Equivalent Visual Sensitivity to Human, Animal and Object Motion In Children with Autism Spectrum Disorder

Background: Relative to typically developing (TD) children, those with Autism Spectrum Disorder (ASD) exhibit atypical processing of human motion. For example, observers with ASD demonstrate equivalent visual sensitivity to point-light depictions of human motion and object motion while typical observers show enhanced sensitivity to human motion. This difference could indicate that observers with ASD show atypical visual processing of (1) human motion, per se, or (2) all biological motions, in general.

Objectives: We sought to determine whether ASD is associated with atypical visual sensitivity to human movement, in specific, or to human and non-human animal movements, in general. To that end, we compared visual sensitivity to point-light displays of human, dog, and tractor motion in matched observers with and without ASD.

Methods: Fifteen children with ASD and 18 TD children participated in the study. Groups were matched on chronological age (ASD=12.08; TD=11.33 years) and Performance IQ (ASD = 98; TD =103). Participants performed a psychophysical coherent motion detection task with masked point-light displays of human, dog, and tractor motion. On each trial, participants viewed a 5-second movie depicting point-lights that were either attached to a locomoting person, dog, or tractor (and thus moved coherently) or were scrambled such that no coherent motion was present. Observers reported with a Yes/No button-press whether the point-lights were “stuck” to a person (or dog or tractor, depending on the block). Participants completed a practice session with unmasked displays prior to the experimental task.

Results: Visual sensitivities (d-prime) to the presence of coherent human, dog, and tractor motion were calculated for each participant with each stimulus type. The TD and ASD groups performed above chance in all conditions (all ps < .005). A group by condition analysis revealed that detection performance across the ASD and TD groups differed for human (t(21.682) = -3.339, p = .003) (degrees of freedom adjusted for unequal group variance) and dog (t(31) = -3.269, p = .003) motion but not tractor (t(31) = -1.661, p = .107) motion. TD children exhibited enhanced detection sensitivity to coherent human (t(17) = 3.874, p = .001) and dog (t(17) = 2.555, p = .020) motions relative to tractor motion, and equivalent sensitivity, within the biological domain, to human and dog motion (t(17) = 1.272, p = .220). Conversely, children with ASD showed no significant differences in detection sensitivity across the three motion types (all ps > .287).

Conclusions: The results of this study suggest that observers with ASD are equally sensitive to biological and non-biological motion whereas TD observers have visual systems that are tuned to biological motion. The results characterize the nature of atypical processing of biological motion in children with ASD and indicate that this disruption does not appear to be specific to human motion.  The typical visual system becomes increasingly tuned to human versus animal motion into adulthood.  Insufficient tuning of the visual system for the detection of the actions of other people may derail social processing in observers with ASD.