Changes in the profile of adults diagnosed as autistic since 2010: population based studies in England and Sweden

This population-based cohort study of over 33 million adults in the UK and Sweden reveals that since 2010, the profile of newly diagnosed autistic individuals has shifted dramatically toward younger people, females, and those with prior mental health conditions like anxiety and ADHD, while diagnoses of epilepsy, psychosis, and intellectual disability have declined.

The Gist

Imagine the landscape of autism diagnoses in the UK and Sweden as a vast, shifting garden. For decades, gardeners (doctors and researchers) have been counting the flowers (people diagnosed with autism). This study, which looked at the garden from 2010 to the early 2020s, found that the garden hasn't just grown bigger; the types of flowers blooming there have changed dramatically.

Here is what the researchers found, broken down into simple concepts:

1. The Garden is Growing Faster in Some Areas

The number of adults getting diagnosed with autism has skyrocketed in both countries. However, the speed of this growth isn't the same everywhere.

• The "Young and Female" Sprout: The fastest-growing patches of the garden are among young adults (ages 16–20) and women. In fact, by the end of the study period, new diagnoses for adult women were actually happening more frequently than for men.
• The "White" and "Swedish-Born" Soil: In the UK, the garden grew fastest in areas with White ethnic groups. In Sweden, it grew fastest among people whose parents were both born in Sweden. Conversely, the garden grew much slower in Black, South Asian, and immigrant communities in these countries. The gap between these groups is widening, suggesting that some people are finding it much harder to get a diagnosis than others.

2. The "Profile" of the Flower Has Changed

If you look at the "autism flower" from 2010 and compare it to the one from 2023, they look quite different. The researchers found a distinct shift in the "background story" of the people being diagnosed.

• The "Heavy Backpack" is Getting Lighter: In the past, people diagnosed with autism often carried a heavy backpack of other serious conditions, such as intellectual disabilities, epilepsy, or psychosis. Today, fewer people diagnosed with autism carry this specific heavy load. The proportion of adults with these conditions before their autism diagnosis has dropped significantly.
• The "Anxiety Umbrella" is Getting Heavier: While the heavy backpack is lighter, a different kind of burden has grown. People diagnosed with autism today are much more likely to have a history of anxiety, depression, ADHD, sleep issues, or eating disorders. It's as if the garden is now filled with flowers that are struggling with the weather (mental health challenges) rather than the soil itself (intellectual disability).

3. Two Different Gardens, Similar Weather Patterns

The study compared two different countries: the UK (using primary care records, like a local neighborhood watch) and Sweden (using national registers, like a massive government database).

• Different Speeds: Sweden generally had higher diagnosis rates than the UK, but the UK saw a massive spike in diagnoses after 2020.
• Same Trends: Despite the different data sources, both countries told the same story: fewer people with intellectual disabilities are being diagnosed, and many more people with anxiety and depression are being diagnosed.

4. What Does This Mean? (The Authors' View)

The researchers are careful not to say exactly why this is happening, but they offer a few possibilities, like different lenses to view the garden through:

• Better Eyes: We might finally be seeing the "hidden" flowers—specifically women and people who are good at masking their struggles—who were previously overlooked.
• Changing Rules: The rules for what counts as a "flower" might have broadened. Doctors might be more willing to label someone as autistic even if they have other mental health issues first.
• The "Key" to Support: Sometimes, getting a diagnosis is the only way to unlock support services. If a person is struggling with anxiety but can't get help, they might be diagnosed with autism because that is the "key" that opens the door to care.

The Bottom Line

The study concludes that the "face" of autism in adults has changed completely since 2010. It is no longer just the story of a child with intellectual disabilities; it is increasingly the story of a young adult, often female, navigating a world with high levels of anxiety and depression.

The authors warn that while this might mean we are finally recognizing more people, it also raises questions: Are we seeing the whole picture, or are we misidentifying other struggles as autism? They emphasize that we need to understand these changes to ensure that the right support reaches the right people, regardless of their background or the specific mix of challenges they face.

Technical Summary

Problem Statement
While the number of autism diagnoses has increased globally in recent decades, the underlying distribution of autistic features in the general population appears stable. This discrepancy raises questions regarding whether the rise in diagnoses reflects a more nuanced recognition of the condition (e.g., in females or those without language delays) or an over-inclusive broadening of diagnostic criteria that may pathologize normal variation or misattribute symptoms of other mental health conditions. Existing literature suggests shifts in the sociodemographic and neuropsychiatric profiles of diagnosed individuals, but there is a need for large-scale, cross-national evidence to determine if these trends are generalizable or localized phenomena. This study investigates whether the sociodemographic and neuropsychiatric characteristics of adults diagnosed with autism in the United Kingdom (UK) and Sweden have changed since 2010.

Methodology
The study employed cross-context population-based cohort designs using routine health records from two nations with distinct healthcare systems.

• Data Sources:
  • UK: The Clinical Practice Research Datalink (CPRD) Aurum, containing primary care records from 1,784 general practices (March 2024 extract). The study included 24,537,039 individuals aged 16 or over, with 141,119 having an autism diagnosis. Data covered the period 2010–2023.
  • Sweden: Linked national registers (Total Population Register, National Patient Register, Cause of Death Register, and LISA). The study included 9,096,874 individuals aged 16 or over, with over 100,817 having an autism diagnosis. Data covered the period 2010–2021.
• Study Population: Individuals entered the study upon meeting criteria (Jan 1, 2010, age 16, and registration duration). Follow-up ended at death, loss of registration, migration, or the study end date.
• Variables:
  • Sociodemographic: Age, sex, ethnicity (UK), parental birth country, parental education, and household income (Sweden). Area-level deprivation was measured via the Index of Multiple Deprivation (IMD) in the UK.
  • Neuropsychiatric: First recorded diagnoses of conditions including intellectual disability (ID), epilepsy, ADHD, anxiety, depression, psychosis, and others.
  • Medications: First recorded prescriptions/dispensations for antipsychotics, antidepressants, antiseizure medications, opioids, gabapentinoids, melatonin, and sedatives.
• Statistical Analysis:
  • Annual age-standardized incidence and prevalence rates were calculated.
  • For incident diagnoses, newly diagnosed autistic adults were matched with up to four non-autistic comparators (age, sex, practice/municipality).
  • Age-standardized proportions of prior neuropsychiatric conditions and medications were calculated for both incident and prevalent (all existing) autistic adults.
  • Standardization used the mid-2022 UK age distribution (16–65 years) for most analyses, with specific adjustments for Swedish parental data (restricted to 16–35 years).
  • Confidence intervals were calculated using Byar's method with Dobson adjustment, Agresti/Caffo for absolute differences, and Wald/Katz-log for rate ratios.

Key Contributions
This study provides a comprehensive, longitudinal analysis of adult autism diagnosis trends across two high-income nations with different healthcare structures. Its primary contributions include:

1. Cross-National Validation: Demonstrating that shifts in the profile of diagnosed adults are consistent across the UK and Sweden, suggesting these are not merely artifacts of a single healthcare system.
2. Dual Perspective: Analyzing both incident (newly diagnosed) and prevalent (all diagnosed) populations to distinguish current diagnostic trends from the cumulative public health burden.
3. Granular Neuropsychiatric Profiling: Moving beyond simple prevalence counts to detail the changing landscape of co-occurring conditions and medication use, specifically highlighting the divergence between conditions that are decreasing (e.g., ID, epilepsy) and those increasing (e.g., anxiety, ADHD).
4. Sociodemographic Disparities: Quantifying widening gaps in diagnosis rates between ethnic groups in the UK and parental backgrounds in Sweden, and noting sex-specific shifts where female diagnosis rates are overtaking male rates in incident cases.

Results

• Incidence Trends: Incident autism diagnoses were consistently higher in Sweden than the UK, though the UK saw a rapid increase after 2020. In both nations, incidence increased fastest for 16–20-year-olds and females. By the end of the study period, incident diagnosis rates for females exceeded those for males in both countries.
• Sociodemographic Shifts:
  • UK: Incidence increased faster in White and Mixed ethnic groups compared to Black, South Asian, or Other groups. In 2023, the age-standardized incidence for White 16–65-year-olds was 11 per 10,000 person-years, compared to 2.2 for South Asian groups. Deprivation patterns were nuanced; while overall incidence was higher in deprived areas, this pattern disappeared for females by the study's end.
  • Sweden: Incidence increased faster among adults with Swedish-born parents compared to those with immigrant backgrounds.
• Neuropsychiatric Profile Changes:
  • Decreasing Conditions: There was a consistent decline in the proportion of newly diagnosed autistic adults with prior records of epilepsy, intellectual disability (ID), and psychotic disorders. For example, in the UK, prior epilepsy records dropped from 10% (2010) to 4% (2023).
  • Increasing Conditions: Conversely, there was a marked increase in prior records of anxiety, depression, ADHD, sleep disorders, eating disorders, and stress-related conditions. In the UK, prior anxiety records rose from 28.7% (2010) to 58.3% (2023).
  • Medications: Trends mirrored diagnoses, with decreases in antiseizure and antipsychotic medications and increases in antidepressants, melatonin, and gabapentinoids.
  • Sex Differences: Mental health conditions were generally more common in females. The reduction in ID diagnoses over time was greater in magnitude for females than males.
• Prevalence: Prevalence increased across all age groups but remained higher in younger adults. The male-to-female prevalence ratio decreased significantly (from 4.1 to 2.2 in the UK; 1.9 to 1.6 in Sweden), though males still constituted the majority of the prevalent population.

Significance and Claims
The paper concludes that the sociodemographic and neuropsychiatric characteristics of adults diagnosed with autism have changed dramatically since 2010. The authors claim these findings indicate that:

1. Diagnostic Inequity: Diagnoses are inequitably distributed, with widening disparities between ethnic groups in the UK and parental backgrounds in Sweden.
2. Diagnostic Complexity: The rising prevalence of co-occurring mental health conditions (anxiety, depression, ADHD) adds complexity to the diagnostic process, suggesting that clinicians are increasingly diagnosing autism in individuals with significant psychiatric comorbidities.
3. Evolving Support Needs: The average health needs of autistic groups are shifting away from those with intellectual disability and epilepsy toward those with co-occurring mental health conditions. This has direct consequences for public health policy and the provision of support services.

The authors explicitly state that it remains unclear whether these changes reflect a "nuanced recognition of autism" (e.g., identifying camouflaged presentations in females) or a "broadening of diagnostic practice" (potentially misattributing other conditions to autism). They emphasize that further research is required to understand the drivers of these changes and their impact on support provision, noting that the current data cannot definitively distinguish between improved recognition and potential misdiagnosis. The study calls for future research co-produced with the autistic community to address these disparities and evolving needs.