A systematic review and meta-regression analysis of social functioning correlates in autism and typical development

Differences in social functioning are a hallmark feature of autism spectrum disorder (ASD), and much research has been devoted to locating cognitive and developmental explanations for this domain. To sort through this literature, we conducted a systematic review and meta-analysis that quantifies the extent to which several of these candidate constructs are associated with social functioning. We gathered 881 effect sizes calculated from 133 unique participant samples, and synthesized Pearson's r correlations between social functioning and three cognitive constructs; (a) theory of mind (ToM), (b) executive function, and (c) central coherence, and five developmental constructs: (d) initiating joint attention, (e) responding to joint attention, (f) imitation, (g) pretend play, and (h) visual fixation to social stimuli. We synthesized effect sizes using robust variance estimation for each putative correlate, for populations with ASD and typical development (TD) separately. We also conducted a series of meta-regressions to determine if sample and study features moderated effect sizes. We found that, in the ASD group, effect size estimates were significant and small (<0.30) for ToM, executive function, and initiating joint attention. Effect size estimates were significant and moderate (0.30 < r < 0.50) for imitation and response to joint attention. In the TD group, effect size estimates for ToM, executive function, and initiating joint attention were significant and small. In a meta-regression collapsed across correlates, we found that effect sizes were significantly larger in the ASD group (P < 0.05) and decreased as mental age increased (P < 0.001). Autism Res 2019, 12: 152-175 © 2018 International Society for Autism Research, Wiley Periodicals, Inc. LAY SUMMARY: In this meta-analysis, we found that correlations between social functioning and several constructs used to explain the developmental or cognitive origins of social functioning were quite low. This could mean that researchers will need to develop new theories about social functioning in ASD.