The development of multisensory integration in high-functioning autism: high-density electrical mapping and psychophysical measures reveal impairments in the processing of audiovisual inputs

Abstract

Successful integration of auditory and visual inputs is crucial for both basic perceptual functions and for higher-order processes related to social cognition. Autism spectrum disorders (ASD) are characterized by impairments in social cognition and are associated with abnormalities in sensory and perceptual processes. Several groups have reported that individuals with ASD are impaired in their ability to integrate socially relevant audiovisual (AV) information, and it has been suggested that this contributes to the higher-order social and cognitive deficits observed in ASD. However, successful integration of auditory and visual inputs also influences detection and perception of nonsocial stimuli, and integration deficits may impair earlier stages of information processing, with cascading downstream effects. To assess the integrity of basic AV integration, we recorded high-density electrophysiology from a cohort of high-functioning children with ASD (7-16 years) while they performed a simple AV reaction time task. Children with ASD showed considerably less behavioral facilitation to multisensory inputs, deficits that were paralleled by less effective neural integration. Evidence for processing differences relative to typically developing children was seen as early as 100 ms poststimulation, and topographic analysis suggested that children with ASD relied on different cortical networks during this early multisensory processing stage.

Keywords: ERPs; auditory; electrophysiology; multimodal; visual.

Figures

Figure 1.

Testing the race model. Mean data across the full data set are presented (N = 118). (a) RTs and standard errors for the multisensory (AV) and unisensory conditions. (b, c) CP distributions for the multisensory (red trace), auditory-alone (blue trace), visual-alone (green trace) stimulus conditions, and the CP predicted by the race model (black trace) as a function of RTs (b) and percentile (c). When the CP for the multisensory condition is greater than that predicted by the race model, race-model violation has occurred. (d) Miller inequality: values greater than zero signify race-model violation. Miller inequality values at the first seven quantiles (equivalent to the 35th percentile) were submitted to t-tests. Asterisks indicate statistically significant race-model violation.

Figure 2.

The percentage of children in each group who show violation of the race model at each of the seven quantiles considered (the first third of the RT distribution). TD, children with typical development; ASD, children with autism spectrum disorder.

Figure 3.

Miller inequality curves are presented for each of the four groups. Values greater than zero signify race-model violation. Miller inequality values at the first seven quantiles (equivalent to the 35th percentile) were submitted to t-tests. Asterisks indicate statistically significant race-model violation.

Figure 4.

Auditory ERPs. Mean ERPs to the auditory-alone condition are presented for each of the four groups. Traces represent the composite signal from two adjacent electrodes, the locations of which are indicated on the head models.

Figure 5.

Visual ERPs. Mean ERPs to the visual-alone condition are presented for each of the four groups. Traces represent the composite signal from two adjacent electrodes, the locations of which are indicated on the head models.

Figure 6.

Multisensory effects at 110 ms. (a) Multisensory (AV) and summed (A + V) ERPs and their difference (AV − (A + V)) are shown for each of the four participant groups. Traces represent the composite of four adjacent fronto-central electrode sites (Fz, FCz, FC1, and FC2). Gray bars highlight the 100–120 ms window of analysis. (b) Voltage maps depict the scalp distribution of the MSI effect at 110 ms poststimulus onset (the difference between the multisensory and summed responses). (c) Multisensory (AV) and summed (A + V) ERPs and their difference (AV − (A + V)) are shown for each of the four participant groups. Traces represent the composite of four adjacent parietal electrode sites (Pz, P1, P2, CPz). (d) Results of the TANOVA analysis. Significant topographical differences between the TD and ASD groups are marked in red, presented separately for young and old age groups.

Figure 7.

SCPs: running t-tests comparing the multisensory and sum ERPs for each of the four groups. Significance is depicted for effects meeting a 0.05 alpha criterion and lasting for at least 10 consecutive data points (19.2 ms at a 512 Hz sampling rate). The color bar indicates directionality of the effects, with white indicating an absence of significant t-values. Time is plotted in the x-axis from −50 to 300 ms. Electrodes are plotted in the y-axis. Starting from the bottom of the graph, the electrodes are divided into sections from posterior to anterior scalp with each color representing 4–5 electrodes, the relative positions of which are located on the corresponding head.

Figure 8.

MSI effects at 150 ms. Multisensory (AV) and summed (A + V) ERPs and their difference (AV − (A + V)) are shown for each of the four participant groups. Traces represent the composite of three adjacent left parieto-occipital sites (PO7, PO3, O1; location indicated with a dashed circle on the left most voltage map). Gray bars highlight the 140–160 ms window of analysis. Voltage maps depict the scalp distribution of the MSI effect (the difference between the multisensory and sum responses) at 150 ms.