Recent research has identified biomarkers for subtypes of ADHD. The mental health field has celebrated this finding, while failing to grasp that adoption of these new categories requires a complete rejection of the current symptom-based system of classification.

Psychiatrists have long sought biological markers for mental health conditions — both to legitimize the conditions that we treat and to more precisely match effective treatments to the underlying causes. Yet years of research in genetics, brain imaging, and neurochemistry have failed, until recently, to uncover valid psychiatric biomarkers.

For many mental health conditions, researchers have found variables which on average differ between those with a condition and those without. People with chronic depression have a smaller hippocampus than those who weren’t depressed. Genome wide analysis finds hundreds of alleles associated with a tiny increased risk of developing schizophrenia. Abnormal activity of dopamine transporters can be measured in the prefrontal cortex and striatum of those with ADHD. But because the range of values varies so widely both among those who have these conditions and in the general population, measurements overlap between the two groups, meaning that none of these potential biomarkers distinguishes which group any individual belongs to.

In early 2026, Nanyang Pan and fellow researchers at Sichuan University revealed biomarkers that describe three distinct, non-overlapping varieties of ADHD, finally giving psychiatry a laboratory measurement that objectively differentiates specific mental health conditions.

Pan et al.’s research looked at patterns in the connectivity between hundreds of different areas in the brain, in hundreds of children. They found four distinct, non-overlapping patterns of neural connectivity, which they called biotypes. One biotype corresponded to the children in their study without ADHD. Three distinct patterns of uncommon wiring were found in the brains of children with ADHD. In both their original, multisite sample, and in a completely independent replication, each of the three ADHD biotypes made up roughly one-third of the individuals with ADHD.

The researchers labeled the three ADHD brain wiring patterns biotypes1, 2, and 3. Symptoms of emotional volatility, along with severe symptoms of inattention and hyperactivity, were found in biotype1 individuals. Individuals in biotype2 scored highly on hyperactive symptom scales of ADHD, and biotype3 individuals had very high inattentive scores.

The popular press, ADHD groups, and mental health organizations widely touted the findings in self-congratulatory tones, claiming that the research affirmed that these new ADHD biotypes “mirrored” the subtypes already identified by our current symptom-based diagnostic system.

On the surface, the new biotypes look like a pretty good match for the current DSM division of ADHD into combined, hyperactive, and inattentive subtypes. A summary from the American Psychiatric Association daily electronic newsletter even noted that the hyperactive and inattentive categories “map neatly” onto the way ADHD is currently diagnosed.

Nothing could be further from the truth.

The three biotypes of ADHD identified by Pan’s team don’t align at all with conventional subtypes of ADHD. For one thing, individuals in their biotype2 didn’t just have high hyperactivity and impulsivity, they also displayed substantial symptoms of inattention. Those in biotype3 didn’t just have high inattention, they also displayed substantial symptoms of hyperactivity and impulsivity.

Pan provided a breakdown of which biotypes were found among each of the symptom-based, DSM ADHD subtypes:

For inattentive subtype ADHD, 25% were biotype1, 45% biotype2, 30% biotype3

For combined subtype ADHD, 40% were biotype1, 40% biotype2, 20% biotype3

For hyperactive subtype ADHD, 33% were biotype2 and 66% were biotype3

The only correspondence between ADHD biotypes and DSM subtypes exists in the unfortunate, occasional description, in Pan’s article, of biotype2 as hyperactive/impulsive and biotype 3 as inattentive. Pan’s ADHD biotypes don’t “mirror” the DSM subtypes — not even with funhouse mirrors. An individual with DSM inattentive ADHD is actually most likely to have “hyperactive” biotype2, but also has a decent likelihood of having “severe” biotype 1 or “inattentive” biotype3.

If these ADHD biotypes pan out, and are replicated not just for subtypes of ADHD, but for ADHD itself, and for other psychiatric conditions, it will revolutionize mental health diagnosis and treatment. It’s also worth pointing out that Pan’s approach, using fMRI and sophisticated computer analysis to assess patterns of functional connectivity between hundreds of brain areas, doesn’t provide an easy office-based diagnosis and is currently relegated to research settings.

Progress in developing more effective mental health treatments has been stymied because our disorders don’t truly reflect underlying distinct biological processes. What we call depression is probably a collection of different biologic conditions. Patients arrive at the final endpoint of symptomatic depression after traveling very different trajectories. The same is likely true of anxiety, schizophrenia, substance abuse, ADHD, and our other psychiatric conditions.

It is disappointing that many mental health experts appear to have missed that Pan’s ADHD biotypes completely refute our current symptom-based categories. Rejecting the old system will necessitate suspending belief in decades of clinical observations and research findings about mental health conditions. If our new categories don’t align with the old DSM criteria, that indicates that many individuals really didn’t fit into the little diagnostic boxes that we had thoughtfully, neatly, and carefully placed them in.

The changes also have the potential to render all previous mental health research uninterpretable and hence obsolete. At best, it will require a lot of thought and cognitive contortion to salvage useful knowledge gleaned from using diagnostic labels that are misaligned with the new biomarker-driven categories.

Other medical fields have successfully undergone revisions of diagnostic categories. Oncologists now identify cancers based on their genetic heritage, rather than on their organ of origin.

In the long run, psychiatrists are well-positioned to adopt the new biomarker paradigm. After all, mental health workers have experience helping people to see their world in new ways, and to escape from restrictive, inaccurate, and harmful conceptions of themselves, their world, and their relationships. Psychiatry’s journey to more precise and accurate diagnoses and treatment needs to start with healing ourselves and learning to look in new ways.