Dr. Leeds and Dr. Andrew Hill on Neurofeedback

I had a conversation on the Drleeds show about how neurofeedback retrains the brain for sleep, anxiety, and addiction. Watch the original conversation. What follows is drawn from that discussion, in my own words, with the mechanisms spelled out.

What does neurofeedback actually do?

Neurofeedback is operant conditioning for your brainwaves. We put sensors on your scalp, read the electrical activity your cortex is already producing, and feed it back to you in real time through sound or visuals. When your brain produces more of a target pattern and less of a problem pattern, the system rewards you. Your brain notices the reward and starts producing the rewarded state more often. You are not consciously steering this. The learning happens below awareness, the same way you learned to balance on a bike without being able to explain the physics.

The approach I use is gentle and iterative. We do not push the brain hard or force a state. We nudge, watch what happens, and adjust over many short sessions. That matters because the brain learns better through repetition and small wins than through a single aggressive intervention. If you want the deeper picture of how this trains the brain's capacity to change, I cover that in Biohacking Plasticity.

How does brain mapping guide the training?

Before training, I map the brain. A QEEG brain map records the EEG across the scalp and compares your patterns to a normative database. That tells me where the activity sits outside the typical range, which frequency bands are over- or under-expressed, and where the training should aim. Two people with the same complaint, say insomnia, can have very different maps. One might show excess high-frequency beta over the frontal cortex. Another might show slow alpha or a sluggish thalamocortical rhythm. The map keeps me from guessing. I cover the phenotype approach in Biohacking with EEG Phenotypes.

Why does a busy brain refuse to sleep?

A man came into our office in his late sixties. He had not had a good night's sleep in roughly 25 years. When I looked at his brain, the story was on the map: a flood of high-frequency beta, the fast EEG activity that goes with cortical arousal, vigilance, and a mind that will not stop talking to itself. That beta is the signature of a cortex stuck in an on state. At night, when the system is supposed to downshift, his frontal cortex kept idling at high RPM. He would lie in bed with his thoughts running.

This is the pattern behind a large share of the chronic insomnia I see in the brain maps. The cortex is over-aroused, the brakes are not engaging, and the person experiences it as a mind that will not turn off. The racing thoughts are not imagined. The arousal is measurable on the EEG, and elevated high-frequency EEG power at sleep onset is a documented feature of primary insomnia (Perlis et al., 2001). If anxiety drives your version of this, I go deeper into the circuitry in Biohacking Anxiety and Neurofeedback for Anxiety.

What happened when we trained his brain?

We worked with him over a few weeks of short, frequent sessions. The goal was to bring that excess beta down and strengthen the patterns that let the cortex disengage. About six weeks in, I walked into the office and he was asleep on the couch. He had called that morning to find out when I was coming in, arrived half an hour early, and put himself to sleep in the waiting area. He wanted to prove to me that he could now turn his mind off and fall asleep at will. After 25 years, he could do it on demand.

Around the month-and-a-half mark, the excess beta had dropped away. He could quiet his mind, he could fall asleep, and his alcohol craving had eased.

Why did the alcohol craving fade too?

The man had been using alcohol the way many people with chronic over-arousal do, as a sedative. When your cortex runs hot and you cannot downshift on your own, alcohol becomes the tool that finally quiets the noise and gets you to sleep. It works in the short term and it builds dependence over years.

When the training brought his arousal down, the brain no longer needed the chemical brake. The craving was tracking the underlying dysregulation. Lower the beta, restore the brain's own capacity to settle, and the pull toward the drink loosens. This is a pattern I have seen across many brain maps, not a randomized trial result, so I hold it at that level of confidence. The mechanism is consistent: addiction often sits on top of a brain that cannot self-regulate arousal, and giving the brain back that capacity removes part of the reason the substance was there. If you are working on a habit loop like this, I lay out the broader strategy in Biohacking Bad Habits.

What circuits make sleep training work?

The frequency band I lean on most for sleep is SMR, the sensorimotor rhythm, around 12 to 15 Hz over the sensorimotor cortex. SMR training strengthens the thalamocortical circuits that also generate sleep spindles, the 12-to-14 Hz bursts that stabilize sleep and gate out interruptions. When you train a brain to produce robust SMR during waking calm, you reinforce the same machinery that keeps sleep intact at night. Research on SMR training has reported improved sleep onset and increased sleep spindle density (Hoedlmoser et al., 2008). That is why SMR training often relates to daytime focus and nighttime sleep at once, both states depend on the thalamus and cortex talking to each other in a stable rhythm. I go through this in detail in SMR Neurofeedback and Biohacking Sleep.

For other presentations I train different targets. Alpha, the 8-to-12 Hz rhythm, acts as the cortex's idle and its braking system. In older adults, the peak alpha frequency slows with age, and training toward a healthy individual alpha frequency has shown cognitive benefit in studies like Angelakis and colleagues (2007). Alpha and alpha-theta protocols also have a track record with anxiety. The protocol follows the map. I explain the alpha mechanism in Decoding Alpha Waves.

Does this require coming into a clinic?

It does not. I coach clients all over the world through remote sessions. We ship equipment, set you up, and run the same protocols you would get in person, monitored in real time. The brain does not care where the chair is. For how that works in practice, see Remote Neurofeedback, and for what it costs and what insurance covers, see How Much Does Neurofeedback Cost.

What you can take from this

If your version of insomnia is a mind that will not stop running at night, that is often an over-aroused cortex producing excess high-frequency beta, and the research suggests it is trainable. The first step is a measurement. A QEEG map tells you whether your sleep problem is high beta, slow alpha, or something else, and that determines the protocol. Sleep, anxiety, and substance cravings frequently share the same root in arousal regulation, which is why settling one can relate to settling the others. A 67-year-old who had not slept well in 25 years napped on my couch to prove the point.