Rates of Audiovisual Speech Integration Covary with Low-Level Multisensory Temporal Processing in ASD Individuals

Background: While the diagnostic criteria for ASD include communication impairments, social deficits, and restricted/repetitive behaviors, sensory impairments are frequently reported as well. These sensory deficits are the focus of ongoing research, with the framework that these contribute to the deficits seen in higher-order cognitive function. In the current study we seek to elucidate these links by investigating how temporal deficits in low-level sensory processing, specifically audiovisual integration, are related to measures of speech processing in ASD individuals.

Objectives: Our objective is to assess the relationship between individual’s multisensory temporal processing and their ability to perceive an audiovisual speech utterance as a single, unified event. We sought to accomplish this by measuring the temporal binding window (TBW), a probabilistic construct that reflects the interval of time within which two sensory signals may be perceptually bound, and relating this to an audiovisual measure of perceptual fusion of speech utterances. Differences between ASD and typically developed (TD) groups were assessed for each measure.

Methods: Participants included 21 TD adults (18-35), 40 TD (7-17) children, and 24 ASD (6-18) children, matched for IQ, visual, and auditory acuity. Participants completed a simultaneity judgment task where visual flashes and auditory beeps were presented at varying asynchronies from 0-400ms in both auditory-first and visual-first configurations. Participants also completed a McGurk task where they reported their perception to congruent /ba/ and /ga/ utterances and an incongruent condition with an auditory /ba/ presented with a visual /ga/ (the McGurk stimulus). The TBWs for both auditory-first and visual-first conditions were calculated, as where rates of McGurk perception. The relationship between these three measures was then correlationally related within individuals.

Results: In TD adults, the visual-first side of the TBW was significantly narrower than the auditory-first TBW. In addition, only the width of the visual-first portion of the TBW (reflecting ecologically valid stimuli) was correlated with individuals’ rates of McGurk perception, such that Individuals with narrow TBWs were more likely to report the McGurk effect. TD children also showed a narrower visual-first TBW, and again only this visual-first TBW was correlated with rates of McGurk perceptions. ASD children, on the other hand, failed to exhibit an asymmetry in the TBW. Importantly, both the visual-first and the auditory-first TBWs were highly correlated with rates of McGurk perception in the ASD group.   

Conclusions: With typical development, the TBW matures to reflect the natural statistics of audiovisual sensory inputs: TD perceptual systems are more tolerant of visual-first inputs. As such, in TD, the temporal processing reflecting such natural perceptions (the visual-first portion of the TBW) is related to our abilities to perceptually bind an incoming auditory and visual signal (the McGurk effect). In ASD individuals, however, this low-level temporal processing is impaired, resulting in a symmetrical TBW. Furthermore, this impairment in low-level temporal processing maps directly onto the ability of these individuals to perceptually fuse speech signals. This suggests that deficits in low-level temporal processes may have a cascading effect, impacting higher-order cognitive functions including, but not limited to, speech communication.