Multisensory Integration and Temporal Synchrony in Autism

Background:  Multisensory integration is the process by which individuals combine information from multiple sources (e.g., vision, audition, touch) to produce a unique and unitary percept that produces a more accurate experience. Individuals with autism frequently show difficulty in multisensory integration tasks, and as a result may have difficulty locating and identifying stimuli in “noisy” sensory environments. This may be especially harmful during early development, when much of what is learned about language and communication depends on the ability to locate a speaker and identify what they are saying. A crucial aspect of multisensory integration in typical individuals is the ability to perceive and utilize temporal synchrony cues between two streams of information. Previous work strongly suggests that temporal synchrony detection may be altered in individuals with autism, but the effect of such a deficit on localization and identification of multisensory events has not been clearly studied.

Objectives:  The present study examined the ability of individuals with autism to utilize temporal cues while locating and discriminating social and nonsocial audiovisual stimuli.

Methods:  Twenty children and adolescents with high functioning autism and twenty well matched controls completed audiovisual localization and discrimination tasks presented with varying levels of audiovisual asynchrony. In Experiment 1, participants viewed side by side videos (of either a person speaking or a ball bouncing), one of which matched the sound they heard and one of which was mismatched at varying levels of asynchrony.  For Experiment 2, participants completed a discrimination task, in which they were asked to discriminate small differences in auditory stimuli during a social and nonsocial task. The videos for the stimuli were presented at varying levels of asynchrony and discrimination thresholds were determined for each level.

Results:  Results from Experiment 1 revealed that individuals with autism showed a differential pattern of sensitivity to audiovisual asynchrony compared to controls when viewing social versus nonsocial stimuli. In particular, when using audiovisual synchrony to locate stimuli, individuals with autism showed a pattern of differentially larger Temporal Windows of Integration (TWI) for nonsocial compared to social stimuli in comparison to the pattern seen in the control group. However, results from Experiment 2 showed that individuals with autism showed enhanced auditory discrimination at baseline, and that they utilized temporal synchrony cues similarly to controls when discriminating audiovisual stimuli.

Conclusions:  In combination, the results of this study place emphasis on the role of orientation and attention in audiovisual integration in autism, suggest differences in processing of social and nonsocial information, and show largely intact processing of temporal cues when task complexity is minimized. These results have implications for our theoretical understanding of the etiology of autism as well as expression and development of clinical symptoms.