November 3, 2026

expert reaction to observational study of changes in characteristics of adults diagnosed with autism over time

Scientists comment on an observational study published in BMJ Mental Health looking at changes in characteristics of adults diagnosed with autism over time.

Dr Hannah Bowers, Behavioural Scientist at the Primary Care Research Centre, University of Southampton, said:

“The paper is methodologically rigorous and written incredibly well, balancing multiple possible explanations for their findings in the context of existing evidence.

“These findings highlight the complexity in primary care where autistic adults are often also experiencing multiple mental health problems. We need to make sure GPs have the right support and training to care for autistic adults in a way that recognises these overlapping difficulties.

“In our own research, autistic adults have described their mental health difficulties as intertwined with their autism and that this makes getting support really challenging. Using large primary care datasets, this paper’s findings echo that lived experience of having multiple interacting difficulties. It is clear that we need primary care, mental health and autism services to work together rather than in isolation.

“Many autistic women have shared their stories of experiencing mental health problems before getting an autism diagnosis later in life. This study adds important large-scale primary care data to that picture. It helps to uncover important patterns in the rise in autism diagnoses and carefully considers different possible explanations for these based on the evidence.

“This paper highlights the need for better integration across primary care, mental health care and autism services. More integrated care could help improve diagnostic services and help people to get the right support for their needs.”

Dr Rachel Moseley, Principal Academic in Psychology, Bournemouth University, said:

“The study confirms what other, recent research has shown: a more diverse group of people are now being diagnosed as autistic. Two papers in JAMA Psychiatry, earlier this year, showed, likewise, that we’re seeing diversification of the genetic and behavioural profile of autistic people (LaBianca et al., 2026; Tarp et al., 2026).

“We are now more proficient than ever at recognising autism in girls, women and people assigned female at birth, whose profiles were therefore not showing up when we looked at autistic populations in past decades. Since 2013, we also recognise that people can be autistic AND have ADHD at the same time – so likewise, these people (‘auDHDers’) simply weren’t showing up in our studies before. We also now recognise that even when someone has a normal IQ and “milder” autistic traits, they can still experience “clinically significant impairments”, thus warranting an autism diagnosis as per current diagnostic thresholds. With all of this, since we are recognising a wider group of people as autistic, this distils the proportion with epilepsy and intellectual disabilities when we’re looking at autistic populations.

“In interpreting this study, it’s vital to highlight what it doesn’t show. Essentially, the study shows how diagnostic practices have changed, but it does not tell us that autism is overdiagnosed or that these diagnostic changes are a bad thing (as the authors acknowledge). Indeed, since we now know that people with less extreme autistic differences still differ markedly from the general population in health and even life expectancy, improved recognition of people who meet diagnostic cut-offs with “clinically significant impairments” could be argued to be highly beneficial at a personal and societal level.

Prof Dame Uta Frith, Emeritus Professor of Cognitive Development, UCL, said:

“This is a very carefully performed population study of officially registered autism diagnoses in the UK and Sweden. The data confirm previous studies of exponential increases in diagnosis over the last decade without reaching a plateau.

“The results provide further evidence of a split between early and late diagnosed autism. Cases with early diagnosed autism are linked with neurological signs, such as epilepsy and cognitive Impairments. Cases with late-diagnosed autism are instead linked with prior mental health problems, such as anxiety and depression. It is precisely this group that accounts for the increase. For example, in the UK, the proportion of adults with a newly recorded autism diagnosis and a prior history of anxiety disorders, increased from 29% in 2010 to 58% in 2023.

“New findings emerging from this study suggest that the late diagnosed group has distinct demographics, which may provide clues to the causes for its rise. They hint at cultural changes in the meaning and use of the autism label rather than biological changes.”

Dr Monique Botha, Associate Professor in Social and Developmental Psychology, Durham University, said:

“This is a large, methodologically strong population-based study documenting substantial changes in the demographic and psychiatric profiles of adults receiving autism diagnoses. Its strengths include the scale of the datasets, cross-national comparisons and identification of inequalities in diagnostic recognition. However, the descriptive findings are more compelling than some of the explanations proposed.

“In the paper, the authors hypothesise that that increasing psychiatric diagnoses among autistic adults may reflect diagnostic overinclusiveness. However, elevated rates of anxiety, depression and trauma-related conditions among autistic people are already well established. Large studies repeatedly show us that autistic adults are more likely to have depression, anxiety, self-harm, and suicidal ideation regardless of whether those diagnoses or the developmental ones came first. These difficulties cannot be considered independently of the social determinants of autistic mental health, including discrimination, exclusion, reduced quality of life and disproportionate exposure to repeated victimisation, including sexual violence.

“This is particularly relevant to autistic women. Research demonstrates that girls with autistic characteristics comparable to boys are less likely to be referred for assessment or diagnosed, contributing to substantial delays in recognition. Such delays increase both exposure to potentially harmful environments and opportunities for psychiatric difficulties to develop and diagnoses to accumulate. Consequently, the increasingly extensive psychiatric histories observed among recently diagnosed autistic adults may partly reflect the cumulative consequences of historical diagnostic inequalities rather than changing diagnostic boundaries.

“The study cannot adequately distinguish between these explanations. Although age-standardised, the analyses do not account for diagnostic delay, differential psychiatric service contact, victimisation or other social determinants of mental health. Nor do they independently establish the validity of autism or preceding psychiatric diagnoses. These limitations are particularly important because the population receiving diagnoses is itself shaped by unequal access to recognition.

“The findings concerning ethnic and migration-related disparities are important and warrant further investigation, particularly into inequalities in referral, assessment and access to support.

“Overall, this is valuable descriptive epidemiology, but its interpretation does not sufficiently engage with established evidence on autistic mental health and the cumulative consequences of diagnostic exclusion. Changes in the psychiatric profiles of diagnosed autistic adults cannot establish that diagnostic thresholds have weakened or that other psychiatric conditions are increasingly being mistaken for autism. Indeed, the reverse possibility, that autism has historically gone unrecognised while psychiatric diagnoses accumulate, is well documented. The study demonstrates a changing diagnosed population, not declining diagnostic validity, and these conclusions should not be conflated.”

* ‘Changes in the profile of adults diagnosed as autistic since 2010: population-based studies in the UK and Sweden’ by Aws Sadik et al. was published in BMJ Mental Health at 23:30 UK time on Tuesday 3 November

DOI: 10.1136/ bmjment-2026-302983

Declared interests

Dr Hannah Bowers: My job is partly funded by NIHR and I am the lead researcher on several NIHR funded projects. (The views expressed are those of the author(s) and not necessarily those of the NIHR or the Department of Health and Social Care.) A small portion of my time is funded by Bristol, North Somerset, and South Gloucestershire Integrated Care Board. The rest of my salary is funded by the University of Southampton Faculty of Medicine.

I was paid in stock options for consulting on written content for an American antidepressant deprescribing app in 2022.

One of the co-authors (D Rai) is a co-investigator on two of my NIHR funded studies but I had no involvement with this work and this data/analysis is not related to my own studies.

Dr Rachel Moseley: “No conflicts of interest”

Prof Dame Uta Frith: “I have no conflict of interest.”

Dr Monique Botha: “No COIs”