Rethinking Gender Differences in Camouflaging: Multivariate Associations Across Compensation, Masking, and Assimilation
Abstract
Background: Studies often report higher camouflaging among autistic women, whereas female-male differences are less consistent outside diagnosed autistic samples. Whether differences persist after accounting for empathizing, systemizing, and autistic traits remains unclear. We examined relationships across Camouflaging Autistic Traits Questionnaire (CAT-Q) strategies in adults without a reported autism diagnosis. Methods: We analyzed digitally collected data from 193 adults (75.1% women; mean age = 22.0 years, SD = 5.53). Multivariate models examined Compensation, Masking, and Assimilation. Sequential models tested Gender differences adjusted for Age, adjustment for Empathizing, Systemizing, and Autistic Traits, and Gender-by-trait interactions. We pooled estimates across 20 imputed datasets. Results: The age-adjusted model showed an omnibus Gender difference across CAT-Q subscales, Wald χ²(3) = 12.75, q = .006. Men reported higher Compensation (β = .596, q = .003) and CAT-Q total scores (β = .470, q = .007), whereas Masking and Assimilation did not differ by Gender. In the trait-adjusted model, the omnibus Gender effect was nonsignificant, Wald χ²(3) = 3.52, q = .368. Empathizing showed the strongest multivariate association with CAT-Q subscales, Wald χ²(3) = 71.89, q < .001; higher Empathizing was associated with lower Compensation (β = −.257, q = .003), Assimilation (β = −.450, q < .001), and total scores (β = −.321, q < .001). Systemizing and Autistic Traits showed no significant main effects after correction. The multivariate Gender × AQ-10 interaction was significant, Wald χ²(3) = 12.34, q = .025. The largest outcome-specific interaction involved Masking, although it did not survive coefficient-level correction (q = .065). Conclusion: Initial differences by Gender, driven by Compensation, attenuated after trait adjustment. Lower Empathizing was associated with greater Compensation and Assimilation, whereas Systemizing and Autistic Traits showed no robust main associations. Exploratory Gender × AQ-10 heterogeneity did not survive outcome-specific correction. Findings support strategy-specific rather than simple gender- or autistic-trait interpretations of CAT-Q scores.
Related
- Replicability of multivariate brain-behaviour associations depends on clinical profile
- Migraine-related differences in connectivity patterns during emotional audiovisual processing revealed by functional connectivity multivariate pattern analysis.
- Associations between multivariate profiles, lifestyle behaviors and reaction time performance in active young adults
- Latent profiles of body composition in older adults with hypertension: associations with physical fitness and differences in blood pressure and fasting plasma glucose
- Sex‐Differences in Infrapatellar Fat Pad Size in Runners After Knee Surgery and Associations With Patellofemoral‐Related Symptoms: TRAIL Cohort Cross‐Sectional Study
BibTeX
@misc{monteiro2026rethinking,
title = {{Rethinking Gender Differences in Camouflaging: Multivariate Associations Across Compensation, Masking, and Assimilation}},
author = {Sara Monteiro and Leonie Ambraß and Elizabeth B. de Sousa Fernandes Perna and Arjan Blokland and Johannes E. A. Stauder},
howpublished = {PsyArXiv},
year = {2026},
month = sep,
doi = {10.31234/osf.io/wrmja_v1},
url = {https://osf.io/preprints/psyarxiv/wrmja_v1/},
note = {CC BY 4.0},
}
RIS
TY - UNPB
TI - Rethinking Gender Differences in Camouflaging: Multivariate Associations Across Compensation, Masking, and Assimilation
AU - Sara Monteiro
AU - Leonie Ambraß
AU - Elizabeth B. de Sousa Fernandes Perna
AU - Arjan Blokland
AU - Johannes E. A. Stauder
PY - 2026
DA - 2026/09/07
PB - PsyArXiv
DO - 10.31234/osf.io/wrmja_v1
UR - https://osf.io/preprints/psyarxiv/wrmja_v1/
AB - Background: Studies often report higher camouflaging among autistic women, whereas female-male differences are less consistent outside diagnosed autistic samples. Whether differences persist after accounting for empathizing, systemizing, and autistic traits remains unclear. We examined relationships across Camouflaging Autistic Traits Questionnaire (CAT-Q) strategies in adults without a reported autism diagnosis. Methods: We analyzed digitally collected data from 193 adults (75.1% women; mean age = 22.0 years, SD = 5.53). Multivariate models examined Compensation, Masking, and Assimilation. Sequential models tested Gender differences adjusted for Age, adjustment for Empathizing, Systemizing, and Autistic Traits, and Gender-by-trait interactions. We pooled estimates across 20 imputed datasets. Results: The age-adjusted model showed an omnibus Gender difference across CAT-Q subscales, Wald χ²(3) = 12.75, q = .006. Men reported higher Compensation (β = .596, q = .003) and CAT-Q total scores (β = .470, q = .007), whereas Masking and Assimilation did not differ by Gender. In the trait-adjusted model, the omnibus Gender effect was nonsignificant, Wald χ²(3) = 3.52, q = .368. Empathizing showed the strongest multivariate association with CAT-Q subscales, Wald χ²(3) = 71.89, q < .001; higher Empathizing was associated with lower Compensation (β = −.257, q = .003), Assimilation (β = −.450, q < .001), and total scores (β = −.321, q < .001). Systemizing and Autistic Traits showed no significant main effects after correction. The multivariate Gender × AQ-10 interaction was significant, Wald χ²(3) = 12.34, q = .025. The largest outcome-specific interaction involved Masking, although it did not survive coefficient-level correction (q = .065). Conclusion: Initial differences by Gender, driven by Compensation, attenuated after trait adjustment. Lower Empathizing was associated with greater Compensation and Assimilation, whereas Systemizing and Autistic Traits showed no robust main associations. Exploratory Gender × AQ-10 heterogeneity did not survive outcome-specific correction. Findings support strategy-specific rather than simple gender- or autistic-trait interpretations of CAT-Q scores.
ER -