All analyses were stratified by sex and clustered by family to account for relatedness. Primary analyses examined discrete childhood trauma variables (CPAT, CSAT, CNAT) and APOEε4 as predictors of intercept and slope of frontal alpha EEG inter- and intra-hemispheric coherence during adolescence and young adulthood using latent growth curve models (LGCM) for those with three or more EEG assessments. LGCM models were computed using Mplus version 8.9 (Muthén & Muthén, 1998-2017), which employs full information maximum likelihood to account for missing data.
