Three Brain Biotypes Cut Across ADHD and Anxiety Diagnoses in Children, Stanford-Led Study Finds

The children in this study were 9 or 10 years old when they lay inside an MRI scanner, and what researchers eventually found in their scans does not line up neatly with the diagnoses printed in their charts. A team led from Stanford Medicine reported on Oct. 16 in Nature Mental Health that children with attention-deficit/hyperactivity disorder, anxiety disorder, or both fall into three distinct brain biotypes, each marked by malfunction in a specific set of brain circuits.

The problem the researchers set out to address is a familiar one in child psychiatry. ADHD and anxiety are among the most common mental health conditions of childhood, and they often overlap. Both are diagnosed with symptom checklists, and a child with one condition frequently shows several symptoms of the other. Yu Zhang, PhD, an assistant professor of psychiatry and behavioral sciences at Stanford and co-senior author, said that overlap hinders understanding of the neurobiology behind the traditional diagnoses. His team wanted to know whether biological differences among these children follow the diagnostic labels used in the clinic. The answer, he said, was not exactly.

The data came from the Adolescent Brain Cognitive Development study, a long-term project on brain development that enrolled more than 11,000 U.S. children at age 9 or 10. From it, the researchers drew 3,508 children who already had a diagnosis of ADHD, anxiety, or both when they enrolled; 693 who were undiagnosed at enrollment but later received one or both diagnoses; and 1,947 controls who never received either. All had fMRI scans and completed cognitive and behavioral tests. To check their results, the team used scans from another 224 children in a second project, the Healthy Brain Network, all of whom had a diagnosis of ADHD or anxiety.

The team built an AI-powered model to analyze functional connectivity, meaning how different brain regions work together. The model drew on diagnoses and behavioral assessments to identify which connectivity patterns were more prominent in children with ADHD or anxiety than in controls. It found two brain-behavior dimensions, cognitive-behavioral dysfunction and emotion-attention dysfunction, each driven by different brain networks. Combined in different ways, they produce three biotypes: externalizing-disinhibited, internalizing-distressed, and mixed.

Each child belonged to only one biotype, but every biotype appeared within every traditional diagnostic category. Still, tendencies emerged. Children in the externalizing-disinhibited group showed more hostility, defiance, impulsivity, aggression, and rule-breaking, and were more likely to have an ADHD diagnosis. Those in the internalizing-distressed group showed more guilt, tension, perfectionism, and anxious-depressed symptoms, and more often had anxiety diagnoses. Children in the mixed group had more attention problems, irritability, and self-injury, and were more likely to have ADHD diagnoses, though all pairings of biotype and diagnosis occurred.

The finding that surprised Zhang most was that the biotypes could be seen in baseline scans of children who had no diagnosis at enrollment but were later diagnosed with ADHD, anxiety, or both. He said this suggests the biotypes might serve as an early indicator, allowing more timely intervention that could slow, or, he hopes, even reverse, progression.

Zhang does not expect biotypes to replace symptom-based diagnosis. He sees the two approaches as complementary, with potential to ease a process that now involves a lot of trial and error. A new medication can take a couple of weeks or months to show a response, and if it fails, doctors switch to another, which he described as slow, costly, and frustrating. He envisions future clinical trials testing which treatments change which patterns of brain activity, and matching treatments to each biotype. "We want to find biomarkers that can really inform the right treatment at the right time," he said.

Yong Jiao, PhD, a postdoctoral scholar at the University of Iowa, is lead author; Zhang shares senior authorship with Jing Jiang, PhD, of Iowa's pediatric neurology department. Contributors also came from the University of Texas at Austin and the University of Pennsylvania Perelman School of Medicine. Funding came from the National Institutes of Health and several foundations. Whether biotype-guided treatment works remains to be tested in those future trials.