Last week I, unfortunately, read an interview with autism researcher Uta Frith in TES Magazine. (Linking for the #Discourse) In the interview, she shared her opinion that autism may not truly be a spectrum and raised concerns that the diagnostic category has expanded too broadly.
As it turns out, I have opinions about her opinion.
Let’s start with some context as to why this even matters. Frith is not a minor figure in autism research. Her work helped shape modern cognitive models of autism, particularly through research on social cognition and theory of mind. Many clinicians and researchers have learned from the frameworks she helped build. Her words have weight and influence.
Which is why I was disheartened by this opinion piece. To be clear, debate about what the diagnostic criteria of any condition should be is welcome; after all this is how we do science. But this interview sets aside a lot of the contemporary research and understanding in autism spectrum disorders. By framing the conversation about who should be included rather than what the criteria should be, Frith seems to speak over autistic voices who have historically been excluded.
When someone with this much influence (let’s be real, she’s been titled Dame because of her work) speaks about autism publicly, the impacts are way beyond the bounds of academic debate. Saying a broad section of the Autistic population is “just overly sensitive” is unkind and not true.
So let’s take a careful look at what the evidence actually says.
Why the Spectrum Concept Exists
Autism is currently defined clinically by differences in social communication combined with restricted or repetitive behaviours. These autistic traits must have existed in early development and cause clinically significant impairment. Their is variety in how these traits can appear and folks can experience them with different intensities.
This variability is precisely why the DSM-5 conceptualized autism as a spectrum disorder.
Two autistic people may share certain traits but experience very different support needs, strengths, and challenges. Some folks may require substantial daily support. Others may live independently but struggle with sensory overload or social communication in certain environments. The level of support needed and impact of someones autistic traits upon their daily life can be variable, not just person to person, but for one person in different seasons of life or even day-to-day.
The spectrum model does not claim that autism is infinitely broad or without boundaries. It acknowledges that autism is heterogeneous.
The Problem With Dismissing Diagnostic Bias
A concerning claim in the interview was the suggestion that there has not been meaningful gender bias in autism diagnosis. The literature suggests otherwise.
Historically, autism has been described as “primarily affecting males,” with prevalence estimates commonly reported around four males for every female diagnosed. (My apologies for the binary language, which reflects the research language not the expansive gender experiences of Autistic people) But researchers increasingly believe that this ratio reflects diagnostic bias rather than true prevalence.
Several factors contribute to this:
Early diagnostic criteria were developed largely from studies of boys 🙃
This has created a reality where are current diagnostic tools often show reduced sensitivity to women, diagnosed with autism, who are then left out of research trials.
Some folks camouflage or mask traits in social environments
Clinicians frequently misattribute autistic traits to other conditions due to lack of knowledge and practitioner bias
Masking, also called camouflaging, despite Firths claim that it has “no scientific basis” is a well-documented phenomenon in autism research and can be measured using standardized psychometric tools. Many autistic individuals consciously or unconsciously modify behaviour to fit social expectations. This strategy can delay diagnosis and is associated with increased mental health risks.
In other words, the apparent rise in diagnoses may reflect recognition of people who were previously missed, rather than expansion of the category to include people who are not autistic.
This is such a big difference and it matters a lot.
The history of autism diagnosis includes not only over-pathologizing difference, but also missing entire groups of people. The group of people being, any one who wasn’t a wealthy white boy child.
The Lost Generation of Autistic Adults
In my own clinical work, I’ve had my own experience of understanding and applying the 2013 DSM-V definition of autism. Twelve years ago, I would have said I had no Autistic patients in my practice.
Today, about one in five of my patients are Autistic.
Some of those individuals were already my patients a decade ago.
They didn’t change. What changed was our understanding.
It took time (Ok it’s still taking time) for clinicians to “catch-up” to the 2013 criteria; to understand the full spectrum of how autistic traits can present; to challenge their own biases and previous learning; and to consider Autism as a possible diagnosis outside of early paediatrics, the only area where it was historically diagnosed.
Many adults today are discovering they are Autistic. This is not diagnostic inflation. It is diagnostic correction.
Sensory Differences Are A part Of Autism
Another point raised in the interview was skepticism about tools such as ear defenders used by autistic children.
Sensory differences are not a peripheral feature of autism. They are part of the diagnostic criteria itself. Many autistic individuals experience hyper- or hypo-reactivity to sensory input, including sound. This makes sense from our understanding of autism which current evidence shows differences in neuronal excitability as a neurobiolgical difference.
For someone with auditory hypersensitivity, environments that seem ordinary to others (think classrooms, grocery stores, public transit) can be overwhelming. Noise-attenuating headphones, earplugs, and similar tools are simple accommodations that help individuals regulate sensory input.
The research base on specific devices is still emerging, but the broader literature on sensory processing differences in autism is extensive. The literature that does exist shows some classroom behaviour changes in children with access to ear defenders, a finding I have found in my clinical practice as well. Dismissing these accommodations minimizes the lived experience of autistic individuals.
And that brings us to the deeper issue.
Who Gets to Define Autism?
For decades, autism research was conducted almost entirely by non-autistic researchers studying autistic people. Autistic voices were largely absent from research design, interpretation, and policy discussions.
That is changing.
Today, autistic researchers, advocates, and clinicians are contributing to a more nuanced understanding of autism, one that incorporates lived experience alongside clinical research. This isn’t a rejection of earlier work but a continuation of scientific progress. When influential figures dismiss contemporary perspectives and autistic voices, it risks recreating the very exclusion the field is trying to move beyond.
With influence comes responsibility. Public commentary about autism should consider not only theoretical debates but also the impact those statements have on the community being discussed.
Science Evolves. So Should Our Models.
The cognitive models that shaped autism research in the late twentieth century were foundational. They gave researchers tools to study social cognition and developmental differences in new ways.
But the field has moved forward.
Today’s research includes:
sensory processing differences
masking and camouflaging behaviours
intersectional diagnostic disparities
dimensional approaches to neurodevelopment
the role of environment and accommodation
These developments do not invalidate earlier work. They expand it. And they reinforce a simple truth: autism cannot be reduced to a single presentation or pathway.
It is, in fact, a spectrum.
Moving the Conversation Forward
Look, disagreement in science is healthy. We absolutely should question our models and refine them as evidence grows. But those conversations must include the people most affected by the conclusions. Autistic individuals have spent decades being misunderstood, misdiagnosed, or excluded from discussions about their own experiences. The responsibility of researchers and clinicians today is not only to advance knowledge, but also to ensure that knowledge is shaped by evidence and by the voices of those living the reality we are trying to understand.
If you’re wanting a clearer, evidence-informed understanding of how autism is being understood today, I’m hosting a live webinar later this month where I’ll walk through the research, what’s changed, and what clinicians and families need to know.