Cognitive‐Behavior Therapy for Social Anxiety Does Not Increase Reports of Camouflaging Behavior in Autistic Adults: Results From an Exploratory Study

ABSTRACT

Camouflaging has been defined as the masking or compensation of autistic traits during social interactions, often as a response to the stigma associated with autism and social expectations. Social anxiety has been closely linked to camouflaging, as autistic individuals may adopt camouflaging strategies to navigate social expectations and to reduce the risk of being negatively evaluated. Understanding the relationship between social anxiety and camouflaging in autism is essential for developing effective clinical interventions. This study investigated the effects of a modified Cognitive Behavioral Therapy (CBT) group intervention, the Engage Program, on camouflaging and social anxiety. Specifically, the objective was to determine whether the CBT program would reduce social anxiety symptoms and if this reduction would also be associated with improvements in camouflaging behaviours. Seventy‐one autistic adults participated in an 8‐week modified CBT group intervention for social anxiety. Camouflaging behaviors were assessed using the CAT‐Q scale, and social anxiety levels were measured using three established and validated self‐report measures. Pre‐ and post‐intervention scores were analyzed to determine change following treatment, and correlations between social anxiety and camouflaging measures were explored. As expected, the CBT intervention program significantly reduced social anxiety symptoms across all measures, showing moderate effect sizes from pre‐ to post‐treatment. However, no significant changes in camouflaging behaviors were observed from pre‐ to post‐treatment. Despite this, reductions in social anxiety symptoms were correlated with decreases in camouflaging behaviours, particularly for the compensation and assimilation subscales. Autistic participants who showed the most benefit from therapy on social anxiety measures also showed the greatest reduction in their camouflaging scores from pre‐ to post‐treatment. These findings suggest that improvements in social anxiety symptoms from CBT are also associated with reductions in camouflaging. Importantly, CBT did not lead to an increase in camouflaging behaviors in autistic adults. This study supports the overall benefits of CBT for autistic adults and suggests a need for more randomized controlled studies.

Summary.

• The Engage Program shows that autistic adults can reduce social anxiety after a group‐therapy intervention without increasing their need to hide or mask their autistic traits.

• This research emphasizes the importance of providing support that builds social confidence while respecting each person's natural way of interacting.

• The findings highlight the need for tailored support programs that address social anxiety while respecting neurodiversity, filling a critical gap in available interventions.

1. Introduction

Camouflaging is the broad term used to describe the masking or compensation of autistic traits during social interactions (Cook et al. ^2021a; Hull et al. ²⁰¹⁷; Lai et al. ²⁰¹⁷; Livingston and Happé ²⁰¹⁷). Social camouflaging strategies may consist of hiding certain behaviors that are judged as socially unacceptable or explicitly displaying a different conduct to appear socially competent (Cook et al. ^2021a; Gould ²⁰¹⁷; Green et al. ²⁰¹⁹; Lai et al. ²⁰¹⁷). It can include conscious efforts to conform to societal norms and perform a social mask (Hull et al. 2017; Livingston and Happé ²⁰¹⁷), but it may also presume unconscious suppression of autistic traits (Lawson 2020), possibly because of internal or external stigma (Cook et al. ^2021a; Pearson and Rose ²⁰²¹; Perry et al. ²⁰²¹).

Camouflaging consists of three key components: compensation, masking, and assimilation, which together offer a thorough framework for understanding the various strategies autistic individuals use to navigate social interactions (Hull et al. 2019; Williams ²⁰²¹). Compensation involves imitating social cues, masking focuses on concealing autistic traits, and assimilation aims to blend in by adjusting behavior to fit social norms. These strategies highlight the effort involved in managing social perceptions and expectations. It is important to note that not all autistic individuals engage in camouflaging behaviors, and some may not always mask their autistic traits in different social contexts or relationships (Cook et al. 2021; Evans et al. ²⁰²³; Hull et al. ^{2017, 2020}).

Petrolini et al. (2023) conducted a comprehensive review of the literature on camouflage, and the findings suggest that the experience of camouflaging for autistic individuals is more nuanced and complex than initially thought. For instance, while some autistic individuals considered camouflaging as an unpleasant but necessary behavior to navigate social settings, others highlighted its importance in facilitating social interactions (Libsack et al. 2021; Cook et al. ²⁰²¹). Indeed, Ai et al. (2022) have recently suggested that camouflaging may mirror the self‐presentation management commonly seen in neurotypicals, but with unique mechanisms and consequences within autistic populations. In this way, camouflaging has been linked to social anxiety, where some aspects of the behavior may be related to fears of negative evaluation, discrimination, and judgment (Lei et al. 2023).

Many autistic individuals report that masking their autism can be mentally and emotionally exhausting (Hull et al. 2017; Nel ²⁰²⁴; Pearson and Rose ²⁰²¹). They often experience cognitive dissonance, discomfort, and fatigue (Hull et al. 2017; Nel ²⁰²⁴; Pearson and Rose ²⁰²¹). Camouflaging to meet social expectations has been associated with increased anxiety (Hull et al. 2021) and elevated stress linked to navigating a stigmatized minority identity, with negative mental health outcomes potentially arising from experiences of discrimination and social stigma (Botha and Frost 2018; Mitchell et al. ²⁰²¹; Turnock et al. ²⁰²²). While camouflaging may co‐occur with social anxiety, it is not clear whether these relationships are causally related in either direction. That is, increased camouflaging behaviors may also lead to social anxiety and vice versa, or there may be no causal relationship between the two. It is also possible that broader contextual factors, such as stigma and discrimination, might separately contribute to both camouflaging and social anxiety and their overlap.

Camouflaging has been linked to anxiety, stress, depression, and a diminished sense of authenticity (Bargiela et al. 2016; Beck et al. ²⁰²⁰; Bernardin, Lewis, et al. ²⁰²¹; Bernardin, Mason, et al. ²⁰²¹; Cage and Troxell‐Whitman ²⁰¹⁹; Hull et al. ²⁰¹⁹). While some studies describe camouflaging as an adaptive coping strategy that enables individuals to navigate social situations, achieve personal goals, and engage more effectively with others (Bradley et al. 2021; Hull et al. ²⁰¹⁷; Miller et al. ²⁰²¹; Petrolini et al. ²⁰²³), other studies highlight detrimental psychological effects, including heightened distress and poorer mental health outcomes (Bradley et al. 2021; Evans et al. ²⁰²³; Lawson ²⁰²⁰).

Despite a growing body of research on camouflaging, the literature remains in its early stages and presents mixed findings regarding its association with mental health symptoms. While some quantitative studies link masking to increased anxiety and depression (Bernardin, Lewis, et al. ²⁰²¹; Bernardin, Mason, et al. ²⁰²¹; Hull et al. ²⁰¹⁹), others find no significant association with depression (Bernardin, Lewis, et al. ²⁰²¹; Bernardin, Mason, et al. ²⁰²¹; Cage and Troxell‐Whitman ²⁰¹⁹; Cassidy et al. ²⁰¹⁹; Evans et al. ²⁰²³). However, masking is consistently associated with psychological distress and higher neuroticism (Boyd et al. 2011; Cage and Troxell‐Whitman ²⁰¹⁹; Dell'Osso et al. ²⁰²¹; Lai et al. ²⁰¹⁷; McKenzie ²⁰⁰⁷).

Qualitative studies reveal the emotional toll of sustained camouflaging (Bradley et al. 2021; Field et al. ²⁰²⁴). Autistic individuals described it as exhausting and isolating, with lasting impacts on mental and physical health, identity, and self‐acceptance. The most harmful aspect suggested by some participants in qualitative studies was how ingrained camouflaging can become, leaving some feeling unable to function without it. Those who no longer felt the need to mask have described this shift as freeing, highlighting the critical role of societal acceptance in supporting mental health. Regarding support for social anxiety specifically, research supports the effectiveness of social anxiety and skills‐based interventions for autistic adults, though the field is in its infancy, particularly for adults (Steinbrenner et al. 2020). Meta‐analyses suggest these programs, particularly those using cognitive‐behavioral therapy (CBT), can improve social engagement, with interventions like the Multimodal Anxiety and Social Skills Intervention (MASSI) enhancing social responsiveness and communication (Spain et al. 2017; White et al. ²⁰¹²). Open‐label studies have also suggested CBT is effective in reducing social anxiety in autistic individuals (Bemmer et al. 2021).

According to cognitive‐behavioral models, social anxiety is maintained by safety behaviors, which are cognitive or behavioral strategies employed to reduce perceived social‐evaluative threat without leaving the social situation (Wong and Rapee 2016). These include avoidance behaviors (e.g., avoiding eye contact) and impression management strategies (e.g., rehearsing conversations or trying to appear more sociable), which can inadvertently sustain anxiety over time (Clark and Wells 1995; Heimberg et al. ²⁰¹⁰). Some researchers have argued that camouflaging is conceptually similar to safety behaviors and is associated with heightened anxiety and depression (Hull et al. 2021; Lei et al. ²⁰²³). Camouflaging may also be more complex, shaped not only by internal fears of negative evaluation but also by stigma and exclusion (Lei et al. 2024). Camouflaging may serve a broader range of functions for autistic individuals, including strategizing around social expectations, increasing safety in social settings, or facilitating access to opportunities (Khudiakova et al. 2024). Camouflaging may also be necessary or even protective in some contexts, complicating the assumption that its reduction is uniformly beneficial.

More importantly, there is a growing debate about whether social anxiety treatments that incorporate skills‐based elements reduce the need for camouflaging. It is also plausible that incorporating skills‐based programs could entrench camouflaging behaviors and interfere with intervention gains by limiting emotional authenticity, therapeutic openness, or opportunities to challenge perceived social threats. The implications of this shift are not yet fully understood, as it may enhance authenticity and well‐being for some, but for others, it may also increase vulnerability to social risk (Cooper et al. 2022).

Camouflaging may not be solely about anxiety regulation in autistic populations; it may also involve identity management, self‐protection, or strategic social presentation, particularly in settings where being openly autistic may feel unsafe (Miller et al. 2021). In some contexts, reduced pressure to camouflage may allow for more authentic, less effortful, and potentially more satisfying social interactions, but this might depend on the environment being accepting and inclusive (Cooper et al. 2022; Lei et al. ^{2024, 2025}). Thus, further exploration is needed into how and when camouflaging functions adaptively versus maladaptively, and how interventions can support both psychological well‐being and social safety. The supportive group environment provided by these programs also plays a crucial role, offering a safe space to practice new skills and receive peer support, which can enhance participants' confidence and reduce anxiety (Bemmer et al. 2021). However, the efficacy of these studies has been brought into question in recent years, particularly when considered through the neurodiversity lens (Ashman et al. 2017; Ke et al. ²⁰¹⁷; Monahan et al. ²⁰²¹; Spain and Blainey ²⁰¹⁵). The understanding of camouflaging and the potential negative impacts of traditional social anxiety and skills programs has evolved, leading to a shift in how these interventions are perceived and implemented. Researchers argue that these interventions have the risk of focusing on “teaching” individuals to conform to neurotypical social norms and subsequently increasing camouflaging behaviors and poor mental health (Bottema‐Beutel et al. ²⁰¹⁸; Monahan et al. ²⁰²¹). This possible instruction in masking may have a detrimental impact on an individual's psychological development, leading to stigma, shame, and interpersonal trauma (Bottema‐Beutel et al. ²⁰¹⁸; Leaf et al. ²⁰²¹; Sandoval‐Norton et al. ²⁰¹⁹).

At the same time, social structuring can be a valuable component of therapy, offering a clear framework for navigating social interactions while respecting individual differences. This structured approach supports confidence and skill development in a manner that respects individual differences and reduces the risk of reinforcing camouflage behaviors. By fostering a supportive environment rather than enforcing neurotypical norms, social structuring can encourage genuine social engagement and personal growth (Bellini and Peters 2008; Linden et al. ²⁰²²; Wilson and Gullon‐Scott ²⁰²⁴).

This study aims to investigate camouflage behaviors both before and after a modified CBT group intervention targeting social anxiety and confidence in autistic adults. The primary objective was to assess changes in camouflage behaviors among participants following the intervention. Additionally, we examined social anxiety levels before and after the intervention and evaluated if changes in social anxiety were related to changes in camouflaging behaviors. This approach provides an opportunity to understand how interventions targeted at addressing social anxiety within the autistic community may impact camouflage behaviors. All analyses and hypotheses reported here are considered exploratory, given that the primary function of these groups is to reduce social anxiety symptoms.

We hypothesized that the intervention would lead to a significant reduction in social anxiety levels, consistent with findings from Bemmer et al. (2021). Additionally, we predicted a decrease in camouflaging behaviors following the intervention and anticipated a positive correlation between reductions in social anxiety and decreases in camouflaging, suggesting that lowering social anxiety symptoms is associated with a corresponding decrease in camouflaging behaviors. We also considered an alternative hypothesis that persistent camouflaging behaviors may limit the effectiveness of social anxiety interventions. That is, individuals who continue to engage in high levels of camouflaging may benefit less from treatment, potentially due to barriers in emotional expression, social authenticity, or cognitive burden. Understanding how these processes interact is critical for informing future adaptations of interventions that are better aligned with the lived experiences of autistic adults.

2. Method

This study was a pre‐post‐examination of a CBT group intervention designed for autistic adults with social anxiety. The intervention, known as the “Engage Program”, was adapted to include social skills components tailored specifically for autistic adults. All social skills elements were implemented to support participants in carrying out exposure‐based tasks and testing their beliefs in real‐life social situations, ensuring they had the necessary skills and confidence to engage with the therapeutic process effectively. Approval for this study was granted by the University of Sydney Human Research Ethics Committee (Approval No. 2015/365). Before their involvement in the study, all participants provided written informed consent.

2.1. Participants

The study recruited a total of 71 eligible individuals who completed both pre‐ and post‐CBT group assessments. Recruitment took place between January 2022 and March 2025, drawing participants from clinical referrals, community sources, local Headspace centres (offering mental health services for 12–25‐year‐olds), and referrals to the Clinic for Autism and Neurodevelopmental (CAN) Research at the Brain and Mind Centre, located at the University of Sydney. Inclusion criteria involved participants seeking help and either (1) receiving an ASD diagnosis within the past 12 months through the Autism Diagnostic Interview‐Revised (ADI‐R) or Autism Diagnostic Observation Schedule‐2 (ADOS‐2) or (2) undergoing the ADOS‐2 at study entry and meeting the cut‐off for autism spectrum on Module 4. Participants were also required to be at least 16 years old and to have completed both pre‐ and post‐intervention assessments.

The exclusion criteria for participant eligibility encompassed several factors: the presence of suicide risk or severe depression, a FSIQ score below 70 as determined by the Wechsler Test of Adult Reading (WTAR), the identification of active psychosis during the intake assessment, acute mental health issues, limited proficiency in English, substance abuse, and significant impairment in vision or hearing that might impede participation in the audio/visual aspects of the program. Participants had the option to withdraw before the group sessions started or at any point during the study. Participants who missed more than three treatment sessions were excluded from this data analysis (n = 5). The study sample consisted of 71 participants, with a mean age of 25.3 years (SD = 5.80), ranging from 17 to 40 years old. The mean Full‐Scale IQ of the sample was 111 (SD = 7.15). Gender distribution was diverse, with 45.1% identifying as male, 39.4% as female, 7.0% as other, and 8.5% preferring not to disclose their gender. A total of 13 intervention groups were conducted at the University of Sydney Brain and Mind Centre, with an average of 6 individuals in each group.

2.2. Intervention Program—The Engage Program

The modified CBT program in this study, “Engage Program,” was based on established treatments for social anxiety in adults (Heimberg 2002; Hofmann and Otto ²⁰⁰⁸; Rapee et al. ²⁰⁰⁹) and adapted to meet the needs of autistic adults, who may struggle with traditional cognitive interventions due to alexithymia and differences in social processing (Cooper et al. 2018; Kinnaird et al. ²⁰¹⁹). The eight‐week program (Bemmer et al. 2021) incorporated structured approaches to building social confidence, covering conversation skills, assertiveness, and conflict management. The primary therapeutic emphasis was on addressing social anxiety, with social skills support scaffolding the exposure component of the therapy. CBT elements such as exposure, cognitive restructuring, and in‐session behavioral experiments were included. Cognitive restructuring was simplified to support behavioral interventions, including role‐plays, exposure tasks, and homework exercises (Guastella and Thomas 2020). Participants could invite a social support person to aid in applying CBT strategies outside of sessions. These adaptations align with recommendations for anxiety interventions for autistic adults (Ainsworth et al. 2020; Bemmer et al. ²⁰²¹; Gaus ²⁰¹⁸; Spain et al. ²⁰¹⁷). The program used here has been previously described (see Bemmer et al. ²⁰²¹).

2.3. Measures

2.3.1. Initial Visit

The Autism Diagnostic Observation Schedule‐2 (ADOS‐2) (Lord et al. 2000) is a widely used tool designed to assess autism spectrum in individuals across different age groups. It involves a semi‐structured observational assessment that evaluates behaviors related to social interaction, communication, and imaginative use of materials. In our study, we utilized Module 4 of the ADOS‐2, which is used for verbally fluent older adolescents and adults. Participants were required to meet specific criteria on the ADOS‐2 to qualify for participating in the research. The Wechsler Test of Adult Reading (WTAR) (Wechsler 2001) is a neuropsychological evaluation tool used to estimate Full‐Scale Intelligence Quotient (FSIQ). It assesses participants' ability to read aloud 50 irregular words of varying difficulty, providing scores with a mean of 100 and a standard deviation of 15, normalized for age.

2.3.2. Camouflaging Behaviors

The Camouflaging Autistic Traits Questionnaire (CAT‐Q) is a self‐report questionnaire measuring camouflaging behaviors with 25 items answered on a seven‐point Likert scale. The questionnaire consists of three subscales: compensation (strategies to compensate for social difficulties), masking (concealing autistic traits or displaying neurotypical behavior), and assimilation (attempts to fit into social situations). The questionnaire has good content validity, good test–retest reliability, and maintains measurement invariance across diagnostic (autistic/non‐autistic) and gender (male/female) groups (Hull et al. 2019).

2.3.3. Social Anxiety and Social Responsiveness

2.3.3.1. Liebowitz Social Anxiety Scale—Self‐Report (LSAS‐SR)

The LSAS‐SR is a 24‐item self‐report questionnaire designed to assess anxiety and avoidance related to social interactions and situations. The self‐report version is widely used in the social anxiety literature due to its strong psychometric properties, as well as in autistic populations (Baker et al. 2002; Boulton and Guastella ²⁰²¹; Oakman et al. ²⁰⁰²). Participants rate their level of fear and anxiety in various social situations, as well as the frequency of avoidance, using a 4‐point Likert scale. Higher scores reflect greater levels of anxiety and avoidance, with a total score of 60 or more indicating generalized social anxiety (Liebowitz 2015; Mennin et al. ²⁰⁰²). The LSAS‐SR provides scores for both fear and avoidance subscales, in addition to a total score.

The LSAS‐SR has demonstrated excellent reliability, with Cronbach's α values of 0.90 for both the fear and avoidance subscales and 0.95 for the total score. It also exhibits good convergent and discriminant validity across clinical and nonclinical populations (Fresco et al. 2001; Oakman et al. ²⁰⁰²). In studies involving autistic adults, the LSAS‐SR has shown robust internal consistency, with Cronbach's α values of 0.94, 0.92, and 0.96 for the fear, avoidance, and total scores, respectively (Kanai et al. 2011).

2.3.3.2. Social Interaction Anxiety Scale (SIAS) and Social Phobia Scale (SPS)

The Social Interaction Anxiety Scale (SIAS) and Social Phobia Scale (SPS) are self‐report measures used to evaluate anxiety related to social interactions. Each scale consists of 20 items, with the SIAS focusing on general social interaction anxiety and the SPS on the fear of being observed or judged. Respondents use a 5‐point Likert scale to indicate how much each statement applies to them. Higher scores on the SIAS indicate increased social interaction anxiety, while higher scores on the SPS reflect greater fear of being negatively judged. A score of 36 or above on the SIAS and 26 or above on the SPS suggests clinically significant levels of anxiety (Peters 2000). These scales also show good convergent and discriminant validity across diverse samples (Mattick and Clarke 1998), including autistic cohorts (Boulton and Guastella 2021). In autistic adults, the SIAS has shown strong internal consistency (Cronbach's α ranging from 0.92 to 0.95) and good convergent validity (Maddox and White 2015; Spain et al. ²⁰¹⁶). The SPS has similarly demonstrated strong psychometric properties in autistic populations, with a Cronbach's α of 0.93 (Maddox and White 2015).

These measures were selected to capture complementary aspects of social anxiety and to address ongoing uncertainty about the most appropriate assessment tools for autistic adults. All three have been previously used in autistic populations, providing a precedent for their use in this context. The LSAS‐SR was employed as the primary outcome measure, as it assesses both fear and avoidance across a range of social and performance situations, with an emphasis on observable behavioral avoidance. It is also recognized in regulatory contexts (e.g., FDA) as a gold standard for treatment evaluation in anxiety research. The SIAS and SPS, though originally developed for neurotypical populations, offer valuable insight into the subjective experience of social fear, specifically, anxiety in interpersonal interactions (SIAS) and fears of being observed or judged by others (SPS). Together, these measures allowed for a multidimensional assessment of social anxiety and improved the robustness and interpretability of our findings in an autistic sample.

2.3.3.3. Social Responsiveness Scale‐2—Adult Self‐Report (SRS‐2)

The SRS‐2 (Constantino and Gruber 2012) is a comprehensive 65‐item instrument designed to assess social functioning and autistic traits in adults. It comprises five subscales, “Awareness,” “Cognition,” “Communication,” “Motivation,” and “Restricted Interests and Repetitive Behaviours” (RRB), in addition to an overall score. Raw scores from the SRS‐2 are transformed into T scores to reflect the severity of social difficulties: ≤ 59 indicates typical functioning, 60–65 suggests mild difficulties, 66–75 points to moderate issues, and ≥ 76 denotes severe challenges. The scale demonstrates excellent internal consistency, with a Cronbach's alpha of 0.94.

2.4. Procedures

Before the eight‐week intervention, participants completed self‐report measures assessing social behavior, trait severity, and mood, including the CAT‐Q (Hull et al. 2019) for camouflaging. Assessments were conducted both pre‐ and post‐intervention. Pre‐intervention measures were administered approximately 1 week before the start of the 8‐week program, and post‐intervention measures were completed during the final week of the intervention. Before the group sessions, participants also attended an initial visit to the Brain and Mind Centre, where they completed the Autism Diagnostic Observation Schedule, 2nd Edition (ADOS‐2), along with other cognitive assessments (detailed in the measures section). These measures were selected based on their reliability in tracking relevant outcomes and their prior use in evaluating traits and treatment responsiveness in autistic adults (Boulton and Guastella 2021; Park et al. ²⁰¹⁹; Spain et al. ²⁰¹⁸).

The modified CBT program included groups of six to eight participants across eight weekly three‐hour sessions, led by two clinically trained facilitators with extensive experience working with autistic adolescents and adults. After each session, facilitators held a 60‐minute debriefing to review participant goals. Each session began with a review of homework and prior topics, allowing participants to discuss their practice of the introduced skills. Facilitators provided adjustments and additional support as needed. Homework varied, including tasks like making phone calls, tracking thoughts and feelings, or planning social activities. After core CBT and social skills training, new homework was assigned. The final 30 min of each session were designated as “coffee time,” during which participants could practice skills in a realistic setting with the support of the facilitators. For further details on the CBT program structure, see Bemmer et al. (2021).

2.5. Community Involvement

Recognizing the importance of incorporating input from autistic community members in research, we consulted the CAN Research Community Consultation Committee throughout the design and implementation of this project. This committee includes autistic adults, including some who had previously participated in the group program described here. They provided valuable feedback on the format and delivery of the intervention and have co‐authored related publications about the CBT program. The committee also reviewed and approved the study aims and design, ensuring they aligned with the priorities and needs of the autistic community. Their involvement helped enhance the relevance and accessibility of the research.

2.6. Statistical Analyses

We conducted an a priori power analysis using G*Power 3 (Faul et al. 2009) to assess differences between two paired‐sample group means using a two‐tailed test, a medium effect size (d = 0.40), and an alpha of 0.05. A total sample size of 41 pairs was necessary to achieve a power of 0.80.

All analyses were two‐tailed, with alpha set at 0.05. For statistical analyses, we used the Statistical Program for Social Science (SPSS), version 26. Data were visually inspected for normality, skewness, and kurtosis values, and Shapiro–Wilk's test of equality of variance was used. The assumption of normality was violated for several variables, including the SRS total scores, SIAS total score, CAT‐Q total score, and the CAT‐Q subscale for compensation (all p < 0.005). As a result, non‐parametric Wilcoxon signed rank tests were conducted for these variables to ensure the accuracy and reliability of the findings.

Paired‐sample t‐tests or Wilcoxon signed‐rank tests were employed to compare pre‐treatment to post‐treatment scores on self‐report questionnaires. Cohen's d was calculated to determine effect sizes, with accepted cut‐offs of 0.2 (small), 0.5 (medium), and 0.8 (large) (Cohen 1984). Correlations between camouflaging behaviors and social anxiety measures were examined using Pearson correlation coefficients to explore relationships between these constructs. To minimize the number of tests conducted, we only looked at total scores for the social anxiety measures and the SRS. Bonferroni adjustments were applied, and p values < 0.006 (0.05/8) were considered statistically significant.

3. Results

Descriptive statistics were computed for age, WTAR estimated IQ (FSIQ) and sex within the sample. Independent‐samples t‐tests revealed no statistically significant differences in age (t(69) = 1.32, p = 0.191) or IQ (t(68) = −0.498, p = 0.620) between males and females. Similarly, there was no statistically significant difference in the distribution of males and females (biological sex assigned at birth) within the complete sample (χ ²(1, N = 71) = 0.014, p = 0.906).

Table 1 presents the descriptive results for the CAT‐Q, LSAS‐SR, SIAS, SPS, and SRS and the results of paired samples t‐tests and Wilcoxon signed‐rank tests pre‐ and post‐intervention. These findings indicate statistically significant reductions in social anxiety and social phobia post‐intervention, with moderate effect sizes across measures. Additionally, enhancements in social performance and social responsiveness were observed, as evidenced by a decrease in the total SRS score.

TABLE 1.

Combined table of descriptive statistics and paired samples tests.

Variable	Pre‐group mean (SD)	Post‐group mean (SD)	Statistic	p	Effect size	Test
CAT‐Q—Total	120.9 (26.25)	118.7 (25.34)	1418	0.210	0.17	Wilcoxon signed
CAT‐Q—Compensation	39.9 (12.61)	39.9 (12.87)	970	0.63	−0.06	Wilcoxon signed
CAT‐Q—Masking	39.9 (8.23)	39.1 (8.24)	1.162	0.249	0.14	Student's t
CAT‐Q—Assimilation	41.1 (9.41)	39.8 (8.39)	1.869	0.066	0.22	Student's t
LSAS‐SR—Total	79.0 (28.78)	67.7 (26.61)	5.15	< 0.001	0.61	Student's t
SPS—Total	34.8 (19.95)	28.1 (16.08)	5.79	< 0.001	0.68	Student's t
SIAS—Total	63.7 (15.60)	57.9 (15.49)	1883	< 0.001	0.61	Wilcoxon signed
SRS—Total	75.3 (8.46)	72.8 (8.17)	1551	< 0.001	0.54	Wilcoxon signed

Note: The test was Student's t for parametric data, and Wilcoxon signed for non‐parametric data. The effect size was measured by Cohen's d when the test was Student's t and was measured by rank biserial correlation when the test was Wilcoxon signed.

Abbreviations: CAT‐Q, Camouflaging of Autistic Traits Questionnaire; LSAS‐SR, Leibowitz Social Anxiety Scale‐Self Report; SIAS, Social Interaction Anxiety Scale; SPS, Social Phobia Scale; SRS‐2, Social Responsiveness Scale 2nd Edition.

The primary aim of this study was to evaluate changes in camouflaging behaviors among autistic adults following a modified CBT group intervention designed to address social anxiety. Analysis of camouflaging, as measured by the CAT‐Q total scores, revealed no changes pre‐ and post‐intervention (Wilcoxon signed = 1418, p = 0.210, effect size = 0.17). The compensation subscale did not exhibit significant change (Wilcoxon signed = 970, p = 0.639, effect size = −0.07); similarly, the masking subscale (t(70) = 1.162, p = 0.249, d = 0.14) and the assimilation subscale (t(70) = 1.869, p = 0.066, d = 0.22) also did not show statistically significant change. While there was no statistically significant change from pre‐ to post‐intervention, the direction of the mean change suggested that, if anything, the CAT‐Q scores were more likely to decrease. Of the 71 participants, 64.8% exhibited a reduction in total CAT‐Q scores post‐intervention. Subscale‐specific decreases were observed in 50.7% (compensation), 52.1% (masking), and 66.2% (assimilation) of participants.

In line with our first hypothesis, significant reductions in social anxiety were observed across multiple measures. The LSAS‐SR total score showed a substantial decrease after the intervention (t(70) = 5.15, p < 0.001, d = 0.61). The SPS total score also showed a significant decline (t(51) = 5.79, p < 0.001, d = 0.68), while the SIAS total score demonstrated a marked reduction (t(70) = 4.81, p < 0.001, d = 0.57).

In terms of social engagement and communication, the SRS‐2 revealed significant improvements. The total SRS score decreased significantly (t(69) = 3.82, p < 0.001, d = 0.45), indicating enhanced social responsiveness. Overall, the results show that the modified CBT group intervention led to significant reductions in social anxiety levels, with moderate effect sizes observed across multiple measures.

To explore the relationship between changes in social anxiety and camouflaging behaviors following the intervention, we conducted correlation analyses of change (pre‐ minus post‐treatment scores), as presented in Table 2. We hypothesized that there would be a positive correlation between reductions in social anxiety and a decrease in camouflaging, suggesting that lowering social anxiety would be associated with a corresponding decline in camouflaging behaviors.

TABLE 2.

Correlations between change in social anxiety (LSAS‐SR, SPS, SIAS) and social functioning (SRS), and change in camouflaging (CAT‐Q).

	CAT‐Q—Total	CAT‐Q—Compensation	CAT‐Q—Masking	CAT‐Q—Assimilation
LSAS‐SR—Total	0.404***	0.329**	0.212	0.442***
SPS—Total	0.204	0.128	0.076	0.317**
SIAS—Total	0.516***	0.465***	0.276**	0.487***
SRS—Total	0.442***	0.506***	0.238*	0.242*

Note: *p < 0.05, **p < 0.01, ***p < 0.001.

Abbreviations: CAT‐Q, Camouflaging of Autistic Traits Questionnaire; LSAS‐SR, Leibowitz Social Anxiety Scale‐Self Report; SIAS, Social Interaction Anxiety Scale; SPS, Social Phobia Scale; SRS‐2, Social Responsiveness Scale 2nd Edition.

The results confirmed this hypothesis, revealing significant correlations between changes in the LSAS‐SR total score and reductions in camouflaging behaviors as measured by the CAT‐Q total score (r(69) = 0.404, p < 0.001). More specifically, significant correlations were observed between changes on the LSAS‐SR total score and changes on the CAT‐Q subscales of compensation (r(69) = 0.329, p < 0.05) and assimilation (r(69) = 0.442, p < 0.001), indicating that reductions in social anxiety were associated with decreases in these specific aspects of camouflaging. However, the correlation between the LSAS‐SR total score and the CAT‐Q masking subscale did not reach statistical significance (r(69) = 0.212, p > 0.076).

Similarly, changes in the SIAS total score were associated with changes in the CAT‐Q total score (r(69) = 0.516, p < 0.001), the compensation subscale (r(69) = 0.465, p < 0.001), the masking subscale (r(69) = 276, p < 0.05) and the assimilation subscale (r(69) = 0.487, p < 0.001), further suggesting that reductions in social anxiety are associated with decreases in camouflaging behaviors.

Considering social integration, changes in the SRS‐2 total score were significantly associated with changes in the CAT‐Q total score (r(69) = 0.442, p < 0.001), indicating that as social and communication interaction challenges decreased, camouflaging behaviors also decreased.

To examine the relationship between changes in camouflaging and improvements in social anxiety, we conducted a series of multiple regression analyses, adjusting for baseline LSAS scores to account for regression to the mean. In the first model, we entered the change in camouflage as the predictor and baseline LSAS as a covariate. The model was significant, F(2, 68) = 18.06, p < 0.001, R ² = 0.35. Both change in CAT‐Q (B = 0.44, p < 0.001) and baseline LSAS (B = 0.28, p < 0.001) were significant predictors of LSAS change scores, suggesting that reductions in camouflaging were associated with greater reductions in social anxiety, independent of initial anxiety levels. In the second model, we tested whether baseline CAT‐Q predicted LSAS change. The overall model was significant, F(2, 68) = 8.72, p < 0.001, R ² = 0.20. However, baseline CAT‐Q was not a significant predictor (B = 0.11, p = 0.27), while baseline LSAS remained significant (B = 0.22, p = 0.02).

To further investigate the relationship between improvements in social anxiety and changes in camouflaging behaviors, we conducted an exploratory analysis comparing CAT‐Q scores changes between the bottom 25% of responders, who showed no improvement (M = −11.35, SD = 8.10) in social anxiety symptoms (as measured by LSAS), and the top 25% of responders, who experienced most improvement (M = 35.56, SD = 10.98) in social anxiety symptoms. An independent samples t‐test revealed that participants in the top responder group showed significantly greater reductions in total camouflaging (M = −6.62, SD = 14.35) compared to the bottom responder group (M = 12.22, SD = 19.97); t(37) = −3.42, p = 0.002, d = −1.10. Significant group differences were also observed across all three CAT‐Q subscales. For Assimilation, top responders had a greater reduction (M = −2.14, SD = 6.13) than bottom responders (M = 3.44, SD = 4.60); t(37) = −3.17, p = 0.003, d = −1.03. Masking scores similarly decreased more for top responders (M = −1.33, SD = 5.87) compared to bottom responders (M = 3.67, SD = 7.60); t(37) = −2.33, p = 0.025, d = −0.75. For Compensation, top responders showed a greater reduction (M = −3.14, SD = 6.47) than bottom responders (M = 5.11, SD = 12.48); t(37) = −2.38, p = 0.022, d = −0.76. These findings suggest that greater reductions in social anxiety were associated with broader and more substantial decreases in camouflaging behaviors. To aid interpretation of these results, we present Supporting Information figures comparing mean changes in CAT‐Q total and subscale scores (assimilation, masking, and compensation) across LSAS responder groups (bottom 25%, middle 50%, and top 25%) (see Supporting Information, pp. 1–4). These patterns support the observed associations between reductions in social anxiety and decrease in camouflaging, although no evidence of non‐linear relationships was found (see Supporting Information, pp. 5–6).

4. Discussion

The present study aimed to investigate the effects of a modified CBT group intervention on camouflaging as measured by the CAT‐Q scale (Hull et al. 2019). To our knowledge, this is the first study to investigate camouflaging levels before and after a CBT group intervention for autistic adults. Understanding camouflaging is crucial due to its profound implications for the mental health of this population (Evans et al. 2023; Hull et al. ^{2017, 2021}). As hypothesized, the intervention led to a reduction in social anxiety symptoms. However, contrary to expectations, there was no statistically significant reduction in camouflaging behaviors as measured by the CAT‐Q. While the direction of change across some camouflaging subscales suggested a trend toward reduction, these changes did not reach significance. Importantly, the intervention did not increase overall camouflaging behaviors, which is a reassuring finding given concerns that social skills interventions may inadvertently promote masking or concealment of autistic traits (Bottema‐Beutel et al. ²⁰¹⁸; Monahan et al. ²⁰²¹). We then explored whether there was an association between changes in camouflaging behaviors and improvements in social anxiety symptoms. As expected, results showed that reductions in social anxiety were significantly correlated with decreases in certain aspects of camouflaging, particularly in the areas of compensation and assimilation. This suggests that while the intervention did not broadly change camouflaging behaviors, individuals who experienced the greatest reduction in social anxiety symptoms were more likely to show decreases in specific camouflaging strategies.

To further investigate the relationship between changes in camouflaging and improvements in social anxiety, we conducted additional regression analyses adjusting for baseline LSAS scores to account for regression to the mean. These analyses revealed that reductions in camouflaging remained significantly associated with reductions in social anxiety even after controlling for initial anxiety levels. The persistence of this effect after controlling for baseline social anxiety also supports the conceptual argument that reductions in camouflaging may signal a meaningful shift in self‐presentation or social coping strategies that align with those who received the most therapeutic gains in social anxiety symptoms. These findings reinforce the value of further investigating camouflaging as a modifiable, though complex, behavioral correlate of social anxiety in autistic adults.

The persistence of camouflaging behaviors in this study, despite reductions in social anxiety, may indicate that some camouflaging behaviors are not maintained by immediate anxiety, but by longer‐term adaptations to chronic experiences of exclusion and stigma (Cage and Troxell‐Whitman ²⁰¹⁹; Hull et al. ²⁰¹⁷). Thus, reductions in social anxiety may reflect increased self‐efficacy or decreased fear of social judgment within the group context, without necessitating the unlearning of camouflaging strategies that may have historically protected individuals from harm or rejection. This pattern contrasts with typical outcomes in CBT for social anxiety in non‐autistic populations, where safety behaviors are conceptualized as anxiety‐maintaining and are expected to reduce alongside symptom improvement (Clark and Wells 1995). In autistic adults, however, camouflaging may serve broader, more enduring functions beyond anxiety regulation, including adapting to social norms and seeking inclusion in predominantly neurotypical spaces (Botha and Frost 2018; Hull et al. ²⁰²¹; Mitchell et al. ²⁰²¹; Turnock et al. ²⁰²²). As such, even when anxiety is alleviated, the perceived need to camouflage may persist, particularly if external social pressures remain unchanged. This suggests that camouflaging is not always responsive to the mechanisms targeted in traditional CBT and emphasizes the importance of contextual, identity‐affirming approaches in intervention design.

Another explanation for why camouflaging behaviors did not significantly change from pre‐ to post‐intervention is that camouflaging may be a deeply ingrained coping mechanism that is not easily altered by short‐term interventions. Camouflaging behaviors may persist due to their long‐standing nature or ongoing social and environmental pressures (Cook et al. 2021). These findings align with prior research indicating the complex nature of camouflaging (Cook et al. 2021; Libsack et al. ²⁰²¹; Petrolini et al. ²⁰²³). For some individuals, camouflaging may become intrinsic over time, evolving from initially induced behaviors (e.g., to fit in socially) into habitual responses. Similarly, navigating social contexts with specific rules may require adaptive behaviors such as learning to adjust actions through feedback sensitivity and response modulation, without explicit instruction on social interaction strategies. Consequently, attempting to address and modify camouflaging behaviors within a brief intervention period can be challenging, as these behaviors may be deeply ingrained and influenced by various contextual factors (Bradley et al. 2021; Cage and Troxell‐Whitman ²⁰¹⁹; Cook et al. ²⁰²¹; Hull et al. ²⁰¹⁷). This underscores the importance of considering the nuanced nature of camouflaging and the potential limitation of short‐term intervention approaches in facilitating meaningful change (Petrolini et al. 2023).

To further explore the interplay between social anxiety and camouflaging behaviors, we conducted an exploratory analysis comparing participants who showed the greatest reductions in social anxiety with those who showed the least improvement. This analysis was guided by our dual hypothesis. First, that reductions in social anxiety would be associated with decreases in camouflaging; and second, that persistent camouflaging behaviors might interfere with the effectiveness of social anxiety interventions. Results show that participants who benefited most from the intervention also showed the greatest declines in total camouflaging scores and across all three subscales, indicating that alleviating social anxiety may relieve some of the pressure to camouflage in social situations. However, the pattern observed in the bottom 25% of responders, who showed no anxiety improvement and, in some cases, increased camouflaging, offers preliminary support for the alternative hypothesis, that persistent camouflaging behaviors may act as a barrier to therapeutic progress.

Traditionally, CBT interventions are based on top‐down techniques, utilizing cognitive restructuring and behavioral modification based on predetermined frameworks. However, Petrolini et al. (2023) suggest a shift toward more relational and bottom‐up approaches, specifically for modified interventions for autistic adults. The “Engage Program” focuses on developing and reinforcing relevant social skills in a flexible way. In other words, participants in the program were encouraged to engage in conversations and social situations that were relevant for each context within the group setting, creating a scaffolding for the development of social skills and coping mechanisms. Our findings demonstrated a significant decrease in social anxiety, highlighting the effectiveness of this relational and more flexible approach in addressing the social challenges faced by autistic adults.

One of the most important aspects of the Engage Program is the peer connection, which facilitates meaningful interactions with other autistic individuals. This aligns with findings by Evans et al. (2023), which demonstrate the effects of participating in the autistic community, where peer connection is associated with lower levels of camouflaging, increased self‐esteem, and a stronger sense of authenticity. However, unlike Evans' findings, our study did not observe significant changes in camouflaging behaviors immediately following the intervention. One possible explanation is that many participants start the intervention feeling isolated and are often still coming to terms with their autism diagnosis. For most, this is their first opportunity to interact meaningfully with other autistic individuals. While the program provides a safe and supportive environment where participants can explore their identities and connect with peers, these processes may take longer to translate into reductions in camouflaging behaviors.

While the structured components of the intervention, such as cognitive restructuring, behavioral experiments, and psychoeducation, were designed to target social anxiety, it is also important to consider the potential impact of the social context in which the intervention was delivered. The group setting provided participants with regular opportunities for peer interaction in an accepting environment, such as during coffee breaks and other unstructured social times. Given the high levels of social isolation reported by many autistic adults, it is possible that simply participating in a welcoming, autistic‐majority space contributed meaningfully to reductions in social anxiety and camouflaging behaviors. Thus, reductions in social anxiety and specific camouflaging behaviors may reflect not only the content of the intervention but also the protective effects of increased social connectedness and a supportive neurodivergent social environment. Relational aspects of group‐based interventions, such as peer validation, shared identity, and social safety, may play a critical role in therapeutic outcomes for autistic adults (Crane et al. 2020; Valderrama et al. ²⁰²²).

In this study, we hypothesized that social anxiety would improve following the CBT group intervention and that this improvement would correlate with a reduction in camouflaging behaviors. The results supported this hypothesis, as significant reductions in social anxiety levels were observed across all measures post‐intervention. Consistent with findings from Bemmer et al. (2021), we observed decreases in social anxiety and improvements in social responsiveness following the intervention, with moderate effect sizes, indicating the intervention was able to reduce social anxiety and improve social responsiveness. As social anxiety decreased, so did camouflaging across multiple subscales, suggesting that interventions that address social anxiety may have a dual benefit of reducing the need for assimilation. This finding aligns with prior research (e.g., Bemmer et al. ²⁰²¹; Hull et al. ²⁰²¹; Lei et al. ²⁰²³), which emphasizes the role of avoidance behaviors in maintaining both social anxiety and camouflaging.

The significant correlations between LSAS and the CAT‐Q subscales of Compensation and Assimilation suggest that individuals with higher social anxiety may be more likely to engage in camouflaging behaviors that involve actively modifying or adapting their social behavior to fit in. However, the absence of a significant correlation between social anxiety and the Masking subscale suggests that masking, which involves concealing autistic traits rather than actively imitating or assimilating, may be driven by different factors (Cage and Troxell‐Whitman ²⁰¹⁹; Hull et al. ²⁰¹⁷). Further research is required to address this. The correlation between reductions in LSAS total scores and decreases in CAT‐Q total and assimilation subscale scores highlights a key mechanism: as social anxiety diminishes, individuals may feel less pressure to conform to neurotypical norms or suppress their autistic traits. This supports the idea that camouflaging is, at least in part, a coping strategy driven by fear of negative evaluation, one that lessens as social anxiety is addressed (Bernardin, Lewis, et al. ²⁰²¹; Bernardin, Mason, et al. ²⁰²¹; Black et al. ²⁰²²; Han et al. ²⁰²¹; Hull et al. ²⁰²¹). The strong association between LSAS reductions, particularly given its focus on avoidance, and decreases in CAT‐Q total scores is especially notable. Avoidance, a core feature of social anxiety, often reinforces the need to camouflage in social settings (Lei et al. 2023). By reducing avoidance behaviors through CBT, individuals may gain confidence in social interactions without relying on masking strategies. This suggests that addressing avoidance is key to both reducing camouflaging and fostering more authentic self‐expression.

Additionally, significant correlations between reductions in SIAS scores and decreases in CAT‐Q compensation, assimilation, and masking subscales indicate that as individuals become less anxious about social interactions, they may rely less on compensatory strategies such as rehearsed scripts and observational learning. This aligns with research showing that those who engage in camouflaging tend to experience heightened anxiety (Cage and Troxell‐Whitman ²⁰¹⁹; Hull et al. ²⁰²¹). As anxiety decreases, individuals may find it easier to express themselves naturally, alleviating the cognitive and emotional strain of camouflaging (Lorenz and Hull 2024).

Overall, these findings reinforce the importance of addressing social anxiety as a pathway to reducing camouflaging behaviors. They also highlight the potential for CBT to not only alleviate social anxiety but also promote greater authenticity and self‐acceptance, particularly in autistic individuals who often experience a significant social and emotional toll from prolonged masking. Future research could explore whether there are reductions in camouflaging over time and whether they lead to improvements in broader outcomes such as self‐esteem and quality of life.

4.1. Strengths and Clinical Utility

The demographic factors investigated in the study were not considered statistically significant in influencing therapy response. There were no clear barriers to treatment, such as age‐related or gender‐specific differences, which provide a foundation for further research and potential application of the intervention in broader clinical settings. The demographic characteristics of our sample show a diverse representation of genders and sexes, and the mean age of participants (25.1 years) suggests that the intervention was tested across a range of adult ages, enhancing the generalizability of our findings.

In clinical practice, the demand for treatments that are both resource‐efficient and effective is paramount, especially considering the prevalence and diversity of psychiatric comorbidities within autistic populations (Moore et al. 2021; Shattuck and Grosse ²⁰⁰⁷). The short‐term group intervention utilized in this study offers a cost‐effective approach to intervention, requiring fewer therapist resources while fostering peer learning and support dynamics. Moreover, group interventions play a vital role in addressing social anxiety and enhancing peer interaction, as they provide opportunities for supported exposure to challenging social situations and development of more confident, authentic social communication (Bemmer et al. 2021).

4.2. Limitations

This study has several limitations that should be considered when interpreting the findings. First, the absence of a control group limits the ability to directly attribute the observed pre‐ and post‐treatment changes in camouflaging and social anxiety symptoms to the group intervention alone. We suggest that future studies employ randomized controlled trial designs to provide stronger evidence for the efficacy of the intervention. Additionally, the study did not examine the long‐term maintenance of treatment effects. Including follow‐up assessments would allow for a longitudinal understanding of the findings over time.

Another consideration is the generalizability of the results. Participants meeting criteria for intellectual disability (ID) were excluded from the study, leaving a gap in understanding how the intervention might be adapted or tailored for autistic adults with ID. Future research should explore the development of accessible interventions for this group. Participant dropout in the early sessions (n = 5) was an important point of consideration, which may have influenced the final sample size and generalizability of findings. Additionally, participants who chose to participate in the program were likely those who were open to and interested in CBT, which may not be representative of the broader autistic population, including individuals who are critical of CBT or who have had negative past experiences with it. Future studies may explore factors contributing to early disengagement in the intervention. Moreover, qualitative research could provide valuable insights into the lived experiences of individuals undergoing the intervention.

We note that we relied on the ADOS‐2 Module 4 as part of the inclusion criteria. It is possible that we could have unintentionally excluded some autistic individuals who engage in high levels of camouflaging and, as a result, did not meet criteria on the ADOS. Alternate methods for standardized recruitment may be considered in the future (Dell'Osso et al. ²⁰²¹; Duvekot et al. ²⁰¹⁶; Whitlock et al. ²⁰²⁰).

An important consideration in interpreting the null findings regarding changes in camouflaging behaviors is the scope of the CAT‐Q as a measurement tool and its construct validity (McKinnon et al. ²⁰²⁴). While the CAT‐Q has been instrumental in operationalizing and quantifying key camouflaging strategies, specifically compensation, masking, and assimilation, it may not fully capture the nuance of camouflaging as described by autistic individuals across diverse contexts. Emerging research (e.g., Livingston et al. ²⁰¹⁸; Cook et al. ^2021b) highlights that camouflaging is a multifaceted and context‐dependent phenomenon that can include a wide range of strategies that might not be covered by the CAT‐Q, such as internal self‐regulation, impression management, or context‐specific behavioral scripts. It is therefore possible that some camouflaging behaviors decreased (or even increased) during the intervention in ways not captured by the CAT‐Q subscales. This underscores the importance of using complementary methods, such as qualitative interviews or expanded measurement tools, to more comprehensively assess camouflaging behavior and its dynamic nature over time. Future studies should consider integrating these approaches to better understand the complexity and variability of camouflaging, especially in response to therapeutic interventions.

Although change scores are a useful method for examining individual‐level improvements over time, they are susceptible to certain limitations, including measurement error and regression to the mean. To address this, we conducted multiple regression analyses adjusting for baseline camouflaging levels, which confirmed that changes in camouflaging significantly predicted reductions in social anxiety, even when controlling for initial scores. This strengthens confidence in our interpretation that camouflaging change was meaningfully associated with symptom improvement. Nonetheless, we recognize that the use of change scores still poses interpretative constraints, and future research should consider more advanced longitudinal methods to better capture the temporal dynamics and causal relationships between camouflaging and social anxiety. Finally, while the intervention aimed to reduce social anxiety and promote authentic social engagement, there is ongoing debate about whether social skills programs risk increasing camouflaging behaviors, which are associated with poorer mental health outcomes. Our findings suggest that camouflaging did not change after the intervention, but further evaluation is needed to balance fostering social skills with preserving participants' autonomy and authenticity. Addressing these limitations in future studies will not only strengthen the evidence base for group interventions targeting social anxiety and camouflaging in autistic adults but also ensure that these programs are inclusive, effective, and aligned with the principles of neurodiversity.

In conclusion, this study provides valuable insights into the effects of a modified CBT intervention on camouflaging behaviors among autistic adults with social anxiety. While no significant post‐intervention changes in camouflaging were observed, findings suggest that camouflaging behaviors may decrease as social anxiety declines. The results also highlight the complexity of camouflaging, which is influenced by both individual and environmental factors. Importantly, the intervention did not lead to an increase in camouflaging, indicating it did not unintentionally reinforce masking behaviors.

Future research should explore alternative intervention strategies, utilize larger sample sizes, and incorporate longitudinal follow‐ups to assess the long‐term impact on camouflaging behaviors. These efforts are essential for refining treatment approaches that address the diverse needs of autistic individuals.

Author Contributions

B.B.R., K.A.B., E.E.T., and A.J.G. conceptualized and coauthored the article. All authors contributed to data screening. B.B.R. contributed to the study design, co‐facilitated the program, data extraction, conducted statistical analyses, interpreted the findings, and was involved in manuscript preparation and review. K.A.B. helped in co‐supervision, statistical analyses, interpretation of findings, and review. E.E.T. contributed to assessment, program co‐facilitator, program design and development, and modification of intervention. A.J.G. was involved in program development, supervision, study conception and design, interpretation of findings, and manuscript review. All authors read and approved the final manuscript.

Conflicts of Interest

The authors declare no conflicts of interest.

Supporting information

Data S1: Supporting information.

Roisenberg, B. B. , Boulton K. A., Thomas E. E., and Guastella A. J.. 2025. “Cognitive‐Behavior Therapy for Social Anxiety Does Not Increase Reports of Camouflaging Behavior in Autistic Adults: Results From an Exploratory Study.” Autism Research 18, no. 9: 1896–1909. 10.1002/aur.70103.

Data Availability Statement

The data that support the findings of this study are available on request from the corresponding author. The data are not publicly available due to privacy or ethical restrictions.