Pupillary Responses to Manipulations of Stimuli Type and Synchrony in Children with Autism Spectrum Disorder
Objectives: The current study is the first to measure pupillary responses to dynamic, temporally manipulated audio-visual stimuli to infer cognitive processes involved in perception of non-social, social, and social-linguistic stimuli. The aims of the current study were threefold:
1) Determine if differences between pupillary responses to non-social and social-linguistic information are the result of difficulties in social or linguistic processing.
2) Determine if processing of asynchronously presented stimuli can be indexed through pupillary change, and if responses differ between typical development (TD) and ASD.
3) Determine if pupillary responses to stimuli relate to clinical symptomatology in ASD.
Methods: Participants included 39 children with ASD (Age; M=12.3, SD=3.2) and 32 TD children (Age; M=12.4, SD=3.0) who were matched for chronological and mental age using the WASI-II (Wechsler, 2011). Each participant was presented with audiovisual stimuli in 6 conditions in a 2x3 design. Stimuli were one of three types, non-social (a video of the children’s game Mousetrap), social non-linguistic (a video of an actress making verbal, non-speech sounds), and social-linguistic (the same actress speaking). Each of these stimulus types were presented synchronously or asynchronously (1000ms delay, collapsed across visual-leading and auditory-leading). Pupillary responses were captured and recorded using a Tobii X60 eye-tracker. Participants’ parents completed the Autism Quotient- Child Version as a measure of symptom severity.
Results: Children with ASD showed smaller pupillary responses to social information than TD children for both the social-linguistic (p=.009) and social non-linguistic (p=.005) conditions. However, there was no group difference in pupillary response between groups for the non-social condition (p=.376; Figure 1). Analysis of temporal processing was more equivocal. No significant Group x Stimulus x Synchrony interaction was apparent in the mean pupillary responses. However, more in-depth waveform analyses showed significant effects of synchrony in both the social non-linguistic and the socio-linguistic conditions (p’s<0.001), but not in the non-social condition (Figure 2). Furthermore, responses to asynchronous stimuli were significantly correlated with symptom severity in the social non-linguistic (p=.01) and social-linguistic (p=.03) conditions, but not the non-social conditions (p=.16) such that smaller pupillary responses were associated with more severe symptomatology.
Conclusions: These data suggest that the atypical response to socio-linguistic sensory information in individuals with ASD is a result of the social components of the information. This held for both overall processing differences seen with synchronous stimuli, and differences observed in multisensory temporal processing. Finally, differences in multisensory temporal processing of social stimuli were significantly correlated with symptom severity, providing convergent evidence that differences in early-stage physiological processing (i.e., pupillary change) may contribute to symptomatology, including communication issues.