Atypical Neural Responses to Direct Gaze from a Live Person in Autism

Friday, May 16, 2014: 11:18 AM
Imperial A (Marriott Marquis Atlanta)
L. A. Harrison1, J. M. Tyszka2, J. Elison3 and R. Adolphs4, (1)Computation and Neural Systems, California Institute of Technology, Pasadena, CA, (2)Biology and Biological Engineering, California Institute of Technology, Pasadena, CA, (3)University of Minnesota, Minneapolis, MN, (4)Humanities and Social Sciences; Biology, California Institute of Technology, Pasadena, CA

In 1943, Leo Kanner observed children with autism having “a far better relationship with pictures of people than with people themselves.” While a large literature documents abnormal processing of faces and eye gaze in people with autism spectrum disorder, the vast majority of stimuli used in those studies have a series shortcomings: they are pictures or videos, rather than people in the flesh. Yet, as Kanner’s quoted observation and our own experience argue, pictures and real people are processed in quite different ways, and can be dissociatively impaired in autism. 

As autism is partially characterized by a persistent deficit in social interaction, it is reasonable to presume deficits, including in gaze, may be accentuated in actual interaction with another person.   Findings from eye tracking studies indicate gaze behavior is influenced by the direct presence of another person. It is unknown whether direct live gaze is differently represented in the brains of individuals with autism.


This study comprised two main objectives. The first objective was to determine whether controls and individuals with autism are sensitive to the difference between live and recorded gaze, namely, whether the neural response to gaze changes as a function of the presence of another person. The second objective of the study was to characterize the differential neural response to live and recorded gaze in autism relative to controls.


We explored these aims using a blocked design fMRI experiment. Participants included 15 males with autism, and 15 age, gender, and IQ-matched controls. In a Live condition, participants monitored the gaze of a live actor sitting behind the bore of the magnet. Audio instructions delivered to the actor’s headphones cued their gaze: actor conditions consisted of direct gaze, averted gaze, or eyes closed. In the Recorded condition, equivalent video recordings of the Live condition were used.  


Networks of activation revealed a set of partially distinct networks activated in autism compared to controls by direct live gaze. The autism and control groups were primarily differentiated by an anterior/posterior separation in the Live but not Recorded condition. Individuals with autism had less anterior, but more posterior activation during the Live condition than controls. Meanwhile, in the Recorded condition, anterior activity in autism was more similar to that of controls. 


We demonstrate that the neural response to live gaze is distinct in autism; an interactive experiment was necessary to capture this difference. This finding provides significant insight into the difficulty individuals with autism face in interacting with other people in everyday life.