Rise in autism diagnoses among girls and women accelerates post-2020, study finds

New research indicates a sharp increase in autism diagnoses among females since 2020, driven by improved recognition, delayed assessments, and changing diagnostic practices, highlighting the need for earlier detection in females.

The sharp rise in new autism diagnoses recorded after 2020 in one large US health system was concentrated among girls and women, not boys and men. In the study, female first-time diagnoses almost doubled from 284 in 2016 to 513 in 2024, while the annual male count stayed in a narrower band and the male incidence trend fell by 3.8% a year across the full period. (jamanetwork.com)

That matters because the headline increase in autism assessments looked very different once the figures were split by sex. The overall rate dropped from 2016 to 2020 and then rose by 7.1% a year from 2020 to 2024, but the paper said that “aggregate growth masked a significant sex disparity”. Erick Messias, chair of psychiatry and behavioural neuroscience at Saint Louis University School of Medicine and a SLUCare physician at SSM Health, led the study, which was published in JAMA Network Open in July. (jamanetwork.com)

The steepest movement was among teenage girls. Their incidence of newly recorded autism diagnoses rose by 22.3% a year from 2020 to 2024, while diagnoses among female children increased by 17.4% a year after 2020; among boys aged under 10, by contrast, incidence declined by 2.0% a year across 2016 to 2024. The pattern was visible beyond childhood too, although the paper said the upward trend among adult women did not reach statistical significance. (jamanetwork.com)

The researchers also found that females were diagnosed later. Across the cohort, girls and women received an autism diagnosis at an average age of 15.7, compared with 12.3 for boys and men; in the child group, the median age at diagnosis was 8.5 for girls and 7.1 for boys, and among adolescents it was 17.6 against 15.8. Those delays matter because diagnosis often determines whether children, teenagers and adults can secure educational help, workplace adjustments or specialist support. (jamanetwork.com)

The timing around the pandemic is suggestive, but the authors were careful not to present Covid-19 as the cause of more autism. When 2020 was removed from the analysis, the same female-heavy rise remained, with the turning point shifting back to 2018; the paper and later coverage pointed instead to catch-up assessments after disrupted screening, wider use of telehealth and growing public awareness that autism can present differently in females. The authors also suggested that online discussion may have helped some women recognise their own traits and seek assessment. (jamanetwork.com)

Another strand of the story is diagnostic housekeeping. The paper tracked both autism spectrum disorder and Asperger disorder codes, and found that use of the Asperger label continued to fall even though ICD-10 billing codes still retained it after DSM-5 removed it as a separate diagnosis in 2013. In the researchers’ view, that decline shows how changes in coding practice can reshape autism statistics without signalling a biological shift in the population. (jamanetwork.com)

The study drew on records from 171,134 people who had at least one clinical encounter every year between 2016 and 2024 within SSM Health’s virtual data warehouse, which covers more than five million patients across Illinois, Missouri, Oklahoma and Wisconsin. That design let the team compare children, adolescents and adults within a single system before and after the pandemic, but it also narrowed the sample to people with continuous contact with care; the authors said that reliance on billing codes, a predominantly White cohort and the likelihood of over-representing patients with stable insurance all limit how far the findings can be generalised. They added a further caution for adults: nearly 70% of the adult cohort was female, which may itself have increased the chance of identifying autism in women who use healthcare more regularly. (jamanetwork.com)

The broader backdrop is that autism prevalence in US eight-year-olds has been rising for years, with CDC surveillance cited in the paper showing an increase from 18.5 per 1,000 children in the 2016 surveillance year to 32.2 per 1,000 by 2022. What this new analysis adds is the suggestion that at least part of the recent growth in recorded diagnoses may reflect a correction of who gets seen, and when, rather than a simple across-the-board surge. The practical challenge now is whether schools, clinicians and health systems can spot those missed female presentations earlier, instead of waiting until adolescence or adulthood. (jamanetwork.com)