A transdiagnostic study of theory of mind in children and youth with neurodevelopmental conditions
A child’s IQ and social-communication skills predict theory-of-mind scores better than the autism, ADHD, or OCD label.
01Research in Context
What this study did
The team gave a video-based mind-reading task to kids and teens. Some had autism, ADHD, or OCD. Some had no diagnosis.
They used the SAT, a short test that shows clips of people acting. The child picks why the actor did it.
The goal was to see if the label on the chart predicted scores better than everyday skills like IQ and talking.
What they found
Kids with autism scored lower on the task than the other groups.
Yet IQ and social-communication trouble explained the score spread better than the diagnosis itself.
In plain words: smart, chatty kids did well, no matter the label.
How this fits with other research
Carter Leno et al. (2021) saw the same pattern. Once they held verbal IQ steady, theory-of-mind scores no longer linked to behavior problems in autistic youth.
Granader et al. (2014) also found that planning and shifting skills, not the autism label, drove ToM scores in bright preschoolers.
These three studies line up like Lego bricks. They all say, "Look at the child’s skills, not the file folder."
Thirion-Marissiaux et al. (2008) looked only at kids with ID. They still saw lower ToM, but language level mattered most. The new paper widens the lens to ADHD and OCD and gets the same answer.
Why it matters
Stop letting a diagnosis decide your goal list. Test the child’s IQ, language, and social use. Pick lessons that grow those skills. A chatty kid with ADHD may need the same ToM drill you planned for an autistic peer.
Want CEUs on This Topic?
The ABA Clubhouse has 60+ free CEUs — live every Wednesday. Ethics, supervision & clinical topics.
Join Free →Run a quick language sample and social-communication checklist before you pick a ToM lesson — use the results, not the chart code, to choose the target.
02At a glance
03Original abstract
Theory of mind (ToM) is fundamental for social interactions, allowing individuals to appreciate that others have their own mental states. Children and youth with neurodevelopmental conditions (e.g., autism, attention-deficit hyperactivity disorder (ADHD), and obsessive-compulsive disorder (OCD)) often show differences in ToM abilities compared to their neurotypical (NT) peers. Given the phenotypic heterogeneity and overlap associated with these conditions, this motivates a transdiagnostic investigation of ToM across neurodevelopmental conditions. Five hundred and fifty-five participants (5–22 years; 193 ADHD, 189 autism, 33 OCD, and 140 NT) were recruited via the Province of Ontario Neurodevelopmental Disorders network. To measure ToM, participants completed the Social Attribution Task (SAT), where participants attribute social stories to videos of moving shapes. The Animation Index (ability to attribute social stories to the videos) and Pertinence Index (how pertinent the attributions are) were calculated from the descriptions. Three analyses were performed: (1) a case-control analysis, comparing the SAT indices amongst the diagnostic groups, (2) a univariate dimensional analysis, examining associations with phenotypic variables (e.g., full-scale IQ, verbal IQ, and social communication difficulties), (3) and a multivariate analysis (partial least squares) that identifies a latent space that describes the associations between the SAT and phenotypic measures. There were no between-group differences in the Animation Index, but the Pertinence Index was significantly lower in autism compared to the other diagnostic categories. Phenotypic variables (full-scale IQ, verbal IQ, and social communication difficulties) were found to be significantly associated with SAT performance across groups, and explained more variance than the diagnostic categories. In the multivariate analysis, the phenotypic variables contributed more strongly to the identified latent component compared to the diagnostic categories. The verbal requirement of the SAT limited the inclusion of non-verbal participants, while the overall cognitive demand limited the participation of those with lower IQs. Additionally, our OCD group was significantly smaller than the other groups, which may have limited our ability to detect OCD-specific effects. In a large sample, we found that transdiagnostic measures, such as IQ and social communication difficulties, are related to SAT abilities across neurodivergent and neurotypical children and youth and better describe differences in SAT performance compared to the individual diagnostic categories. Although poorer performance on ToM tasks has been classically associated with autism, this study highlights that transdiagnostic, phenotypic variables are a stronger predictor of SAT performance than diagnostic group. The online version contains supplementary material available at 10.1186/s13229-025-00671-3.
Molecular Autism, 2025 · doi:10.1186/s13229-025-00671-3