Non-invasive Brain Stimulation (NIBS) techniques are used to modulate brain activity in a safe and well-tolerated way . In particular, Transcranial direct current stimulation (tDCS), uses low-intensity electrical currents to modulate cortical excitability in a polarity-specific manner . Classical tDCS montages consist of two rectangular sponge electrodes with a contact area of ∼25-35 cm2, where electrical current between 0.5mA and 4mA flows from a positively charged electrode (anode) to a negative one (cathode) passing through various tissue compartments including skin, muscle, bone, cerebrospinal fluid and brain. Due to its safety and relatively low-cost, tDCS experiments have been widely carried out to investigate human neurophysiology and to test its application as a new potential therapeutic solution for neurological and psychiatric conditions. To ensure adequate understanding of the observed effects, however, researchers need to rely on valid and approved control placebo conditions, a fundamental requirement in randomized controlled trials. Traditional standard sham protocols consist on an initial ramp up of the current, followed by a short stimulation period (usually for 5-60 seconds) and a final ramp down [, , ], (i.e., Fade In of current, brief real Stimulation, Fade-Out; commonly known as “FISSFO” protocol), an approach thought to cause sensory stimulation similar to real tDCS without affecting cortico-spinal excitability . However, both these assumptions (i.e., adequate blinding and absence of effects on the brain) are still under examination. FISSFO sham has been considered effective in providing a proper blinding when compared with real tDCS at 1mA for 20 minutes , becoming the standard for sham tDCS . The rationale stems from participants’ reports in which the cutaneous perceptions that generally cue subjects on tDCS being effectively delivered (i.e., tingling or itching sensation), have been mostly reported during the first 30-60 seconds of stimulation to then gradually decrease, possibly due to habituation . However, a recent investigation has revealed that even naïve subjects (N=192) are capable of distinguishing classic sham stimulation (FISSFO) from active tDCS when delivered at 1 mA for 20 minutes over the left dorsolateral prefrontal cortex (DLPFC) . Prior experiments had already suggested blinding inefficacy when real tDCS is applied at 1.5-2 mA, even for only 10 minutes [9,10]. Accordingly, non-naïve subjects seem more capable of distinguishing real from sham tDCS  and extreme individual variability has been reported with regard to sensibility to stimulation intensity and duration, with subjects being able to perceive tDCS even at very low intensity (i.e., 400 μA) .
On the other hand, additional sham protocols have been developed with modified durations of ramp up/down, or even delivering constant low intensity currents (0.016 or 0.034 mA) [7,12,13]. However, these approaches have not been properly tested on large sample of patients/subjects, with no data on the effects of such alternative sham protocols on the brain, while inconsistent results on many neurophysiological parameters have been documented when adopting such modified approaches .
Beyond the single or double blinding efficacy of FISSFO and related approaches , an element of interest is the question of whether tDCS effects are due to cortical interaction of the generated electric fields or from peripheral nervous system (PNS) stimulation. Since the ramp-up/ramp-down method for blinding decreases both cortical and peripheral stimulation, they cannot help disentangling cortical and peripheral effects. In addition, cortical effects of the brief period of real stimulation during sham protocols may not completely be excluded .
An additional challenge is the fact that the induced tDCS electric field is conditioned by the heterogeneity of cortical and non-cortical tissues, as well as by the complexity of cortical geometry . In recent years, this has been addressed by the use of multichannel tDCS systems in combination with realistic finite element modeling of current propagation in the head derived from subject neuroimaging data (e.g. MRI, fMRI) [17,18]. The rationale for multifocal stimulation resides on both the need for more targeted stimulation of the cortex, as well as the notion that brain regions operate in networks and communicate with each other’s through modulatory interactions [, , ]. Realistic physical models provide a crucial element for better experimental understanding and control of the electric fields generated by tDCS.
In the present study, we investigate a novel approach to sham stimulation based on controlled shunting of currents via a model-based quantification of transcutaneous and transcranial effects. Specifically, the novel sham tDCS solution benefits from the use of an optimization algorithm allowing tDCS montages to be tailored in such a way that zero or very low magnitude electric fields are delivered on the brain, while medium to high intensity currents are maintained in at least some scalp electrodes, thus eliciting scalp sensations necessary for blinding. Notably, this allows to maintain the stimulation ON for the entire duration of sham tDCS, therefore inducing scalp sensations similar to real tDCS, while avoiding known limitations of the FISSFO protocol. We hypothesize that such montage (Active Sham, ActiSham hereafter) (i) will generate scalp sensations similar to a Multifocal (real) tDCS montage based on the same electrodes’ location and identical stimulation intensity/duration; and that (ii) ActiSham will not induce changes in cortico-spinal excitability (CSE), as assessed through Motor Evoked Potentials (MEPs) induced by Transcranial Magnetic Stimulation (TMS) as an index of corticospinal excitability. If successful, this and similar other approaches for improved sham stimulation could contribute to more efficient design of future tDCS research studies and clinical trials.
Fourteen subjects participated in 4 randomized tDCS stimulation visits, spaced 7±3 days to ensure no carryover effects. The tDCS conditions were: real Bifocal-tDCS, Bifocal-Sham, real Multifocal-tDCS and ActiSham. Each session lasted approximately 90 minutes during which participants seated in a comfortable chair with their eyes open. To measure changes in corticospinal excitability, single pulse TMS was applied over the left primary motor cortex (M1) at the beginning and the end of each stimulation session. Somatosensory sensations elicited by tDCS were addressed by means of ad-hoc questionnaires. See dedicated sections below for further details about tools and procedures.
Fourteen healthy right-handed naïve subjects (25.4 years ± 2.1; 5 males) were recruited at the University Campus of Siena, School of Medicine (Siena, Italy). Possible contraindications to either TMS or tDCS were assessed by means of a screening questionnaire . Exclusion criteria included: history of seizures, head injury, pacemakers or other implanted medical devices, metallic objects in the head, hearing impairments, medications altering cortical excitability or other significant medical concerns. All participants gave written informed consent prior to participating to the study. The research proposal and associated methodologies were approved by the local ethical committee in accordance with the principles of the Declaration of Helsinki.
tDCS sessions lasted 15 minutes, with electrode types, scalp montages and stimulation intensities customized for each tDCS protocol (Figure 1). Transcranial stimulation was delivered using a “Starstim 8” brain stimulator controlled via Bluetooth using a laptop computer (Neuroelectrics, Barcelona, Spain). For canonical Bifocal-tDCS (active or sham), stimulation was delivered through traditional 5×7 cm rectangular sponge electrodes, with a contact area of 35 cm2 (SPONSTIM, Neuroelectrics, Barcelona, Spain). Before current delivery, electrodes were soaked with 15 ml of sterile sodium chloride solution (0.9%). For Multichannel stimulation conditions (real and ActiSham), current was instead delivered using circular Ø 20 mm PISTIM electrodes (Neuroelectrics, Barcelona, Spain) with an Ag/AgCl core and a gel/skin contact area of 3.14 cm2. Electrodes were filled with a conductive gel before the tDCS intervention. To further improve current conductivity, the scalp was gently rubbed with an alcohol solution at the beginning of each session. Electrodes were inserted in a neoprene cap with available positions following the 10/20 EEG system.