Temporal Profiles of Social Attention Are Different Across Development in Autistic and Neurotypical People
Abstract
Background
Sociocommunicative difficulties, including abnormalities in eye contact, are core diagnostic features of autism spectrum disorder (ASD). Many studies have used eye tracking to measure reduced attention to faces in autistic people; however, most of this work has not taken advantage of eye-tracking temporal resolution to examine temporal profiles of attention.
Methods
We used growth curve analysis to model attention to static social scenes as a function of time in a large (N= 650) sample of autistic participants and neurotypical participants across a wide age range (6-30 years).
Results The model yielded distinct temporal profiles of attention to faces in the groups. Initially, both groups showed a relatively high probability of attending to faces, followed by decline after several seconds. The neurotypical participants, however, were significantly more likely to return their attention to faces in the latter part of each 20-second trial, with increasing probability with age. In contrast, the probability of returning to the face in the autistic participants remained low across development. In participants with ASD, more atypical profiles of attention were associated with lower Vineland Adaptive Behavior Scales communication scores and a higher curvature in one data-driven cluster correlated with symptom severity.
Conclusions
These findings show that social attention not only is reduced in ASD, but also differs in its temporal dynamics. The neurotypical participants became more sophisticated in how they deployed their social attention across age, a pattern that was significantly reduced in the participants with ASD, possibly reflecting delayed acquisition of social expertise.
Summary
Social Attention - our tendency of focusing on cues with social relevance in the environment, such as eyes and faces - has been long studied by comparing the grand-average of people with some characteristic in common. By measuring the overall Social Attention levels of autistic people, and comparing it with the same measure from neurotypical people, researchers were able to draw general differences, for example, that autistic people, on average, look less at other people’s faces and eyes. However, the same analysis revealed outstanding heterogeneity: while less Social Attention characterised Autism as a group, it was difficult to explain why this result could not be easily replicated.
This research tried to capitalise on heterogeneity and implement a method that could help to understand the changes of Social Attention. We unwrapped Social Attention, by looking at the second-by-second change of the looking time on the face, and focused on age, asking ourselves if we could situate this average difference in time.
We found that we could quantify a late peak (after 10-15 seconds of observation passed already) of the looking time to the face, emerging with age (between 6 and 30 years!). This trend was significantly stronger in neurotypical people.
We think that this finding might be related to the specialisation of the social brain towards face processing, derived from accumulating experience with other people. Autistic people might experience external social environment and/or self-directed experience to a less extent, thus not developing a peak as strong as the neurotypicals in this specific time-window.
Commentary
Daniel P. Kennedy, “Illuminating Autism Spectrum Disorder With Eye Tracking”, Biological Psychiatry: Cognitive Neuroscience and Neuroimaging, Volume 6, Issue 8, 2021, Pages 765-766, ISSN 2451-9022
Teresa Del Bianco
Postdoctoral Researcher
Scientist researching brain and child development and neurodiversity.