Early Salient Network Connectivity and Attention to Faces in Infants at High-Risk for ASD

Background: Attenuated attention to social stimuli during infancy, including faces, has been recognized as a risk marker for developing ASD (Chawarska, Macari, & Shic, 2012; Jones & Klin, 2013). However, little is known about its neural underpinnings, which may inform the ontogeny of autism symptomatology. Mounting evidence implicates early alterations in functional brain connectivity in the emergence of ASD (Emerson et al., 2017). Of particular interest is the Salience Network (SN), a functional brain network involved in detecting and guiding attention to the most salient aspects of one’s internal and external environment. Altered SN connectivity complement behavioral observations of heightened attention to non-social sensory inputs at the expense of developmentally relevant social stimuli in ASD (e.g., Green et al., 2016). As faces represent a highly salient class of stimuli for young infants, SN connectivity may relate to trajectories in face-looking in infants at high-risk for ASD.

Objectives: The current study aims to determine whether patterns of SN connectivity at 6 weeks predict individual trajectories in social attention to faces during the first year in infants at high- (HR) and low-risk (LR) for ASD.

Methods: 50 HR and LR infants participated in a longitudinal study; risk status was determined by virtue of having at least one older sibling with a confirmed ASD diagnosis. At 6 weeks, infants completed an 8-minute resting-state fMRI scan during natural sleep. Eye-tracking was conducted at 3-, 6-, 9-, and 12-months of age while infants viewed video excerpts of Charlie Brown and Sesame Street. Percent fixation to the character’s faces was recorded and longitudinally analyzed with a Bayesian hierarchical linear model. SN connectivity at 6-weeks was identified using the right anterior insula (rAI) seed derived from an infant atlas (Shi et al., 2011), and estimates of individual rate of change in face-looking from 3- to 12-months was used as a covariate in the SN connectivity analysis.

Results: The eye-tracking analyses revealed a significant interaction between risk and age, such that LR infants increasingly attended to faces at a faster rate than HR infants (95% CI: 0.38-10.54; p<0.001). When eye-tracking data were incorporated in the 6-week resting state fMRI analyses, greater functional connectivity between the hub of the SN (i.e., rAI) and regions associated with social processing (i.e., anterior cingulate cortex, orbitofrontal cortices, inferior frontal gyrus) in LR infants, relative to HR infants, predicted greater increases in attention to faces (Z>2.3, p<0.01).

Conclusions: Six-week-old patterns of SN connectivity predicted developmental trends in social attention to faces, suggesting that very early functional connections between regions implicated in salience detection and social processing may be foundational to the perceived salience of faces observed in typical development. Conversely, the lack of a relationship between SN connectivity and social attention to faces in HR infants may imply a disruption in normative processes underlying the development of social attention; we found HR infants trailed LR infants in normative face-looking trajectories. Differences in SN connectivity at 6 weeks strongly suggest that neural underpinnings of ASD likely begin before birth, with cascading effects on social attention.