Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2022 Apr 1.
Published in final edited form as: Biol Psychiatry Cogn Neurosci Neuroimaging. 2020 Nov 25;6(4):439–448. doi: 10.1016/j.bpsc.2020.11.006

Transcranial Direct Current Stimulation to the Left Dorsolateral Prefrontal Cortex Improves Cognitive Control in Patients With Attention-Deficit/Hyperactivity Disorder: A Randomized Behavioral and Neurophysiological Study

Laura Dubreuil-Vall 1, Federico Gomez-Bernal 1, Ana C Villegas 1, Patricia Cirillo 1, Craig Surman 1, Giulio Ruffini 1, Alik S Widge 1, Joan A Camprodon 1
PMCID: PMC8103824  NIHMSID: NIHMS1692132  PMID: 33549516

Abstract

BACKGROUND:

Attention-deficit/hyperactivity disorder (ADHD) is a neurodevelopmental disorder associated with significant morbidity and mortality that may affect over 5% of children and approximately 2.8% of adults worldwide. Pharmacological and behavioral therapies for ADHD exist, but critical symptoms such as dysexecutive deficits remain unaffected. In a randomized, sham-controlled, double-blind, crossover mechanistic study, we assessed the cognitive and physiological effects of transcranial direct current stimulation (tDCS) in 40 adult patients with ADHD in order to identify diagnostic (cross-sectional) and treatment biomarkers (targets).

METHODS:

Patients performed three experimental sessions in which they received 30 minutes of 2 mA anodal tDCS targeting the left dorsolateral prefrontal cortex, 30 minutes of 2 mA anodal tDCS targeting the right dorsolateral prefrontal cortex, and 30 minutes of sham. Before and after each session, half the patients completed the Eriksen flanker task and the other half completed the stop signal task while we assessed behavior (reaction time, accuracy) and neurophysiology (event-related potentials).

RESULTS:

Anodal tDCS to the left dorsolateral prefrontal cortex modulated cognitive (reaction time) and physiological (P300 amplitude) measures in the Eriksen flanker task in a state-dependent manner, but no effects were found in the stop signal reaction time of the stop signal task.

CONCLUSIONS:

These findings show procognitive effects in ADHD associated with the modulation of event-related potential signatures of cognitive control, linking target engagement with cognitive benefit, proving the value of event-related potentials as cross-sectional biomarkers of executive performance, and mechanistically supporting the state-dependent nature of tDCS. We interpret these results as an improvement in cognitive control but not action cancellation, supporting the existence of different impulsivity constructs with overlapping but distinct anatomical substrates, and highlighting the implications for the development of individualized therapeutics.

Attention-deficit/hyperactivity disorder (ADHD) is associated with functional impairment and high morbidity and mortality in youth and adulthood (1). Epidemiologic studies suggest that ADHD may affect over 5% of children and approximately 2.8% of adults worldwide (2,3). While there is emerging evidence that available psychopharmacology and cognitive behavioral therapy interventions can address executive functioning deficits in individuals with ADHD (4–7), these have been and still remain critical symptoms closely associated with functional impairment yet with suboptimal (or null) response to current therapies; further research could identify the place of transcranial direct current stimulation (tDCS) as an alternative or complementary intervention.

tDCS is emerging as a promising tool in human neuroscience research and for the treatment of neuropsychiatric disorders and of dysexecutive syndromes in particular (8,9). Previous studies show that tDCS targeting the dorsolateral prefrontal cortex (DLPFC) modulates domains of executive function (10,11), specifically those affected in ADHD (12). None of these studies, however, combined behavioral and physiological measures.

The Eriksen flanker task (EFT) (13) and the stop signal task (SST) (14) are well-established experimental tasks used to assess impaired executive functions in ADHD (15,16). Although both capture inhibitory control processes, the EFT primarily assesses interference cognitive control (the ability to resist or resolve distracting interference that is irrelevant to the task), while the SST measures action cancellation (the ability to suppress dominant, automatic, already-initiated responses) (17–19).

Human electrophysiological studies assessing event-related potentials (ERPs) have established relevant signatures of executive function during these tasks. Specifically, the P200, N200, P300, and error-related negativity (ERN)/error-related positivity (Pe) characterize the attentional and inhibitory functions that break down during conflict owing to dysexecutive deficits and impulsivity (20–24). Previous literature also found associations between ADHD symptom scores, ERP amplitudes, and poorer task performance in ADHD, which supports the use of these ERPs as correlates of executive function in ADHD (24).

The P200 is an early component that usually appears in all trials of the EFT and the SST in the range of 150 to 275 ms. Although there is a wide range of factors affecting the characteristics of the P200, its amplitude usually reflects a more basic level of selective attentional processing of visual stimuli (25,26). The N200 is a negative-going wave that usually peaks in the incongruent trials of the flanker task in a range of 200 to 350 ms poststimulus. Although there are mixed results in the literature for the N200 component in ADHD (27), its amplitude is generally related to the degree of conflict prompted by a given stimulus, or the extent to which individuals are distracted by task-irrelevant (flanker) information compared with task-relevant (target stimulus) information (20,28), requiring greater deployment of attentional resources. The P300 appears 250 to 500 ms after the stimulus and reflects the conflict postprocessing and behavioral inhibition of the incorrect prepotent response in incongruent trials of the EFT and Stop trials of the SST (21,29,30). The ERN and Pe are response-locked ERPs that appear in both the EFT and the SST. The ERN is a negative deflection in the ERP that occurs following error commission, and it is time-locked to an individual’s response. It typically peaks between 0 and 150 ms after the erroneous response begins and is thought to be a marker of response conflict that occurs during error commission (22,31). The ERN is often followed by a positive peak (Pe), a positive deflection that can peak 100 to 200 ms after making the incorrect response. The Pe amplitude is thought to reflect the perception or recognition of the error: the more awareness of the error, the larger the amplitude (23,32).

In this study, we tested 40 patients with ADHD during three experimental visits and compared the effect of anodal tDCS targeting the left DLPFC versus anodal tDCS targeting the right DLPFC versus sham. Immediately before and after tDCS, half of the patients performed the EFT and the other half performed the SST, while we measured behavioral (reaction time [RT] and accuracy) and neurophysiological (ERPs) responses. Our previous research showed that tDCS targeting the left DLPFC in healthy adults led to a significant decrease in RT correlated with a modulation of N200 and P300 amplitude in the flanker task (11). Our aims here were to assess 1) the role of DLPFC laterality in ADHD deficits in interference cognitive control (EFT) and action cancellation (SST), 2) the physiological dynamics sustaining the modulation of executive function by tDCS, and 3) the impact of state-dependent dynamics of tDCS effects.

METHODS AND MATERIALS

Trial Design

A randomized, sham-controlled, double-blind, crossover study was performed at Massachusetts General Hospital (Boston, MA). Recruitment started in July 2015 and ended in March 2018. The study was approved by the Partners HealthCare Institutional Review Board and is registered on ClinicalTrials.gov (NCT04175028). The full protocol is available upon request.

Participants

Forty-four adult patients with a primary diagnosis of ADHD were recruited from the Division of Neuropsychiatry, the Behavioral Neurology Unit, and the Adult ADHD Research Program at Massachusetts General Hospital and randomized. See Table 1 for demographic and clinical characteristics and see Table S1 for inclusion/exclusion criteria. All participants gave informed and written consent for participation.

Table 1.

Participant Characteristics

EFT SST
Demographics
 Age, Years, Mean (SD) 43.85 (14.78) 31.2 (13)
 Female, n (%) 10 (50) 10 (50)
 White, n (%) 16 (80) 16 (80)
 Single, n (%) 10 (50) 15 (75)
Baseline Score, Mean (SD)
 CHRT 0.6 (1.89) 0.1 (0.45)
 PHQ9 10.05 (8.05) 6.7 (5.6)
 QIDS 9.33 (6.14) 10.29 (5.45)
 ASRS 62.6 (9.17) 58.20 (19.58)
Prior Medication, n (%)a
 No medication 11 (55) 7 (35)
 Adderall 2 (15) 4 (20)
 Vyvanse 2 (10) 3 (15)
 Claritin 0 1 (5)
 Concerta 1 (5) 1 (5)
 Lisinopril 0 1 (5)
 Lamictal 0 1 (5)
 Nortryptiline 0 1 (5)
 Verapamil 1 (5) 0
 Aspirin 1 (5) 0
 Levothyroxine 1 (5) 0
 Modafinil 1 (5) 0
Open in a new tab

ASRS, Adult ADHD Self-Report Scale; CHRT, Concise Health Risk Tracking scale; EFT, Eriksen flanker task; PHQ9, Patient Health Questionnaire-9; QIDS, Quick Inventory of Depressive Symptomatology; SST, stop signal task.

a

Patients either were off stimulant medications or, if undergoing treatment with stimulants, were asked to discontinue 2 days prior to the experiment, under physician-guided protocol, and allowed to resume afterward.

Intervention

For each session, 2 mA of anodal stimulation was applied for 30 minutes with Ag/AgCl electrodes (contact area = 3.14 cm2) using the hybrid tDCS-electroencephalography (EEG) Starstim system (Neuroelectrics, Boston, MA) (also used for EEG recording). The duration of the ramp-up and ramp-down at the beginning and the end of the stimulation was set to 15 seconds. During the stimulation period, the subject was instructed to sit and relax with eyes open. See Figure 1 for details about the stimulation montage.

Figure 1.

Figure 1.

Open in a new tab

Modeling of the normal component of the electrical field (V/m) created by the montage targeting the left dorsolateral prefrontal cortex (DLPFC) and right DLPFC. Specifically, the anodal electrode was placed on the scalp at the F4 (for right DLPFC stimulation) or F3 (for left DLPFC stimulation) positions, according to the International 10–20 system for electroencephalography. The cathode was placed in the contralateral supraorbital region (Fp1 or Fp2). The four electrodes were always placed at both sides for all stimulation conditions (left, right, and sham) to ensure the blinding of the patient and the operator. For the sham condition, the current was applied only for the 15-second ramp-up phase at the beginning and the end of a 30-minute sham stimulation period, to simulate the potential experience of local tingling sensation that real stimulation produces but without sustained effect on cortical activity. The stimulation is usually not noticeable between the ramp-up and the ramp-down for either active or sham transcranial direct current stimulation, thus ensuring the blinding of the patient. The modeling is based on a finite element model included in the Starstim’s software NIC.

The order of stimulation administration (sham, left, or right) was randomized across subjects using a permutation-based randomization list generated by a computer to avoid any confounding order effects across sessions. The experimenter and the subject were blinded by using the double-blind mode in Starstim’s software NIC, which blinds the user to the type of stimulation used (active tDCS targeting left/right DLPFC or sham stimulation) after a 4-digit password is introduced by the administrator.

Outcomes

Immediately before and after tDCS, half of the patients (n = 20) completed the EFT (Figure 2A), in which subjects must respond to the direction of a central arrow that is surrounded (“flanked”) by distracting arrows that can have either the same orientation (congruent trials) or the opposing orientation (incongruent trials) as the central one. Participants were instructed to press the left or right arrow button following the direction of the central arrow, ignoring the flanker arrows. The accuracy of correct/incorrect responses and the RT for each stimulus were measured.

Figure 2.

Figure 2.

Open in a new tab

Flanker task and stop signal task scheme. (A) The flanker task consisted of 140 trials in two blocks of 70. Each subject had a different, fully random sequence of congruent and incongruent trials, with two congruent trials for each incongruent trial, in order to build a tendency toward congruent responses and thus increase the difficulty of conflict detection in incongruent trials. The task had a total duration of 10 minute, with 1 minute of training before the task started. The flanker arrows were first presented alone for duration of 136 ms, 114 ms, 92 ms, 70 ms, or 48 ms, and were then joined by the target arrow for 62 ms, 52 ms, 42 ms, 32 ms, or 22 ms, respectively (values were adjusted to the psychometric “sweet spot” in which each patient achieved a performance in the range of 80%–85%). These values were calibrated just once at the first session for each participant to avoid confounding the outcomes, so the same values were used for all sessions within participants. Stimulus presentation was followed by a black screen for 500 ms. The time window for participants’ response was 600 ms after target onset. If the participant did not respond within the response window, a screen reading “TOO SLOW!” was presented for 300 ms. Participants were told that if they saw this screen, they should speed up. If a response was made before the deadline, the “TOO SLOW!” screen was omitted, and the black screen remained on screen for the 300-ms interval. Finally, each trial ended with presentation of the fixation cross for an additional randomly chosen duration (200, 300, or 400 ms) in order to avoid any habituation or expectation by the subject. Thus, trial durations varied between 1070 and 1400 ms. (B) The stop signal task consisted of 160 Go trials (80%) and 40 Stop trials (20%). There were only two types of Go trials: “A” and “Z.” The “A” or “Z” stimuli were first presented for 100 ms and they were followed by a black screen for 500 ms. Patients had to press the left mouse button whenever the “A” stimulus was presented and press the right mouse button whenever the “Z” stimulus was presented. For the Stop trials, the stop signal initially appeared 400 ms after the “A” or “Z” stimuli, and was adjusted dynamically according to the subject’s performance, increasing or decreasing by 50 ms after a successful or unsuccessful answer, respectively, within a range of 50–500 ms in order to yield approximately 50% successful inhibition of the Stop trials (Figure S1). RT, reaction time.

The other half of the patients (n = 20) performed the SST (Figure 2B), in which participants had to provide a response as quickly as possible when letters “Z” or “A” appear (Go trial). However, in some trials, the “A” or “Z” stimuli were followed by the stop signal “X” (Stop trials), which appeared with varying adaptive delays from the Go stimulus. In these trials, participants must withhold their response. We measured the accuracy of correct/incorrect responses, the RT for Go trials, and the time it takes for the participant to withhold their response in the Stop trials (stop signal RT [SSRT]). See the Supplement for more details.

During the tasks, EEG was recorded from 7 positions (Fp1, Fp2, F3, Fz, F4, P3, and P4) with Ag/AgCl electrodes at a sampling frequency of 500 samples/s. EEG data were referenced to the right mastoid. Offline processing was then performed using EEGlab (version 13.5.4b) (33). Independent component analysis was utilized to identify and remove activity associated with blinks, eye movements, and other artifacts. Data were filtered from 1 to 20 Hz to remove nonneural physiological activity (skin/sweat potentials) and noise from electrical outlets. Trials were epoched within a time frame of 200 ms before and 800 ms after the stimulus onset. The mean of the prestimulus baseline [−200, 0] ms was then subtracted from the entire ERP waveform for each epoch to eliminate any voltage offset. After rejecting trials that had at least a sample above ±150 μV, the remaining trials were averaged for each time point and stimulation condition.

Statistical Analysis

Based on similar studies (11,34), we estimated a sample size of 20 subjects for each block of tasks to provide 85% power to detect an effect size of d = 0.6 (α = 5%), while 25 subjects would provide 90% power and 30 subjects would provide 95% power for the same estimated effect size and α = 5%. We estimated a 20% attrition rate based on previous studies.

RT, accuracy of correct responses, ERP amplitudes, cross-sectional biomarkers, and state dependencies were all modeled and analyzed using generalized linear models with mixed effects. See the Supplement and Tables S2 to S8 for more details.

RESULTS

As shown in Figure S0, from the 20 patients assigned to each task block for analysis, 2 patients from the EFT and 1 patient from the SST were discarded as outliers due to extreme movement artifacts in the EEG data, thus leaving 18 patients in the EFT group and 19 in the SST group. Attrition rate was lower than anticipated. No important harms or unintended side effects were reported.

Flanker Task

Cognitive Results.

There was a significant stimulation type × time point × trial type interaction in RT (β = −9.99 ms; 95% confidence interval [CI], 3.50 to 16.48 ms; p = .03), indicating that the stimulation type × time point interaction was significantly different for incongruent versus congruent trials. After post hoc tests, we found that this difference resulted from the lack of significant changes in congruent trials for any of the stimulation conditions (Figure S2A), while for incongruent trials, left-sided stimulation led to a significantly faster RT compared with sham (left/sham × PRE/POST [β = −16.1 ms; 95% CI, −22.8 to −9.3 ms; p < .0001]), and right-sided stimulation did not have any significant effect compared with sham (right/sham × PRE/POST [β = −5.3 ms; 95% CI, −13.7 to 3.1 ms; p = .390]) (Figure 3A). The effect of left-sided stimulation was also significantly greater than right-sided stimulation (left/right × PRE/POST [β = 10.7 ms; 95% CI, 1.26 to 20.2 ms; p =.0183]). None of the stimulation conditions led to significant changes in accuracy compared with sham, for both incongruent and congruent trials (Figure 3B and Figure S2B).

Figure 3.

Figure 3.

Open in a new tab

Flanker task results. (A) Mean reaction time (RT) and (B) accuracy for incongruent trials and p values with multivariate correction. Error bars indicate confidence intervals. Significance indicated as *p < .05, ***p < .001. (C) Grand average event-related potentials time-locked to incongruent stimuli. Waveforms correspond to the average of the F3, Fz, and F4 positions. The red circle indicates the significant amplitude changes compared with sham. See Figure S3 for event-related potentials at individual electrodes. (D) Scalp topographies of POST-PRE difference of P200, N200, and P300 amplitude (μV). Averaging time window for P300 = [260, 360] ms. Averaging time window for N200 = [180, 230] ms.

Event-Related Potentials.

Both left-sided stimulation (β = 2.15 μV; 95% CI, 0.31 to 3.99 μV; p = .022) and right-sided stimulation (β = 2.37 μV; 95% CI, 0.53 to 4.20 μV; p = .011) led to a significant P300 amplitude increase compared with sham for incongruent trials (Figure 3C).

There were no significant changes in the N200 after left-sided or right-sided stimulation compared with sham, but the reduction in N200 amplitude after left-sided stimulation was significantly different from right-sided stimulation (β = −2.43 μV; 95% CI, −4.64 to −0.22 μV; p = .027). Note that most P200, N200, and P300 amplitude changes occurred primarily around the area of the anodal stimulation electrode (F3 or F4), matching the laterality of the stimulation, especially for left-sided stimulation (Figure 3D).

There were no significant changes in P200 amplitude for incongruent (Figure 3B) or congruent (Figure S4A) trials. The ERN and Pe did not show significant differences either (Figure S4B).

ERP Cross-Correlation With RT.

The amplitudes of P200, N200, and P300 were significantly correlated with RT for incongruent trials in a cross-sectional trial-by-trial basis: the greater the P200 amplitude (β = −0.26 ms/μV; 95% CI, −0.51 to −0.004 ms/μV; p = .046) and the P300 amplitude (β = −0.25 ms/μV; 95% CI, −0.50 to −0.004 ms/μV; p = .046), the faster the RT, and the smaller the N200 amplitude, the faster the RT (β = −0.54 ms/μV; 95% CI, −0.79 to −0.29 ms/μV; p < .0001).

State Dependencies.

Table S9 shows the effect of variables at baseline (before stimulation) on the change on the same (and other) variables after stimulation (i.e., state-dependent relationships). Figure 5 shows the scatterplots of the significant predictors. The change in P300 and P200 after stimulation is conditioned by the amplitude of P300 and P200 at baseline, while the change in N200 is conditioned by the amplitude of N200 and P300 at baseline. We also found no significant differences in these relationships before versus after stimulation, indicating that tDCS did not significantly modulate the relationship between RT and ERP amplitudes (RT × P200 [β = 0.72; 95% CI, −0.79 to 2.32; p = .374], RT × P300 [β = 1.29; 95% CI, −0.33 to 2.91; p = .121], RT × N200 [β = −0.79; 95% CI, −2.22 to 0.87; p = .295]).

Figure 5.

Figure 5.

Open in a new tab

State dependencies. Scatterplots, regression lines, and confidence intervals for significant state dependencies in the Eriksen flanker task (EFT). (Top row) Change in P300 as a function of P300 (left) and N200 (right) at baseline in the EFT. (Middle row) Change in N200 amplitude as a function of P300 (left) and N200 (right) amplitude at baseline in the EFT. (Bottom row) Change in P200 amplitude as a function of P300 (left) and P200 (right) amplitude at baseline in the EFT. tDCS, transcranial direct current stimulation.

Stop Signal Task

Cognitive Results.

The RT for Go trials significantly increased after left-sided stimulation compared with sham (β = 8.32 μV; 95% CI, 2.18 to 14.47 μV; p = .0044) (Figure 4A), but there were no significant changes in the SSRT for Stop trials (Figure 4B). There were also no significant changes in accuracy for any of the stimulation conditions and for any of the trial types (Figure S5).

Figure 4.

Figure 4.

Open in a new tab

Stop signal task results. (A) Mean reaction time (RT) for Go trials. (B) Stop signal reaction time (SSRT) for Stop trials. Error bars indicate confidence intervals. Significance indicated as *p < .05, **p < .01. (C) Grand average event-related potentials time-locked to Go trials. The red circle indicates the significant amplitude changes compared with sham. (D) Grand average event-related potentials time-locked to Stop trials. Waveforms correspond to the average of the F3, Fz, and F4 positions.

Event-Related Potentials.

For Go trials, P200 amplitude significantly increased after left-sided stimulation compared with sham (β = 0.51 μV; 95% CI, 0.09 to 0.92 μV; p = .0160) (Figure 4C). For Stop trials, there were no significant changes in the P200, N200, and P300 (Figure 4D). There were no significant changes in the ERN or Pe (Figure S6).

ERP Cross-Correlation With RT.

The amplitude of P200 was also significantly correlated with RT for Go trials—i.e., the greater the P200 amplitude, the slower the RT (β = 1.08 ms/μV; 95% CI, 0.69 to 1.47 ms/μV; p < .001).

State Dependencies

No significant state dependencies were found for the SST (Table S9).

DISCUSSION

Flanker Task

Our results confirmed that anodal tDCS targeting the DLPFC improves RT in the EFT in patients with ADHD, similar to what we previously described in healthy control subjects (HCs) (11). Specifically, we describe that anodal tDCS targeting the left DLPFC results in a significant reduction in RT in incongruent trials, compared with a nonsignificant change after sham or right DLPFC anodal modulation. Compared with our previous study with HCs, the baseline RT (before stimulation) was slower in patients with ADHD, and the size effect of the improvement in RT after left-sided stimulation was larger for ADHD (β = −16.1 ms) than for HCs (β = −8.37 ms).

The faster RT of incongruent trials in patients with ADHD was correlated with a significant increase in P300 amplitude, in this case for both left- and right-sided anodal tDCS. Larger P300 amplitudes were associated with effective conflict postprocessing and cognitive control, with the subsequent behavioral inhibition of incorrect prepotent responses (21,29,35). We thus interpret the increase in P300 amplitude as a modulation of conflict resolution and interference control processes, leading to more efficient inhibition of distractors and competing responses (i.e., faster RT). Given that the P300 increased significantly after left- and right-sided stimulation, but behavior (RT) only changed after left-sided stimulation, we hypothesize that the physiological effect of right-sided stimulation was not sufficient to trigger a significant behavioral change, suggesting a greater role for the left DLPFC in the modulation of executive function.

Although in previous research with HCs we found a significant decrease in N200 amplitude after left-sided stimulation (11), indicative of an improvement in selective attention in a context of conflict resolution, in the current study the decrease in N200 amplitude was not significant compared with sham. Similarly, previous studies could not reliably confirm between-group differences for the N200 component, as they found heterogeneous results for N200 alterations in ADHD (27). While this may be partially explained by the higher intraindividual variability in populations with ADHD than in control populations (36), it highlights the need to better understand the underlying physiological differences between patients with ADHD and HCs leading to different tDCS effects.

McGough et al. (37) also found executive function improvements using a transcranial nerve stimulation protocol designed to primarily target cutaneous nerves. Both this stimulation and our tDCS intervention likely activate the prefrontal cortex (directly or indirectly via activation of brainstem nuclei), although transcranial nerve stimulation is bilateral and uses a very different temporal pattern, making the direct comparison between transcranial nerve stimulation and tDCS not trivial. It may be interesting in future studies to test whether these approaches have common mechanisms.

Stop Signal Task

In the SST, proactive inhibition is defined as the advanced preparation to halt action in the anticipation of an imminent stop signal in Go trials, requiring greater selective attention in the visual search for the stop signal to appear. Reactive inhibition is defined as the performance of outright stopping in response to the appearance of a stop signal in Stop trials (38). In the current study, we found that tDCS to the left DLPFC led to a significant increase in the time patients withheld their response in Go trials waiting for the stop signal to appear, which was correlated with a significant increase in P200 amplitude. There is a wide range and diversity of factors that have been found to affect the characteristics of the P200, but its amplitude is generally associated with selective attention to visual stimuli (25). We thus interpret the increase in P200 amplitude as a modulation in selective attention when searching for the stop signal to appear, with the subsequent improvement in proactive inhibition. However, the lack of significant changes in the SSRT suggests no effects on reactive inhibition. Although there have been positive results in tDCS studies using the SST in a healthy population targeting other areas (39–45), previous tDCS studies with patients with ADHD using the SST and targeting the DLPFC have also found a lack of significant effects on the SSRT and accuracy (34,46). These results support the formulation of inhibitory control and impulsivity as complex multimodal processes with subdomai n specificity (e.g., impulsivity of thought, action, affect) captured by different experimental tasks and with different anatomical representation (47). These findings confirm the hypothesis that that there is a dissociation between action cancellation or the ability of suppressing prepotent responses that have already been initiated, captured by the SST, and interference cognitive control or the ability of resisting distractors and resolving interference between competing responses, captured by the EFT (19,48–53). From a translational perspective, this also implies that the DLPFC may be a good substrate to improve interference cognitive control and proactive inhibition, but if the goal is to improve action cancellation (e.g., tics, compulsions), one should consider alternative windows into the circuitry (18,47).

Cross-Sectional Biomarkers

Our findings indicate that the amplitude of P200, N200, and P300 on a trial-by-trial basis is correlated cross-sectionally with RT in the EFT [as previously described in HCs (11)] and the SST, thus supporting the interpretation of the observed tDCS effects (i.e., small N200 and large P300 are associated with more adaptive responses, and thus changes in this direction should be therapeutic) and highlighting their value as a potential diagnostic, monitoring, or surrogate biomarker (54) for cognitive performance.

State Dependence

Our results indicate that the effect of tDCS in the flanker task depends on the individual’s electrophysiological state at baseline (before stimulation). Specifically, we found that low baseline P300 amplitudes and small N200 baseline amplitudes (associated with impaired cognitive performance) are correlated with greater P300 amplitude increases (and N200 decreases) after tDCS, which are associated with improvement in cognitive performance. This suggests that tDCS leads to greater modulation (improvement) of physiology in subjects with baseline physiological signatures indicative of less adaptive processing, as they allow greater range of modulation. As expected, sham did not show any significant state dependencies, suggesting that potential effects were not due to regression to the mean. However, we note that these are just correlations and not a definitive proof of causality.

These effects are possibly explained by the principle of state dependency, a phenomenon by which the response of a system to an external intervention is affected not only by the properties of that intervention (e.g., stimulation parameters), but also by the internal state of the system. The state-dependent characteristics of tDCS have important implications for treatment development: beyond stimulation parameters, clinical trials may benefit from controlling patients’ state (before, during, or after stimulation) to minimize response variability and maximize the therapeutic effects of tDCS.

Limitations

Our results show modulation of executive function in the context of highly controlled experimental tasks, and therapeutic benefits should be confirmed in future clinical trials using relevant clinical and functional outcome measures after several repeated tDCS sessions aiming to induce longer-lasting procognitive and pro-executive plastic changes. Because we did not control for handedness, a remaining question is whether stimulation to the dominant hemisphere in left-handed individuals would be a confound in the results, which should be addressed in future studies. In addition, future trials should also assess the effects of the duration of stimulation as well as changes in executive functions and ERP with ADHD symptom outcomes.

We also acknowledge a significant age difference (EFT: 43.85 ± 14 years; SST: 31.2 ± 13 years; p = .0073) between the EFT and the SST groups; the lack of appropriately age-matched groups resulted because the two cohorts were recruited prospectively in different time frames (though with the same exact protocol and hardware) and then analyzed together retrospectively to address the proposed questions. While mean ages are well after periods of brain maturation when myelination patterns and ADHD symptoms are thought to be persistently established, and well before a geriatric threshold when other types of biological changes (including normal aging) may affect cognition, the wide age range (18–67 years of age) may also introduce some heterogeneity in the results. Thus, future prospective validation studies should use larger homogeneous cohorts, with more restricted age ranges and randomized age-matched groups.

Because there is a modest correlation between neuro-psychologically determined executive deficits and molar measures such as the Behavior Rating Inventory of Executive Function, the sample is likely to include patients with fewer impairments in those functions measured by the EFT and SST tasks, which may obscure therapeutic effects in those with cognitive impairments. Future clinical trials should include a more homogeneous sample of individuals with impairments in those functions measured by the EFT and SST, as tDCS effects were most significant for those individuals.

We did not perform any assessment of the effectiveness of the patient blinding, as the sham protocol used in the current study has been proven to be effective (55), but we acknowledge the need to include this type of assessment in future trials.

For the EFT group, it is also worth noting that although the 16-ms decrease in RT after left-sided stimulation compared with sham was significant (p = .0001) and greater than the 8 ms we found in HCs (11), we acknowledge that it may still be considered small and could be affected by the ADHD heterogeneity, estimation errors, and transformations to the data. While the timing precision of the Presentation software (https://www.neurobs.com/menu_presentation/menu_features/features_overview) is <0.1 ms and thus should not introduce significant estimation errors, we plotted the effects at the individual level (Figure S7) to discard other sources of errors. This figure shows that while there was a high between-subject variability (as expected owing to the inherent ADHD heterogeneity), within-subject standard errors were very small for most patients, and there was a relatively reliable individual-level effect that was not outlier-driven, thus supporting our conclusions. However, further studies with larger samples should be carried out in order to minimize between-subjects variability and other potential sources of errors.

While the field has established differences between the constructs captured by the EFT and the SST, the nomenclature used to define the overlap and differences at the cognitive and behavioral levels remains equivocal and often contradictory. This seems to be more than a simple problem of semantics and reflects deficits in the core formulation of the subtleties across executive constructs. We have opted for descriptive terms previously used in the literature (e.g., interference cognitive control and action cancellation) but acknowledge that other terminologies may be considered.

Conclusions

This study indicates that anodal tDCS over the left DLPFC using a simple bipolar montage has procognitive effects in dysexecutive patients with ADHD associated with the modulation of physiological signatures of cognitive control (i.e., treatment target), supporting specific hypotheses and stratgies for neuromodulation treatment development under an experimental therapeutics framework aiming to link target engagement (cognitive and physiological) with clinical benefit. In addition, we provide mechanistic support for the state-dependent nature of the effects of tDCS, highlighting the importance of controlling (or at least measuring) the neural states before (and possibly during) stimulation as a relevant therapeutic strategy beyond choices regarding the neuromodulation parameter space. We also provide empirical evidence supporting the value of the P200, N200, and P300 as cross-sectional biomarkers of cognitive performance across tasks, and across populations if taken together with our previous similar report in healthy subjects (11). Last, we interpret these results as an improvement in interference cognitive control (captured by the EFT) but not in action cancellation (assessed by the SST), supporting the hypothesis of the existence of different impulsivity constructs with overlapping but distinct anatomical substrates and therapeutic strategies.

Supplementary Material

Supplementary

ACKNOWLEDGMENTS AND DISCLOSURES

This work was supported by National Institutes of Health Grant Nos. R01 MH112737 (to JAC), R21 DA042271 (to JAC), R21 AG056958 (to JAC), and R21 MH115280 (to JAC).

Footnotes

A previous version of this article was published as a preprint on medRxiv: doi: https://doi.org/10.1101/2020.01.13.20017335.

GR is co-founder of Neuroelectrics, a company that manufactures the tDCS technology used in the study. LD-V is an employee at Neuroelectrics. JAC is a member of the scientific advisory board for Apex Neuroscience Inc. AW has patent applications pending related to cognitive enhancement through brain stimulation and new methods of transcranial electrical stimulation. CS reports that, within the past 12 months, he has received research support from Shire/Takeda Pharmaceuticals; has served as a consultant to Adlon, Shire, Sunovion, Supernus, and Teva Pharmaceuticals; and has received book royalties for “Fast Minds: How to Thrive If You Have ADHD (or Think You Might) as well as ADHD in Adults: A Practical Guide to Evaluation and Management. All other authors report no biomedical financial interest or potential conflicts of interest.

ClinicalTrials.gov: Neuromodulation of Executive Function in the ADHD Brain; https://clinicaltrials.gov/ct2/show/NCT04175028; NCT04175028.

REFERENCES

  • 1.Barkley RA, Fischer M (2019): Hyperactive child syndrome and estimated life expectancy at young adult follow-up: The role of ADHD persistence and other potential predictors. J Atten Disord 23:907–923. [DOI] [PubMed] [Google Scholar]
  • 2.Fayyad J, Sampson NA, Hwang I, Adamowski T, Aguilar-Gaxiola S, Al-Hamzawi A, et al. (2017): The descriptive epidemiology of DSM-IV Adult ADHD in the World Health Organization World Mental Health Surveys. Atten Defic Hyperact Disord 9:47–65. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 3.Polanczyk GV, Willcutt EG, Salum GA, Kieling C, Rohde LA (2014): ADHD prevalence estimates across three decades: An updated systematic review and meta-regression analysis. Int J Epidemiol 43:434–442. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 4.Biederman J, Petty C, Fried R, Fontanella J, Doyle AE, Seidman LJ, et al. (2006): Impact of psychometrically defined deficits of executive functioning in adults with attention deficit hyperactivity disorder. Am J Psychiatry 163:1730–1738. [DOI] [PubMed] [Google Scholar]
  • 5.Brown TE, Landgraf JM (2010): Improvements in executive function correlate with enhanced performance and functioning and health-related quality of life: Evidence from 2 large, double-blind, randomized, placebo-controlled trials in ADHD. Postgrad Med 122:42–51. [DOI] [PubMed] [Google Scholar]
  • 6.Biederman J, Mick E, Fried R, Wilner N, Spencer TJ, Faraone SV (2011): Are stimulants effective in the treatment of executive function deficits? Results from a randomized double blind study of OROS-methylphenidate in adults with ADHD. Eur Neuropsychopharmacol 21:508–515. [DOI] [PubMed] [Google Scholar]
  • 7.Bilder RM, Loo SK, McGough JJ, Whelan F, Hellemann G, Sugar C, et al. (2016): Cognitive effects of stimulant, guanfacine, and combined treatment in child and adolescent attention-deficit/hyperactivity disorder. J Am Acad Child Adolesc Psychiatry 55:667–673. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 8.Philip NS, Nelson BG, Frohlich F, Lim KO, Widge AS, Carpenter LL (2017): Low-intensity transcranial current stimulation in psychiatry. Am J Psychiatry 174:628–639. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 9.Brunoni AR, Nitsche MA, Bolognini N, Bikson M, Wagner T, Merabet L, et al. (2012): Clinical research with transcranial direct current stimulation (tDCS): Challenges and future directions. Brain Stimul 5:175–195. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 10.Sarkis RA, Kaur N, Camprodon JA (2014): Transcranial direct current stimulation (tDCS): Modulation of executive function in health and disease. Curr Behav Neurosci Rep 1:74–85. [Google Scholar]
  • 11.Dubreuil-Vall L, Chau P, Ruffini G, Widge AS, Camprodon JA (2019): tDCS to the left DLPFC modulates cognitive and physiological correlates of executive function in a state-dependent manner. Brain Stimul 12:1456–1463. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 12.Rubia K (2018): Cognitive neuroscience of attention deficit hyperactivity disorder (ADHD) and its clinical translation. Front Hum Neurosci 12:100. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 13.Eriksen BA, Eriksen CW (1974): Effects of noise letters upon the identification of a target letter in a nonsearch task. Percept Psychophys 16:143–149. [Google Scholar]
  • 14.Lappin J, Eriksen C (1966): Use of a delayed signal to stop a visual reaction-time response. J Exp Psychol 72:805–811. [Google Scholar]
  • 15.Mullane JC, Corkum PV, Klein RM, McLaughlin E (2009): Interference control in children with and without ADHD: A systematic review of flanker and Simon task performance. Child Neuropsychol 15:321–342. [DOI] [PubMed] [Google Scholar]
  • 16.Alderson RM, Rapport MD, Kofler MJ (2007): Attention-deficit/hyperactivity disorder and behavioral inhibition: A meta-analytic review of the stop-signal paradigm. J Abnorm Child Psychol 35:745–758. [DOI] [PubMed] [Google Scholar]
  • 17.Logan GD, Van Zandt T, Verbruggen F, Wagenmakers EJ (2014): On the ability to inhibit thought and action: General and special theories of an act of control. Psychol Rev 121:66–95. [DOI] [PubMed] [Google Scholar]
  • 18.Boisseau CL, Thompson-Brenner H, Caldwell-Harris C, Pratt E, Farchione T, Barlow DH (2012): Behavioral and cognitive impulsivity in obsessive-compulsive disorder and eating disorders. Psychiatry Res 200:1062–1066. [DOI] [PubMed] [Google Scholar]
  • 19.Verbruggen F, Logan GD (2008): Response inhibition in the stop-signal paradigm. Trends Cogn Sci 12:418–424. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 20.Danielmeier C, Wessel JR, Steinhauser M, Ullsperger M (2009): Modulation of the error-related negativity by response conflict. Psychophysiology 46:1288–1298. [DOI] [PubMed] [Google Scholar]
  • 21.Clayson PE, Larson MJ (2011): Conflict adaptation and sequential trial effects: Support for the conflict monitoring theory. Neuropsychologia 49:1953–1961. [DOI] [PubMed] [Google Scholar]
  • 22.Larson MJ, Clayson PE, Clawson A (2014): Making sense of all the conflict: A theoretical review and critique of conflict-related ERPs. Int J Psychophysiol 93:283–297. [DOI] [PubMed] [Google Scholar]
  • 23.Falkenstein M, Hoormann J, Christ S, Hohnsbein J (2000): ERP components on reaction errors and their functional significance: A tutorial. Biol Psychol 51:87–107. [DOI] [PubMed] [Google Scholar]
  • 24.Marquardt L, Eichele H, Lundervold AJ, Haavik J, Eichele T (2018): Event-related-potential (ERP) correlates of performance monitoring in adults with attention-deficit hyperactivity disorder (ADHD). Front Psychol 9:485. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 25.Philips S, Takeda Y (2009): An EEG/ERP study of efficient versus inefficient visual search. Proc Annu Meet Cogn Sci Soc 31:383–388. [Google Scholar]
  • 26.Sur S, Sinha V (2009): Event-related potential: An overview. Ind Psychiatry J 18:70–73. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 27.Kaiser A, Aggensteiner P-M, Baumeister S, Holz NE, Banaschewski T, Brandeis D (2020): Earlier versus later cognitive event-related potentials (ERPs) in attention-deficit/hyperactivity disorder (ADHD): A meta-analysis. Neurosci Biobehav Rev 112:117–134. [DOI] [PubMed] [Google Scholar]
  • 28.Yeung N, Ralph J, Nieuwenhuis S (2007): Drink alcohol and dim the lights: The impact of cognitive deficits on medial frontal cortex function. Cogn Affect Behav Neurosci 7:347–355. [DOI] [PubMed] [Google Scholar]
  • 29.Clayson PE, Larson MJ (2011): Effects of repetition priming on electrophysiological and behavioral indices of conflict adaptation and cognitive control. Psychophysiology 48:1621–1630. [DOI] [PubMed] [Google Scholar]
  • 30.Cortese S (2012): The neurobiology and genetics of attention-deficit/hyperactivity disorder (ADHD): What every clinician should know. Eur J Paediatr Neurol 16:422–433. [DOI] [PubMed] [Google Scholar]
  • 31.Coles MGH, Rugg MD (1995): Event-related brain potentials: An introduction. In: Electrophysiology of Mind: Event-Related Brain Potentials and Cognition. New York, NY: Oxford University Press, 1–26. [Google Scholar]
  • 32.Nieuwenhuis S, Ridderinkhof KR, Blom J, Band GP, Kok A (2001): Error-related brain potentials are differentially related to awareness of response errors: Evidence from an antisaccade task. Psychophysiology 38:752–760. [PubMed] [Google Scholar]
  • 33.Delorme A, Makeig S (2004): EEGLAB: An open source toolbox for analysis of single-trial EEG dynamics including independent component analysis. J Neurosci Methods 134:9–21. [DOI] [PubMed] [Google Scholar]
  • 34.Allenby C, Falcone M, Bernardo L, Wileyto EP, Rostain A, Ramsay JR, et al. (2018): Transcranial direct current brain stimulation decreases impulsivity in ADHD. Brain Stimul 11:974–981. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 35.Enriquez-Geppert S, Konrad C, Pantev C, Huster RJ (2010): Conflict and inhibition differentially affect the N200/P300 complex in a combined go/nogo and stop-signal task. Neuroimage 51:877–887. [DOI] [PubMed] [Google Scholar]
  • 36.Castellanos FX, Sonuga-Barke EJ, Scheres A, Di Martino A, Hyde C, Walters JR (2005): Varieties of attention-deficit/hyperactivity disorder-related intra-individual variability. Biol Psychiatry 57:1416–1423. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 37.McGough JJ, Sturm A, Cowen J, Tung K, Salgari GC, Leuchter AF, et al. (2019): Double-blind, sham-controlled, pilot study of trigeminal nerve stimulation for attention-deficit/hyperactivity disorder. J Am Acad Child Adolesc Psychiatry 58:403–411. e3. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 38.Castro-Meneses LJ, Johnson BW, Sowman PF (2015): The effects of impulsivity and proactive inhibition on reactive inhibition and the go process: Insights from vocal and manual stop signal tasks. Front Hum Neurosci 9:529. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 39.Hogeveen J, Grafman J, Aboseria M, David A, Bikson M, Hauner KK (2016): Effects of high-definition and conventional tDCS on response inhibition. Brain Stimul 9:720–729. [DOI] [PubMed] [Google Scholar]
  • 40.Cunillera T, Fuentemilla L, Brignani D, Cucurell D, Miniussi C (2014): A simultaneous modulation of reactive and proactive inhibition processes by anodal tDCS on the right inferior frontal cortex. PLoS One 9: e113537. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 41.Ditye T, Jacobson L, Walsh V, Lavidor M (2012): Modulating behavioral inhibition by tDCS combined with cognitive training. Exp Brain Res 219:363–368. [DOI] [PubMed] [Google Scholar]
  • 42.Jacobson L, Javitt DC, Lavidor M (2011): Activation of inhibition: Diminishing impulsive behavior by direct current stimulation over the inferior frontal gyrus. J Cogn Neurosci 23:3380–3387. [DOI] [PubMed] [Google Scholar]
  • 43.Cai Y, Li S, Liu J, Li D, Feng Z, Wang Q, et al. (2016): The role of the frontal and parietal cortex in proactive and reactive inhibitory control: A transcranial direct current stimulation study. J Cogn Neurosci 28:177–186. [DOI] [PubMed] [Google Scholar]
  • 44.Stramaccia DF, Penolazzi B, Sartori G, Braga M, Mondini S, Galfano G (2015): Assessing the effects of tDCS over a delayed response inhibition task by targeting the right inferior frontal gyrus and right dorsolateral prefrontal cortex. Exp Brain Res 233:2283–2290. [DOI] [PubMed] [Google Scholar]
  • 45.Kwon YH, Kwon JW (2013): Response inhibition induced in the stop-signal task by transcranial direct current stimulation of the pre-supplementary motor area and primary sensoriomotor cortex. J Phys Ther Sci 25:1083–1086. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 46.Cosmo C, Baptista AF, de Araujo AN, do Rosario RS, Miranda JG, Montoya P, et al. (2015): A randomized, double-blind, sham-controlled trial of transcranial direct current stimulation in attention-deficit/hyperactivity disorder. PLoS One 10:e0135371. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 47.Bari A, Robbins TW (2013): Inhibition and impulsivity: Behavioral and neural basis of response control. Prog Neurobiol 108:44–79. [DOI] [PubMed] [Google Scholar]
  • 48.Verbruggen F, Liefooghe B, Notebaert W, Vandierendonck A (2005): Effects of stimulus-stimulus compatibility and stimulus-response compatibility on response inhibition. Acta Psychol 120:307–326. [DOI] [PubMed] [Google Scholar]
  • 49.Verbruggen F, Liefooghe B, Vandierendonck A (2004): The interaction between stop signal inhibition and distractor interference in the Flanker and Stroop task. Acta Psychol 116:21–37. [DOI] [PubMed] [Google Scholar]
  • 50.Friedman NP, Miyake A (2004): The relations among inhibition and interference control functions: A latent-variable analysis. J Exp Psychol Gen 133:101–135. [DOI] [PubMed] [Google Scholar]
  • 51.Aron AR, Robbins TW, Poldrack RA (2004): Inhibition and the right inferior frontal cortex. Trends Cogn Sci 8:170–177. [DOI] [PubMed] [Google Scholar]
  • 52.Derrfuss J, Brass M, von Cramon DY (2004): Cognitive control in the posterior frontolateral cortex: Evidence from common activations in task coordination, interference control, and working memory. Neuroimage 23:604–612. [DOI] [PubMed] [Google Scholar]
  • 53.Chambers CD, Bellgrove MA, Gould IC, English T, Garavan H, McNaught E, et al. (2007): Dissociable mechanisms of cognitive control in prefrontal and premotor cortex. J Neurophysiol 98:3638–3647. [DOI] [PubMed] [Google Scholar]
  • 54.FDA-NIH Biomarker Working Group (2016): BEST (Biomarkers, EndpointS, and other Tools). Silver Spring, MD: Food and Drug Administration. Bethesda, MD: National Institutes of Health. [PubMed] [Google Scholar]
  • 55.Dinn W, Göral F, Adigüzel S, Karamürsel S, Fregni F, Aycicegi-Dinn A (2017): Effectiveness of tDCS blinding protocol in a sham-controlled study. Brain Stimul 10:401. [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Supplementary
NIHMS1692132-supplement-Supplementary.pdf (485.5KB, pdf)

RESOURCES