Relative Contributions of Speech and Gesture on the ADOS “Demonstration Task” in Adolescents with High-Functioning Autism

Background: Gesture, as a form of nonverbal communication, is thought to be significantly impaired in autism spectrum disorders (ASD). Gestural impairments are considered in both the social and communicative domains of the DSM diagnostic criteria, and are assessed on multiple items of the ADOS and the ADI-R. Decreased frequency of gesture, as well as poor integration of gesture and speech, is considered more symptomatic. Despite the clinical significance of gesture in ASD, it has received scant empirical attention. As such, the origin of the gesture impairment in ASD remains unknown. It is unclear whether gesture is specifically impaired or whether gesture impairments reflect the broader communicative impairment seen in ASD. Although early studies found reduced frequency of gestures in ASD, most studies find no group differences when controlling for the frequency of communicative acts.

Objectives: We investigate the formal qualities of co-speech gesture in high-functioning adolescents with ASD, focusing on 1) modality (gesture, speech or both) choice for event description and 2) gesture form. Gesture is an important communicative tool, and an understanding of how individuals with ASD gesture is a critical first step toward designing interventions to improve nonverbal communicative skills.

Methods: This study examines the spontaneous production of co-speech gestures in a sample of 12 high-functioning adolescents with ASD ages 13-17, compared with 12 chronological age-, gender-, and IQ-matched typically-developing (TD) adolescents (group differences, all F’s < 2, all p’s > .18). Participants were filmed during the demonstration task of the ADOS, in which they were asked to “show and tell” the experimenter how to brush her teeth. The speech and gestures used to describe four specific tooth-brushing events were coded, with a focus on errors in gesture formation (as in Dewey et al., 2007).

Results: Both groups encoded approximately the same number of events (ASD M = 2.6/4, TD M = 3.2/4) collapsed across modalities. Although most events were represented in both modalities (such that gesture complemented the information presented in speech), the TD adolescents had a bias toward representing these events as speech, whereas the ASD group had a bias toward representing these events as gesture (group by modality interaction, p = .06). Interestingly, no differences were found in the “structural quality” of gestures; that is, the gestures of the ASD group were all as well-formed as those of controls.

Conclusions: During a structured communication task, in which participants are explicitly instructed to “show and tell” the experimenter how to complete an often-enacted, familiar task, we find that high-functioning adolescents with ASD do not show a decreased frequency of gesture production relative to overall communication, nor do their gestures reflect motor planning difficulties. Rather, adolescents with ASD may rely more heavily on gesture to convey information, whereas TD adolescents may rely more heavily on speech. High-functioning adolescents with ASD may be more inclined to demonstrate familiar actions in gesture that are difficult for them to represent with words. Importantly, these adolescents were able to communicate information with gesture even when that information was absent from speech.