Diagnostic criteria for autism would suggest that different autism features tend to co-occur and that these are distinct from other aspects of neurodevelopmental disability, e.g. language impairment. However, it is possible that diagnostic criteria represent a skewed picture of the autism phenotype as it manifests in the world. I will present a study in which we investigated the coherence of the autism phenotype, comparing groups selected for neurodevelopmental features (a diagnosis of autism) and genetic aetiology (a diagnosis of sex chromosome trisomy; SCT).

Autism features showed a similar factor structure and were distinct from language ability in both groups, supporting the overall coherence of the autism phenotype. However, the SCT group was more likely to show clinically-significant difficulties in just some aspects of autism and to show fewer non-social autism features for their social-communication disabilities. Therefore, at a subtler level, it seemed that the autistic group over-selected for even profiles of difficulty and for disproportionately elevated non-social autism features. This analysis supports longstanding questions about the diagnostic validity of autism, indicating that diagnostic criteria may not map exactly on the autism phenotype as it manifests “naturally.”

Please register in advance for each seminar. After registering, you will receive a confirmation email containing information about joining the meeting. If you could share with anyone who might be interested in your research groups / departments that would be great. Contact catherine.manning@psy.ox.ac.uk for any questions or suggestions for future speakers.