The costs of masking

In studies mostly of autistic adults without intellectual disability, those who report more camouflaging also report, on average, more anxiety and depression and lower well-being. A 2024 meta-analysis pooled 16 studies, and reviewed 6 more, with 5,897 autistic participants in all. It found moderate correlations between camouflaging and anxiety, depression and social anxiety, and a small negative correlation with well-being.¹ Earlier systematic reviews, and a 2026 scoping review of 48 studies, point the same way.^2,3,4 In 733 autistic children and adolescents aged 4 to 17, a larger gap between parent and clinician ratings went with more anxiety, depression and physical complaints.⁵

The meta-analysis's authors judged every study too small to detect small effects, and many used measures that had not been well validated. The strength of the link did not differ significantly with study quality, participants' age, or the share of women or of people with a clinical diagnosis.¹ Similar links have been found in non-autistic people.⁴ A critical review of 389 studies found the main questionnaire reliable but its validity mixed: its scores overlap with social anxiety and are not specific to autism, and late-diagnosed autistic women are overrepresented while people with intellectual or language disabilities or higher support needs are underrepresented.⁵¹ In 59 autistic young people aged 13 to 18, a self-report questionnaire and a discrepancy measure of camouflaging were only weakly related, and the link was not statistically significant.⁵⁴ More on these limits is in Measuring masking.

The link is not the same for everyone. In 352 Dutch autistic adults aged 30 to 84, it was strong in only a small subgroup and small or absent in most.⁶ In an online study of 305 autistic adults without intellectual disability, it was small once autistic traits and age were taken into account, and stronger for anxiety than for depression.⁷ In 113 autistic adults with social anxiety who were interested in anxiety treatment, camouflaging added nothing to the prediction of depression, distress or disability once social anxiety and social responsiveness were counted.⁸ In a Japanese sample of 210 adults, both very little and very much camouflaging went with poorer mental health, while in a UK sample of 305 the relationship ran in a straight line; both samples reported their own diagnoses.⁹ The scoping review found the link appeared stronger for assimilation, trying to fit in, than for the other parts of camouflaging.⁴ In 627 autistic adults, assimilation statistically carried the link between autistic traits and every well-being measure, in data collected at one time.¹⁰ Feeling connected to the autistic community went with better well-being in 196 autistic people, but did not weaken the link between masking and poorer well-being.¹¹

Which comes first is still unknown. Almost every study measured camouflaging and mental health at the same time, and qualitative accounts suggest the influence may run both ways.^4,12 The only study among those reviewed here that followed people over time tested 332 of the Dutch adults twice, about two years apart. People who camouflaged more at the start showed a small decrease in mental health difficulties, and people who camouflaged less a small increase. Earlier difficulties did not clearly predict later camouflaging. The authors found no strong evidence that camouflaging is followed by worse mental health.¹³ Two other studies fit an influence running from anxiety to camouflaging as well as the reverse. In 71 autistic adults given group therapy for social anxiety, without a control group, camouflaging did not change on average, but those whose social anxiety fell most also camouflaged less afterward.⁵² In 205 Taiwanese adolescents, 100 of them autistic, social anxiety and camouflaging each showed indirect links with depression through the other, and the link between camouflaging and depression was weaker in the autistic adolescents.⁵³ A 2025 commentary describes recent numerical findings as inconsistent.¹⁴

Among 164 autistic adults in an online survey, 99 of them women, camouflaging was one of three factors that went with higher suicidality scores, alongside self-injury and unmet support needs.¹⁶ In 180 university students, camouflaging went with feelings of defeat and entrapment, but did not add to lifetime suicidal thoughts once those feelings were counted; statistically, it lay on the path from autistic traits to suicidal thoughts.¹⁷ In 160 undergraduates, 87% women and not selected for an autism diagnosis, camouflaging went with feeling that one does not belong and with lifetime suicidality.¹⁸ These three studies come from one research group.

Two independent studies point the same way. In 58 women with autistic traits, most without a diagnosis, camouflaging went with distress and, at above-average levels, with suicidal thoughts.¹⁹ In 471 women living in Spain, 72 of them autistic, the assimilation part of camouflaging went with suicidal thoughts, measured by a single questionnaire item, and anxiety statistically accounted for part of the link.²⁰ In open answers to an online survey that also included non-autistic people, only autistic respondents described masking leading to suicidal thoughts.²¹

Exhaustion comes up repeatedly when autistic adults describe camouflaging.^22,23 In one survey of 277 autistic adults, some of them self-identified, the researchers concluded from open answers that the amount of time spent camouflaging seemed most harmful to mental health.²³ In 262 autistic people, camouflaging in many settings, and switching between settings, both went with poorer mental health.²⁴ Ten autistic adults in Australia said that camouflaging in job interviews took a lot of effort and brought stress, anxiety and exhaustion.²⁵ The tiredness is not unique to autistic people: in the survey that included non-autistic respondents, they too described masking as exhausting.²¹

Autistic adults describe autistic burnout as long-lasting exhaustion, loss of skills and lower tolerance of sensory input. In one study, doing things in an autistic way, including unmasking, was among what they said helped recovery.²⁶ A 2025 systematic review of 48 studies found camouflaging among the contributors to burnout that autistic people named.²⁷ In 92 autistic adults, burnout-exhaustion partly accounted for the link between camouflaging and depression, in data collected at one time.²⁸ In accounts from 11 autistic adults, many described "dual masking", hiding the signs of burnout as well as their autistic traits, which they said added strain, delayed recognition and undermined recovery.²⁹

Autistic adults often describe camouflaging as a threat to their sense of who they are. In a qualitative study of 92 autistic adults, its consequences included threats to self-perception.²² In the 277-person survey, the dangers people named included "loss of identity and acceptance of self".²³ A review of 58 studies, mostly qualitative and mostly of white, female, late-diagnosed adults, listed low self-esteem and identity confusion among its consequences.³ Eleven autistic adults in Singapore recalled camouflaging growing more complex over the years, until for some it became part of who they were.³¹ After interviews and focus groups with 10 autistic adolescents who were female or gender-diverse, the authors concluded that camouflaging harmed the young people's understanding of themselves and their well-being.³²

Studies with numbers point the same way, at a single time point. In 342 autistic adults recruited through autistic social media, more masking went with lower self-esteem, lower authenticity, more anxiety and depression, more reported past interpersonal trauma, and less involvement in the autistic community.³³ In 463 university students not selected for autism, camouflaging and a less clear sense of self statistically linked autistic traits to depressive symptoms.³⁴ The feeling is not confined to autistic people: in the survey that included non-autistic respondents, people in every group said masking left them disconnected from their true sense of identity.²¹

Masking can work, at least in the short term. In the 277-person survey, the positive aspects autistic adults named included greater access to social spaces and protection from harm, and many saw camouflaging as necessary to survive in a world built for non-autistic people.²³ A review of 58 studies described it, among its consequences, as a practical way of exerting agency and control.³ In a UK survey of 285 autistic, 88 other neurodivergent and 99 neurotypical adults, people in all three groups saw masking at work as an adaptive response, usually used to guard against harm to their social standing or employment.³⁵

Studies of observed behavior show the same double edge. In a study of 161 autistic young people aged 10 to 16, mostly boys, a comparison of about 74 of them found that those whose observed autistic traits were low despite poor theory-of-mind test scores showed more vocal expressiveness and better rapport in a structured social interaction than those poor on both.³⁶ In an earlier study of 136 autistic adolescents grouped the same way, the better-compensating group also reported more anxiety.³⁷ A 2025 commentary proposes that camouflaging may also help people get jobs and form relationships.¹⁴

The gains have limits. A review of eight qualitative studies found camouflaging was a main way autistic adolescents and adults pursued friendship. For most it was exhausting and often ineffective, and friendships made this way often lacked closeness.³⁸

One author's count of more than 120 imaging studies of the social brain in autism found that 45% included only males and another 24% only one or two females.³⁹

In one UK group of adults without intellectual disability, autistic women's brain responses during tasks about the self and about other people's minds did not differ from non-autistic women's, while autistic men's responses were lower than non-autistic men's. In the 28 autistic women, more camouflaging, measured by the discrepancy method, went with a stronger response in the ventromedial prefrontal cortex when thinking about themselves.⁴⁰ In the same group, a resting-state measure proposed, on the basis of modeling and mouse experiments, as a proxy for excitation-inhibition balance in the medial prefrontal cortex was atypical in autistic men but not in autistic women, and a more typical value went with more camouflaging in the 25 autistic women only.⁴¹ Neither finding has been repeated by another group. These two studies and the first discrepancy study all come from the same research group and overlapping groups of adults, so they count as one line of evidence.⁴²

A US study of 45 autistic adults without intellectual disability found resting-state connectivity patterns linked to camouflaging scores that differed between women and men. The analyses were exploratory, the sample was small once split by sex, and the finding has not been repeated.⁴³ In 110 Chinese university students with no reported diagnosis, among those with more autistic traits, more camouflaging went with stronger connectivity between two large brain networks, the default mode network and the cingulo-opercular network. That connectivity statistically accounted for part of the link between camouflaging and depression and anxiety scores. The study measured everything once, so this mediation does not show cause.⁴⁴

Two EEG studies looked at responses to faces. In 44 adolescents, 17 with an autism diagnosis and mostly boys, "passing" (a high clinician-rated autism score alongside low teacher ratings) went with faster early brain responses to faces and, for subtle expressions only, smaller later emotional responses. The effects were small.⁴⁵ In 41 autistic adults with average to high IQ, split into smaller groups of women and men, autistic women's automatic early responses to faces fell between those of autistic men and non-autistic women. That study did not measure camouflaging, and its authors use the finding to urge caution about reading autistic women's social skills as camouflaging.⁴⁶

It has been proposed, from autistic people's accounts and clinicians' observations, that camouflaging takes considerable mental effort, which could add to stress, anxiety and depression.⁴² One study offers indirect support. In 136 autistic adolescents, those whose social behavior, as observed by a clinician, was good despite poor theory-of-mind test scores had higher IQ, better executive function and more self-reported anxiety than those poor on both.³⁷ A synthesis of 13 qualitative studies found that the strategies that seemed most "successful" involved heavy self-monitoring, were very mentally demanding or very habitual, and seemed to affect mental health more.⁴⁷ Other accounts place more of the cost outside the person: masking as a response to stigma, and camouflaging as impression management that depends on the setting and the people present, where feeling compelled to manage impressions takes a toll.^48,49 On these accounts, the effort would vary with how accepting the setting is.

One way to picture the proposal starts from what autistic adults in one online survey described: rules for conversation prepared ahead of time, and scripts.^22,50 If reading a face or choosing a reply in a non-autistic style is worked out step by step, like doing sums by hand rather than knowing the answer, it would take conscious effort, and that could contribute to the tiredness people report. Non-autistic people also describe masking as tiring, which an autism-specific account of effort would not explain.²¹ This is an illustration of the effort proposal, not a finding, and it sits alongside the stigma account rather than replacing it. No study found as of October 2026 has measured it during real conversations.

Autistic writers describe the cost first as tiredness. Megan Griffith compares masking to a "customer service voice" and adds, "you have to use it all the time, and it's utterly exhausting."^c1 Sol Smith writes that "the feedback loop of anticipating an interaction, rehearsing it, watching it closely, and then reviewing it will eventually break you down."^c2 Orion Kelly, an autistic writer, calls masking "physically, emotionally, and mentally exhausting".^c11 Devon Price argues that masking contributes to burnout, depression, anxiety and suicidal thoughts.^c3 The research above finds links with all of these, but not, so far, which comes first.

Several writers describe a cost that others do not see. Ralph G., an autistic contributor to a clinicians' guide, writes of the "vulnerability behind the camouflage", naming "The burnout, the headaches, the crying, the disassociation, the pain, the exhaustion."^c12 Georgina, writing in Luke Beardon's What Works for Autistic Adults, describes a cycle of masking to get a job, burning out in it, losing it and being sent back to training: "Repeat until I can no longer face the world."^c13 Daniel Burford, a late-diagnosed autistic writer, looks back on decades of masking: "The cost was burnout that you thought was just 'being an adult.'"^c14

Many write about losing track of who they are. Megan Anna Neff writes that "Masking can result in a diffuse sense of self, making it challenging to understand who you truly are."^c4 Price, describing the years before learning they were autistic, writes: "I was all defense mechanisms, with nothing left inside worth defending."^c5 Smith recalls that with age "my mask became less and less comfortable, and I increasingly missed some identity I never had."^c6 Lynne Soraya, who learned to pass so well that even family did not see the strain, writes that "you can never be sure that people who claim to like you really do."^c15 Luke Beardon, a senior lecturer in autism, grants that masking can help people get by in the short term, but warns that "Ongoing masking can eventually lead you to lose track of who you are".^c16 Griffith offers a gentler reading of the same discovery: "What if your masked self is just another facet of who you are?"^c7

Writers also describe masking hiding distress from others and from themselves. Price writes, "We conceal our suffering as part of our masking, and so we may be unable to admit to ourselves when we are having a difficult time."^c8 Amy Pearson, an autistic psychologist, and Kieran Rose, an autistic trainer and researcher, propose a reason this can happen. Many autistic people find their own emotions hard to identify (see Emotions and empathy on alexithymia), and masking pulls a person further from their internal states, so stress may build until it is "only noticeable once it is too late to prevent severe outcomes", such as burnout.^c17

Few writers call masking simply bad. Griffith calls it "sort of a double-edged sword" that can open opportunities but takes a psychological, and sometimes physical, toll.^c1 Price writes that "Sometimes, masking is necessary. But the fact that it is necessary does not make it good."^c9 Sarah Nannery calls masking in itself "just a tool", and places the harm in the pressure to fit everyone to one mold.^c10