The CAT-Q in a new culture: does "masking" survive the journey from London to Tehran?

Camouflaging is one of the more clinically loaded ideas to enter adult autism assessment in the past decade. It names the effortful work many autistic adults do to pass as non-autistic: rehearsing facial expressions, suppressing stimming, scripting conversations, forcing eye contact. The construct matters because camouflaging is linked to delayed diagnosis, exhaustion, and elevated distress, and because it can hide exactly the traits a clinician is trying to detect. The Camouflaging Autistic Traits Questionnaire (CAT-Q), developed in the United Kingdom, has become the default self-report tool for measuring it. But a self-report scale is only as good as the population it was built on, and camouflaging is, almost by definition, a culturally shaped behaviour. What counts as a normative social performance in London is not what counts in Tehran.

This study asks the right question: when you carry the CAT-Q across a major cultural and linguistic boundary, does it still measure the same thing? The authors used a standard forward-backward translation with expert review, then administered the Persian CAT-Q to 1,100 people, retaining 948 after screening. The structural result is reassuring. Confirmatory factor analysis recovered the original three-factor model – compensation (actively generating social behaviour), masking (hiding autistic features), and assimilation (forcing oneself through social situations) – with fit indices in the acceptable range. The construct did not collapse into a single undifferentiated "effort" factor, nor did it fragment.

The more useful finding for practice is the measurement invariance result. The authors compared adults with high versus low autistic traits and found configural, metric, and scalar invariance. In plain terms: the items mean the same thing to both groups, the items relate to the underlying trait with the same strength, and a given score corresponds to the same level of camouflaging regardless of group. Without scalar invariance, comparing camouflaging scores between high- and low-trait adults would be comparing two different rulers. That is the technical precondition for any clinical claim that one person camouflages "more" than another.

There is a soft spot, and the authors are honest about it. The assimilation subscale showed the weakest reliability. This echoes a recurring pattern in the wider CAT-Q literature, where assimilation often behaves less coherently than compensation or masking. It may be that "forcing yourself to be in social situations" is too heterogeneous a behaviour to hang together cleanly, or that it overlaps with social anxiety rather than camouflaging proper. Either way, a clinician reading a Persian CAT-Q profile should weight the assimilation subscale with appropriate caution.

Why the construct travels but the subscale wobbles

The split result – a robust overall structure with one shaky subscale – is more informative than a uniformly clean validation would have been. It suggests that the core of camouflaging (deliberately producing or suppressing social signals) is a relatively portable human strategy, while the more diffuse experience of pushing oneself through social contact is shaped by local expectations and may blur into adjacent constructs. A measure that travels perfectly is often a measure that is not measuring anything specific.

What this means at the desk

For practitioners outside the Anglophone world, this is one fewer reason to default to an unvalidated English instrument. A Persian-speaking adult presenting with burnout, late-recognised autistic traits, or a history of "managing" socially can now be assessed with a tool that has documented psychometric standing in their language. The clinical use is not diagnostic; it is to surface the hidden labour that flat behavioural observation misses.