Fig 1.
Measures of social cognition and social affiliation.
All images are for illustrative purposes only and do not reflect items appearing in the actual test measures. Measures (1) and (2) used images adapted from the Ekman and Friesen database [40]. Illustration images adapted from CC-BY licensed images originally appearing in Skelly and Decety, 2012 [41]. Illustration image for measure (4) reprinted from Duchaine & Nakayama 2006 [19] under a CC BY license, with permission from Bradley C. Duchaine, original copyright 2006. Reprinted with permission. The illustration image for measure (4) was adapted from a public domain (CC0) photograph: http://www.pdpics.com/photo/3564-drum-seller/
Table 1.
Sample Characteristics.
Fig 2.
Adversity prevalences and relationships with PCA-derived components.
In the top part of the figure, bars represent the number of participants who were coded as exposed to each type of adversity. The bottom of the figure shows a heatmap relating each adversity to a set of four principal components derived from principal components analysis with varimax rotation. Darker shades of red represent higher loading of an adversity on a particular component.
Table 2.
Summary scores and internal reliability for each social cognitive or social functioning measure.
Fig 3.
Relationship between PCA-derived adversity components and six domains of social cognition and social affiliation.
Results are given in terms of standardized regression coefficients, after controlling for influence of age, sex, and race/ethnicity on each dependent measure. Solid lines give 95% confidence intervals for each effect size estimate. Three asterisks indicate associations that were significant at p < 0.0021 (Bonferroni corrected for all comparisons), two asterisks indicate associations significant at p < 0.0125 (Bonferroni corrected for number of orthogonal comparisons in each model), and one asterisk indicates nominal significance at p < 0.05. Theory of mind ability, social motivation, and social support were all robustly associated with parental maltreatment—with reductions in scores across all three measures. Parental maladjustment was most associated with reduced theory of mind ability, parental neglect with reduced social support in adulthood, and sexual abuse / institutional care with reduced social motivation in adulthood. None of the face discrimination or face recognition memory showed more than nominal associations with childhood adversity.
Fig 4.
Associations between individual adversity exposure and social cognition / social affiliation measures.
Results are given in terms of Cohen’s d by exposure (exposed versus non-exposed), controlling for influence of age, sex, and race/ethnicity on each dependent measure. Solid lines give 95% confidence intervals for Cohen’s d. Three asterisks indicate associations that were significant based on independent samples-t-test at p < 0.00033 (Bonferroni corrected for all comparisons), two asterisks indicate associations significant at p < 0.002 (Bonferroni corrected for number of adversities), and one asterisk indicates nominal significance at p < 0.05. As in the previous analysis, a number of childhood adversities robustly predicted theory of mind ability, social motivation, and social support, but none predicted face discrimination or face memory at above nominal levels of significance.