control over such variables. Finally, this study did not explore other potential mechanisms contributing to the observed effects, such as neuroplasticity or functional connectivity changes, which could provide a more comprehensive understanding of the underlying neural processes.
PDF Links
PubReader
ePub Link
XML Download
Full text via DOI
Download Citation
Print
