AI Can Accurately Flag Pediatric ADHD: New Study - ADDitude

May 19, 2026

Artificial intelligence tools can accurately estimate a child’s risk of having ADHD years before a typical evaluation and diagnosis, finds a new study published in Nature Mental Health.¹ The potential consequences of this analysis were highlighted by another new study, which found that children with ADHD who were diagnosed at earlier ages had higher GPAs and were far less likely to drop out of high school, compared with kids who were diagnosed in adolescence.

The first study was conducted by a team of researchers at Duke Health who pre-trained an AI model to analyze patients’ medical history using the electronic health records (EHRs) of 720,000 individuals. They further refined the model, training it with the records of 140,000 children with and without ADHD such that the model was able to identify developmental, behavioral, and clinical patterns associated with a diagnosis of ADHD.

Using data available at age 5, the AI model predicted which children would be diagnosed with ADHD between ages 5 and 9 with 92% accuracy. The model’s accuracy remained consistent over various demographics, with subjects of differing sex, race, ethnicity, and insurance status.

“Early diagnosis is critical, yet demographic and clinical disparities can delay detection,” write the study’s authors. “EHR-based predictive models could help providers reliably identify children with ADHD in a timely manner.”

“We have this incredibly rich source of information sitting in electronic health records,” said the study’s lead author, Elliot Hill, Ph.D., M.D., in a press release. “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.”

Early Diagnosis Linked to Higher GPAs Later

“Connecting families with timely, evidence-based interventions is essential for helping [students] achieve their goals and laying a foundation for future success,” wrote Naomi Davis, Ph.D., co-author of the Nature Mental Health study regarding the potential impact of predictive diagnostic models in ADHD.

This assertion was validated by the findings of another new study, published in JAMA Psychiatry.² Using Finland’s national registry data for individuals born throughout the 1990s, researchers found that receiving an ADHD diagnosis at an earlier age was associated with higher GPAs and lower high school dropout rates, compared with receiving diagnoses at older ages.

After adjusting for sociodemographic variables, the researchers found that, compared with those who received a diagnosis in mid-adolescence (ages 14-16), children who received an early diagnosis (ages 4-6):

• Had significantly higher GPAs (by .6 to .7 points on a 10-point scale)
• Were three times less likely to drop out of high school
• Were four times more likely to complete an academic track in high school (2.5 times more likely for girls)

The Long-Term Benefits of Early Treatment

Early and consistent ADHD treatment helps children with more than academic performance, explained Adelaide S. Robb, M.D., in the ADDitude webinar, “Beyond the Core Symptoms of ADHD.” “One 10-year follow-up study also found that stimulants, especially when taken in early childhood, have protective effects on the rates of developing comorbidities like depression, anxiety, and ODD,” she wrote. “It has real, long-term impacts on areas like higher education and employment, involvement with the law, risky behavior, and overall quality of life.”

In an April 2026 ADDitude webinar titled, “The Brain Chemistry of ADHD,” Greg Mattingly, M.D., echoed this sentiment, explaining that treating ADHD early can have long-lasting benefits: “There is now data out there… that treatment isn’t just about the symptoms now; it’s about preventing damage in the brain so you don’t develop secondary issues like anxiety, depression, emotional dysregulation, and insomnia.”

View Article Sources