Figure 1: Intervention frequency in percent of cases by trauma group. Schema mode interventions were most frequent. Significant differences between the trauma groups were observed for the use of skills training. Bonferroni-adjusted Fisher tests between the trauma groups: * p< .05.

We then tested for higher symptom severity and less decrease in symptom severity in the trauma group (Hypothesis 2a, 2b). Adding trauma group as a factor significantly improved the models for anxiety and depression symptoms (BAI: F (1, 345.86) = 4.23, p = .041, BDI-II:F (1, 1,264.71) = 5.33, p = .021), indicating differences in symptom severity between the trauma groups across time. As expected, group means for the trauma group were higher than for the no-trauma group on BAI and BDI-II at the start and end of therapy, see Figure 2. Adding time as a factor improved the models for anxiety and depression symptoms (BAI: F (1, 190.53) = 118.27, p < .001, BDI-II: F (1, 401.71) = 356.67, p < .001), indicating a general decrease in symptom severity over time across both trauma groups. However, contrary to our expectation, adding the interaction effect between the trauma group and time point did not improve the models (BAI: F (1, 141.76) = 0.62, p = .431, BDI-II: F (1, 346.90) = 0.01, p = .938), indicating no evidence for differential symptom trajectories in the trauma group compared to the no trauma group.