Episode Summary
I joined the team at NeuroNoodle for a holiday Q&A about neurofeedback, the rush toward ketamine for depression, and why so many ADHD diagnoses turn out to be something else entirely. You can watch the original conversation for the full discussion. What follows is drawn from my own answers in that session, cleaned up and organized so you can use it.
Why am I cautious about ketamine and other dissociative drugs?
Psychiatrists are excited about ketamine for a simple reason: it produces a rapid lift in depression, and almost nothing else does that. Standard antidepressants take weeks. Ketamine moves the needle in hours. When you have very few tools that work fast on depression, a fast tool gets a lot of attention.
I am the conservative voice on this. The thing I like about doing neurofeedback well is that you watch the live EEG and see whether the brain is following along in a good direction, a defensive direction, or whether you have it wrong. You get feedback in the moment. With most medications, you take a hit and you do not know where you are going until the cycle runs through, and for many drugs the compound stays in the brain and body for a long time. Some effects are permanent.
Before anyone goes near a dissociative, I want to look at the EEG. Two specific findings raise the risk of a bad ketamine experience, and they are the same features that predict medication failure: epileptiform content and beta spindles.
What does the EEG show before someone uses a dissociative?
Beta spindles greater than 20 microvolts indicate hyper-excitable cortex. These are not a little fast activity riding on top of the record. They form an actual spectral peak, often in the low 20s and sometimes above 30 Hz. Beta spindles were first described in the early 1930s in Lennox's lab by Gibbs and Gibbs as a form of epilepsy, because the spindles were enormous. The tonic phase of a tonic-clonic seizure shows up as beta spindles so large and organized they read as poly-spikes.
A beta spindle focus sitting in the cortex is an Achilles heel waiting to fire. Add the stimulating side effect of a dissociative and you can have enough to trigger it, not in a good way. My grandmother's rule applies: do not dive under the water unless you know what is under the surface. Look at the EEG, weigh cost against benefit, and decide whether you are walking into an enhanced risk.
A practical note from the panel as well: if someone has a history of alcohol problems, a colleague at Yale advises against dissociatives, since the dissociative effect can resemble alcohol intoxication. We already know a lot about alcohol volume and tolerance, including the mu and kappa opioid receptor variants that make alcohol roughly 25% more reinforcing for some people, and we still fail to apply that knowledge in standard care. I would rather measure first. You can wear a breathalyzer to learn what a given blood alcohol actually feels like in your body. That kind of measurement pierces the mystery and gives you agency. I want the same posture with ketamine.
For more on the circuits behind cravings and reactive behavior, see biohacking bad habits and biohacking anxiety.
What does depression actually look like in the brain?
Depression is a lump category in the DSM. When I worked at McLean Hospital doing MRI imaging on people trying to withdraw from antidepressants, we were tracing structures like the globus pallidus looking for a change over a month or two. The psychiatric truism is that if you try to come off an antidepressant and relapse twice, you should probably stop trying. We were hunting for the structural correlate. We did not find a clean one.
Real findings do show up in depression. Hippocampal volume changes. Frontal asymmetry. Cortical thinning. The cortisol state strips tissue and reduces BDNF in the hippocampus, which thins the temporal lobe. The markers are there if you know where to put the spotlight, though the causal direction remains unsettled.
When you look for thalamocortical dysrhythmia, you can also find it pointing at the anterior cingulate. That same spot shows up for reward-deficiency conditions, addiction, OCD, and for direct physiologic problems like tinnitus after hearing loss, phantom pain, and movement disorders. It is affect regulation and cognitive regulation, not a distinct marker for one diagnosis. The EEG earns its value prognostically. It tells you where you are, where you need to go, and how to get there. Diagnosis is a different question.
How do you talk to a psychiatrist who is against neurofeedback?
If a friend on antidepressants wants their psychiatrist to consider neurofeedback, the move is simple: have the psychiatrist get a brain map of themselves. Psychiatrists love data and think mechanistically. Once they see their own QEEG brain map, they start referring clients.
About half of psychiatrists I encounter hold an anti-neurofeedback view, and the other half ask what it is. Either way you can create converts by educating. Fifteen years ago that number was effectively zero, and the reaction was combative rather than skeptical. The long arc of neuroscience is bending toward the biological. The APA's biological-psychiatry and biological-psychology sections are the parts that have grown.
When a client gets a dismissive reaction from a provider, I do not waste energy fighting it. I tell the client they are the boss. They are hiring all of us for our knowledge, and they get to choose. Most clients who find neurofeedback are extremely savvy, because they have had to be. They have tried things, educated themselves, refined their understanding of what is happening, and often arrive with a diagnostic picture that differs from what they were first told.
The same principle works with schools. When a teacher or counselor pressures a parent to medicate, I suggest the parent thank them for flagging the concern, say they are getting assistance through their own doctor, and then let the next report card speak for itself. Agency is the way in. We help you understand your own phenomena and take control, without picking a fight with insurance companies or the half of psychiatry that resists. For the insurance question specifically, see is neurofeedback covered by insurance.
Why is ADHD so often a misdiagnosis?
ADHD is largely a misnomer. It functions as an umbrella for "we do not know why this kid, or adult, is not doing what we want in this moment." People with this pattern are paying attention to more things than you think they should be. They have executive control of attention difficulties, not an intentional problem.
And it is not low IQ. More often the opposite. IQ breaks down into processing speed, implicit learning, and working memory. Most people with the ADHD pattern have disinhibition plus high processing speed, which gives them a leg up on at least a third of the regulation equation.
What gets labeled ADHD is frequently something else. I see adults and kids with the diagnosis who perform executive-function testing and brain mapping perfectly fine in their frontal regions, but their alpha speed is dragged down, their delta is elevated, and they do not sleep at night. They can sit for 20 minutes and perform beautifully on a continuous performance test while the map shows fatigue and stress features everywhere. Fatigue and stress mimic ADHD constantly. Trauma response and generalized anxiety do the same, especially in kids, where they show up as inattention or as a brittle, reactive state rather than a constant one.
How much of "ADHD" is actually a sleep disorder?
A solid paper in the literature estimates that 25 to 50% of the ADD/ADHD population has an underlying sleep disorder. It breaks into a few types.
The first is a primary disorder of vigilance. The child sits down and cannot stay awake. They look inattentive because they drop into stage one as soon as they sit, and in 10 minutes of eyes-closed EEG you watch them fall precipitously into stage two. This is essentially narcolepsy without cataplexy, an orexin problem that modulates wakefulness drive. These kids respond well to modafinil. Speeding up their alpha or chasing frontal theta misses the actual mechanism entirely.
A second group has a circadian rhythm delay. On a state-by-state basis, the CDC measures about 10% less ADD/ADHD in the desert Southwest than on the East Coast, where the sun is not as bright. That circadian-delay group is treatable by bright sun resetting their rhythm. Living in enough morning light handles a meaningful slice of them. For the home version of this, see biohacking your morning and biohacking sleep.
Then there are nighttime breathing problems that produce daytime sleepiness, and occasionally epileptiform content during sleep. Disordered sleep yields poor attentional skill across the board.
The value of brain mapping is that when a child arrives with the label, we can see which region, which efficiency or inefficiency, is producing the behavior. For the SMR work that ties daytime focus and nighttime sleep together through the same thalamocortical circuits, see SMR neurofeedback. For the broader framework, see does neurofeedback work for ADHD.
Why are so many women getting new ADHD diagnoses in midlife?
A large and growing segment of my client base over the past decade is middle-aged women receiving a brand-new ADHD diagnosis. The mechanism is sleep. As sleep dysregulates with menopause, a latent ADHD pattern breaks through and gets in the way. These women go to the doctor and usually get an ADHD diagnosis rather than a sleep diagnosis.
Women a generation or two back were not assessed at all. They were written off as the "ditsy female" and given no academic support. When I show one of these women the part of her brain producing the pattern, the relief is visible immediately. She has spent decades believing she was not smart, and the map says otherwise.
What does neurofeedback do for autism?
Autism is disproportionately represented in the neurofeedback field, probably at least a quarter of the client base, because it produces visible change where little else does. I see movement in sensory processing, seizure activity, social engagement, obsessive patterns, sleep, and executive function. Language change is less reliable, though I have seen that too.
Language has a developmental window. The left-right hemisphere laterality finishes around ages 8 to 9 in girls and 9 to 10 in boys. After that the brain prunes away phonemes it has not heard, which is why you lose the ability to hear and produce new speech sounds and why accents form. With the little ones, before that window closes, we can do real work on speech. Because we rarely see them as older kids, parents often never learn what negative trajectory was avoided.
That window is also a superpower you can give any child. Make it a reward to watch their favorite cartoon dubbed in Farsi, Arabic, Mandarin, and Spanish, with the English subtitles on. A couple of hours here and there keeps those phonemes intact, so if they decide to learn the language later they can hear and speak it like a native. That is left-right hemisphere integration built through play.
What is the elevated alpha at T6 actually telling you?
T6 elevated alpha is a phenotypic pattern that cuts across the DSM, so it is not specific to any one diagnosis. The right parietal area handles spatial perception, and T6 specifically picks up face detection and the reading of emotional body language. You expect to see that spot off in PTSD, sometimes in anxiety, in Asperger's and autism, in sensory integration difficulties, and in reactive attachment. The common thread is emotional perceptual skill. When that foundation is off, it limits or distorts what is built on top of it, the way Wernicke's area underwrites speech comprehension.
The goal is to normalize the spectrum at that site, not to chase alpha alone. Alpha is an index frequency for the brain and tends to push back when you train it in a linear fashion. The frequency of alpha varies widely between people. Neurology does not assume 8 to 12 Hz is alpha; it looks for the rhythm at the back of the head that attenuates when you open your eyes, and that can sit anywhere from the low single digits up into the mid-teens. There are also overlapping functional alphas, idling alpha and preparatory alpha, that look similar but run slower or faster. You read the whole spectrum, find which other bands are cramped or dysregulated, and map what the pattern is doing for that person. For more on this band specifically, see decoding alpha waves and the framework in biohacking with EEG phenotypes.
What does slow EEG content after a head injury mean?
When someone comes in with a lot of slow content after a head injury, that points to a more severe injury involving white matter, not just gray matter. Bob Thatcher's work comparing quantitative MRI and quantitative EEG in mild TBI showed that delta in the EEG corresponded with white matter changes on MRI, while alpha and beta changes corresponded with gray matter changes. Crush injuries produce delta. Shear injuries produce beta hotspots, because you lose the inhibitory interneurons and the cortex runs unchecked.
I see a striking amount of bilateral temporal delta and theta lately, and not all of it is direct impact. A lot of it tracks with the neck. When someone reports eighth cranial nerve symptoms like tinnitus or vestibular issues, I increasingly refer them for upper cervical analysis. I do not recommend chiropractic bone-cracking for this. I recommend gentle upper-cervical traction practitioners who do sophisticated x-rays and hold tension to create slow shifts. I am seeing a surprising rise in cervical instability, POTS, and dysautonomia presentations, often with a neck component.
Is QEEG reliable for diagnosing head injury?
The QEEG is not validated as a diagnostic instrument for TBI. For that specific question, the evoked response potential is the right tool.
Post-concussion looks the same as TBI on a map. Apnea looks like TBI. Mold, Lyme, chemotherapy, and sleep deprivation from trauma all look like brain fog and all trip the same TBI visibility on the standard discriminants. When I reach the data page that says "injury" on it, I tell people not to believe the label. The software is using an injury population as an index for brain fog.
Two more cautions. To run a TBI discriminant validly, the EEG first has to be read as normal by a qualified professional. If the software operator does not have that credential, the discriminant is functioning as a screener, and we found a 50% false positive rate when it is used that way. The software author also reports a 20% false negative rate, meaning one in five people who did sustain a head injury will be told they did not. I do not trust those labels. I watch how the metric shifts across time, which is genuinely informative as fatigue or jet lag resolves. The severity index at the bottom is weighted heavily toward slow content, delta and theta, because slow content tracks how much white matter is involved.
Can neurofeedback reduce inflammation?
Inflammatory markers do drop during training, and I have seen inflammation reduce. Working with a functional or medical doctor to get inflammation under control before neurofeedback is your main tool is the right sequence. Photobiomodulation also works for some inflammatory processes, with the same caveat about sequencing.
Training a swollen brain carries real risks. Inflammation produces swelling, and the brain has nowhere to expand inside the skull. It compresses tissue and the blood vessels inside it, reducing flow. That compression produces ischemia, which shows up as the slow edge of alpha, locally if the problem is local and diffusely if it is systemic, as in conditions like lupus.
On the laser-versus-LED question for photobiomodulation, no good side-by-side study exists. "My frequency is better than your frequency" gets claimed by a lot of devices. If you make the claim, provide the data.
What about the thyroid showing up in the EEG?
Band training does not affect the thyroid, though activity below 1 Hz might touch it. The more clinically useful direction is the reverse: thyroid problems influence the EEG. Hashimoto's thyroiditis is an autoimmune condition where a peroxidase antibody binds T3 and T4 out of bioavailability. They still register on a standard T3/T4 blood test, but they are not available at the cellular level. Hashimoto's produces gigantic paroxysmal slow bursts in the EEG, a dramatic abnormality that is frequently mistaken for something else. If you suspect a T3/T4 problem, see an endocrinologist and ask for the peroxidase antibody.
The bottom line
The throughline across all of this is the same posture: measure before you act. A label like ADHD or TBI often hides a sleep disorder, a trauma response, a thyroid problem, or a neck injury. Brain mapping shows you where the inefficiency lives and how to train it, and the right next step is to get a map and read the whole spectrum rather than chase a single band or a single drug. If you want to go deeper on a specific pattern, start with the QEEG brain mapping guide and the phenotype framework, then match the training to what your own brain is actually doing.