Fig. 8: Deviations of children and adults are linked to the social environment during early development.

To capture regional variation of deviation scores across 34 cortical regions, we applied Principal Component Analysis (PCA) to individual deviations (Fig. 1e). Next, we assessed the association between the first principal component (PC1) of regional deviations and early developmental social factors, including socio-economic status (SES) and quality of mother-infant relationship (Parent-Infant Relationship Index, PIRI; BLS-26 only) using Spearman rank correlation. a In preterm children of the ABCD-10 cohort, the association between SES with the first principal component (PC1) of regional cortical thickness (CTh) (left, Spearman rho(380) = –0.002, p = 0.977, p_FDR = 0.977, CI = [–0.102,0.098], two-sided) and of regional surface area (SA, right, Spearman rho(294) = 0.269, p = 0.001, p_FDR = 0.006, CI = [0.160,0.371], two-sided) is shown. b In preterm adults of the BLS-26 cohort, the association between CTh PC1 with SES (left, Spearman rho(94) = –0.217, p = 0.033, p_FDR = 0.088, CI = [–0.400, –0.018], two-sided) and with PIRI (Spearman rho(89) = − 0.211, p = 0.044, p_FDR = 0.088, CI = [–0.400, –0.006], two-sided) is illustrated. Orange lines and shades in panels (a, b) represent the linear regression line with 95% confidence intervals based on 10,000 bootstrap resamples. c In adults, the association between gestational age (GA) and CTh PC1 scores was moderated by SES (interaction term SES x GA in linear regression model: β = 0.104, p = 0.035, SE = 0.057, CI = [–0.0090,0.217], one-sided). Conditional effects of SES on GA are shown on the right. A low (red, effect = 0.246, p = 0.35 × 10 × 10^–4, one-sided) or middle (yellow, effect = 0.143, p = 0.001, one-sided) SES significantly moderates the relationship between SES and GA, whereas a high SES (green, effect = 0.039, p = 0.303, one-sided) does not. Source data are provided as a Source Data file.