Abstract
Despite the growing interest in precision medicine-based therapies for Alzheimer's disease (AD), little research has been conducted on how individual AD risk factors influence changes in cognitive function following transcranial direct current stimulation (tDCS). This study evaluates the cognitive effects of sequential tDCS on 63 mild cognitive impairment (MCI) patients, considering AD risk factors such as amyloid-beta deposition, APOE ε4, BDNF polymorphism, and sex. Using both frequentist and Bayesian methods, we assessed the interaction of tDCS with these risk factors on cognitive performance. Notably, we found that amyloid-beta deposition significantly interacted with tDCS in improving executive function, specifically Stroop Word-Color scores, with strong Bayesian support for this finding. Memory enhancements were differentially influenced by BDNF Met carrier status. However, sex and APOE ε4 status did not show significant effects. Our results highlight the importance of individual AD risk factors in modulating cognitive outcomes from tDCS, suggesting that precision medicine may offer more effective tDCS treatments tailored to individual risk profiles in early AD stages.
Similar content being viewed by others
Introduction
Alzheimer's disease (AD), characterized by the buildup of amyloid beta (Aβ) and tau proteins, profoundly impacts cognitive abilities and daily life1. Mild Cognitive Impairment (MCI), a condition that may precede dementia, is marked by noticeable cognitive decline that is not severe enough to interfere significantly with daily life or independent function2. However, MCI often progresses to more severe forms of dementia, including AD2. Current treatments for MCI are limited and show modest effectiveness3. Non-pharmaceutical interventions, such as cognitive stimulation, physical activity, and dietary modifications, demonstrate promising results but are challenging to maintain consistently in MCI patients4,5,6.
Monoclonal antibodies targeting Aβ, aimed at plaque removal, have shown modest cognitive decline reduction by 25–30% in AD patients7,8 but are less effective in those with MCI9, indicating the need for broader intervention strategies. This highlights the complexity of AD and the necessity for accessible treatments that not only enhance cognitive abilities but also mitigate neurodegenerative processes.
Therefore, alternative treatments accessible to individuals with MCI are required. There is increasing interest in non-invasive brain stimulation methods, especially transcranial direct current stimulation (tDCS)10. Along with its affordability and portability, tDCS modulates cortical neuron excitability, influences neuroplasticity, and potentially aids in the removal of Aβ by affecting the brain-blood barrier (BBB)11,12,13. tDCS transiently increases BBB permeability, facilitating Aβ clearance primarily through nitric oxide release13. This mechanism involves the opening of endothelial junctions and modulation of the basement membrane, enhancing Aβ transport across the BBB13,14. Additionally, tDCS influences the expression of proteins critical for BBB integrity, further promoting Aβ clearance13.
Studies have shown that tDCS can improve cognitive performance, although results are mixed. While some studies report improvements in Mini-Mental State Examination (MMSE) scores and recognition memory in AD patients15,16,17, others show no significant cognitive benefits18. Previous studies often target the dorsolateral prefrontal cortex (DLPFC), and the extent of cognitive improvement differs based on factors such as session number, stimulation intensity, and duration19,20. In the context of DLPFC laterality, past research on MCI and dementia patients has predominantly targeted the left DLPFC due to its significant role in cognitive functions such as verbal memory, working memory, and executive functions19,20. Although studies directly comparing the left and right DLPFC are limited, prior research has shown that the left DLPFC significantly modulates cognitive and physiological correlates of executive function more effectively than the right DLPFC21.
Research on MCI patients is limited but suggests potential benefits of tDCS on memory and attention19,22, although meta-analyses present mixed results19,23, necessitating cautious interpretation.
Precision medicine, which tailors treatment strategies based on an individual's genetics, biomarkers, and clinical data24, shows promise in optimizing tDCS responses. Factors such as Aβ deposition25, APOE ε4 allele26, Brain-Derived Neurotrophic Factor (BDNF) levels27,28, and sex differences significantly influence AD progression and treatment responses29. Despite these insights, few studies have explored the impact of tDCS based on individual AD risk factors.
Our preliminary study on MCI patients indicated that applying tDCS to the DLPFC results in varied effects on brain functionality, linked to individual factors such as Aβ deposition and APOE ε4 allele status30. Previous research suggests that BDNF Val66Met polymorphism and sex differences also significantly influence neuroplasticity and neurodegeneration after tDCS application31,32,33. These findings highlight the necessity of customizing tDCS treatment protocols based on individual risk factors.
In our ongoing study, we examined neuropsychological performance alterations in MCI patients following a 10-session sequential anodal tDCS application on the left DLPFC, considering individual AD risk factors such as Aβ deposition, APOE ε4 status, BDNF polymorphism, and sex. We employed both Bayesian and frequentist methods to leverage their complementary strengths. The Bayesian approach integrates prior knowledge and quantifies uncertainty34,35, while the frequentist approach emphasizes empirical data and hypothesis testing. This dual methodology ensures a comprehensive analysis of tDCS effects, considering the complexity of AD risk factors.
Materials and methods
Participants
The study recruited participants from the Brain Health Center at Yeoui-do St. Mary’s Hospital, affiliated with College of Medicine, the Catholic University of Korea. The selection criteria were as follows: (1) alignment with Petersen's MCI criteria36, including Seoul-Instrumental Activities of Daily Living (S-IADL) scores below 8 to confirm independence in activities of daily living37, and (2) a Clinical Dementia Rating (CDR) of 0.5. Exclusion criteria were as follows: (1) prior alcohol or drug misuse, head injuries, or mental health conditions; (2) use of psychotropic drugs such as cholinesterase inhibitors, N-Methyl-D-Aspartate receptor antagonists, antidepressants, benzodiazepines, and antipsychotics; (3) tDCS or MRI contraindications (e.g., ferromagnetic or coiled metal implants); and (4) any skin disorder that compromised skin integrity over the scalp. Additionally, only participants with Hamilton Depression Rating Scale (HAMD) scores of 7 or below were included, ensuring they were within the normal range for depression38. All participants were naive to tDCS, meaning they had not previously undergone tDCS treatment. Two experts in geriatric psychiatry oversaw this selection process. Participants agreed to review their medical records. All evaluations were performed at the Brain Health Center of Yeoui-do St. Mary’s Hospital, College of Medicine, the Catholic University of Korea. The study adhered to the principles of the Declaration of Helsinki and was approved by the Institutional Review Board of the Catholic University of Korea (SC19DEST0012). Written informed consent was obtained from all the participants.
Study protocol
In this single-arm, prospective study, we implemented a design that did not involve a sham condition. The patients underwent ten tDCS sessions administered at their homes, scheduled five times per week for two weeks. This number of sessions was chosen in alignment with findings from prior clinical research, in where ten tDCS sessions demonstrated efficacy in treating AD and MCI17,22,39,40, also considering adherence factors in older patients. Neuropsychological evaluations and MRI scans were conducted at the Brain Health Center of Yeouido St. Mary's Hospital, both within a fortnight before initiating tDCS and following the final 10th session. In addition, participants were subjected to [18F] flutemetamol (FMM) PET-CT scans and APOE and BDNF genetic testing, all within four weeks prior to the first tDCS session. Neuropsychological examiners and the participants remained uninformed about the FMM-PET, APOE, and BDNF genotyping results. A detailed schematic illustrating the experimental procedures, as featured in our previous study, can be referred to for further clarification32. Registered with the Clinical Research Information Service of the Korea Disease Control and Prevention Agency (KCT0006020), this research spanned from May 2020 to February 2022 at the Brain Health Center. The authors declare no ethical or financial conflict of interests with respect to the manufacturers of any of the equipment used in the study.
Transcranial direct current stimulation application
In this procedure, a steady direct current of 2 mA was applied for 20 min using a stimulator compatible with MRI technology (YDS-301N, YBrain, Seoul, Republic of Korea). The NEUROPHET tES LAB software (version 3.0; Neurophet, Seoul, Republic of Korea) was used for brain modeling, calculating the tDCS-induced E-field, and determining each participant’s optimal electrode location based on individual brain structural characteristics to stimulate a specific target area. A recent simulation study demonstrated that the optimized tDCS electrode position increased the E-field over the left DLPFC by 55.28% compared to the conventional 10–20 EEG-based system41. This emphasizes the potential of simulation software to enhance stimulation efficacy by optimizing electrode placement according to individual brain structures. For the creation of the brain model and later analysis of the tDCS-induced E-field, all participants underwent a T1-weighted MR image at baseline. The software analyzed and segmented the T1-weighted MR images to reconstruct a 3D model of each participant’s brain, including structures such as skin, skull, cerebral gray and white matter, cerebellum gray and white matter, cerebrospinal fluid (CSF), and ventricles. The electrical conductivity values pre-programmed in the software were as follows: skin at 0.465 S/m, skull at 0.010 S/m, cerebral and cerebellar gray matter at 0.276 S/m, cerebral and cerebellar white matter at 0.126 S/m, and CSF and ventricles at 1.65 S/m42. After the brain model was created, an investigator assigned landmarks (nasion, inion, and both preauricular points) to the model to determine the tDCS electrode locations accurately.
The DLPFC was targeted for stimulation, with the anode positioned above the left DLPFC and the cathode over the contralateral supraorbital zone. Both electrodes were disk type with a radius of 3 cm. The tDCS intensity of 2 mA was input into the software. Based on these initial locations, tDCS parameters, and the conductivity of the brain tissues, the software analyzed the optimal electrode position for each participant. The software “moved” the electrodes around the target area to determine the location that generated the maximum E-field-induced by tDCS. Finally, the software provided the necessary guides to find the optimized electrode position on the participant’s head. The trained personnel were then provided with the optimized electrode positions to apply the stimulation. The device operation was facilitated by trained personnel who visited patients' homes during each session. Measurements were taken from one preauricular point to the center of the electrode, intersecting this point with a line from the vertex to the nasion. The distance from both preauricular points to the center of the electrode was noted, and its position relative to the landmarks was verified before each session began. Furthermore, 15 min into the session, the staff reassessed the electrode's placement for precision. The same staff member was assigned to each participant throughout the ten sessions.
Neuropsychological assessment
All participants underwent cognitive assessments using the Korean adaptation of the Consortium to Establish a Registry for Alzheimer’s Disease (CERAD-K)43. This included Korean versions of several tests: Verbal Fluency (VF), the 15-item Boston Naming Test, and the Korean version of MMSE (MMSE-K)44, as well as tests for word list memory (WLM), recall, recognition, constructional praxis, and recall (CR). We also summed the scores of word-list memory, recall, and recognition to derive the total memory score. The comprehensive CERAD-K score was the aggregate of all scores, excluding the MMSE-K and CR. Cognitive function was evaluated using the Korean Montreal Cognitive Assessment (MoCA-K). For executive function, we utilized the Korean Stroop Word-Color Test (K-SWCT), which measures reaction control in letter and color reading scenarios45, and Trail Making Test B, which assesses the time taken to sequence numbers and letters. Detailed descriptions of the assessments are provided in the Supplementary Material.
Aβ deposition
The methodology for acquiring and processing [18F]-FMM PET images to evaluate Aβ deposition, as well as the calculation of standardized uptake value ratios (SUVRs) to quantify deposition intensity, is comprehensively detailed in the Supplementary Material. A threshold of 0.62 was employed to differentiate between Aβ positive (Aβ +) and Aβ negative (Aβ −) accumulations, aligning with previous FMM-PET studies46. It is crucial to note that 'negative accumulation' denotes subthreshold deposition rather than a complete absence of amyloid deposition.
APOE genotyping
The approach for APOE genotyping is detailed in the Supplementary Material. According to our protocol, we would exclude subjects who possessed the APOE ε2 allele due to its observed protective role47. Participants were classified based on the presence of the APOE ε4 allele; those with at least one ε4 allele were grouped as APOE ε4 carriers, whereas those without any ε4 alleles were designated as non-carriers.
BDNF genotyping
The methodology for BDNF genotyping is described in the Supplementary Material. In the context of the BDNF Val66Met polymorphism (rs6265), we categorized participants based on existing genetic research48,49 into groups: those carrying at least one Met66 allele were labeled as Met carriers, whereas those without any Met66 alleles were considered Met non-carriers.
Magnetic resonance imaging acquisition and processing
Detailed information regarding magnetic resonance imaging acquisition and processing is available in the Supplementary Material.
Statistical analysis
Statistical analyses were performed using the R software (version 4.3.2) and jamovi (version 2.3.28) (https://www.jamovi.org). Assumptions of normality were tested for continuous variables using the Kolmogorov–Smirnov test in R software. All data demonstrated a normal distribution and were standardized by z-score transformation for the analysis.
Repeated-measures ANOVA was used to predict the impact of the effect modifier-by-tDCS interaction (effect modifier × tDCS) on neuropsychological performance scores, with tDCS (pre- and post-tDCS) as a repeated-measures factor and Aβ deposition, APOE ε4 carrier status, BDNF Val66Met polymorphism status, and sex as the between-subject factor (a potential effect modifier). Additionally, age, years of education, adjusted hippocampal volume, and other nonincluded effect modifiers were incorporated as covariates. Adjusted hippocampal volume = raw hippocampal volume − β × (intracranial volume [ICV] − mean ICV), where β is the regression slope of the raw hippocampal volume versus ICV50. This method was applied separately to determine both the right- and left-adjusted hippocampal volumes, and their averages were used in the analysis. For the analysis of within-subject effects, when more than one effect modifier demonstrated an interaction with tDCS application, we included the multiple effect modifiers × tDCS interaction term (effect modifier 1 × effect modifier 2 × tDCS) in the analysis.
In this study, we applied a Bayesian perspective to replicate repeated measures ANOVA, which quantifies the impact of the effect modifier × tDCS interaction on neuropsychological performance scores and determines whether these interactions exist. The purpose of this study was to measure the degree of support for the hypothesis. JASP’s (version 0.18.2) default Jeffreys-Zellner-Siow priors were applied to the parameters, and each hypothesis under investigation was assumed to be equally likely. Instead of a P-value, this analysis provides the Bayes factor (BF). For example, if BF10 is ‘3,’ the data are three times more likely under model '1' than under model '0.’ The BF is interpreted as a continuous measure of evidence. In interpreting BF10, the following qualitative strength of evidence statements are recommended: a BF10 between 1 and 3 suggests anecdotal evidence, a BF10 in the range of 3 to 10 indicates moderate evidence, a BF10 between 10 and 30 is indicative of strong evidence, a BF10 from 30 to 100 denotes very strong evidence, and a BF10 greater than 100 is reflective of extreme evidence in favor of model'151.' In Repeated-Measures ANOVA based model comparisons, we calculate the ‘prior model probability’ of each hypothesis before observing the data (P[M]) and the 'posterior model probability' (P[M|data]) of each hypothesis after observing the data. The BF for the model (BFM) was then determined by comparing the two probabilities. Specifically, it is calculated by dividing the posterior odds of the model by prior odds. This calculation reflects the degree to which the observed data shift our belief from the prior to the posterior. A higher BFM value indicated that the data provided substantial support to the model, significantly altering our initial understanding. Conversely, a lower BFM suggests that the data did not significantly change the prior beliefs about the model's likelihood. In addition, the error rate of the numerical method was used to calculate BF. A lower error rate indicates less variability in the BF across Markov Chain Monte Carlo (MCMC) sampling. Typically, an error rate of less than 20 is considered acceptable.
Furthermore, the BF for inclusion (BFinclusion) was calculated to assess the strength of evidence for incorporating the effect modifier × tDCS interaction within our model 51. The methodology for deriving this value is grounded in Bayesian statistics and involves comparing models with and without the interaction in question. First, we established two models: one including the effect modifier × tDCS interaction (inclusion) and another excluding it (exclusion). This comparative framework allows for an assessment of the interaction's contribution to the model. The prior probabilities [P(inclusion)] for each model were then determined based on the existing knowledge and theoretical considerations. These probabilities represent our initial belief in the plausibility of each model before analyzing the data. After data collection and processing, we calculated the posterior probabilities [P(inclusion|data)] for both the models. These probabilities reflect the updated likelihood of each model, in light of the new evidence provided by our data. The BFinclusion was computed using the following formula:
In this formula, \(\text{P}\left(\text{inclusion}|\text{data}\right)\) and \(\text{P}\left(\text{exclusion}|\text{data}\right)\) are the posterior probabilities of the models with and without interaction, respectively. This calculation provides the ratio of these probabilities, adjusted by the ratio of their prior probabilities. The resulting BFinclusion value serves as an indicator of the evidence strength for including the effect modifier × tDCS interaction in the model. The resulting BFinclusion value serves as an indicator of the evidence strength for including the effect modifier × tDCS interaction in the model. A BFinclusion value greater than 1 suggests evidence favoring the model with the interaction. Conventionally, BFinclusion: 1–3 is anecdotal evidence, 3–10 indicates moderate evidence, 10–30 indicates strong evidence, 30–100 indicates very strong evidence, and BFinclusion > 100 indicates extreme evidence. Additionally, considering the differential risk of AD and differences in tDCS current intensity at the target site based on sex29,52, the effects of tDCS were analyzed separately for males and females using repeated-measures ANOVA based on both frequentist and Bayesian perspectives. Covariates included age, education years, adjusted hippocampal volume, Aβ deposition, APOE ε4 carrier status, and BDNF Val66Met polymorphism status.
Finally, partial correlation analysis was performed to evaluate the association between baseline [18F] FMM SUVRPONS and differences in neuropsychological performance scores, adjusting for age, sex, years of education, adjusted hippocampal volume, APOE ε4 carrier, and BDNF Val66Met polymorphism status. Both global and regional FMM SUVRs were assessed in this analysis. Regional SUVRs were evaluated in the frontal (FL), superior parietal (PL), lateral temporal (TL), and anterior (ACC) and posterior cingulate cortex regions (PCC). All data were standardized using z-score transformation for partial correlation analysis. All statistical analyses used a two-tailed P-value < 0.05 to define statistical significance. We also extended the Bayesian approach, applied in the Repeated-Measures ANOVA, to partial correlation analysis. In the context of our Bayesian correlation matrix, we utilized BF10 to assess the strength of the correlations between variables. This approach focuses on evaluating the presence and magnitude of associations, employing BF10 to quantify evidence supporting the existence of a relationship between pairs of variables. Unlike more complex model structures, such as regression or ANOVA, where BFinclusion is relevant, the Bayesian correlation matrix simplifies the analysis to the calculation of BF10, reflecting the direct evaluation of correlations without the need to consider model inclusion factors.
Results
Baseline demographic and clinical data
In this study, out of the initial 70 participants who were eligible based on specific criteria, seven did not complete the study. This dropout was due to six individuals opting out and one experiencing a minor side effect from tDCS, specifically a tingling sensation beneath the electrode. Consequently, 63 participants completed the study and were included in the final analysis. For a more detailed visualization of participant flow in this study, refer to the flowchart available in the original research article32. Table 1 shows the baseline demographic data for the participants who completed the study. The number and distribution of participants based on AD risk factors are described in the Supplementary Material.
Changes in neuropsychological performance scores according to effect modifiers
Several important interactions were observed in the present study. First, a significant interaction was found between Aβ deposition and tDCS, manifesting as increased CERAD-K SWCT scores in patients with Aβ- MCI (CERAD-K SWCT, P = 0.009, Fig. 1A). Moreover, a BDNF polymorphism × tDCS interaction was also identified, which was shown by an increased CERAD-K WLM score in Met non-carriers (P = 0.028, Fig. 1B). A table containing all the statistical comparisons shown in Fig. 1 can be found in the Supplementary Results.
Differential impact of anodal tDCS on neuropsychological performance scores according to effect modifiers: (A) Aβ deposition and (B) BDNF polymorphism. Both traditional Repeated Measures ANOVA and Bayesian Repeated Measures ANOVA were used to predict the impact of the effect modifier-by-tDCS interaction (effect modifier*tDCS) on neuropsychological performance (between-subject factors: Aβ deposition, APOE ε4 carrier status, BDNF polymorphism, sex), adjusting for age, years of education, adjusted hippocampal volume, and between-subject factors not showing interaction. P-value, the statistical significance of the effect modifier-by-tDCS interaction, as determined by traditional Repeated Measures ANOVA; BFinclusion, the Bayesian factor value derived from Bayesian Repeated Measures ANOVA. The vertical bars in the graph represent confidence intervals. CERAD-K, the Korean version of the Consortium to Establish a Registry for AD; Total CERAD-K scores, the sum of all subcategory scores, excluding MMSE-K and CR cores; WLM, word list memory.
Regarding the Aβ deposition × tDCS interaction on the CERAD-K SWCT score, Bayesian re-analysis was performed. The results indicated that the P(M) of the interaction was 0.200, suggesting a 20% belief in its presence. After data analysis, P(M|data) increased to 0.842 or 84.2%, significantly enhancing the likelihood of this interaction being real. BFM and BF10 were 21.270 and 17.088, respectively. This suggests strong evidence for the impact of the Aβ deposition × tDCS interaction on the CERAD-K SWCT score (Table 2A). The BFinclusion regarding the Aβ deposition × tDCS interaction for the CERAD-K SWCT score was calculated to be 21.27, indicating strong evidence for including the Aβ deposition × tDCS interaction in the model. [P(inclusion) = 0.200, P(inclusion|data) = 0.842] (Table 3A).
In contrast, for the Total CERAD-K score, repeated measures ANOVA did not reveal any significant Aβ deposition × tDCS interaction. However, the Bayesian reanalysis provides a different perspective. P(M) remained at 0.200, reflecting 20% initial belief in the interaction. The P(M|data) increased to 0.546 or 54.6%, indicating a notable increase in the probability of the interaction's existence. The BFM was calculated to be 4.809, and the BF10 was an impressive 833.16. This indicates extreme evidence supporting the impact of the Aβ deposition × tDCS interaction on the Total CERAD-K score (Table 2B). The BFinclusion regarding the Aβ deposition × tDCS interaction for the Total CERAD-K score was calculated to be 4.81, indicating moderate evidence for Aβ deposition × tDCS interaction in the model. [P(inclusion) = 0.200, P(inclusion|data) = 0.546] (Table 3B).
For the re-analysis of the BDNF polymorphism × tDCS interaction on the CERAD-K WLM score, a Bayesian approach was adopted. The P(M) value for this specific interaction was determined to be 0.200, indicating an initial 20% conviction of its existence. Subsequent to the data review, P(M|data) increased to 0.478 or 47.8%, markedly elevating the perceived reality of this interaction. The BFM was computed to be 3.67, while the evidence in favor of the interaction, as denoted by BF10, reached a significance of 2357.392. These results strongly support the substantial influence of the BDNF polymorphism × tDCS interaction on the CERAD-K WLM score, as detailed in Table 2C. Furthermore, the calculated BFinclusion for this interaction was 3.67 (Table 3C). This result, while indicating moderate evidence, suggests a noteworthy inclination towards incorporating the BDNF polymorphism × tDCS interaction into our model, corroborated by the calculated probabilities [P(inclusion) = 0.200 and P(inclusion|data) = 0.478]. Despite these significant findings, no other effect modifiers (sex and APOE ε4 carrier status) × tDCS interaction was observed for any of the neuropsychological performance scores.
The results of cognitive function changes following tDCS application for each sex are as follows. In the analysis based on frequentist methodology, no significant cognitive function changes were observed following tDCS application in both males and females. However, in the Bayesian analysis, only females showed improvements with extreme evidence strength in the CERAD-K WLM, total memory, and total scores (BF10 values were 65,463.09, 1766.62, and 912.49, respectively) following tDCS application. Additionally, no significant interactions were found between tDCS and multiple effect modifiers, including Aβ deposition, APOE ε4 status, BDNF polymorphism, and sex.
Association between baseline Aβ deposition and changes in neuropsychological performance scores
Concerning the relationship between the difference in neuropsychological performance scores and baseline global Aβ deposition, there was a negative association with the difference in MoCA-K score, CERAD-K WLR, SWCT, and Total scores (Fig. 2, MoCA-K, Pearson’s r = − 0.285, P = 0.031; CERAD-K WLR, Pearson’s r = − 0.264, P = 0.047; CERAD-K SWCT, Pearson’s r = − 0.322, P = 0.015; Total CERAD-K, Pearson’s r = − 0.274, P = 0.039).
Associations between amyloid-beta deposition and changes in neuropsychological performance scores. Partial correlation analysis adjusted for age, sex, years of education, adjusted hippocampal volume, APOE ε4 carrier status, and BDNF polymorphism. Bayes Factor (BF10) to quantify the strength of correlations between variables. The BF10 values represent the evidence ratio in favor of the alternative hypothesis over the null hypothesis. MoCA-K, the Korean version of the Montreal Cognitive Assessment; CERAD-K, the Korean version of the Consortium to Establish a Registry for AD; WLR, word list recall; Total CERAD-K scores, the sum of all subcategory scores excluding the MMSE-K and CR cores.
In the Bayesian analysis, the associations between baseline global [18F] FMM SUVRPONS and CERAD-K WLR and Total CERAD-K scores, with BF10 values of 1.270 and 1.675, respectively, fell into the category of anecdotal evidence. These results indicate a minimal level of evidence for a negative correlation with these specific neuropsychological tests. Conversely, the MoCA-K and SWCT scores provided substantial evidence. The negative correlation between baseline global Aβ deposition and MoCA-K score had a BF10 value of 9.131, indicating moderate evidence. For the SWCT, the negative association with baseline global Aβ deposition was supported by a BF10 value of 18.671, reflecting strong evidence.
Regarding the relationship between the difference in neuropsychological performance scores and baseline regional Aβ deposition, frequentist analysis revealed significant associations between specific cognitive functions and regional SUVRs. However, the Bayesian re-analysis did not show associations with a strength of moderate evidence or higher. The detailed results are as follows: Associations between the CERAD WLM scores and regional SUVRs showed significant findings with the PCC (Pearson’s r = − 0.338, P = 0.011, BF10 = 0.913) and TL (Pearson’s r = − 0.279, P = 0.038, BF10 = 0.712). For the MMSE-K scores, significant correlations were identified with the PL (Pearson’s r = − 0.294, P = 0.028, BF10 = 1.03), PCC (Pearson’s r = − 0.273, P = 0.042, BF10 = 1.12), and TL (Pearson’s r = − 0.282, P = 0.035, BF10 = 1.12). In the case of CERAD-K CR scores, significant relationships were noted with the ACC (Pearson’s r = − 0.352, P = 0.008, BF10 = 2.45), FL (Pearson’s r = − 0.355, P = 0.007, BF10 = 1.94), PL (Pearson’s r = − 0.304, P = 0.023, BF10 = 1.46), and TL (Pearson’s r = − 0.381, P = 0.004, BF10 = 2.67). Regarding the total memory scores, significant correlations were observed with the PL (Pearson’s r = − 0.278, P = 0.038, BF10 = 1.50), PCC (Pearson’s r = − 0.281, P = 0.036, BF10 = 1.84), and TL (Pearson’s r = − 0.349, P = 0.008, BF10 = 2.26).
Discussion
In this study, we aimed to evaluate whether a 2-week application of sequential tDCS alters neuropsychological performance scores in patients with MCI and whether these changes depend on individual risk factors for AD. Furthermore, to assess differences in cognitive function, we also considered the correlation with baseline Aβ deposition. Our analysis employed both frequentist and Bayesian methodologies, allowing for a comprehensive interpretation of the data and a more complete understanding of the neuropsychological impacts and AD risk factors.
Regarding executive function, our frequentist analysis indicated a significant interaction between AD risk factors and tDCS in MCI patients. This was particularly evident in the enhanced K-SWCT scores among participants with negative Aβ deposition following tDCS sessions. Bayesian analysis corroborated this finding, presenting strong evidence of the substantial effect of Aβ deposition × tDCS interaction on executive function, thereby challenging the null hypothesis. Executive function, encompassing high-order cognitive abilities, is pivotal for adaptive, goal-oriented behaviors and thoughts and plays a significant role in various aspects of life53. The SWCT addresses key aspects of executive functioning, including inhibitory control and cognitive flexibility54. Central to these functions, the DLPFC has been identified as a key neural hub55, with studies indicating its activation across various executive function subdomains56. In line with this, tDCS targeting the left DLPFC has been explored as a potential modulator of executive function. Research has delved into how tDCS applied to this area might enhance executive functioning, highlighting changes in neurophysiological markers, such as increased P300 and decreased N20021. Moreover, it is plausible that the role of tDCS in modulating cortical excitability and neuroplasticity is more significant before Aβ accumulates to levels forming plaques. This suggests that tDCS could be more effective in enhancing executive functions in the Aβ negative group. Given these findings, the potential for tDCS to facilitate Aβ clearance may be more pronounced before the formation of detectable Aβ plaques as assessed by Aβ-PET imaging. Therefore, future studies should consider using fluid-based Aβ biomarkers to evaluate the relationship between dynamic Aβ accumulation and the efficacy of tDCS, thereby providing a more complete understanding of its therapeutic potential.
When considering patients with MCI, the effects of tDCS on executive function are mixed. Some studies have not found significant benefits from sequential anodal tDCS to the left DLPFC (10 sessions)57, while others have reported improvements58. These varying results could be influenced by factors beyond the sample size and session frequency, such as the absence of consideration of AD risk factors, including Aβ deposition. Prior studies, although primarily involving cognitively unimpaired older adults, have shown that Aβ deposition is independently associated with poorer executive function performance in contrast to memory performance, which is more closely related to tau pathology59. These findings support the notable interaction between Aβ deposition and tDCS observed in our study, suggesting an influence on executive function changes. This suggests a potential benefit in considering key AD risk factors, such as Aβ deposition, when applying brain region-specific stimulation. In a clinical context, such thoughtful considerations could subtly yet meaningfully impact the efficacy of interventions, such as tDCS, especially regarding the modulation of executive function in patients with MCI.
The outcomes of global cognition were somewhat different. While the Repeated-Measures ANOVA did not demonstrate a significant interaction effect on the overall CERAD-K score, our Bayesian reanalysis revealed substantial evidence supporting the existence of this interaction. This finding suggests its relevance in predictive models of alterations in global cognitive scores. Recent meta-analyses have yielded mixed results regarding the effect of tDCS on global cognition in patients with MCI19,60. Variability in stimulation protocols and a lack of consideration of AD risk factors, such as Aβ deposition, complicate these findings. While some analyses indicate that anodal tDCS may enhance general cognition in AD-related MCI and dementia, the heterogeneity of the study designs calls for cautious interpretation and the need for additional research. In terms of cognitive function domains, executive function was notably associated with Aβ deposition59,61. Despite its relatively weaker correlation with cognition in the prodromal phase of AD compared to tauopathy or neurodegeneration62, the link between Aβ deposition and global cognition has been reported with small-to-moderate effect sizes63,64. Our study's Bayesian analysis, which revealed the differential impact of sequential anodal tDCS on global cognition based on Aβ deposition status, underscores the potential of addressing these inconsistencies by factoring in AD risk modifiers. Furthermore, considering the role of tau pathology as a predictor of global cognitive decline and its interaction with Aβ deposition, future studies should incorporate comprehensive tau pathology data. This approach promises to deepen our understanding of the effects of tDCS on global cognitive function within the AD spectrum.
Additionally, in our study, sequential tDCS increased CERAD-K WLM scores in BDNF Met non-carriers, indicating a notable interaction between BDNF polymorphisms and tDCS, particularly affecting memory performance. This observation was further validated by Bayesian analysis, which provided substantial evidence for the existence of an interaction and its predictive relevance in immediate memory changes. This finding aligns with the known mechanisms of after-effects of tDCS, including modulation of NMDA receptors and the induction of synaptic plasticity through long-term potentiation (LTP) and depression (LTD)65. The role of BDNF in these processes is critical, especially in the regulation of protein synthesis at glutamatergic synapses, which is fundamental for LTP- and LTD-like neuroplasticity66,67. Research has consistently shown that epigenetic regulation and polymorphisms of BDNF, particularly the BDNF Val66Met polymorphism, influence synaptic plasticity68, which is a key factor in cognitive functioning69. Met carriers of BDNF typically exhibit reduced secretion of this neurotrophin and have been found to perform poorly in memory tasks compared to non-carriers, a pattern also observed in patients with amnestic MCI70. Our results suggest that the application of sequential anodal tDCS may have differential effects on memory performance based on the BDNF Met carrier status, with non-carriers showing more substantial improvements. This highlights the potential of personalized tDCS interventions that consider genetic variations, such as BDNF polymorphisms. To further elucidate these effects, future research could benefit from directly measuring BDNF release and assessing LTP and LTD, thereby providing a clear understanding of the mode of action of tDCS in cognitive modulation.
Additionally, considering the differential risk of AD and differences in tDCS current intensity at the target site based on sex52, we analyzed the effects of tDCS separately for males and females using both frequentist and Bayesian approaches. While the frequentist approach did not show significant cognitive changes post-tDCS in either sex, Bayesian analysis revealed significant improvements in overall cognitive function and memory domains in females only, with extreme evidence. Previous research has shown that the age-dependent sex difference in tDCS current intensity, due to cerebral atrophy, might lead to higher tDCS current intensity at the target site in older females compared to males52. Although the differences in the effects of tDCS application on cognitive function changes by sex have not been specifically explored, this factor could explain the significant improvements in memory and overall cognitive function observed in females in our study. Therefore, further research is necessary to explore the sex-specific cognitive effects of tDCS in larger samples.
Lastly, in our research, frequentist analysis revealed that individuals with higher baseline Aβ deposition experienced a more pronounced decline in global cognition and memory performance following the application of sequential anodal tDCS to the left DLPFC. In particular, Bayesian analysis provided substantial evidence for these cognitive changes, specifically in the SWCT and MoCA-K scores. The MoCA-K, with its heightened sensitivity and specificity for MCI detection in older adults71, assesses a wide range of cognitive domains, including those linked to frontal lobe functions, such as executive function, abstraction, and verbal fluency72. Thus, it is especially effective in identifying subtle cognitive deficits that may be missed by broader cognitive screenings. Furthermore, the SWCT is a key measure of executive function, reflecting the hubal function of the prefrontal cortex53,54. Considering previous findings that highlight a more pronounced association between Aβ deposition and executive function compared to other cognitive areas59,61, it seems plausible that baseline Aβ deposition significantly influences the observed changes in executive function-related cognitive performance post-tDCS application. These findings highlight the intricate interplay between baseline neurobiological factors, such as Aβ deposition, and the cognitive effects of tDCS, particularly in executive function domains. They underscored the importance of considering individual neurobiological profiles when evaluating tDCS responses, emphasizing the role of specific biomarkers, such as Aβ deposition.
Recent studies indicate that regional Aβ burden, particularly in the prefrontal cortex, temporal lobe, and posterior cingulate cortex, is significantly associated with cognitive deficits73,74. These findings highlight the importance of regional Aβ deposition in understanding cognitive decline. Regional Aβ measures, rather than global ones, may be more closely linked to specific cognitive deficits, especially in memory and executive functions73,74. In our study, baseline Aβ deposition in regions like the PCC, FL, TL, and PL showed associations with changes in cognitive domains following tDCS application. However, these associations did not reach moderate evidence in Bayesian analysis. Specifically, our findings showed a negative association with changes in overall cognitive function and memory domains, which did not fully align with previous studies that emphasized deficits in memory and executive functions. Further research is needed to confirm these findings and explore the underlying mechanisms.
One of the limitations of our study pertains to the duration of anodal tDCS application. In this study, we assessed changes in cognitive function following a 2-week period of tDCS. Previous studies have suggested that the duration of tDCS application could act as a moderator of its effects on memory31,75. Consequently, a longer application period of anodal tDCS in patients might reveal more pronounced differences in cognitive function changes associated with AD risk factors. Therefore, future studies should consider extending the duration of tDCS treatment to explore these potential differences more comprehensively. Additionally, the learning effect of performing cognitive function tests at 2-week intervals also need to be considered when interpreting our results. Such an effect may confound the true impact of the intervention. Moreover, the small sample size may not adequately fulfill the requirements of various statistical methods employed, potentially increasing the likelihood of generating false positive results. While we incorporated Bayesian analysis to help address this issue, future studies should aim to include a larger sample size to ensure the robustness and reliability of the current findings.
Furthermore, the absence of a sham-stimulation group in our study design limited our ability to distinguish between the specific effects of tDCS and placebo responses. While this could be seen as a limitation in an intervention study, our primary focus was on exploring whether the application of tDCS shows significant interactions with AD-related effect modifiers on cognitive function. Nevertheless, incorporating a sham control in future research could significantly enhance the understanding of tDCS's efficacy and provide more robust validation of our findings.
Lastly, our study explored how tDCS affects cognitive function in MCI patients depending on AD risk factors. However, our subjects included individuals negative for Aβ deposition and lacked information on tauopathy. As such, it encompassed groups categorized under normal AD biomarkers or non-AD pathological changes76. This aspect should be considered when interpreting our results, as the participants do not exclusively represent MCI due to AD.
In conclusion, our study offers significant contributions to the understanding of sequential anodal tDCS effects on cognitive function, particularly in the context of AD risk factors. We demonstrated that Aβ deposition significantly influences changes in executive function following tDCS, a finding strongly supported by our Bayesian analysis. While traditional methods did not show a significant impact on overall cognitive function, the Bayesian approach revealed a clear interaction with Aβ deposition, suggesting its potential role in influencing comprehensive cognitive changes post-tDCS. Additionally, our findings indicated a substantial association between baseline Aβ deposition and changes in executive function-related cognitive aspects after anodal sequential tDCS application. The interaction with BDNF polymorphisms has also emerged as a significant factor, particularly in the modulation of the memory domain. While some results presented moderate evidence of clinical relevance, they underscore the need for further research with expanded sample sizes and sham-controlled designs. Crucially, the results of our study lend support to the precision medicine approach in the application of tDCS for individuals at high risk of dementia. By identifying specific AD risk factors, such as Aβ deposition and BDNF polymorphism, that modulate the response to tDCS, this research paves the way for more tailored and effective interventions. This is particularly relevant in addressing the complex challenges of managing and treating cognitive decline in dementia risk groups. In essence, the application of a Bayesian approach in our study not only facilitated the discovery of clinically significant insights that might have been undetected by traditional methods but also reinforced the importance of integrating various analytical perspectives. Such an approach is invaluable in advancing our understanding of intricate interactions in cognitive neuroscience, especially concerning neurodegenerative diseases such as AD, and in the development of precision medicine strategies.
Data availability
The datasets generated or analyzed during the current study are not publicly available due to the Patient Data Management Protocol of Yeouido Saint Mary’s Hospital but are available from the corresponding author upon reasonable request.
References
Scheltens, P. et al. Alzheimer’s disease. Lancet 388, 505–517 (2016).
Gauthier, S. et al. Mild cognitive impairment. Lancet 367, 1262–1270 (2006).
Anderson, R. M., Hadjichrysanthou, C., Evans, S. & Wong, M. M. Why do so many clinical trials of therapies for Alzheimer’s disease fail?. Lancet 390, 2327–2329 (2017).
Jean, L., Bergeron, M. E., Thivierge, S. & Simard, M. Cognitive intervention programs for individuals with mild cognitive impairment: Systematic review of the literature. Am. J. Geriatr. Psychiatry 18, 281–296 (2010).
Lautenschlager, N. T., Cox, K. & Kurz, A. F. Physical activity and mild cognitive impairment and Alzheimer’s disease. Curr. Neurol. Neurosci. Rep. 10, 352–358 (2010).
Singh, B. et al. Association of mediterranean diet with mild cognitive impairment and Alzheimer’s disease: A systematic review and meta-analysis. J. Alzheimers Dis. 39, 271–282 (2014).
Mintun, M. A. et al. Donanemab in early Alzheimer’s disease. N. Engl. J. Med. 384, 1691–1704 (2021).
van Dyck, C. H. et al. Lecanemab in early Alzheimer’s disease. N. Engl. J. Med. 388, 9–21 (2023).
Swanson, C. J. et al. A randomized, double-blind, phase 2b proof-of-concept clinical trial in early Alzheimer’s disease with lecanemab, an anti-Aβ protofibril antibody. Alzheimers Res. Ther. 13, 80 (2021).
Liu, C. S. et al. Using transcranial direct current stimulation to treat symptoms in mild cognitive impairment and Alzheimer’s disease. Neurodegener. Dis. Manag. 7, 317–329 (2017).
Medeiros, L. F. et al. Neurobiological effects of transcranial direct current stimulation: A review. Front. Psychiatry 3, 110 (2012).
Meinzer, M. et al. Transcranial direct current stimulation in mild cognitive impairment: Behavioral effects and neural mechanisms. Alzheimers Dement 11, 1032–1040 (2015).
Shin, D. W. et al. In vivo modulation of the blood-brain barrier permeability by transcranial direct current stimulation (tDCS). Ann. Biomed. Eng. 48, 1256–1270 (2020).
Luo, Y. et al. Transcranial direct current stimulation alleviates neurovascular unit dysfunction in mice with preclinical Alzheimer’s disease. Front. Aging Neurosci. 14, 857415 (2022).
Ferrucci, R. et al. Transcranial direct current stimulation improves recognition memory in Alzheimer disease. Neurology 71, 493–498 (2008).
Boggio, P. S. et al. Temporal cortex direct current stimulation enhances performance on a visual recognition memory task in Alzheimer disease. J. Neurol. Neurosurg. Psychiatry 80, 444–447 (2009).
Khedr, E. M. et al. A double-blind randomized clinical trial on the efficacy of cortical direct current stimulation for the treatment of Alzheimer’s disease. Front. Aging Neurosci. 6, 275 (2014).
Bystad, M. et al. Transcranial direct current stimulation as a memory enhancer in patients with Alzheimer’s disease: A randomized, placebo-controlled trial. Alzheimers Res. Ther. 8, 13 (2016).
Majdi, A., van Boekholdt, L., Sadigh-Eteghad, S. & Mc Laughlin, M. A systematic review and meta-analysis of transcranial direct-current stimulation effects on cognitive function in patients with Alzheimer’s disease. Mol. Psychiatry 27, 2000–2009 (2022).
Chen, J., Wang, Z., Chen, Q., Fu, Y. & Zheng, K. Transcranial direct current stimulation enhances cognitive function in patients with mild cognitive impairment and early/mid alzheimer’s disease: A systematic review and meta-analysis. Brain Sci. 12, 562 (2022).
Dubreuil-Vall, L., Chau, P., Ruffini, G., Widge, A. S. & Camprodon, J. A. tDCS to the left DLPFC modulates cognitive and physiological correlates of executive function in a state-dependent manner. Brain Stimul. 12, 1456–1463 (2019).
Manenti, R., Sandrini, M., Gobbi, E., Binetti, G. & Cotelli, M. Effects of transcranial direct current stimulation on episodic memory in amnestic mild cognitive impairment: A pilot study. J. Gerontol. B Psychol. Sci. Soc. Sci. 75, 1403–1413 (2020).
Inagawa, T. et al. A meta-analysis of the effect of multisession transcranial direct current stimulation on cognition in dementia and mild cognitive impairment. Clin. EEG Neurosci. 50, 273–282 (2019).
Isaacson, R. S. et al. The clinical practice of risk reduction for Alzheimer’s disease: A precision medicine approach. Alzheimers Dement. 14, 1663–1673 (2018).
Lukiw, W. J., Vergallo, A., Lista, S., Hampel, H. & Zhao, Y. Biomarkers for Alzheimer's Disease (AD) and the application of precision medicine. J. Pers. Med. 10 (2020).
Harrison, J. R., Mistry, S., Muskett, N. & Escott-Price, V. From polygenic scores to precision medicine in Alzheimer’s disease: A systematic review. J. Alzheimers Dis. 74, 1271–1283 (2020).
Chaieb, L., Antal, A., Ambrus, G. G. & Paulus, W. Brain-derived neurotrophic factor: Its impact upon neuroplasticity and neuroplasticity inducing transcranial brain stimulation protocols. Neurogenetics 15, 1–11 (2014).
Ng, T. K. S., Ho, C. S. H., Tam, W. W. S., Kua, E. H. & Ho, R. C. Decreased Serum Brain-Derived Neurotrophic Factor (BDNF) Levels in Patients with Alzheimer's Disease (AD): A Systematic Review and Meta-Analysis. Int. J. Mol. Sci. 20 (2019).
Ferretti, M. T. et al. Sex differences in Alzheimer disease—The gateway to precision medicine. Nat. Rev. Neurol. 14, 457–469 (2018).
Kang, D.-W. et al. Impact of transcranial direct current stimulation on cognitive function, brain functional segregation, and integration in patients with mild cognitive impairment according to amyloid-beta deposition and apoe ε4-allele: A pilot study. Brain Sci. 11, 772 (2021).
Puri, R. et al. Duration-dependent effects of the BDNF Val66Met polymorphism on anodal tDCS induced motor cortex plasticity in older adults: A group and individual perspective. Front. Aging Neurosci. 7, 107 (2015).
Kang, D. W. et al. Impact of transcranial direct current stimulation on white matter microstructure integrity in mild cognitive impairment patients according to effect modifiers as risk factors for Alzheimer’s disease. Front. Aging Neurosci. 15, 1234086 (2023).
Fehring, D. J. et al. Investigating the sex-dependent effects of prefrontal cortex stimulation on response execution and inhibition. Biol. Sex Differ. 12, 1–15 (2021).
Wagenmakers, E.-J. et al. Bayesian inference for psychology. Part I: Theoretical advantages and practical ramifications. Psychonomic Bull. Rev. 25, 35–57 (2018).
Temp, A. G. M. et al. A Bayesian perspective on Biogen’s aducanumab trial. Alzheimers Dement 18, 2341–2351 (2022).
Petersen, R. C. Mild cognitive impairment as a diagnostic entity. J. Internal Med. 256, 183–194 (2004).
Ku, H.-M. et al. A study on the reliability and validity of Seoul-Instrumental Activities of Daily Living (S-IADL). J. Korean Neuropsychiat. Assoc. 189–199 (2004).
Zimmerman, M., Martinez, J. H., Young, D., Chelminski, I. & Dalrymple, K. Severity classification on the Hamilton depression rating scale. J. Affect. Disorders 150, 384–388 (2013).
Cotelli, M. et al. Anodal tDCS during face-name associations memory training in Alzheimer’s patients. Front. Aging Neurosci. 6, 38 (2014).
Roncero, C. et al. Inferior parietal transcranial direct current stimulation with training improves cognition in anomic Alzheimer’s disease and frontotemporal dementia. Alzheimer’s Dementia Transl. Res. Clin. Intervent. 3, 247–253 (2017).
Lee, J., Kim, T., Kim, H. & Kim, D. Electric field strength-based anode positioning of transcranial direct current stimulation targeting at left dorsolateral prefrontal cortex: In silico study. Brain Stimul. Basic Transl. Clin. Res. Neuromodul. 16, 284 (2023).
Mendonca, M. E. et al. Transcranial DC stimulation in fibromyalgia: Optimized cortical target supported by high-resolution computational models. J. Pain 12, 610–617 (2011).
Lee, J. H. et al. Development of the Korean Version of the Consortium to Establish a Registry for Alzheimer’s Disease Assessment Packet (CERAD-K) clinical and neuropsychological assessment batteries. J. Gerontol. Ser. B Psychol. Sci. Soc. Sci. 57, P47–P53 (2002).
Jh, P. Standardization of Korean version of the Mini-Mental State Examination (MMSE-K) for use in the elderly. Part II. Diagnostic validity. Korean J. Neuropsych. Assoc. 28, 125–135 (1989).
Kim, T. Y. et al. Development of the Korean stroop test and study of the validity and the reliability. J. Korean Geriatrics Soc. 8, 233–240 (2004).
Thurfjell, L. et al. Automated quantification of 18F-flutemetamol PET activity for categorizing scans as negative or positive for brain amyloid: Concordance with visual image reads. J. Nucl. Med. 55, 1623–1628 (2014).
Li, Z., Shue, F., Zhao, N., Shinohara, M. & Bu, G. APOE2: Protective mechanism and therapeutic implications for Alzheimer’s disease. Mol. Neurodegenerat. 15, 63 (2020).
Carballedo, A. et al. Reduced fractional anisotropy in the uncinate fasciculus in patients with major depression carrying the met-allele of the Val66Met brain-derived neurotrophic factor genotype. Am. J. Med. Genet. B Neuropsychiat. Genet. 159, 537–548 (2012).
Han, E.-J. et al. Evidence for association between the brain-derived neurotrophic factor gene and panic disorder: A novel haplotype analysis. Psychiatry Invest. 12, 112 (2015).
Raz, N. et al. Regional brain changes in aging healthy adults: General trends, individual differences and modifiers. Cereb Cortex 15, 1676–1689 (2005).
van Doorn, J. et al. The JASP guidelines for conducting and reporting a Bayesian analysis. Psychon Bull. Rev. 28, 813–826 (2021).
Bhattacharjee, S. et al. Sex difference in tDCS current mediated by changes in cortical anatomy: A study across young, middle and older adults. Brain Stimul. 15, 125–140 (2022).
Diamond, A. Executive functions. Annu. Rev. Psychol. 64, 135–168 (2013).
Scarpina, F. & Tagini, S. The stroop color and word test. Front. Psychol. 8, 557 (2017).
Miller, E. K. & Cohen, J. D. An integrative theory of prefrontal cortex function. Annu. Rev. Neurosci. 24, 167–202 (2001).
Niendam, T. A. et al. Meta-analytic evidence for a superordinate cognitive control network subserving diverse executive functions. Cogn. Affect. Behav. Neurosci. 12, 241–268 (2012).
Satorres, E. et al. Home-based transcranial direct current stimulation in mild neurocognitive disorder due to possible Alzheimer’s disease. A randomised, single-blind, controlled-placebo study. Front. Psychol. 13, 1071737 (2022).
Gomes, M. A. et al. Transcranial direct current stimulation (tDCS) in elderly with mild cognitive impairment: A pilot study. Dement. Neuropsychol. 13, 187–195 (2019).
Tideman, P. et al. Association of β-amyloid accumulation with executive function in adults with unimpaired cognition. Neurology 98, e1525–e1533 (2022).
Saleh, O. et al. The effects of transcranial direct current stimulation on cognitive function for mild cognitive impairment: A systematic review and meta-analysis of randomized controlled trials. Aging Clin. Exp. Res. 35, 2293–2306 (2023).
Farrell, M. E. et al. Association of emerging β-amyloid and tau pathology with early cognitive changes in clinically normal older adults. Neurology 98, e1512–e1524 (2022).
Ossenkoppele, R. et al. Associations between tau, Aβ, and cortical thickness with cognition in Alzheimer disease. Neurology 92, e601–e612 (2019).
Baker, J. E. et al. Cognitive impairment and decline in cognitively normal older adults with high amyloid-β: A meta-analysis. Alzheimers Dement. (Amst.) 6, 108–121 (2017).
Saridin, F. N. et al. Brain amyloid β, cerebral small vessel disease, and cognition: A memory clinic study. Neurology 95, e2845–e2853 (2020).
Stagg, C. J. & Nitsche, M. A. Physiological basis of transcranial direct current stimulation. Neuroscientist 17, 37–53 (2011).
Stuchlik, A. Dynamic learning and memory, synaptic plasticity and neurogenesis: An update. Front. Behav. Neurosci. 8, 106 (2014).
Rossini, P. M., Ferilli, M. A., Rossini, L. & Ferreri, F. Clinical neurophysiology of brain plasticity in aging brain. Curr. Pharm. Des. 19, 6426–6439 (2013).
Egan, M. F. et al. The BDNF val66met polymorphism affects activity-dependent secretion of BDNF and human memory and hippocampal function. Cell 112, 257–269 (2003).
Toh, Y. L., Ng, T., Tan, M., Tan, A. & Chan, A. Impact of brain-derived neurotrophic factor genetic polymorphism on cognition: A systematic review. Brain Behav. 8, e01009 (2018).
Cechova, K. et al. Impact of APOE and BDNF Val66Met gene polymorphisms on cognitive functions in patients with amnestic mild cognitive impairment. J. Alzheimers Dis. 73, 247–257 (2020).
Pinto, T. C. C. et al. Is the montreal cognitive assessment (MoCA) screening superior to the mini-mental state examination (MMSE) in the detection of mild cognitive impairment (MCI) and Alzheimer’s Disease (AD) in the elderly?. Int. Psychogeriatr. 31, 491–504 (2019).
Song, C. S., Lee, H. S. & Chun, B. Y. Comparison of montreal cognitive assessment in korean version for predicting mild cognitive assessment in 65-year and over individuals. Occup Ther. Int. 2022, 4108434 (2022).
Bhattarai, P. et al. Predicting cognitive dysfunction and regional hubs using Braak staging amyloid-beta biomarkers and machine learning. Brain Inform. 10, 33 (2023).
Stevens, D. A. et al. Regional amyloid correlates of cognitive performance in ageing and mild cognitive impairment. Brain Commun. 4, fcac016 (2022).
Dedoncker, J., Brunoni, A. R., Baeken, C. & Vanderhasselt, M. A. A Systematic review and meta-analysis of the effects of transcranial direct current stimulation (tDCS) over the dorsolateral prefrontal cortex in healthy and neuropsychiatric samples: influence of stimulation parameters. Brain Stimul. 9, 501–517 (2016).
Jack, C. R. Jr. et al. NIA-AA research framework: Toward a biological definition of Alzheimer’s disease. Alzheimer’s Dementia 14, 535–562 (2018).
Funding
This research was supported by a National Research Foundation of Korea (NRF) grant funded by the Korean government (Ministry of Science and ICT) (No. 2019R1C1C1007608), a grant from the Korea Health Technology R&D Project through the Korea Health Industry Development Institute (KHIDI) funded by the Ministry of Health & Welfare, Republic of Korea (No. HI22C0467), Basic Medical Science Facilitation Program through the Catholic Medical Center of the Catholic University of Korea funded by the Catholic Education Foundation, and a research fund of Seoul St. Mary’s Hospital, the Catholic University of Korea (No. ZC23TISI0860).
Author information
Authors and Affiliations
Contributions
D.W.: Conceptualization, methodology, data curation, writing–original draft, visualization, formal analysis, and funding acquisition. S.M.: Methodology, data curation, writing, reviewing, and editing. Y.H.: Software, Investigation. S.: Methodology, Data Curation. T.: Methodology, Data Curation. D.: Methodology, Data Curation. C.U.: Conceptualization, Supervision. H.K.: conceptualization, methodology, writing review and editing, supervision, and project administration.
Corresponding author
Ethics declarations
Competing interests
Hyun Kook Lim, TaeYeong Kim, and Donghyeon Kim were employed by NEUROPHET Inc. The remaining authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as potential conflicts of interest. The data processing services provided by NEUROPHET Inc. were used to enhance the quality and analysis of the brain imaging data collected during the study. The authors declare that the research outcomes and conclusions remain unbiased and are not influenced by any commercial interests associated with NEUROPHET Inc.'s products or services.
Additional information
Publisher's note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Supplementary Information
Rights and permissions
Open Access This article is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License, which permits any non-commercial use, sharing, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if you modified the licensed material. You do not have permission under this licence to share adapted material derived from this article or parts of it. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by-nc-nd/4.0/.
About this article
Cite this article
Kang, D.W., Wang, SM., Um, Y.H. et al. Effects of transcranial direct current stimulation on cognition in MCI with Alzheimer's disease risk factors using Bayesian analysis. Sci Rep 14, 18818 (2024). https://doi.org/10.1038/s41598-024-67664-9
Received:
Accepted:
Published:
DOI: https://doi.org/10.1038/s41598-024-67664-9
Keywords
Comments
By submitting a comment you agree to abide by our Terms and Community Guidelines. If you find something abusive or that does not comply with our terms or guidelines please flag it as inappropriate.
