Objectives: Our goal was to examine whether eye gaze processing in infants at-risk predicts autism diagnosis at three years of age. Furthermore, we characterize processes of risk and resilience in the infant brain at 6-months for the entire group of infants at-risk that go on to develop variable outcomes in toddlerhood. We use a data-driven approach to ascertain whether event-related potential (ERP) findings fit with hypothetical profiles corresponding to a range of outcomes including canalization.
Methods: Participants were from the British Autism Study of Infant Siblings (BASIS). One hundred infants (54 high-risk sibs and 50 low-risk controls) were included in the analysis. When aged between 6 and 10 months, ERPs were recorded while the infants viewed various gaze and face contrasts. Outcome measures at 2- and 3-years of age included a range of standardized measures combined with expert clinical judgment to ascertain outcome classification.
Results: Atypical response to gaze shifts at 6-months characterized the group that went on to develop autism at 3-years of age. This is the first study to demonstrate that subtle signs observed within the first year of life are associated with emerging behavioral symptoms of autism in toddlerhood.
On the whole, relative to the control group, the high-risk siblings group showed both similarities and differences in the amplitude and latency of components related to gaze processing. However, within the high-risk group, the variation in individual infant’s ERP response characteristics was associated with variable outcome measures. Atypical response to static gaze characterized the group of infants at-risk who went on to exhibit typical outcomes. On several additional measures, the three groups were indistinguishable.
Conclusions: As a group, infants at-risk for autism show differences in certain neural components related to the processing of face and gaze. Moreover, individual differences in the infant ERP could be mapped onto behavioral characteristics of the same infants when they reach toddlerhood. The findings help to highlight the potential scientific and clinical utility of infant ERP measures. Specifically, we propose three hypothetical profiles as targets for future research with infant siblings. Infants within the first year of life may exhibit risk markers that are (a) compounded and amplified over time leading to cascading effects across multiple domains of functioning and a diagnosis, (b) expressed subtly and remain stable leading to characteristics of the broader autism phenotype, or (c) canalized through influences of protective genetic and/or environmental factors leading to typical behavioral outcome.