Aetiological relationship between language performance and autistic-like traits in childhood: a twin study
Article first published online: 31 DEC 2010
2007 Royal College of Speech & Language Therapists
International Journal of Language & Communication Disorders
Volume 42, Issue 3, pages 273–292, May-June 2007
How to Cite
Dworzynski, K., Ronald, A., Hayiou-Thomas, M., Rijsdijk, F., Happé, F., Bolton, P. F. and Plomin, R. (2007), Aetiological relationship between language performance and autistic-like traits in childhood: a twin study. International Journal of Language & Communication Disorders, 42: 273–292. doi: 10.1080/13682820600939002
- Issue published online: 31 DEC 2010
- Article first published online: 31 DEC 2010
- Received 25 October 2005; Accepted 21st July 2006
- domains of autism spectrum disorders (ASDs);
- early language performance;
- twin modelling
Background: Impairments in language and communication are core features of autism spectrum disorders (ASDs). The basis for this association is poorly understood. How early language is related to each of the triad of impairments characteristic of ASDs is also in need of clarification.
Aims: This is the first study that aims to determine the extent to which shared genetic and environmental factors underlie the association between early language performance and autistic-like traits (ALTs) in middle childhood.
Methods & Procedures: Data came from a population-based twin sample (n = 6087 pairs) assessed prospectively at 2, 3, 4 and 8 years. ALTs measured by the Childhood Asperger Syndrome Test (CAST) at 8 years were investigated in relation to language assessed by the MacArthur Communicative Development Inventory (CDI) at 2, 3 and 4 years. Multivariate model fitting techniques were used to analyse the origins of this association.
Outcomes & Results: Total CAST scores, as well as Social and Communication subscales, at 8 years were weakly but significantly negatively correlated with language ability at 2, 3 and 4 years. Correlations between language and restrictive and repetitive behaviours and interests (RRBI) were not significant. The phenotypic correlations between language and Social and Communication ALTs were almost entirely mediated by shared genetic influences. There were specific genetic influences on early language that were not shared with ALTs, and specific genetic influences on ALTs not shared with earlier language performance.
Conclusions: This is the first study to demonstrate shared genetic influences in relation to language performance as an early antecedent of later ALTs. These results support the idea that the triad of core features in ALTs are aetiologically heterogeneous, with early language relating to social and communication impairments but not RRBIs.