Abstract
Sensory sensitivities are estimated to occur in 90% of individuals with autism, but little is known about the genetic underpinnings of these putative differences. Here, we sought to test how individuals with genetic conditions which are highly penetrant genetic causes of autism allocate visual attention in 360° real-world scenes using a traditional multi-level salience model.10 controls enrolled in this study, along with individuals with Fragile X condition (N = 8) and 16p.11.2 deletion syndrome (N = 4). Participants viewed 40 trials of real-world, 360° panoramic scenes using a head-mounted display. Viewing behavior was measured using an in-headset eye-tracker. During each trial (20s), participants actively explored one novel 360° scene, using head turns to change their viewpoint. We compared individuals’ gaze behavior using a traditional multi-level salience model. Gaze-behavior in individuals with 16p.11.2 deletions was relatively less driven by feature-level salience (t=2.86, p=0.015) and relatively more driven by meaning-level salience (t=−2.19, p=0.048). However, this pattern of gaze behavior was not observed for individuals with Fragile X, who showed comparable attentional allocation to controls at both levels (feature-level: t=2.02, p=0.062; meaning-level: t=−0.50, p=0.62). Further, the observed differences in gaze behavior between individuals with 16p and controls were not driven by basic differences in fixation number (t=0.95, p=0.36) or duration (t=−0.16, p=0.88). Similarly, individuals with Fragile X exhibited comparable fixation numbers and durations as controls (both t>0.40, both p>0.13). Our results suggest that attention in real-world scenes is guided by different features for individuals with genetic conditions linked to autism and controls. Together, these results suggest that quantitative measurements of real-world viewing may help distinguish between genetic subgroups of autism.