Next, we examined the association between childhood trauma exposure and the intercepts and slopes of EEGc. The intercept estimates individual variation in starting values, which may reflect different types of exposures prior to the measurement period, as well as other trait differences. The slope reflects how change in EEGc occurs over time, on average. Given the known co-occurrence of different types of trauma exposures, the models allowed for simultaneous testing of associations between trauma exposure types and EEGc intercept and slope. The models also evaluated whether intercept and growth parameters in EEGc were associated with AUDsx in the full sample. In a subset of the sample that endorsed exposure to a potentially traumatic event (i.e. met Criterion A for PTSD diagnosis) at any point from baseline to follow-up 3, we added PTSDsx as an additional dependent variable to account for the co-occurrence of AUD and PTSD. Figure 3 is an illustration of the path diagram for the analytic model.
