Visual Function in Adults with High Versus Low Autism Quotient Scores

Background: Although hyper- and hypo-sensitivities to various forms of sensory input are commonly reported in individuals with autism spectrum disorder (e.g., Iarocci & McDonald, 2006), it remains unclear how or why these problems occur.

Objectives: Our goal is to determine whether there are differences in sensory processing in people with high- versus low-levels of autistic-like traits. In the current study, we focus on vision.  Research implicates visual differences between people with and without autism spectrum disorder (ASD), although results are equivocal (see Simmons et al., 2009 for review). By assessing vision in typical adults with high- and low-levels of autistic-like traits, we hope to circumvent problems associated with studying people with ASD, including small sample sizes and various confounds.

Methods: Participants were adults aged 18-29 years. Each participant completed an online survey: The Autism Quotient (AQ). Participants with the highest and lowest 10% of AQ scores were invited to participate in two lab-based sessions. To date, 13 low AQ (M= 8.0, SD=2.3) and 18 high AQ individuals (M=31.0, SD=5.0) have been seen. Participants reported no history of visual problems. Visual screening showed no group differences in logMAR acuity, stereoacuity on the Titmus test, nor fusion on the Worth 4 dot test. In the first session, participants completed the Dunn Sensory Profile and the Broad Autism Phenotype Questionnaire (BAPQ). We also measured participants’ sensitivity to form and motion using psychophysical threshold tasks. Tasks included first-order local motion and pattern discrimination, second-order local motion and pattern discrimination, glass patterns (global form discrimination) and random-dot kinematograms (global motion discrimination). These tasks target lower- and higher-levels of the dorsal and ventral pathways of the visual system. In a separate visit, visual evoked potentials (VEPs) were recorded from each participant in response to pattern reversals and motion onset.

Results: AQ scores correlated strongly with BAPQ scores, r(30)=.76, p< .001. Scores on the Dunn Sensory Profile differed between the high and low AQ groups. Specifically, the high AQ group was both more hypo- and hyper-sensitive to sensory information (p’s<.05 and .001, respectively). The high AQ group also had higher levels of sensory avoidance behaviour (p<.001) and lower levels of sensory seeking behaviour (p<.01). In contrast, we found no group differences in sensitivity to form or motion using threshold tasks. Preliminary analyses of VEPs showed group differences in response to motion onset but not pattern reversal. Specifically, for motion onset, the high AQ group had reduced peak amplitude for the N2 waveform.

Conclusions: We show that typical adults with autistic-like traits have sensory processing styles similar to people with ASD (Crane et al., 2009). VEPs suggest group differences in the processing of motion, but not pattern. Group differences were not large enough to be evident using threshold tasks, at least under the tasks demands of the current study. It is possible that the equivocal findings of atypical vision in ASD reflect task difficulty and/or subtle differences in visual processing that are difficult to measure in high functioning people with ASD or typical adults with autistic-like traits.