Social camouflaging and pragmatic self-perception are key aspects of communication in neurodivergent populations. However, little is known about how these processes manifest in adults with autism, high intellectual ability (HIA), or twice-exceptionality (2e: HIA+autism). This study aimed to examine group-based differences in communicative adaptation and perceived pragmatic competence across these profiles.
Materials and methodsA total of 109 Spanish-speaking adults participated. Social camouflaging was assessed using the Camouflaging Autistic Traits Questionnaire (CAT-Q), and pragmatic self-perception was measured with the Communicative Competence Profile (CCP). Statistical analyses explored differences across groups and by gender.
ResultsSignificant differences emerged in all CAT-Q dimensions. Autistic participants reported the highest levels of camouflaging, followed by the 2e group, while only the compensation subscale distinguished autism from 2e. Regarding pragmatic self-perception, the HIA group scored highest, suggesting stronger communicative confidence, whereas autistic adults scored lowest. The 2e profile consistently occupied an intermediate position, though closer to autism, especially in subtle pragmatic aspects. Gender effects were limited overall but more noticeable within the Autism group.
ConclusionsFindings reveal distinct patterns of social camouflaging and pragmatic self-perception across neurodivergent profiles. Twice-exceptional adults show intermediate but autism-leaning traits, underscoring the complexity of their communicative adaptation. Results emphasize the need for inclusive frameworks that account for diverse communication styles within neurodivergence.
El camuflaje social y la autopercepción pragmática son aspectos clave de la comunicación en poblaciones neurodivergentes. Sin embargo, se sabe poco sobre cómo se manifiestan estos procesos en adultos con autismo, alta capacidad intelectual (HCI) o doble excepcionalidad (2e: HCI+autismo). Este estudio tuvo como objetivo examinar las diferencias entre grupos en la adaptación comunicativa y la competencia pragmática percibida en estos perfiles.
Material y métodosUn total de 109 adultos hispanohablantes participaron en el estudio. El camuflaje social se evaluó mediante el Camouflaging Autistic Traits Questionnaire (CAT-Q), y la autopercepción pragmática se midió con el Communicative Competence Profile (CCP). Los análisis estadísticos exploraron diferencias entre grupos y por género.
ResultadosSe encontraron diferencias significativas en todas las dimensiones del CAT-Q. Los participantes autistas reportaron los niveles más altos de camuflaje, seguidos por el grupo 2e, mientras que solo la subescala de compensación distinguió al autismo del 2e. En cuanto a la autopercepción pragmática, el grupo HCI obtuvo las puntuaciones más altas, lo que sugiere mayor confianza comunicativa, mientras que los adultos autistas puntuaron más bajo. El perfil 2e se situó de manera consistente en una posición intermedia, aunque más cercana al autismo, especialmente en aspectos pragmáticos sutiles. Los efectos de género fueron en general limitados, pero más notables dentro del grupo con autismo.
ConclusionesLos hallazgos revelan patrones diferenciados de camuflaje social y autopercepción pragmática en los distintos perfiles neurodivergentes. Los adultos con doble excepcionalidad muestran características intermedias pero con tendencia hacia el autismo, lo que subraya la complejidad de su adaptación comunicativa. Los resultados enfatizan la necesidad de marcos inclusivos que contemplen estilos de comunicación diversos dentro de la neurodivergencia.
In recent years, attention to neurodiversity has increased significantly within scientific, educational, and clinical communities, reflecting a growing interest in better understanding the heterogeneity of neurodivergent profiles. Within this context, the twice-exceptional profile has gained prominence. This term refers to the coexistence of high intellectual abilities (HIA) with an additional condition such as autism, ADHD, learning difficulties, or mental health-related conditions (Cain et al., 2019; Madaus et al., 2022; Nicpon et al., 2011; Sharkey & Nickl-Jockschat, 2023)
One of the Twice-exceptional combinations that has gained greater visibility – partly due to its dissemination on social media – is the coexistence of HIA and autism. These twice-exceptional profiles present unique challenges stemming from the interaction between intellectual strengths and the features associated with co-occurring (Khan & Khan, 2023), which can manifest in diverse ways. In some cases, HIA may mask or compensate for certain autistic traits, making them less noticeable and leading individuals to develop compensatory strategies that hinder recognition (Reis et al., 2000). In other cases, autistic features may overshadow giftedness, resulting in incomplete or inaccurate diagnoses (Baum et al., 2014; Sharkey & Nickl-Jockschat, 2023). There are also situations in which both conditions obscure one another, leaving the individual's needs and challenges entirely unrecognized (Mullet & Rinn, 2015). Given this variability, it is essential that assessment and intervention processes identify both giftedness and coexisting conditions simultaneously to ensure comprehensive and effective support (Madaus et al., 2022; Ryan & Waterman, 2018).
Within this framework, communication emerges as a domain where challenges may arise in twice-exceptional profiles. Specifically, individuals on the autism spectrum tend to interact and interpret social norms in distinctive ways, leading to communication styles that may diverge from conventional expectations regarding verbal and non-verbal language use (Baron-Cohen & Lombardo, 2017; Uddin, 2022). These characteristics may involve differences or difficulties in interpreting social cues, managing turn-taking in conversations, adapting discourse based on context or interlocutor, and understanding or using inferential language (Mahmoud Al-Zoubi, 2024; Kissine, 2012).
On the other hand, individuals with HIA also display a complex communicative profile, often characterized by an advanced vocabulary and verbal expression. Their conversational style may be overly centered on personal interests, show a tendency toward literal interpretation, and present difficulties in understanding implicit social rules (França-Freitas et al., 2014). Despite having grammatically sophisticated discourse and high linguistic creativity, these features can lead to misunderstandings or misinterpretations in social interactions (Gali et al., 2019).
Despite the importance of communication as a cross-cutting domain, few studies have explored the communicative particularities of twice-exceptional individuals. The existing literature suggests a potential tension between the strengths associated with HIA and the challenges related to autism, possibly resulting in a distinct communicative profile (Conejeros-Solar et al., 2018; Sharkey & Nickl-Jockschat, 2023; Uddin, 2022).
These communicative tensions, along with the desire to fit into social contexts that often lack awareness of neurodiversity, may lead many twice-exceptional individuals to develop subtle, often invisible, adaptation strategies. This highlights the relevance of the concept of social camouflaging, defined as the set of strategies used (consciously or unconsciously) by many twice-exceptional individuals to navigate communicative and pragmatic difficulties in social settings (Hull et al., 2017). Camouflaging involves minimizing or concealing neurodivergent traits to better conform to social expectations.
In the case of autism, camouflaging may include memorizing social scripts, imitating facial expressions and gestures, and suppressing self-regulatory behaviors such as repetitive movements (Hull et al., 2021; Livingston et al., 2019; Miller et al., 2021). For individuals with HIA, camouflaging may involve downplaying their intellectual strengths by adjusting their discourse and behavior to meet social norms, potentially leading to over-adaptation that conceals their true abilities and interests (Gómez León, 2020). Among twice-exceptional individuals, camouflaging becomes even more complex, as these strategies can combine and vary depending on the social context and situational demands (Cross et al., 2019; Gómez León, 2020). Moreover, compensatory strategies may begin in early childhood and persist into adulthood, complicating diagnosis, distorting self-perception, and limiting access to appropriate support (Ginapp et al., 2023).
These dynamics of communication and camouflaging not only impact clinical recognition but also influence how individuals adapt their behavior in social settings. Recent research indicates that certain groups, such as women, face additional sociocultural pressures and heightened expectations regarding social and communicative skills (Cain et al., 2019; Sharkey & Nickl-Jockschat, 2023; Uddin, 2022).
Within this context, the present study aims to examine camouflaging strategies among adults with twice-exceptional profiles, autism, and isolated high intellectual ability (HIA), with the goal of identifying potential differences based on diagnostic profiles. Additionally, it seeks to compare communication self-perception across these groups to determine whether differences in pragmatic awareness exist based on diagnosis. A third aim is to explore the potential relationship between social camouflaging and pragmatic self-awareness within each group, and to assess whether the strength or direction of this association differs across diagnostic profiles. A gender perspective is also incorporated to explore whether there are significant differences in camouflaging strategies and communication perception between men and women within each diagnostic group. To address these aims, the study proposes null hypotheses assuming no significant differences between profiles, and alternative hypotheses suggesting the existence of such differences. In line with these aims, the present study formulates the following specific hypotheses:
- 1.
Significant differences will be found across the three diagnostic groups in all dimensions of social camouflaging, with autistic and twice-exceptional adults expected to display higher scores than individuals with high intellectual ability (HIA).
- 2.
The twice-exceptional profile will exhibit an intermediate camouflaging pattern, more closely resembling the autistic group than the HIA group.
- 3.
Regarding pragmatic self-perception, adults with HIA are expected to report the highest levels of communicative competence, followed by the 2e group, while autistic adults will report the lowest levels
- 4.
Gender differences are expected to emerge primarily within the Autism group and, to a lesser extent, within the twice-exceptional group, particularly in variables related to camouflaging.
- 5.
A positive association is expected between social camouflaging and pragmatic self-awareness within the autistic and twice-exceptional groups, with weaker or non-significant associations anticipated in the HIA group.
A total of 109 participants (50 women, 50 men, and 9 individuals with diverse gender identities) took part in this study, with a mean age of 32.1 years (SD=7.29). Participants were divided into three groups based on their neurodivergent profiles. The first group consisted of 46 individuals diagnosed with twice-exceptionality (2e: HIA+autism), including 20 women, 21 men, and 5 individuals with diverse gender identities, with a mean age of 31.3 years (SD=7.0). The second group included 33 individuals with a diagnosis of high intellectual ability (HIA), comprising 20 women and 13 men. Of these, 18 presented with complex talents and 15 with a more balanced gifted profile, with a mean age of 33.2 years (SD=6.8). The third group consisted of 30 individuals diagnosed with autism, including 10 women, 16 men, and 4 individuals with diverse gender identities, with a mean age of 31.8 years (SD=7.98) (Table 1).
InstrumentsTwo questionnaires were used to collect data on the dependent variables: camouflaging strategies and communication self-perception. All questionnaires were administered online via a Microsoft Forms survey due to the geographic dispersion of the sample, and responses were self-administered.
First, the Camouflaging Autistic Traits Questionnaire (CAT-Q), developed by Hull et al. (2019) and preliminarily adapted to Spanish by Mosquera et al. (2022), was used to assess the frequency with which neurodivergent individuals employ social camouflaging strategies. It consists of 25 items grouped into three subscales: (1) compensation, which reflects strategies used to mask communication differences; (2) masking, which includes items related to the suppression of spontaneous, atypical behaviors; and (3) assimilation, which involves actions aimed at blending in or “passing” in social contexts. Each item is rated on a 7-point Likert scale (from “strongly disagree” to “strongly agree”), providing both a global score and subscale-specific scores. The Spanish version underwent a preliminary translation and validation process (Mosquera et al., 2022), suggesting it is suitable for use in exploratory studies such as this one.
Second, the Pragmatic Awareness Questionnaire (CCP), developed by Rodríguez Muñoz (2012), was employed to assess adults’ self-perceived pragmatic competence in everyday communicative situations. The goal of this instrument is to identify differences in awareness of language use, with particular attention to contextual appropriateness, interpretation of communicative intent, use of implicit language, and adherence to social interaction norms. Importantly, the CCP has shown to be useful not only in autistic populations but also in a variety of non-autistic groups, including adults with typical development (Rodríguez Muñoz, 2012), primary school teachers (Rodríguez-Muñoz, 2018), individuals with Williams syndrome (Moraleda Sepúlveda & López Resa, 2024), university students without clinical diagnoses (Zambrano-Cruz et al., 2025), and clinical populations such as Borderline Personality Disorder and Bipolar Disorder (Muriel et al., 2023). These applications demonstrate the broad suitability of the instrument for assessing pragmatic awareness across diverse communicative profiles.
The questionnaire consists of 25 items, phrased as statements about the individual's own communicative behavior, and rated on a 5-point Likert scale (ranging from 1=“very poorly” to 5=“very well”). Item-level interpretation is used in the analysis, and the self-report format makes it particularly well-suited for studies focusing on subjective perceptions of language use in neurodivergent adults.
ProcedureTo conduct the study, an initial contact form was developed using Microsoft Forms. The form was distributed via social media and accounts related to giftedness and neurodivergent communities, aiming to reach Spanish-speaking adults. The inclusion criteria were: being 18 years or older, being a Spanish-speaking adult, and presenting a formal clinical diagnosis of HIA, Autism or 2E. Exclusion criteria included the absence of verifiable diagnostic documentation, self-diagnosis without a professional report, and incomplete or invalid questionnaire responses (e.g., uniform response patterns or atypical completion times). Participants were assured of confidentiality and informed of their right to withdraw from the study at any time without consequence. Upon accessing the form, participants were presented with an information sheet outlining the study's purpose, data use, and participants’ rights.
As part of data quality assurance, clinical profiles were verified through an individual follow-up interview conducted after informed consent was obtained. These interviews were held online via Microsoft Teams, following a standardized protocol to ensure consistency across participants. Each session lasted approximately 25 minutes and allowed the researcher to verify inclusion and exclusion criteria, confirm the existence of a formal clinical diagnosis issued by a qualified professional, and assess functional proficiency in Spanish required to understand the study materials. These brief interviews collected relevant diagnostic history, including age at diagnosis, IQ scores, and the clinical tools used in the assessment process. Participants were also asked to present written diagnostic evidence (e.g., a clinical report or certificate) during the interview.
Diagnostic documentation varied depending on the group. Individuals classified as HIA typically presented results from standardized intelligence tests (most commonly Wechsler scales (Wechsler, 2008) or RIAS-2 Reynolds & Kamphaus, 2015)) and at least one creativity measure such as CREA (Corbalán et al., 2003). Participants in the Autism group provided diagnostic evidence based on instruments such as ADOS-2 (Lord et al., 2012) and ADI-R (Rutter et al., 2003), complemented by questionnaires assessing alexithymia, adaptive functioning, anxiety or depression; in some cases, cognitive testing (e.g., IQ assessments) was also included in their reports. Finally, participants identified as 2E (twice-exceptional) presented documentation covering both domains (HIA and autism) therefore including all the aforementioned measures. This step ensured that all participants had a formal diagnosis provided by a qualified professional, thereby reinforcing the internal validity of the study and excluding unverified cases.
The form included both assessment instruments: first the CAT-Q, followed by the CCP. The estimated completion time was approximately 10 minutes, and data collection remained open from January to early May 2025. To enhance data integrity, the form was configured to allow only one response per device, and periodic monitoring was conducted to detect potential duplicate, incomplete, or inconsistent entries. After closing the data collection phase, all responses were coded, securely stored, and processed in accordance with current data protection regulations. The dataset was exported to a secure spreadsheet and cleaned by removing entries that did not meet the inclusion criteria. The final sample consisted exclusively of adults with confirmed clinical diagnoses and complete responses on both scales.
Statistical analysisData analysis was conducted using R software. Descriptive statistics were first calculated for all variables of interest. For quantitative variables, measures of central tendency (mean, median) and dispersion (standard deviation, interquartile range) were computed.
Next, the normality of quantitative variables was assessed through multiple procedures. The Shapiro–Wilk test was applied (suitable for n≤50), and additional indicators such as skewness and kurtosis coefficients were analyzed. Variables were considered to approximate a normal distribution if skewness and kurtosis values fell within the ±2 range. Visual inspections using histograms and Q–Q plots were also performed to corroborate the statistical findings.
Given that normality assumptions were not met (p<.05), non-parametric techniques were used for inferential analyses. The Kruskal–Wallis test was employed to compare median values across diagnostic groups. When significant overall differences were found, post hoc comparisons were conducted using the Dwass–Steel–Critchlow–Fligner test, which is appropriate for multiple comparisons following a Kruskal–Wallis analysis. Additionally, correlations between variables were examined using Spearman's rho, given the non-parametric nature of the data.
The study protocol received prior approval from the Social Research Ethics Committee of the University of Castilla-La Mancha (Approval Code: CEIS-2025-114946).
ResultsThe results from the Camouflaging Autistic Traits Questionnaire (CAT-Q) provided insight into the use of social camouflaging strategies across its three assessed dimensions: compensation, masking, and assimilation, in addition to the total score. Scores were analyzed based on participants’ diagnostic group (autism, HIA, and twice-exceptional) and gender identity (women, men, and non-binary individuals).
Overall, the highest scores across all three dimensions and the total scale were observed in the Autism group, followed by the twice-exceptional group, while the HIA group showed the lowest scores. Regarding gender, women and non-binary individuals consistently reported higher levels of social camouflaging compared to men.
Table 2 presents the mean scores and standard deviations for each CAT-Q dimension and the total score, disaggregated by diagnostic group and gender identity.
Mean scores and standard deviations by diagnostic group and gender on the CAT-Q.
| Group | Mean | |||
|---|---|---|---|---|
| Women | Man | Non-Binary | Total | |
| Compensation | ||||
| Autism | 79 (36.6) | 28.2 (11.7) | 52.8 (31.8) | 45.5 (30.5) |
| HIA | 18 (2.94) | 11.4 (4.34) | 17 (2.15) | 18.1 (3.33) |
| Twice-exceptionality | 40.2 (17.6) | 21.1 (6.35) | 35 (18.5) | 30.6 (16.1) |
| Masking | ||||
| Autism | 56.5 (30.8) | 30.8 (9.51) | 57.4 (37.5) | 45 (30) |
| HIA | 56.5 (2.18) | 16.1 (4.78) | 14.5 (3.70) | 15.7 (3.18) |
| Twice-exceptionality | 39.9 (12.4) | 22.9 (4.73) | 32.5 (9.47) | 31.1 (12.2) |
| Assimilation | ||||
| Autism | 45.5 (16.9) | 30.8 (11.6) | 84.3 (43.1) | 53.6 (37.4) |
| HIA | 18.8 (3.01) | 9.3 (3.43) | 15 (1.13) | 18.7 (3.53) |
| Twice-exceptionality | 61.4 (11.1) | 25.1 (4.53) | 53.3 (12.8) | 43.3 (19.7) |
| Total | ||||
| Autism | 181 (68.5) | 89.8 (19.9) | 195 (77.8) | 144 (74.6) |
| HIA | 52.4 (5.74) | 33.7 (11.2) | 46.5 (3.96) | 52.5 (7.84) |
| Twice-exceptionality | 142 (33.9) | 69.1 (7.44) | 121 (40.2) | 105 (43.2) |
To determine whether the differences observed between groups were statistically significant, non-parametric analyses were conducted using the Kruskal–Wallis test for each of the CAT-Q dimensions and the total score. The results indicated statistically significant differences among the groups for all variables analyzed (Table 3).
Kruskal–Wallis analysis across the three groups.
| Kruskal–Wallis | Post hoc | ||||
|---|---|---|---|---|---|
| H | p | Group | W | p | |
| Compensation | 47.71 | <.001 | 1–2 | −8.68 | <.001 |
| 1–3 | −3.88 | .017 | |||
| 2–3 | 7.55 | <.001 | |||
| Masking | 64.08 | <.001 | 1–2 | −9.09 | <.001 |
| 1–3 | −2.34 | .222 | |||
| 2–3 | 10.18 | <.001 | |||
| Assimilation | 55.45 | <.001 | 1–2 | −7.79 | <.001 |
| 1–3 | −2.06 | .314 | |||
| 2–3 | 9.98 | <.001 | |||
| Total | 3.42 | .181 | 1–2 | −3.05 | <.001 |
| 1–3 | −1.82 | .051 | |||
| 2–3 | −1.16 | <.001 | |||
Note: Group 1 corresponds to Autism; Group 2 corresponds to HIA; and Group 3 corresponds to Twice-exceptionality.
Specifically, post hoc tests revealed that participants in the Autism group scored significantly higher than those in the HIA group across the total CAT-Q score (p<.001) and in the compensation (p<.001), masking (p<.001), and assimilation (p<.001) dimensions.
Similarly, when comparing the twice-exceptional group (2e) with the HIA group, significant differences were found across all dimensions and in the total score. The 2e group scored higher than the HIA group on the total score (p<.001), as well as on compensation (p<.001), masking (p<.001), and assimilation (p<.001).
In contrast, the differences between the autism and 2e groups were less clear. No statistically significant differences were found between these two groups in the total score (p=.051), masking (p=.222), or assimilation (p=.314). However, in the compensation dimension, the Autism group scored significantly higher than the 2e group (p=.014).
Regarding gender differences within neurodivergent profiles, statistically significant differences were found in the Autism group between men and non-binary individuals in the compensation, masking and total score and between men and women in all dimensions. In the twice-exceptional group, statistically significant gender differences were observed between women and men in compensation, masking and assimilation. No statistically significant differences were found when comparing the rest of the genders in the different groups. These statistical results are presented in Table 4.
Kruskal–Wallis analysis across gender in three groups.
| Autism | HIA | 2e | |||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Kruskal–Wallis | Post hoc | Kruskal–Wallis | Post hoc | Kruskal–Wallis | Post hoc | ||||||||||
| H | p | Group | W | p | H | p | Group | W | p | H | p | Group | W | p | |
| Compensation | 17.23 | <.001 | 1–2 | −5.14 | <.001 | .587 | .745 | 1–2 | .029 | 1 | 13.53 | .001 | 1–2 | −5.151 | <.001 |
| 1–3 | 2.31 | 0.233 | 1–3 | −1.48 | .545 | 1–3 | .289 | .977 | |||||||
| 2–3 | 4.07 | 0.011 | 2–3 | −.28 | .979 | 2–3 | 2.49 | .183 | |||||||
| Masking | 11.83 | .003 | 1–2 | −3.89 | .016 | .497 | .780 | 1–2 | −.147 | .994 | 15.022 | <.001 | 1–2 | −5.33 | <.001 |
| 1–3 | 2.52 | .176 | 1–3 | −1.227 | .661 | 1–3 | .289 | .977 | |||||||
| 2–3 | 3.61 | .029 | 2–3 | −.422 | .952 | 2–3 | 2.49 | .183 | |||||||
| Assimilation | 9.33 | .009 | 1–2 | −3.85 | .018 | 3.341 | .188 | 1–2 | .147 | .994 | 19.99 | .002 | 1–2 | −5.075 | <.001 |
| 1–3 | .304 | .975 | 1–3 | −2.618 | .153 | 1–3 | −.867 | .813 | |||||||
| 2–3 | 3.14 | 0.067 | 2–3 | −2.24 | .250 | 2–3 | 2.129 | ,288 | |||||||
| Total | 16.23 | <.001 | 1–2 | −4.99 | .001 | 3.080 | .214 | 1–2 | −.065 | .999 | 5.14 | .077 | 1–2 | −2.907 | .099 |
| 1–3 | 1–3 | 2.54 | .170 | 1–3 | .169 | .992 | |||||||||
| 2–3 | 3.48 | 0.037 | 2–3 | 2.12 | .291 | 2–3 | 2.103 | .297 | |||||||
Regarding self-perception of communication, the Pragmatic Awareness Questionnaire (CCP) results indicated that 2e individuals perceived more communicative difficulties than those with HIA but fewer than those with autism (Fig. 1).
The Kruskal–Wallis test revealed statistically significant differences among the three groups across all 26 CCP items, with p<.001 in 25 out of 26 items. The only exception was the item on Lexical cohesion, which showed more moderate significance (H=8.66, p=.013). To determine where these differences lay, post hoc comparisons were conducted using the Dwass–Steel–Critchlow–Fligner procedure.
When comparing the autism and HIA groups, statistically significant differences were found across all 26 CCP items. In every case, the p value was ≤.010, with most items showing p<.001; the highest p values in this comparison were observed for Lexical cohesion (W=−4.12, p=.010) and Understanding humour (W=−4.26, p=.007). This pattern suggests a clear distinction between the two groups across most pragmatic competencies.
The comparison between the autism and 2e groups also revealed significant differences in the majority of items. However, several comparisons did not reach statistical significance, including Body posture (W=2.67, p=.142), Movements of hands and arms (W=5.06, p=.397), Lexical cohesion (W=−1.60, p=.494), Interpretation of ambiguous expressions (W=−1.76, p=.427), Intelligibility and paralanguage (W=−8.64, p=.078), Social proxemics (W=−2.95, p=.093), Lexical competence (W=−4.54, p=.253), and Gaze (W=−2.33, p=.227). For the remaining items, Wvalues ranged from 3.94 to 11.15, with p values below .05, indicating broad differentiation between the two groups.
In contrast, the comparison between the HIA and 2e groups revealed a more heterogeneous pattern. While most items showed significant differences (p<.001), several did not, including Lexical cohesion (W=3.23, p=.058), Syntactic cohesion (paragrammatisms) (W=2.44, p=.195), Coherence (textual superstructures) (W=−7.99, p=.195), and Understanding humour (W=0.98, p=.767). These results indicate that certain pragmatic domains show greater similarity between HIA and 2e individuals, whereas others differentiate them clearly. Table S5 (Supplementary Material) presents the relevant statistics from the non-parametric analyses and pairwise comparisons.
Regarding gender, no statistically significant differences were observed across any of the twenty-six items of the CCP questionnaire in Group 2e. Kruskal–Wallis tests conducted on each item yielded p-values greater than .05, indicating that the scores of women, men, and individuals with diverse gender identities did not differ significantly. These findings suggest that, within this group, gender did not constitute a differentiating factor in the competencies assessed (Fig. 2).
In contrast, a single statistically significant gender-related difference was identified in the HIA group, specifically in the item assessing contextual or interlocutor appropriateness (H=6.83, df=2, p=.033). Post hoc analysis revealed that this difference was concentrated in the comparison between men and individuals with diverse gender identities, which was statistically significant (W=−3.789, p=.020). Comparisons between women and men (p=.960), as well as between women and gender-diverse individuals (p=.107), did not reach statistical significance. This result suggests a specific difference in this particular item but does not indicate a generalized pattern of gender-based differences within the HIA group (Tables 5 and 6).
Finally, in the Autism group, statistically significant gender differences were observed in several CCP items. The Kruskal–Wallis test yielded significant results for social proxemics (H=7.710, p=.021), gaze behavior (H=12.491, p=.002), lexical cohesion (H=9.600, p=.008), and humor comprehension (H=6.656, p=.036). No significant differences were found in the remaining items. Post hoc pairwise comparisons indicated statistically significant differences across several combinations. For social proxemics and gaze behavior, differences emerged between women and gender-diverse individuals (p=.021 and p<.001, respectively), with marginal differences observed between men and gender-diverse individuals (both p=.069). For lexical cohesion, the difference was significant between women and gender-diverse individuals (p=.006), while for humor comprehension, the difference was observed between women and men (p=.044). Collectively, these findings suggest that, within the Autism group, gender-related differences (though not generalized) were consistent across multiple items, particularly in comparisons involving women and individuals with diverse gender identities.
To examine the association between camouflaging and pragmatic self-awareness, Spearman's rank-order correlations were computed between the four CAT-Q indices (Compensation, Masking, Assimilation, and the total score) and the CCP items across the three diagnostic groups. In total, 29 significant correlations (p≤.05) were identified (Fig. 3).
In the Autism group, camouflaging showed selective but meaningful associations with several pragmatic dimensions, particularly in nonverbal communication and higher-order discourse skills. The clearest patterns emerged for Compensation, which correlated significantly with Comprehension (r=.471, p=.017), Social proxemics (r=.479, p=.015), and Facial expression (r=.533, p=.006). Masking was also associated with Facial expression (r=.492, p=.013). Assimilation correlated with Linguistic appropriateness to the situation/interlocutor (r=.422, p=.036). Notably, several significant negative correlations appeared for the total CAT-Q score, including Social proxemics (r=−.420, p=.037) and Facial expression (r=−.574, p=.003), suggesting that greater overall camouflaging may relate to reduced pragmatic metacognitive insight. Overall, effects were modest and less consistent than anticipated.
In the HIA group, the associations between camouflaging and pragmatic performance were generally weak and mostly nonsignificant. Only isolated effects emerged, such as the positive correlations between Assimilation and Interactional proxemics (r=.483, p=.004) and Facial expression (r=.399, p=.021), as well as between the total CAT-Q score and Interactional proxemics (r=.384, p=.028). These associations did not form a coherent pattern across pragmatic dimensions.
In contrast, the 2e group exhibited the strongest and most consistent links between camouflaging and pragmatic performance. Compensation correlated positively with Physical contact (r=.346, p=.019), Gaze (r=.388, p=.008), Coherence at the macrostructural level (r=.395, p=.007), Thematic progression (r=.301, p=.042), Maintaining and following up on the conversation (r=.301, p=.042), and Maxim of relevance (r=.306, p=.039). Masking showed significant associations with Coherence (superstructures; r=.303, p=.040), Thematic progression (r=.305, p=.040), and Maxim of relevance (r=.311, p=.035). Assimilation correlated with Physical contact (r=.414, p=.004), Gestures (r=.365, p=.013), Gaze (r=.321, p=.030), Syntactic cohesion/Paragrammatisms (r=.303, p=.040), Coherence (macrostructure; r=.338, p=.021), Thematic progression (r=.305, p=.039), and Maintaining and following up on the conversation (r=.291, p=.050). In contrast, the total CAT-Q score correlated negatively with several pragmatic indicators, including Physical contact (r=−.381, p=.009), Gaze (r=−.382, p=.009), Lexical cohesion (r=−.396, p=.006), and Coherence (superstructures; r=−.315, p=.033), reinforcing the notion that higher global camouflaging may be linked to reduced pragmatic metacognitive insight in this group. The complete correlation matrix is presented in the Supplementary Material.
DiscussionThe results obtained through the CAT-Q questionnaire revealed statistically significant differences across all evaluated dimensions, confirming the influence of diagnosis on camouflaging patterns.
The group with an exclusively autistic diagnosis showed the highest levels of camouflaging in all dimensions compared to the HIA group. This trend aligns with findings reported by Montejano Fluchaire (2023) and Ruggieri (2024), who emphasize that autistic individuals develop a wide repertoire of strategies to mask their communicative differences in order to fit into neurotypical contexts. This intensive use of camouflaging has been associated with a high emotional cost, including fatigue, anxiety, and a diminished sense of authenticity (Hull et al., 2017; Lai et al., 2017).
Although few studies have examined camouflaging and pragmatic abilities jointly, several lines of research point to meaningful links between both constructs. Camouflaging often relies on cognitive processes associated with pragmatic monitoring, contextual interpretation, and conversational adjustment (Hull et al., 2019; Livingston et al., 2019). Likewise, pragmatic research has highlighted the cognitive load required for interpreting social cues, managing implicit meanings, and adapting communication across contexts, skills whose effortful deployment may underpin or intensify camouflaging behaviors (Cola et al., 2022; Uddin, 2022). These converging findings suggest that studying camouflaging and pragmatic awareness together is theoretically coherent and may provide insight into the mechanisms that support social adaptation in neurodivergent adults.
The 2e group, in turn, displayed significantly higher scores than the HIA group across all dimensions but did not differ significantly from the autistic group, except in the compensation dimension, where a significant difference was observed. This suggests that the camouflaging pattern of adults with 2e closely resembles that of the autistic profile, particularly in terms of masking and assimilation, and diverges markedly from the HIA profile. This proximity to the autistic profile could be attributed to the predominant influence of autism spectrum traits on social behavior, prompting these individuals to adopt similar strategies in response to social expectations (Cage & Troxell-Whitman, 2019; Heuving et al., 2025).
The only dimension where significant differences emerged between the 2e and autistic groups was compensation, with autistic individuals reporting higher use of this strategy. This difference may be understood through the distinct social trajectories of each profile. Autistic individuals may have faced greater external pressure to conform to neurotypical norms, leading to the development of more deliberate and visible compensatory strategies (Hull et al., 2017). In contrast, individuals with a 2e profile may have received different social feedback, where certain behavioral traits were interpreted as signs of intelligence rather than communicative differences (Assouline et al., 2012; Pérez Tejera et al., 2017). Thus, it is not necessarily the case that one profile needs to compensate more than another, but rather that the use of these strategies is shaped by context, recognition, and opportunities for identity validation (Miller et al., 2021). Consequently, the lower compensation scores in the 2e group may reflect more internalized, less conscious, or simply different forms of adaptation that are nonetheless significant in their emotional and social impact (Assouline et al., 2012).
These findings point toward a promising avenue for future research: exploring how pragmatic awareness interacts with camouflaging across diagnostic profiles. Examining these constructs simultaneously (through conversational analyses, pragmatic judgment tasks, and self-report measures) may help clarify whether specific pragmatic skills support compensatory behaviors or, conversely, amplify the cognitive and emotional demands associated with camouflaging. Such work would deepen understanding of the adaptive mechanisms underlying social behavior in neurodivergent adults, particularly within twice-exceptional populations.
Taken together, the results reveal a stepped pattern: the autistic profile demonstrates the highest level of social camouflaging, the HIA profile the lowest, and the 2e profile occupies an intermediate position – although notably closer to the autistic end. This distribution suggests that, in social contexts, the influence of autism tends to outweigh that of HIA traits. In this sense, it can be argued that autism more strongly modulates the social behavior of individuals with a 2e profile, shaping an adaptive style that largely mirrors autistic camouflaging (Alaghband-rad et al., 2023).
The third analytical axis focused on the subjective perception of communication, assessed through the Pragmatic Awareness Questionnaire (CCP). The results revealed statistically significant differences in nearly all questionnaire items, underscoring the sensitivity of diagnostic category in shaping communicative self-perception.
The HIA group showed the highest scores, indicating greater confidence in their ability to adapt to communicative environments. However, this does not imply an absence of communicative tensions or misunderstandings. As illustrated by item 12 (interpretation of ambiguous expressions), no significant differences were found compared to the 2e group, suggesting that their communication may not always align with implicit social expectations (Pérez Tejera et al., 2017).
Conversely, the autistic group scored the lowest, reflecting a higher self-perception of difficulty in pragmatic abilities. This does not necessarily equate to overt communication difficulties but rather points to a more challenging communicative experience, likely shaped by repeated instances of misunderstanding, rejection, or persistent effort to meet external expectations (Khan & Khan, 2023; Russell et al., 2025). The autistic group differed significantly from the HIA group in all items, positioning each group within a distinct communicative profile.
The 2e profile, once again, occupied an intermediate position, albeit closer to the autistic group. This intermediate placement may reflect diverse communicative experiences, where certain HIA-associated traits facilitate communication in more structured contexts, while the presence of autistic-like communicative styles may lead to misunderstandings in neurotypical social settings (Melogno et al., 2016). This pattern suggests that, although some skills enable a degree of communicative camouflaging, they do not eliminate social barriers or the burden of striving to be understood (Cage & Troxell-Whitman, 2019; Melogno et al., 2016).
Importantly, beyond these group-level differences, the correlation analyses revealed that camouflaging and pragmatic awareness are not entirely independent processes. Although the strength and consistency of the associations varied by diagnostic profile, the Autism and 2e groups showed the clearest convergences, with higher camouflaging linked to specific pragmatic behaviors and, at times, to lower metacognitive insight. These patterns (consistent with proposals that camouflaging relies on effortful pragmatic monitoring and contextual interpretation (Hull et al., 2019; Livingston et al., 2019; Cola et al., 2022) suggest that both variables may interact in shaping communicative experiences, particularly in adults whose profiles include autistic traits.
It is particularly noteworthy that differences across groups did not appear uniformly across all items. For instance, no significant differences were found between the 2e and autistic groups in items related to body posture, hand and arm movements, lexical competence, ambiguous expressions, and humor comprehension. These areas may represent points of heightened sensitivity within the autistic profile, where even individuals with HIA do not fully compensate for the divergence from social expectations. In contrast, on more structural items related to discourse organization, the 2e group scored higher than the autistic group, likely supported by the cognitive and linguistic strengths associated with HIA (Gali et al., 2019). This trend is also observed in comparisons between the 2e and HIA groups, as no significant differences were found in items related to syntactic cohesion and linguistic appropriateness to the interlocutor, further reinforcing the similarity between the two profiles in these linguistic features (Gali et al., 2019; Seitz et al., 2024).
Overall, these results should not be interpreted as a hierarchy of competencies, but rather as a map of communicative perceptions shaped by individual, contextual, and identity-related factors. From this perspective, the need to revisit normative communicative standards becomes evident, advocating for the recognition of diverse communicative styles and the creation of environments that respect rather than penalize deviations from the norm. In this regard, the 2e profile presents multiple facets, illustrating that compensatory mechanisms do not always fully address the more subtle aspects of social language. In the pragmatic domain in particular, autistic traits tend to outweigh the linguistic strengths associated with HIA. However, in certain contexts, HIA-related abilities may mask communicative differences, potentially contributing to delayed or incomplete diagnoses (Khindey et al., 2025; Leedham et al., 2020).
Beyond the theoretical relevance of these findings, the results also offer practical insights for supporting adults across the three neurodivergent profiles examined. For autistic adults, the high levels of camouflaging and the lower self-perception of pragmatic competence underscore the need for interventions that prioritize communicative authenticity and reduce pressures to perform neurotypical social behavior. Recent work highlights that sustained masking is associated with emotional exhaustion, reduced well-being, and identity conflict (Hull et al., 2017, 2021; Leedham et al., 2020), supporting the value of neurodiversity-affirming practices that promote acceptance of divergent communication styles (Cage & Troxell-Whitman, 2019). Speech-language interventions informed by these approaches may help individuals develop communication strategies that enhance clarity and comfort without reinforcing socially prescribed masking demands.
For adults with high intellectual ability (HIA), the findings suggest that communicative strengths may coexist with subtler pragmatic vulnerabilities, especially in contexts requiring inferential processing, flexibility, or implicit interpretation. Prior research shows that gifted individuals may display literal interpretations, challenges with implicit rules, and difficulties adapting to social norms (França-Freitas et al., 2014; Gali et al., 2019). Therefore, intervention programs targeting metapragmatic awareness, contextual interpretation, and discourse flexibility may help reduce misunderstandings that arise despite their high linguistic proficiency.
Twice-exceptional (2e) adults combine elements of both profiles, presenting communicative patterns that align with autistic individuals in subtle pragmatic dimensions while also drawing on strengths associated with high intellectual ability. Literature on 2e populations emphasizes that cognitive advantages do not necessarily mitigate social-pragmatic demands and may even mask underlying difficulties, delaying recognition and support (Assouline et al., 2012; Heuving et al., 2025). As such, practical interventions for this group should adopt an integrated perspective that differentiates adaptive strategies from emotionally costly camouflaging (Livingston et al., 2019; Miller et al., 2021). Supporting identity validation, reducing compensatory pressures, and facilitating environments that respect communicative diversity may be especially beneficial for adults navigating the dual demands inherent to twice-exceptionality.
Finally, gender plays a critical role in understanding camouflaging patterns. Recent research has consistently shown that autistic women tend to engage in camouflaging more frequently and more elaborately than autistic men (Hull et al., 2019; Lai et al., 2017; Livingston et al., 2019). This heightened social pressure and adaptive demand have been linked to delayed diagnoses and increased emotional strain. Although the present sample limits firm conclusions, incorporating gender-sensitive analyses into future studies may clarify how gender modulates both camouflaging strategies and pragmatic self-perception across neurodivergent profiles.
Despite the relevance of these findings, the interpretation of group and gender-related differences must be approached with caution due to the limited sample size, particularly in gender subgroups within each diagnostic profile. Although the analyses allowed for meaningful comparisons, the reduced number of participants in some categories may constrain the generalizability of the results and the stability of observed effects. Future studies would benefit from larger and more balanced samples to further clarify the patterns identified.
ConclusionTwice-exceptionality (2e) presents unique challenges that often elude traditional diagnostic categories. When both high intellectual capacity (HIC) and autism converge, the resulting profile frequently fails to align with conventional clinical or educational expectations. This discrepancy can hinder early identification and limit access to appropriate support systems.
Within this framework, speech and language therapy plays a critical role – not only through direct intervention targeting communicative abilities, but also by contributing to the recognition and support of diverse identity trajectories. Historically, the field of speech-language pathology has focused primarily on language development; however, its scope has progressively expanded to include pragmatic and socio-communicative dimensions.
In 2e profiles, this pragmatic dimension takes on particular significance. The core issue is not necessarily language impairment, but rather a qualitatively different mode of engaging with the world. As such, intervention requires more than technical expertise – it demands an ethical commitment. This involves shifting from a corrective paradigm to one grounded in listening, respect, and accompaniment.
Such an approach to speech-language therapy does not aim to mold individuals into predefined norms, but rather to equip them with meaningful communicative tools that honor their authentic selves. Understanding these nuances is especially critical in clinical practice, as it enables the development of more tailored, respectful, and effective interventions. These interventions should not be driven by a desire to normalize, but by a commitment to understand and to integrate difference. In doing so, speech and language therapy can become a space where individuals with twice-exceptionality not only improve their communication but also find validation and recognition of their identities.
CRediT authorship contribution statementAll authors have substantially contributed to the study design, data collection and analysis, interpretation of the results, and manuscript writing. Additionally, all have reviewed and approved the final version submitted.
Ethical considerationsThis study was approved by the Social Research Ethics Committee of the University of Castilla-La Mancha (approval code: CEIS-2025-114946). All procedures were conducted in accordance with the Declaration of Helsinki, current legislation, and applicable institutional guidelines.
Informed consentAll participants were informed of the study's objectives and provided written informed consent prior to participation. It is guaranteed that no personal data or images allowing the identification of participants are included in the manuscript.
Declaration of generative AI and AI-assisted technologies in the writing processNo generative artificial intelligence tools were used in the writing or scientific analysis of this manuscript.
FundingThis work was funded by the University of Castilla-La Mancha as part of support for open access research (Open Access funding). The institution had no involvement in the study design, data collection or analysis, manuscript writing, or the decision to submit for publication.
Conflict of interestThe authors declare that there are no financial, personal, or institutional conflicts of interest that could have influenced the results of this research.
Data availabilityThe data supporting the findings of this study are available through the corresponding author upon reasonable request.
Open Access funding provided by the University of Castilla-La Mancha.
The following are the supplementary data to this article:
