Using Event-Related Potentials to Explore Age-Related Changes In Infant Responses to Faces

Background: Within the first year of life, there are important changes in an infant’s neural response to social stimuli.  One replicated finding documents age-related changes in infant response to familiar versus unfamiliar faces, such that younger infants show a larger neural response to familiar faces, whereas older infants show a larger neural response to unfamiliar faces.  This shift is thought to be associated with important milestones in social development and occurs around the first birthday.  It is not known whether infants at risk for ASD – a population at risk for early social impairments – exhibit this pattern of age-related changes in neural response to faces.

Objectives: The objective was to determine whether children at risk for ASD show the anticipated age-related pattern of neural response to faces.  The expected pattern was characterized by a shift from a larger response to familiar (i.e, mother’s) face at younger ages, followed by a larger response to unfamiliar (i.e., stranger’s) face at older ages. 

Methods: Data were collected from infants at high risk for ASD (HRA) and low risk control (LRC) infants at 6, 9 and 12 months of age.  ERPs were recorded using a 64- or 128-channel Hydrocel Geodesic Sensor Net.  All signals were referenced to a single vertex electrode, sampled at 250 Hz, and filtered using a bandpass of  0.1 to 100 Hz.  Stimuli included images of the infant’s mother and an unfamiliar stranger.  The component of interest was the Nc, a negative-going fronto-central deflection occurring 400-850 milliseconds after stimulus onset.  Mean amplitude was calculated from a region of interest that included 14 electrodes.

Results: A 3 (age: 6, 9, 12) X 3 (hemisphere: left, midline, right) X 2 (condition: mother, stranger) mixed model examined effects on mean amplitude of the Nc.  There was a main effect of hemisphere F(2, 174) = 5.75, p < .01, such that the neural response was smaller over the right hemisphere than over the left hemisphere (p<.01) and midline leads (p=.01).  In addition, an age X condition X group interaction was found.  Follow-up tests revealed no within- or between- subject differences at 6 months of age.  By 9 months of age, the LRC group demonstrated the anticipated effect of condition (i.e., larger response to mother’s face than stranger’s face, t = 2.00, p=.04), where the HRA group did not show a difference across conditions.  Analyses at 12 months indicated that the LRC revealed the expected age-related shift, with a larger response to stranger than mother (t=2.87, p<.01), while the HRA group again showed no difference across conditions.

Conclusions:   In the present investigation, the control infants at low risk for ASD showed the anticipated age-related changes in response to familiar and unfamiliar faces, showing differential responses starting at 9 months of age.  However, the infants at high risk for ASD did not show a differential response between their mother’s face and a stranger’s face at any age, suggesting that the processing of important social stimuli may be atypical in children with a family history of ASD.