The shocking state of autism diagnosis in the UK

Autism is under-diagnosed. This fact has been a truth within our online Autistic spaces for as long as I’ve been in them. Despite mounting evidence that many adults are struggling through life without access to formal identification, many of us face discriminatory and financially inaccessible systems that feel as though they are boxing us out of knowing ourselves. The time has come to deconstruct this debacle, this is the sad state of autism diagnosis.

What is the average age of autism diagnosis in the UK?

Trying to get am accurate number for the average age of autism diagnosis is difficult because much of the research focuses on children and young people (more on this later). As an example, Brett et al (2016) found the median age at diagnosis to be 55 months (or 5 years and 7 months). The problem with this study was that it only looked at children and young people, and approached autism using outdated subtypes.

A meta-analysis by van ‘t Hof (2020) looked more broadly and found the average age of diagnosis to have a mean age of 6 years (the ages ranged from 2.5 to 19.5 years). While this is better, it still fails to capture people who are inevitably diagnosed much later that might have changed the averages. In truth, I have struggled to find any research that provides robust findings on the average age of an autism diagnosis.

Gender and ethnicity differences for age at autism diagnosis

This was once again hampered by an overwhelming research investment into children, with very little robust data on adults. Petrou et al (2018) found that when age of diagnosis was 5 years or more, girls recieved their diagnosis later. For males and females, toileting issues and aggression were associated with earlier diagnosis. On the flip-side, having co-occuring diagnoses was associated with later diagnosis.

Three things stood out to me in the aforementioned research into gender differences among children

  • Gender was treated as a binary. This is fallacious as gender diversity exists among children and young people.
  • Toileting issues and aggression were associated with earlier diagnosis. I believe this highlights the stereotypes that exist within the diagnostic system.
  • Co-occurring diagnoses were associated with earlier diagnosis. When we consider that Autistic people tend to have multiple co-occurring diagnoses, it worries me how many of us may be denied access to our identity.

In terms of race and ethnicity, Roman-Urrestarazu et al (2021) found that while the general prevalence of autism among school age children was 1.76%, this figure increased to 2.11% for Black children. Despite this, I have struggled to find robust research into the average age of diagnosis for different ethnicities and races. Despite this increased prevalence, Tromans et al (2021) found that the identification of autism was generally lower for minority ethnic groups. This may explain the prior research’s finding that prevalence rates of autism in Roma communities is only 0.85%.

Is autism under-diagnosed?

Given the lack of good understanding around BIPOC and gender diverse experiences of autism, and the aforementioned impact of co-occurring diagnoses delaying autism diagnosis, it is reasonable to assume that autism is in fact under-diagnosed.

Under-diagnosis is most apparent when we look at age related differences in diagnosis. O’Nions et al (2023) found huge age-related disparities, with 2.94% of their co-hort aged 10-14 years having a diagnosis, compared to just 0.02% of people aged 70 or over. This same study estimated that out of a population of 56.5 million, 463,500 are diagnosed Autistic compared to a further 435,700 to 1,197,300 are undiagnosed. This means that if everyone in the UK was properly and accurately assessed, we could expect the number of autism diagnoses to rise by up to 258%.


The diagnostic environment of the UK is currently crippled by racial and gender bias, ageism, and stereotypes that lead diagnosticians to believe that autism only exists in aggressive little boys. Added on top of this is the fact that co-occurring diagnoses appear to be delaying autism diagnosis when they should, in fact, be an indicator that autism should be considered as a factor.

To my mind, this issue has arisen from dominant neuronormative narrative perpetuated by neurotypical researchers, most of whom come from a privileged Western culture. In order to level the playing field, we need to be encouraging research and diagnostic practice that is directly informed by not just Autistic people, but Autistic people from the diverse intersections that so many of us exists upon.

The UK is lauded for its universal healthcare. It’s time that the state of diagnosis in the UK reflected that.