Visual Social Attention in Infants at Risk for Autism Spectrum Disorders Differs Between Schematic and Live-Action Social Scenes

Background: Atypical visual attention to faces is a hallmark of autism spectrum disorders (ASD). Studies of infant siblings of children with ASD have investigated the extent to which this behavior serves as an early marker of the disorder. Atypical spontaneous attention to infant-directed social stimuli has been demonstrated in 6 month-old infants with ASD. Given that difficulties in gaze processing in ASD may stem from an aversion to eye contact, it is possible that altered patterns of visual social attention are a consequence of a model’s direct gaze in infant-directed bids. Here, we asked whether visual fixation patterns reliably distinguish infants at risk for ASD under less demanding social contexts.

Objectives:   The current study investigates social attention among infants at high and low risk for developing ASD while viewing two classes of dynamic social stimuli. These include schematic and live-action representations of social actors interacting with each other without direct intentions to engage the viewer.

Methods:   Six and 9 month old infants [high risk (HR) n = 17; low risk (LR) n = 29] viewed two 2-minute full audiovisual video segments taken from a cartoon (A Charlie Brown Christmas) and a children’s television program (Sesame Street). Videos were matched for duration, action sequences, motion, social interactions, and musical and linguistic content.  Eye movements were recorded with a Tobii T60 eye-tracker. ASD diagnoses were rendered using clinical best estimate and DSM criteria when children were 36 months old, assigning the HR-infant siblings into ASD (HR-ASD; N = 7), non-ASD delayed (HR-DD; N =3) and typical outcome groups (HR-TD; N = 7). The principle dependent variable was the proportion of infants’ attention to faces in the stimuli.

Results: Attention to faces when viewing the Charlie Brown and Sesame Street stimuli were similar within HR and LR groups, which did not differ by age. Infants exhibited greater attention to schematic faces than live-action faces (HR: t(20) = 6.046, p < 0.001; LR: t(24) = 10.579, p < 0.001). Attention to faces in Charlie Brown was similar among HR-ASD, HR-TD and LR infants. HR infants selectively attended less to faces in the Sesame Street stimuli than LR infants (p < 0.01). Among LR infants, attention to faces in Charlie Brown was highly correlated with attention to faces in Sesame Street (p < 0.05). However, this relation did not hold in the HR group. Attention to faces among the HR-ASD infants was strongly correlated with 24-month ADOS Social Affect scores, (r = .787, p = 0.008).

Conclusions: The infants exhibited greater visual attention toward schematic depictions of faces than real ones. Our preliminary results suggest that while attenuated attention to realistic social stimuli may be a behavioral endophenotype for ASD, attention to faces per se may not be sufficient for predicting outcome in HR-infants. The positive association between level of social impairment and social attention suggests that faces may provide different reward value for infants at risk for ASD. Additional research should further characterize active and passive looking behavior in infants at risk for ASD.