Understanding the associations between neurodevelopmental features and internalising and externalising behaviours: a transdiagnostic approach

Background: Internalising and externalising behaviours – significant markers for lifetime psychiatric vulnerability – are elevated in children with neurodevelopmental diagnoses, including autism. Although neurodevelopmental features of autism are dimensions that span the population, limited research has examined their differential patterns of association with internalising and externalising behaviours in children without specific, categorically-defined diagnoses. Evidence of such associations outside of a traditional diagnostic context may enable more targeted support for children’s individual needs, irrespective of diagnoses. The current study aimed to characterise the relationship between neurodevelopmental features found in autism – restricted and repetitive behaviours (RRB) and social communication difficulties – and internalising and externalising behaviours in children from mainstream school who experience emotional, behavioural, or cognitive challenges.

Method: We recruited 136 6-7-year-olds without known clinical conditions but with school-identified emotional, behavioural or cognitive difficulties. The Repetitive Behaviour Questionnaire-2 assessed RRBs, the pragmatics scale from the Revised Children’s Communication Checklist-2 assessed social communication, and the Strengths and Difficulties Questionnaire examined internalising and externalising behaviours.

Results: Simultaneous hierarchical linear regression analysis identified differential associations when adjusting for covariance between internalising and externalising. Social communication made a stronger contribution than RRBs to variance in externalising behaviours (Fchange(1,131)=11.84, p<.001). However, for internalising behaviours, RRBs made the strongest contribution (FChange(2,131)=8.19, p<.001). The insistence on sameness subdomain of RRB predicted variance in internalising but not externalising behaviours independently of social communication while the repetitive sensory and motor behaviour subdomain predicted variance in externalising but not internalising behaviours, but only when social communication was not included.

Conclusion: These findings will inform future research aimed at understanding the co-occurrence of traits across diagnostic boundaries. Evidence that RRBs and social communication are differentially associated with internalising and externalising behaviours may identify target areas for the support of children with emotional and behavioural difficulties, a group whose co-occurring neurodevelopmental features are often under-recognized.