Objectives: Prior work studying infants at-risk for ASD has pointed to general deficits in attention disengagement and work with older ASD individuals has showed deficits in emotion recognition. In order to follow up on these findings, the present study asks whether infants at-risk for ASD will a) show attention disengagement to faces that is modulated by emotional expression; and b) whether disengagement is influenced by scanning patterns to core facial features.
Methods: Using a Tobii eye-tracker, eye movements were recorded in 9-month-old infants at-risk for ASD (by virtue of having at least one older sibling with ASD) during both a looking time task and an overlap task. During the looking time task, infants were presented with images of fear, happy, and neutral faces. Eye-tracking data captured duration of looking to each face, as well as duration of looking to core features of the face (i.e., eyes and mouth). During the overlap task, infants were presented with a centrally-located face displaying a fear, happy, or neutral expression, and this stimulus remained present while a peripheral target appeared on the right or left side of the screen. Eye-tracking data examined latency to disengage from the face as a function of emotional expression. These preliminary findings are based on 12 9-month-old infants at-risk for ASD.
Results: Contrary to prior work with typically-developing infants that finds an attentional bias for fear faces, the facial expression of the central stimulus had no significant effect on attention disengagement latencies in at-risk infants, F(2, 22) = 0.77, p = .48. It took equally long for infants to disengage attention from fear faces (M = 549 ms) as compared to happy, M = 514 ms, t(11) = 1.024, p = 0.33, and neutral faces, M = 557, t(11) = -0.19, p = 0.85. Individual variability in the fear bias was then assessed as a function of visual scanning patterns to fear, happy, and neutral faces during the looking time task. Results indicate that infants who spend a greater proportion of time looking to the eyes and mouth when viewing fear faces show longer disengagement latencies to fear faces as compared to happy and neutral (r = .64, p = .046).
Conclusions: The present work highlights the importance of examining individual differences in these at-risk samples in order to better characterize the heterogeneity of ASD. This approach provides a window into the mechanisms by which early visual attention could influence lower level attentional biases.