Heritability of Pragmatic Language in Autism Spectrum Disorder: A Study of Twins

Friday, May 16, 2014
Atrium Ballroom (Marriott Marquis Atlanta)
V. Y. Kang1, K. Levesque1, A. Anderson1, A. Kresse1, S. Faja1, E. E. Neuhaus1, R. Bernier1 and S. J. Webb2, (1)University of Washington, Seattle, WA, (2)Psychiatry and behavioral sciences, University of Washington, Seattle, WA

Environmental and genetic factors both influence the development of Autism Spectrum Disorder (ASD) (Hallett et al., 2012). Some evidence for a strong genetic component in ASD comes from twin studies, which have found rates of concordance for the disorder to be higher among monozygotic (MZ) twins relative to dizygotic (DZ) twins (Hallmayer et al., 2011). Additional support comes from the study of the broader autism phenotype (BAP) in which unaffected first degree relatives of individuals with autism display some ASD traits (Constantino et al., 2010; Toth et al., 2007).

Studies have shown that genetic factors have a strong influence on communication in individuals with autism (Hallett et al., 2012). Specifically, deficits in pragmatic language are not only one of the consistent impairments found in ASD, but also a result of a genetic component (Tager-Flusberg, 2005; Baron-Cohn, 1988; Bishop et al., 2006). Impairment in pragmatic language includes overtalkativeness, talking in a stereotypical manner, or being unable to consider others’ viewpoints (Bishop et al., 2006). As pragmatic language demonstrates an understanding of social cues and norms, it could be used as a method of measuring one’s communication skills (Lam et al., 2012).  


This study will investigate pragmatic language impairments in a twin ASD population. If pragmatic language has a genetic component, there should be a higher degree of similarity in pragmatic language ability within MZ twin pairs compared to DZ twin pairs.


19 MZ twin pairs (14 TD, 5 ASD) and 17 DZ pairs (5 TD, 12 ASD) between the ages of 5 and 22 participated in the study (M = 10.6, SD = 4.02).

Individuals were video-recorded while they completed a 15-20 minute interview from the ADOS. The videos were rated by coders blind to diagnosis. Coders assessed subjects’ pragmatic language using the Pragmatic Rating Scale-Modified (PRS-M) (Ruser, 2007), which scores subjects on both verbal (e.g., clarity, referencing emotions, grammar) and non-verbal (e.g., eye contact, gestures) aspects of pragmatic language.


Typically developing (TD) individuals showed significantly better pragmatic language (evidence by lower PRS-M sum of scores) (M = 9.21, SD = 3.48) compared to the ASD group (M = 13.52, SD = 3.07) (F (1) = 16.19, p = .00) confirming that pragmatic language is an impairment in the ASD sample.

The intra-class correlation for MZ and DZ pairs were analyzed with SPSS. PRS-M sum scores of MZ twin pairs showed a strong correlation (ICC (19) = .85, p = .00). PRS-M sum scores were not related within twin pairs for DZ twins (ICC (17) = .45, p = .13).  


Pragmatic language ability showed strong concordance within MZ twins but not DZ twins, suggesting that it may have a genetic basis. Given its potential genetic basis and its sensitivity to ASD impairment, it may have potential to be an endophenotype of ASD.