Gesture Production Across Multiple Input Modalities In ASD

Background:  Imitation deficits are a robust finding in individuals with autism spectrum disorders (ASD), and previous findings suggest that imitation performance in ASD appears to be task dependent (Hamilton, 2008; Mostofsky et al., 2006; Smith & Bryson, 2007). However, modality-specific effects on gesture production are not well understood in ASD. Gestural production deficits affect the ability to produce signs and gestures necessary for social communication in children with autism.  Further, the neurocognitive mechanisms affecting gestural processing in ASD are not well established.

Objectives:  1) To explore modality-specific effects on gesture production in a group of individuals with ASD using three tasks of meaningful gestures (transitive gestures, intransitive gestures, and pantomimes) across verbal, visual, and tactile modalities to compare their performance to a matched group of typically developing peers and 2) to investigate the role of underlying cognitive mechanisms including visuomotor integration (VMI), working memory (WM), and visuoperceptual processing (VP) on gestural production performance in ASD.

Methods: 19 individuals with ASD and 23 TD controls (mean age of 12.0 and 12.1 respectively) were tested on five tasks of gesture production comprising intransitive gestures in verbal and visual modality; pantomimes in verbal, visual, tactile, and imitation modalities; and transitive gestures in spontaneous object use and imitation of object use. All gestures were videotaped, coded, and analyzed. Here we report analyses of the gesture production tasks across all modalities.

Results:  Binomial regression models were fitted to determine whether there was an effect of autism and/or an effect of modality in the production performance of transitive gestures, intransitive gestures, and pantomimes in the ASD and control groups. Results revealed that the odds of successful gesture performance among typically developing children were 9.34 fold higher than among autistic children. Transitive gesture production was performed better than imitation; intransitive gesture imitation was performed better than visual or verbal modalities; and pantomime imitation was performed the worst followed by tactile, visual, and verbal modalities. After controlling for the underlying cognitive mechanisms, the ASD group continued to show significantly poorer production performance than the TD group. Higher VP and listening recall were associated with higher odds of success [OR=1.06 and OR=1.02 respectively]. An additive model measuring the effect of ADOS and SCQ on gesture production in each modality was fitted. The results indicated that lower ADOS scores were associated with higher success but the effect of SCQ did not reach statistical significance. Gesture production performance was a significant predictor of the characteristic features of autism as measured by the ADOS.

Conclusions:  Our results indicate that individuals with ASD demonstrated impairments across gesture types and modalities and that these findings suggest that transitive, intransitive gestures and pantomimes were performed differently across the tested modalities. These results may inform future therapeutic interventions that include the use of multiple modality communication strategies, especially in nonverbal children with autism. Production performance was a significant predictor of the characteristic features of autism as measured by the ADOS, suggesting that disorders of praxis processing may be a defining feature of ASD.