Swedish study shows autism equally common in girls and boys by adolescence
A study recently published in the British Medical Journal tracked 2.7 million people born in Sweden between 1985 and 2022. Researchers from the Karolinska Institutet analysed national registers covering diagnosis rates from birth through to adulthood.
The findings: by adolescence, autism diagnosis rates between girls and boys approach a one-to-one ratio.
For decades, autism has been understood as occurring at a 4:1 male-to-female ratio. This statistic appears in clinical guidance, research papers, neurodiversity statistics, funding allocation decisions, and public understanding of autism as predominantly affecting boys.
The Swedish data shows that ratio was an artifact of diagnostic exclusion, not biological reality.
Girls were diagnosed at an average age of 14.3 years. Boys received diagnoses during childhood. The gap wasn’t about autism emerging later in girls — it was about assessment tools systematically missing female presentation until adolescence, when masking becomes unsustainable and girls finally meet male-normed diagnostic thresholds.
The researchers noted that “the male-to-female ratio for autism has decreased over time and with increasing age at diagnosis. This male-to-female ratio may be substantially lower than previously thought.”
Translation: we weren’t measuring autism prevalence accurately. We were measuring how well diagnostic criteria recognised male autistic presentation whilst systematically excluding girls who didn’t present identically.
The 4:1 autistic boys-girls ratio was diagnostic bias, not biological reality
Diagnostic criteria were developed and normed primarily on male autistic children. Assessment tools operationalised autism as it typically manifests in boys — direct social communication difficulties, obvious repetitive behaviours, and narrow interests that align with stereotypically male domains.
Female autistic presentation often differs. Not because autism itself differs by sex, but because social conditioning produces different compensatory strategies. Girls are socialised more intensively toward social compliance, emotional regulation performance, and relationship maintenance. Autistic girls develop sophisticated masking and camouflaging strategies earlier and more comprehensively than boys.
These strategies make them appear “less autistic” through the lens of male-normed standards of cantankerousness.
An autistic girl who maintains eye contact through learned script, participates in social groups through mimicry, and restricts her special interests to socially acceptable domains like animals or books doesn’t trigger the diagnostic red flags built around male presentation. But the neurological reality is that of a neurodivergent individual working significantly harder to achieve superficially neurotypical performance, but the assessment framework reads this as “not autistic enough.”
The medical model mistake: assuming male presentation equals universal autism presentation. Then pathologising anyone who doesn’t conform to that template as either not autistic or “mildly” autistic.
The Swedish data proves what autistic women have been saying for decades. The autism was always there. The diagnostic tools just refused to recognise it.
The fourteen year diagnostic delay, and its implications for girls
The average diagnostic delay for girls is 14.3 years.
That’s fourteen years of “something’s wrong with me” without a framework to understand it. Fourteen years of internalising dysfunction rather than recognising neurological difference. Fourteen years locked out of support systems because you don’t have the diagnostic label required to access them.
I was diagnosed autistic at 21. Not fourteen, but the mechanism is identical regardless of my biology — years of knowing I didn’t fit, years of failed attempts to perform neurotypicality, years of accumulating trauma from environments fundamentally incompatible with my neurology.
The harm isn’t just delayed access to formal support. It’s the psychological damage of believing you’re broken when you’re actually just different from the diagnostic template that excluded you.
For girls, this inevitably compounds. Puberty intensifies social expectations and peer relationship complexity precisely when masking becomes cognitively unsustainable. The autistic girl who managed primary school through mimicry and script-following hits secondary school and the system collapses.
That’s when diagnosis rates “catch up” — not because autism emerged, but because the girl finally can’t maintain the performance that kept her invisible to male-normed assessment.
The Swedish researchers found that over the 35-year study period, autism was diagnosed in 1.1% of individuals assigned male at birth and 0.8% assigned female. Those percentages converge as diagnostic age increases, demonstrating systematic catch-up during adolescence.
But catch-up for whom?
"Catch-up effect" language reveals the fundamental framing error
The research paper describes a “catch-up effect” — girls’ diagnosis rates catching up to boys’ during adolescence.
This language reveals the assumption embedded in the analysis. Girls are catching up. As if they were behind. As if autism developed later in them and they’re only now reaching the level boys achieved earlier.
What actually caught up: diagnostic tools to female autistic presentation.
The “catch-up” isn’t biological. It’s not girls developing autism during adolescence. It’s diagnosticians finally recognising autism they systematically excluded during childhood by using assessment criteria that operationalised male presentation as universal standard.
Dr Rachel Moseley, principal academic in psychology at Bournemouth University, noted that “this paper is timely and provides important support for what autism researchers have known for a long time: that autism is markedly underdiagnosed in people assigned female at birth.”
The research doesn’t prove autism is equally common. It proves autism was always equally common, and assessment tools were systematically biased toward recognising it in males whilst excluding it in females who presented differently.
Which, of course, is something autistic females have always known. So the only “catch up” here is surely those who were slow to know this.
When the researchers write that the male-to-female ratio “may be substantially lower than previously thought,” they’re documenting diagnostic failure at scale. Not changing biology. Not new autism emerging. Just finally measuring what was always there but remained invisible to tools calibrated exclusively for male presentation.
The 4:1 ratio was never about autism. It was about diagnostic exclusion masquerading as epidemiology. And girls spent/spend fourteen years, on average, without understanding their own neurology because of it.
