Increased Pupil Size to Emotional Faces in Infants at High Risk for Autism As an Early Predictor of Atypical Development

Thursday, May 15, 2014: 1:55 PM
Imperial A (Marriott Marquis Atlanta)
J. B. Wagner1, R. J. Luyster2,3, H. Tager-Flusberg4 and C. A. Nelson2, (1)Department of Psychology, College of Staten Island, CUNY, Staten Island, NY, (2)Boston Children's Hospital, Boston, MA, (3)Emerson College, Boston, MA, (4)Boston University, Boston, MA

Past work with individuals with ASD has found differences in both visual attention and autonomic measures of face processing, including reduced time on core features and heightened sympathetic arousal to faces.  Prospective work with infant siblings of children with ASD, a group with as high as a 1 in 5 chance of also developing the disorder, provides an opportunity to look for early markers of atypical face processing, some of which might be predictive of later ASD outcome or other developmental difficulties.


To understand the origins of face processing differences in individuals with ASD,  visual scanning and pupil diameter, a measure of sympathetic arousal, were examined in response to emotionally-salient faces in infant siblings of children with ASD.  These measures were then examined alongside 18-month social-communicative outcomes and ASD classification to look for early markers that might predict atypical development.


At 9-months-old, 38 infants at high risk for ASD (HRA) and 30 low-risk controls (LRC) were presented with static images of fearful, happy, and neutral faces.  Each face was presented twice for 10 seconds.  A Tobii eye-tracker captured visual attention and pupil size while infants viewed each face.

A set of longitudinal analyses followed HRA and LRC infants through 18 months to ask whether any features of attention and arousal might relate to later outcomes in these infants, as measured by the Communicative and Symbolic Behavior Scales (CSBS) and the Autism Diagnostic Observation Schedule (ADOS). 


Analyses of visual attention to the three emotional faces (both for overall time on faces and for proportion of time on eyes vs. mouth) revealed no influence of group on scanning patterns for LRC and HRA.  Analysis of pupil size revealed a significant difference between LRC and HRA, with greater pupil size in HRA than LRC (p= .042), suggesting that HRA infants show increased arousal to the emotionally-salient stimuli.  

A set of correlations examined 9-month-old attention and arousal to faces and social-communicative behavior at 18 months as measured by the CSBS.  In LRC, significantly greater attention to eyes at 9 months was related to higher CSBS social scores at both 12 and 18 months (rs > .51, ps < .03).  In HRA, there was no relation between attention to eyes and later CSBS scores; however, larger pupil size at 9 months was significantly correlated with worse overall CSBS outcomes at 18 months (r = -.44, p = .028).  A preliminary set of analyses also examined ASD classification at 18 months and found that the group difference at 9 months for pupil size between HRA and LRC was driven by the subset of HRA infants (N=9) who received a positive 18-month ADOS classification (p= .027).


Taken together, the CSBS and ADOS relations to pupil size in HRA suggest the importance of studying early arousal mechanisms as predictors of later developmental difficulties in ASD and the broader autism phenotype, particularly in the context of emotionally-salient stimuli, as is the focus of the present scientific panel.