Dose dependency of transcranial direct current stimulation: implications for neuroplasticity induction in health and disease

There has been a rapid growth in the use of non‐invasive brain stimulation to induce neuroplasticity and modulate human brain function over the past decade. An example of this is transcranial direct current stimulation (tDCS), which involves the application of weak electrical currents between at least two electrodes placed over the scalp. When applied to the primary motor cortex (M1), tDCS can induce lasting polarity‐dependent changes in excitability, with facilitation following anodal stimulation and suppression following cathodal stimulation. These changes share similarities with the long‐term potentiation (LTP) and long‐term depression (LTD) observed in animal models, and importantly, there is some evidence that tDCS can have positive therapeutic effects for patients with various neurological and psychiatric conditions.

Of the various techniques developed for non‐invasively inducing neuroplasticity in the human brain, the advance of tDCS in particular has been met with significant enthusiasm, due in no small part to its relative low cost, compact design and ease of use. However, as is becoming increasingly evident for many neuroplasticity‐inducing non‐invasive brain stimulation protocols, its after‐effects can be extremely variable between individuals, not just in magnitude and duration, but in direction also. This inter‐individual variability of tDCS after‐effects is problematic for its translation to clinical practice, and thus, determining the causes of this variability is of critical importance.

Many factors have been identified that could contribute to tDCS response variability (Ridding & Ziemann, 2010); however, surprisingly little attention has been paid to the optimal dose of stimulation. While increasing the duration and/or intensity of stimulation might be assumed to maximise the therapeutic effects of tDCS, there is now growing evidence in healthy human populations suggesting a dose–response relationship for tDCS that is not linear, but instead is more complex than initially thought (Batsikadze et al. 2013; Kidgell et al. 2013; Monte‐Silva et al. 2013).

In a recent article published in The Journal of Physiology, Jamil et al. (2017) sought to further examine the dose dependency of tDCS by systematically measuring the neuroplastic response to stimulation across the full DC intensity range. Healthy participants received either anodal or cathodal tDCS applied for 15 min to left M1 at five different intensities: sham, 0.5, 1.0, 1.5 and 2.0 mA. Single‐pulse transcranial magnetic stimulation (TMS)‐elicited motor‐evoked potentials (MEPs) were recorded before and at multiple time points up to 2 h following stimulation to quantify M1 plasticity. The results showed no evidence for a linear relationship between increasing stimulation intensity and the induced neuroplastic response for either anodal or cathodal tDCS. Across the early time points (0–30 min) post‐stimulation, anodal tDCS induced MEP facilitation for all active intensities relative to sham, with no difference between stimulation intensities. While a difference between stimulation intensities was observed for cathodal tDCS, this was due to a reduced MEP suppression at higher intensities (1.5 and 2.0 mA), with active stimulation at 1.0 mA producing the optimal response.

The non‐linear relationship between tDCS intensity and the induced plasticity response has been shown in previous studies. For instance, Kidgell et al. (2013) compared the effects of anodal tDCS applied for 10 min at 0.8, 1.0 and 1.2 mA, and found no difference in MEP facilitation between stimulation intensities. In another study, Batsikadze et al. (2013) compared the M1 plasticity response to cathodal tDCS applied for 20 min at either 1.0 or 2.0 mA. While 1.0 mA stimulation induced the expected suppression in MEP amplitudes, cathodal stimulation at 2.0 mA reversed the polarity of after‐effects from suppression to facilitation.

Further evidence for the non‐linear dose–response relationship for tDCS has been shown in studies comparing different stimulation durations. As well as examining different spacing intervals for repeated applications of tDCS, Monte‐Silva et al. (2013) compared a single application of 1.0 mA anodal tDCS, applied for either 13 or 26 min. Rather than increasing the facilitatory response, doubling the duration of stimulation reversed the polarity of effects from MEP facilitation to suppression. The MEP suppression induced by 26 min of anodal tDCS was abolished by flunarizine, suggesting that the polarity reversal following prolonged stimulation was probably calcium channel dependent. Given the known role of calcium in determining LTP/LTD induction, it is possible that similar calcium channel‐dependent mechanisms may be responsible for the differential response to cathodal tDCS at different intensities, although this remains to be tested.

In addition to investigating the influence of different stimulation intensities on tDCS response, Jamil et al. (2017) also examined different sources of inter‐individual variability in their data. Consistent with previous research (Labruna et al. 2016), they found that the facilitatory response to 1.0 mA anodal tDCS was greatest in those participants with higher sensitivity to TMS (i.e. requiring a lower TMS intensity to elicit an MEP with 1 mV peak‐to‐peak amplitude). This relationship is probably the result of inter‐individual differences in anatomical (e.g. skull thickness, cortical morphology, etc.) and physiological (e.g. neurotransmitter availability and receptor distribution) factors that similarly affect the efficacy of both TMS and tDCS (Labruna et al. 2016). Interestingly, Jamil et al. (2017) found there was a tendency for participants with lower sensitivity to TMS (that is, requiring higher TMS intensities to elicit a 1 mV MEP response) to respond with a greater facilitatory response to anodal tDCS at higher intensities of 1.5 and 2.0 mA, although this did not reach statistical significance. Nevertheless, this may lead to intriguing possibilities for using the stimulus–response characteristics of TMS as a way of identifying the optimal stimulation intensity and individualising tDCS dose.

While the work of Jamil et al. (2017) has provided some useful new information regarding tDCS dose–response relationships, there remains much to be investigated. For instance, while the effects of varying stimulation intensity and duration have each been studied, it is still unclear how these variables might interact to affect tDCS response. Further complicating matters is the possible influence of electrode size and montage, as well as different repetition intervals for multiple stimulation sessions (Monte‐Silva et al. 2013). Although the results of Jamil et al. (2017) and others suggest that intensities of 0.5–2.0 mA for anodal and 1.0 mA for cathodal may be sufficient to produce the desired MEP response to tDCS at the group level, it is clear that such generalised approaches will not work for every participant. On this basis, further studies aimed at individualising tDCS parameters based on individual anatomy or physiology are required.

Finally, while the current research has been largely focused on tDCS effects on M1 in healthy young adults, whether the dose–response relationship differs for non‐motor cortical targets and in patient populations remains an open question. While investigating the neurophysiological effects of tDCS outside M1 presents its own challenges, continual progress is being made in the combination of TMS and tDCS with neuroimaging modalities such as functional magnetic resonance imaging, positron emission tomography, and electroencephalography, providing several new and exciting insights. The application of these novel techniques to examine the dose dependency of tDCS will be essential for advancing treatment options for a variety of neurological and psychiatric disorders, in particular those characterised by pathophysiological changes in non‐motor regions, such as depression, schizophrenia and dementia.

The findings of Jamil et al. (2017) highlight the complexity of the tDCS dose–response relationship, and may lead to novel solutions for individualising stimulation and reducing inter‐individual response variability. While the obstacles associated with characterising the optimal dose for all possible stimulation parameters and cortical targets are not trivial, this line of research remains necessary for tDCS to reach its full therapeutic potential.

Additional information

Competing interests

None declared.

Funding

M.R.G. is supported by an NHMRC‐ARC Dementia Research Development Fellowship (National Health and Medical Research Council of Australia (NHMRC) and Australian Research Council (ARC); 1102272). B.H. is supported by a research fellowship from the NHMRC (1125054)