Visual Sensitivity to Angry Point-Light Walkers Varies as a Function of the Observer's Autistic Traits

Background: Previous research has shown that typical adults demonstrate an “anger superiority effect” in their heightened visual sensitivity to potentially threatening human actions. For example, typical adults are more sensitive to the presence of angry human gaits than to the presence of happy, fearful, sad, or neutral gaits. The visual perception of emotional human motion depends upon the Superior Temporal Sulcus and the amygdala.  Autism is associated with anatomical and functional abnormalities in these “social brain” areas. This suggests that the anger superiority effect may be modulated by observers’ autistic traits.

Objectives: Do typical adult observers vary in their visual sensitivity to emotional gaits as a function of their autistic tendencies?

Methods: Twenty-three participants were recruited from the Rutgers student body. They completed the Autism-Spectrum Quotient (AQ) questionnaire that measures the magnitude of autistic traits in typical adults with normal intelligence. Then participants completed a psychophysical task.  Emotional human gaits were depicted in point-light movies.  Each point-light walker displayed anger, happiness, fear, sadness, or a neutral emotional state and was hidden within a cloud or mask of identically moving points.  On half of the trials, the starting locations of the points defining the walker were scrambled so that the walker was “absent.” Participants viewed each masked point-light movie for 3 seconds and reported with a button press whether a walker was “present” or “absent” in the mask. Emotion was neither mentioned nor explicitly judged.

Results: AQ scores fell in the expected normal range (Mean = 17.35, SD = 4.69). A median split of AQ scores divided participants into low and high AQ groups. Each group’s performance on the psychophysical task was analyzed as the mean D-prime for each emotional gait. D-primes were computed by subtracting the standardized rate of false alarms from the standardized rate of hits. Overall, walker detection performance was best with angry walkers. However, there was a significant effect of AQ group on visual sensitivity to angry walkers (F(21,1) = 4.719, p < .05) as the low AQ (fewer autistic traits) group showed a significant “anger superiority effect” while the high AQ (more autistic traits) group did not. Notably, AQ score did not influence sensitivity to walkers exhibiting any other emotions. Results from a separate control study with inverted walkers support the linkage between AQ and threat detection.

Conclusions: Psychophysical experiments examined the relationship between social skills and perceptual sensitivity to emotional human movement. The results indicate that typical adult observers with high AQ scores (more autistic traits) do not demonstrate heightened sensitivity to point-light displays of angry, and thus potentially threatening, human gaits. Conversely, typical observers with fewer autistic traits (low AQ) exhibit selectively enhanced visual sensitivity to such potential threat. These results (1) support the idea that autistic traits extend into the general population, (2) compliment and extend findings of impaired visual analysis of human motion in individuals with autism, and (3) highlight the critical connection between social skills and visual sensitivity to negative emotions.