Does Neurofeedback Work for ADHD? What Blinded Studies Show

In the first months after the diagnosis I was open to anything that said "a solution without medication." Neurofeedback was on that list too, and it seemed very reasonable to me: you train your brain, no pills, no side effects, permanent.

Then I looked at the evidence side and saw that it was not so reasonable. This article is the record of that look.

What the procedure does

Neurofeedback is based on measuring the brain's electrical activity with electrodes placed on the scalp and giving the person instant feedback. The measuring tool is the EEG.

The logic is this: certain wave patterns look different in ADHD, the person sees their own activity on a screen through a game or a bar, and over time learns to pull that pattern in the desired direction.

It sounds appealing, and part of it is true: brain activity can be measured, feedback can be given, people can learn to change things. The question that needs asking is a different one: does this change carry over to the symptoms in daily life?

What the meta-analysis found

The meta-analysis by Cortese and colleagues (2016) combined 13 randomized controlled trials on the subject. A total of 520 participants.

The finding has two layers, and the second layer is decisive:

First layer: In assessments made by the people closest to the treatment setting, a significant improvement was seen. That is, the people who knew the procedure was being done reported improvement.

Second layer: In blinded assessments and in studies compared with a sham procedure, no significant effect was found.

The conclusion the study reached is that evidence from well-controlled studies does not support neurofeedback as an effective treatment.

Seeing improvement and improvement happening are not the same thing. Blinded assessment measures the gap between them.

Why blinded assessment matters so much

This is a matter that applies not only to neurofeedback but to all the non-medication approaches in ADHD.

The meta-analysis by Sonuga-Barke and colleagues (2013) showed that in non-medication approaches the effects shrink markedly when blinding is done.

The reason is human and nothing to blame: when you go for weeks to an expensive, lengthy procedure onto which hope has been pinned, the tendency to see change grows stronger. This is so for everyone. In the glossary the general name for it is placebo. The comparison that separates the effect of expectation from the effect of the method is the sham treatment arm.

I have seen on myself how strong this is. In the weeks when I give money and time to a method, I have no difficulty at all finding evidence that the method works. I write a day when I feel good into the method's column, and a day when I feel bad I explain away with "I was tired anyway."

The risk of delaying treatment. The real harm in this area generally does not come from the procedure itself. When starting a method that will take months, postponing treatments with strong evidence is a real cost. Especially in children, school years do not come back.

The "it did so much good for us" stories

You will hear this sentence a lot, and it describes real experiences. The person telling it is not lying.

But in the period during which a session is attended for weeks, other things happen too, and they are all written under the same column:

Time passes. Symptoms fluctuate; anything started in a bad period seems to have worked once things return to the average.

A routine is set. Going somewhere at a fixed time twice a week is a structure on its own. The effect of the structure is written into the method's column.

Attention is given. Someone listens to you regularly, asks about your progress, gives hope. The effect of this is not to be belittled, and it has nothing to do with any electrode.

Other things change as well. In the same period sleep pattern, medication, workload or season may have changed.

So the difference between "it worked" and "the method worked" can only be told by comparative studies. Your own experience is real for you; it is not evidence for anyone else.

What the guidelines say

The NICE guideline (NG87) does not include neurofeedback among the approaches it recommends in the treatment of ADHD.

This is not a ban or a condemnation; it means the evidence is not enough to make a recommendation. In the glossary this is called the evidence-based treatment criterion: for a method to get onto the list, it must be shown to work; it is not enough that it has not been shown not to work.

It also needs to be said that the evidence can change

I want to say one thing clearly here, because it is easy to read this article as "neurofeedback is nonsense," and that would not be right.

Neurofeedback is not a single thing. There are different protocols, different numbers of sessions, different target waves, and they are all put in the same bag. Research continues too. The picture today is this: no effect can be shown in well-controlled studies. Tomorrow better-designed studies may say something else.

So the sentence here is not "never." It is "there is not enough evidence today." Separating these two helps when making your own decision too.

When deciding

If you are still thinking of trying it, there are questions to ask:

• What is the cost, how many sessions? Usually dozens of sessions and a high total cost are being talked about.
• Where else could this money and time go? There are options with far stronger evidence: medication treatment, cognitive behavioral therapy, coaching, adjusting the environment.
• Who delivers it? A person authorized in the health field, and what training have they had?
• What is being promised? Promises like "you won't need medication" or "it will be permanently fixed" are warning signs.
• How will it be measured? Has the criterion for success been set beforehand, or will it be judged only by impression?

Let me expand on the last item, because it is what makes self-deception hardest. Before starting, write down three concrete items: how many appointments you missed, how many deliveries you were late with, how many minutes it takes you to sit down to a task. Measure them at the start, in the eighth week and at the end. Numbers, not impressions.

The same questions apply to other methods

Attention training apps, computer-based cognitive training programs, various devices, headsets that promise "brain waves."

The common criterion is this: is there a study done with blinded assessment and compared with a sham procedure? If not, caution is needed. I covered the cycle of buying apps in the article on app hunting in ADHD, and the same problem with supplements in the article on whether omega-3 helps ADHD.

I compared the options with stronger evidence in the article on medication or therapy in ADHD, and what coaching is and is not in the article on what ADHD coaching is. The decision on diagnosis and medication is a physician's job.

Where I stopped

I did not try it. The reason is not that I think the method is harmful; it is that I chose to put that money and those forty sessions into things whose effects have been measured.

For anyone who wants to try, I think the real question is this too: can this time and money go to an option whose effect has been shown? If the answer is yes, going there first is less risky. If the answer is no, then at the very least it needs to be tried with measurement.

There is also a rule I set for myself, and let me write it here too: when I try a method, I do not change anything else at the same time. If I do both at once, I can never know which one worked, and I usually credit the one I put more effort into. In me this is a bias that works in favor of the expensive one.