Martijn van Teffelen

Chapter 6 126 manipulated in both benign and hostile directions in students (Hawkins & Cougle, 2013). This study also showed that participants experienced less anger when confronted with a provocative insult following benign interpretation training. The second study evidenced that eight sessions of benign interpretation training in a relatively small sample of aggressive boys resulted in greater reductions of HIB, anger and self-reported aggression compared to untrained controls (Vassilopoulos et al., 2014). A third randomized sham-controlled study in non-treatment seeking people with alcohol use disorder and elevated levels of trait anger showed that eight sessions of CBM-I resulted in greater improvements in interpretation bias, trait anger and self-reported anger expression (Cougle et al., 2017). The fourth study compared the effects of benign versus hostile training in students and showed that benign training resulted in an increase in prosocial interpretations and a reduction in anger and self-reported verbal aggression, whereas hostile training did not result in significant changes (AlMoghrabi et al., 2018). A fifth and final study in non-treatment seeking college students with major depressive disorder compared an eight- session CBM-I training to a sham control condition and demonstrated that CBM-I resulted in greater improvement in interpretation bias and anger control. However, no effects were found on depressive interpretation bias, depressive symptoms, or trait anger (Smith et al., 2018). Taken together, preliminary evidence suggests that CBM-I may be moderately efficacious in reducing hostile interpretation bias. To the best of our knowledge, amethodologically stringent randomized sham-controlled clinical trial in (adult) people with clinical levels of hostility is currently lacking. Besides its potential efficacy as a stand-alone intervention, recent evidence suggests that CBM-I may perhaps additionally augment the efficacy of existing therapy protocols. Two studies showed that offering CBM-I prior to (computerized) cognitive behavior therapy enhanced training effects on anxiety symptoms compared to sham-training prior to cognitive behavior therapy (Beard et al., 2019; Butler et al., 2015). The idea of offering CBM-I prior to existing therapy protocols may be even more attractive for populations with increased levels of hostility. That is, patients with increased hostility levels are at an increased risk to engage in hostile interactions with therapists (von der Lippe et al., 2008), which negatively impacts working alliance (Gülüm et al., 2018). One study showed that working alliance positively mediated the relationship between low levels of hostility and treatment outcome in dialectical behavior therapy for borderline personality disorder (Hirsh et al., 2012). Offering CBM-I prior to psychotherapy may therefore have beneficial effects on psychotherapy on top of its general effects, through increased working alliance levels. Taken together, preliminary evidence suggests that CBM-I potentially reduces HIB, anger, self-reported aggression, and working alliance. In the present work we developed an eight- session CBM-I intervention and compared its effects to an active control condition in two trials. The first study (Study 1) served as a feasibility trial to establish whether CBM-I alters hostile and benign biases and state anger in the desired direction followed by a randomized sham-