Understanding camouflaging, stigma, and mental health for autistic people in Japan

1 Background : Camouflaging refers to behaviors in which autistic individuals mask their 2 autistic characteristics and “pass” as non-autistic people. It is postulated that camouflaging is 3 a response to stigma, and preliminary evidence supports this hypothesis. However, research 4 on this topic outside of Western countries is limited. This study replicated and extended 5 previous work in the West that examined the relationships between camouflaging, stigma, 6 and mental health of autistic adults, with a Japanese sample. 7 Methods: Two-hundred eighty-seven autistic people living in Japan (146 men, 120 women, 8 14 non-binary, 5 other gender identities, 2 preferred not to say; mean age = 37.5 years, 9 standard deviation = 9.8 years) completed an online survey on camouflaging, perceived 10 stigma, coping strategies for stigma, mental well-being, generalized anxiety, social anxiety, 11 and depression. We used hierarchical multiple regression analyses to investigate the 12 relationships between camouflaging and stigma and coping strategies for stigma. Mediation 13 analyses were also employed to examine whether camouflaging mediated the relationships 14 between stigma and autistic people’s mental health.


Background
Autistic people have specific styles of social communication and interaction and unique preference in interests and activities.They often experience mental health problems, 1 with the lifetime prevalence of mental health conditions being 79%, higher than in non-autistic people (41%). 2,35][6][7][8] Social camouflaging refers to behaviors in which autistic people conceal their social differences to get by in a predominantly neurotypical world. 7,9It includes several types of behaviors, such as masking autistic behaviors like repetitive hand movements, learning social cues from TV programs, movies, and books, and assimilating with non-autistic people by forcing interaction. 10,11stic people are both benefited and disadvantaged by camouflaging.It helps some autistic people achieve social goals, such as establishing close relationships with others, getting jobs or an education, and building their careers. 7,12,13These social achievements foster self-confidence and increased feeling of connectedness to others among autistic people. 12ver, camouflaging is also related to substantial costs to autistic people.Qualitative studies have described how camouflaging can contribute to fatigue, burnout, a sense of falsifying their identity, and feelings of "false relationships" with others. 4,7Quantitative studies have shown associations between camouflaging and depression, generalized anxiety, social anxiety, and suicidality. 5,6Additionally, camouflaging makes autistic traits harder to recognize and could prevent autistic people from receiving a timely diagnosis. 14,15ough the relationship between camouflaging and mental health of autistic people has not been established clearly, in prior qualitative studies, most autistic people emphasized the disadvantages of camouflaging over its advantages. 7,12,16Ai et al. 17 proposed that camouflaging is a type of impression management (IM) used by various neurotypes, 18 and suggested the possibility that autistic people are more negatively affected by IM/camouflaging compared with neurotypicals.IM refers to behavior in which people attempt to adjust their behavior to create a particular impression of themselves. 18,19IM encourages success in interpersonal relationships, 20 while being related to loneliness and lower levels of life satisfaction. 21IM (camouflaging) can contribute to mental loads for both autistic and non-autistic individuals; 22,23 however, in a qualitative study, 23 only autistic individuals mentioned that masking is linked with dangerous thoughts such as suicidal ideation.
Ai et al. 17 noted the reasons why IM/camouflaging has different impacts on autistic people and neurotypicals.
First, as members of a stigmatized group, autistic people might be compelled to camouflage.Stigma refers to negative attitudes toward characteristics that do not conform to culturally established norms. 24When members of a certain group perceive stigma that the general population has toward them (public stigma), they develop the awareness of stigma or a belief that others hold stigmatizing thoughts toward one's condition (perceived stigma).As members of the stigmatized group begin to endorse and apply these stigmatized beliefs to themselves (internalized stigma), this is associated with lower self-esteem and lower quality of life. 25,268][29] To them, impression management is not behavior aimed at being "better than others," but rather at not being rejected by society, which has been linked with severe depression and reduced quality of life. 29,305][36][37][38] These stigmas can even promote bullying and difficulty finding employment. 39Autistic people might be forced to camouflage to avoid imminent threats, such as violence, which could heighten their anxiety toward the possibility of their camouflage failing, which could, in time, exhaust them.Perry et al. 8 quantitatively investigated the relationship between perceived stigma and camouflaging using social identity theory (SIT). 40Social identity is one's perception of the social group one belongs to.SIT proposes that people are motivated to create, maintain, and protect the positivity of their social identity to maintain positive self-esteem. 41When a group is stigmatized, and the self-esteem of in-group members is threatened, they could adopt two different coping strategies.Those who use "collective strategies" seek to restore positive social identity, redefining the stigmatized group as valuable and centrally defining aspects of identity. 42They resist social norms and values underpinning stigma 43 and try to improve the in-group's status by restructuring oppressive cultural and structural systems (e.g., by participating in social movements). 44In contrast, those who use "individualistic strategies" try to hide, minimize, or overcome stigmatized characteristics and move from stigmatized ingroup to a higher status out-group. 42For autistic people, individualistic strategies include masking their autistic characteristics, denying being autistic, and trying to correct their characteristics to become less autistic. 8,42f the two strategies, using individualistic strategies seems similar to camouflaging, although camouflaging does not include minimizing or trying to overcome autistic characteristics.If camouflaging is highly related to using individualistic strategies (a response to stigma), one can infer that perceived stigma facilitates camouflaging.Perry et al. 8 examined the relationship between perceived stigma, individualistic strategies, collective strategies, and camouflaging.Further, they proposed that camouflaging mediates the relationship between stigma and autistic people's mental well-being, considering the idea that camouflaging is a response to stigma.This hypothesis was based on Botha and Frost's 45 proposition that stigma severely impacts autistic individuals' mental well-being and is associated with lower quality of life and lower self-esteem. 45,46Botha and Frost 45 interpreted these negative effects of stigma within the framework of the minority stress model.The model proposes that people with minority identities are exposed to higher stress burdens, such as stigma or lack of social support, resulting in greater physical and mental health problems. 47The hypothesis of Perry et al. 8 suggested that there are indirect pathways wherein stigma affects mental well-being through increased camouflaging, in addition to the direct effects of stigma that the minority stress model posits.
Perry et al. 8 conducted an online survey of 223 autistic adults (130 women, 53 men, 39 non-binary or other gender identities, 1 preferred not to say; mean age (M) = 34.19years, standard deviation (SD) = 11.00)mainly living in the UK and North America.Multiple regression analyses showed that perceived stigma was positively related to camouflaging, supporting their hypothesis that camouflaging is a response to stigma.As for strategy use, individualistic strategy use was positively associated with camouflaging, suggesting that the two are highly related.However, collective strategy use, which could seem to be the opposite of individualistic strategies, was also positively related to camouflaging.Contrary to the hypothesis, camouflaging did not mediate the relationship between stigma and mental wellbeing.The mediation of camouflaging between stigma and mental health warrants further investigation by considering other aspects of mental health, including depression, generalized anxiety, and social anxiety, which relate to camouflaging. 6e the study by Perry et al. 8 is meaningful, being the first to interpret camouflaging in the framework of SIT, it has some limitations.One of these is the regional, racial, and ethnic bias of participants.Most were Caucasians living in the UK or North America-almost none living in Asian countries.Since social-cultural factors differ between Asia and the West, the style in which autistic people react to stigma and the impacts of camouflaging could differ.
In East Asia, people emphasize "inter-dependence"; they respect cooperation with others and avoid rejection by not disturbing the harmony of the group. 45,46In such environments, people try to find their faults that do not conform to others' expectations and correct them to achieve self-fulfillment. 47Over-adaptation behavior-an attitude in which an individual attempts to conform to the demands of the environment suppressing their personal demands 48 is a well-known concept in Japan, 52 and is a good reflection of this nature.In Western countries, "independence" is respected, where the self is a unique reality separate from others. 49Under such a self-view, people find attributes to be proud of in themselves and try to gain self-esteem by expressing them outwardly. 50Although there are certain criticisms of perceiving national characteristics in this way, 53,54 it is possible that Asian autistic individuals are more sensitive to stigma and camouflage at a higher level.6][57] Someki et al. 56 compared the levels of autism-related stigma and knowledge of autism that Japanese and American college students had.Their finding showed that Japanese students exhibited greater social distance and less knowledge than those in U.S. In addition, a multinational comparison study 58 showed that nearly 70% of parents in Japan believe that autistic individuals differ from their peers in intellectual ability.Such beliefs could contribute to higher levels of stigma. 56,59Outside Japan, Yu et al. 57 showed that only 57-65% of Chinese people have enough knowledge about autism, and 38% supported the discrimination against autistic people.In an environment where the level of autism-related stigma is higher and harmony with others is more expected than in the West, perceived stigma might be more facilitative of camouflaging.Moreover, Japanese people prefer indirect forms of communication compared with those in the West. 60Tezuka stated that the mutual complementarity between speaker and listener makes Japanese communication effective. 61This mutual complementarity may make effective camouflaging more difficult as it may require more prudence and mental resources, which may be linked with poor mental well-being.These differences in socio-cultural backgrounds may influence the relationship between stigma, camouflaging, and the mental health of autistic individuals.Therefore, verifying whether the results of Perry et al. 8 are replicated in Japan-where the cultural background and experience of stigma could differ-and whether the level of stigma, camouflaging, and coping strategies used by autistic people are different, is meaningful.
This study is a quasi-replication of Perry et al. 8 that verifies whether their results are replicated in Japan.We also re-examined the mediation of camouflaging, adding several different mental health questionnaires to the ones they used.Further, we examined whether the degree to which stigma affects camouflaging differs between samples from different cultural backgrounds, specifically the Japanese and Western cultures.
We hypothesized that, in Japan, (1) perceived stigma is more positively related to camouflaging than in the West; (2) both individualistic and collective strategy use positively relate to camouflaging; (3) and camouflaging mediates the relationship between stigma and autistic people's mental health (mental well-being, depression, generalized anxiety, and social anxiety).

Participants
Individuals who self-reported receiving a diagnosis of autism by physicians according to the Diagnostic and Statistical Manual of Mental Disorders fifth edition 62 were eligible to participate.Since the inclusion criteria were verified only by participants' self-reports, we also used the Japanese version of the Autism-Spectrum Quotient 10-item short version 63 to measure autistic characteristics, and excluded those who scored below the cut-off point of 7 (n = 163 excluded).We also excluded those who had received the diagnosis from someone other than a physician1 or had self-diagnosed (none were excluded).Thus, there were 287 participants.
To recruit participants, we sent one email newsletter to individuals who belonged to a private firm's research panel and another to individuals registered with a support facility that helped autistic people find employment.We recruited participants and collected data between January and March 2022.We obtained informed consent from all participants.

Materials and procedure
This study was approved by the Research Ethics Committee at Chiba University (Reference No: M10319).This study did not include autistic consultation, engagement, or coproduction.
Potential participants accessed the research website and were screened to check whether they were eligible for the study.The questionnaires were presented in the order mentioned below.Those who scored below the cut-off on the AQ-J-10 did not complete the questionnaires after that point.
In selecting the scales, we chose the same scale as Perry et al. 8 when possible.In quantifying participants' autistic traits, we chose the AQ-J-10 because the scale that Perry et al. 8 used (Ritvo Autism and Asperger Diagnostic Scale) 64 was not validated and standardized in Japanese.Moreover, in the absence of similar measures in Japanese, we translated the stigma consciousness scale, 65 the individualistic strategy use scale, 42 and the collective strategy use scale, 42 which were used by Perry et al. 8 When scales had to be translated, the Japanese version was back-translated and verified by native English speakers.The translated versions of the scales are shown in Supplemental material 1.

Demographic information
Participants first provided their demographic information, including their age, age at diagnosis, gender, ethnicity, residency, highest level of education, and employment status.

Autism-Spectrum Quotient 10-item short version (AQ-J-10)
The AQ-J-10 is a short version of the Autism Quotient, which quantifies autistic traits. 63,66icipants rated 10 items, such as "I prefer to do things with others rather than on my own," on a four-point Likert scale.Participants scored zero or one point per sentence, depending on their ratings.The total score could range between 0 and 10, and higher scores indicated higher levels of autistic traits.With a cut-off value of 7, the AQ-J-10 has a high negative predictive value (0.97) for autism without intellectual disability. 63The scale had poor internal consistency in this study (α = 0.448).

Patient Health Questionnaire-9-Japanese version
The Patient Health Questionnaire-9 (PHQ-9) 67 is a nine-item, self-administered measure to assess depressive symptoms in the past two weeks.Participants rated each item (e.g., "Little interest or pleasure in doing things") on a Likert scale of 0 (not at all) to 3 (nearly every day).
The total score could range from 0 to 27, with higher scores suggesting more depressive symptoms.Ten is the threshold for the presence of major depressive disorder. 67We used the PHQ-9 Japanese version developed by Muramatsu et al. 68 The scale had good internal consistency in this study (α = 0.867).

Generalized Anxiety Disorder-7-Japanese version
The General Anxiety Disorder-7 (GAD-7) 69 comprises seven items related to generalized anxiety symptoms.Participants rated items such as "Feeling nervous, anxious, or on edge" on the same Likert scale as the PHQ-9.Total scores could range from 0 to 21, with 10 as the cutoff for the presence of generalized anxiety disorder.We used the Japanese version of the GAD-7. 70The scale had good internal consistency in this study (α = 0.899).

Liebowitz Social Anxiety Scale-Japanese version
The Liebowitz Social Anxiety Scale (LSAS) 71 is a 24-item self-reported measure of social anxiety symptoms over the past week.Items were related to various social situations, and participants rated the extent that they feared and avoided the situation separately on a Likert scale of 0 to 3. The total scores could range between 0 and 144, and higher scores suggested more social anxiety.Asakura et al. 72 developed and validated the Japanese version.The scale had excellent internal consistency in this study (α = 0.972).

Camouflaging Autistic Traits Questionnaire (CAT-Q)-Japanese version
This 25-item measure was designed to quantify camouflaging for autistic people. 11The scale has three subscales: masking, compensation, and assimilation.The masking subscale contains items related to hiding autistic characteristics and pretending to non-autistic (e.g., "I always think about the impression I make on other people").The compensation subscale includes items regarding strategies to compensate for difficulties in social situations (e.g., "When I am interacting with someone, I deliberately copy their body language or facial expressions").
Assimilation contains strategies to fit in with others in social situations (e.g., "In social situations, I feel like I'm 'performing' rather than being myself.").Items were rated on a seven-point scale (strongly disagree to strongly agree).Total scores range from 25 to 175.
Those with higher scores camouflaged more.The Japanese version of Hongo et al. 73 was reliable and valid, and the scale had good internal consistency in this study (α = 0.888).

Warwick-Edinburgh Mental Wellbeing Scale-Japanese version
Participants' mental well-being was assessed using the Warwick-Edinburgh Mental Wellbeing Scale (WEMWBS), which has 14 items. 74Participants rated items such as "I've been feeling relaxed" on a five-point Likert scale ranging from "none of the time" to "all of the time."The total scores could be between 14 and 70, with higher scores suggesting more positive mental well-being.We used the WEMWBS's Japanese version that Suganuma et al. 75 developed, which has excellent internal consistency (α = 0.905).

Stigma consciousness scale
The stigma consciousness scale comprised five items related to perceived stigma 65 adapted by Perry et al. 8 for autistic people.Participants rated items such as "Stereotypes about autistic people have not affected me personally" (reversed item) and "Most people do not judge someone based on their being autistic" (reversed) on a Likert scale of 0 (strongly disagree) to 3 (strongly agree).The total score could be between 0 and 15, with higher scores indicating a stronger consciousness of social stigma.Since there was no Japanese version of this scale, we translated it into Japanese.It had moderate internal consistency (α = 0.606).

Individualistic strategy use
We translated the 13-item scale of individualistic strategy use 42 adapted for autistic people. 8icipants reported the extent that they agreed with each item on a seven-point Likert scale.
The scale includes three subscales: "concealing status" (e.g., "I try to hide my autistic traits from others in certain situations"), "deny/minimize" (e.g., "I don't think of myself as an autistic person."),and "attempt to overcome" (e.g., "I do not need to be "cured" of autism;" reversed item).The total score could be between 13 and 91, and those with higher scores used individualistic strategies more frequently.It had poor internal consistency (α = 0.598).

Collective Strategy Use
We modified and translated the collective strategy use scale 42 into Japanese. 8The scale comprised 13 items related to pride in the autistic community, "I have a lot of pride in the autistic community", or social change activism, "I am an autistic rights activist."Each item was rated on a seven-point agreement scale, with total scores ranging from 13 to 91.Those who scored high used collective strategies more frequently.It had excellent internal consistency (α = 0.905).

Design and data analysis
Being a quasi-replication of Perry et al. 8 , this study had a cross-sectional correlational design.We performed all statistical analyses using SPSS statistics version 28.0 and mediation analyses using PROCESS add-on version 3.3. 76For hierarchical linear multiple regression analyses, we created two dummy codes for gender: "male versus female" and "male versus non-binary."We set male as the reference category and coded it zero because most participants identified as male.
To test hypothesis one, we performed a hierarchical linear multiple regression analysis to test whether perceived stigma significantly correlated with camouflaging.We set camouflaging as the dependent variable and entered demographic variables (gender, age, age at diagnosis, and autistic traits) as independent variables at Step 1, and perceived stigma at Step 2. All demographic variables have been previously found to be related to camouflaging and were therefore controlled for in our analyses. 77In examining the extent that stigma affects camouflaging, we compared Cohen's f 2 of stigma in our study with that of Perry et al. 8 We considered f 2 values of .02 as a small effect, .15 as a medium effect, and .35as a large effect. 78For comparison, we used the values showed in Perry et al. 8 since we did not have access to the dataset used by them.
We tested hypothesis 2 by performing a hierarchical linear multiple regression analysis with camouflaging as the dependent variable.We entered demographic variables as above at Step 1 and individualistic and collective strategy use at Step 2 as independent variables.
The "masking" subscale of the CAT-Q and the "concealing status" subscale of the individualistic strategy scale contain similar items; thus, the degree of correlation between the two may be affected by the overlap of items between the scales.To examine this effect, we also performed the above analysis replacing the individualistic strategy scale with that without the "concealing status" subscale.
To test hypothesis 3, we used four mediation models.In all models, stigma was the independent variable, and camouflaging was the mediating variable.The dependent variables were mental well-being (Model 1), depression (Model 2), generalized anxiety (Model 3), and social anxiety (Model 4).
As this was a quasi-replication study, we employed the same analysis method as Perry et al. 8 However, in testing Hypothesis 1 and 2, we used hierarchical multiple regression analysis instead of multiple regression analysis, to determine how much variance in camouflaging was accounted for by perceived stigma or coping strategies.We considered two-tailed p-values of .05 as significant and 95% confidence intervals (CIs) as appropriate.
Using G*power and referring to the effect size in Perry et al. 8 , the adequate sample size for hierarchical multiple regression analyses (f 2 = 0.06, α = .05,1-β = 0.80) was 234.The variance inflation factor was between 1.05 and 3.61 for hierarchical multiple regression analyses, indicating no multicollinearity.

Results
All variables used in the analysis were normally distributed with skewness and kurtosis in the range of -2 to 2. 79 Levene's tests showed that there was homogeneity of variance across all variables.For mediation analyses, variables were normally distributed, and variance for all variables was homogeneous.Further, no missing values existed.The average, SD, range, kurtosis, and skewness of all variables are presented in Table 1.The correlation matrix is also presented in Supplemental Table 1.

Hypothesis 1: Perceived stigma and camouflaging
The results of hierarchical multiple regression analysis are presented in Table 2.In Steps 1 and 2, the model accounted for 9.2 and 17.2% of the variation in camouflaging, respectively.
The likelihood ratio test showed that the goodness-of-fit model of Step 2 was significantly better than that of Step 1 (likelihood ratio  2 (1) = 26.30,p < .001).Stigma was a significant predictor of camouflaging; greater stigma scores suggested greater camouflaging.Moreover, identifying as female or non-binary, younger age, and older age at diagnosis were significantly related to higher degree of camouflaging.Cohen's f 2 of stigma (0.09) was small, and comparable to that in Perry et al. 8 (0.07).

Hypothesis 2: Camouflaging and strategy use
The results of hierarchical multiple regression analysis are presented in Table 3A.In Steps 1 and 2, the model accounted for 9.2% and 19.4% of the variation in camouflaging, respectively.The likelihood ratio test showed that the goodness-of-fit model of Step 2 was significantly better than that of Step 1 (likelihood ratio  2 (2) = 33.07,p < .001).Greater individual and collective strategy use and identifying female or non-binary were significant predictors of more camouflaging.
The results of the same analysis replacing the individualistic strategy use scale with that without "concealing status" subscale is shown in Table 3B.In Steps 1 and 2, the model accounted for 9.2% and 9.9% of the variation in camouflaging, respectively.The likelihood ratio test revealed that the goodness-of-fit model of Step 2 was not significantly better than that of Step 1 (likelihood ratio  2 (2) = 2.13, p = .345).Older age at diagnosis and identifying as female or non-binary were significant predictors of more camouflaging.However, the correlation between the CAT-Q and the individualistic strategy use scale (without the "concealing status" subscale) was not significant.

Discussion
This study examined the relationship between perceived stigma, camouflaging, and the mental health of Japanese autistic people using the SIT framework.Replicating Perry et al. 8 , we found that higher perceived stigma contributed to more camouflaging, and both individualistic and collective strategy use were positively related to camouflaging.
Camouflaging did not mediate the relationship between stigma and mental well-being; however, it mediated the relationships between stigma and depression, generalized anxiety, and social anxiety.One point to note is that these findings should be interpreted with caution because the internal consistencies of the three questionnaires in this study (the AQ-J-10, the stigma consciousness scale and the individualistic strategy use scale) are not good.
Although causal relationships cannot be inferred from our findings, the results support the hypothesis that camouflaging is a response to stigma.In perceiving stigma, autistic people might be motivated to camouflage and avoid bullying, abuse, or rejection by others. 7,35This finding aligns with that in Perry et al. 8 and previous discussions. 4,80The fact that stigma is closely related to camouflaging is of great importance.Often, support available for autistic individuals is based on the biomedical model and focuses on reducing "maladaptive behaviors" by intervening with being autistic. 81,82Contrastingly, the current findings support the view of the social model, which considers the difficulties of people with disabilities to be due to mismatches between their characteristics and the environment. 83To reduce the negative role of camouflaging, it is essential to create societies where autistic people are accepted as they are and not forced to become "less autistic."Although there is still no consistent knowledge of factors associated with reducing autism-related stigma, several previous studies have suggested that accurate knowledge about autism and positive contact experiences with autistic individuals could reduce stigma. 84,85In Japan, where knowledge of autism is significantly less than in the West, 56 education about autism will play an especially important role.
We hypothesized that in Japan, where the level of autism-related stigma is higher and inter-dependence is emphasized, stigma is more facilitative of camouflaging.Contrary to our hypothesis, there was no difference in the effect size of perceived stigma on the level of camouflaging between Japan and the West.This finding showed that the relationship between stigma and camouflaging could be generalizable across different cultural groups.
Investigations of camouflaging outside of the West have only just begun.In addition, there is still insufficient insight into the differences in how people perceive autism and their attitudes toward autistic individuals depending on their cultural backgrounds, such as an emphasis on inter-dependence.Further studies are warranted to examine the differences in contributing factors of camouflaging owing to cultural differences.
Regarding hypothesis 2, individualistic strategy use was a positive predictor of camouflaging, supporting the hypothesis that camouflaging is highly related to individualistic strategies and a response to stigma.Further, collective strategy use-a concept that theoretically opposes individualistic strategy use-was also positively related to camouflaging: people with stronger and more positive autistic identity were more likely to camouflage.This finding aligned with that of Perry et al. 8 , although Cohen's f 2 in our sample (0.017) was relatively smaller than that in the Western sample (0.042).Further study is required to determine why collective strategy use positively predicted camouflaging; however, one possibility is the "double bind" in which autistic people are caught.For example, using mediation analysis, Cage and Troxell-Whitman 86 showed that higher autistic identity contributed to less camouflaging via autistic individuals disclosing their diagnosis.However, when disclosure was controlled for, higher autistic identity directly increased camouflaging.This competitive mediation suggests that people with higher autistic identity have a dilemma; they want to disclose their diagnosis and stop camouflaging; however, it is hard to do so.Cage and Troxell-Whitman 86 cited fear of stigma as a reason they cannot stop camouflaging.
Disclosure of autistic status could foster to further stigma 45,87 and negative labeling. 33n environment wherein disclosing an autistic diagnosis is unsafe, autistic people could have no choice but continue to camouflage, even if they are proud of being autistic.
Notably, those who mostly use collective strategies might have different motivations to camouflage than those who use individualistic strategies.They might perceive, but do not internalize autism-related stigma, and consider camouflaging simply as a tool to get along with others.Therefore, the psychological burden of camouflaging could be relatively small for them.The finding that collective strategy use was positively correlated to positive wellbeing in this study suggests this.This study did not consider differences in the impacts of camouflaging by motives, and further study is needed.
As for hypothesis 3, perceived stigma was significantly and negatively related to mental well-being, depression, generalized anxiety, and social anxiety for autistic people.
Moreover, camouflaging mediated the relationship of stigma with depression, generalized anxiety, and social anxiety completely or partially.This suggests that an indirect pathway exists in which stigma impairs autistic people's mental health, although stigma also directly affects generalized anxiety and social anxiety.These findings offer a new perspective on interpreting mental health problems in autistic individuals, supporting the minority stress model. 45rastingly, we found that camouflaging did not mediate the relationship between stigma and mental well-being.Moreover, camouflaging was not related to mental well-being significantly in our analyses.One possible reason for camouflaging not being related to mental well-being is that well-being is a more general concept that includes positive feelings, life satisfaction, autonomy, good relationships with others, and positive self-esteem. 88,89itative studies have suggested that camouflaging links to a sense of career accomplishment, good relationships with colleagues and friends, and pride in masking skills 4,12,16 while contributing to depression and exhaustion, 4,6 thus undermining any perceived "successful camouflaging" and off-setting any possible benefits of masking. 90her reason is that we used a standardized measure of well-being that could not fully capture autistic-specific aspects of well-being, which could differ from neuro-normative conceptualizations. 91Our finding on well-being is consistent with that in Perry et al. 8 , contrary to the hypothesis that camouflaging is negatively related to mental well-being in Japan, where camouflaging requires greater mental resources.The relationship between camouflaging and the overall well-being of autistic individuals has rarely been examined and warrants further investigation.It is also important to consider whether these offsets occur within or between individuals.

Limitations
This study had several limitations.First, because this was an online survey, we could not ascertain whether participants had a formal diagnosis of autism.One participant reported the scales to include similar items.However, the overlap of items might result in significant correlation in the multiple regression analysis.
Fourth, this study employed a cross-sectional design, and we could not clarify causal relationships between variables.There are criticisms that most studies on camouflaging are cross-sectional, 77,[97][98][99] and some studies suggest reverse causality between some variables (e.g., pre-existing anxiety could promote camouflaging; camouflaging reaffirms the stigma that autistic characteristics must be hidden). 98,100To resolve this limitation of cross-sectional studies, longitudinal studies assessing perceived stigma, camouflaging, and mental health of autistic individuals are needed.
Finally, examining the relationships between camouflaging and mental health, we focused on camouflaging behavior itself and did not consider contexts for camouflaging or consequences of camouflaging.Some autistic individuals might consider camouflaging simply a skill for survival, and camouflage consciously.They might camouflage only in certain situations, such as at job interviews, and be themselves around people who accept their autistic characteristics.In that case, they might gain more confidence in their adaptability and more positive well-being by camouflaging.However, for those who believe that they must be "normal" and camouflage unknowingly, camouflaging would have significant negative impacts.They might mask their characteristics at every turn, even with their families.In addition, whether camouflaging works effectively may also affect the relationships between camouflaging and mental health.If camouflaging does not result in good social adaptation, it will not promote social well-being or resilience, 101 but rather will only consume mental resources.The CAT-Q does not take the contexts and effectiveness of camouflaging into account, and this study did not involve scales that evaluate social adaptation or social well-being.Researchers recently began studying the factors influencing the impact of camouflaging, 16 and further quantitative research is needed.

Conclusion
This study replicated the findings in Perry et al. 8 It showed that camouflaging is closely related to stigma among Japanese autistic adults.Beyond replication, it also showed that camouflaging mediates the association of stigma with depression, generalized anxiety, and social anxiety.These findings show how stigma is deeply problematic for autistic people across different cultural groups and highlight the importance of focusing on the social environment to reduce stigma and the negative role of camouflaging on mental health of autistic people.Patient Health Questionnaire-9; GAD-7: Generalized Anxiety Disorder-7; LSAS: Liebowitz Social Anxiety Scale; AQ-J-10: Autism-Spectrum Quotient 10-item short version; CAT-Q:

Figures
Figures 1A-1D present the mediation diagram of each model.

Table 3A
Hierarchical multiple regression model examining the predictive value of individualistic and collective strategy use, age, age at diagnosis, gender, and autistic characteristics in predicting camouflaging