Transcranial Direct Current Stimulation

Abstract

Transcranial direct current stimulation (tDCS) is a non-invasive neuromodulation technique that passes a weak, constant electrical current between two scalp electrodes to shift the excitability of the underlying cortex. This article traces the method from Bindman and colleagues' animal work on polarizing currents to the modern revival by Nitsche and Paulus that established the polarity rule: anodal stimulation raises cortical excitability while cathodal stimulation lowers it. It sets out the subthreshold physiological basis, the after-effects that outlast stimulation and depend on synaptic plasticity, and the cognitive applications from working memory to numerical competence. It weighs the efficacy debate that a large null-result review provoked, surveys the therapeutic guidelines and the ELECT-TDCS depression trial, and covers the safety record. Three demonstrations model the polarity rule, the after-effect time course, and the current-density calculation that governs dose.

Keywords: transcranial direct current stimulation, neuromodulation, cortical excitability, anodal, cathodal

A nine-volt battery held to the scalp does nothing useful; one or two milliamps of steady current passed between two saline electrodes placed with anatomical care will measurably raise the excitability of the cortex beneath the anode and lower it beneath the cathode. Transcranial direct current stimulation is built on that narrow, reproducible fact. It does not force neurons to fire; it nudges their resting membrane potential closer to or further from the threshold at which they would fire on their own, biasing spontaneous activity rather than driving it (Nitsche & Paulus, 2000). The method is cheap, portable, and painless, which explains both its rapid spread through cognitive neuroscience and the scrutiny its more ambitious claims have attracted (Filmer et al., 2014).

What Transcranial Direct Current Stimulation Is

Transcranial direct current stimulation is a form of neuromodulation, classified by MeSH under electric stimulation therapy, in which a low-intensity direct current, typically one to two milliamps, flows continuously from an anode to a cathode placed on the scalp. Unlike transcranial magnetic stimulation, which induces currents strong enough to discharge neurons, tDCS is a subthreshold technique: the current is far too weak to make a resting neuron fire, and it works instead by polarizing the membrane, raising or lowering the resting membrane potential so that the cell is more or less likely to respond to its ordinary synaptic inputs (Stagg & Nitsche, 2011). It biases the probability of firing rather than commanding it.

Because the effect is a bias, its sign depends on the direction of current flow relative to the neuron. The organizing generalization, sometimes called the polarity rule, is that stimulation under the anode depolarizes the resting membrane and raises cortical excitability, while stimulation under the cathode hyperpolarizes it and lowers excitability. That rule was first demonstrated cleanly in the human motor cortex, where the size of a muscle response evoked by a fixed magnetic pulse, the motor evoked potential, rises after anodal and falls after cathodal stimulation (Nitsche & Paulus, 2000). The rule is a starting point, not a law: in cognitive tasks, where the geometry of the stimulated region and the state of the network both matter, the neat anodal-excitatory, cathodal-inhibitory dichotomy often breaks down (Jacobson et al., 2012).

Figure 1

The Polarity Rule of tDCS

Historical Origins

The idea that a steady current changes how the brain behaves is older than the modern method. Nineteenth-century investigators applied galvanic currents to the head, and mid-twentieth-century animal experiments made the mechanism precise: Bindman, Lippold, and Redfearn showed in the rat that brief polarizing currents applied to the cerebral cortex changed the rate of spontaneous firing during the current, and, crucially, that the change persisted for hours after the current stopped (Bindman et al., 1964). That dissociation, an online effect during stimulation and a lasting after-effect once it ends, defined the two phenomena every later account would have to explain, and it located the method firmly in the physiology of neuronal polarization rather than in any special property of the current itself.

The technique lapsed for decades before its decisive modern revival at the turn of the millennium. Nitsche and Paulus demonstrated in humans that a few minutes of weak transcranial direct current over the motor cortex produced polarity-specific, reversible changes in excitability, measured as the size of the motor evoked potential, with anodal current raising it and cathodal current lowering it (Nitsche & Paulus, 2000). A companion study showed that lengthening the stimulation produced excitability elevations that outlasted the session by up to an hour, re-establishing the after-effect Bindman had reported and setting the dose parameters the field still uses (Nitsche & Paulus, 2001). By the end of the decade the method had matured enough to warrant a consensus statement of its principles, protocols, and open questions (Nitsche et al., 2008).

Physiological Basis

The two phenomena Bindman separated have two mechanisms. The immediate, online effect is a straightforward consequence of polarization: the applied field shifts the resting membrane potential of cortical neurons, and because a depolarized neuron sits closer to threshold, anodal current makes spontaneous and evoked firing more likely while cathodal current makes it less likely. This effect appears and disappears with the current and does not itself require any change in the synapses (Stagg & Nitsche, 2011). The first demonstration makes the polarity rule concrete: it lets the reader switch between an anodal and a cathodal electrode and see the modelled shift in resting membrane potential toward or away from firing threshold, reinforcing that the sign of the effect is set by the electrode, not by the current strength.

The after-effect is different in kind, and the evidence points to synaptic plasticity. The lasting excitability changes are blocked by drugs that block NMDA-receptor function and are shaped by the balance of the excitatory transmitter glutamate and the inhibitory transmitter GABA, which is the signature of mechanisms akin to long-term potentiation and long-term depression, the synaptic changes that also underlie learning (Stagg & Nitsche, 2011). Magnetic resonance spectroscopy studies find that anodal stimulation reduces local GABA while cathodal stimulation reduces glutamatergic activity, tying the macroscopic polarity rule to a specific neurochemical shift. This plasticity framing matters for cognition: if tDCS after-effects are LTP-like, then pairing stimulation with a task that is itself driving plasticity should interact with what the brain is already doing, which is why stimulation is increasingly delivered during, rather than before, the behaviour it aims to change (Polania et al., 2018). The second demonstration renders the reported dissociation between the two phenomena as a schematic curve: very brief stimulation changes excitability only while the current flows, but longer sessions engage plasticity and produce after-effects that outlast the session and saturate, letting the reader move a duration marker along the curve.

Dose and Montage

What reaches the brain is not the number on the stimulator dial but the current density at the cortex, and confusing the two is the commonest error in reading the literature. Current density is the current divided by the electrode area, so the same two milliamps delivered through a small electrode produces a far stronger field than through a large one. Because the polarizing effect scales with the field, electrode size, position, and the arrangement of the two electrodes, the montage, jointly determine both where the current flows and how strong it is (Woods et al., 2016). A montage is never one active electrode: current must return through the other, so the reference electrode also stimulates the tissue beneath it, and a poorly placed reference can produce effects the experimenter did not intend.

Table 1. Standard tDCS parameters and why each governs the delivered dose (Woods et al., 2016).

The third demonstration builds the current-density calculation that any honest dose comparison must start from, letting the reader vary current, electrode area, and duration and read off the density and charge density that actually reach tissue.

Cognitive Applications

The claim that pushed tDCS from a physiological curiosity into a tool of cognitive neuroscience was that it could change cognition, not just muscle twitches. The first influential demonstration was Fregni and colleagues' report that anodal stimulation of the left dorsolateral prefrontal cortex improved performance on a working memory task relative to sham stimulation, the first evidence that the method could enhance a higher cognitive function in healthy people (Fregni et al., 2005). The finding was widely taken up, both because working memory is a bottleneck for so much of cognition and because a portable, cheap enhancer of it would be consequential.

A second landmark extended the reach into numerical cognition. Cohen Kadosh and colleagues trained participants on artificial number symbols while stimulating the parietal cortex and found that stimulation produced specific, long-lasting changes in numerical competence that were still detectable months later, linking tDCS to the parietal substrate of the approximate number system and to numerical learning (Cohen Kadosh et al., 2010). Across these and many subsequent studies, tDCS became a method for testing causal claims about brain function: if stimulating a region changes a task, that region is implicated in the task in a way correlational imaging cannot establish (Filmer et al., 2014). The applications now span attention, learning, decision making, and executive function, though the reliability of the effects varies markedly across domains.

The Efficacy Debate

Enthusiasm outran evidence, and the correction was sharp. Horvath, Forte, and Carter conducted a quantitative review of single-session tDCS in healthy adults and, restricting themselves to outcomes reported by more than one group, found no reliable effect on any of them, a null result that directly challenged the accumulating claims of cognitive enhancement (Horvath et al., 2015). The review was contested on methodological grounds, its narrow inclusion criteria among them, but it did lasting good: it forced the field to confront small samples, flexible analyses, and the wide variability between individuals that a headline effect size conceals.

Part of the resolution came from recognizing that the clean motor-cortex polarity rule does not transfer intact to cognition. A meta-analysis by Jacobson, Koslowsky, and Lavidor found that the anodal-excitatory, cathodal-inhibitory pattern was robust in motor and somatosensory studies but far weaker and less consistent in cognitive ones, so that expecting cathodal stimulation to reliably impair a cognitive task was expecting more than the physiology supports (Jacobson et al., 2012). The current consensus treats single-session cognitive effects in healthy people as real but small and state-dependent, sensitive to baseline performance, task difficulty, and individual anatomy, which is a more defensible position than either the early enthusiasm or the strong null (Polania et al., 2018).

Clinical Use and Safety

Alongside the cognitive work, tDCS developed a clinical literature, most maturely in depression. The evidence-based guidelines assembled by Lefaucheur and colleagues surveyed the therapeutic applications and issued graded recommendations, finding the strongest support for anodal stimulation of the left prefrontal cortex in major depression while judging most other indications to be at the level of possible rather than probable efficacy (Lefaucheur et al., 2017). The most rigorous single trial, ELECT-TDCS, compared tDCS against the antidepressant escitalopram and against placebo in a large randomized design; it found tDCS superior to placebo but inferior to the drug, and it did not meet the pre-specified criterion for non-inferiority to escitalopram, a result that is neither a vindication nor a dismissal but a realistic placement of the method's clinical value (Brunoni et al., 2017).

The safety record, within established parameters, is strong. The evidence-based safety update by Bikson and colleagues reviewed the accumulated human data and concluded that conventional tDCS protocols, delivered at conventional current densities for conventional durations, have not produced serious adverse effects, the most common complaints being transient tingling, itching, and mild skin reactions under the electrodes (Bikson et al., 2016). That reassurance is explicitly conditional: it holds for the parameter space that has been studied, and the same current density that is safe through a large electrode can burn the skin through a small one or a dry sponge, which is exactly why dose is measured at the tissue and why the technical guides insist on it (Woods et al., 2016).

Worked Example

The dose that matters is the current density at the cortex, and the third demonstration reports the same quantities the arithmetic below produces. Take a common protocol: a current of two milliamps delivered through an electrode of thirty-five square centimetres for twenty minutes. Two questions follow, and both are answered by simple division and multiplication.

The first is current density, the current divided by the electrode area. Two milliamps divided by thirty-five square centimetres is 0.057 milliamps per square centimetre, which sits in the middle of the conventional range and is the number a safety limit is actually written against, not the raw two milliamps. Now halve the electrode to seventeen and a half square centimetres at the same two milliamps: the current density doubles to 0.114 milliamps per square centimetre, a stronger and more focal dose from an unchanged dial setting, which is the whole reason electrode area cannot be treated as a free parameter. The second quantity is charge density, the current density multiplied by the duration in seconds. Twenty minutes is 1,200 seconds, so 0.057 milliamps per square centimetre times 1,200 seconds gives about 68.6 millicoulombs per square centimetre. This is the cumulative dose the tissue receives, and it is the figure that governs the margin against the charge-density thresholds established in animal safety work, thresholds that conventional protocols stay far below (Bikson et al., 2016). The lesson is that two protocols reading two milliamps on the dial can deliver very different doses to the brain, and only the density and the charge density reveal which.

Discussion

The settled core of tDCS is small but solid. A weak direct current polarizes cortical neurons, anodal stimulation raises and cathodal stimulation lowers excitability in the motor system, and sustained stimulation produces after-effects that depend on NMDA-receptor-mediated, LTP-like synaptic plasticity (Nitsche & Paulus, 2000; Stagg & Nitsche, 2011). Around that core sits a far more contested periphery: the cognitive effects are real but smaller, less reliable, and more state-dependent than the first wave of enthusiasm implied, and the strong null-result review that provoked the reckoning did the field a service even where its conclusions were too sweeping (Horvath et al., 2015; Jacobson et al., 2012).

The open problems are increasingly well defined. Individual variability is the central one: the same montage produces different current distributions in different heads because skull thickness, gyral folding, and cerebrospinal-fluid geometry all steer the current, so a fixed protocol delivers a variable dose (Polania et al., 2018). The clinical evidence, most developed in depression, places the method as better than placebo but not yet a first-line treatment (Brunoni et al., 2017; Lefaucheur et al., 2017). What is not in doubt, running from Bindman's polarizing currents in the rat to the current generation of individualized, task-paired protocols, is that a current too weak to make a neuron fire can nonetheless change, measurably and lastingly, how that neuron behaves.

Current Directions

The most active current work is aimed squarely at the variability problem. Computational current-flow modelling, built on each participant's own structural scan, is being used to predict where the current will actually concentrate and to design montages that hit an intended target rather than a nominal one, moving the field from fixed electrode placements toward individualized dosing (Polania et al., 2018). High-definition montages using arrays of small electrodes trade some intensity for far greater focality, and the technical guides now treat electrode configuration as a design variable to be optimized rather than a convention to be inherited (Woods et al., 2016). In parallel, the clinical program continues to test tDCS in larger, better-controlled trials and in combination with drugs and behavioural therapy, with depression remaining the leading indication and the ELECT-TDCS design a model for the rigour required (Brunoni et al., 2017). Two themes recur: pairing stimulation with the task it aims to change, so that the LTP-like after-effect acts on a network already engaged, and individualizing the dose so that the current density at the target, not the setting on the dial, is what is held constant across people.

Common Misconceptions

tDCS makes neurons fire.

It does not. The current is subthreshold: it shifts the resting membrane potential toward or away from firing threshold, biasing the likelihood that a neuron responds to its own synaptic inputs, rather than driving action potentials the way transcranial magnetic stimulation can (Stagg & Nitsche, 2011).

Anodal always helps and cathodal always hurts.

The anodal-excitatory, cathodal-inhibitory rule is robust in the motor cortex but unreliable in cognition, where a meta-analysis found the pattern far weaker and less consistent. Effects depend on the region, the task, and the individual (Jacobson et al., 2012).

The dose is the number of milliamps on the stimulator.

The dose that reaches the brain is the current density, the current divided by the electrode area, so identical current settings through different electrodes deliver different doses. Density, not raw current, is the quantity that matters and the one safety limits are written in (Woods et al., 2016).

Glossary

After-effect.

A change in cortical excitability that outlasts the period of stimulation, first reported by Bindman in the rat and dependent on synaptic plasticity rather than on ongoing polarization.

Anodal stimulation.

Stimulation delivered under the anode, the positive electrode, which depolarizes the resting membrane and, in the motor cortex, raises cortical excitability.

Cathodal stimulation.

Stimulation delivered under the cathode, the negative electrode, which hyperpolarizes the resting membrane and, in the motor cortex, lowers cortical excitability.

Charge density.

The current density multiplied by the duration of stimulation, expressing the cumulative charge delivered per unit area; the quantity against which tissue-safety thresholds are set.

Cortical excitability.

The readiness of cortical neurons to respond to input, indexed in the motor system by the size of the motor evoked potential; the primary quantity tDCS modulates.

Current density.

The applied current divided by the electrode area, typically 0.03 to 0.08 milliamps per square centimetre; the dose that actually reaches tissue, as distinct from the raw current.

Direct current.

A steady, constant electrical current flowing in one direction, as opposed to alternating current; the defining feature of tDCS and the source of its polarity-specific effects.

GABA.

The principal inhibitory neurotransmitter, reduced under the anode; its balance with glutamate shapes the plasticity that underlies tDCS after-effects.

Glutamate.

The principal excitatory neurotransmitter, acting at NMDA receptors whose function is required for the lasting, LTP-like after-effects of stimulation.

Long-term potentiation.

A lasting strengthening of synaptic transmission following activity; the after-effects of anodal tDCS resemble it and share its dependence on NMDA-receptor function.

Montage.

The spatial arrangement of the two electrodes, including the position of the reference; it jointly determines where current flows and how strong the field is at the target.

Motor evoked potential.

The muscle response elicited by a fixed transcranial magnetic pulse; its amplitude is the standard readout of motor cortical excitability and the measure that established the polarity rule.

Neuromodulation.

The broad class of techniques that alter neural activity by applied electrical, magnetic, or chemical means; tDCS is a non-invasive electrical member of it.

Polarity rule.

The generalization that anodal stimulation increases and cathodal stimulation decreases cortical excitability; reliable in the motor cortex, weaker in cognitive applications.

Resting membrane potential.

The voltage across a neuron's membrane at rest; tDCS shifts it toward or away from firing threshold, which is the immediate, online mechanism of the technique.

Sham stimulation.

A control condition in which the current is briefly ramped up and off so that participants feel the initial tingling without receiving a sustained dose; the placebo comparison in tDCS trials.

Key Researchers

Marom Bikson. Professor of biomedical engineering at the City College of New York; his work on current-flow modelling and electrode design underpins the technical understanding of tDCS dose, and he led the evidence-based safety update that codified the field's safety parameters. ORCID - Google Scholar - Faculty Page

Andre R. Brunoni. Psychiatrist and neuroscientist now at the University of Texas Southwestern Medical Center, previously at the University of Sao Paulo; lead investigator of the ELECT-TDCS trial that provided the most rigorous randomized comparison of tDCS against an antidepressant in major depression. ORCID - Google Scholar - Faculty Page

Felipe Fregni. Director of the Spaulding Neuromodulation Center and professor at Harvard Medical School; lead author of the first report that anodal prefrontal tDCS enhances working memory in healthy people, the study that opened the cognitive-enhancement literature. ORCID - Google Scholar - Faculty Page

Roi Cohen Kadosh (b. 1976). Head of the school of psychology and professor of cognitive neuroscience at the University of Surrey; his parietal-stimulation studies established that tDCS could produce specific, long-lasting changes in numerical competence, extending the method decisively into higher cognition. ORCID - Google Scholar - Wikipedia

Michael A. Nitsche. Head of the department of psychology and neurosciences at the Leibniz Research Centre for Working Environment and Human Factors in Dortmund; with Walter Paulus he co-authored the foundational study that revived tDCS by demonstrating polarity-specific changes in human motor cortex excitability, and he remains among the field's most cited authorities on its mechanisms. ORCID - Google Scholar - Faculty Page

Walter Paulus (b. 1953). German neurologist and clinical neurophysiologist at the University Medical Center Gottingen and past president of the International Federation of Clinical Neurophysiology; co-author of the Nitsche and Paulus revival studies and a central figure in establishing the dose parameters and safety framework of non-invasive brain stimulation. ORCID - Google Scholar - Wikipedia

Charlotte J. Stagg. Professor of human neurophysiology at the Wellcome Centre for Integrative Neuroimaging, University of Oxford; her spectroscopy and imaging work tied the macroscopic polarity rule to specific changes in glutamate and GABA, articulating the physiological basis of tDCS. ORCID - Google Scholar - Wikipedia

Frequently Asked Questions

What is transcranial direct current stimulation? It is a non-invasive neuromodulation technique that passes a weak, constant electrical current, typically one to two milliamps, between two scalp electrodes to raise or lower the excitability of the underlying cortex (Nitsche & Paulus, 2000).

Does tDCS make neurons fire? No. The current is subthreshold, far too weak to trigger action potentials. It shifts the resting membrane potential toward or away from firing threshold, biasing how likely neurons are to respond to their own synaptic inputs (Stagg & Nitsche, 2011).

What is the difference between anodal and cathodal stimulation? Anodal stimulation, under the positive electrode, depolarizes the membrane and generally raises excitability; cathodal stimulation, under the negative electrode, hyperpolarizes it and lowers excitability. The rule is reliable in the motor cortex and weaker in cognition (Jacobson et al., 2012).

Can tDCS improve memory or intelligence? Some studies report enhanced working memory and numerical learning, but a major review found single-session cognitive effects in healthy people to be unreliable. The current view is that effects are real but small and highly dependent on task and individual (Fregni et al., 2005; Horvath et al., 2015).

Is tDCS used to treat depression? Yes, it is the most developed clinical application. Evidence-based guidelines give it the strongest support for major depression, and the ELECT-TDCS trial found it better than placebo but inferior to the antidepressant escitalopram (Brunoni et al., 2017; Lefaucheur et al., 2017).

Is tDCS safe? Within conventional parameters, the safety record is strong; the evidence-based safety update found no serious adverse effects, with tingling, itching, and mild skin reactions being the common complaints. Safety is conditional on staying within studied current densities and durations (Bikson et al., 2016).

How does the dose of tDCS get measured? By current density, the current divided by the electrode area, rather than by the raw milliamps. The same current through a smaller electrode delivers a stronger, more focal dose, which is why density and charge density are the quantities that matter (Woods et al., 2016).

How long do the effects last? Brief stimulation produces changes only during the current, but a longer session produces after-effects that can outlast it by up to an hour, because sustained stimulation engages synaptic plasticity resembling long-term potentiation (Nitsche & Paulus, 2001).

References