A study in JAMA Psychiatry has been going around with a headline version: there are three types of ADHD, and brain scans can tell them apart. The study sorted children with ADHD into three groups based on the structure of their brains, and the groups lined up loosely with different symptom patterns. It did not create a new diagnosis or a test, and it was not about treatment.
Didn't ADHD already have three types?
Yes, and that is the source of most of the confusion. The diagnostic manual clinicians use already describes three presentations of ADHD: predominantly inattentive, predominantly hyperactive-impulsive, and combined. Those are defined by symptoms, the things a person and the people around them notice.
The new study's three groups are different. The researchers didn't start from symptoms at all. They started from brain scans, let a statistical method find groups in the brain data, and only then looked at how each group's symptoms compared. Their groups ended up with names that echo the familiar ones, which is why posts about the study often read as if it confirmed the manual's three types. It didn't. It found three brain-based groupings that partly resemble them.
What did the researchers do?
The team, based at West China Hospital in Chengdu with collaborators in the US and Australia, used structural MRI scans from research sites in China and the US. Their first sample was 446 children with ADHD and 708 children without it, aged 6 to 18. They then checked the result in a second, separate sample from the Healthy Brain Network, a large US research database: 554 children with ADHD and 123 without.
For each child, they built a map of how similar different brain regions are to one another in structure, then compared each map with what is typical for that child's age. That comparison step, called normative modeling, lets researchers see how far each child's brain sits from the typical range, region by region, instead of averaging everyone together. A clustering method then grouped the children by their patterns of difference.
What were the three groups?
The authors named them by the symptom pattern that stood out in each:
Severe-combined with emotional dysregulation. The most widespread differences, centered on a circuit linking the medial prefrontal cortex and the pallidum, a deep brain structure involved in movement and motivation.
Predominantly hyperactive/impulsive. Differences centered on a circuit linking the anterior cingulate cortex and the pallidum.
Predominantly inattentive. Differences mainly in the superior frontal gyrus, toward the front and top of the brain.
The symptom differences between the groups were real but small: which group a child fell into accounted for under 2 percent of the variation in their symptom scores. In the second sample, the differences in hyperactivity and impulsivity held up. The differences in inattention did not.
What can't it tell us yet?
The authors are careful about this themselves. They write that their three groups may be points along a spectrum rather than separate kinds of ADHD. They note that the children were not medication-free, that leaving out children with other conditions may make the results less like a real clinic, and that the brains of children with and without ADHD overlapped a great deal. The links the paper draws to brain chemistry came from comparing its maps with separate brain-chemistry data, and the authors say those links cannot inform treatment. ADHD Evidence, a site that reviews ADHD research for the public, reads it the same way.
It is also a study of children. Whether adults sort the same way is unknown.
Why does a study like this matter?
ADHD looks different from one person to the next, and from one day to the next in the same person. Clinicians have known that for a long time. A study like this is an early attempt to find where some of that variety shows up in the brain, which could one day help match support to the person. That day is not here yet, and the authors present the work as groundwork.
Where does Beatwired fit?
Beatwired is a practice tool, not a medical treatment. Each session starts by asking how you are feeling today (overwhelmed, balanced, or under-stimulated) and suggests a pattern for the day you are having.
Sources
Pan N, Long Y, Qin K, et al. Mapping ADHD Heterogeneity and Biotypes by Topological Deviations in Morphometric Similarity Networks. JAMA Psychiatry. 2026;83(5):478-490. Published online February 25, 2026. doi:10.1001/jamapsychiatry.2026.0001 (open access)
ADHD Evidence. Finding order in the complexity of ADHD: a brain imaging study identifies three neurobiological subtypes, March 31, 2026.
Ramsay JR. Let's ask brains what ADHD looks like. Psychology Today, April 2026.