Things you wish you had known before your first tDCS/tACS study

As you may be aware of, scientists enjoy controversies. After all, spirited discussions are often the source of innovation, intellectual advances, and inspiration. Also, let's face it, sometimes they are not. In any case, recent papers have started to seriously question some of the assumptions that build the foundation of many tDCS/tACS studies. Some of the objections are likely appropriate, others I would disagree with (this is material for another post). One thing that is clear to me is that the field has grown so fast and attracted so many researchers with little background in using brain stimulation, or even neuroscience, that I often see studies that are unfortunately tainted by some newbie mistakes. So here is my mission: let's talk about some of these challenges to improve the quality of our studies.

• Blinding: Studies need to be double-blind or triple-blind or whatever you want to call it. In any case, neither participants not researchers that interact with researchers should know the stimulation condition. The placebo effect is an incredibly powerful force you do not want to mess with. Also, you need to make sure that "the blind worked." You need to ask both your research participants and your study personnel what study condition they guess they experienced (and why). This is important because you do feel tDCS/tACS! Carefully chosen "active sham" palcebo stimulation can help. If your device cannot do that, then I see you over at my company Pulvinar Neuro, which sells the device we developed to address this issue.
• Be prepared to fight very hard to get a negative result published. Be honest, pre-define your outcomes (in the US this is now required by federal code, see clinicaltrials.gov), admit to yourself that your study was negative, and do not do some magic statistics to claim some really weird four-way interaction. It will be frustratingly hard because editors fear it will reduce their journal's impact factor, but please do so. Otherwise the "file drawer" effect, which could be also called "biased editor", effect will introduce such a large publication bias that can sink an entire promising field!
• Have the appropriate control conditions. If you do tACS, I strongly recommend a control frequency in addition to the placebo condition such that you can make a statement about a frequency-specific effect, which is at the heart of how we think tACS works. If you do tDCS, make sure to include an anodal and cathodal group.
• Make sure to include some type of target engagement measurement (EEG, fMRI, etc). Any behavioral findings without some evidence of changes in neuronal activity (even if quite indirect, still way better than nothing) is hard to interpret.
• Run your electric field simulations before you do your study to know where your electric current goes. There are now a number of great tools available, both free and commercial. Use at least one of them! BTW, NIMH will not consider tDCS/tACS grants anymore if they do not include such modeling.

Now do not get me wrong, I am as "guilty" as anyone else. Our field has rapidly developed and we all have published papers that do not quite match these standards. However, I encourage you to think about these things, especially if you are new to the field. We will discuss all this and much more in our May 21st tACS/tDCS workshop at UNC as part of the Carolina Neurostimulation Conference. I hope to see you there!

Take care,

Flavio