The Function of Gesture: Socially-Oriented Vs. Process-Oriented Gestures in Autism Spectrum Disorder

Background: Diagnostic criteria for autism spectrum disorder (ASD) specify the presence of deficits in the understanding and use of gestures. While some research suggests that individuals with ASD gesture less frequently than typically developing (TD) peers, other reports suggest that gestures in ASD occur as frequently, but are poorly synchronized with verbal referents. This discrepancy in findings may reflect differences in the function of gesture: some gestures have a primarily social-communicative function, while others are process-oriented (i.e., assisting the speaker with lexical access or other cognitive functions). Given the sociocommunicative deficits in ASD, we might expect deficits in socially-oriented gestures in ASD, but no differences in process-oriented gestures.

Objectives: Our objective was to compare production of socially-oriented gestures on the ADOS and process-oriented gestures in a picture naming task, in ASD.

Methods: Adolescents with ASD (n = 13) and TD (n = 14) completed two tasks: the ADOS (Module 3 or 4) and a modified version of the Boston Naming Test. Groups did not differ in age, gender, full scale or verbal IQ, or receptive vocabulary.

Participants named each of 20 picture stimuli on the Boston Naming Test (BNT).  Gestures elicited by this task were coded by a trained research assistant. Item A9 from the ADOS, which measures “Descriptive, Conventional, Instrumental or Informational Gestures,” was rated by a research-reliable clinician. Gesture rates were compared by group; we also examined correlations between gesture performance and symptom severity on the ADOS and SCQ.  

Results: While ASD and TD groups differed on gesture scores on the ADOS (ASD M = 0.571, TD M = 0.154, F(1, 25) = 5.738, p = 0.024, h² =  0.187), gesture frequency on the BNT did not differ (ASD M = 5.000, TD M = 4.077, F(1, 25) = 0.310, p = 0.583, h² =  0.012). There were also no differences for gesture subtypes (iconic, beat, or conventional gestures, all p’s > 0.237). That is, the groups differed in production of socially-oriented gestures (on the ADOS), but did not differ in process-oriented gestures (on the BNT). Furthermore, gesture rates on the BNT were uncorrelated with symptom severity on the SCQ or on the ADOS (combined communication + social interaction domain scores), all r’s< .45, all p’s > .17.

Conclusions: Adolescents with ASD and TD produced comparable numbers of gestures during a lexical access task, suggesting no group differences in production of process-oriented gestures. In contrast, youth in the ASD group had noticeably different gestures during the ADOS. The ADOS likely elicits more socially-oriented gestures; in addition, this measure collapses the dimensions of gesture frequency and quality onto one item. Results suggest that separating gesture performance into more specific categories such as process-oriented vs. socially-oriented gestures or quality vs. frequency of gesture on the ADOS may help to refine our understanding of gestures and, more broadly, social interaction in ASD.