Sensory Profiles and Bisensory Facilitation: Is Unimodal Processing Impaired, but Audio-Visual Integration Intact in ASD?

Background:

Atypical responses to sensory stimuli are frequently seen in children with autism spectrum disorders (ASD), implying that sensory information may be processed abnormally. In previous studies, Dunn’s (1997) Sensory Profile scores have been found to differ between children with autism and their typically developing (TD) peers (Kern et al., 2007). It has been further suggested that individuals with ASD have difficulty integrating auditory- visual information in verbal conditions (Smith & Bennetto, 2007). Data on sensory integration in nonverbal tasks are less consistent. We used a nonverbal auditory-visual paradigm, for which bisensory facilitation effects (reduced response times when congruent stimuli are presented in both modalities) have been reported in typical adults.  

Objectives:

To examine whether bisensory facilitation is affected in autism and whether facilitation may be related to abnormal sensory processing. 

Methods:

Participants were 13 boys (mean age 13.8, sd=2.2 ) with ASD and 13 TD boys (mean age 14.0, sd=2.3), matched for nonverbal IQ. Participants completed the Adolescent/Adult Sensory Profile questionnaire prior to the experimental task, which included three stimulus types: auditory (a high or low tone), visual (a dot appearing in a top or bottom box that was presented throughout), and bisensory (congruent auditory and visual stimuli). Participant responded by pressing a “high” or “low” button on a button-press device. Stimuli were presented in two different conditions: a Repetitive Task (REPT) consisting of repeated presentation of a given stimulus type in separate runs, and a Modality Switching Task (MST) with randomly alternating stimulus types.   

Results:

There was no group difference in RT, F(1,22)=1.43, p=.30, or accuracy , F(1,22)=1.75, p=.20.  Both groups showed a bisensory facilitation effect, with a decrease in reaction time during the bimodal condition, compared to either unimodal condition, F(1,22)=56.95, p<.01.  For the Sensory Profile, a one-way ANOVA showed that the ASD group had significantly higher scores for the Low Registration quadrant than the TD group, F=9.98, p=.005. In the ASD group (but not the TD group), scores on this quadrant were significantly correlated with response times for auditory trials of the REPT, r=.77, p=.005, and for both unimodal and bimodal conditions of the MST (auditory r=.62, p=.02, visual r=.57, p=.04, bimodal r=.62, p=.02) .
Conclusions:

There was no evidence of impaired sensory integration in the ASD group, as both groups showed bisensory facilitation. However, sensory scores were abnormal in our ASD group, consistent with previous studies. Specifically, higher scores on the Low Registration quadrant of the Sensory Profile reflect slowed detection of sensory stimuli (Brown & Dunn, 2002),  as supported by our finding of correlation with response times for auditory stimuli and with generally slowed response on the modality switching task, for the ASD group. The pattern of results suggests that low-level sensory impairment is more common and more severe in ASD than any potential defects in sensory integration.