Correlates of Social Functioning in Autism Spectrum Disorder and Typical Development: A Systematic Review and Meta-Regression Analysis

Background: Differences in social functioning are a core feature of autism spectrum disorder (ASD). Several psychological constructs have been put forth as correlates of social functioning, in order to provide developmental and/or theoretical explanations for the social phenotype in ASD, as well as in typical development (Carter et al., 2013).

Objectives: The goal of this study is to synthesize literature reporting Pearson’s r effect sizes between social functioning and 8 putative correlates, including (a) central coherence, (b) executive functioning, (c) initiating joint attention (IJA), (d) responding to joint attention (RJA), (e) theory of mind (ToM), (f) imitation, (g) symbolic play, and (h) visual attention patterns to faces. Using a meta-regression framework, we generated summary effect sizes for correlations between social functioning and each construct, for children with ASD and for children who are TD. Further, we determined if summary effect sizes differ between populations with ASD and TD.

Methods: We conducted a structured search using ERC, ERIC, Medline, PsychINFO, and ProQuest Dissertations and Theses databases. To generate summary effect sizes and to conducted meta-regressions, we used a robust standard error approach to account for the nesting of effect sizes within studies.

Results: Currently, 100 studies (nearly half from grey literature) comprising a total of 725 effect sizes have been located. Effect sizes were calculated on a total of 9,906 participants. Average chronological ages of participants was 6 years 8 months for TD children, and 16 years 5 months for children with ASD. Summary effect sizes were weak (range = .05 - .48) across all constructs for both ASD and TD groups. In the TD group, summary effect sizes were significant and positive for Pearson’s r correlations calculated on social functioning and (a) executive function, (b) IJA, (c) RJA, and (d) ToM. The effect size for Play was positive, but not significantly different from zero. There were too few effect sizes available to calculate summary effects for central coherence and visual fixation in this group. In the ASD group, summary effect sizes were significant and positive for Pearson’s r correlations calculated on social functioning and (a) executive function, (b) imitation, and (c) ToM. The effect sizes for central coherence, IJA, and visual fixation were positive, but not significantly different from zero. There were too few effect sizes available to calculate summary effects for RJA and symbolic play in this group. An illustration of these results can be found in Figure 1. Across all constructs, effect sizes were larger in the ASD group as compared to the TD group, even after controlling for age (p = 0.02). We did not find any evidence of publication bias.

Conclusions: The results of this study indicate there are only modest associations between the several psychological correlates proposed to undergird social functioning, but that associations are generally stronger in ASD. Future studies should be conducted to further investigate these associations so that summary effects can be calculated, to better understand potential moderators of these associations, and to propose novel constructs that may better explain social functioning.