Autistic people sometimes camouflage their behaviour to appear non-autistic. This meta-analysis rigorously tests the relationship between autistic traits and camouflaging, examining contributing person- and study-level variables. Following Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines, we searched PubMED, PsycINFO, Web of Science, and ProQuest Dissertations in April 2025. All quantitative designs examining autistic traits and camouflaging on a continuum were included. Fifty papers (N = 16, 895; ages 10-90) contributed to a three-level meta-analysis (accounting for dependent effect sizes). Study quality was medium to high with no evidence of publication bias. Results revealed a moderate association between autistic traits and camouflaging (r = 0. 34, 95% CI: 0. 30-0. 39), comparable across sexes. Depression, but not anxiety or social anxiety, moderated the relationship. Age was not a moderator, but the association was stronger in general population samples (vs diagnosed), with self-reported autism measures (vs observational), and when using the discrepancy method for camouflaging (vs self-report). Among camouflaging subdomains, assimilation showed the strongest effect, followed by compensation and masking. Limited sample diversity constrains inferences across the full spectrum. This meta-analysis provides a clearer understanding of when, why, and how autistic traits are related to camouflaging, with important research and clinical implications. No funding was obtained for this study. Registration: https: //osf. io/uswtr/? viewₒnly=277aec07cdfc402dae75f4900f291253Lay AbstractUnderstanding the autistic trait and camouflaging relationship is critical to identify who is most vulnerable to camouflaging and the way in which autism and camouflaging measurement may influence our understanding of this phenomenon. This directly impacts clinical diagnosis and support, as camouflaging contributes to diagnostic delay and poorer mental health outcomes, creating a cycle of continued camouflaging. Our findings may help to establish the foundation needed to develop targeted interventions. We completed a systematic search to identify all studies that assessed the relationship between autistic traits and camouflaging. In total, 50 studies met all inclusion criteria. The first author extracted data related to participant characteristics (age, gender, diagnostic status, mental health), autistic trait characteristics, and the camouflaging measurement characteristics. The 50 contributing studies included a total of 16, 895 participants (61% female). These data show that the more autistic traits a person has, the more camouflaging they engage in; this relationship is evident for both males and females, and the strength of this relationship does not vary across the adult lifespan. People from the general population show an increase in the strength of this relationship, compared to those diagnosed, and the relationship changes based on how autistic traits and camouflaging are measured and conceptualised. Mental health did not have a clear impact on the overall relationship. There is a nuanced relationship between autistic traits and camouflaging, the strength of which is dependent on specific person-related (diagnostic status and depression) and study-related factors (autistic trait measurement type, camouflaging measurement type, and camouflaging subdomain). Autistic traits are most strongly linked to behaviours that help people to assimilate (try to fit in and appear 'normal'), followed by strategies to compensate for social differences. The act of hiding autistic traits was the least related. Because the relationship between autistic traits and camouflaging was weaker for diagnosed autistic people, further work is needed to test why this occurs. In addition, clinicians must be aware of the potential for camouflaging to disrupt the diagnostic process, and campaigns that aim to reduce stereotypes of autism and promote acceptance of neurodiversity may help to reduce the stigma that drives camouflaging.
Greig et al. (Fri,) studied this question.