Social Cognitive Atypicalities Associated with Preterm Birth: A Challenge to the Early Diagnosis of Autism

Background: Prospective research into the early signs of autism has focused largely on the infant siblings of children with a confirmed diagnosis. These infants have approximately 18 times higher likelihood of an autism diagnosis compared with the general population. This research has uncovered a number of early features which seem to be associated with later autism outcome. However many such features (e.g. reduced fixation on eyes) are contested and some are associated with autism-siblinghood but not with autism diagnosis specifically.  This has led to a questioning of the previously-convincing hypothesis that early autism is characterised by reduced social interest. Exploring early social cognition in preterm infants gives insight into both an alternative route to autism (since preterm birth is associated with approximately 4 times higher likelihood of diagnosis) as well as the association of early cognitive traits with general developmental delay (which is present in about 50% of infants born preterm).

Objectives: We tested the hypothesis that the early social cognitive phenotype associated with preterm birth differs from that of term controls. Our goal is to consider whether instead of being autism-specific, some early social cognitive atypicalities are generic markers of delayed development or associated with intellectual disability rather than autism.

Methods: We have recruited a large sample (n=92) of infants born preterm (less than 32 weeks gestational age, and weighing less than 1500g at birth). Of these, 38 have been assessed at approximately 9 months corrected age along with 37 typically-developing infants matched on age and gender.

Tasks were eye-tracker based assessments of social cognition previously used in comparative studies of populations with and without autism. The tasks increased in the complexity of the stimulus presented, but were all free-viewing tasks. Stimuli in order of complexity were: neutral, static faces; faces presented alongside other objects in a grid-like array; people depicted in naturalistic photographs alongside another photograph without people.

Results: We calculated looking time scores to social and non-social areas of interest. The preterm group fixated the eyes of a neutral face significantly less than the control group (mean difference = 0.62s, p=.036).  They also looked less at a face in a grid-like array (mean difference = 0.45s, p=.05) and looked less at naturalistic photos with social content (mean difference = 0.34s, p=.048). There were no differences in latency to fixate social areas of interest. In addition, preterm children showed a normal sized attentional disengagement effect in the gap-overlap task, indicating that generalised attention deficits were not contributing to differences in social attention. Final results will be presented from a predicted sample of approximately n=60 in each group.

Conclusions: In each task preterm infants showed atypical fixation on social content, a pattern previously thought to be specific to children who later receive an autism diagnosis, but perhaps instead associated with general cognitive delay. Final results will be considered in light of their impact on our understanding of early autism and specifically consequences for attempts to create early diagnostic assessments.