Atypical Face Scanning in Unaffected High-Risk Infant Siblings

Background:

Toddlers with Autism Spectrum Disorder (ASD) exhibit an atypical distribution of attention when examining faces. It is not clear, however, when these abnormalities first manifest, and whether these attentional alterations are shared by unaffected first-degree relatives.  

Objectives:

To examine developmental changes in the scanning of static, dynamic, and speaking faces in the first year of life in low-risk typically developing (TD) infants (N=24) and in infants who, due to genetic liability, are at high-risk for ASD (N=31).  The latter group was further subdivided based on 18-month assessment into those who developed symptoms of autism spectrum disorder (ASD, N=5) and those who appeared unaffected (NAF, N=26). An additional group of infants with questionable phenotype at 12 or 18 months were excluded from this analysis.

Methods:

Infants were tested at 3, 6, 9, and 12 months.  Each infant viewed three types of faces: Static, Dynamic, and Speaking.  Responses were recorded with an eyetracker. Average percentages of time spent monitoring specified face regions were examined as a function of age, group, and display.  ASD infant comparisons were qualitative due to small N.  

Results:

Attention to faces in general and to their internal features (eyes, nose, and mouth) increased rapidly from 3 to 6 months in TD and NAF groups.  Regardless of age, infants primarily monitored the eyes in the Static condition, less in the Dynamic condition, and least during Speech. The pattern was reversed for the mouth region, that was primarily attended to during Speech and least during the Static condition.  Specifically for mouth viewing, age differentially affected attention between groups.  In TD infants, it increased rapidly between 3 and 9 months, likely reflecting reliance on visual cues in encoding speech sounds.  Between 9 and 12 months attention to the mouth dropped significantly in TD group possibly due to a lesser reliance on lip-reading in speech perception.  However, in NAF infants, attention to mouth continued to increase from 3 to 12 months.  NAF infants also spent more time examining outer facial features than TD controls.  In comparison, responses of infants with ASD were variable and different from NAF and TD infants, with differences becoming particularly apparent at 9 months in the condition involving infant-directed speech.   

Conclusions:

TD and NAF infants modulate their scanning strategies depending on the context in which faces are presented.  Amount of attention directed to the mouth varies by age, most likely due to its role in speech perception.  Though, unaffected high-risk infants exhibit a less mature strategy reflected in more time spent examining the outer regions of the face.  At 12 months they spent more time examining the mouth than TD controls, suggesting that they might still rely on visual processing of articulatory movements for sound discrimination.  These findings suggest that even apparently symptom-free infants show mildly atypical face scanning patterns associated with their genetic risk status possibly linked to purported vulnerabilities in language development.  Considering high variability amongst symptomatic infants, their results will be discussed individually in a context of symptom severity and levels of developmental skills.