To See but Not to See: Visual Perception Mediates Imitation in ASD

Background: The ability to imitate is an early and important developmental milestone providing the foundation for turn taking, joint attention and theory of mind. Imitation deficits are well documented in individuals with Autism Spectrum Disorder (ASD). Impairments in imitation performance have been identified in meaningful gestures (i.e., goal-directed action), meaningless gestures and even style (e.g., harsh versus gentle). The Self-Identification theory suggests that individuals with autism fail to imitate the style of the examiner due to a lack of interpersonal connectedness. Underlying cognitive factors, including the relationship of visual perception to the imitation of style, have not been fully explored. To our knowledge, this is the first study to design perceptual tasks assessing discrimination of style in ASD. 

Objectives: A case-control study was conducted to: a) determine if individuals with ASD are successful in imitating and discriminating the goal and the style of actions; and b) to investigate the role of underlying cognitive mechanisms including visuomotor integration (VMI), working memory (WM), visuoperceptual processing (VP) and discrimination of style in predicting group membership.

Methods: Nineteen adolescents with high functioning autism (HFA) and 22 typically developing controls (MA = 12.1, SD = 2.4 and MA =12.2, SD = 1.9 respectively) were tested on four newly designed tasks of discrimination of style (e.g., harsh versus gentle and fast versus slow) of object use, intransitive (social) gestures and pantomimes. Four imitation tasks, adapted from Hobson and Lee (1999), were tested: Strumming a pipe rack on a stick, using a stamp and ink pad, flattening a toy frog with a roller and pressing a bobble head. 

Results: No significant between group differences were evident in imitating the goal; however, the ASD group was significantly poorer at imitating the style of the action:  pipe rack and stick, [X ² (1, N = 41) = 5.13, p = .038; stamp and ink pad X ² (1, N = 41) = 5.26, p = .036; frog and roller x² (1, N = 40) = 19.93, p = <. 001 and atypical use of the bobble-head, x² (1, N = 41) = 15.33, p = .003]. Results from a series of multivariate logistic regression models suggested that visuomotor integration and discrimination of the style of intransitive gestures successfully predicted group membership, [OR = .89, p - .013; OR = .38, p = .020]. Results of the discrimination tasks showed significant between-group differences in discrimination of transitive gesture style, t₃₉ = -2.54, p = .015, Cohen’s d =-.79, and intransitive gesture style, t₃₉ = -2.99, p = .005, Cohen’s d = -.92. 

Conclusions: Our findings support and extend the results of Hobson and Lee (1999) revealing significant difference in the imitation of style in ASD and highlight that impairments in visuomotor integration as well as the ability to discriminate the style of gestures may account for this imitative deficit. These findings replicate and extend the original study of Hobson and Lee (1999) suggesting that a primary deficit in visual perception may underlie a deficit in imitation of style in ASD.