What if ADHD is three different brain types, not one?
If you've been told you have ADHD, you got handed one of three DSM labels: primarily inattentive, primarily hyperactive-impulsive, or combined. Those buckets describe behavior. They were never built to describe the brain, and anyone who works in QEEG brain mapping or neurofeedback already knows the labels don't fit individuals cleanly. When you look at the quantitative EEG, you see a spread of individual phenotypes that point toward DSM categories without landing in them precisely.
A paper out of JAMA Psychiatry (Pan, Long, Quinn, and colleagues, February 2025) gives this some structural weight. The authors used morphometric similarity network analysis, a method that maps how brain regions co-organize structurally rather than just measuring activity at single locations. Think of it as the imaging cousin of what we do in QEEG, where amplitude and connectivity together tell a more useful story than amplitude alone. They found three biotypes. Across 25,000 brain maps I've read, these three crystallize patterns I've watched for years.
This matters because of where it was published. A pattern discussed at neurofeedback conferences stays in the back room. A large-scale network analysis in JAMA Psychiatry changes how psychiatrists and parents think.
Biotype one: the inattentive, frontal-slow pattern
The first biotype is the inattentive presentation, the brain I'd describe as left-dominant. The study characterized it as general frontal slowing, a structural shift toward slower activity. On the QEEG you see frontal theta, frontal alpha, and elevated theta/beta ratios, or in Monastra's framing, a high theta-to-alpha ratio. The mechanism is underactivation. The engine won't start, won't stay in gear, won't follow through. The old 1987 label "ADD" fits it well.
Frontal theta and frontal alpha are not one thing, and separating them is clinically important. At least two generators produce excess slow activity. One is a midline pinch of the anterior cingulate, which handles error monitoring, conflict resolution, and selecting among competing thoughts and behaviors. The other sits over the precentral areas and governs activation and the supervision of behavior. The inattentive subtype carries a complicated frontal signature, not simply a sleepy frontal lobe. It often carries executive fatigue, energy-regulation problems, and anxiety features in the same map.
This group also tends to show weak SMR and weak low-beta tone across the sensorimotor strip, which is why sleep maintenance suffers alongside focus. The same left-side beta that keeps you awake and in gear during the day helps stabilize the depth of sleep at night.
This biotype is the classic stimulant responder. You can build activation tone with stimulants, train beta up over the left frontal sites, reduce alpha excess, and use focused attention practice to anchor sustained attention. For the mechanics of how alpha throttles cortical engagement, see my breakdown of alpha waves and cortical idling.
Biotype two: the hyperactive-impulsive, disinhibited pattern
The second biotype is primarily hyperactive-impulsive, a more right-dominant phenomenon. On the QEEG you see hyperarousal, excess beta, sometimes a paradoxical mix of fast and slow activity, and heavy involvement of the anterior cingulate, often with faster alphas. The study tied this to the right hemisphere yoked into the cingulate and pallidal regions.
The mechanism is failed inhibition. The right inferior frontal gyrus signals down through the basal ganglia to pump the brakes on behavior. Without good control over that circuit, you can't stop the thing: can't stop saying it, moving, fidgeting. This is the disinhibited person who can't sit still.
Sleep takes a hit here through the same circuit. SMR when you're asleep goes by another name, sigma, or sleep spindles. Weak inhibitory tone over the sensorimotor cortex means fewer spindles, which means trouble sustaining deep sleep when the environment is noisy. Failure of executive SMR while awake produces fidgeting; failure of SMR while asleep produces fragile sleep.
SMR training is your primary move with this pattern, building inhibitory tone over the right sensorimotor cortex. If this is your biotype and you want to meditate, a moving or active practice will serve you better than sitting on a cushion anchoring to nothing, because this brain has the hardest time with stillness. In my phenotype model I call these regions the supervisor and the problem child.
Biotype three: the emotional dysregulation pattern
The third biotype is the one that changes the picture. The combined ADHD label in the DSM implies inattentiveness and impulsivity bolted together. This biotype is structurally distinct, centered on the medial prefrontal cortex and pallidal networks, with emotional regulation fundamentally altered.
For decades, emotional dysregulation, anxiety, low motivation, and procrastination were treated as comorbidities riding alongside ADHD, not part of it. Shaw and colleagues (2014) argued that emotional dysregulation should be considered a core ADHD feature, implicating the dorsolateral PFC and orbitofrontal cortex interacting with the amygdala and ventral striatum. That's reward and threat circuitry, the machinery that tags memories with danger. This new study adds real weight to that argument.
The lived version: rejection feels like getting punched. Frustration escalates fast. Small disappointments trigger disproportionate responses. If you've been told you're too sensitive or too much, and you also carry an ADHD label, you may be looking at this phenotype. It offers a reasonable structural account of what people call rejection sensitivity dysphoria.
This is the group I see medicated poorly or overmedicated. Push stimulants into a brain with a hot frontal midline or low-power delta from poor sleep, and you can drive anxiety, tics, intrusive thoughts, and worse sleep, which spirals. About 20 to 30 percent of people don't respond well to stimulants, and a mismatch between treatment and biotype is a significant reason why. I also suspect a fraction of people labeled bipolar II are this deeply dysregulated ADHD flavor with strong executive, mood, and anxiety components layered together. That's a clinical observation from my own practice, not an established finding.
Treatment here gets individualized. From a brain-mapping view, I look hard at the cingulates and the pre- and postcentral gyri on both sides, the "plus sign" of tissue, to find the biggest pinches. The cingulates often respond to NAC. Methylation matters too: MTHFR, COMT, and related variants can leave these tissues cramped, and a targeted B-vitamin strategy built around your genetics can loosen them. The data on methylation and overlapping executive, anxiety, and sensory features is suggestive rather than settled, but I think there's something real there.
If social processing is also off, you may see a flat affect or missed sarcasm without full autism, mapping to the right tempoparietal junction behind the right ear. For more on that integration circuit, see my piece on sensory and social processing.
Why matching treatment to biotype changes outcomes
If ADHD is three things wearing one label, treating it as one thing manufactures non-responders. Match the intervention to the brain:
- Frontal-slow inattentive: stimulants, left-frontal beta uptraining, alpha reduction, focused-attention practice.
- Hyperactive-impulsive: SMR uptraining for inhibitory tone, active meditation rather than seated stillness.
- Emotional dysregulation: target the cingulates and the approach/avoid circuitry, consider NAC and a genetics-informed methylation strategy, and avoid stacking more stimulant onto an already hot midline.
This is why I map before intervening. I can't tell you what matters most to you, but I can tell you what's most unusual in your brain compared to average, and starting from the most atypical features gets you a long way. For executive symptoms, I typically see about a standard deviation of change every couple of months, with the brain map, the performance test, and lived experience moving together. If you want to know whether neurofeedback is a fit for ADHD specifically, I've written a full neuroscientist's guide to neurofeedback for ADHD.
How do you tell ADHD apart from autism on the EEG?
These categories overlap inside individuals, so the buckets aren't clean. ADHD and autism can both show excess alpha, theta and delta, and difficulties with sensory, social, and executive processing. A few things help distinguish them. Run a continuous performance test (a go/no-go task) alongside the QEEG; the pattern of impulsivity and inattention looks different behaviorally. I tend to see more primary midline-central involvement in ADHD and more cramping behind the right ear in autism, sometimes with tic and anxiety phenomena down the midline.
Thirty years ago you couldn't be diagnosed with both. Now we recognize a continuum, and the internet's "AuDHD" label captures that blend even though it isn't in the DSM. I treat it roughly as higher-functioning autism mixed with ADHD. Bring the same person to an ADHD therapist and they get an ADHD diagnosis; bring them to an autism therapist and they get an autism diagnosis. The more useful question is how the individual's resources actually work, because once you know how your brain is built you can train it.
What about auditory processing that looks like ADHD?
Alpha blobs over the auditory tissue, around the posterior and right temporal regions, can produce auditory processing lag that gets misdiagnosed as inattentive ADHD. I see this regularly, at least one person in twenty of my new intakes. The performance test looks clean except for a couple of very particular auditory measures, and the brain map looks typical except for those localized alpha blobs. High alpha lower down in the temporal areas can also show up with vertigo, tinnitus, or auditory processing difficulties. Excess beta at the right tempoparietal junction is my marker for an intense, hard-to-filter incoming signal, what I call the princess and the pea.
Can neurofeedback help narcolepsy and daytime sleepiness?
I've worked with truly diagnosed narcolepsy maybe eight to ten times, and sleep regulation improved in most of those people, though not in the older clients. I suspect a larger group carries excessive daytime sleepiness that gets the narcolepsy label and a prescription for modafinil. The mechanism that helps is timing: the left precentral beta tone keeps you awake when you're awake and stabilizes sleep depth when you're asleep. Train beta up on the left while someone's awake and you reduce micro-sleeping; their nighttime sleep then consolidates as well.
True narcolepsy involves the hypothalamus, orexin neurons, and peptide regulation, so I'd also investigate circadian cues, morning routine, and evening fasting before assuming the EEG is the whole story. If sleep is your bottleneck, I've laid out the broader approach in biohacking sleep.
Making the labels work for you
This more refined view of ADHD only becomes possible once you already have a framework for what you're looking at. Operating in the functional space of networks, where you can model the circuit, measure it, and watch it change, holds together better than the behavioral buckets in the DSM and is more workable day to day than purely molecular or genetic accounts. Grab the diagnostic labels and the brain-map models as far as they fit you and as far as they hand you a strategy. A label you just sit with does nothing.
If you want to see which of these patterns is driving your own attention, sleep, and emotional regulation, get a QEEG brain map and look at the data with someone who can read it.