Continuity,Change,and Psychosocial Predictors of Camouflaging in Autistic Adults: A Three-Time Point Follow-Up Study Over 2 Years

Abstract

Background:

Many autistic people camouflage their autistic traits to fit into non-autistic environments. While psychosocial factors influence camouflaging, its development and relationship with these factors over time remain unclear due to limited prospective follow-up studies. This study examined (a) continuity and change in camouflaging over 2 years, and (b) whether sociocultural predictors at Time 1 (T1: perceived stigma, vulnerability events, cultural tightness-looseness, and independent and interdependent self-construal) and psychological factors at Time 2 (T2: fear of negative evaluation [FNE], self-esteem, and autistic identity) predicted camouflaging at Time 3 (T3).

Methods:

Participants completed self-report measures of camouflaging and psychosocial experiences online at three time points across 2 years. Of 231 T1 participants, 135 completed T2, and 104 completed T3. After data cleaning and addressing missing data, the final sample included 226 autistic adults (18–77 years; 60.6% female). We assessed continuity and change using mean scores, intraclass correlation coefficients (ICCs), and latent growth modeling, and conducted a path analysis of earlier (T1 and T2) psychosocial predictors of later (T3) camouflaging.

Results:

Camouflaging showed considerable stability in individual rankings (ICCs = 0.80–0.86) and at group mean levels, with growth modeling suggesting a slight overall decline. The path model of T1 sociocultural and T2 psychological factors explained 32.2% of the variance in T3 camouflaging. Earlier FNE, autistic identity, and self-construal variables predicted later camouflaging, directly or indirectly.

Conclusion:

This is the first prospective study examining camouflaging over 2 years. Findings highlight the overall stability of camouflaging in adulthood and the role of earlier psychosocial predictors in later camouflaging, emphasizing the need to address sociocultural and psychological pressures on autistic people. Future research could use ecological momentary assessment and longer term studies to explore camouflaging across different social contexts and life stages.

Community Brief

Why is this an important issue?

Many autistic people adapt their behaviors to fit into non-autistic social environments, a process called camouflaging. While this can help build social connections and avoid stigma, it can often contribute to mental health issues such as anxiety, depression, and burnout. Although previous research shows that various psychological and social factors relate to camouflaging, we still do not fully understand how camouflaging changes over time or the extent to which it is affected by these factors, as there are no prospective studies with multiple time points to date.

What was the purpose of this study?

This study explored how camouflaging behaviors change or remain stable over 2 years. It also looked at whether earlier sociocultural factors (such as perceived stigma, negative life events, cultural emphasis on conformity, and the desire to fit in or stand out) and psychological factors (such as fear of being negatively judged, self-esteem, and autistic identity) affect the extent to which individuals will engage in later camouflaging behaviors.

What did the researchers do?

We asked participants to complete online measures of camouflaging, well-being, and psychosocial experiences at three time points across 2 years. We analyzed data from 226 autistic adults (18–77 years; 60.6% female), to understand how camouflaging behaviors changed over time. We also looked at the relationships between psychosocial factors and camouflaging across three time points.

What were the results and conclusions of the study?

Camouflaging behaviors reported by participants were generally stable over 2 years. Earlier psychosocial factors, such as fear of negative judgment, autistic identity, and the desire to fit in or stand out, were linked to later camouflaging.

What is new or controversial about these findings?

This study is the first to track changes in camouflaging across three time points. It shows that autistic adults’ camouflaging behaviors remain relatively stable over 2 years. It also shows that earlier psychological and social pressures predict later camouflaging behaviors, emphasizing the need to address and reduce these pressures in autistic people’s lives.

What are potential weaknesses in the study?

Our investigation of stability in camouflaging across 2 years of adulthood may have missed substantial changes in camouflaging occurring over longer intervals or across significant developmental periods, such as adolescence. Also, participant demographics—mainly White, female, and highly educated—may limit how well the results apply to other groups of autistic people. Longer term studies could explore how camouflaging behaviors develop or change across different life stages and social contexts, as well as how they vary across cultures and among autistic people with diverse backgrounds.

How will these findings help autistic adults now or in the future?

The findings show that it is important to understand and address the psychological and social pressures related to camouflaging. By promoting acceptance and fostering supportive environments, autistic individuals can feel safer expressing their authentic selves without fear of judgment, which will likely improve their mental health and well-being.

Keywords

autism camouflaging masking psychosocial prospective

Many autistic individuals consciously or unconsciously adjust their autistic behaviors to fit within non-autistic social environments by adopting camouflaging ¹ or masking strategies.^1–2 These strategies involve hiding or suppressing autistic characteristics, adopting non-autistic personas, compensating for social differences, or blending in while concealing discomfort.^2–3 Camouflaging offers practical benefits, enabling autistic people to establish social connections and access life opportunities, and is necessary to defend against stigma for social survival.^4–5 However, camouflaging also often comes at a cost, as it is associated with various negative impacts on mental health and well-being, including depression, anxiety, suicidality, autistic burnout,² identity confusion, lower quality of life, and reduced access to diagnosis and supports.^6–9

Background

Continuity and change in camouflaging

While most camouflaging research has primarily focused on autistic adults, both qualitative and quantitative literature suggest that camouflaging behaviors occur across the lifespan, from childhood through adolescence and adulthood.^5,10–15 However, our understanding of how camouflaging may change over time remains limited. To our knowledge, all existing quantitative studies on camouflaging are cross-sectional, with no prospective longitudinal follow-up research available in the literature to date.

A retrospective qualitative study by Loo et al.¹³ provides significant insights into the developmental trajectories of camouflaging behaviors among autistic adults, revealing both continuity and change. Loo and colleagues interviewed 11 Singaporean autistic adults aged 22–45 to explore their recollections of how their camouflaging behaviors developed and evolved over time, shaped by their psychological and sociocultural experiences. Participants recalled beginning to camouflage their autistic traits in later childhood and adolescence and continuing to do so in adulthood, often motivated by concerns for safety, social acceptance, and pragmatic reasons. Supporting the notion of continuity in camouflaging, some autistic adults reported engaging in camouflaging for so long that it became an internalized, automatic, and habitual process that was challenging to alter or inhibit (see also other articles^1,4,14,16,17).

Conversely, Loo et al.¹³ highlighted changes in the pressures to engage in camouflaging, noting that some participants felt less pressure to camouflage as they gained self-acceptance and received acceptance from others regarding their autistic differences. This finding aligns with other studies in which autistic adults describe shifting away from camouflaging toward disclosure, self-advocacy, and/or seeking accommodations as they gain self-awareness, embrace their autistic identity, or recognize camouflaging’s negative impact on emotional well-being.^{4,9,16,18–22} In addition, some autistic adults reported that camouflaging becomes increasingly difficult with age and unsustainable over extended periods due to exhaustion and autistic burnout.^{2,4,5,14,23–27} Overall, these studies suggest that while camouflaging shows some elements of continuity across time, it may also change with development, experiences, and social contexts, influenced by psychological and social factors.

Psychosocial correlates of camouflaging

Various psychosocial experiences influence the extent to which autistic individuals camouflage. Ai et al.²⁸ proposed that camouflaging is a form of impression management that is dynamic, transactional, and context-dependent, motivated by the desire for belonging, social opportunities, and stigma mitigation. Supporting this, Zhuang et al.’s⁹ mixed-methods systematic review of 58 studies with predominantly autistic participants identified several psychosocial correlates of camouflaging, including social norms and pressures, social acceptance and rejection, and self-esteem and identity.

Building on Ai et al.²⁸ and Zhuang et al.,⁹ Zhuang et al.²⁹ analyzed self-report cross-sectional data from 225 autistic adults to examine an integrated model of psychosocial predictors of camouflaging (see Fig. 1 in Zhuang et al.²⁹), which included individual psychological factors (fear of negative evaluation [FNE], self-esteem, and autistic identity) and broader sociocultural predictors (perceived stigma, vulnerability events, cultural tightness-looseness, and independent and interdependent self-construal³). Zhuang et al.²⁹ hypothesized that greater camouflaging would be associated with (a) greater FNE, as camouflaging is often used to mitigate negative judgments^4,5,12,16; (b) lower self-esteem, reflecting feelings of shame and internalized stigma^5,13,14,30; (c) greater perceived autism stigma, with camouflaging acting as a defense mechanism against discrimination^28,31–34; (d) more vulnerability events, as individuals camouflage to protect themselves from further victimization^4,5,13,35; (e) greater cultural tightness contexts, where stronger social norms and less tolerance for deviation increase the pressure to conform³⁶; and (f) greater interdependent self-construal, reflecting a greater desire to fit in and assimilate with others.³⁷ Furthermore, Zhuang et al.²⁹ hypothesized that fewer camouflaging behaviors would be associated with (a) a stronger autistic identity, as individuals with greater acceptance of their autistic identity may be more likely to disclose it and thus feel less need to camouflage^5,16,20,38; and (b) greater independent self-construal, driven by a desire to express uniqueness and autonomy.³⁷ Finally, they examined whether the psychological factors (FNE, self-esteem, and autistic identity) mediated the relationships between the sociocultural factors and camouflaging.

FIG. 1.

Hypothesized model of earlier (T1 and T2) psychosocial predictors of later (T3) camouflaging. T1 = Time 1; T2 = Time 2; T3 = Time 3; indep SC = independent self-construal; interdep SC = interdependent self-construal; FNE = fear of negative evaluation.

Path analysis findings from Zhuang et al.²⁹ partially supported the hypothesized model, with the final model explaining about 25% of the variance in camouflaging. Greater vulnerability events and lower independent self-construal were directly associated with increased camouflaging. Sociocultural factors were indirectly linked to camouflaging through individual psychological factors. FNE, in particular, emerged as a strong and important predictor, mediating most relationships between sociocultural factors and camouflaging. However, contrary to initial hypotheses, higher self-esteem and stronger autistic identity were associated with increased camouflaging in the model.

Zhuang et al.²⁹ was the first study to examine an integrated, comprehensive model of psychosocial predictors of camouflaging. Nevertheless, its cross-sectional design limits the ability to draw conclusions about temporal sequences and the directionality of effects.^39–40 Furthermore, some findings, such as the positive associations of self-esteem and autistic identity with camouflaging, were contrary to expectations, warranting further investigation.

The present study

This study had two main aims. First, we examined continuity and change in camouflaging among autistic adults over 2 years. Given the lack of prospective studies on camouflaging and the mixed evidence from existing literature suggesting both continuity and change, this aim was exploratory. As such, we did not formulate specific hypotheses.

Second, we tested and extended Zhuang et al.’s²⁹ cross-sectional model of psychosocial predictors of camouflaging using prospective data collected at three time points. This allowed us to temporally sequence sociocultural factors, individual psychological factors, and camouflaging, thereby strengthening directional inferences in longitudinal mediation.^41–42 Specifically, we examined whether sociocultural factors at Time 1 (T1)—perceived stigma, vulnerability events, cultural tightness-looseness, and independent and interdependent self-construal—predicted camouflaging at Time 3 (T3), mediated by individual psychological factors at Time 2 (T2), namely FNE, self-esteem, and autistic identity (see Fig. 1).

Considering that camouflaging is often motivated by psychosocial reasons, it is important to understand whether and how earlier sociocultural and psychological predictors may influence later engagement in camouflaging. By examining these associations prospectively, we aimed to clarify the directionality of temporal influences and understand how psychological and sociocultural factors may interact to impact camouflaging over time. Investigating the psychosocial mechanisms underlying camouflaging could inform the development of targeted supports to reduce related stressors, thereby alleviating the pressure on autistic individuals to camouflage and potentially enhancing their mental well-being.

Methods

Participants and procedure

Ethical approval was obtained from the University of Western Australia Human Research Ethics Committee (reference number: 2021/ET000065). We initially recruited participants as part of a cross-sectional study²⁹ via social media, autism associations, and Prolific (https://prolific.co; a UK-based online research platform) between May and August 2021 (T1). Eligibility criteria included being at least 18 years old, able to complete an online survey in English, and having a professional diagnosis or self-identifying as autistic. We included self-identified autistic individuals to increase the representation of those facing obstacles to professional diagnoses^43–45 or who remain undiagnosed, potentially due to camouflaging.⁴⁶

Subsequently, we emailed invitations to T1 participants who had agreed to be contacted for future research, inviting them to complete two follow-up surveys: one between May and August 2022 (T2), and another between July and September 2023 (T3). The average intervals between successive time points were 11.6 months between T1 and T2 (SD = 0.9, range = 9–14 months) and 12.4 months between T2 and T3 (SD = 0.7, range = 11–15 months).

To safeguard anonymity while enabling follow-up, Prolific participants provided their Prolific-associated emails, which were used solely for follow-up invitations and were deleted after we matched responses across time points using Prolific IDs. For non-Prolific participants, personal emails were stored in a secure Qualtrics mailing list linked to a randomly generated login ID, then deleted after matching.

At all three time points, participants provided written informed consent and completed demographic questions and self-report measures via a Qualtrics-hosted survey (https://www.qualtrics.com). They could select “prefer not to say” for any question. After completing the survey, participants reviewed an online debrief form containing information on autism-specific services and community support organizations. Each participant received reimbursement worth AUD20 via direct bank transfer or online gift cards.

To reduce participant burden and fatigue given the number of measures at T1,²⁹ (a) we selected measures based on their relevance and brevity to balance data richness with minimizing fatigue; (b) participants could save their progress and return to the survey later, allowing flexible completion; and (c) we shortened the T2 and T3 follow-up surveys by excluding questions or measures considered to be relatively stable over time.

An initial 259 participants accessed our survey at T1. We excluded the data of 28 T1 participants⁴ due to the following: (a) incomplete or unsubmitted responses (n = 16), (b) not meeting inclusion criteria (n = 6), (c) a reCAPTCHA score of <0.5 (n = 3), (d) indicating that their data were invalid for research use (n = 1), and (e) participants selecting “prefer not to say” for over 20% of responses (n = 2; see also Zhuang et al.²⁹) Of the remaining 231 T1 participants, 135 completed T2 and 104 completed T3, with response rates of 58.4% and 45.0%, respectively. Furthermore, we excluded data from five participants due to inconsistencies in reporting demographic information across time points. All other participants at T2 and T3 passed our data checks.⁵ In the current study, we included all participants with valid T1 data, and after addressing attrition-related missing data (see the Data Analyses section), the final sample consisted of 226 autistic adults.

To assess potential attrition bias, we conducted independent-samples t-tests and chi-square tests comparing participants who completed only T1 with those who also completed T2 and T3 on key baseline characteristics (age, gender, autistic traits, camouflaging scores, and autism, mental health, and neurodevelopmental diagnoses). No statistically significant differences were found between retained and nonretained participants (all ps > 0.05), and all effect sizes were small (all |d| < 0.20; all |φ| < 0.13), suggesting that the two groups were broadly comparable.

At T1, the mean age of our participants was 35.8 years (SD = 11.9; range = 18–77). Most were female, White, employed, university-educated, and living in Australia (see Table 1 for participant characteristics). More than 70% reported having at least one co-occurring mental health condition in the last 5 years, most commonly anxiety or depression. Most reported a professional autism diagnosis (n = 184; 81.4%), with the mean age of diagnosis being 28.5 years (SD = 14.2, range 2–61), while 42 (18.6%) self-identified as autistic. There were no significant differences between professionally diagnosed and self-identified participants in autistic traits reported, Wald χ²(1) = 0.55; p = 0.46. In addition, 96.5% of participants had a Broad Autism Phenotype Questionnaire (BAPQ) total score above 3.15, the suggested screening cutoff score for this measure.⁴⁷

Table 1.

Participant Characteristics at Time 1 (N = 226)

Characteristics	N	%
Gender
Female	137	60.6
Male	64	28.3
Other genders^a	22	9.7
Preferred not to say	3	1.3
Ethnicity
White	195	86.3
Asian	13	5.8
Multiethnic	13	5.8
Black/African/Caribbean	4	1.8
Hispanic/Latino	1	0.4
Country of residence
Australia	128	56.6
The United Kingdom	43	19.0
The United States	28	12.4
New Zealand	8	3.5
Canada	8	3.5
Other European countries	8	3.5
Singapore, Hong Kong	3	1.3
Highest educational level
Undergraduate degree	79	35.0
Postgraduate degree	56	24.8
Secondary school	49	21.7
Trade/technical certificate	35	15.5
Primary school	3	1.3
Other qualifications (e.g., diploma, A-level certificate)	4	1.8
Employment status
Employed full-time	63	27.9
Employed part-time	52	23.0
Unemployed	41	18.1
Student	37	16.4
Homemaker/caregiver	18	8.0
Unpaid work (e.g., volunteering, unpaid internship)	8	3.5
Retired	6	2.7
Preferred not to say	1	0.4
Living arrangements
Living with partner and/or child(ren)	102	45.1
Living with parents/relative(s)	59	26.1
Living alone	42	18.6
Living with friend(s)/housemate(s)	21	9.3
Other	2	0.9
Other neurodevelopmental diagnoses
Attention-deficit/hyperactivity disorder	56	24.8
Specific learning disorder	17	7.5
Intellectual disability	3	1.3
Other (e.g., cerebral palsy, developmental coordination disorder)	3	1.3
Preferred not to say	6	2.7
Mental health diagnoses (in the last 5 years)
Anxiety	129	57.1
Depression	114	50.4
Post-traumatic stress disorder	28	12.4
Obsessive-compulsive disorder	13	5.8
Personality disorders	7	3.1
Bipolar disorder	5	2.2
Eating disorders	5	2.2
Other (e.g., psychotic disorders, dissociative disorders)	11	4.9
Preferred not to say	4	1.8

Other genders included nonbinary (n = 13), agender (n = 3), genderqueer (n = 2), gender neutral (n = 1), gender fluid (n = 1), gender curious (n = 1), and demifemale (n = 1).

Measures

We administered the same self-report measures at all three time points.

Autistic traits

The 36-item BAPQ⁴⁷ assesses autistic traits using a 6-point Likert-type scale (1 = very rarely, 6 = very often; mean item-total score range 1–6). The BAPQ has demonstrated strong reliability (α = 0.95–0.96) and validity, with good sensitivity and specificity in distinguishing between autistic and non-autistic individuals.^3,47,48

Camouflaging

The 25-item Camouflaging Autistic Traits Questionnaire (CAT-Q)³ assesses camouflaging behaviors on a 7-point Likert-type scale (1 = strongly disagree, 7 = strongly agree, range 25–175). It demonstrates a consistent factor structure across autistic and non-autistic adults, male and female, and has excellent internal consistency (α = 0.91–0.94).^3,49

Individual psychological factors

Fear of Negative Evaluation

The 8-item Brief Fear of Negative Evaluation-II Scale (BFNE-II)⁵⁰ assesses anxiety about negative judgment from others on a 4-point Likert-type scale (0 = not at all characteristic of me, 4 = extremely characteristic of me, range 0–32). The BFNE-II exhibits moderate convergent and discriminant validity, with strong internal consistency (α = 0.95–0.96) in both clinical and nonclinical groups.^50–51

Self-Esteem

The 10-item Rosenberg Self-Esteem Scale⁵² evaluates global self-worth on a 4-point Likert-type scale (1 = strongly disagree, 4 = strongly agree, range 10–40). It demonstrates excellent internal consistency (α = 0.90–0.92) in autistic adults.^53–55

Autistic Identity

The 14-item Multicomponent In-group Identification Scale (MIIS)⁵⁶ has been adapted to measure autism identification using a 7-point Likert-type scale (1 = strongly disagree; 7 = strongly agree; mean item-total score range 1–7). The adapted MIIS has good to excellent internal consistency (α = 0.84–0.91).^53,57–59 We used the terms “autistic individual/person/community” in our adaptation.

Proximal sociocultural factors

Perceived Stigma

The 5-item Perceived Stigma Measure⁶⁰ was originally developed to assess stigma associated with having Asperger’s syndrome. We updated the language by replacing “having Asperger’s syndrome” with “being autistic/on the autism spectrum.” We also modified the original “yes/no” format to a 5-point Likert-type scale (0 = never, 4 = very often, range 0–20) for a broader response range. The modified scale demonstrated good internal consistency in our study (see Table 2).

Table 2.

Descriptive Statistics, Reliabilities, and Intercorrelations Among Variables in Path Analyses (N = 226)

	Stigma (T1)	Vul events (T1)	Cul tightness (T1)	Indep SC (T1)	Interdep SC (T1)	FNE (T2)	Self-esteem (T2)	Autistic identity (T2)	Cam (T3)
Vul events (T1)	0.34***	—
Cul tightness (T1)	0.04	0.06	—
Indep SC (T1)	−0.11	0.04	0.10	—
Interdep SC (T1)	−0.05	−0.05	0.17**	−0.12	—
FNE (T2)	0.16*	0.09	0.08	−0.41***	0.25***	—
Self-esteem (T2)	−0.21**	−0.12	0.06	0.48***	−0.24***	−0.50***	—
Autistic identity (T2)	−0.09	0.01	0.08	0.26***	0.17*	−0.04	0.20**	—
Cam (T3)	0.07	0.13	0.16*	−0.37***	0.22**	0.42***	−0.26**	0.15	—
M	13.3	22.4	4.8	4.3	4.4	20.9	24.3	4.7	129.5
SD	4.5	10.2	0.8	0.8	0.8	9.0	6.7	1.1	21.0
Range	0–20	3–48	1–6	1.9–6.7	2.3–6.8	0–32	10–40	1.4–7	67–168
Cronbach’s alpha (α)	0.84	0.92	0.70^a	0.71	0.71	0.97	0.93	0.92	0.91

Vul events = vulnerability events; cul tightness = cultural tightness; indep SC = independent self-construal; interdep SC = interdependent self-construal; FNE = fear of negative evaluation; cam = camouflaging; *p < 0.05; **p < 0.01; ***p < 0.001; correlations ≥0.3 in bold.

Item 4 of the Tightness-Looseness Scale (TLS) (“people in this country have a great deal of freedom in deciding how they want to behave in most situations”) was removed due to its low item-total correlation (0.14), which improved Cronbach’s alpha from 0.64 to 0.70. As the only reverse-scored item, it may have introduced construct-unrelated method variance and was identified as having the lowest item loading in Gelfand et al.’s (2011) original validation.

Vulnerability Events

The 60-item Vulnerability Experiences Quotient (VEQ)⁶¹ measures the frequency of negative experiences across 10 life domains, including education, employment, finances, and victimization. Participants selected “yes” (= 1 for presence) or “no” (= 0 for absence) for each experience, with three social support items reverse-scored (range 0–60). The VEQ shows strong internal consistency (α = 0.89) among autistic adults.⁶¹

Distal sociocultural factors

Cultural Tightness-Looseness

The 6-item Tightness-Looseness Scale³⁶ evaluates the strength of social norms and tolerance for norm violations using a 6-point Likert-type scale (1 = strongly disagree, 6 = strongly agree, mean item-total score range 1–6). It has demonstrated strong internal consistency (α = 0.80–0.85) and validity in large cross-national studies.^36,62,63

Independent and Interdependent Self-Construal

The Self-Construal Scale⁶⁴ consists of two 12-item subscales measuring independent and interdependent self-construals, rated on a 7-point Likert-type scale (1 = strongly disagree, 7 = strongly agree, mean subscale score range 1–7). Both subscales demonstrate acceptable internal consistency (α = 0.69–0.81) in ethnically diverse student populations.^64–66

Data analyses

We used Mplus version 8.11 for all model-based analyses. We treated “prefer not to say” responses and responses lost due to T2/T3 attrition as missing, addressing them using a two-step approach. First, we applied person-mean imputation for item-level missingness, with a 50% cutoff for proration. Then, we addressed scale-level missingness using robust maximum likelihood estimation (MLR) or Bayesian estimation in Mplus, which includes all available data without biasing estimates or standard errors.⁶⁷ We used MLR in all analyses except for mediation analyses, where asymmetrical confidence intervals (CIs) for indirect effects are not available. For these analyses, we used Bayesian estimators to appropriately account for the asymmetrical distribution of the standard errors of the indirect effects.⁶⁸ Both estimators account for non-normality in variable distributions.

We assessed continuity and change in camouflaging in three ways. First, we conducted pairwise dependent-samples comparisons using Wald tests in Mplus to compare mean camouflaging scores across the three time points. Second, to evaluate the relative stability of individuals’ rankings of camouflaging behaviors across time points, we calculated intraclass correlation coefficients (ICCs) using SPSS version 30, based on a two-way mixed-effects model assessing consistency of single measurements. We also examined the relative stability of psychosocial variables by calculating ICCs using the same SPSS procedure. Finally, we used a latent growth model in Mplus to examine the sample’s mean camouflaging trajectory (i.e., slope) over time and assess variability in individual trajectories. We specified this model with linear slopes, designating T1 as the intercept and allowing the intercept and slope factors to correlate.

To investigate earlier psychosocial predictors of later camouflaging, we conducted prospective path analyses in Mplus. Building on Zhuang et al.’s²⁹ cross-sectional model, we specified sociocultural factors at T1, individual psychological factors at T2, and camouflaging at T3 (see Fig. 1). In line with our primary aim, we present in the Results section a model that did not control for baseline T1 variables, allowing for the examination of predictive associations between earlier psychosocial factors and later T3 camouflaging without adjusting for prior levels of any dependent variables. However, to complement the primary path analysis, we also conducted a secondary path analysis controlling for T1 camouflaging and T1 psychological variables (FNE, self-esteem, and autistic identity). This additional model, reported in the Supplementary Data (see Supplementary Fig. S1), provided a more conservative test of these associations by accounting for baseline camouflaging and psychosocial predictor levels and focusing on residual change over time.

We assessed model fit in all models using the comparative fit index (CFI ≥ 0.90 for acceptable fit; >0.95 for excellent fit), standardized root mean square residual (SRMR ≤ 0.08 for acceptable fit), and root-mean-square error of approximation (RMSEA ≤ 0.08 for acceptable fit; <0.06 for excellent fit).^69–70 For mediation models using Bayesian estimation, we also evaluated the posterior predictive p value, where values greater than 0.05 suggest acceptable fit, values close to 0.50 indicate excellent fit, and the associated 95% CI should include zero to indicate adequate fit.^68,71

Authors’ positionality

Our approach was influenced by our psychological training, academic and professional experiences, and, for some authors, lived experiences as autistic or neurodivergent individuals. Our research team has neurodiversity-affirming values and aligns with the social model of disability.

Community involvement

In testing and expanding Zhuang et al.’s²⁹ cross-sectional camouflaging model using prospective data, we engaged two autistic advisers who contributed to the initial model development and measure selection. In this study, both advisers continued to contribute to the formulation of research questions and the interpretation of findings, ensuring alignment with their firsthand experiences and perspectives on camouflaging within the autistic community. The advisers were compensated for their time and are coauthors of this work.

Results

Table 2 presents descriptive statistics, internal consistencies (ranging from acceptable to excellent; 0.70 ≤ α ≤ 0.97), and intercorrelations among variables in the path analyses. T1 cultural tightness, T1 interdependent self-construal, and T2 FNE were positively correlated with T3 camouflaging, while T1 independent self-construal and T2 self-esteem were negatively correlated with T3 camouflaging, with small- to medium-effect sizes.⁶

Most psychological and sociocultural factors demonstrated moderate to high stability over 2 years, with ICCs ranging from 0.73 to 0.94 across pairwise comparisons among the three time points (see Supplementary Data). Perceived stigma showed lower stability (ICCs = 0.42–0.62), while cultural tightness exhibited low and inconsistent stability, with some ICCs near zero and CIs overlapping zero.

Continuity and change in camouflaging

Mean camouflaging scores showed a slight decline over 2 years (T1 = 132.1, T2 = 130.7, T3 = 129.5), but pairwise Wald test comparisons showed that these differences were not statistically significant (T1 vs. T2: Wald χ²(1) = 0.72, p = 0.398; T2 vs. T3: Wald χ²(1) = 0.60, p = 0.440; T1 vs. T3: Wald χ²(1) = 1.57, p = 0.210). Similarly, ICCs between CAT-Q total scores were very large across the three time points: T1–T2, ICC = 0.86, 95% CI 0.81–0.90; T2–T3, ICC = 0.85, 95% CI 0.78–0.89; and T1–T3, ICC = 0.80, 95% CI 0.71–0.86, indicating substantial stability in individuals’ rankings of camouflaging behaviors over time. ICCs of CAT-Q total scores were also very large and comparable across gender and age groups.⁷

The linear latent growth model demonstrated an excellent fit to the data (χ²[1] = 0.01; p = 0.934; SRMR = 0; CFI = 1; RMSEA = 0 [90% CI, 0–0.05]). The mean slope was just statistically significant and negative, indicating an average linear decline in camouflaging across the three assessments (estimate = −1.21, SE = 0.62, p = 0.049). However, slope variability did not reach statistical significance (estimate = 32.69, SE = 17.76, p = 0.066), indicating no significant differences in camouflaging trajectories across participants. In addition, there was a strong, inverse correlation (r = −0.55) between the intercept and slope of camouflaging, indicating that individuals with higher initial camouflaging levels showed greater decline over time, consistent with regression to the mean.

Earlier psychosocial predictors of later camouflaging

The path analysis of T1 sociocultural and T2 individual psychological factors as predictors of T3 camouflaging, either directly or indirectly, demonstrated excellent fit (PPP = 0.50, 95% CI [−10–38.41]; CFI = 1, 90% CI [0.98–1]; RMSEA = 0, 90% CI [0–0.04]), explaining 32.2% of the variance in T3 camouflaging (see Fig. 2). Consistent with the study’s main aim, we did not control for earlier camouflaging or psychological predictors in this model to allow examination of the predictive associations of earlier psychosocial factors with later T3 camouflaging without adjusting for prior levels of either camouflaging or the psychological factors. This approach also enabled comparison with the cross-sectional path analyses reported in Zhuang et al.²⁹

FIG. 2.

Path model of T1 sociocultural predictors predicting T3 camouflaging via T2 psychological mediators. Solid lines indicate significant paths, while dashed lines indicate nonsignificant paths. Standardized coefficients with confidence intervals are shown only for significant paths. Values within circles represent the proportion of variance explained for each variable. T1 = Time 1; T2 = Time 2; T3 = Time 3; indep SC = independent self-construal; interdep SC = interdependent self-construal; FNE = fear of negative evaluation.

Consistent with the cross-sectional findings of Zhuang et al.,²⁹ higher T1 independent self-construal was directly associated with lower T3 camouflaging and indirectly through its association with lower T2 FNE. In addition, higher T1 independent and interdependent self-construal scores were associated with a stronger T2 autistic identity, which, in turn, was associated with higher T3 camouflaging.

Contrary to the cross-sectional findings, T1 perceived stigma, T1 vulnerability to negative life events, T1 cultural tightness, and T2 self-esteem were not significantly associated with T3 camouflaging either directly or indirectly. Greater T1 interdependent self-construal was also not associated with higher T3 camouflaging via T2 FNE.

To complement the primary path analysis, we conducted a secondary path analysis that retained the same T1 sociocultural and T2 psychological predictors as the main model and additionally controlled for T1 camouflaging and T1 psychological predictors (FNE, self-esteem, and autistic identity; see Supplementary Fig. S1). This more conservative model tested whether earlier psychosocial factors predicted T3 camouflaging above and beyond baseline levels of camouflaging and psychological predictors. In this model, most T1 predictors were no longer significant, likely reflecting the strong temporal stability of camouflaging and psychological factors, which leaves less residual variance for other predictors to explain. Notably, only T1 independent self-construal uniquely and significantly negatively predicted T3 camouflaging.

Discussion

To our knowledge, this study is the first to examine camouflaging in autistic adults using a prospective design with three time points. Building on the cross-sectional findings of Zhuang et al.,²⁹ we investigated continuity and change in camouflaging and its psychosocial predictors over 2 years. Our findings suggest considerable stability in camouflaging behaviors, with some evidence of a slight decline during this period. The path model of T1 sociocultural factors and T2 psychological mediators predicting T3 camouflaging demonstrated excellent fit and explained 32.2% of the variance in T3 camouflaging. While we replicated several of the cross-sectional pathways from Zhuang et al.,²⁹ others did not hold over time.

Overall stability in camouflaging over 2 years

Our findings indicate substantial continuity in camouflaging over 2 years, both in individual rankings and group mean levels, as evidenced by strong ICCs and nonsignificant differences in mean scores across time points. We also observed a slight average linear decline in camouflaging using growth modeling, which was relatively consistent across individuals. This overall pattern—characterized by substantial continuity alongside some evidence of a small decline—aligns with the limited existing literature to date, highlighting both stability and change in camouflaging.¹³

The stability of camouflaging among autistic adults may be attributed to prolonged engagement in these coping mechanisms, which become deeply ingrained and automatic over time.^1,4,13,14,16 Camouflaging can be seen as an instinctual response to social trauma, driven by a deep-seated need for safety and social survival, making it difficult to unlearn and unmask.^5,72–74 In addition, our findings align with research showing that, following an increase in childhood and adolescence, camouflaging remains elevated throughout adulthood among autistic individuals, in contrast to a decline in non-autistic people.⁷⁵ While non-autistic people may become less concerned about social impressions with age due to increased emotional stability, autistic adults often face persistent external pressures, such as stigma and the need for social connections, which reinforce camouflaging as essential for societal acceptance.^1,2,13,75

Concurrently, the small decline in camouflaging observed in our growth modeling analyses aligns with autistic adults’ reports of reduced camouflaging with age as they gain self-awareness and understanding and adopt alternative strategies such as disclosure, seeking accommodations, and advocacy.^13,19

Earlier psychosocial predictors of later camouflaging

The path analysis (Fig. 2) replicated and extended several pathways from Zhuang et al.’s²⁹ cross-sectional findings, whereby earlier psychosocial predictors measured prospectively exerted enduring influences over 2 years and remained linked to later camouflaging. Notably, FNE, previously identified as an important cross-sectional predictor, remained a significant earlier predictor of later camouflaging in this study. This aligns with qualitative studies highlighting how autistic individuals develop camouflaging strategies in response to early negative social experiences, such as bullying, rejection, and discrimination.^2,5,24 These experiences may foster fear of social rejection and contribute to camouflaging that begins in childhood or adolescence and persists into adulthood, as autistic individuals anticipate social threats and avoid negative evaluations.^{1,2,4,5,13,73}

Furthermore, our finding that earlier FNE predicted later camouflaging supports the notion of camouflaging as a long-term protective stigma-mitigation strategy used by autistic individuals to navigate their marginalized status.^28,32,34 Given FNE’s established role in social anxiety,⁷⁶ this result aligns with research showing significant links between camouflaging and social anxiety.^77–80 Collectively, these findings suggest that camouflaging may share similarities with social anxiety-related safety behaviors, with FNE acting as a common driver of both.^78,81,82 However, it is important not to conflate the two—while safety behaviors are typically responses to perceived but often disproportionate threats, camouflaging in autistic individuals often stems from accumulated negative social experiences and very real social threats, including FNE.^9,83

In addition to FNE, earlier T1 independent and interdependent self-construal scores, and T2 autistic identity, demonstrated persistent associations with T3 camouflaging over 2 years. These factors, related to self-perception and identity, may be deeply ingrained and stable, reflecting deeply held beliefs about oneself and one’s relationships with the social world and the autistic community. This is supported by the high correlations (above 0.70) among these variables across time points, which suggest their temporal stability and may partially account for their consistent associations with camouflaging over time.

Furthermore, consistent with Zhuang et al.,²⁹ both independent and interdependent self-construals were linked to a stronger autistic identity, which, in turn, predicted greater T3 camouflaging. Autistic individuals who value individuality may be more inclined to embrace their unique autistic identity, while those prioritizing group cohesion may also report a stronger autistic identity if they consider the autistic community as their primary group affiliation. Our finding that earlier autistic identity predicted later camouflaging does not appear to be consistent with qualitative reports suggesting that autistic individuals may feel less need to camouflage as they better understand and accept their autistic identity.^4,5,13,16 However, the present finding builds on previous quantitative cross-sectional studies using the CAT-Q, which also reported a positive relationship between camouflaging, autistic identity, and community affiliation,^33,38,84,85 suggesting that this relationship persists over time. Individuals with a stronger autistic identity may have heightened awareness of their autistic differences and the associated stigma they experience, leading them to camouflage more.^38,84 The seemingly counterintuitive association between autistic identity and camouflaging underscores the powerful influence of sociocultural pressures and highlights the tension that many autistic people experience navigating between their personal and public identities.^38,85 In situations where personal acceptance conflicts with sociocultural expectations, these external pressures can be so overwhelming and pervasive that autistic individuals may feel compelled to camouflage, despite their strong autistic identity.^28,86

Integrating this study’s findings on the associations between FNE, autistic identity, and camouflaging, our findings suggest two potentially distinct routes to camouflaging. The first route, driven by FNE and often involving internalized stigma and shame, may lead autistic individuals to camouflage to mitigate a negative self-image and minimize negative evaluation.^5,13 The second route may involve camouflaging that coexists with a strong autistic identity, characterized by pride, advocacy, and connection to the autistic community.³³ Here, camouflaging may be a pragmatic response to perceived, but not internalized, stigma and may be adopted out of practical necessity to navigate a predominantly neurotypical society.^9,85 Notably, our study did not replicate Zhuang et al.’s²⁹ cross-sectional association between higher self-esteem and increased camouflaging. While the zero-order correlation between T2 self-esteem and T3 camouflaging was significant and negative, the effect size of this association was small and became nonsignificant after accounting for other psychosocial factors in the model. Future research should continue to investigate the links between self-esteem and camouflaging and explore how various psychosocial factors interact to influence these dynamics over time.

In addition, the direct and indirect cross-sectional associations between perceived stigma, vulnerability to negative life events, and cultural tightness with camouflaging, reported in Zhuang et al.,²⁹ were not replicated in path modeling over 2 years using data from three time points. In the cross-sectional model of Zhuang et al.,²⁹ these pathways generally showed smaller effects, with most path coefficients below 0.20. The absence of enduring associations in this follow-up study may reflect the context- and time-specific nature of these factors, which could fluctuate with changes in one’s environment, as suggested by the lower ICCs over time for perceived stigma and cultural tightness, compared with the higher ICCs observed for other psychosocial factors (see Supplementary Data). Furthermore, the temporal dynamics between the psychosocial factors and camouflaging may unfold over different time frames. For example, factors such as perceived stigma and vulnerability to negative events may be stronger predictors of camouflaging when measured closer in time. These factors may exert more immediate influences, with camouflaging acting as a short-term adaptive response to recent negative events or perceived stigma, which could explain their diminishing influence over an extended period.

In the more conservative path model controlling for T1 camouflaging and psychological factors (see Supplementary Data), T1 predictors were no longer significant except for T1 independent self-construal, which remained significantly negatively associated with T3 camouflaging, over and above T1 camouflaging and the other psychosocial predictors. When considered in relation to the findings of the primary (unadjusted) model, this pattern of findings reflects the strong temporal stability of camouflaging and psychological variables (ICCs = 0.73–0.86): many of the prospective associations across the 2-year study period observed in the primary model likely capture relatively stable between-person differences, alongside any residual change. Notably, independent self-construal uniquely predicted residual change in camouflaging, potentially highlighting the role of earlier independent self-construal in predicting later camouflaging and underscoring the importance of sociocultural orientation in understanding camouflaging, which has been largely underexplored in the literature to date.

Limitations and future directions

The overall stability in camouflaging observed in this study may be influenced by our assessment approach, which, although prospective, asked participants to recall their camouflaging behaviors at approximately 1-year intervals, potentially missing the dynamic nature of camouflaging. Moreover, the 2-year follow-up may have been too brief to capture meaningful variations in camouflaging, given the relative stability of psychosocial contexts in adulthood, such as more established relationships and consistent work environments, compared with the fluidity of earlier developmental stages.⁸⁷

Future research could use ecological momentary assessment to capture real-time camouflaging, providing a finer grained understanding of how it varies across different contexts and closely spaced time intervals (see, e.g., Scheeren et al., in press⁸⁸). In addition, investigating camouflaging over extended periods, particularly during key developmental stages such as adolescence, life transitions, or events such as receiving an autism diagnosis, could provide insights into how it evolves with changing psychosocial contexts. Large cross-sectional studies recruiting autistic individuals from various age cohorts could explore developmental and age-related differences in camouflaging. Although a few studies have examined age associations with camouflaging, these samples are often small, typically including 100 or fewer autistic participants.^75,89–91

Furthermore, the generalizability of our findings may be limited by the characteristics of our participants, most of whom were White, female, and highly educated. Given that our study identified cultural orientation as an earlier predictor of later camouflaging, future studies should explore cultural variations in camouflaging. Although studies in non-Western, collectivistic cultures, such as in Japan^85,92–94 and Taiwan,^95–96 are increasing, most research, including ours, has recruited participants from presumably individualistic countries such as the United Kingdom, the United States, and Australia.^46,97 To our knowledge, only two studies to date have conducted direct cross-cultural comparisons,^98–99 with the larger study limited by a similar, predominantly White, and highly educated participant demographic from developed, individualistic nations.⁹⁸

In addition, future studies with larger and more diverse samples should examine whether key sociodemographic variables—such as age, gender, education level, or employment status—moderate the stability of camouflaging or its psychosocial predictors. Such research could also benefit from longitudinal data-analytic techniques, such as cross-lagged panel modeling, to disentangle reciprocal and contemporaneous relationships between camouflaging and its psychosocial correlates over time.

Attrition also poses a limitation to our prospective follow-up study design. Of the 231 participants with valid baseline data at T1, retention rates were 58.4% at T2 and 45.0% at T3, reflecting a substantial reduction in sample size over time. Although analyses showed no significant baseline differences between those who completed follow-ups and those who did not, and missing data were addressed using MLR estimation or Bayesian estimation, the reduced sample size over time may still influence the generalizability of the findings. Future studies could benefit from using statistical techniques such as inverse probability weighting to more rigorously address potential attrition-related bias.

Finally, future studies should explore whether different motivations for camouflaging are associated with different mental health and well-being outcomes, as these relationships likely vary across individuals. For example, camouflaging driven by FNE may perpetuate social anxiety in the long term, similar to how safety behaviors maintain social anxiety over time in Clark and Wells’¹⁰⁰ cognitive model of social anxiety.^78,101 In contrast, autistic individuals with a strong autistic identity might view camouflaging as a pragmatic tool for social interaction, thus experiencing less negative psychological impact.⁸⁵ Alternatively, others with a strong autistic identity who camouflage may experience cognitive dissonance between their values and outward behaviors, negatively affecting their well-being. Indeed, some autistic individuals have reported feeling they are denying their autistic identity and “betraying” the autistic community by camouflaging.² Future longitudinal research should investigate how specific psychosocial motivations—such as managing FNE or navigating identity conflict—predict mental health outcomes and identify protective factors that allow some individuals to camouflage without significant psychological cost.

Implications

Our findings emphasize the role of individual psychological factors and broader sociocultural influences in shaping camouflaging behaviors. They highlight a nuanced interplay in which earlier sociocultural factors, such as self-construals, influence later individual psychological factors, such as FNE and autistic identity, which in turn affect camouflaging. Taken together, our findings call for a “two-pronged approach” that promotes broader systemic change while simultaneously supporting the needs of autistic individuals.⁸⁶(p11),101

Regarding systemic change, the substantial stability of camouflaging over 2 years suggests that societal norms continue to exert significant pressures on autistic individuals to conform. According to the minority stress model, autistic people, as a devalued minority, experience daily discrimination, internalized stigma, and heightened FNE, which drive camouflaging as a protective strategy and negatively impact mental health and well-being.^{28,77,102,103} Our finding that earlier FNE significantly predicts later camouflaging underscores the need for a neurodiversity-affirming culture to alleviate these psychosocial pressures.^77,101 Potential interventions include antistigma education and training for the public and professionals, as well as inclusive media representations of autistic diversity to challenge harmful stereotypes.^34,104,105 By fostering greater acceptance and inclusivity, these initiatives can help create environments where autistic individuals feel safe to express themselves authentically without fear of judgment.

Regarding individual supports, professionals working with autistic individuals should recognize the potential long-term impact of psychosocial factors on camouflaging behaviors. Our finding that earlier FNE, often stemming from adverse experiences such as bullying and victimization,^2,5,13 can contribute to later camouflaging suggests the need for trauma-informed therapeutic approaches. As camouflaging is often a survival-driven response rather than a voluntary choice,¹⁰⁶ professionals supporting autistic people need to provide corrective experiences that foster safety, collaboration, and agency. While autistic individuals often face systemic barriers and may have limited control over their environments, supports should help them navigate these constraints and make purposeful decisions to shape safe environments and relationships.^107–108 Furthermore, our finding that a stronger earlier independent self-construal predicted reduced camouflaging later suggests that environments embracing diversity and supporting self-expression are crucial for fostering autonomy and healthy identity formation for autistic people.²¹ Given the potential overlap between camouflaging and social anxiety,^77–80 strategies used to support socially anxious individuals—such as narrative therapy and imagery rescripting—may also help autistic individuals process and unlearn the harms associated with social trauma.^109–111

Overall, our findings highlight the complexities in the relationship between psychosocial motivations and camouflaging. These complexities emphasize the need for professionals to (a) collaborate with autistic individuals to develop a shared understanding of the contexts, reasons, and functions of camouflaging—whether driven by FNE or pragmatic social goals—and its implications for mental well-being^80,101; (b) support autistic individuals sensitively and respectfully in navigating the tension between autistic identity pride and sociocultural pressures; and (c) work alongside autistic individuals to identify and differentiate between helpful coping strategies, necessary camouflaging, and unhelpful or unnecessary safety behaviors, and decide which behaviors to keep, modify, or eliminate based on social context.

Footnotes

Acknowledgments

The authors thank Dawn-Joy Leong, multi-artist, independent researcher, and autistic consultant from Singapore, for her contributions to the initial conceptualization of the broader camouflaging research project within which this follow-up study was embedded.

Author Disclosure Statement

All authors declare that they have no conflicts of interest.

Funding Information

S.Z. is funded by the University of Western Australia International Fee Scholarship and University Postgraduate Award. D.W.T. is supported by a Macquarie University Research Fellowship (MQRF0001113). In addition, this work is supported by a start-up grant awarded to I.M. (PG10301041) by the University of Western Australia.

Authorship Confirmation Statement

S.Z.: Conceptualization, methodology, investigation, data curation, formal analysis, visualization, writing—original draft, writing—review and editing, and project administration. M.B.: Investigation, data curation, and writing—review and editing. S.R.: Conceptualization, validation, and writing—review and editing. L.D.: Conceptualization, validation, and writing—review and editing. K.N.-G.: Conceptualization, methodology, resources, formal analysis, supervision, and writing—review and editing. M.M.: Conceptualization, methodology, supervision, and writing—review and editing. D.W.T.: Conceptualization, methodology, supervision, and writing—review and editing. I.M.: Conceptualization, methodology, supervision, writing—review and editing, funding acquisition, and project administration. The article has been submitted solely to Autism in Adulthood.

Supplemental Material

References

Cage

, Troxell-Whitman

. Understanding the reasons, contexts and costs of camouflaging for autistic adults. J Autism Dev Disord. 2019; 49(5):1899–1911; doi: 10.1007/s10803-018-03878-x

Hull

, Petrides

, Allison

, et al. “Putting on my best normal”: Social camouflaging in adults with autism spectrum conditions. J Autism Dev Disord. 2017; 47(8):2519–2534; doi: 10.1007/s10803-017-3166-5

Hull

, Mandy

, Lai

M-C

, et al. Development and validation of the Camouflaging Autistic Traits Questionnaire (CAT-Q). J Autism Dev Disord. 2019; 49(3):819–833; doi: 10.1007/s10803-018-3792-6

Bradley

, Shaw

, Baron-Cohen

, Cassidy

. Autistic adults’ experiences of camouflaging and its perceived impact on mental health. Autism Adulthood. 2021; 3(4):320–329; doi: 10.1089/aut.2020.0071

Chapman

, Rose

, Hull

, Mandy

. “I want to fit in… but I don’t want to change myself fundamentally”: A qualitative exploration of the relationship between masking and mental health for autistic teenagers. Res Autism Spectr Disord. 2022; 99:102069; doi: 10.1016/j.rasd.2022.102069

Klein

, Krahn

, Howe

, et al. A systematic review of social camouflaging in autistic adults and youth: Implications and theory. Dev Psychopathol. 2025; 37(3):1320–1334; doi: 10.1017/s0954579424001159

Khudiakova

, Russell

, Sowden-Carvalho

, Surtees

. A systematic review and meta-analysis of mental health outcomes associated with camouflaging in autistic people. Res Autism Spectr Disord. 2024; 118:102492; doi: 10.1016/j.rasd.2024.102492

Remnélius

, Neufeld

, Isaksson

, Bölte

. Does camouflaging cause reduced quality of life? A co-twin control study. J Autism Dev Disord. 2024; doi: 10.1007/s10803-024-06583-0

Zhuang

, Tan

, Reddrop

, et al. Psychosocial factors associated with camouflaging in autistic people and its relationship with mental health and well-being: A mixed methods systematic review. Clin Psychol Rev. 2023; 105:102335; doi: 10.1016/j.cpr.2023.102335

10.

Dean

, Harwood

, Kasari

. The art of camouflage: Gender differences in the social behaviours of girls and boys with autism spectrum disorder. Autism. 2017; 21(6):678–689; doi: 10.1177/1362361316671845

11.

Bernardin

, Lewis

, Bell

, Kanne

. Associations between social camouflaging and internalising symptoms in autistic and non-autistic adolescents. Autism. 2021; 25(6):1580–1591; doi: 10.1177/1362361321997284

12.

Jedrzejewska

, Dewey

. Camouflaging in autistic and non-autistic adolescents in the modern context of social media. J Autism Dev Disord. 2022; 52(2):630–646; doi: 10.1007/s10803-021-04953-6

13.

Loo

, Teo

, Liang

, et al. Exploring autistic adults’ psychosocial experiences affecting beginnings, continuity and change in camouflaging over time: A qualitative study in Singapore. Autism. 2024; 28(3):627–643; doi: 10.1177/13623613231180075

14.

Miller

, Rees

, Pearson

. “Masking is life”: Experiences of masking in autistic and nonautistic adults. Autism Adulthood. 2021; 3(4):330–338; doi: 10.1089/aut.2020.0083

15.

Wood-Downie

, Wong

, Kovshoff

, et al. Sex/Gender differences in camouflaging in children and adolescents with autism. J Autism Dev Disord. 2021; 51(4):1353–1364; doi: 10.1007/s10803-020-04615-z

16.

Cook

, Crane

, Bourne

, Hull

, Mandy

. Camouflaging in an everyday social context: An interpersonal recall study. Autism. 2021; 25(5):1444–1456; doi: 10.1177/1362361321992641

17.

Pryke-Hobbes

, Davies

, Heasman

, et al. The workplace masking experiences of autistic, non-autistic neurodivergent and neurotypical adults in the UK. PLoS One. 2023; 18(9):e0290001; doi: 10.1371/journal.pone.0290001

18.

Halsall

, Clarke

, Crane

. “Camouflaging” by adolescent autistic girls who attend both mainstream and specialist resource classes: Perspectives of girls, their mothers and their educators. Autism. 2021; 25(7):2074–2086; doi: 10.1177/13623613211012819

19.

Livingston

, Shah

, Happé

. Compensatory strategies below the behavioural surface in autism: A qualitative study. Lancet Psychiatry. 2019; 6(9):766–777; doi: 10.1016/s2215-0366(19)30224-x

20.

Lilley

, Lawson

, Hall

, et al. ‘A way to be me’: Autobiographical reflections of autistic adults diagnosed in mid-to-late adulthood. Autism. 2022; 26(6):1395–1408; doi: 10.1177/13623613211050694

21.

Øverland

, Andersen

, Orm

, et al. “I have tried to fit in, but I just want to be me”: Development toward identity formation in young autistic adults. Autism Adulthood. 2024; doi: 10.1089/aut.2024.0156

22.

Schneid

, Raz

. The mask of autism: Social camouflaging and impression management as coping/normalisation from the perspectives of autistic adults. Soc Sci Med. 2020; 248:112826; doi: 10.1016/j.socscimed.2020.112826

23.

Angulo

, Chan

, DeThorne

. Life is a stage: Autistic perspectives on neurotypicality. Autism Adulthood. 2019; 1(4):276–285; doi: 10.1089/aut.2019.0024

24.

Bernardin

, Mason

, Lewis

, Kanne

. “You must become a chameleon to survive”: Adolescent experiences of camouflaging. J Autism Dev Disord. 2021; 51(12):4422–4435; doi: 10.1007/s10803-021-04912-1

25.

Higgins

, Arnold

, Weise

, et al. Defining autistic burnout through experts by lived experience: Grounded Delphi method investigating #AutisticBurnout. Autism. 2021; 25(8):2356–2369; doi: 10.1177/13623613211019858

26.

Mantzalas

, Richdale

, Adikari

, Lowe

, Dissanayake

. What is autistic burnout? A thematic analysis of posts on two online platforms. Autism Adulthood. 2022; 4(1):52–65; doi: 10.1089/aut.2021.0021

27.

Raymaker

, Teo

, Steckler

, et al. “Having all of your internal resources exhausted beyond measure and being left with no clean-up crew”: Defining autistic burnout. Autism Adulthood. 2020; 2(2):132–143; doi: 10.1089/aut.2019.0079

28.

, Cunningham

, Lai

M-C

. Reconsidering autistic ‘camouflaging’ as transactional impression management. Trends Cogn Sci. 2022; 26(8):631–645; doi: 10.1016/j.tics.2022.05.002

29.

Zhuang

, Bougoure

, Leong

, et al. Examining an integrated path model of psychological and sociocultural predictors of camouflaging in autistic adults. Autism. 2025; 29(9):2328–2343; doi: 10.1177/13623613241262110

30.

Evans

, Krumrei-Mancuso

, Rouse

. What you are hiding could be hurting you: Autistic masking in relation to mental health, interpersonal trauma, authenticity, and self-esteem. Autism Adulthood. 2024; 6(2):229–240; doi: 10.1089/aut.2022.0115

31.

Botha

, Dibb

, Frost

. “Autism is me”: An investigation of how autistic individuals make sense of autism and stigma. Disabil Soc. 2022; 37(3):427–453; doi: 10.1080/09687599.2020.1822782

32.

Pearson

, Rose

. A conceptual analysis of autistic masking: Understanding the narrative of stigma and the illusion of choice. Autism Adulthood. 2021; 3(1):52–60; doi: 10.1089/aut.2020.0043

33.

Perry

, Mandy

, Hull

, Cage

. Understanding camouflaging as a response to autism-related stigma: A social identity theory approach. J Autism Dev Disord. 2022; 52(2):800–810; doi: 10.1007/s10803-021-04987-w

34.

Turnock

, Langley

, Jones

CRG

. Understanding stigma in autism: A narrative review and theoretical model. Autism Adulthood. 2022; 4(1):76–91; doi: 10.1089/aut.2021.0005

35.

Trundle

, Jones

, Ropar

, Egan

. The forensic implications of camouflaging: A study into victimisation and offending associated with autism and pathological demand avoidance. AIA. 2023; 9(2):116–131; doi: 10.1108/aia-02-2022-0006

36.

Gelfand

, Raver

, Nishii

, et al. Differences between tight and loose cultures: A 33-nation study. Science. 2011; 332(6033):1100–1104; doi: 10.1126/science.1197754

37.

Markus

, Kitayama

. Culture and the self: Implications for cognition, emotion, and motivation. Psychol Rev. 1991; 98(2):224–253; doi: 10.1037/0033-295x.98.2.224

38.

Cage

, Troxell-Whitman

. Understanding the relationships between autistic identity, disclosure, and camouflaging. Autism Adulthood. 2020; 2(4):334–338; doi: 10.1089/aut.2020.0016

39.

Jose

. The merits of using longitudinal mediation. Educ Psychol. 2016; 51(3–4):331–341; doi: 10.1080/00461520.2016.1207175

40.

, Yoshida

, Kaufman

, Mathur

. A brief primer on conducting regression-based causal mediation analysis. Psychol Trauma. 2023; 15(6):930–938; doi: 10.1037/tra0001421

41.

Caemmerer

, Hennessy

, Niileksela

. Third variables in longitudinal research: Application of longitudinal mediation and moderation in school psychology. J Sch Psychol. 2024; 103:101283; doi: 10.1016/j.jsp.2024.101283

42.

Preacher

. Advances in mediation analysis: A survey and synthesis of new developments. Annu Rev Psychol. 2015; 66(1):825–852; doi: 10.1146/annurev-psych-010814-015258

43.

de Broize

, Evans

, Whitehouse

, et al. Exploring the experience of seeking an autism diagnosis as an adult. Autism Adulthood. 2022; 4(2):130–140; doi: 10.1089/aut.2021.0028

44.

Lai

M-C

, Baron-Cohen

. Identifying the lost generation of adults with autism spectrum conditions. Lancet Psychiatry. 2015; 2(11):1013–1027; doi: 10.1016/S2215-0366(15)00277-1

45.

Lewis

. A mixed methods study of barriers to formal diagnosis of autism spectrum disorder in adults. J Autism Dev Disord. 2017; 47(8):2410–2424; doi: 10.1007/s10803-017-3168-3

46.

Libsack

, Keenan

, Freden

, et al. A systematic review of passing as non-autistic in autism spectrum disorder. Clin Child Fam Psychol Rev. 2021; 24(4):783–812; doi: 10.1007/s10567-021-00365-1

47.

Hurley

, Losh

, Parlier

, Reznick

, Piven

. The Broad Autism Phenotype Questionnaire. J Autism Dev Disord. 2007; 37(9):1679–1690; doi: 10.1007/s10803-006-0299-3

48.

Sasson

, Lam

, Childress

, et al. The Broad Autism Phenotype Questionnaire: Prevalence and diagnostic classification. Autism Res. 2013; 6(2):134–143; doi: 10.1002/aur.1272

49.

Hannon

, Mandy

, Hull

. A comparison of methods for measuring camouflaging in autism. Autism Res. 2023; 16(1):12–29; doi: 10.1002/aur.2850

50.

Carleton

, Collimore

, Asmundson

GJG

. Social anxiety and fear of negative evaluation: Construct validity of the BFNE-II. J Anxiety Disord. 2007; 21(1):131–141; doi: 10.1016/j.janxdis.2006.03.010

51.

Carleton

, Collimore

, McCabe

, Antony

. Addressing revisions to the Brief Fear of Negative Evaluation scale: Measuring fear of negative evaluation across anxiety and mood disorders. J Anxiety Disord. 2011; 25(6):822–828; doi: 10.1016/j.janxdis.2011.04.002

52.

Rosenberg

. Society and the adolescent self-image. Princeton University Press; 1965. doi:10.1515/9781400876136

53.

Cooper

, Smith

, Russell

. Social identity, self‐esteem, and mental health in autism. Euro J Social Psych. 2017; 47(7):844–854; doi: 10.1002/ejsp.2297

54.

Corden

, Brewer

, Cage

. Personal identity after an autism diagnosis: Relationships with self-esteem, mental wellbeing, and diagnostic timing. Front Psychol. 2021; 12:699335; doi: 10.3389/fpsyg.2021.699335

55.

Nguyen

, Ownsworth

, Nicol

, Zimmerman

. How I see and feel about myself: Domain-specific self-concept and self-esteem in autistic adults. Front Psychol. 2020; 11:913; doi: 10.3389/fpsyg.2020.00913

56.

Leach

, Van Zomeren

, Zebel

, et al. Group-level self-definition and self-investment: A hierarchical (multicomponent) model of in-group identification. J Pers Soc Psychol. 2008; 95(1):144–165; doi: 10.1037/0022-3514.95.1.144

57.

Cooper

, Russell

, Lei

, Smith

. The impact of a positive autism identity and autistic community solidarity on social anxiety and mental health in autistic young people. Autism. 2023; 27(3):848–857; doi: 10.1177/13623613221118351

58.

Cooper

, Cooper

, Russell

, Smith

. “I’m proud to be a little bit different”: The effects of autistic individuals’ perceptions of autism and autism social identity on their collective self-esteem. J Autism Dev Disord. 2021; 51(2):704–714; doi: 10.1007/s10803-020-04575-4

59.

Maitland

, Rhodes

, O’Hare

, Stewart

. Social identities and mental well-being in autistic adults. Autism. 2021; 25(6):1771–1783; doi: 10.1177/13623613211004328

60.

Shtayermman

. An exploratory study of the stigma associated with a diagnosis of Asperger’s syndrome: The mental health impact on the adolescents and young adults diagnosed with a disability with a social nature. J Hum Behav Soc Environ. 2009; 19(3):298–313; doi: 10.1080/10911350902790720

61.

Griffiths

, Allison

, Kenny

, et al. The Vulnerability Experiences Quotient (VEQ): A study of vulnerability, mental health, and life satisfaction in autistic adults. Autism Res. 2019; 12(10):1516–1528; doi: 10.1002/aur.2162

62.

Aktas

, Gelfand

, Hanges

. Cultural tightness-looseness and perceptions of effective leadership. J Cross-Cult Psychol. 2016; 47(2):294–309; doi: 10.1177/0022022115606802

63.

Eriksson

, Strimling

, Gelfand

, et al. Perceptions of the appropriate response to norm violation in 57 societies. Nat Commun. 2021; 12(1):1481; doi: 10.1038/s41467-021-21602-9

64.

Singelis

. The measurement of independent and interdependent self-construals. Pers Soc Psychol Bull. 1994; 20(5):580–591; doi: 10.1177/0146167294205014

65.

Christopher

, Skillman

. Exploring the link between self‐construal and distress among African American and Asian American college students. J of College Counseling. 2009; 12(1):44–56; doi: 10.1002/j.2161-1882.2009.tb00039.x

66.

Singelis

, Triandis

, Bhawuk

, Gelfand

. Horizontal and vertical dimensions of individualism and collectivism: A theoretical and measurement refinement. Cross-Cult Res. 1995; 29(3):240–275; doi: 10.1177/106939719502900302

67.

, Gu

, Fukui

. Combining proration and full information maximum likelihood in handling missing data in Likert scale items: A hybrid approach. Behav Res Methods. 2022; 54(2):922–940; doi: 10.3758/s13428-021-01671-w

68.

Muthén

, Asparouhov

. Bayesian structural equation modelling: A more flexible representation of substantive theory. Psychol Methods. 2012; 17(3):313–335; doi: 10.1037/a0026802

69.

Bentler

. Comparative fit indexes in structural models. Psychol Bull. 1990; 107(2):238–246; doi: 10.1037//0033-2909.107.2.238

70.

, Bentler

. Cutoff criteria for fit indexes in covariance structure analysis: Conventional criteria versus new alternatives. Struct Equ Modelling. 1999; 6(1):1–55; doi: 10.1080/10705519909540118

71.

Asparouhov

, Muthén

. Advances in Bayesian model fit evaluation for structural equation models. Struct Equ Model. 2021; 28(1):1–14; doi: 10.1080/10705511.2020.1764360

72.

Lawson

. Adaptive morphing and coping with social threat in autism: An autistic perspective. J Intellect Disabl Diagn Treat. 2020; 8(3):519–526; doi: 10.6000/2292-2598.2020.08.03.29

73.

Pearson

, Rose

, Rees

. ‘I felt like I deserved it because I was autistic’: Understanding the impact of interpersonal victimisation in the lives of autistic people. Autism. 2023; 27(2):500–511; doi: 10.1177/13623613221104546

74.

Seers

, Hogg

. “Fake it ‘till you make it”: Authenticity and wellbeing in late diagnosed autistic women. Feminism Psychol. 2023; 33(1):23–41; doi: 10.1177/09593535221101455

75.

Remnélius

, Bölte

. Camouflaging in autism: Age effects and cross-cultural validation of the camouflaging autistic traits questionnaire (CAT-Q). J Autism Dev Disord. 2024; 54(5):1749–1764; doi: 10.1007/s10803-023-05909-8

76.

Rapee

, Heimberg

. A cognitive-behavioural model of anxiety in social phobia. Behav Res Ther. 1997; 35(8):741–756; doi: 10.1016/s0005-7967(97)00022-3

77.

, Cunningham

, Lai

M-C

. Camouflaging, internalised stigma, and mental health in the general population. Int J Soc Psychiatry. 2024; 70(7):1239–1253; doi: 10.1177/00207640241260020

78.

Lei

, Leigh

, Charman

, Russell

, Hollocks

. Understanding the relationship between social camouflaging in autism and safety behaviours in social anxiety in autistic and non‐autistic adolescents. J Child Psychol Psychiatry. 2024; 65(3):285–297; doi: 10.1111/jcpp.13884

79.

McKinnon

, Bougoure

, Zhuang

, Tan

, Magiati

. Exploring the construct validity of the Camouflaging Autistic Traits Questionnaire: A factor analytic study. Autism. 2025; 29(3):642–658; doi: 10.1177/13623613241287964

80.

Pyszkowska

. It is more anxiousness than role-playing: Social camouflaging conceptualisation among adults on the autism spectrum compared to persons with social anxiety disorder. J Autism Dev Disord. 2025; 55(9):3154–3166; doi: 10.1007/s10803-024-06416-0

81.

Spain

, Zıvralı Yarar

, Happé

. Social anxiety in adults with autism: A qualitative study. Int J Qual Stud Health Well-being. 2020; 15(1):1803669; doi: 10.1080/17482631.2020.1803669

82.

Wilson

, Gullon-Scott

. Social anxiety in autistic people: Does the Clark and Wells model fit? J Autism Dev Disord. 2024; 54(10):3908–3920; doi: 10.1007/s10803-023-06108-1

83.

Spain

. Social anxiety. In: Chaplin

, McCarthy

, Spain

., eds. A clinician’s guide to mental health conditions in adults with autism spectrum disorders: Assessment and interventions. Jessica Kingsley Publishers; 2019:128–146.

84.

Cage

, Cranney

, Botha

. Brief report: Does autistic community connectedness moderate the relationship between masking and well-being? Autism Adulthood. 2022; 4(3):247–253; doi: 10.1089/aut.2021.0096

85.

Tamura

, Cage

, Perry

, et al. Understanding camouflaging, stigma, and mental health for autistic people in Japan. Autism Adulthood. 2025; 7(1):52–65; doi: 10.1089/aut.2023.0035

86.

Moore

, Cassidy

, Rodgers

. Exploring the mediating effect of camouflaging and the moderating effect of autistic identity on the relationship between autistic traits and mental wellbeing. Autism Res. 2024; 17(7):1391–1406; doi: 10.1002/aur.3073

87.

Erikson

. The life cycle completed: A review. W. W. Norton & Company; 1982.

88.

Scheeren

, Nieuwenhuis

, Crane

, Roke

, Begeer

. Masking, social context and perceived stress in autistic adults: An ecological momentary assessment study. Autism. In press. Available from: https://research.birmingham.ac.uk/en/publications/masking-social-context-and-perceived-stress-in-autistic-adults-an [Last accessed: June 25, 2025].

89.

Hull

, Petrides

, Mandy

. Cognitive predictors of self-reported camouflaging in autistic adolescents. Autism Res. 2021; 14(3):523–532; doi: 10.1002/aur.2407

90.

Lai

M-C

, Lombardo

, Ruigrok

, et al.; MRC AIMS Consortium. Quantifying and exploring camouflaging in men and women with autism. Autism. 2017; 21(6):690–702; doi: 10.1177/1362361316671012

91.

Lai

M-C

, Lombardo

, Chakrabarti

, et al.; MRC AIMS Consortium. Neural self-representation in autistic women and association with ‘compensatory camouflaging’. Autism. 2019; 23(5):1210–1223; doi: 10.1177/1362361318807159

92.

Hongo

, Oshima

, Guan

, et al. Reliability and validity of the Japanese version of the camouflaging autistic traits questionnaire. Autism Res. 2024; 17(6):1205–1217; doi: 10.1002/aur.3137

93.

Funawatari

, Sumiya

, Iwabuchi

, Senju

. Double-edged effects of social strategies on the well-being of autistic people: Impact of self-perceived effort and efficacy. Brain Sci. 2024; 14(10):962; doi: 10.3390/brainsci14100962

94.

Funawatari

, Sumiya

, Iwabuchi

, Nishimura

, Komeda

, Senju

. Camouflaging in autistic adults is modulated by autistic and neurotypical characteristics of interaction partners. J Autism Dev Disord. 2024; doi: 10.1007/s10803-024-06481-5

95.

Kuo

, Ni

, Liu

. The associations between self-rated autistic traits, social camouflaging, and mental health outcomes in Taiwanese anime, comics and games (ACG) doujin creators: An exploratory study. BMC Psychol. 2024; 12(1):531; doi: 10.1186/s40359-024-02019-7

96.

Liu

, Chen

, Ni

, Lai

M-C

. Exploring camouflaging by the Chinese version camouflaging autistic traits questionnaire in Taiwanese autistic and non-autistic adolescents: An initial development. Autism. 2024; 28(3):690–704; doi: 10.1177/13623613231181732

97.

Cook

, Hull

, Crane

, Mandy

. Camouflaging in autism: A systematic review. Clin Psychol Rev. 2021; 89:102080; doi: 10.1016/j.cpr.2021.102080

98.

Keating

, Hickman

, Geelhand

, et al. Cross-cultural variation in experiences of acceptance, camouflaging and mental health difficulties in autism: A registered report. PLoS One. 2024; 19(3):e0299824; doi: 10.1371/journal.pone.0299824

99.

Oshima

, Takahashi

, Tamura

, et al. The association between social camouflage and mental health among autistic people in Japan and the UK: A cross-cultural study. Mol Autism. 2024; 15(1):1; doi: 10.1186/s13229-023-00579-w

100.

Clark

, Wells

. A cognitive model of social phobia. In: Heimberg

, Liebowitz

, Hope

, Schneier

., eds. Social phobia: Diagnosis, assessment, and treatment. The Guilford Press; 1995:69–93.

101.

Lei

, Cooper

, Hollocks

. Psychological interventions for autistic adolescents with co-occurring anxiety and depression: Considerations linked to autism social identity and masking. Autism in Adulthood. 2024; doi: 10.1089/aut.2024.0005

102.

Botha

, Frost

. Extending the minority stress model to understand mental health problems experienced by the autistic population. Soc Ment Health. 2020; 10(1):20–34; doi: 10.1177/2156869318804297

103.

Rivera

, Bennetto

. Applications of identity-based theories to understand the impact of stigma and camouflaging on mental health outcomes for autistic people. Front Psychiatry. 2023; 14:1243657; doi: 10.3389/fpsyt.2023.1243657

104.

Gillespie-Lynch

, Bisson

, Saade

, et al. If you want to develop an effective autism training, ask autistic students to help you. Autism. 2022; 26(5):1082–1094; doi: 10.1177/13623613211041006

105.

Jones

, DeBrabander

, Sasson

. Effects of autism acceptance training on explicit and implicit biases toward autism. Autism. 2021; 25(5):1246–1261; doi: 10.1177/1362361320984896

106.

Giroux

, Courcy

, Nadig

. Social camouflage in autism: An analysis of decision making. Autism Adulthood. 2024; doi: 10.1089/aut.2023.0077

107.

Klein

, Macoun

. Person-environment fit and social camouflaging in autism. New Ideas Psychol. 2025; 76:101112; doi: 10.1016/j.newideapsych.2024.101112

108.

Pearson

, Rose

. Autistic masking: Understanding identity management and the role of stigma. Pavilion Publishing and Media Ltd; 2023.

109.

Lloyd

, Marczak

. Imagery rescripting and negative self-imagery in social anxiety disorder: A systematic literature review. Behav Cogn Psychother. 2022; 50(3):1–18; doi: 10.1017/s135246582200008x

110.

Thunnissen

, de Jong

, Rijkeboer

, Voncken

, Nauta

. Interventions targeting negative mental imagery in social anxiety: A systematic review and meta-analysis of characteristics and outcomes. Clin Psychol Psychother. 2024; 31(3):e2996; doi: 10.1002/cpp.2996

111.

White

, Epston

. Narrative means to therapeutic ends. W. W. Norton & Company; 1990.

112.

Livingston

, Happé

. Conceptualising compensation in neurodevelopmental disorders: Reflections from autism spectrum disorder. Neurosci Biobehav Rev. 2017; 80:729–742; doi: 10.1016/j.neubiorev.2017.06.005

113.

Milton

, Sims

. How is a sense of well-being and belonging constructed in the accounts of autistic adults? Disabil Soc. 2016; 31(4):520–534; doi: 10.1080/09687599.2016.1186529

114.

Arnold

, Higgins

, Weise

, et al. Confirming the nature of autistic burnout. Autism. 2023; 27(7):1906–1918; doi: 10.1177/13623613221147410

115.

Cohen

. Statistical power analysis for the behavioural sciences (2nd ed.). Routledge; 1988. doi:10.4324/9780203771587

116.

Cohen

. A power primer. Psychol Bull. 1992; 112(1):155–159; doi: 10.1037/0033-2909.112.1.155

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.13 MB

0.00 MB

0.09 MB