Gaze Preference and Underlying Brain Responses to Dynamic Eye Movement in Individuals with ASD Across Development

Background: Impairments in eye contact are a notable symptom of autism spectrum disorder (ASD). Past research has shown individuals with ASD demonstrate atypical gaze responses and brain activity to passive electroencephalography (EEG) paradigms. Brain-behavior responses during dynamic social paradigms are essential to understand individuals with ASD in more ecologically valid contexts. Understanding these relationships in the context of child development is critical for the creation of clinically practicable biomarkers.

Objectives: The primary aim of the study was to explore brain-behavior relationships in ASD during an interactive social neuroscience paradigm. Specifically, this study assessed the relationship between the N170 and gaze behavior during a gaze-contingent paradigm in children and adults with ASD.

Methods: Participants included individuals with ASD (n = 78) and typically developing (TD) individuals (n = 56) matched on age (range: 8–36 years) and IQ (range of standard scores: 71–158). EEG was recorded with a 128-channel Geodesic Sensor Net, and eye-tracking (ET) was recorded with an EyeLink-1000 remote camera system. ET and EEG were co-recorded while the participant underwent a gaze-contingent viewing paradigm. Participants viewed a cross hair followed by a face. Once the participant looked at the eyes, the face responded by looking at (direct-gaze) or away from (averted-gaze) the participant. Behavioral data, time spent looking at facial features, and neural responses, N170 amplitude and latency, were collected.

Results: TD individuals looked significantly more at the left eye than between the eyes [F(1, 108) = 11.19, p < .01], while individuals with ASD did not demonstrate a gaze preference. Across groups, increased time spent looking at the left eye was associated with faster N170 response in the averted-gaze condition (r = -.222, p = .03). Gaze-condition effects were present in both groups; specifically, the N170 amplitude to direct-gaze was more negative than to averted-gaze [F(1, 108) = 6.95, p = .01)], and N170 latencies were faster in response to direct-gaze [F(1, 108) = 8.55, p < .01]. A main effect of hemisphere revealed N170 amplitudes were more negative in the right hemisphere than the left for both ASD and TD groups across conditions [F(1, 108) = 6.72, p = .01]. In the averted condition, right and left N170 amplitudes were highly correlated in the TD group (r = .47, p < .01) but not in the ASD group (r = .11, p = .37). Across groups, increasing age was related to faster N170s in the right hemisphere in the direct condition (r = -.25, p < .01).

Conclusions: This study demonstrated differences in gaze preference and brain responses to direct and averted gaze. The robust relationship between gaze and diagnostic status revealed atypical viewing patterns in ASD. Differences may be attributable to variation in cross-hemisphere connectivity as demonstrated by synchronization of the N170 amplitude in TD individuals but not those with ASD. Both groups showed developmental effects, with efficiency of gaze perception increasing with chronological age. These results add to the body of evidence supporting N170 as a promising social-communicative biomarker.