Dr. Andrew Hill is a renowned expert in neurofeedback and cognitive enhancement, serving as the founder of Peak Brain Institute. With a background in neuroscience, Dr. Hill has pioneered innovative techniques to treat addiction, anxiety, and sleep disorders. His work emphasizes gentle, iterative brain training approaches over invasive methods. He engages clients globally, offering accessible virtual sessions, and frequently shares insights on brain health through podcasts and live streams.
Episode Summary
This conversation originally aired on Drleeds. You can watch the original conversation. What follows is drawn from my answers in that discussion, written up in my own voice.
What is neurofeedback, and how does it work?
Neurofeedback is a form of biofeedback. We measure the brain in real time, and when the brain briefly shifts in the right direction, we applaud that shift with auditory and visual feedback. It's operant conditioning applied to brain waves. The behavior we're shaping happens to be involuntary, so you're not controlling it on purpose. The brain learns from the feedback the way it learns from any reward signal.
The field is about 57 years old. It was first worked out on cats, and cats are terrible at following instructions. That tells you something important: this is not a participatory mental effort. You can't feel your brain waves and you can't really steer them voluntarily. I've trained people who are non-verbal, people with no language, people with significant cognitive impairment, and the learning still happens. Your job is to pick what you're working on and then notice the after-effect.
Some people remember an older system where you focused to move a balloon or drop a tone in pitch. That's a specific technique called slow cortical potential training, where you do exert some voluntary control. The vast majority of modern neurofeedback runs the other way, off involuntary shifts in the EEG.
What does brain mapping show before you train?
We start with a QEEG, or quantitative EEG. It's an assessment of the resting patterns one brain makes compared to an average population. People are weird, so reading one brain in isolation is hard. The population comparison surfaces what's sticking out so you can understand your resources.
A brain map can surface attention patterns, anxiety flavors, sleep architecture and the consequences of poor sleep, sensory and social irritability, brain fog, and speed of processing. I always pair the QEEG brain map with continuous performance testing, a go/no-go style attention task. Both the brain and the performance get compared to age-matched samples. Out of that, we build a workout plan and go after specific brain waves. People train three or four times a week, about half an hour at a time. You start feeling it in three or four sessions, and each session leaves an after-effect you can grade against your sleep and stress.
If you want a deeper look at how reading your own resting EEG predicts function, I've written about EEG phenotypes separately.
Why does daily drinking change the brain, and can training reverse it?
Drink alcohol several times a week for a few weeks or months and the brain starts manufacturing more glutamate. The GABA released by alcohol is so sedating that the brain pushes back to defend the GABA-to-glutamate balance. Withdraw the alcohol months or years later and you're left shaky, unable to downregulate, unable to soothe internally. Even a few months of chronic drinking produces a chronic overarousal state: anxious, shaky, activated. Craving follows because alcohol drops the activation tone. Sleep onset, sleep quality, and sleep depth all degrade.
Look at these brains and you'll see lots and lots of beta waves, and the beta is hyper-coherent. It's stuck together and won't let go. Deep sleep is hard to reach, and the transitions in and out of sleep are the hardest part.
I treat alcohol abuse as an acquired problem in the brain. It starts with learning driven by anxiety, impulsivity, difficult emotions, maybe trauma, maybe sleep trouble. Over time that learning locks the brain into a dysregulated state, and the brain doesn't back out of hyperarousal on its own.
For these clients I combine two techniques. The first is SMR or low beta training. The second is alpha-theta.
How SMR training rebuilds executive function and sleep
SMR is sensorimotor rhythm. If you've watched a cat in a window stalking birds, body liquid and still, attention laser-locked, you've seen SMR. The same low beta tone helps you sit still, fall asleep, stay asleep, resist seizures, and resist impulses.
Executive function is a big lever in disrupted relationships with substances. It lives largely on the strip of tissue running ear to ear, the precentral gyrus, which handles voluntary behavior. In people struggling with decision-making, whether from chronic alcohol, concussion, ADHD, or chronic sleep loss, you see too much theta (disinhibited, automatic) and not enough controllable low beta. With alcohol, fast-frequency beta dominates instead.
So we place a wire on the scalp at one spot, clip the ears, and set the computer to track the low beta and the theta moment to moment. When theta drops and beta rises, the computer applauds. Move the wrong way and the feedback stops, which the brain finds mildly annoying. The trick is moving the goalposts every few seconds so we only reward movement in the right direction. Two or three sessions in, the brain says "oh, you want some beta," and hands you a surge of it for a few hours. You walk around feeling subtly different, then it fades, then it happens again next session. That iterative push is the whole game. The thalamocortical circuits that generate SMR overlap with the ones that build sleep spindles, which is why this single protocol tends to improve both daytime focus and nighttime sleep. I've written more on SMR neurofeedback if you want the detail.
How alpha-theta training calms craving
On top of SMR, alcohol clients usually get alpha-theta. This brings people to the hypnagogic edge of sleep, the place where creative ideas bubble up while you're half asleep. With eyes closed, we reward the brain waves that surge in that deep, non-linear relaxation state. In that state you're producing growth hormone, T-cells, and relaxing oscillations. After even 20 sessions of alpha-theta, cravings drop away for a lot of people. You can watch resting beta move from three or four standard deviations off the chart back toward typical. Four to six weeks in, people who couldn't relax without major medication can calm down at will and fall asleep at will.
I saw this first at an addiction center in Beverly Hills. One writer had been drinking for 25 years, a bottle of wine on the beach while he worked, then more wine plus lorazepam and zolpidem to sleep, then waking disoriented hours later. He came to us cold after a long detox. About six weeks in I walked into the office and he was asleep on the couch. He'd come in early, on purpose, to prove he could now turn his mind off and fall asleep at will. At roughly 67, it was the first decent sleep in 25 years. The beta had dropped away, the craving had eased, and at the one-year mark he'd stayed sober longer than he had in three decades.
The pattern matters here. Binge drinkers don't show this. Most college students binge drink and most transition out without a problem. Daily drinking for even a few months is what changes the brain aggressively, and I still see that shaky active beta in people who quit drinking 10 or 15 years ago. Healing on time alone is slow. Neurofeedback, plus supports like sauna, cold exposure, and a mu-opioid blocker such as naltrexone to reduce the reward of alcohol, speeds the re-regulation. You can take control over your brain.
Can neurofeedback help with benzodiazepine withdrawal?
Most people on benzodiazepines are physically dependent, and they often get treated as if they're addicted. Pulled off too fast, a person can be left with long, protracted withdrawal. Even a careful, self-paced taper involves real adjustments, because the same GABA-glutamate rebound shows up.
Neurofeedback can help the system re-regulate. The biggest job is managing what flares during withdrawal: pain, anxiety, sleep disruption. Those are tractable. You can train breakthrough pain and breakthrough anxiety directly.
One honest caveat. High doses of benzodiazepines impair learning, which gets in the way of training. Across something like eight to ten thousand people I've worked with, there are only three or four whose brains didn't move. Two were on very large doses of methadone or heroin, and two were on very high doses of benzodiazepines. The benzodiazepine class also makes the brain look anxious. It produces little beta spindles, fast-frequency waves that resemble the friction of anxiety. So the brain is generating anxiety you can't feel while the drug is on board. At a low dose, or off the drug for a while, training works fine. At very high doses you may need clearance time before a session.
What about stimulants, cannabis, and tolerance?
Neurofeedback wipes away acquired tolerance. It's somewhat reliable for benzodiazepines and very reliable for stimulants and cannabis. If you do neurofeedback, the drugs start working two, three, four, five times stronger after a few weeks. That matters because moderation requires low tolerance. With a high tolerance it's hard to moderate; with a low tolerance it's hard not to. You stop reaching for the highest dose.
Stimulant abuse usually starts from impulsivity or trouble managing executive function, on top of the dopamine reward. Walk back the tolerance and you also address the reasons people reach for prescription stimulants in the first place: impulsivity, inattentiveness, slow processing, trouble waking up. Neurofeedback tends to make permanent changes in those over a few months. On a bell curve of executive function, you can usually move a full standard deviation every other month, roughly every 25 sessions. Give someone real control over impulsivity and bring anxiety back to a natural response, and a lot of drug use drops away on its own, because people would rather thrive than crash through the forest. I've covered the same machinery for ADHD and neurofeedback and for anxiety circuits elsewhere.
What does a brain map actually do for you, beyond training?
When I show you your suffering in a brain map, the perseveration, rumination, hypervigilance, busy mind, social irritability, sensory integration, it jumps off the QEEG. Seeing it doesn't solve the problem instantly. What it does is make the problem harder to be overwhelmed by, harder to feel ashamed or guilty about. Understanding how it works in your brain starts giving you traction.
Mapping also shows you how your substances work. Run a clean map and one with caffeine, alcohol, your stimulant, or cannabis on board, and you can see the effect on speed of processing, anxiety features, executive function, and performance. With antidepressants, about 10 to 12 days after starting an SSRI you can see the brain changes that will be felt a month later, so you get an early read on direction. None of this tells you anything you don't already know. You look at your contrast map and say "yeah, that's exactly how I feel with that." It crystalizes your relationship with these compounds so you can make better choices.
Do binaural beats, mind machines, or light glasses actually work?
Binaural beats do not create the brain waves they claim. The human brain doesn't show a frequency-following effect when you play two tones and ask it to track the difference. That effect is demonstrated in some animals and has never been demonstrated in humans, and plenty of research has looked. Binaural beats and monaural beats are a form of meditation, no more. If you enjoy a particular audio track, fine, you're doing an audio meditation, which can help. Don't spend hundreds or thousands on the software expecting it to do what the label says. When someone uses the word "quantum" in a health and wellness context, unless you're sitting inside an MRI, treat it as a red flag.
Old-school audio-visual entrainment, the flashing-glasses mind machines, do very little. Light entrainment may produce a weak driving effect that vanishes the instant you stop, with no after-effect. The one light technique with real promise is different: pulsed red light at specific frequencies delivered to the scalp, which feeds the mitochondrial electron transport chain so it makes more ATP. I've written about photobiomodulation on its own. A few more sacred cows worth retiring: blue-blocking glasses are mostly meaningless, since the light that matters for the blue signal is first thing in the morning. And non-native EMF doesn't do anything detectable to the human brain. It takes an enormous amount of energy to push current through skull, bone, and fat. Even ECT current mostly travels around the skull, not through it.
Is polyvagal theory and heart rate variability training worth it?
Polyvagal theory is a valid framework, and a bit more complicated than any individual needs to internalize. You can think about the vagus nerve as the integrator between gut, heart, and brain in both directions. The practical handle is heart rate variability, the beat-to-beat timing of the heart, which reflects that ascending and descending traffic. What happens in the vagus doesn't stay in the vagus.
HRV biofeedback is legitimate and one of my favorite tools. It's voluntary: clip an ear sensor, breathe so your heartbeat mathematically lines up with your breath, which is called coherence. Training the peripheral nervous system this way trains the brain. It works alone or alongside neurofeedback and other therapy.
On the Safe and Sound Protocol, I'm not deeply familiar with the theory. What I notice is that the practitioners we share clients with tell those clients not to do neurofeedback, on the belief that their method is too powerful to combine with anything. My position is that you should always support your brain with several tools at once.
How does neurofeedback compare to TMS, ketamine, and psychedelics?
TMS uses magnetic stimulation to change the likelihood of firing in the area it targets. The literature supports it for depression, which is about the only thing it's well-proven for, and it works for about two-thirds of people. The effect takes weeks to arrive, the procedure is somewhat painful and expensive, and once depression lifts you get a six-to-nine-month window before it creeps back. I'm a scientist, not a clinician, and my read is that TMS is best for acute, medication-resistant depression to create breathing room while you build other support.
Neurofeedback works more slowly and gently, and you don't feel the process. After three or four months, whatever you've changed becomes your new baseline. If you get a side effect, you might feel wired or tired afterward and it wears off by the next day, so you adjust and continue. The aggressive zapping technologies, TMS, tDCS, tACS, can work, but they don't outperform classic neurofeedback and they're harder to control if side effects show up. I lean toward gentle, iterative, predictable change.
Ketamine and ECT both look to me like turning the computer off and on to force a reboot and a burst of plasticity. ECT done with modern ultra-low currents works very well for medication-resistant depression. With ketamine, the reliable lift seems tied to the dissociative event, which is why the esketamine inhalers produce a weaker effect than full ketamine infusions.
On the plasticity argument behind ketamine and psilocybin microdosing: you're remodeling dendrites all the time anyway. One piano lesson moves every cell in the hand area of cortex. Someone in their 60s or 70s still makes six or seven hundred neurons a day. You don't need more plasticity, you need to shape it. I'd rather you meditate 10 or 20 minutes a morning, go for a walk, do some neurofeedback. The thread through a lot of substance-seeking is an external locus of control, hunting outside yourself for a fix and jumping from drug to drug or medication to medication every two weeks. That switching kindles and traumatizes the central nervous system with overlapping withdrawal and toxic effects. If any of these interventions are used, the set, setting, support, and integration afterward matter more than which drug you found, and none of it should happen outside a clinical setting. For more on how the brain remodels itself, see my piece on neuroplasticity.
How do you get neurofeedback if you don't live near a clinic?
There are only about 15,000 people in the world who do neurofeedback. At Peak Brain Institute we have offices in New York, St. Louis, Los Angeles, Orange County, Stockholm, and London, and about 80% of our clients never set foot in them. We ship a brain mapping system to your home, you put on the cap and fill it with gel, and a coach guides the map live with real-time data acquisition. Coaches then teach you over a couple of weeks to place the wires and set up your own training. Every client gets a private channel and seven-day-a-week communication. In-office clients train three times a week; home clients train four times a week, so they get more sessions in the same window. The hardware and software each cost under a grand now, down from twelve thousand dollars of computers when I started 25 years ago, so we simply lend you a preconfigured kit.
A practical note on the EEG itself: a thick head of dark hair gives the cleanest signal. A bald scalp adds noise because the pulse moves the skin and the fat layer insulates. Very thick hairstyles or natural dreadlocks mean we place individual wires rather than a cap, which I do regularly. The map is the awkward part. Training is just a couple of ear clips and one or two wires, so you can do homework or your taxes while you train and check in afterward. We were about half remote in 2019, jumped to roughly 75% in 2020, and stayed there. If you're curious about doing this from your own home, I've described the remote neurofeedback process in full.
Get into your brain, learn how it works, and make small progressive changes. You can find me at peakbraininstitute.com, and I run a weekly livestream on YouTube at Hill Drill where I do neurofeedback and answer biohacking questions live.