You got your diagnosis in 2022, or 2023, or last year, and somewhere in the relief there was also a smaller, quieter thought: why is everyone I know getting diagnosed at the same time as me. It wasn't your imagination, and it wasn't a trend you happened to fall into by coincidence. There is now a number attached to what you noticed, and it is bigger than most people assumed.
A 2026 study of 171,134 patients in a US healthcare system found that new autism diagnoses among women rose by an average of 19% per year after 2020, while diagnoses among men stayed flat or declined over the same period, across children, adolescents and adults alike. The researchers concluded this reflects improved recognition of autism presentations that have historically been missed in women, not an actual increase in how many people are autistic. Women in the study were also still diagnosed a full 3.4 years later than men on average, age 15.7 versus 12.3.
What the research shows
- Across 171,134 patients in a US health system from 2016 to 2024, autism diagnoses among females rose at an annual rate of 19.0% after 2020, while diagnoses among males remained stable or declined over the same period. Messias et al. (2026)1
- Women received an autism diagnosis a mean of 3.4 years later than men (age 15.7 versus 12.3), a gap that persisted even as overall diagnosis rates shifted. Messias et al. (2026)1
- A systematic review and meta-analysis found that diagnostic tools historically validated on predominantly male samples under-detect autism in women, partly because women use significantly more camouflaging and compensation strategies that conceal traits during clinical assessment. Cruz et al. (2024)2
- Research on autism-related content on TikTok found the platform functions as a space where people, especially those who had not previously considered an autism diagnosis, discover the possibility and refine their self-understanding through exposure to others' experiences. Alper et al. (2025)3
The Number Behind What You Already Noticed
In 2026, researchers at Saint Louis University published an analysis of 171,134 patient records spanning 2016 to 2024 inside a single US healthcare system, tracking exactly when autism diagnoses happened, and to whom1. The pattern that emerged was not subtle. Diagnoses among men held roughly steady across the whole eight-year window, sometimes even ticking downward. Diagnoses among women, by contrast, climbed sharply after 2020, at an average pace of 19% a year, across every age band the researchers looked at: children, teenagers, and adults.
Nineteen percent a year, compounding, is not a gentle drift. It is the kind of number that shows up as a genuinely different world in the space of five years, and it lines up almost exactly with what has been visible anecdotally for a while: waiting lists full of adult women, community spaces filling with people diagnosed well into their thirties, forties and fifties, and a very specific, very common experience of realising you are far from the only one who missed this for decades.
Why the Line Moved in 2020, Specifically
The researchers were careful about what this trend does and does not mean. Their conclusion was that the increase reflects better recognition of autism presentations that clinicians have historically missed in women, not a genuine rise in how many women are autistic1. The autistic population did not suddenly grow. The visible, diagnosed population did, because more of the people who had always been there finally got seen.
What changed in 2020 is not mysterious. It was the year most of the world spent an unprecedented amount of time on video calls, indoors, off normal routines, and online. For a lot of women, that disruption removed some of the scaffolding that had been quietly holding a masked presentation together for years, workplace routines, social obligations, the constant low-level performance of being fine, and left more room to notice what was underneath. It was also the year autism content on platforms like TikTok genuinely took off, turning what used to be a rare, difficult-to-access diagnostic conversation into something people could encounter by accident, in their own words, from other people who sounded exactly like them3.
The Bias That Made This Take So Long
None of this would have been necessary if the diagnostic process had worked evenly in the first place. A systematic review of the evidence found a clear and persistent bias toward diagnosing autism in males, built into tools that were validated on predominantly male samples decades ago and never adequately corrected2. Women showing comparable underlying symptom severity were consistently less likely to be flagged, in large part because they used far more camouflaging and compensation strategies, learned early and practised for decades, that made their traits harder for a clinical checklist to catch.
This is the piece that turns a statistic into something personal. The 3.4-year gap in average diagnosis age is not a rounding error, it is the accumulated cost of tools that were never built to see you clearly in the first place1. If you spent years being told you were "too social" or "too empathetic" to be autistic, that assessment was working exactly as designed, and the design was the problem.
What the Data Doesn’t Say
It is worth being precise here, because this kind of finding gets misused in both directions. The data does not say autism is being overdiagnosed in women, or that the surge represents a fad, a trend, or a wave of inaccurate self-identification. The researchers' own reading is the opposite: this is decades of underdiagnosis finally correcting, not new error being introduced1. The number of autistic women in the world has not changed. The number who have finally been able to get a name for it has.
It is also worth naming what the study cannot tell you: it cannot tell you whether your own diagnosis is "part of a trend" in any way that makes it less real. A pattern in population-level data describes millions of separate, individual, entirely valid diagnoses. You are not a statistic borrowing validity from other people's timing. You are one of the people the old system was built to miss, diagnosed in the specific window when the system finally started catching up.
“When I saw the number of women diagnosed the same year as me, I felt something shift that I didn’t expect. Not relief exactly. More like proof. All those years of being told it was probably anxiety, or a personality thing, and it turns out there was a very good reason nobody caught it. The reason wasn’t me.”
— Autistic adult, HeyASD community
What Finally Getting the Right Name Changes
For a lot of women diagnosed since 2020, the diagnosis itself was less of a discovery and more of a confirmation with a date attached. The traits were always there. What changed was access to a framework that finally matched them, built by better clinical awareness, by a female autism phenotype that is finally being studied properly, and, more informally, by a wave of other women online describing an inner experience that suddenly sounded exactly like yours.
That confirmation opens the door to something the years before it rarely allowed: understanding the specific cost of decades spent masking a presentation nobody was looking for, and starting the much longer process of figuring out who you are underneath it.
This exact moment, the one where a late diagnosis reframes decades of being told you were "too something" to be autistic, is the starting point The Unmasking Years was written from.
Key points
- A 2026 study of 171,134 patients found autism diagnoses in women rose 19% a year after 2020, while diagnoses in men stayed flat or declined.
- Women were still diagnosed 3.4 years later than men on average (15.7 versus 12.3), even during the surge.
- Researchers attribute the rise to better recognition of presentations long missed in women, not a genuine increase in how many women are autistic.
- Diagnostic tools were historically validated on predominantly male samples and consistently under-detect women who camouflage their traits.
- Post-2020 disruption to routines and the rise of autism content on platforms like TikTok both plausibly contributed to more women recognising themselves.
- The trend describes a system correcting itself, not a wave of inaccurate diagnoses. Your own diagnosis does not borrow or lose validity from population-level timing.
Questions about the rise in autism diagnoses among women
Why have autism diagnoses in women increased so much since 2020?
A 2026 study of 171,134 patients found diagnoses among women rose 19% a year after 2020 while male diagnosis rates stayed flat. Researchers attributed this to improved recognition of autism presentations long missed in women, driven partly by pandemic-era disruption to routines and partly by a surge in online autism content that helped many women recognise themselves for the first time.
Does this mean autism is overdiagnosed in women now?
No. The researchers behind the study were explicit that the increase reflects correction of decades of underdiagnosis, not new diagnostic error. Diagnostic tools were historically built around male presentations, so women with genuine autism were consistently missed. The surge represents more accurate identification, not inflated numbers.
Why are women still diagnosed later than men on average?
Even during the recent surge, women were diagnosed a mean of 3.4 years later than men (15.7 versus 12.3). This gap persists because diagnostic tools remain better calibrated to traits and presentations more common in men, and because camouflaging, learned and practised over years, makes autistic traits harder for a standard clinical assessment to catch in women.
What role did the COVID-19 pandemic play in rising diagnoses?
The pandemic removed routine social and workplace scaffolding for many people, which for some women reduced the constant effort of masking enough to notice what had been underneath it for years. This is a plausible contributing factor rather than a proven single cause, and it sits alongside greater diagnostic awareness and increased online visibility of autism as likely contributors to the same trend.
Is social media responsible for the rise in autism diagnoses among women?
Research on autism content on platforms like TikTok found it functions as a genuine space for self-discovery, where people encounter others describing experiences that match their own and consider an autism diagnosis for the first time as a result. It is one plausible contributor among several, not proof that diagnoses are being driven by trend-following rather than genuine recognition.
Why were autism diagnostic tools biased toward men in the first place?
Early autism research and diagnostic criteria were developed using predominantly male samples, so the tools were calibrated to traits and presentations more typical in men. A systematic review found women showing comparable underlying symptom severity were consistently less likely to be diagnosed, largely because camouflaging behaviours common in women make traits harder for those tools to detect.
Does my diagnosis being part of a wider trend make it less valid?
No. A population-level statistic describes a pattern across millions of separate diagnoses, each one individually assessed and individually real. Being diagnosed in a year when many other women were also diagnosed does not mean you borrowed validity from a trend; it means you were one of many people the old diagnostic system was built to miss, identified in the window when recognition finally improved.
Is the diagnostic gap between men and women closing?
The gap in total diagnosis numbers is narrowing as female diagnoses rise while male diagnoses stay flat, but the age-at-diagnosis gap (3.4 years later for women) has not closed at the same pace. Closing that second gap will likely require diagnostic tools and clinician training that specifically account for how autism presents in women, not just a general rise in diagnosis rates.
If you’ve lived this.
Nobody has added anything here yet. You are welcome to be the first.