Autistic Traits and Camouflaging: A Meta-Analysis.

Every included study but one was cross-sectional, so this shows that camouflaging and autistic traits go together, not which comes first. The samples were mostly female, mostly highly educated and largely from the UK and the USA, with many participants diagnosed in adulthood. The findings may not hold for the whole autistic population, for people with intellectual disability, or for children. Only a small number of the contributing studies included children and adolescents, so nothing firm can be said about younger ages. The studies disagreed with each other far more than chance would explain. The authors calculate that a new study could find anything from a large positive link to a small negative one, so the single average figure hides a lot of variation. Both autistic traits and camouflaging were mostly measured by self-report questionnaires, which may overlap in what they ask and inflate the link. The discrepancy method shares items with the trait measure by design, so its very large figure is probably inflated. The authors could not put all the moderators into one model because missing data would have dropped a non-representative set of studies, so it is unclear how these factors combine. Only 22% of the included studies used an appropriate sample frame and only 42% justified their sample size.

The authors received no financial support for the research, authorship or publication of the article, and declared no competing interests. The review was pre-registered on the Open Science Framework, and the paper documents several changes made to the plan afterwards, including adding social anxiety as a moderator, adding two extra age analyses, and changing how camouflaging measurement was analysed.

Do not treat this as evidence that camouflaging causes poor mental health, or as a reason to push clients to stop doing it. Only one of the included studies measured people at more than one point in time, so the analysis can show that camouflaging and autistic traits go together but not which comes first, and no study here tested whether changing camouflaging changes anything else.