Gaze and social functioning associations in autism spectrum disorder: A systematic review and meta-analysis.

Almost every included study measured gaze and social functioning at the same point in time, so this paper cannot say whether gaze differences cause the social difficulties, follow from them, or both. Only a small slice of the evidence could be pooled: 9 studies for the face, 8 for the eyes and 7 for the mouth, chosen because they reported an ADOS global score with a fixation or looking-time measure. Many of the other 95 studies could not be combined. The confidence intervals are wide, and for the face and mouth the studies varied more than chance would explain, so the size of these correlations is uncertain. Thirty-five of the 95 studies examined this relationship as a side question rather than their main aim, often with small samples and cross-sectional designs. Samples were mostly male (from 53% to 100% male) and no study examined whether the relationship differs between males and females. Who rated the autism symptom measure changed the size of the association, so part of the result may reflect the rater rather than the person being rated. The authors could not assess publication bias, because funnel plots do not cope with several estimates coming from the same study.

The authors declare no conflicts of interest. The work was funded by the Australian Research Council.

Do not read the face and eye correlations as evidence that teaching a child to look at faces or eyes will improve their social functioning. Nearly every study measured gaze and social functioning at the same time, so the direction of the relationship cannot be established from these data, and the pooled result for the mouth was negligible. The authors also note that gaze measures are not yet an established biomarker for autism.