Masking is two kinds of work

If you’ve read anything about autistic masking in the last few years, you’ve read this paper second-hand. It’s the study that gave camouflaging a working definition, and it’s the one most writing on the subject traces back to, including ours.

That makes it the right place to start. The model it produced holds up. What gets attached to it often doesn’t, and the difference matters if you’re trying to understand your own experience rather than a headline.

By 2017, clinicians and autistic adults had been describing camouflaging for years. Autistic women in particular reported hiding their traits well enough to be missed by diagnosis. But the research literature had no agreed description of what camouflaging was, why people did it, or what it did to them, and almost everything written about it was about women and girls. The study set out to build a model from autistic adults’ own accounts, across genders, that later research could test.

Laura Hull and colleagues at University College London and Cambridge recruited 92 autistic adults through a Cambridge research database and adverts on social media. Fifty-five were women, 30 were men, and 7 described their gender another way. Ages ran from 18 to 79, and the average age at diagnosis was in the mid to late thirties, so most had spent much of their adult life undiagnosed. Everyone reported a clinical diagnosis; three people who had self-identified were excluded.

Participants filled in an online questionnaire with 43 questions, 20 of them open-ended: whether they camouflaged, why, what it involved, and what happened as a result. Six people said they had never camouflaged, and their comments were kept in. The team analysed the written answers thematically, looking for patterns across the responses rather than counting anything, and checked the themes back with six participants.

Two things about that design matter for everything that follows. The answers were written, not spoken in interviews, so there was no follow-up probing. And the study was advertised as being about “coping behaviours in social situations”, so the sample leans towards people for whom camouflaging was a live enough experience to write about.

The responses fell into seven themes across three stages, which is the part of the paper that travelled. The theme names are the participants’ own words.

Motivations. People camouflaged to “hide in plain sight”: to fit in, avoid bullying and worse, and get and keep jobs. And they camouflaged “to know and be known”: to get past the first awkward stages of a friendship or relationship so a real connection could follow. These weren’t alternatives. Many participants described both.

What camouflaging involves. Two kinds of work. Masking is hiding: suppressing stims, controlling reactions to sensory overload, presenting a persona or a character, and copying how other people talk and move. Compensation is performing: forcing eye contact, producing facial expressions to order, following rules for conversation such as asking questions to keep the spotlight off yourself, and preparing scripts in advance.

Consequences. Three themes. “I fall to pieces”: exhaustion, described as mental, physical, and emotional, followed by anxiety and replaying the conversation afterwards. Many needed time alone to recover. “People have a stereotyped view”: camouflaging well got people jobs and relationships, but it also meant being told they couldn’t be autistic, having a diagnosis delayed or doubted, and being handed more than they could carry because nobody could see the effort. “I’m not my true self”: feeling like a liar, relationships that felt false, and, for some, losing track of which parts of themselves were the performance.

Roughly equal proportions of the men and women said they camouflaged, and the authors found no consistent differences between them. The one gendered pattern they noted was that the minority who felt relieved or satisfied afterwards were mostly men.

Not how common camouflaging is. The study was advertised as being about social coping, and the authors say themselves that this probably drew people who camouflage. The sample can describe the experience. It can’t estimate how many autistic adults share it.

Not whether women camouflage more. The study wasn’t designed to test that, and no statistical comparison was made. The sample had more women than men, which reflects who signed up. Measuring the amount needed a validated questionnaire, and this study is what the questionnaire was later built from.

Not that camouflaging causes burnout, anxiety, or depression. Participants described exhaustion. The study also collected measures of depression, social anxiety, and quality of life, but didn’t analyse them in this paper, and nobody was followed over time or compared with a control group. The link between camouflaging and mental health has been studied since, with mixed and mostly correlational results. This paper is the origin of the hypothesis, not evidence for it.

Not a picture of autistic adults in general. Participants could complete a long written survey in English, were connected to research or community networks, and were mostly diagnosed as adults. The authors note the sample can’t speak for autistic people with intellectual disability, or for anyone who can’t express themselves in written English.

Not a study of ADHD masking. The paper is about autistic adults. ADHD masking as a distinct construct hasn’t had its own study of this kind.

The paper’s real contribution is vocabulary. Masking and compensation as two distinct kinds of work, and exhaustion, disbelief, and a lost sense of self as three distinct costs, is a model that holds up because autistic people recognise it. It came from their words.

One thing to keep straight: the paper uses “camouflaging” for the whole and “masking” for one half of it. Most people now say masking for all of it, and so do we. When you see the paper cited, that’s the shift you’re watching.

The safest way to carry it is as description, not proof. “Autistic adults describe masking as exhausting and identity-eroding” is a sentence this paper supports. “Research shows masking causes burnout” is not, and the paper never claims it.

The framing is worth noticing, too. The paper sits inside a clinical frame: camouflaging is partly of interest because it hides autism from diagnosis. But the authors end somewhere else, arguing that outcomes “can more fundamentally rely on how the social contexts treat” autistic people, and that the answer is a better fit between person and environment. Read that way, it’s a study of what non-autistic environments demand. That’s the reading we hold.

It represents autistic adults who camouflage, know they do, and can describe it in writing: mostly British, mostly diagnosed in adulthood, with a high proportion of women and non-binary people, which the authors count as a strength given how under-represented those groups had been. It doesn’t represent autistic people with intellectual disability, people who don’t identify camouflaging in themselves, undiagnosed people, or non-English-speaking communities.

Longitudinal work: does the amount of camouflaging predict later burnout, once you account for the environment people are camouflaging in? Studies that measure the environment rather than only the person. And work on whether reducing the demand to camouflage, rather than teaching people to camouflage less, changes the costs.

Some of this has started. The same team built a questionnaire from this model two years later, and a 2021 systematic review by Cook and colleagues pulls together what followed. That review is the better citation for anyone making a claim about camouflaging and mental health.