Narrative Abilities In Boys with Autism and Fragile X Syndrome

Background: Several studies have found that narratives produced by children with autism spectrum disorders (ASDs) do not differ from comparison groups of children with intellectual disability (ID) and typical development (TD) on basic levels of story production such as length and reference to main events (Diehl et al., 2006; Loveland et al., 1990). However, children with ASD exhibit difficulties in other key components of narrative, including evaluative devices (which are used to communicate the narrator’s perspective), and the use of causal language in particular (Capps et al., 2000; Diehl et al., 2006; Losh & Capps, 2003; Tager-Flusberg, 1995). They are also more likely to include irrelevant or inappropriate utterances (Diehl et al., 2006; Loveland et al., 1990). Much less research has been conducted on narrative abilities in children with fragile X syndrome (FXS). Keller-Bell & Abbeduto (2007) found that narrative profiles of children with FXS (without comorbid autism) resembled those of TD children more than children with Down syndrome (DS). No research to date has investigated how the narrative abilities of children with ASD, FXS with comorbid autism, and FXS without autism compare to children with ID or TD children. 

Objectives: To compare the narrative abilities of children with autism spectrum disorder (ASD), FXS with ASD (FXS-ASD), FXS only (FXS-O), Down syndrome (DS), and typical development (TD).

Methods: Narratives were elicited from 101 boys using the wordless picturebook Bed Full of Cats (Keller, 2003). Groups included children with ASD (n=20, mean CA=9.36 years, sd=2.0 years), FXS-ASD (n=21, mean CA=10.6 years, sd=2.7 years), FXS-O (n=21, mean CA=9.68 years, sd=2.56 years), DS (n=19, mean CA=10.85 years, sd=2.14 years), and TD (n=19, mean CA=4.72 years, sd=1.04 years) who were similar on non-verbal mental age, expressive language age, and receptive language age. Narratives were transcribed and coded for length, syntactic complexity, evaluative devices, story structure, and inappropriate and off-topic utterances. 

Results: One-way between-group analyses of covariance were conducted to compare narrative variables across groups. After adjusting for non-verbal IQ, and expressive and receptive language ability, significant group differences emerged in the use of evaluative language F(4, 90) = 2.53, p = 0.046, and the number of story episodes included in narrative F(4, 90) = 2.80, p = 0.031. In particular, individuals with FXS-O performed most poorly, whereas individuals with FXS-ASD performed more similarly to individuals with ASD-O. Further examination of patterns of evaluative language use revealed that the ASD and FXS groups (both ASD and FXS-O) showed particular difficulty with causal explanations of emotion in particular. Moreover, children with FXS-ASD used no causal language at all when referring to such states. 

Conclusions: Findings suggest that children with FXS without autism exhibit more severe narrative impairments than those with ASD. Moreover, the narratives of the FXS-ASD and ASD groups shared many of the same characteristics, including similar use of evaluative devices, and particular difficulties with causal explanations of emotions. Taken together, these findings contribute to existing literature documenting an overlap of autism and FXS.