Ordinary least squares regression was used to test the association of rs604300, ELS, and threat-related amygdala habituation. To maintain variability but constrain the influence of extreme outliers, prior to analyses all variables were Winsorized (i.e., outliers more than ± 3 standard deviations from the mean were set at ± 3 standard deviations from the mean; habituation n = 1; ELS n = 2). Sex and two ancestrally informative principal components were entered as covariates for all analyses. As in CATS and SAGE, additional covariates were entered representing the interactions between rs604300 and each covariate of no interest as well as ELS and each covariate of no interest (Keller, 2014).
