Tell Me More: Storytelling in Autism Reflects Motor, Executive, and Social Impairments

Background: Narratives provide naturalistic insight into social-communicative and cognitive functioning. A recent meta-analysis shows that children with autism spectrum disorders (ASD) demonstrate impairments in narrative structuring (Baixauli et al., 2016). Reciprocity of movement (e.g., mimicry of interaction partners, imitation) is also crucial for social development. While mounting research indicates motor deficits and impaired social reciprocity in ASD, how these impairments are linked across domains is unclear. Associations among storytelling and measures of motor, executive, and social impairments can help construct the broader picture of behavior and cognition in ASD.

Objectives: This work examines: (a) diagnostic differences in storytelling, as measured by event inclusion and first-person speech, and (b) associations among storytelling and measures of motor, social, and executive function (EF). Event inclusion reflects memory and narrative structuring skill; first-person usage reflects flexibility of speech and perspective-taking. We hypothesized that: (a) compared with typically-developing (TD) children, children with ASD would include less events and use less first-person speech, and, (b) across ASD and TD groups, storytelling measures would be associated with measures of core social, cognitive, and motor impairments of ASD.

Methods: Twenty-eight children (16 ASD: 1f; 12 TD: 3f; M_age=10.04), matched on age and IQ, of a total target sample of forty, participated in a “memory game”. Children watched a recording of a trained actor, who narrated a brief story while performing specific actions (yawning, arm scratching, face rubbing). Children then retold the story. Sessions were coded for spontaneous mimicry of the narrator’s actions. Children’s retell was evaluated for proportion of total story events included, and for first-person usage as a proportion of total utterances. Additional measures included: Social Responsiveness Scale (SRS-2), a praxis gesture imitation task, and mimicry of narrator (social functioning); Physical and Neurological Examination for Subtle Signs (PANESS, motor functioning); and Behavior Rating Inventory of Executive Function (BRIEF) and Wechsler Intelligence Scale for Children (WISC-V) Digit Span sub-scale (EF).

Results: The ASD group reproduced fewer events (p=.004) and used less first-person speech (p=.004) than did the TD group, suggesting poorer narrative structuring in ASD. Bivariate correlations (Figure 1) across groups revealed that event inclusion was correlated with motor (PANESS), r(26)=-0.57, p=.005, and social-communicative (SRS-2) function, r(26)=-0.58, p=.002, the latter being largely driven by the ASD group. Strong correlations of event inclusion with EF and praxis did not survive multiple-comparison corrections; event inclusion was not correlated with mimicry. Due to low dimensionality of first-person speech data, MANOVA was conducted with presence/absence of first-person speech as the predictor variable (Table 1). Children who produced first-person speech had higher motor control (PANESS: p=.03), social-communication (SRS-2: p=.02), and EF scores (BRIEF: p=.01). First-person speech was not associated with praxis or mimicry.

Conclusions: Using measures of event inclusion and first-person speech, this study demonstrated deficiencies in narrative structuring in children with ASD. The findings reveal that storytelling performance is strongly associated with measures of motor and social-communicative functioning. EF was associated specifically with first-person speech usage. Future analyses with the full dataset will examine how storytelling is associated with ASD symptom severity.