Definition, Measurement, and Validation of Resilience and Canalization in the Early Autism Phenotype

Background: Autism symptoms in toddlerhood result from early emerging brain function differences observed across multiple systems. These initially subtle differences become compounded and amplified due to atypical interactions within developing brain systems and with the external environment (Elsabbagh and Johnson, 2010). The reasons why this compounding occurs in some infants but not in others, especially in cases of familial risk, remains unknown. While a majority of infants at risk will not develop autism as toddlers, they exhibit highly variable pathways. Some have suggested the preliminary possibility that at least in some cases of familial risk, the infant brain may spontaneously re-organize in order to prevent the onset of disabling symptoms from emerging later on (Elsabbagh, 2012). No study has yet operationalized or validated putative canalization.

Objectives: We previously reported a characteristic brain function response to static faces displaying direct vs. averted gaze in infants at-risk who go on to exhibit typical outcomes in toddlerhood. The control group and infants who later develop autism showed no differentiation between direct and averted gaze. We interpreted this pattern as reflecting recruitment of unique and potentially protective neural mechanisms in this group, i.e., canalization. The objective of the current study is to validate this preliminary possibility by investigating potential associations between these brain function mechanisms and more naturalistic measures of parent-child interaction. If these neural differences indeed reflect protective operations, they should be associated with more optimal individual infant behavior within the context of dyadic interactions. 

Methods: Participants were drawn from the British Autism Study of Infant Siblings and included 104 (54 at-risk and 50 control) 7-month-old infants. Approximately 30% of the infants in the at-risk group received an autism diagnosis at 3-years. Data from two tasks were used for the current study. In the first task, ERPs were recorded while infants viewed images of female faces with static direct vs. averted gaze. The second task was a Parent-child interaction (PCI) assessment, where a 6-min episode of unstructured play was videotaped during which the parent-infant dyad was seated on a floor mat in a room with a small set of toys.

Results: Our findings suggest that P400 response differentiation to direct vs. averted gaze, a pattern unique to the infants at-risk who exhibit typical outcomes in toddlerhood, was associated with parent-child interactions. Specifically, a larger P400 response difference is associated with enhanced infant positive affect (r= .551 p <.001) and dyadic mutuality  (r= .553; p=.009). The same associations were not observed in the control group nor in infants at-risk who went on to develop ASD, reflecting recruitment of unique neural mechanisms early in life for a subset of infants who ‘canalize’ the impact of risk.  

Conclusions:  There is enormous value in characterizing and understanding factors promoting brain reorganization; such cases potentially inform how early environments can be modified to prevent disabling consequences from emerging. Our findings suggest that patterns of canalization are measurable early on and associated with more optimal behavior of the infant within their familial environments.