An fMRI Exploration of Atypical Multisensory Perception in Autism Spectrum Disorders

Background: Multisensory perception plays an integral role in early cognitive, perceptual, language, and social development. Relevant to the current study, temporal synchrony is one of the first properties used by infants to integrate their sensory experiences (Lewkowicz, 1999).

Sensory atypicalities may be core deficits of individuals with autism spectrum disorders (ASD) and contribute to downstream perceptual and social-cognitive consequences, including hyper- and hypo-sensitivities, poor eye contact, poor imitation skills, repetitive self-stimulatory behavior, and difficulties with social and communication abilities (Bertone, et al., 2005). Abnormalities in the integration of sensory information may underlie these sensory atypicalities (Iarocci & McDonal, 2006). In individuals with ASD, these multisensory perception difficulties may be limited to, or most apparent for, linguistic-based stimuli (Bebko et al., 2006).  However, previous studies have not specified whether the ostensibly linguistic-specific deficit observed in ASD is actually due to the linguistic nature of the stimuli, or to the social nature of the stimuli that is inherent in audiovisual presentations of a person talking.

The posterior superior temporal sulcus (pSTS) appears to be critical for the temporal integration of audiovisual stimuli (Calvert, 2001). In addition to its central role in multisensory temporal processes, studies have shown anatomical (Levitt, 2003) and functional (Pelphrey, 2008) abnormalities within the pSTS of participants with ASD.

Objectives: The present objective was to explore the neurological basis of deficits in multisensory integration in individuals with ASD. Importantly, the current study delineated linguistic and social processing using three distinct categories of stimuli: social-linguistic, social-nonlinguistic, and nonsocial-nonlinguistic stimuli. Further, temporal variations of the stimuli were presented (synchronous vs. asynchronous). The current study identified neurological regions of interest that demonstrated multisensory enhancement, including the pSTS, to compare multisensory integration between individuals with and without ASD across all content and synchrony conditions.

Methods: Fifteen participants with ASD were matched on chronological age and full sale IQ to 18 participants without ASD (comparison group). All participants were between 18-29 years of age and had average to above average IQ. Participants underwent an fMRI scan while viewing the stimuli videos with concurrent eye-tracking to ensure that participants attended to the stimuli.

Results: Preliminary results indicate that for the comparison group, peak activation in the pSTS is reduced for temporally synchronous relative to temporally asynchronous audiovisual presentations across all content conditions. However for participants with ASD, this reduced activation for the temporally synchronous conditions was only equivalent to the comparison group for the nonsocial-nonlinguisitc and social-nonlinguistic conditions. The ASD group showed less reduction in activation relative the comparison group in the social-linguistic condition.

Conclusions: If the preliminary results hold, this would indicate that atypical neural processing in the pSTS might underlie the linguistic-specific multisensory perception impairment in ASD. The reduced activation to all synchronous stimuli may reflect increased processing efficiency associated with multisensory binding. Thus, individuals with ASD do not display the same increased processing efficiency for linguistic information, which may result in downstream impairments, including socio-communication and language deficits.