I run a weekly livestream where I train my own brain on camera and then take questions from whoever shows up. This week the topic was brain fog. What follows is the teaching from that session, reorganized for reading. I have kept the questions but stripped the names, so anything that started as an audience question appears here as a general point.
If you have a brain, I can help you understand it. Brain mapping and neurofeedback are my two main tools, and brain fog responds to both more reliably than almost anything else I treat.
What is brain fog, and what does it look like in the brain?
Brain fog has many causes: apnea, concussion, chronic stress wrecking your sleep, mold, Lyme, COVID, a metabolic hit like a near-drowning. At a high level, the EEG signature looks roughly the same regardless of cause.
The most useful marker is alpha speed. Your peak alpha frequency, the dominant oscillation in the 8 to 12 Hz band, normally stays synchronized within a hemisphere. Under fatigue and metabolic load, that peak slows and spreads out. It stops holding together cleanly across a hemisphere. That spreading produces the lived experience of fog: delayed recall, tip-of-the-tongue hunting, forgetting what you heard a minute ago. People read it as a memory problem. It is closer to a speed problem. Information loads in and out of working memory too slowly, words stay slippery and short-lived, and your brain struggles to hand them off between regions. For the mechanics of why this band does this, I wrote up the idle-and-brakes role of alpha in Decoding Alpha Waves.
Delta tells you where someone sits in the arc of an injury. Right after an insult, a new concussion, a fresh viral hit, mold exposure, delta runs elevated and fast. That is cleanup on aisle five, metabolism pushing into the foreground while the body tries to heal. A year out from a concussion, delta in those same tissues can drop into low power. When delta runs slow, below average peak delta speed, the quality of deep rest is simply absent.
There is also a coherence picture. In alpha you often see a front-to-back waterfall of blue lines, hypercoherent and locked in phase, when deep sleep is not happening and the fog is bleeding into waking hours. My working read is that this reflects anterior and posterior cingulate hypercoupling: the orienting and selecting circuits refusing to relax relative to each other, so they compensate for fatigue by constantly discerning and selecting. Treat that as a model, not a fact. Delta under fog tends to run hypercoherent and stuck in phase as well, unless the person is deep into healing.
One point on gamma, since people ask about frontal gamma hypocoherence. On standard equipment you are almost never measuring real gamma. Amplitude drops as frequency rises, the one-over-f relationship, so 10 microvolts of delta is one wave while 10 microvolts of gamma is fifty or sixty waves. Every tissue layer between cortex and scalp acts as a high-pass filter and washes the gamma out. Any fatigue drops the whole EEG amplitude and pushes gamma below the noise floor of your filters. The blue gamma you see in tired, concussed, or foggy people is overall signal amplitude dropping and masking real gamma activity, full stop.
Does the cause of brain fog change the treatment?
Mostly no. The fog itself is the part you work on. I have been mapping brains for over a decade and running neurofeedback for about 25 years, and I have before-and-after maps on the same people across COVID, injuries, mold, Lyme, and re-exposures. The same signature shows up again and again no matter the trigger. New infections and new injuries tend to latch onto old inflammatory areas, which is why I watch a healed concussion flare again with a fresh hit: the new injury reactivates the neurovascular inflammatory response in old scar tissue.
The exception is when something is actively keeping the fire lit. An active mold exposure, Lyme with co-infections, a recent post-viral fatigue that is still knocking someone over. There you proceed carefully, because these people overtrain fast. You also see more dysautonomia and POTS in younger people now, and more vestibular complaints and migraines riding along with that chronic inflammatory picture. When the cause is still active and serious, stabilize the deepest layer first, go gentle, and lean harder on the metabolic and vascular tools than on aggressive EEG work.
Which neurofeedback protocols clear brain fog?
The sensorimotor strip is the workhorse. I can move up and down the central strip and train large portions of the brain without ever leaving it, because the strip connects down through the thalamus and out to wide areas of cortex. That makes it both powerful and predictable.
C3 on the left, beta. The left precentral gyrus handles stabilizing executive function, clear calm focus, and staying alert when things are boring so you do not wander off. Left-side beta manages your vigilance tone. Its job is to sustain vigilance and to sustain the absence of vigilance, which is why C3 beta makes you more awake when awake and deeper asleep when asleep. I trained 14.5 to 17.5 Hz this session while inhibiting 4 to 7 Hz theta and a fast band above. If you push the frequency too fast or train too long, you overactivate: sounds too loud, lights too bright, a wired edge. The cleanest readout is sleep. Too fast a C3 and you struggle to fall asleep but sleep great once you do. Too slow and you get sleep maintenance problems.
C4 on the right, SMR. Right central SMR at roughly 11.5 to 14.5 Hz targets supervisory attention, the circuit that monitors whether attention is on task. SMR only behaves as SMR on the sensorimotor strip. The same frequency elsewhere is just regular beta or fast alpha. It calms and stabilizes the body and quiets the wired, can't-shut-it-off feeling. More detail on that band is in SMR Neurofeedback, including why training waking SMR strengthens the same thalamocortical circuits that generate sleep spindles, so daytime focus and nighttime sleep both improve.
CZ at the vertex runs SMR a little slower and targets sleep onset specifically, where C3 is more about staying asleep. If someone is so wired from compensatory strain that they cannot relax, the vertex is where I go.
Brain fog often produces a strained-but-activated state. Deeply depleted, yet the brain pushes up against the fatigue and you feel wired and tired at once, like butter scraped over too much toast. You match the protocol to that presentation. Distractible from fatigue points you right-side. Can't-fall-asleep activation points you to the vertex. I dial in the frequency and site individually first, judging by the subjective after-effect, and only then combine sites into a contingent two-channel protocol, which requires both locations to come online together and produces a different resource than either site alone. My standard combination runs C3, C4, and CZ together, tuned per person.
A practical wiring note for people running their own systems. The ears are active signal sources. Subtracting C4 from the same-side ear, A2, drops out shared signal and gives you a smaller, more spatially specific reading. Subtracting C4 from the opposite ear, A1, leaves a larger signal with more whole-brain contribution. My rough rule, mostly clinical lore: left-ear reference for left sites, linked or both ears for midline, right-ear reference for right sites.
The more frontal you train, the more delayed, individual, and less predictable the effect. There is no thermostat for mood, even though the left-approach and right-avoid asymmetry is real. There are thermostats for sustained attention, impulsivity, and sleep. Fix the legs of the stool at the central strip and the frontal complaints, including mood, often resolve without ever training the frontal lobe. You will even see the frontal map change without having trained there. That is the safer route unless someone is in crisis.
When should you avoid alpha and theta training for fog?
Rewarding alpha or theta usually makes brain fog worse. Avoid it when you are foggy.
There is one exception. With a post-viral phenomenon or an active infection, alpha can help, because eyes-closed PZ alpha training raises CD4-positive T-cells in immunocompromised people. I have used eyes-closed PZ alpha as an immune-support protocol for years and find it accelerates recovery from immune-mediated hits. So even foggy, you may want alpha if the goal is clearing the immune response driving the fog.
For migraine specifically, the tool is passive infrared hemoencephalography. You wear a forehead sensor that reads heat coming off the brain and learn to improve vascular toning through biofeedback on those dynamics. Migraines are a neurovascular event, so this works well, often within a couple of weeks, and I only have people run it five to ten minutes. I run it with a scaled live continuous signal rather than thresholded reward, treating it as skill-transfer training rather than operant conditioning.
What biohacks clear brain fog outside neurofeedback?
Start with the boring foundations. A lot of headache is muscle tension, neck tension, hydration, or electrolytes. A lot of fog is gut-driven: a low-grade food sensitivity, poor motility or absorption, bloating, and the metabolic cost of all that.
Beyond the basics, the tools I reach for:
- Hormetic stressors. Saunas, ice baths, and hyperbaric medicine if you have access to big tools.
- Nutritional ketosis. Drop carbs, raise protein, or use exogenous ketones.
- Photobiomodulation. Red light therapy, which I covered in Brain Biohacking with Photobiomodulation.
- Nootropics. Choline sources like citicoline to push speed of processing, racetams for short-term memory, and newer omega-3 forms such as lyso-phosphatidylcholine-bound DHA.
When illness or injury is active, lean harder into the metabolic and vascular set, add the infrared blood-flow headset and red light, and go slow. The goal early on is modest: a sleep change, a little pain or fog relief. Then gently layer in more. You cannot get away with overtraining a post-viral or dysautonomia case the way you can with an anxious or ADHD brain, where you have more time and a less reactive system. For the wider framework on stacking cognitive tools, see Biohacking Intelligence and the fog-specific writeup at Biohacking Brain Fog. If you want the map before you train, start with QEEG Brain Mapping.
A note on personality and neurofeedback
People ask whether neurofeedback changes personality, and whether Big Five traits show up in the EEG. The EEG I work with is changeable. EEG phenotypes are stable across months and years, not decades, because they reflect regulatory resources the brain is constantly tuning: attention, stress, sleep, sensory and social processing, speed, and stability against fog, seizure, and migraine. Personality sits at a level both above and below that. After neurofeedback I see large changes in executive function, anxiety, speed of processing, and even creativity, while personality stays roughly the same.
The interesting edge case is trauma. Alpha-theta work can look like a personality shift when it removes the constraints that a trauma response, learned helplessness, or acquired alexithymia placed on someone. The same with severe ADHD: give an impulsive, anxious person a brain that can actually follow through on the choice to listen and be present, and they may look far more conscientious. That is someone regulating toward their own center. In 25 years I have never watched a quirky, gifted person lose the quirk while gaining control over their anxiety or attention.
Where to start
Work the central strip: left-side beta at C3, SMR at C4, and SMR at the vertex, combined into a sleep-and-focus protocol you tune by the subjective after-effect. Avoid rewarding slow waves unless you are managing an active infection. Add saunas, ice, red light, and targeted nootropics, or come at it purely through the body with deep nutritional ketosis. Move slowly, watch for signs of overtraining, and track how each change lands. Fog has many causes and almost always shifts when you go after it methodically, but the rate at which you can push depends on how inflamed and reactive the system is underneath.