Summary: A new study demonstrates that artificial intelligence can accurately estimate a child’s risk of developing ADHD years before a clinical diagnosis occurs. By mining “hidden patterns” in routine Electronic Health Records (EHR) from birth through early childhood, the AI identifies combinations of developmental and behavioral markers that human clinicians might overlook during brief visits.
This tool is designed to act as a “clinical safety net,” ensuring at-risk children receive early evaluations and support during critical developmental windows.
Key Facts
- The Dataset: Researchers analyzed the medical history of over 140,000 children, creating a massive comparative baseline of those with and without ADHD.
- Early Detection: The AI model analyzes data from birth and becomes highly accurate at estimating future risk by age 5, well before the average age of diagnosis.
- Equitable Performance: A critical finding was the model’s consistent accuracy across various demographics, including sex, race, ethnicity, and insurance status, suggesting it could help reduce existing disparities in ADHD care.
- Support, Not Diagnosis: The tool is explicitly not an “AI doctor.” Its purpose is to flag children who should be prioritized for screening by primary care providers or specialists.
- Better Outcomes: Early identification is directly linked to improved academic, social, and long-term health outcomes, as it allows for evidence-based interventions before a child falls behind.
Source: Duke University
Attention-deficit/hyperactivity disorder (ADHD) affects millions of children, yet many go years without a diagnosis, missing the chance for early support that can change long-term outcomes even when early signs are present.
In a new study, Duke Health researchers found that artificial intelligence tools can analyze routine electronic health records to accurately estimate a child’s risk of developing ADHD years before a typical diagnosis. By reviewing patterns in everyday medical data, the approach could help flag children who may benefit from earlier evaluation and follow-up.
The research, published in Nature Mental Health on April 27, highlights how powerful insights can come from information already collected during regular health care visits to help support early decision making by primary care providers.
“We have this incredibly rich source of information sitting in electronic health records,” said Elliot Hill, lead author of the study and data scientist in the Department of Biostatistics & Bioinformatics at Duke University School of Medicine.
“The idea was to see whether patterns hidden in that data could help us predict which children might later be diagnosed with ADHD, well before that diagnosis usually happens.”
To arrive at the findings, researchers analyzed electronic health records from more than 140,000 children, with and without ADHD. They trained a specialized AI model to look at medical history from birth through early childhood. The model learned to recognize combinations of developmental, behavioral, and clinical events that often appeared years before an ADHD diagnosis was made.
The model was highly accurate at estimating future ADHD risk in children age 5 and older, with consistent performance across patient characteristics like sex, race, ethnicity, and insurance status.
Importantly, the tool does not make a diagnosis. It identifies children who may benefit from closer attention by their pediatric primary care provider or an earlier referral for ADHD assessment by a specialist.
“This is not an AI doctor,” said Matthew Engelhard, M.D., Ph.D., in Duke’s Department of Biostatistics & Bioinformatics, and senior author of the study. “It’s a tool to help clinicians focus their time and resources, so kids who need help don’t fall through the cracks or wait years for answers.”
The researchers note that earlier identification for screening could lead to earlier diagnosis and therefore earlier support, which is linked to better academic, social, and health outcomes for children with ADHD. They also emphasize the need for further studies before such tools are used in clinical settings.
“Children with ADHD can really struggle when their needs aren’t understood and adequate supports are not in place,” said study author, Naomi Davis, Ph.D., associate professor in the Department of Psychiatry and Behavioral Sciences. “Connecting families with timely, evidence-based interventions is essential for helping them achieve their goals and laying a foundation for future success.”
Hill and Engelhard have also researched the use of AI models in predicting potential risks and causes for mental illness in adolescents.
In addition to Hill Engelhard, and Davis, the authors for this study include De Rong Loh, Benjamin A. Goldstein, and Geraldine Dawson.
Funding: The study was supported by grants from the National Institute of Mental Health (K01-MH127309, UL1 TR002553) and National Center for Advancing Translational Sciences.
Key Questions Answered:
A: The AI looks at the timing and combination of events, such as specific developmental delays, sleep disruptions, or frequent visits for behavioral concerns, that may seem insignificant on their own but together form a “risk signature” for ADHD.
A: No. Senior author Dr. Matthew Engelhard emphasizes that this is a resource-management tool. It helps pediatricians know which kids need a closer look so they don’t “fall through the cracks” while waiting years for a standard evaluation.
A: Primary care providers often have very limited time with patients. AI can scan thousands of pages of a child’s medical history in seconds, highlighting relevant clinical trends that might have happened years prior or with a different physician.
Editorial Notes:
- This article was edited by a Neuroscience News editor.
- Journal paper reviewed in full.
- Additional context added by our staff.
About this AI and ADHD research news
Author: Stephanie Lopez
Source: Duke University
Contact: Stephanie Lopez – Duke University
Image: The image is credited to Neuroscience News
Original Research: Closed access.
“Fetal and postnatal metal metabolism-related changes in brain function are associated with childhood behavioral deficits” by Elliot D. Hill, De Rong Loh, Naomi O. Davis, Benjamin A. Goldstein, Geraldine Dawson & Matthew Engelhard. Nature Mental Health
DOI:10.1038/s44220-026-00628-2
Abstract
Attention deficit hyperactivity disorder (ADHD) is a common neurodevelopmental condition that can negatively impact long-term outcomes for individuals. Early diagnosis is critical, yet demographic and clinical disparities can delay detection.
Using electronic health records (EHRs) from a cohort of over 720,000 patients, we pretrained an EHR foundation model. We then fine tuned it to predict the likelihood of ADHD diagnosis and timing from birth until age 9 years in a pediatric cohort of over 140,000 patients.
By age 5 years, the model achieved a time-dependent area under the receiver operating characteristic curve of 0.92 at a 4-year time horizon. Overall, the model maintained its performance across patients with differing demographics, including sex, race, ethnicity and insurance status.
Our feature importance analysis found that ADHD was strongly associated with developmental, behavioral and psychiatric conditions. Our results suggest that EHR-based predictive models could help providers reliably identify children with ADHD in a timely manner.
