Advertisement

Enhancing cognitive control training with transcranial direct current stimulation: a systematic parameter study

  • Simone Weller
    Affiliations
    Department of Psychiatry and Psychotherapy, Neurophysiology & Interventional Neuropsychiatry, University of Tübingen, Calwerstraße 14, 72076, Tübingen, Germany
    Search for articles by this author
  • Michael A. Nitsche
    Affiliations
    Department of Psychology and Neurosciences, Leibniz Research Centre for Working Environment and Human Factors, Ardeystraße 67, 44139, Dortmund, Germany

    University Medical Hospital Bergmannsheil, Bochum, Germany
    Search for articles by this author
  • Christian Plewnia
    Correspondence
    Corresponding author.
    Affiliations
    Department of Psychiatry and Psychotherapy, Neurophysiology & Interventional Neuropsychiatry, University of Tübingen, Calwerstraße 14, 72076, Tübingen, Germany
    Search for articles by this author
Open AccessPublished:July 17, 2020DOI:https://doi.org/10.1016/j.brs.2020.07.006

      Highlights

      • Transcranial direct current stimulation supports training of cognitive control.
      • Stimulation is highly dependent on the chosen parameters.
      • Varying stimulation parameters showed non-linearity of effects.
      • TDCS may function as a supportive mode for supporting cognitive enhancement.

      Abstract

      Background

      Cognitive control (CC) is an important prerequisite for goal-directed behaviour and efficient information processing. Impaired CC is associated with reduced prefrontal cortex activity and various mental disorders, but may be effectively tackled by transcranial direct current stimulation (tDCS)-enhanced training. However, study data are inconsistent as efficacy depends on stimulation parameters whose implementations vary widely between studies.

      Objective

      We systematically tested various tDCS parameter effects (anodal/cathodal polarity, 1/2 mA stimulation intensity, left/right prefrontal cortex hemisphere) on a six-session CC training combined with tDCS.

      Methods

      Nine groups of healthy humans (male/female) received either anodal/cathodal tDCS of 1/2 mA over the left/right PFC or sham stimulation, simultaneously with a CC training (modified adaptive Paced Auditory Serial Addition Task [PASAT]). Subjects trained thrice per week (19 min each) for two weeks. We assessed performance progress in the PASAT before, during, and after training. Using a hierarchical approach, we incrementally narrowed down on optimal stimulation parameters supporting CC. Long-term CC effects as well as transfer effects in a flanker task were assessed after the training period as well as three months later.

      Results

      Compared to sham stimulation, anodal but not cathodal tDCS improved performance gains. This was only valid for 1 mA stimulation intensity and particularly detected when applied to the left PFC.

      Conclusions

      Our results confirm beneficial, non-linear effects of anodal tDCS on cognitive training in a large sample of healthy subjects. The data consolidate the basis for further development of functionally targeted tDCS, supporting cognitive control training in mental disorders and guiding further development of clinical interventions.

      Keywords

      Introduction

      Continuously changing environments require dynamic adaptation by means of filtering and evaluating internal and external stimuli to orchestrate goal-directed behaviour. This is especially important for situations in which distractions might influence efficient responses. Important information is maintained, while non-relevant stimuli must be suppressed or ignored. Dysfunctions of cognitive control (CC) processes are at the core of many psychopathological conditions [
      • Fitzgerald P.B.
      • Laird A.R.
      • Maller J.
      • Daskalakis Z.J.
      A meta-analytic study of changes in brain activation in depression.
      ,
      • Baune B.T.
      • Miller R.
      • McAfoose J.
      • Johnson M.
      • Quirk F.
      • Mitchell D.
      The role of cognitive impairment in general functioning in major depression.
      ], comprise the intentional selection of thoughts, emotions, and behaviours based on current task demands [
      • Dixon M.L.
      Cognitive control, emotional value, and the lateral prefrontal cortex.
      ] involving functions of attention, memory, and emotional control [
      • Ryman S.G.
      • Shaikh A.A.E.
      • Shaff N.A.
      • Hanlon F.M.
      • Dodd A.B.
      • Wertz C.J.
      • et al.
      Proactive and reactive cognitive control rely on flexible use of the ventrolateral prefrontal cortex.
      ], and are associated with altered patterns of brain activation [
      • McTeague L.M.
      • Huemer J.
      • Carreon D.M.
      • Jiang Y.
      • Eickhoff S.B.
      • Etkin A.
      Identification of common neural circuit disruptions in cognitive control across psychiatric disorders.
      ,
      • Miskowiak K.W.
      • Petersen C.S.
      Neuronal underpinnings of cognitive impairment and - improvement in mood disorders.
      ]. The prefrontal cortex (PFC), particularly the dorsolateral prefrontal cortex (dlPFC), is known to be highly involved in CC processes [
      • Burgess P.W.
      • Stuss D.T.
      Fifty years of prefrontal cortex research: impact on assessment.
      ] by means of processes related to working memory [
      • Barbey A.K.
      • Koenigs M.
      • Grafman J.
      Dorsolateral prefrontal contributions to human working memory.
      ], encoding of task relevant rules and responses [
      • Brunoni A.R.
      • Vanderhasselt M.-A.
      Working memory improvement with non-invasive brain stimulation of the dorsolateral prefrontal cortex: a systematic review and meta-analysis.
      ], and emotion regulation [
      • Calkins A.W.
      • McMorran K.E.
      • Siegle G.J.
      • Otto M.W.
      The effects of computerized cognitive control training on community adults with depressed mood.
      ].
      Transcranial direct current stimulation (tDCS) has been put forward as a means to influence these processes by modulating the likelihood of neuronal firing in response to a stimulus [
      • Nitsche M.A.
      • Paulus W.
      Excitability changes induced in the human motor cortex by weak transcranial direct current stimulation.
      ]. At the macroscopic level, within the common and safe range of stimulation parameters (1–2 mA, up to 30 min of stimulation [
      • Grossman P.
      • Woods A.J.
      • Knotkova H.
      • Bikson M.
      Safety of transcranial direct current stimulation.
      ]), it is supposed that anodal tDCS predominantly enhances, while cathodal tDCS mainly reduces the excitability and spontaneous activity of the targeted and connected areas [
      • Cambiaghi M.
      • Velikova S.
      • Gonzalez-Rosa J.J.
      • Cursi M.
      • Comi G.
      • Leocani L.
      Brain transcranial direct current stimulation modulates motor excitability in mice.
      ]. This polarity-dependent modulation of brain activity by tDCS has a remarkable potential to influence corresponding cognition and behaviour [
      • Zwissler B.
      • Sperber C.
      • Aigeldinger S.
      • Schindler S.
      • Kissler J.
      • Plewnia C.
      Shaping memory accuracy by left prefrontal transcranial direct current stimulation.
      ,
      • Kronberg G.
      • Bridi M.
      • Abel T.
      • Bikson M.
      • Parra L.C.
      Direct current stimulation modulates LTP and LTD: activity dependence and dendritic effects.
      ,
      • Ruf S.P.
      • Fallgatter A.J.
      • Plewnia C.
      Augmentation of working memory training by transcranial direct current stimulation (tDCS).
      ]. However, tDCS does not induce cortical activity per se. It develops its effects particularly in interaction with spontaneous neuronal activity [
      • Fritsch B.
      • Reis J.
      • Martinowich K.
      • Schambra H.M.
      • Ji Y.
      • Cohen L.G.
      • et al.
      Direct current stimulation promotes BDNF-dependent synaptic plasticity: potential implications for motor learning.
      ,
      • Bikson M.
      • Rahman A.
      Origins of specificity during tDCS: anatomical, activity-selective, and input-bias mechanisms.
      ]. This activity-dependent influence on brain networks allows for a ‘functional targeting’ of stimulation when tDCS is directly coupled with the respective cognitive or behavioural process [
      • Bikson M.
      • Brunoni A.R.
      • Charvet L.E.
      • Clark V.P.
      • Cohen L.G.
      • Deng Z.-D.
      • et al.
      Rigor and reproducibility in research with transcranial electrical stimulation: an NIMH-sponsored workshop.
      ], where the target regions are activated (i.e. by a task) and further specifically modulated by the stimulation [
      • Guleyupoglu B.
      • Schestatsky P.
      • Edwards D.
      • Fregni F.
      • Bikson M.
      Classification of methods in transcranial Electrical Stimulation (tES) and evolving strategy from historical approaches to contemporary innovations.
      ]. Correspondingly, tDCS effects have been found especially in neuronal correlates of task features that were active during stimulation [
      • Schroeder P.A.
      • Nuerk H.-C.
      • Plewnia C.
      Prefrontal neuromodulation reverses spatial associations of non-numerical sequences, but not numbers.
      ]. Therefore, the combination of tDCS with task training is suggested to have a synergistic ‘neuroenhancing’ effect that is currently subject of extensive research [
      • Polanía R.
      • Nitsche M.A.
      • Paulus W.
      Modulating functional connectivity patterns and topological functional organization of the human brain with transcranial direct current stimulation.
      ,
      • Segrave R.A.
      • Arnold S.
      • Hoy K.
      • Fitzgerald P.B.
      Concurrent cognitive control training augments the antidepressant efficacy of tDCS: a pilot study.
      ,
      • Brem A.-K.
      • Almquist J.N.-F.
      • Mansfield K.
      • Plessow F.
      • Sella F.
      • Santarnecchi E.
      • et al.
      Modulating fluid intelligence performance through combined cognitive training and brain stimulation.
      ,
      • Ward N.
      • Paul E.
      • Watson P.
      • Cooke G.E.
      • Hillman C.H.
      • Cohen N.J.
      • et al.
      Enhanced learning through multimodal training: evidence from a comprehensive cognitive, physical fitness, and neuroscience intervention.
      ]. However, available data are still inconsistent as efficacy depends on stimulation parameters that vary widely between studies. For a meaningful clinical application, a sustainable enhancement of adaptive plasticity would be most desirable [
      • Cramer S.C.
      • Sur M.
      • Dobkin B.H.
      • O’Brien C.
      • Sanger T.D.
      • Trojanowski J.Q.
      • et al.
      Harnessing neuroplasticity for clinical applications.
      ]. Based on this notion, a specific activation of the CC network and concomitant tDCS holds promise to provide new treatment strategies for cognitive and behavioural disorders [
      • Plewnia C.
      • Schroeder P.A.
      • Wolkenstein L.
      Targeting the biased brain: non-invasive brain stimulation to ameliorate cognitive control.
      ,
      • Kekic M.
      • Boysen E.
      • Campbell I.C.
      • Schmidt U.
      A systematic review of the clinical efficacy of transcranial direct current stimulation (tDCS) in psychiatric disorders.
      ,
      • Vicario C.M.
      • Salehinejad M.A.
      • Felmingham K.
      • Martino G.
      • Nitsche M.A.
      A systematic review on the therapeutic effectiveness of non-invasive brain stimulation for the treatment of anxiety disorders.
      ]. In a plethora of studies, stimulation has already shown to enhance CC by changing emotion regulation processes [
      • Feeser M.
      • Prehn K.
      • Kazzer P.
      • Mungee A.
      • Bajbouj M.
      Transcranial direct current stimulation enhances cognitive control during emotion regulation.
      ], improving frustration tolerance [
      • Plewnia C.
      • Schroeder P.A.
      • Kunze R.
      • Faehling F.
      • Wolkenstein L.
      Keep calm and carry on: improved frustration tolerance and processing speed by transcranial direct current stimulation (tDCS).
      ], modulating emotional vulnerability [
      • Chen N.T.M.
      • Basanovic J.
      • Notebaert L.
      • MacLeod C.
      • Clarke P.J.F.
      Attentional bias mediates the effect of neurostimulation on emotional vulnerability.
      ], dissolving attentional biases [
      • Wolkenstein L.
      • Plewnia C.
      Amelioration of cognitive control in depression by transcranial direct current stimulation.
      ], augmenting working memory training [
      • Ruf S.P.
      • Fallgatter A.J.
      • Plewnia C.
      Augmentation of working memory training by transcranial direct current stimulation (tDCS).
      ], and increasing multitasking capacity [
      • Nelson J.
      • McKinley R.A.
      • Phillips C.
      • McIntire L.
      • Goodyear C.
      • Kreiner A.
      • et al.
      The effects of transcranial direct current stimulation (tDCS) on multitasking throughput capacity.
      ]. However, reliability of results and the plausibility of approaches leaves room for improvement, not at least because studies often yield varying results even for similar tasks [
      • Krause B.
      • Cohen Kadosh R.
      Not all brains are created equal: the relevance of individual differences in responsiveness to transcranial electrical stimulation.
      ,
      • Nilsson J.
      • Lebedev A.V.
      • Rydström A.
      • Lövdén M.
      Direct-current stimulation does little to improve the outcome of working memory training in older adults.
      ,
      • Berryhill M.E.
      • Martin D.
      Cognitive effects of transcranial direct current stimulation in healthy and clinical populations: an overview.
      ,
      • Wörsching J.
      • Padberg F.
      • Goerigk S.
      • Heinz I.
      • Bauer C.
      • Plewnia C.
      • et al.
      Testing assumptions on prefrontal transcranial direct current stimulation: comparison of electrode montages using multimodal fMRI.
      ]. Therefore, reliable knowledge about the efficacy of parameter settings is mandatory for further advancements [
      • Chase H.W.
      • Boudewyn M.A.
      • Carter C.S.
      • Phillips M.L.
      Transcranial direct current stimulation: a roadmap for research, from mechanism of action to clinical implementation.
      ].
      To this aim, we systematically tested different standard stimulation parameters (anodal/cathodal tDCS with 1/2 mA to the left/right dlPFC) in 162 healthy subjects, combining repeated CC training (6 sessions within 2 weeks) with tDCS, and additionally analysed pre- and post-training assessments. We applied a modified adaptive paced auditory serial addition task (PASAT) to challenge and train CC [
      • Tombaugh T.N.
      A comprehensive review of the paced auditory serial addition test (PASAT).
      ]. This task requires continuous updating of working memory with parallel distracting performance feedback; it is known to activate CC [
      • Plewnia C.
      • Schroeder P.A.
      • Kunze R.
      • Faehling F.
      • Wolkenstein L.
      Keep calm and carry on: improved frustration tolerance and processing speed by transcranial direct current stimulation (tDCS).
      ], critically involves resources within the PFC [
      • Audoin B.
      • Ibarrola D.
      • Duong M.V.A.
      • Pelletier J.
      • Confort-Gouny S.
      • Malikova I.
      • et al.
      Functional MRI study of PASAT in normal subjects.
      ], and adapts task difficulty to individual performance [
      • Flegal K.E.
      • Ragland J.D.
      • Ranganath C.
      Adaptive task difficulty influences neural plasticity and transfer of training.
      ]. We hypothesized that adding anodal but not cathodal tDCS to PASAT-induced neuronal activity of the dlPFC [
      • Romero Lauro L.J.
      • Rosanova M.
      • Mattavelli G.
      • Convento S.
      • Pisoni A.
      • Opitz A.
      • et al.
      TDCS increases cortical excitability: direct evidence from TMS–EEG.
      ,
      • Antal A.
      • Polania R.
      • Schmidt-Samoa C.
      • Dechent P.
      • Paulus W.
      Transcranial direct current stimulation over the primary motor cortex during fMRI.
      ] can enhance cognitive training effects [
      • Au J.
      • Katz B.
      • Buschkuehl M.
      • Bunarjo K.
      • Senger T.
      • Zabel C.
      • et al.
      Enhancing working memory training with transcranial direct current stimulation.
      ,
      • Pope P.A.
      • Brenton J.W.
      • Miall R.C.
      Task-specific facilitation of cognition by anodal transcranial direct current stimulation of the prefrontal cortex.
      ], improve performance of the PASAT or similar, even more challenging tasks [
      • Ruf S.P.
      • Fallgatter A.J.
      • Plewnia C.
      Augmentation of working memory training by transcranial direct current stimulation (tDCS).
      ,
      • Plewnia C.
      • Schroeder P.A.
      • Kunze R.
      • Faehling F.
      • Wolkenstein L.
      Keep calm and carry on: improved frustration tolerance and processing speed by transcranial direct current stimulation (tDCS).
      ,
      • Pope P.A.
      • Brenton J.W.
      • Miall R.C.
      Task-specific facilitation of cognition by anodal transcranial direct current stimulation of the prefrontal cortex.
      ,
      • Benwell C.S.Y.
      • Learmonth G.
      • Miniussi C.
      • Harvey M.
      • Thut G.
      Non-linear effects of transcranial direct current stimulation as a function of individual baseline performance: evidence from biparietal tDCS influence on lateralized attention bias.
      ,
      • Schroeder P.A.
      • Plewnia C.
      Beneficial effects of cathodal transcranial direct current stimulation (tDCS) on cognitive performance.
      ,
      • Wiegand A.
      • Sommer A.
      • Nieratschker V.
      • Plewnia C.
      Improvement of cognitive control and stabilization of affect by prefrontal transcranial direct current stimulation (tDCS).
      ,
      • Katsoulaki M.
      • Kastrinis A.
      • Tsekoura M.
      The effects of anodal transcranial direct current stimulation on working memory.
      ], and that higher stimulation intensity does not increase efficacy [
      • Batsikadze G.
      • Moliadze V.
      • Paulus W.
      • Kuo M.-F.
      • Nitsche M.A.
      Partially non-linear stimulation intensity-dependent effects of direct current stimulation on motor cortex excitability in humans: effect of tDCS on cortical excitability.
      ]. Furthermore, we wanted to test if the laterality of stimulation matters. Therefore, PASAT performance under eight different tDCS conditions (combined N = 119) was compared to a sham intervention group (N = 43). Analyses were conducted hierarchically, allowing us to narrow down the responsible factors for the most efficient combination of CC training and tDCS.

      Materials and methods

      The study was performed in accordance with the Declaration of Helsinki, approved by the University of Tübingen local ethics committee, and conducted in the University Hospital Tübingen, Department of Psychiatry and Psychotherapy. Subjects were recruited through flyers, online forums, and the e-mail distribution list of the University of Tübingen. Before inclusion, each subject gave informed written consent.

      Experimental design

      Subjects

      Out of 192 eligible subjects, 163 right-handed subjects finished all experimental sessions and 162 were eventually included in the data analysis of this single-blind between-subject study (127 females, 35 males; ages between 18 and 39; mean age = 23.20 years, standard deviation = 3.98 years). Dropouts were caused by inability to adhere to the strictly timed training schedule, as shifting appointments was not allowed and resulted in termination of participation. Minimum group size was set to 15 subjects per group in accordance with previous studies with similar sample sizes, interventions, and significant outcomes, since standard procedures for calculation of sample size are not available for linear mixed effect models which were fitted to analyse our data [
      • Brunoni A.R.
      • Vanderhasselt M.-A.
      Working memory improvement with non-invasive brain stimulation of the dorsolateral prefrontal cortex: a systematic review and meta-analysis.
      ,
      • Kalu U.G.
      • Sexton C.E.
      • Loo C.K.
      • Ebmeier K.P.
      Transcranial direct current stimulation in the treatment of major depression: a meta-analysis.
      ]. Exclusion criteria were diagnosed neurological or psychiatric disorders, achromatopsia, metallic implants or tattoos near electrode sites, consumption of tobacco to an equivalent of ten or more cigarettes per day, German language skills lower than CEFR level B, and simultaneous brain stimulation from other sources during attendance of this study. Subjects chose between monetary compensation or course credits. To create an incentive for increased effort in solving the task, participants were informed that the top twelve performers additionally received a bonus pay-out at the end of the study.

      TDCS procedure and experimental groups

      Stimulation was delivered by a CE-certified direct current stimulator (DC-Stimulator MC, NeuroConn GmbH, Ilmenau, Germany), version 1.3.8, and two rectangular rubber electrodes (5 × 7 cm). Experimental groups were specified by stimulation polarity, current intensity, and electrode laterality. Polarity-dependent effects were examined by placing either the anode or cathode over the PFC. Stimulation intensity was varied by applying a current of either 1 mA or 2 mA, resulting in densities of 0.03 or 0.06 mA/cm2 respectively. Impedances were kept below 8 kΩ. Laterality was defined by positioning the electrode centre either over F3 or F4 according to the international 10–20 system, with the foreside oriented towards the nasion, the backside with the attached cable oriented towards the inion. The extracephalic electrode was mounted over the opposite lateral deltoid muscle to avoid confounding effects of opposing stimulation polarities on brain physiology which can occur with bipolar cephalic tDCS [
      • Nitsche M.A.
      • Cohen L.G.
      • Wassermann E.M.
      • Priori A.
      • Lang N.
      • Antal A.
      • et al.
      Transcranial direct current stimulation: state of the art 2008.
      ]. Consequently, these parameter combinations resulted in nine groups (referred to as conditions) that each subject was randomly assigned to: sham (S), anodal/1 mA/left PFC (A1L), anodal/1 mA/right PFC (A1R), anodal/2 mA/left PFC (A2L), anodal/2 mA/right PFC (A2R), cathodal/1 mA/left PFC (C1L), cathodal/1 mA/right PFC (C1R), cathodal/2 mA/left PFC (C2L), cathodal/2 mA/right PFC (C2R).
      Electrode surfaces were coated with conductive electrode paste (Ten 20 conductive Neurodiagnostic Electrode Paste, Weaver and Company, Aurora, Colorado), placed on skin areas previously prepared with 70 % ethanol and mildly abrasive peeling gel (Nuprep Skin Prep Gel, Weaver and Company, Aurora, Colorado), and secured by a fabric cap and tape.
      For verum stimulation, the current was ramped up within 5 s to start, and down within 5 s to end the stimulation. Stimulation at target intensity comprised 19:10 min. For sham tDCS, a short stimulation block before and after the training was applied. For the first block, the current was ramped up within 5 s, kept at target intensity (1 or 2 mA) for 40 s, then ramped down within 5 s. After 18:30 min the second block was initiated and the current was ramped up slowly within 39 s, kept at target intensity for 10 s and ramped down within 1 s. This was done to distinctly mark the end of the second sham phase so that subjects of the sham procedure would feel similar sensations as the subjects treated with continuous tDCS. All subjects started the task 1 min after the stimulation was initiated, ensuring that verum stimulation spanned the entire time subjects worked on the task, while sham stimulation blocks were only active before and after completion of the task, which is too short to induce after-effects [
      • Nitsche M.A.
      • Paulus W.
      Excitability changes induced in the human motor cortex by weak transcranial direct current stimulation.
      ,
      • Fritsch B.
      • Reis J.
      • Martinowich K.
      • Schambra H.M.
      • Ji Y.
      • Cohen L.G.
      • et al.
      Direct current stimulation promotes BDNF-dependent synaptic plasticity: potential implications for motor learning.
      ].

      CC training: PASAT

      Our version of the Paced Auditory Serial Addition Task (PASAT) was implemented in PsychoPy 1.82.01 [
      • Peirce J.W.
      Generating stimuli for neuroscience using PsychoPy.
      ]. Subjects were seated in front of a computer screen while hearing random single digit numbers (1–9) over headphones. They were instructed to sum up the current digit and the digit that preceded it by two trials (nth + nth−2; ’two-back’), hence deviating our task from the ’one-back’ (nth + nth−1) version. This was done to increase mental load as well as to prevent subjects from reaching their peak performance too early and therefore having no room for further improvement. Answers were given immediately after each stimulus presentation on a keyboard marked with all possible results (2–18). Only the usage of one finger of the right hand was permitted, preventing subjects from tagging numbers. The added benefit of the keyboard setup compared to the more conventional use of a mouse cursor was to reduce the intricate element of dexterity, therefore linking the reaction times closer to cognitive abilities.
      At the beginning of each session, the interval between digit presentations was 3 s. After four consecutive correct answers, the interval was lowered by 0.1 s, while after four consecutive wrong answers, it was increased by 0.1 s. Feedback on each calculation was given together with the succeeding digit by means of a green (correct) or red (incorrect/missed) computer screen (Fig. 1A). Sessions were divided into three blocks of 5 min of continuous digit presentations, separated by 30 s rest. The total number of trials within each session was not limited and intervals between digit presentations were carried over from preceding blocks within, but not between, sessions. This adaptive presentation speed at fixed time windows qualifies the total number of correct answers in each session as the most sensible performance parameter.
      Fig. 1
      Fig. 1A During the PASAT, subjects heard single digit numbers over headphones and were asked to sum up the current digit (nth) and the second-to-last digit that preceded it (nth−2): e. g. digits at timepoints C and A, D and B, E and C, and so on. Answers were given on a keyboard using one hand only, subsequent feedback appeared simultaneously with the following digit presentation. Consecutive correct answers shortened the interval between digit presentations, wrong answers prolonged it. B During the flanker task, subjects were asked to indicate the direction of a centre stimulus surrounded by distractors. The three go-conditions, consisting of congruent (e. g. <<<<<<<), incongruent (e. g. >>><>>>), and neutral distractors (e. g.===>===), were accompanied by one no-go-condition (XXX < XXX). C The study was divided into nine sessions. Sessions were spread in a way that training days alternated with days without any intervention. Between weeks, there were two resting days. EHI = Edinburgh Handedness Inventory; PANAS = Positive and Negative Affect Schedule; QCM = Questionnaire on Current Motivation; PASAT = Paced Auditory Serial Addition Task; Flanker = Flanker task. D Analysis of the PASAT data was divided by stimulation parameters, starting with broad classifications that were further split into specified intervention parameter sub-categories. This hierarchical analysis was only continued out for groups where significant findings were obtained. E Analyses were additionally made for all stimulation parameter combinations separately in one single model to assess whether effects would still be present in this analysis. All groups were compared to sham, not between each other.
      Subjects were asked to answer correctly as fast as possible and resume quickly after mistakes or failure to respond in time. Each session began with 11 supervised practice trials which were excluded from analyses.

      Transfer: flanker task

      The flanker task [
      • Eriksen B.A.
      • Eriksen C.W.
      Effects of noise letters upon the identification of a target letter in a nonsearch task.
      ] was used to measure possible far transfer effects to CC processes, as successful task performance requires the subject to effectively suppress automated responses in favour of a closely but quickly evaluated response, and hence requires appropriate CC capabilities and conscious processing. It was implemented in PsychoPy 1.82.01. Subjects were seated in front of a computer screen and were instructed to respond to a target, an arrowhead pointing either to the left (<) or right (>), by indicating its direction via button press with the left or right index finger respectively. Each target was surrounded by three distractors on each side, defining the type of the stimulus. This stimulus type could either be congruent (distractors pointing in the same direction as the target: <<<<<<< or >>>>>>>) or incongruent (distractors pointing in the opposite direction to the target: <<<><<< or >>><>>>). In addition to these two conditions, the task included a neutral condition (===<=== or ===>===), during which the direction of the target was also to be indicated, and a pure no-go-condition (XXX<XXX or XXX>XXX) during which no button had to be pressed (Fig. 1B). An experimental session involved two blocks consisting of 88 trials each. Each of the aforementioned conditions was randomly displayed 11 times during the task. Before each stimulus presentation, a white dot appeared for 0.3 s to facilitate eye fixation at the target’s position and minimise response time. Stimuli were displayed for maximally 2 s, but proceeded as soon as subjects gave their response. For incorrect answers, an error message was displayed for 0.6 s, for right answers the screen remained black for 0.6 s. Before each session, subjects went through 15 supervised practice trials which were excluded from analyses.

      Study timeline

      Subjects attended nine sessions: eight sessions were conducted within four weeks while the last session took place three months later (Fig. 1C). The first session (pre-training, week 1) included the initial collection of the demographic data as well as the first PASAT run without tDCS, a 5-min pause, and eventually the flanker task. The second to seventh sessions (training, week 2 and 3) included the PASAT as well as simultaneous tDCS. The eighth (post-training, week 4) and ninth (follow-up, three months after post-training) sessions followed the same general procedure as the first session, without the initial assessment questionnaires. A day of training combined with tDCS always alternated with a day without any stimulation or training. Between weeks, there were at least two, at most three intervention-free days [
      • Woods A.J.
      • Antal A.
      • Bikson M.
      • Boggio P.S.
      • Brunoni A.R.
      • Celnik P.
      • et al.
      A technical guide to tDCS, and related non-invasive brain stimulation tools.
      ].
      During pre- and post-training, we also recorded EEG of groups who underwent treatment of the right PFC (A1R, A2R, C1R, and C2R). Blood samples were collected from all subjects after the follow-up session to assess a possible impact of genetic factors. These data will be published separately.

      Questionnaires

      Right-handedness was ascertained by the ten-item Edinburgh-Handedness-Inventory (EHI) [
      • Oldfield R.C.
      The assessment and analysis of handedness: the Edinburgh inventory.
      ]. Only subjects scoring laterality quotients equal or higher 60 were included in the study [
      • Papadatou-Pastou M.
      Chapter 7 - handedness and cognitive ability: using meta-analysis to make sense of the data.
      ]. Other sample characteristics, such as age, or educational level were gathered in a custom questionnaire (see Table 1). To account for initial interest and perceived challenge of the task, the Questionnaire on Current motivation (QCM) [
      • Rheinberg F.
      • Vollmeyer R.
      • Burns B.
      QCM: a questionnaire to assess current motivation in learning situations.
      ] was applied. It assesses four factors which are defined as follows: anxiety (assumptions of failure attributable to the pressure created by the task), probability of success (successfulness in the task), interest (appreciation of the task), and challenge (how demanding the task is perceived). To evaluate changes of affective states, subjects rated the twenty adjectives provided in the Positive and Negative Affect Schedule (PANAS) [
      • Watson D.
      • Anna L.
      • Tellegen A.
      ] both, before and after the PASAT. Possible adverse effects were reported on a custom five-point Likert-type scale. At the end of post-training, blinding was assessed and subjects stated whether they thought they had received ‘sham’ or ‘verum’ stimulation during their training. Questionnaires were provided in their respective German versions.
      Table 1Demographic group characteristics. If applicable, means and standard deviations (M(SD)) are listed, otherwise the number of subjects belonging to each parameter are shown. One Subject was removed from analyses (group C2R) as performance deviated more than 2 SD from all other subjects. a: Fisher’s Exact test; b: Kruskal-Wallis H test; c: Welch ANOVA; d: one-factorial ANOVA.
      LevelLabelTest statistic
      PolarityShamAnodalAnodalAnodalAnodalCathodalCathodalCathodalCathodal
      IntensitySham1 mA1 mA2 mA2 mA1 mA1 mA2 mA2 mA
      LateralityShamLeftRightLeftRightLeftRightLeftRight
      GroupSA1LA1RA2LA2RC1LC1RC2LC2R
      Subjects (N)431515151515151514
      Sex (m/f)a11/322/134/112/133/125/104/112/132/12p = 0.863
      Ageb

      (M(SD))
      22.767 (3.611)22.800 (4.092)24.733 (5.147)23.533 (3.270)22.267 (3.283)25.600 (3.924)23.133 (4.389)23.533 (4.853)21.071 (2.336)χ2 (8, N = 162) = 15.211, p = 0.055
      EHI-Scorec

      (M(SD))
      0.872 (0.153)0.900 (0.107)0.927 (0.116)0.886 (0.146)0.960 (0.074)0.900 (0.146)0.960 (0.063)0.893 (0.127)0.986 (0.036)F(8, 151) = 1.961, p = 0.055
      Academic Degreea (high school/middle school)42/115/015/015/015/015/015/015/014/0p = 1.000
      QCM: anxietyd3.516 (1.141)3.293 (1.331)4.027 (1.071)3.640 (0.882)3.107 (1.195)3.720 (1.121)3.627 (1.331)3.200 (1.694)4.071 (1.016)F(8, 153) = 1.174, p = 0.318, η2 = 0.058)
      QCM: probability of successd4.192 (1.266)4.050 (1.477)4.133 (1.109)4.567 (1.155)3.989 (1.441)3.933 (1.513)4.267 (0.810)4.143 (1.709)3.875 (1.108)F(8, 152) = 0.381, p = 0.929, η2 = 0.020
      QCM: interestd4.065 (1.282)3.907 (0.959)4.440 (1.127)4.333 (1.024)3.737 (1.383)3.840 (1.157)4.053 (0.987)3.867 (1.608)4.157 (1.017)F(8, 153) = 0.566, p = 0.805, η2 = 0.029
      QCM: challenged5.064 (1.047)5.100 (0.687)5.583 (0.497)5.450 (0.902)5.533 (0.640)5.433 (0.848)5.250 (0.675)5.017 (1.571)5.125 (0.663)F(8, 153) = 1.006, p = 0.434, η2 = 0.050
      Menstruating during experimenta

      (yes/no)
      17/145/88/37/59/36/47/47/510/2p = 0.786
      Hormonal contraceptive (women only)a

      (yes/no)
      16/168/55/67/54/84/66/57/57/5p = 0.978
      Smokera

      (yes/no)
      10/334/111/142/122/133/120/150/141/13p = 0.182

      Statistical analyses

      Threshold for type I error was set to 5 % for all analyses. In case of necessary correction of these values, we provide the adjusted thresholds in the respective results section. Reported values refer to two-tailed tests.

      CC training: PASAT

      All statistical analyses were done with IBM SPSS Statistics Software version 24 [
      • Armonk N.Y.
      IBM corp. IBM SPSS statistics for windows, version 24.0.0.1.
      ], R version 3.5.1 [
      R Core Team. R
      A language and environment for statistical computing.
      ], and packages nlme [
      R Core Team
      Nlme: linear and nonlinear mixed effects models.
      ] together with reghelper [
      • reghelper Hughes J.
      Helper funtions for regression analysis.
      ] in particular. Type I error was corrected by the Bonferroni-Holm method for each step of analysis, the respective corrected levels are presented in the results section.
      We used the number of correct trials as measure of performance (ncorr). In the adaptive PASAT, correct responses are followed by faster digit presentation and thus a higher number of correct trials. Comparing the number of correct trials within a fixed task duration allowed to measure performance during constant stimulation periods throughout all sessions.
      To test for differences in pre-training performance (session one), a one-factorial ANOVA with condition as between-subjects factor and ncorr(pre) as within-subject factor was used.
      To analyse tDCS effects on performance gain during training, sessions including tDCS (training sessions) and sessions without tDCS (pre-training, post-training, and follow-up) were investigated separately. The experimental groups were compared to the sham group by hierarchical analysis. First, the two polarity groups (anodal [A; N = 60], cathodal [C; N = 59]) were each compared to sham (S; N = 43). If significantly different, the respective group was then further split by intensity (1 mA [1; N = 30], 2 mA [2; N = 30]) and then, again if significant, by stimulation laterality (left PFC [L; N = 15], right PFC [R; N = 15]). For each of these planned comparisons, sham was used as comparator (Fig. 1D). Additionally, we analysed a full model with groups split by all three parameters within a single step and compared these eight groups to sham (Fig. 1E).
      A linear mixed model was applied to each hierarchical level since it allows to analyse performance gain over time while accounting for variability explained by the variables as well as accounting for variation not explained by these variables. In the model, ncorr was used as the dependent variable. For planned contrasts, performance of the sham group was used as the reference as we were interested in changes compared to a non-stimulated sample. Fixed effects in each model were the experimental groups at their respective hierarchical level (polarity: S, A, C; intensity: S, 1, 2; laterality: S, L, R; and lastly condition for the full model at the lowest level: S, A1L, A1R, A2L, A2R, C1L, C1R, C2L, C2R), session during which the measurement was taken (time: session two to seven), the interaction condition x time, and lastly pre-training performance as a regression coefficient (ncorr(pre): number of correct trials during pre-training), resulting in the following model: ncorr ∼ group x time + ncorr(pre). Random effects were measurement timepoint and individual subject: ∼1 + time | subject. To compare strength of effects, non-standardised (B) and standardised beta coefficients (β) were computed for each model. These models allow for the comparison of training gains, forming a slope over the course of all sessions, corresponding to how much participants improved in their performance over time.
      To investigate the stability of effects at post-training and follow-up, data of conditions that were found to be significantly different from sham during the training phase were additionally compared to sham by t-tests. Here, Cohen’s d was calculated to assess for effect sizes.

      Transfer: flanker task

      For analysis of the flanker task we used neutral, congruent and incongruent trials. Outliers ( ±3 SD deviation from overall reaction time mean), incorrect trials as well as the respective subsequent trial (which could be subject to inflated reaction times due to potential post error slowing) were removed from the data. Grouping factors of the flanker trials were chosen according to congruency, resulting in three trial type groups: congruent, incongruent, and neutral trial conditions. The repeated measures ANOVA included the dependent variable mean reaction time pooled by trial type, and the main effects trial type, condition (S, A1L, A1R, A2L, A2R, C1L, C1R, C2L, C2R), time (session in which measurement was taken), as well as the interactions time x condition and time x condition x trial type.

      Questionnaires

      Comparisons regarding data gathered from the anamnesis and questionnaires preceding the experimental procedure to ensure group homogeneity were performed on the nine conditions (S, A1L, A1R, A2L, A2R, C1L, C1R, C2L, C2R) as independent factors and the respective questionnaire outcome as dependent variable. Fisher’s Exact Test was used to analyse the distribution across groups of the following variables: gender, educational degree, stage of menstrual cycle for women, application of hormonal contraception, and smoking behaviour. For the distribution of age across groups, the Kruskal-Wallis H test was used. Due to non-homogenous variances, handedness was analysed by Welch ANOVA. QCM items were analysed by one-factorial ANOVAs. Effects of PASAT and tDCS on PANAS scores were analysed by repeated measures ANOVAs, with PANAS scores in the nine conditions before and after PASAT completion (session) as well as before and after the course of training (training). Positive affect (PA) and negative affect (NA) were analysed separately. Adverse effects were analysed item-wise by one-factorial ANOVA. Blinding was evaluated through χ2-test.

      Results

      Sample characteristics

      Composition of the nine groups did not differ regarding demographic data. See Table 1 for a detailed overview of group compositions and the respective test statistics. No differences were found between groups for pre-training performance in the PASAT (conditions: F(8, 153) = 0.939, p = 0.486, η2 = 0.047).

      Effects of tDCS on cognitive control training

      Fig. 2 shows the performance (ncorr) for each level of analysis (polarity [Fig. 2A], intensity [Fig. 2B], laterality [Fig. 2C]). Main effect time was significant (p < 0.001), illustrating that subjects’ performance continually improved in all intervention conditions. Pre-training performance turned out to be a predictor for overall performance increase, ncorr(pre) was significant in all analyses (p < 0.001): subjects performing better during pre-training also showed higher performance gains. Please see Supplementary Table 1 for the numeric values.
      Fig. 2
      Fig. 2Shown are the number of correct trials (sum of the respective session) for each analysis level (A: polarity; B: intensity; C: laterality). Performance of the sham group is shown as a line plot to facilitate comparison. All experimental groups improved significantly over time. In each planned comparison, sham was used as the reference and no comparisons were performed between stimulation groups. A For the first analysis level, the anodal group proved to benefit significantly over sham, therefore this group was further divided by their respective intensity levels (B), for which 1 mA showed significant effects compared to sham. After then additionally dividing this group by laterality (C), we were able to deduce that the left side of this subgroup provided significant performance gain over sham while the right side showed a trend in comparison to sham. Groups that benefited from tDCS and therefore exhibited increased performance are outlined in bold. Error bars show standard error of the mean.

      TDCS-enhanced training: hierarchical analysis

      We found an effect for anodal polarity (t(806) = 2.27, p = 0.0235; B = 2.66, SE = 1.17; β = 0.04, SE = 0.02), indicating that anodal tDCS applied during PASAT training enhances performance gains compared to sham stimulation. Cathodal polarity did not yield an effect (t(806) = 0.82, p = 0.4129; B = 0.96, SE = 1.17; β = 0.02, SE = 0.02). The Bonferroni-Holm corrected threshold for level of significance for this level of analysis is 0.025.
      Subsequently narrowing down the effects of anodal stimulation, intensities of 1 and 2 mA were compared to sham tDCS. Here, 1 mA (t(511) = 2.71, p = 0.0069; B = 3.69, SE = 1.36; β = 0.06, SE = 0.02) showed a highly significant increase in performance gain compared to sham, but 2 mA did not support training (t(511) = 1.19, p = 0.2345; B = 1.62, SE = 1.36; β = 0.02, SE = 0.02). The Bonferroni-Holm corrected threshold for level of significance for this level of analysis is 0.01.
      Finally, testing the effect of stimulation laterality (left, right) in the subsample of 1 mA anodal tDCS, we found a significant effect of the left (t(362) = 2.53, p = 0.0117; B = 4.40, SE = 1.74; β = 0.06, SE = 0.02), but not the right PFC (t(362) = 1.72, p = 0.0869; B = 2.98, SE = 1.74; β = 0.04, SE = 0.02). The Bonferroni-Holm corrected threshold for level of significance for this level of analysis is 0.0167.

      TDCS-enhanced training: single group analysis (Supplementary Table 5)

      Complementary, testing each of the applied tDCS conditions against sham tDCS within a single model indicated that 1 mA anodal tDCS to the left PFC significantly increased PASAT performance gains with training (t(800) = 2.55, p = 0.0111; B = 4.40, SE = 1.73; β = 0.04, SE = 0.02). The Bonferroni-Holm corrected threshold for level of significance for this level of analysis is 0.0125.

      TDCS-enhanced training: sensitivity analysis (Supplementary Table 6)

      To control for possible placebo effects, a sensitivity analysis was performed, including only subjects who assumed to have received verum tDCS. The result coincides with the overall analysis, as the A1L group still shows significantly improved performance gains over sham (t(670) = 1.98, p = 0.0485; B = 3.91, SE = 1.98; β = 0.04, SE = 0.02). The Bonferroni-Holm corrected threshold for level of significance for this level of analysis is 0.05.

      Pre-, post-training, and follow-up

      To investigate the stability of tDCS effects, post-training and follow-up performance were compared for each condition to sham on the different levels of analysis. No persistent effects were found for the overall group of anodal polarity. After training with 1 mA anodal tDCS, superior effects in comparison with sham stimulation were measurable at post-training (p = 0.043) and follow-up (p = 0.017). At the single group level (1 mA anodal tDCS to the left PFC), enhanced performance was observed at the post- (p = 0.049) but not the follow-up session. See Table 2 for the statistics of the conducted t-tests; for the sake of completeness, we show calculations for all sessions.
      Table 2t-test statistics showing stability of tDCS effects. Depicted are mean number of correct trials at the respective level (ncorr (level)), differences in number of correct trials compared to the sham group (Δ_ncorr), standard deviations (SD), standard errors of the mean (SEM), and test statistics.∗p < 0.05; ∗∗p < 0.01; ∗∗∗p < 0.001.
      Stability of tDCS effects over time
      Polarity level: S vs. A (N = 43 vs. 60)
      Sessionncorr (S)ncorr (A)Δ_ncorrSD (S)SD (A)SEM (S)SEM (A)tdfpCohen’s d
      1118.67122.45−3.77634.50334.1265.2624.406−0.5511010.5830.110
      2156.95160.13−3.18045.81038.9676.9865.031−0.3791010.7050.075
      3182.67186.05−3.37651.25142.3817.8165.518−0.3641000.7170.072
      4196.51206.18−9.67246.76944.0267.1325.684−1.0711010.2870.213
      5208.14218.93−10.79449.65640.9117.5725.282−1.2071010.2300.237
      6217.42230.68−13.26549.21738.2937.5054.944−1.5381010.1270.301
      7222.84238.70−15.86351.06039.2887.7875.072−1.70775.5240.0920.348
      8229.70239.19−9.48951.61037.8127.8704.923−1.02273.1800.3100.210
      9219.53234.02−14.48250.53340.5907.7065.240−1.6111010.1100.316
      Intensity level: S vs. A1 (N = 43 vs. 30)
      Sessionncorr (S)ncorr (A1)Δ_ncorrSD (S)SD (A1)SEM (S)SEM (A1)tdfpCohen’s d
      1118.67127.47−8.79234.50327.4305.2625.008−1.162710.2490.282
      2156.95164.73−7.78045.81031.0586.9865.670−0.86570.9630.3900.199
      3182.67195.00−12.32651.25136.2867.8166.625−1.133710.2610.278
      4196.51215.07−18.55546.76932.7687.1325.983−1.99370.9940.050∗0.460
      5208.14229.57−21.42749.65632.6407.5725.959−2.22470.8020.029∗0.510
      6217.42239.37−21.94849.21728.4997.5055.203−2.40368.9920.019∗0.546
      7222.84250.10−27.26351.06028.8537.7875.268−2.90068.4710.005∗∗0.657
      8229.70248.73−19.03651.61026.5897.8704.854−2.05966.1660.043∗0.464
      9219.53243.70−24.16550.53334.0037.7066.208−2.44270.9350.017∗0.561
      Laterality level: S vs. A1L (N = 43 vs. 15)
      Sessionncorr (S)ncorr (A1L)Δ_ncorrSD (S)SD (A1L)SEM (S)SEM (A1L)tdfpCohen’s d
      1118.67126.73−8.05934.50328.5425.2627.369−0.812560.4200.255
      2156.95158.60−1.64745.81027.9546.9867.218−0.16440.6320.8710.043
      3182.67195.47−12.79251.25137.4287.8169.664−0.886560.3800.285
      4196.51210.20−13.68846.76935.1847.1329.084−1.034560.3060.331
      5208.14224.53−16.39449.65634.2347.5728.839−1.181560.2430.384
      6217.42239.67−22.24849.21726.7577.5056.909−2.18145.4460.034∗∗0.562
      7222.84249.07−26.22951.06029.5727.7877.635−2.40542.8230.021∗∗0.629
      8229.70249.13−19.43651.61021.6207.8705.582−2.01453.9370.049∗0.491
      9219.53239.20−19.66550.53333.5937.7068.674−1.399560.1670.458

      Changes of affective state (PANAS)

      Analyses of PA revealed that the interaction of session and training was significant, indicating a slight decrease over the course of the experiment (F(1, 146) = 12.048, p < 0.001, η2 = 0.076). For NA, session was significant (F(1, 148) = 28.634, p < 0.001, η2 = 0.162), with increase in NA immediately after completion of the PASAT within the session. NA decreased between sessions as reflected by significance of training (F(1, 148) = 13.567, p < 0.001, η2 = 0.084). The interaction of session and training was also significant, pointing towards a reduction of PASAT-induced NA with training (F(1, 148) = 17.272, p < 0.001, η2 = 0.105). See Table 3 for the reported mean scores.
      Table 3PANAS sum scores (standard deviations in parentheses). Missing Ns resulted from subjects not reporting a score for at least one adjective of the PANAS, hence the overall score was not calculated.
      Timepoint within sessionBefore PASATAfter PASATBefore PASATAfter PASATBefore PASATAfter PASATBefore PASATAfter PASAT
      SessionPositive affectNegative affect
      Pre-trainingPost-trainingPre-trainingPost-training
      S (N = 41)30.049 (6.797)30 (7.029)26.439 (7.991)26.073 (8.563)12.286 (2.075)14.952 (4.288)12.595 (4.591)12.714 (3.293)
      A1L (N = 15)30.533 (5.705)31.6 (6.905)27.067 (7.216)24.667 (8.516)14.467 (6.457)13.467 (5.167)13.667 (6.651)13.667 (5.778)
      A1R (N = 14)28.5 (4.848)29.643 (7.792)26.857 (7.655)26.714 (9.042)11.571 (1.651)15.286 (3.931)11.429 (1.604)12.071 (3.025)
      A2L (N = 14)31.643 (7.045)30.786 (7.557)29.071 (7.322)29.5 (9.59)11.8 (2.731)14.667 (4.865)12.067 (4.284)12.4 (3.996)
      A2R (N = 15)26.867 (5.54)32.2 (7.002)25.733 (8.336)27.933 (7.166)12.071 (3.339)13.929 (5.595)11.714 (1.939)12.143 (2.248)
      C1L (N = 15)31.667 (6.102)29.133 (5.397)31.733 (6.819)27.267 (6.017)12.267 (2.463)15.4 (5.889)11.4 (1.502)13.467 (5.963)
      C1R (N = 12)23.333 (4.774)27.75 (7.06)25.5 (4.89)25.333 (6.499)12.2 (2.624)16.133 (5.475)12.133 (2.696)13.133 (3.796)
      C2L (N = 15)31.467 (5.693)34.2 (6.951)29.467 (8.236)28.533 (8.175)12 (2.353)13.143 (4.4)12.071 (3.339)12.143 (2.958)
      C2R (N = 14)26.714 (6.031)31.643 (6.476)24.571 (4.274)27.286 (4.697)11.538 (2.367)13.385 (3.948)11.231 (2.204)12.538 (5.06)

      Effects of tDCS and CC training on the flanker transfer task

      In all analyses and under all conditions, we found highly significant effects of time, trial type, and the interaction time x trial type, indicating that subjects improved over time and that the trial type consistently affected subjects’ reaction times with congruent trials eliciting faster reaction times than incongruent trials. No effects of tDCS on flanker task performance were found. For an overview of the flanker scores please refer to Supplementary Table 7, for test statistics to Supplementary Table 8.

      Adverse effects

      There were statistically significant differences between groups for itching sensations under the electrode surface (F(8, 153) = 2.081, p = 0.041, η2 = 0.098). Least significant post hoc analysis showed that the sensations were significantly stronger in A2L, A2R, and C2R compared to S. Significant differences were also found for overall itching sensations (F(8, 153) = 2.910, p = 0.005, η2 = 0.132), with post hoc analysis revealing that sensations were stronger in A2L, A2R, and C2R compared to S. Overall, these results show that higher intensities generally caused more noticeable sensations.
      No differences were found for tingling sensations on the head area (F(8, 153) = 1.364, p = 0.217, η2 = 0.067), fatigue (F(8, 153) = 1.456, p = 0.178, η2 = 0.071), headache (F(8, 153) = 0.987, p = 0.448, η2 = 0.049), nausea (F(8, 153) = 0.667, p = 0.720, η2 = 0.034), and other miscellaneous effects (F(8, 153) = 1.488, p = 0.166, η2 = 0.072). See Supplementary Table 9 for the reported values regarding the magnitude of the perceived sensations.

      Blinding

      Of the sham group (N = 43), 29 subjects (67.44 %) guessed that they received verum stimulation. Out of the 119 subjects that actually received verum tDCS, 107 (89.92 %) guessed correctly. Of note, in the verum group, ratios did not differ regarding stimulation intensity (Fig. 3): 53/60 subjects (1 mA), 54/59 subjects (2 mA), χ2 (1, N = 119) = 0.33, p = 0.563, Cramer’s V = 0.053.
      Fig. 3
      Fig. 3Number of subjects and their respective guesses on blinding. Regardless of stimulation intensity, subjects were able to discern whether they received verum tDCS or not. The majority of subjects in the sham group also assumed to have received verum stimulation.

      Discussion

      To provide comprehensive evidence for the beneficial effects of tDCS on CC training, and the critical dependency of these effects on stimulation parameters, we systematically tested multiple tDCS parameter combinations applied concurrently to a challenging modified two-back adaptive PASAT training in a large sample of healthy subjects (N = 162). By means of hierarchical analysis we found that: i) CC training gains were enhanced by anodal - but not cathodal - tDCS and superior to sham stimulation, confirming the polarity-dependence of tDCS; ii) stimulation intensity of 1 mA was superior to sham tDCS, supporting previous findings that the influence of stimulation intensity on task-related cognitive processing and plasticity is not linear [
      • Batsikadze G.
      • Moliadze V.
      • Paulus W.
      • Kuo M.-F.
      • Nitsche M.A.
      Partially non-linear stimulation intensity-dependent effects of direct current stimulation on motor cortex excitability in humans: effect of tDCS on cortical excitability.
      ]; iii) this effect was particularly salient when applying 1 mA anodal tDCS to the left PFC, suggesting spatial specificity as right-sided stimulation missed the significance level; and iv) the effects were observable until three months after training.
      These results are particularly valuable to identify reliable stimulation protocols for further development of translational applications [
      • Giordano J.
      • Bikson M.
      • Kappenman E.S.
      • Clark V.P.
      • Coslett H.B.
      • Hamblin M.R.
      • et al.
      Mechanisms and effects of transcranial direct current stimulation.
      ,
      • Polanía R.
      • Nitsche M.A.
      • Ruff C.C.
      Studying and modifying brain function with non-invasive brain stimulation.
      ]. Current lack of conclusive evidence might be primarily due to the multitude of studies with small sample sizes, ill-defined mechanistic models, varying parameter settings, and experimental designs [
      • Polanía R.
      • Nitsche M.A.
      • Ruff C.C.
      Studying and modifying brain function with non-invasive brain stimulation.
      ]. With this study, we addressed these requirements by investigating a large sample in a standardised intervention, identifying reliably effective stimulation conditions. We tested the effects on adaptive plasticity measured by means of training effects in contrast to single session interventions. Although there is no doubt about the malleability of cognitive functions with transcranial brain stimulation [
      • Duecker F.
      • de Graaf T.A.
      • Sack A.T.
      Thinking caps for everyone? The role of neuro-enhancement by non-invasive brain stimulation in neuroscience and beyond.
      ,
      • Colzato L.S.
      • Hommel B.
      The future of cognitive training.
      ], beneficial neuroplastic effects of tDCS as reflected in lasting improvements or amelioration of goal-directed behaviour are still under debate [
      • Bestmann S.
      • de Berker A.O.
      • Bonaiuto J.
      Understanding the behavioural consequences of noninvasive brain stimulation.
      ]. Therefore, the primary endpoint of our study was performance-gain during a training paradigm. Harnessing state-dependency of tDCS effects [
      • Zwissler B.
      • Sperber C.
      • Aigeldinger S.
      • Schindler S.
      • Kissler J.
      • Plewnia C.
      Shaping memory accuracy by left prefrontal transcranial direct current stimulation.
      ,
      • Schroeder P.A.
      • Plewnia C.
      Beneficial effects of cathodal transcranial direct current stimulation (tDCS) on cognitive performance.
      ,
      • Hsu T.-Y.
      • Juan C.-H.
      • Tseng P.
      Individual differences and state-dependent responses in transcranial direct current stimulation.
      ,
      • Li L.M.
      • Violante I.R.
      • Leech R.
      • Ross E.
      • Hampshire A.
      • Opitz A.
      • et al.
      Brain state and polarity dependent modulation of brain networks by transcranial direct current stimulation.
      ], functional targeting [
      • Bikson M.
      • Rahman A.
      Origins of specificity during tDCS: anatomical, activity-selective, and input-bias mechanisms.
      ] of the cognitive control networks was implemented by stimulating strictly concurrent to PASAT training, in accordance with previous studies which showed a boosting effect of tDCS on task-relevant networks [
      • Polanía R.
      • Nitsche M.A.
      • Paulus W.
      Modulating functional connectivity patterns and topological functional organization of the human brain with transcranial direct current stimulation.
      ].
      Regarding stimulation polarity, the dichotomy of anodal (activity-enhancing) and cathodal (activity-decreasing) effects has often been challenged in research [
      • Garnett E.O.
      • Malyutina S.
      • Datta A.
      • Ouden D-B den
      On the use of the terms anodal and cathodal in high-definition transcranial direct current stimulation: a technical note.
      ]. Here, we provide new evidence for polarity-dependence of tDCS effects on CC, showing that only anodal stimulation over the PFC boosted training gains. Additionally, our data support the notion that tDCS causes non-linear effects, e. g. higher intensities do not necessarily elicit more prominent outcomes in a cognitive task, which was so far shown mainly for physiological tDCS effects [
      • Batsikadze G.
      • Moliadze V.
      • Paulus W.
      • Kuo M.-F.
      • Nitsche M.A.
      Partially non-linear stimulation intensity-dependent effects of direct current stimulation on motor cortex excitability in humans: effect of tDCS on cortical excitability.
      ,
      • Jamil A.
      • Batsikadze G.
      • Kuo H.-I.
      • Labruna L.
      • Hasan A.
      • Paulus W.
      • et al.
      Systematic evaluation of the impact of stimulation intensity on neuroplastic after-effects induced by transcranial direct current stimulation.
      ]. None of the higher-intensity groups analysed in our study showed significantly improved performance gains compared to sham intervention. Particularly relevant for the clinical context, these findings imply that a simple increase of stimulation intensity does not necessarily enhance efficacy but, by contrast, might compromise efficacy of the respective intervention. Nevertheless, clinical populations may require higher stimulation intensities due to pathology- or medication-dependent impairments of neuroplasticity [
      • Fuchsova B.
      • Alvarez Juliá A.
      • Rizavi H.S.
      • Frasch A.C.
      • Pandey G.N.
      Altered expression of neuroplasticity-related genes in the brain of depressed suicides.
      ,
      • Schwippel T.
      • Papazova I.
      • Strube W.
      • Fallgatter A.J.
      • Hasan A.
      • Plewnia C.
      Beneficial effects of anodal transcranial direct current stimulation (tDCS) on spatial working memory in patients with schizophrenia.
      ]. Finally, laterality of tDCS modulating potentially lateralised cognitive functions might be critical for yielding effects, as shown by our present and previous studies [
      • Ruf S.P.
      • Fallgatter A.J.
      • Plewnia C.
      Augmentation of working memory training by transcranial direct current stimulation (tDCS).
      ]. However, conclusive evidence for clear hypotheses on CC functions is scarce [
      • Stephan K.E.
      • Marshall J.C.
      • Friston K.J.
      • Rowe J.B.
      • Ritzl A.
      • Zilles K.
      • et al.
      Lateralized cognitive processes and lateralized task control in the human brain.
      ,
      • Vanderhasselt M.-A.
      • De Raedt R.
      • Baeken C.
      Dorsolateral prefrontal cortex and Stroop performance: tackling the lateralization.
      ]. Consistently, although we observed superior effects of 1 mA anodal tDCS to the left PFC, the trendwise effect of 1 mA anodal tDCS to the right PFC does not allow drawing definite conclusions regarding the prefrontal lateralization of CC processes.
      Since the PASAT was used to target CC of emotionally relevant information, affective states were assessed regularly. The PASAT was indeed frustrating as reflected by an increased NA after completion of the task within a session. Interestingly, the magnitude of NA changes decreased with training, which might be due to habituation, but could also indicate improved CC which might support keeping focus on the task at hand and not the distracting, often negative, feedback, hence lowering frustration elicited by the task. Nevertheless, our data do not reveal an effect of tDCS on the affective responses measured with the PANAS as indicated by our previous trials [
      • Plewnia C.
      • Schroeder P.A.
      • Kunze R.
      • Faehling F.
      • Wolkenstein L.
      Keep calm and carry on: improved frustration tolerance and processing speed by transcranial direct current stimulation (tDCS).
      ,
      • Wiegand A.
      • Sommer A.
      • Nieratschker V.
      • Plewnia C.
      Improvement of cognitive control and stabilization of affect by prefrontal transcranial direct current stimulation (tDCS).
      ]. Apparently, the effect of tDCS on acute mood states in healthy subjects is less consistently measurable than on cognitive performance. Additionally, not limited to tDCS but also observable in rTMS trials, emotional states in healthy subjects are quite resistant against modulatory interventions. This might be due to a low precision of assessment tools and high variability of affective reactions to the PASAT interacting with stimulation.
      Although in our study effects of tDCS on CC training gains were unambiguous and partially lasting for at least three months, transfer effects as tested with the flanker task were absent. However, the flanker task is rather far from the trained and improved two-back PASAT and transferability of cognitive training gains to other cognitive domains is generally under debate [
      • Martin D.M.
      • Liu R.
      • Alonzo A.
      • Green M.
      • Player M.J.
      • Sachdev P.
      • et al.
      Can transcranial direct current stimulation enhance outcomes from cognitive training? A randomized controlled trial in healthy participants.
      ,
      • Baniqued P.L.
      • Allen C.M.
      • Kranz M.B.
      • Johnson K.
      • Sipolins A.
      • Dickens C.
      • et al.
      Working memory, reasoning, and task switching training: transfer effects, limitations, and great expectations? Akyürek E, editor.
      ]; our study sample of healthy subjects at the prime of their cognitive abilities poses additional challenges for improvements. Yet, the lack of transfer effects argues for the specificity of tDCS for trained tasks in accordance with the notion of a synergistic effect of training and stimulation-activated networks [
      • Bikson M.
      • Rahman A.
      Origins of specificity during tDCS: anatomical, activity-selective, and input-bias mechanisms.
      ,
      • Plewnia C.
      • Schroeder P.A.
      • Wolkenstein L.
      Targeting the biased brain: non-invasive brain stimulation to ameliorate cognitive control.
      ].
      A relevant limitation of this study is effective blinding: verum tDCS was detected by most participants, independent from stimulation intensity. In the sham group, most but nonetheless fewer subjects assumed having received verum stimulation, following results of studies highlighting challenges of proper tDCS blinding [
      • Greinacher R.
      • Buhôt L.
      • Möller L.
      • Learmonth G.
      The time course of ineffective sham blinding during low-intensity (1mA) transcranial direct current stimulation.
      ,
      • Turi Z.
      • Csifcsák G.
      • Boayue N.M.
      • Aslaksen P.
      • Antal A.
      • Paulus W.
      • et al.
      Blinding is compromised for transcranial direct current stimulation at 1 mA for 20 minutes in young healthy adults.
      ]. Recent findings show that blinding aided by topical anaesthetics was successful up to 3 mA [
      • Mosayebi Samani M.
      • Agboada D.
      • Jamil A.
      • Kuo M.-F.
      • Nitsche M.A.
      Titrating the neuroplastic effects of cathodal transcranial direct current stimulation (tDCS) over the primary motor cortex.
      ]. Nevertheless, as established by sensitivity analysis, our results remain basically the same when individuals that judged their stimulation as ‘not real’ were removed. Furthermore, to warrant a safe and valid stimulation of three subjects in parallel, this study was conducted in a single-blind fashion. To avoid experimenter bias, randomisation was performed prior to pre-training measures, instructions were strictly read from scripts, and other interactions with the subjects were limited to a minimum.
      In conclusion, our study confirms polarity-dependent, non-linear, beneficial effects of optimised tDCS on CC training. Based on systematic testing of parameters in a large group of healthy subjects, these data provide a solid basis for further developments in the neuroscientific and clinical use of electrical brain stimulation. Linking tDCS with cognitive tasks may allow for targeted enhanced brain network retraining, opening new perspectives for cognitive enhancement and efficient treatment of symptoms in various psychiatric disorders.

      Author contributions

      CP designed the study. SW performed the experiments. SW and CP analysed the data, interpreted the results, and wrote the manuscript; SW, MAN, and CP participated in the result interpretation and finalised the paper. All authors read and approved the final manuscript.

      CRediT authorship contribution statement

      Simone Weller: Methodology, Formal analysis, Investigation, Software, Data curation, Writing - original draft, Writing - review & editing, Visualization. Michael A. Nitsche: Writing - review & editing. Christian Plewnia: Conceptualization, Methodology, Formal analysis, Resources, Writing - original draft, Writing - review & editing, Supervision, Project administration, Funding acquisition.

      Declaration of competing interest

      SW and CP declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest. MAN is scientific advisor for Neuroelectrics, and NeuroDevice.

      Acknowledgements

      This study was supported by the GCBS research consortium funded by the German Federal Ministry of Education and Research (FKZ: 01EE1403D) and was registered under ClinicalTrials.gov (NCT04108663).

      Appendix A. Supplementary data

      The following is the Supplementary data to this article:

      References

        • Fitzgerald P.B.
        • Laird A.R.
        • Maller J.
        • Daskalakis Z.J.
        A meta-analytic study of changes in brain activation in depression.
        Hum Brain Mapp. 2008 Jun 1; 29: 683-695
        • Baune B.T.
        • Miller R.
        • McAfoose J.
        • Johnson M.
        • Quirk F.
        • Mitchell D.
        The role of cognitive impairment in general functioning in major depression.
        Psychiatr Res. 2010 Apr; 176: 183-189
        • Dixon M.L.
        Cognitive control, emotional value, and the lateral prefrontal cortex.
        Front Psychol. 2015; 6
        • Ryman S.G.
        • Shaikh A.A.E.
        • Shaff N.A.
        • Hanlon F.M.
        • Dodd A.B.
        • Wertz C.J.
        • et al.
        Proactive and reactive cognitive control rely on flexible use of the ventrolateral prefrontal cortex.
        Hum Brain Mapp. 2018; 40: 955-966
        • McTeague L.M.
        • Huemer J.
        • Carreon D.M.
        • Jiang Y.
        • Eickhoff S.B.
        • Etkin A.
        Identification of common neural circuit disruptions in cognitive control across psychiatric disorders.
        Am J Psychiatr. 2017 Mar 21; 174: 676-685
        • Miskowiak K.W.
        • Petersen C.S.
        Neuronal underpinnings of cognitive impairment and - improvement in mood disorders.
        CNS Spectr. 2018 Sep 10; : 1-24
        • Burgess P.W.
        • Stuss D.T.
        Fifty years of prefrontal cortex research: impact on assessment.
        J Int Neuropsychol Soc. 2017 Oct; 23: 755-767
        • Barbey A.K.
        • Koenigs M.
        • Grafman J.
        Dorsolateral prefrontal contributions to human working memory.
        Cortex. 2013 May; 49: 1195-1205
        • Brunoni A.R.
        • Vanderhasselt M.-A.
        Working memory improvement with non-invasive brain stimulation of the dorsolateral prefrontal cortex: a systematic review and meta-analysis.
        Brain Cognit. 2014 Apr; 86: 1-9
        • Calkins A.W.
        • McMorran K.E.
        • Siegle G.J.
        • Otto M.W.
        The effects of computerized cognitive control training on community adults with depressed mood.
        Behav Cognit Psychother. 2015 Sep; 43 (05): 578-589
        • Nitsche M.A.
        • Paulus W.
        Excitability changes induced in the human motor cortex by weak transcranial direct current stimulation.
        J Physiol. 2000 Sep; 527: 633-639
        • Grossman P.
        • Woods A.J.
        • Knotkova H.
        • Bikson M.
        Safety of transcranial direct current stimulation.
        in: Knotkova H. Nitsche M.A. Bikson M. Woods A.J. Practical guide to transcranial direct current stimulation. Springer International Publishing, Cham2019: 167-195
        • Cambiaghi M.
        • Velikova S.
        • Gonzalez-Rosa J.J.
        • Cursi M.
        • Comi G.
        • Leocani L.
        Brain transcranial direct current stimulation modulates motor excitability in mice.
        Eur J Neurosci. 2010 Feb; 31: 704-709
        • Zwissler B.
        • Sperber C.
        • Aigeldinger S.
        • Schindler S.
        • Kissler J.
        • Plewnia C.
        Shaping memory accuracy by left prefrontal transcranial direct current stimulation.
        J Neurosci. 2014 Mar 12; 34: 4022-4026
        • Kronberg G.
        • Bridi M.
        • Abel T.
        • Bikson M.
        • Parra L.C.
        Direct current stimulation modulates LTP and LTD: activity dependence and dendritic effects.
        Brain Stimulat. 2017 Jan 1; 10: 51-58
        • Ruf S.P.
        • Fallgatter A.J.
        • Plewnia C.
        Augmentation of working memory training by transcranial direct current stimulation (tDCS).
        Sci Rep. 2017; 7: 1-11
        • Fritsch B.
        • Reis J.
        • Martinowich K.
        • Schambra H.M.
        • Ji Y.
        • Cohen L.G.
        • et al.
        Direct current stimulation promotes BDNF-dependent synaptic plasticity: potential implications for motor learning.
        Neuron. 2010 Apr; 66: 198-204
        • Bikson M.
        • Rahman A.
        Origins of specificity during tDCS: anatomical, activity-selective, and input-bias mechanisms.
        Front Hum Neurosci. 2013; 7: 688
        • Bikson M.
        • Brunoni A.R.
        • Charvet L.E.
        • Clark V.P.
        • Cohen L.G.
        • Deng Z.-D.
        • et al.
        Rigor and reproducibility in research with transcranial electrical stimulation: an NIMH-sponsored workshop.
        Brain Stimulat. 2018 May 1; 11: 465-480
        • Guleyupoglu B.
        • Schestatsky P.
        • Edwards D.
        • Fregni F.
        • Bikson M.
        Classification of methods in transcranial Electrical Stimulation (tES) and evolving strategy from historical approaches to contemporary innovations.
        J Neurosci Methods. 2013 Oct 15; 219: 297-311
        • Schroeder P.A.
        • Nuerk H.-C.
        • Plewnia C.
        Prefrontal neuromodulation reverses spatial associations of non-numerical sequences, but not numbers.
        Biol Psychol. 2017 Sep 1; 128: 39-49
        • Polanía R.
        • Nitsche M.A.
        • Paulus W.
        Modulating functional connectivity patterns and topological functional organization of the human brain with transcranial direct current stimulation.
        Hum Brain Mapp. 2011; 32: 1236-1249
        • Segrave R.A.
        • Arnold S.
        • Hoy K.
        • Fitzgerald P.B.
        Concurrent cognitive control training augments the antidepressant efficacy of tDCS: a pilot study.
        Brain Stimulat. 2014 Mar; 7: 325-331
        • Brem A.-K.
        • Almquist J.N.-F.
        • Mansfield K.
        • Plessow F.
        • Sella F.
        • Santarnecchi E.
        • et al.
        Modulating fluid intelligence performance through combined cognitive training and brain stimulation.
        Neuropsychologia. 2018 Apr 9; 118: 107-114
        • Ward N.
        • Paul E.
        • Watson P.
        • Cooke G.E.
        • Hillman C.H.
        • Cohen N.J.
        • et al.
        Enhanced learning through multimodal training: evidence from a comprehensive cognitive, physical fitness, and neuroscience intervention.
        Sci Rep. 2017 Jul 19; 7: 1-8
        • Cramer S.C.
        • Sur M.
        • Dobkin B.H.
        • O’Brien C.
        • Sanger T.D.
        • Trojanowski J.Q.
        • et al.
        Harnessing neuroplasticity for clinical applications.
        Brain. 2011 Jun 1; 134: 1591-1609
        • Plewnia C.
        • Schroeder P.A.
        • Wolkenstein L.
        Targeting the biased brain: non-invasive brain stimulation to ameliorate cognitive control.
        Lancet Psychiatry. 2015 Apr; 2: 351-356
        • Kekic M.
        • Boysen E.
        • Campbell I.C.
        • Schmidt U.
        A systematic review of the clinical efficacy of transcranial direct current stimulation (tDCS) in psychiatric disorders.
        J Psychiatr Res. 2016 Mar 1; 74: 70-86
        • Vicario C.M.
        • Salehinejad M.A.
        • Felmingham K.
        • Martino G.
        • Nitsche M.A.
        A systematic review on the therapeutic effectiveness of non-invasive brain stimulation for the treatment of anxiety disorders.
        Neurosci Biobehav Rev. 2019 Jan 1; 96: 219-231
        • Feeser M.
        • Prehn K.
        • Kazzer P.
        • Mungee A.
        • Bajbouj M.
        Transcranial direct current stimulation enhances cognitive control during emotion regulation.
        Brain Stimulat. 2014 Jan; 7: 105-112
        • Plewnia C.
        • Schroeder P.A.
        • Kunze R.
        • Faehling F.
        • Wolkenstein L.
        Keep calm and carry on: improved frustration tolerance and processing speed by transcranial direct current stimulation (tDCS).
        PloS One. 2015 Apr 2; 10 (e0122578)
        • Chen N.T.M.
        • Basanovic J.
        • Notebaert L.
        • MacLeod C.
        • Clarke P.J.F.
        Attentional bias mediates the effect of neurostimulation on emotional vulnerability.
        J Psychiatr Res. 2017 Oct 1; 93: 12-19
        • Wolkenstein L.
        • Plewnia C.
        Amelioration of cognitive control in depression by transcranial direct current stimulation.
        Biol Psychiatr. 2013 Apr; 73: 646-651
        • Nelson J.
        • McKinley R.A.
        • Phillips C.
        • McIntire L.
        • Goodyear C.
        • Kreiner A.
        • et al.
        The effects of transcranial direct current stimulation (tDCS) on multitasking throughput capacity.
        Front Hum Neurosci. 2016 Nov 29; 10: 589
        • Krause B.
        • Cohen Kadosh R.
        Not all brains are created equal: the relevance of individual differences in responsiveness to transcranial electrical stimulation.
        Front Syst Neurosci. 2014; 8: 25
        • Nilsson J.
        • Lebedev A.V.
        • Rydström A.
        • Lövdén M.
        Direct-current stimulation does little to improve the outcome of working memory training in older adults.
        Psychol Sci. 2017 Jul 1; 28: 907-920
        • Berryhill M.E.
        • Martin D.
        Cognitive effects of transcranial direct current stimulation in healthy and clinical populations: an overview.
        J ECT. 2018 Aug 8; 34: e25-e35
        • Wörsching J.
        • Padberg F.
        • Goerigk S.
        • Heinz I.
        • Bauer C.
        • Plewnia C.
        • et al.
        Testing assumptions on prefrontal transcranial direct current stimulation: comparison of electrode montages using multimodal fMRI.
        Brain Stimulat. 2018 May 4; 11: 998-1007
        • Chase H.W.
        • Boudewyn M.A.
        • Carter C.S.
        • Phillips M.L.
        Transcranial direct current stimulation: a roadmap for research, from mechanism of action to clinical implementation.
        Mol Psychiatr. 2019 Aug 27; : 1-11
        • Tombaugh T.N.
        A comprehensive review of the paced auditory serial addition test (PASAT).
        Arch Clin Neuropsychol. 2006 Jan 1; 21: 53-76
        • Audoin B.
        • Ibarrola D.
        • Duong M.V.A.
        • Pelletier J.
        • Confort-Gouny S.
        • Malikova I.
        • et al.
        Functional MRI study of PASAT in normal subjects.
        Magn Reson Mater Phys Biol Med. 2005 May; 18: 96-102
        • Flegal K.E.
        • Ragland J.D.
        • Ranganath C.
        Adaptive task difficulty influences neural plasticity and transfer of training.
        Neuroimage. 2018 Dec; 188: 111-121
        • Romero Lauro L.J.
        • Rosanova M.
        • Mattavelli G.
        • Convento S.
        • Pisoni A.
        • Opitz A.
        • et al.
        TDCS increases cortical excitability: direct evidence from TMS–EEG.
        Cortex. 2014 Sep; 58: 99-111
        • Antal A.
        • Polania R.
        • Schmidt-Samoa C.
        • Dechent P.
        • Paulus W.
        Transcranial direct current stimulation over the primary motor cortex during fMRI.
        Neuroimage. 2011 Mar 15; 55: 590-596
        • Au J.
        • Katz B.
        • Buschkuehl M.
        • Bunarjo K.
        • Senger T.
        • Zabel C.
        • et al.
        Enhancing working memory training with transcranial direct current stimulation.
        J Cognit Neurosci. 2016 May 11; 28: 1419-1432
        • Pope P.A.
        • Brenton J.W.
        • Miall R.C.
        Task-specific facilitation of cognition by anodal transcranial direct current stimulation of the prefrontal cortex.
        Cerebr Cortex. 2015 Nov; 25: 4551-4558
        • Benwell C.S.Y.
        • Learmonth G.
        • Miniussi C.
        • Harvey M.
        • Thut G.
        Non-linear effects of transcranial direct current stimulation as a function of individual baseline performance: evidence from biparietal tDCS influence on lateralized attention bias.
        Cortex. 2015 Aug 1; 69: 152-165
        • Schroeder P.A.
        • Plewnia C.
        Beneficial effects of cathodal transcranial direct current stimulation (tDCS) on cognitive performance.
        J Cogn Enhanc. 2017 Mar 1; 1: 5-9
        • Wiegand A.
        • Sommer A.
        • Nieratschker V.
        • Plewnia C.
        Improvement of cognitive control and stabilization of affect by prefrontal transcranial direct current stimulation (tDCS).
        Sci Rep. 2019 May 1; 9: 6797
        • Katsoulaki M.
        • Kastrinis A.
        • Tsekoura M.
        The effects of anodal transcranial direct current stimulation on working memory.
        in: Vlamos P. GeNeDis 2016. Springer International Publishing, Cham2017: 283-289
        • Batsikadze G.
        • Moliadze V.
        • Paulus W.
        • Kuo M.-F.
        • Nitsche M.A.
        Partially non-linear stimulation intensity-dependent effects of direct current stimulation on motor cortex excitability in humans: effect of tDCS on cortical excitability.
        J Physiol. 2013 Apr 1; 591: 1987-2000
        • Kalu U.G.
        • Sexton C.E.
        • Loo C.K.
        • Ebmeier K.P.
        Transcranial direct current stimulation in the treatment of major depression: a meta-analysis.
        Psychol Med. 2012 Sep; 42: 1791-1800
        • Nitsche M.A.
        • Cohen L.G.
        • Wassermann E.M.
        • Priori A.
        • Lang N.
        • Antal A.
        • et al.
        Transcranial direct current stimulation: state of the art 2008.
        Brain Stimulat. 2008 Jul; 1: 206-223
        • Peirce J.W.
        Generating stimuli for neuroscience using PsychoPy.
        Front Neuroinf. 2009; 2
        • Eriksen B.A.
        • Eriksen C.W.
        Effects of noise letters upon the identification of a target letter in a nonsearch task.
        Percept Psychophys. 1974 Jan; 16: 143-149
        • Woods A.J.
        • Antal A.
        • Bikson M.
        • Boggio P.S.
        • Brunoni A.R.
        • Celnik P.
        • et al.
        A technical guide to tDCS, and related non-invasive brain stimulation tools.
        Clin Neurophysiol. 2016 Feb 1; 127: 1031-1048
        • Oldfield R.C.
        The assessment and analysis of handedness: the Edinburgh inventory.
        Neuropsychologia. 1971 Mar; 9: 97-113
        • Papadatou-Pastou M.
        Chapter 7 - handedness and cognitive ability: using meta-analysis to make sense of the data.
        in: Forrester G.S. Hopkins W.D. Hudry K. Lindell A. Progress in brain research. 238. Elsevier, 2018: 179-206
        • Rheinberg F.
        • Vollmeyer R.
        • Burns B.
        QCM: a questionnaire to assess current motivation in learning situations.
        Diagnostica. 2001 Jan 1; : 47
        • Watson D.
        • Anna L.
        • Tellegen A.
        Development and validation of brief measures of positive and negative affect: the PANAS scales. vol. 8. 1988
        • Armonk N.Y.
        IBM corp. IBM SPSS statistics for windows, version 24.0.0.1.
        2016 ([Released])
        • R Core Team. R
        A language and environment for statistical computing.
        R Found Stat Comput. 2018; ([Internet]) (Available from:)
        • R Core Team
        Nlme: linear and nonlinear mixed effects models.
        R Found Stat Comput. 2018; ([Internet]) (Available from:)
        • reghelper Hughes J.
        Helper funtions for regression analysis.
        R Found Stat Comput. 2018; ([Internet]) (Available from:)
        • Giordano J.
        • Bikson M.
        • Kappenman E.S.
        • Clark V.P.
        • Coslett H.B.
        • Hamblin M.R.
        • et al.
        Mechanisms and effects of transcranial direct current stimulation.
        Dose-Response. 2017 Mar 1; 15 (1559325816685467)
        • Polanía R.
        • Nitsche M.A.
        • Ruff C.C.
        Studying and modifying brain function with non-invasive brain stimulation.
        Nat Neurosci. 2018 Feb; 21: 174-187
        • Duecker F.
        • de Graaf T.A.
        • Sack A.T.
        Thinking caps for everyone? The role of neuro-enhancement by non-invasive brain stimulation in neuroscience and beyond.
        Front Syst Neurosci. 2014 Apr 29; 8: 71
        • Colzato L.S.
        • Hommel B.
        The future of cognitive training.
        ([Internet])in: Strobach T. Karbach J. Cognitive training. Springer International Publishing, Cham2016: 201-211
        • Bestmann S.
        • de Berker A.O.
        • Bonaiuto J.
        Understanding the behavioural consequences of noninvasive brain stimulation.
        Trends Cognit Sci. 2015 Jan 1; 19: 13-20
        • Hsu T.-Y.
        • Juan C.-H.
        • Tseng P.
        Individual differences and state-dependent responses in transcranial direct current stimulation.
        Front Hum Neurosci. 2016; 10: 643
        • Li L.M.
        • Violante I.R.
        • Leech R.
        • Ross E.
        • Hampshire A.
        • Opitz A.
        • et al.
        Brain state and polarity dependent modulation of brain networks by transcranial direct current stimulation.
        Hum Brain Mapp. 2019 Feb 15; 40: 904-915
        • Garnett E.O.
        • Malyutina S.
        • Datta A.
        • Ouden D-B den
        On the use of the terms anodal and cathodal in high-definition transcranial direct current stimulation: a technical note.
        Neuromodulation Technol Neural Interface. 2015; 18: 705-713
        • Jamil A.
        • Batsikadze G.
        • Kuo H.-I.
        • Labruna L.
        • Hasan A.
        • Paulus W.
        • et al.
        Systematic evaluation of the impact of stimulation intensity on neuroplastic after-effects induced by transcranial direct current stimulation.
        J Physiol. 2017; 595: 1273-1288
        • Fuchsova B.
        • Alvarez Juliá A.
        • Rizavi H.S.
        • Frasch A.C.
        • Pandey G.N.
        Altered expression of neuroplasticity-related genes in the brain of depressed suicides.
        Neuroscience. 2015 Jul 23; 299: 1-17
        • Schwippel T.
        • Papazova I.
        • Strube W.
        • Fallgatter A.J.
        • Hasan A.
        • Plewnia C.
        Beneficial effects of anodal transcranial direct current stimulation (tDCS) on spatial working memory in patients with schizophrenia.
        Eur Neuropsychopharmacol. 2018 Dec 1; 28: 1339-1350
        • Stephan K.E.
        • Marshall J.C.
        • Friston K.J.
        • Rowe J.B.
        • Ritzl A.
        • Zilles K.
        • et al.
        Lateralized cognitive processes and lateralized task control in the human brain.
        Science. 2003 Jul 18; 301: 384-386
        • Vanderhasselt M.-A.
        • De Raedt R.
        • Baeken C.
        Dorsolateral prefrontal cortex and Stroop performance: tackling the lateralization.
        Psychon Bull Rev. 2009 Jun 1; 16: 609-612
        • Martin D.M.
        • Liu R.
        • Alonzo A.
        • Green M.
        • Player M.J.
        • Sachdev P.
        • et al.
        Can transcranial direct current stimulation enhance outcomes from cognitive training? A randomized controlled trial in healthy participants.
        Int J Neuropsychopharmacol. 2013 Oct; 16 (09): 1927-1936
        • Baniqued P.L.
        • Allen C.M.
        • Kranz M.B.
        • Johnson K.
        • Sipolins A.
        • Dickens C.
        • et al.
        Working memory, reasoning, and task switching training: transfer effects, limitations, and great expectations? Akyürek E, editor.
        PloS One. 2015 Nov 10; 10e0142169
        • Greinacher R.
        • Buhôt L.
        • Möller L.
        • Learmonth G.
        The time course of ineffective sham blinding during low-intensity (1mA) transcranial direct current stimulation.
        Eur J Neurosci. 2019; 50: 3380-3388
        • Turi Z.
        • Csifcsák G.
        • Boayue N.M.
        • Aslaksen P.
        • Antal A.
        • Paulus W.
        • et al.
        Blinding is compromised for transcranial direct current stimulation at 1 mA for 20 minutes in young healthy adults.
        Eur J Neurosci. 2019; 50: 3261-3268
        • Mosayebi Samani M.
        • Agboada D.
        • Jamil A.
        • Kuo M.-F.
        • Nitsche M.A.
        Titrating the neuroplastic effects of cathodal transcranial direct current stimulation (tDCS) over the primary motor cortex.
        Cortex. 2019 Oct; 119: 350-361