Visual Biases and Attentional Inflexibilities Differentiate Those at Elevated Likelihood of Autism: An Eye-Tracking Study

INTRO: Autism Spectrum Disorder (ASD) is characterized by a complex and vacillating behavioral and clinical profile making diagnosis difficult at earlier ages; however, some have successfully identified symptoms of ASD as early as 9-12 months, often using prospective samples (Lee et al., 2019; Ozonoff et al., 2010). Earlier identification of ASD and markers of ASD-risk is vital as it allows for earlier access to early intervention during critical developmental periods, to optimize developmental and participation outcomes. A key diagnostic marker of ASD is social deficits, therefore it is particularly advantageous to identify early markers related to social attention in infants who may be at elevated likelihood of ASD (EL-ASD). This will enable these infants and their caregivers to receive occupational therapy services specifically targeted to their social participation needs earlier in life. We believe that social attentional biases may have a significant impact on optimal functioning. In this paper, we address the following aims: (1) Evaluate the role of attentional biases (social versus nonsocial) in differentiating EL-ASD from those with lower likelihood (LL-ASD) (2) identify possible strategies occupational therapists can use to promote improved occupational engagement and social participation. METHODS: Infants (11 to 16 months) were recruited into one of two groups: EL-ASD based on the dual cut-off criteria of the First Years Inventory Lite (FYI-LITE V3.1b), (Baranek, Watson, Crais, Turner-Brown, & Reznick, 2014; n=21) and LL-ASD (n=20). Eye gaze patterns were collected via a Gazepoint GP3 HD eye tracker while infants passively viewed a CDS paradigm that consisted of a baseline video (classical music and minimal visual stimuli), three CDS videos (a puppet narrative, a book story-telling, and an adult playing with a toy using non-sensical speech), and a second baseline. Areas of interest (AOIs) were mapped for each scene. We examined differences in social and non-social attention, overall attention and number of discrete fixations by AOI using t-tests. To evaluate dyadic attention, switching attention between two stimuli, longitudinal patterns of fixation were examined by identifying the frequency of patterns of interest (e.g., fixating on the book, then face; fixating on the face then other portions of the screen; fixating on a toy then face). RESULTS: There were no differences in overall attention to the screen (across all scenes) between groups. Int