A study of more than 2.1 million children and adolescents in Denmark suggests that part of the rise in ADHD and autism diagnoses reflects a change in who gets diagnosed, not only a rise in the underlying conditions. The research, led by the Barcelona Institute for Global Health (ISGlobal) and Aarhus University and published in JAMA Psychiatry, found that children diagnosed more recently look much more like the general population than those diagnosed a decade earlier. The comparison covers diagnoses made between 2012 and 2022, and the authors argue that the shift in who is being diagnosed deserves as much attention as the rising counts themselves.
Magnus Elias Tarp, the first author, said people diagnosed more recently resemble the general population more closely than those diagnosed a decade ago. Senior author Oleguer Plana-Ripoll said the rise “should not be interpreted solely as evidence of an increase in the underlying occurrence” of ADHD or autism.
Nationwide registry data and a matched comparison
The team drew on nationwide Danish registry data covering everyone born between January 1994 and the end of 2022, a population of 2,194,951 children followed up to age 18. According to the abstract in JAMA Psychiatry, the design was a matched case-control study. Cases were 71,317 young people diagnosed with ADHD or autism spectrum disorder between 2012 and 2022, of whom 37.1% were female and whose median age at diagnosis was 11.4 years. Each was compared with matched peers who never received either diagnosis, 713,170 in all.
The researchers then asked how well 19 characteristics recorded before diagnosis, among them parental factors, family history, perinatal factors and patterns of healthcare use, separated the diagnosed from the undiagnosed. They did this year by year, so that the strength of each association could be tracked across the decade, and the analysis ran from April 2025 to June 2026. If diagnoses had simply risen because more children carried the same risks, the associations would have held steady. If the net had widened to include children without those markers, the associations would weaken, and that is the pattern the registry data show.
Low birth weight as the clearest signal
Low birth weight showed the change most starkly. In 2012 and 2013, children born small were 54% more likely to receive an ADHD or autism diagnosis (odds ratio 1.54, 95% confidence interval 1.41 to 1.68). By 2020 to 2022 the odds ratio had fallen to 1.17 (1.10 to 1.24), a 17% higher likelihood. A risk factor that once picked out the children who were diagnosed had lost much of its power to do so.
The ISGlobal and Aarhus release frames the pattern as a population drifting toward the average: associations between diagnosis and early-life risk factors narrowed across the board. The abstract adds a difference between conditions. ADHD showed more attenuation than autism, with the association with household income declining 6.5% a year for ADHD against 0.7% a year for autism.
Broader identification, not a verdict on risk factors
The authors list several reasons the diagnosed group may have widened: greater awareness, improved identification by healthcare services, wider access to diagnostic services and evolving clinical thresholds. The ScienceDaily report on the paper quotes the team’s conclusion that rising rates “may partly reflect broader identification and changing diagnostic patterns, rather than only an increase in the underlying conditions.”
The researchers are careful about what that does not mean. In their words, “the key message is not that these risk factors are no longer important.” The associations weakened rather than vanished, and the odds ratio for low birth weight still sits above 1 in the latest period.
The design carries limits that the team states openly. It cannot isolate which of the possible drivers, such as awareness, access or shifting thresholds, accounts for how much of the change, and it describes Denmark only. Coverage of the paper by Sense About Science and a SciTechDaily write-up repeat the same calls for replication in other health systems before the result is generalized.
The practical stakes sit with planners and researchers. Studies of long-term outcomes that treat a diagnosis as a fixed label may be comparing groups that differ from one decade to the next, and service planners who read rising counts as rising disease burden may be misreading a shift in who is identified. The release lists consequences for healthcare planning and outcome research, and the authors say diagnostic trends need recalibrated interpretation: a rising count tells planners how many children are being identified, which is a separate quantity from how many have the condition.
Denmark’s registries show the diagnosed group converging on the general population, but they do not say how much of the surge in diagnoses that convergence accounts for. The authors point to replication in other health systems as the next test, and until another country’s data are run the same way, the portion of the rise that reflects widened identification stays unmeasured.
This article was produced with the assistance of AI and reviewed by Morning Overview editors prior to publication.
More from Morning Overview
- Early electric-car owners are hitting battery and screen failures no one warned them about
- A magnitude 5.3 quake struck off the Oregon coast this week
- Amazon’s Prime refunds are rising to $200 as millions more customers become eligible
- A handful of car engines are so tough mechanics say they almost never wear out
