Eye Gaze in Autism Spectrum Disorder: A Review of Neural Evidence for the Eye Avoidance Hypothesis.

The review included only eleven neuroimaging studies, and eight of them supported the eye avoidance hypothesis. The included samples were mostly male and mostly higher-functioning autistic participants, so the findings may not apply to all people on the autism spectrum. The studies used pictures or short movies of faces, not real conversation, so they do not show what happens in clinic. Only one social-brain study compared gaze cuing with free viewing, so it cannot establish that reduced eye gaze caused brain differences. No study directly measured physiological arousal, such as heart rate or pupil size, during gaze cuing. Alexithymia and other individual differences were not controlled, and they may explain some mixed eye-gaze findings. Region-of-interest analyses were criticised for inflating group differences, and the review excluded most region-of-interest studies unless they compared free viewing with gaze cuing.

The paper reports funding from the University of Western Australia. It does not report a commercial funder or author conflict disclosure.

Do not conclude that teaching eye contact will improve social communication. This review reports brain activity differences during face viewing, not outcomes from a tested therapy, and it calls for more gaze-cuing research before causal claims.