A large study published in JAMA Psychiatry has found evidence of an ADHD autism diagnosis shift over the past decade: children and adolescents receiving these diagnoses today resemble the general population more closely than those who received the same diagnoses ten years ago. The findings, drawn from data on more than 2.1 million children and adolescents in Denmark, suggest that rising diagnosis rates partly reflect changes in who is being identified, not solely an increase in the underlying conditions.
What the Danish registry data show
The study was led by the Barcelona Institute for Global Health (ISGlobal), a centre supported by the ‘la Caixa’ Foundation, and Aarhus University. According to Life Science Network, the research drew on nationwide Danish registry data covering individuals born between 1 January 1994 and 31 December 2022, with each person followed up to age 18. Data analysis ran from April 2025 through June 2026.
The researchers compared more than 71,000 young people diagnosed with attention deficit/hyperactivity disorder (ADHD) or autism spectrum disorder (ASD) between 2012 and 2022 against more than 713,000 individuals who had not received either diagnosis. Previous research had established several characteristics that appear more often among people diagnosed with ADHD or ASD: being born prematurely or at low birth weight, having parents with lower levels of education or lower household income, having a parent with a history of psychiatric disorders, and making heavier use of healthcare services before diagnosis. The team wanted to know whether those patterns had held as diagnosis rates climbed.
They had not held, at least not at the same strength. Young people diagnosed with ADHD or ASD were still more likely than their undiagnosed peers to carry many of these characteristics, but the size of the differences narrowed consistently across the study period. Children born at low birth weight were 54% more likely to receive an ADHD or ASD diagnosis than those born at normal weight at the start of the period; by the end, that figure had fallen to 17%. Comparable declines were recorded for associations involving premature birth and several measures of socioeconomic disadvantage.
The ADHD autism diagnosis shift and what drives it
Magnus Elias Tarp, a PhD student at Aarhus University and first author of the study, is careful about what the pattern does and does not mean. ‘The key message is not that these risk factors are no longer important,’ he says. ‘What we found is that people diagnosed in recent years resemble the general population more closely than those who received the same diagnoses a decade ago.’
The researchers offer several possible explanations for the shift, without claiming to have measured the contribution of any one of them. Awareness of ADHD and ASD has grown. Healthcare and education systems may be identifying cases more effectively. Access to diagnostic evaluations may have expanded. And the threshold at which neurodevelopmental characteristics are considered clinically impairing may itself have moved. The study design cannot disentangle these influences.
What it can say is that the ADHD autism diagnosis shift recorded in Danish data means rising diagnosis counts cannot straightforwardly be read as evidence that the underlying occurrence of either condition has increased by the same margin. The two things, diagnosis rate and true prevalence, are not interchangeable, and this study offers one quantified illustration of why.
Oleguer Plana-Ripoll, researcher at ISGlobal and Aarhus University and senior author of the study, is direct about what the findings do not show. ‘They do not show that these conditions are being overdiagnosed or that they have become less severe,’ he says. ‘Rather, they indicate that the population receiving these diagnoses has changed over time, and this needs to be taken into account when interpreting current trends.’ He adds that further research in other countries would be needed to determine whether the findings replicate elsewhere.
Implications for research and service planning
The study’s authors say the results carry practical implications beyond academic debate about prevalence. If the characteristics of the diagnosed population are shifting, that has consequences for how healthcare, education, and social services are planned and resourced for people with these diagnoses.
There is a subtler methodological concern too. Some research has suggested that outcomes for people with ADHD or ASD are improving over time. The authors caution that apparent improvements may partly reflect changes in the composition of the group receiving diagnoses, rather than genuine changes in the prognosis of the conditions themselves. Separating the two effects is not straightforward, and the Danish findings add a reason to be careful when drawing that kind of conclusion.
The study was published in JAMA Psychiatry under the title ‘Changes in Characteristics Associated With ADHD and ASD Diagnoses Over Time’, with a DOI of 10.1001/jamapsychiatry.2026.2722. The authors call for replication in other national datasets before any broader generalisations are drawn.
