Visual Attention to Faces Is Related to Social Referencing: A Study of 16-Month-Old Infants at High and Low Risk for Autism

Background:   To understand the origins of autism spectrum disorder (ASD) and allow for earlier diagnosis, research has examined infants who are at heightened genetic risk based on their having an older sibling with ASD.  Up to 20% of infant-siblings may later receive an ASD diagnosis (Ozonoff et al., 2011).  We focus on lower levels of social interest across contexts as possible early signs of ASD.  Social referencing requires infants to distribute their attention between an object of interest and a social partner, but few prospective studies have investigated early social referencing deficits in ASD (Cornew et al., 2012).  Moreover, no studies have examined how underlying attentional demands relate to social referencing. 

Objectives:   The current study investigated visual attention and social referencing among 16-month-old infants with (high-risk, HR) and without (low-risk, LR) an older sibling with ASD.    We also examined whether visual attention differences (e.g., attention to faces during a videotaped social scene from Mr. Rogers) were related to variability in social referencing.  It was expected that HR infants would demonstrate less visual attention to faces and lower rates of social referencing compared to LR infants, and that visual attention differences to dynamic social scenes would be associated with lower rates of social referencing during the naturalistic interaction. 

Methods:   A prospective study design was used to compare HR infants (n=36) with LR infants (n=33).  Eye-tracking data were collected while 16-month-olds viewed a dynamic social scene of an adult (Mr. Rogers) speaking and showing toys. These same infants also participated in a social referencing paradigm in which the infant was introduced to an appealing but somewhat frightening novel toy in the presence of an experimenter and caregiver.  

Results:   During the dynamic social scene, HR infants spent a lower proportion of time looking at the upper half of the face (M=.13, SD=.23) compared to LR infants (M=.27, SD=.28) (F(1, 51)=4.46, p=.04, partial η²=.08).  HR infants also referenced the adults less frequently (M=1.69, SD=1.45) than did LR infants (M=2.71, SD=2.37) (F(1, 67)=4.72, p=.03, partial η²=.07).  In addition, infants who spent more time looking at the upper half of the face during the social scene were also more likely to reference an adult when shown the ambiguous toy (r=.35, p=.03).  In contrast, a higher proportion of looking time to the objects was associated with lower rates of social referencing (r=-.31, p=.04).   

Conclusions:  High risk infants showed lower social interest as reflected in both less visual attention to the upper half of the face and less social referencing at 16 months of age.  Importantly, attention to the upper half of the face was positively associated with social referencing, whereas attention to objects was negatively correlated with social referencing.  This study is the first to report an association between visual attention to faces and objects and social referencing during social interaction.  Further research is needed to understand how these measures relate to other indices of social interest and engagement and whether they will prove to be useful early signs of ASD.