Episode Summary
This article comes from my conversation on "Neurofeedback Works, But Here's the Catch." You can watch the original conversation. What follows are my own observations from twenty-five years and 25,000-plus brain maps of clinical work.
Does neurofeedback actually change the brain?
Yes. I have watched it happen thousands of times. People come in carrying real suffering: anxiety, trauma, autism, disregulated sensory processing, problem drinking. Over a few months of training, many of them change, and the change is often large.
When you see this often enough, you stop treating each case as a miracle. The brain is suffering, the brain is transforming, you observe both with the same steady attention. That detachment is the healthy version of an experienced practitioner. It comes from watching the same arc play out across hundreds of brain maps and learning to trust the process without getting swept up in it.
If you want the research basis for that clinical confidence, I have laid it out separately in Is Neurofeedback Legitimate? A Research Overview.
What does a good neurofeedback practitioner actually do?
Good practitioners build a deep working model of the brain. They iterate. They assess before they train, they re-assess as training progresses, and they adjust the plan based on what the EEG is doing. I would argue that experienced neurofeedback clinicians understand how the brain operates in real time as well as anyone on the planet, because they are watching it regulate, fail to regulate, and learn to regulate again, session after session.
The work resembles personal training more than it resembles taking a pill. A good trainer looks at your starting point, picks the right exercises, watches your form, and changes the program when something is not working. A neurofeedback clinician does the same thing with cortical rhythms. The assessment defines the target. The target defines the protocol.
This is why I start clients with a QEEG brain map. The map tells me where the dysregulation lives, which frequency bands are over- or under-expressed, and which sites to train.
Why does one-size-fits-all neurofeedback cause problems?
A small slice of the field, maybe 10 to 20 percent, runs automatic, one-size-fits-all systems. These devices auto-adjust, skip the assessment entirely, and the people selling them tend to make three claims: neurofeedback works, it cannot cause harm, and you can do as much of it as you want. All three claims are false.
Here is the mechanism. Neurofeedback is operant conditioning of brain activity. You reward the brain for producing more of one rhythm and less of another, and the brain learns to shift its baseline in that direction. If you push a frequency band the wrong way for a given brain, you train the brain to regulate in a direction it did not need. That produces side effects.
Train down too much fast-wave activity in someone who already runs slow, and you can deepen brain fog or flatten arousal. Push SMR or beta in someone who is already over-aroused and wired, and you can worsen anxiety or sleep. Run alpha-theta deep-state work in someone who is not stable enough to handle it, and you can stir up material they were not ready for. The brain you build depends entirely on the direction you reward. Build it in the wrong direction and you get a worse-regulated brain, not a better one.
Can neurofeedback cause side effects?
Yes, and pretending otherwise is the most dangerous thing in this field. The claim that neurofeedback "cannot cause harm" is what lets people run unlimited sessions on automatic systems with no map and no plan. The same plasticity that makes training work makes it possible to train in the wrong direction. The reward signal does not know whether the direction is good for you. That judgment comes from the assessment and the clinician, not the machine.
If you want the underlying biology, I have written about how the brain's adaptive plasticity cuts both ways. The system that lets you reshape a rhythm is the same system that will happily reshape it the wrong way.
How much neurofeedback is the right amount?
There is no fixed answer, and "as much as you want" is the wrong one. The right dose depends on the protocol, the site, and the individual brain. Some people need more sessions of a gentle protocol. Some need a small number of sessions of something more targeted, then a reassessment. The plan should change as the brain changes.
This is the core of individualized training. Different people need different protocols at different sites, and the protocol that helps one anxious brain can aggravate another. The map and the ongoing measurement tell you which is which.
What should you ask before starting neurofeedback?
Ask whether the practitioner does a real assessment before training. A QEEG or equivalent measurement should define your protocol, not a default setting on a device.
Ask whether they reassess. Training is iterative. If nobody is measuring change, nobody is steering.
Ask what they would do if you developed a side effect. A clinician who has trained enough brains knows that side effects happen and knows how to back them out. Anyone who tells you side effects are impossible has told you they are not watching for them.
Neurofeedback is a heavily individualized form of personal training for your brain. It produces real change, and it can go wrong the same way personal training goes wrong when you do the wrong exercise for your body. The protective factor is the same in both cases: an experienced practitioner who assesses, picks the right target, watches your response, and adjusts. Start with the map, and let the measurements guide every step after that.