Anger is a negative emotional state associated with high levels of neuroticism, and likely plays a significant role in the development and maintenance of psychological disorders (Cassiello-Robbins & Barlow, 2016). However, the presence and consequences of anger in psychopathology is rarely examined (Cassiello-Robbins & Barlow, 2016). Limited research suggests that comorbid anger is associated with worse treatment outcomes for the principal diagnoses of PTSD and depression (Fava et al., 1991; Forbes et al., 2008). Based on this preliminary evidence, anger appears to be an important and understudied emotion in the treatment of psychiatric problems. The current study investigates whether anger impacts the length of time it takes individuals receiving psychotherapy to reach clinically reliable change (CRC) on overall psychological symptomatology. We hypothesized that adults who enter therapy with elevated levels of anger will take longer to reach CRC compared to non-angry patients.
The study’s sample includes 422 adults receiving psychotherapy at a community-based mental health training clinic. Participants included 129 males (30.6%) and 254 women (60.2%); 39 (9.2%) did not report gender. The age of sample ranged from 18 to 73 years (median = 29, mean = 33.2).
Anger was assessed via a subset of anger-items on the Outcomes Questionnaire-45.2 (OQ-45.2), a measure of mental health symptom occurrence (Beckstead et al., 2003). The anger subscale consists of four items that were dichotomized to reflect whether an anger symptom is persistent (occurring Frequently or Always; coded as 1) or not (occurring Never, Rarely, or Sometimes; coded as 0). Participants who endorsed at least one persistent anger symptom at baseline were classified as having elevated anger.
Overall symptomology was measured using the OQ-45.2 total score, minus the anger items. The OQ-45.2 was scored using a binary method in which a symptom was considered “present” and coded as 1 if the symptom occurred at least Sometimes (Gelin, & Zumbo, 2003). Participants completed the OQ-45.2 bi-weekly throughout treatment.
Clinically reliable change (CRC) in overall symptomatology was defined as a 1-standard-deviation (or more) decrease in total OQ-45.2 score from baseline. The primary outcome was reliable-change survival, defined as the number of weeks between the baseline assessment and first occurrence of CRC on the OQ-45.2.
Analyses were conducted in SPSS. To estimate reliable-change survival functions for patients with and without elevated anger, the cox regression method was used with anger entered as a covariate. Cox regression analyses also estimated the hazard ratio associated with anger. Data for participants who did not achieve CRC were censored at the time of the survival analysis.
Overall, 33.2% of the total sample achieved CRC on the OQ-45.2 over the course of treatment. At baseline, 25.1% of the sample endorsed elevated anger. Cox regression analyses revealed that patients with elevated anger took longer (more weeks in therapy) to achieve CRC than non-angry patients. Patients with elevated anger were two times less likely to achieve CRC that non-angry individuals (odds ratio = .478, p = .006, 95% CI = .28 - .81). Clinical implications and future directions will be discussed.
