Modern Psychological Studies

Volume 12 Number 1 Article 5

2006

Effectiveness of treatment interventions for adults suffering from anger disorders: a literature review

Amy Levitt Muhlenberg College

Mark J. Sciutto Muhlenberg College

Follow this and additional works at: https://scholar.utc.edu/mps

Part of the Psychology Commons

Recommended Citation Levitt, Amy and Sciutto, Mark J. (2006) "Effectiveness of treatment interventions for adults suffering from anger disorders: a literature review," Modern Psychological Studies: Vol. 12 : No. 1 , Article 5. Available at: https://scholar.utc.edu/mps/vol12/iss1/5

This articles is brought to you for free and open access by the Journals, Magazines, and Newsletters at UTC Scholar. It has been accepted for inclusion in Modern Psychological Studies by an authorized editor of UTC Scholar. For more information, please contact [email protected]. Effectiveness of Treatment In- terventions for Adults Suffer- Amy Levitt ing from Anger Disorders: A Advisor: Professor Mark J. Sciutto, Ph.D. Literature Review

Muhlenberg College

Anger is a Common Problem medical problems that result from suppressed Anger plays a significant role in everyday anger are artery disease, cancer and hypertension life. Defined, it is a feeling "produced by active (Greer & Morris, 1975). processes, including remembrances of the past, expectations of the future, awareness of current Anger Should be More Frequently Addressed behavioral and physiological responses and in Treatment Outcome Literature comparison of actual to desired behaviors and Even though anger has been shown to be judgments of what others are thinking" detrimental to oneself and others, various treat- (Kassinove & Sukhodolsky, 1995, p. 177). ment outcome research expressing methods to Referred to as a harmful, phenomenological or decrease its prevalence have been extremely internal feeling state, anger encompasses percep- scarce. Therapists find it particularly odd that tual and cognitive distortions such as anger is a neglected topic in treatment literature misattributions of blame, feelings of injustice, since anger is a frequent subject in psycho- subjective labeling, negative physiological therapy. Evidence for this lack of support for alterations and tendencies to engage in socially treatment outcome research is that the primary destructive behavior. The totality of cognitive interest in the past twenty-five years based on and feeling states differentiates anger from other keywords in PsycINFO has been on anxiety and similar feelings such as sadness and anxiety. depression (Chambless & Hollon, 1998). The Sometimes anger is brief and fairly intense. neglect of anger in past literature renders little However, at other times, it is unrelenting, severe empirical help to clinical practitioners. and highly disruptive in one's life as well the As the detrimental effects of anger have been lives of loved ones, co-workers and others in known for a long time, it is imperative that a one's social network. Overt anger, also known review of treatment outcome research be per- as aggression, can lead to a negative self-con- formed to assess treatments — pharmacological cept, pessimistic judgments by others, feelings of and psychological — with the greatest efficacy. low self-worth, inability to function in the To prove efficacy, researchers should establish workplace, physical and verbal injury to self or empirically supported treatments (ESTs), i.e. others, destruction of property and family con- treatments shown to be valuable in controlled flict (Deffenbacher & Stark, 1992). Anger is research with a clearly defined population also problematic when suppressed. Examples of (Chambless & Hollon, 1998). Controlled re- 30 search allows the experimenter to infer that the dependent on his or her interpretations, thoughts effects of an experiment are due to the treatment and self-statements of real-life experiences, and not to chance or confounding factors such as desire to help patients recognize and dispute the incidence of conditions with very similar their irrational and imprecise thoughts to con- participants in terms of attitudes or physical struct more adaptive cognitions (Beck, 1976). health, or passage of time (Campbell & Stanley, This type of treatment for anger has produced an 1963). average effect size of .93, which is a large Efficacy is also best demonstrated in random- treatment effect. ized clinical trials (RCTs) — group designs in Another form of treatment that has been which individuals are randomly assigned to a proven effective in the healing of anger is Anger treatment (Chambless & Hollon, 1998). Also Management Training (Richardson & Suinn, critical to prove efficacy is replication by two or 1972). The client is asked to re-construct past more independent researchers. Replication helps experiences that have provoked angry feelings, protect from drawing incorrect assumptions of such as an argument with a significant other or the efficacy of a treatment based on one unusual co-worker. The therapist helps the patient finding. arrange these experiences in a hierarchical To identify a population for research, many fashion, with increasing levels of severity. have used the Diagnostic and Statistical Manual Concurrently, the therapist teaches the client of Mental Disorders (DSM-IV-TR) because the ways to relax him or herself. Once capable to diagnostic labels in the manual are standardized, display these relaxation techniques, the client is allowing reliable diagnoses to be made that are instructed to re-enact the least anger-provoking familiar to both clinicians and researchers. experience. Repetition ensues, with the patient Another way to identify a population for re- honing skills in associating the anger-provoking search is through cutoff scores on reliable and experience with relaxation, both in therapy and valid questionnaires or interviews identifying the at home. Relaxation is successively paired with problem of interest. However, the question arises increasing levels of anger-provoking contexts. that some internal experiences of the human The authors illustrated that Anger-management being that a researcher may desire to investigate training also has shown to be very effective, with such as anger have not been classified in the an average effect size of 1.01. DSM-IV-TR. Given the potential severity of anger-related problems, it is surprising that anger The Present Study: Comparison of Effect has received such little attention in the psycho- Sizes for Recent Treatment Outcome Re- logical literature. It is therefore imperative to search on Anger, 1995 - 2005 compare and contrast treatment intervention Despite the previous lack of treatment out- outcomes for this common physiological, behav- come studies on anger, there appears to be an ioral, cognitive and socially reinforced experi- increase in the quality of the literature and the ence, with the goal of not only finding the most number of studies within the past ten years. effective treatment programs when anger is Effect sizes for these studies and others were excessive and disruptive but possibly a separate computed to assess which measures are most categorization in the DSM-IV-TR manual. effective in the assessment of anger (self-report, physiological and behavioral), which popula- What We Know about Treatment Strategies tions were most likely to comply with treatment for Anger (voluntary versus mandated/offender), which We do have some information regarding the studies included a control/comparison group, the effectiveness of various treatments for anger. most frequently indicated treatment (Cognitive- Cognitive therapies, based on the hypothesis that Behavioral, Anger Management Training and the emotional functioning of an individual is pharmacotherapy), the range of follow-up period 31 (smallest, largest, and median duration), the sixty-one individuals participated in Anger average treatment period (smallest, largest, and Management Training, with an average sample median duration), average sample sizes and size of 140.3 per study. Other treatments used dropout, effect sizes for various measures of were Dialectical Behavior Therapy, in which the anger (anger-in, anger-out, anger control, aggres- therapist discusses recent unhealthy or danger- sion/violence) and comparison of effect sizes of ous behaviors with the client in a hierarchical treatments solely targeting anger versus anger/ manner, from its precursors to cognitive and aggression. Therefore, the purpose of this litera- behavioral barriers that prevented the client from ture review was to assess more recent research to choosing more positive behaviors, to exploring outline factors that might affect treatment out- various alternative options that the client can use come for anger, with an ultimate goal of deter- in the future. This therapy had a sample size of mining which treatments are most efficacious for eight participants. The two pharmacotherapies the treatment of anger. used were Citalopram and Topiramate. Citalopram, a selective serotonin reuptake Method inhibitor (S SRI) that is taken primarily for depression, contained forty-five participants. Materials Topiramate, an anticonvulsant, included twenty- Effect sizes, which measure the magnitude of nine participants. Both pharmacotherapies a treatment effect, were computed using a pro- significantly increased anger control. gram called POWPAL and ZUMASTAT The longest duration of a study was eleven (Gorman, Primavera, & Allison, 1995; ZumaStat months, while the shortest time period was three Statistical Programs [ZSP]). days. The average duration of a study was 14.9 weeks. The longest follow-up period was three Procedure years, seven months. This study treated anger Effect sizes were computed as the standard- with Cognitive-Behavior Therapy. Studies ized mean difference statistic or Cohen's d, containing the shortest follow-up periods were where a positive effect size signified increase of mostly cognitive behavioral in nature, with an the measure and a negative effect size character- average follow-up period of eight weeks from ized a decrease of symptoms. Tests used to the end of treatment. obtain effect sizes were t-tests, analysis of Of the studies which assessed both anger and variance and correlation coefficients. aggression, the average effect size was d = -0.84. Studies which assessed only anger had a nega- Results tive effect size of d = -1.05. Nine studies con- tained voluntary populations, while fourteen Twenty-two studies were found addressing were comprised of mandated/offender/forensic the efficacy of recent treatment interventions for populations. Thirteen studies used a control adults suffering from anger disorders. Results group, mostly in the form of a waiting-list can be found in Table 1. Out of three scales used control. The treatment that most frequently used to document outcome — physiological, behav- a control group was Cognitive-Behavior ioral and self-report — twenty studies used self- Therapy, which consisted of mostly mandated, report, one study used a physiological measure offender populations. and fifteen studies used behavioral measures. Nine studies did not use a control group. Of The largest sample size was 419 and the smallest these, Cognitive-Behavior therapy was most was five. The average sample size was 71.3. frequently used. Out of the Cognitive-Behavior Nine-hundred-twenty-six individuals partook in Therapy studies that did not use a control group, the Cognitive-Behavior therapy, with an average the population most frequently assessed were size of 61.7 individuals per study. Five-hundred- involuntary. Many of these mandated samples 32 were individually recommended or referred to Management Training across population was d = treatment through court order, as part of an -1.04. These findings indicate that Anger Man- inpatient treatment program in a psychiatric agement Training is slightly more effective for hospital, and through recommendations from decreasing anger in adults. counselors, allied health providers and case- workers. Discussion Average dropout rate across treatment was 14.2 individuals, a rather large dropout given the After critically analyzing existing contempo- majority of studies had a small sample size. rary research on treatment interventions for Cognitive-Behavior therapy had an average adults with anger disorders, it is evident that dropout rate of 10.4 individuals. Average drop- there is not a significantly large disparity in out rate for Anger Management Training was 41 efficacy of Anger Management Training and participants. Cognitive-Behavior Therapy. Given that the For offender populations, the most common sample sizes are widely varied and that there are treatment was Cognitive-Behavior therapy. In unequal numbers of studies for each treatment, this population, the most effective measurement one cannot justifiably compare the two. There for anger was through the Anger Management were almost twice as many individuals in Anger Questionnaire. The smallest effect size for direct Management Training than Cognitive-Behavior measurements of anger in an offender population Therapy. However, results do indicate that an was Anger-In. The average effect size for Anger adult with an anger disorder in treatment does Control was d = .74. The average effect size for differ significantly from the controls. In other the Novaco Anger Scale was d = -1.62. The words, treatment does seem to be a main facilita- average effect size for depression was d = -0.79. tor of positive change, since individuals in the For voluntary populations, the most com- treatment conditions have larger effect sizes for monly cited treatment was Cognitive-Behavior decreases in anger than in the control groups. therapy. In this population, the most effective The average effect sizes for state and trait anger measurement of anger was state-anger. The are large for voluntary populations using Cogni- lowest effect size for direct measurements of tive-Behavior Therapy, most likely because they anger in a voluntary population was STAXI-Trait have a sincere motivation to develop healthy Anger. The average effect size for State-Anger cognitions and peaceful behaviors towards was d = -1.99. The average effect size for Anger themselves and others in their immediate envi- Control was d = 1.14. The average effect size ronment. for Trait-Anger was d = -1.54. The average Perhaps Cognitive-Behavior Therapy was effect size for Anger-Out was d = -0.94. most frequently used with offender populations Anger Management Training and Cognitive- because distorted cognition is one of the most Behavior Therapy both addressed anger and prominent traits of chronic offenders (Beck, aggression with equal frequency. The average 1999). Likewise, individuals may want to effect size for Anger Management Training that volunteer for this treatment due to its high addressed both anger and aggression was effectiveness. Anger control appeared to in- d = -0.42. The average effect size for Cognitive- crease more as an effect of Cognitive-Behavior Behavior Therapy that focused on anger and therapy than Anger Management Training. This aggression was d = -0.23. Therefore, Anger may be because this treatment is structured, Management Training is most effective for focused and active. The client takes an active treating both anger and aggression. role in not only realizing that one's cognitions Overall, the average effect size for Cognitive- are the problem and not the event itself which Behavior Therapy across population was facilitates the client's strategizing of ways to d = -0.70. The average effect size for Anger restructure his or her cognitions to view ob 33 stacles in life in a healthier, constructive, peace- anger. Only twenty-two studies were identified ful manner. As compared to Anger Management in which anger was the target of treatment. Training, which teaches control by learning However, one positive sign of current research is relaxation techniques, Cognitive Behavior that the populations are more relevant to the Therapy engages in a concept known as disputa- disorder. In other words, current studies have tion, where the client essentially attempts to targeted mainly offender populations, individuals defend his or her irrational beliefs and eventually who have been sexual abusers and excessive realizes for him or herself that the beliefs are, in alcohol users, instead of fairly harmless college fact, irrational. Therefore, once the client is students who volunteered to participate in outside of the supportive therapeutic alliance, he treatment research. or she develops the mental strength and cogni- Future research should attempt to gather tive awareness to approach challenges in life larger sample sizes, less reliance on self-report with a healthier perspective, as opposed to measures, longer follow-up period as treatment simply being instructed problem-solving skills effects may dwindle as a function of exposure to and relaxation practices. However, perhaps the a less supportive environment than therapy short follow-up period with Cognitive-Behav- provides. Future research should also attempt to ioral was due to the misconception that the client facilitate the use of matched control groups has vastly improved in therapy and will continue (particularly with offender populations to detect to do so. Another possible reason is the cost- possible confounding variables due to history effectiveness of a short follow-up period for such effects) and training of therapist on ways to a long, costly treatment. boost morale and consequently decrease dropout Compared to literature from ten or more years rate. Given that the effect size is larger for the ago, the average effect sizes of the most effective treatment of both anger and aggression as op- treatments are extremely similar. This is quite posed to only anger, perhaps researchers should high and indicates that there are effective thera- focus their efforts on addressing both issues peutic approaches available for clinicians work- through less reliance on self-report and more ing with adult clients suffering from anger behavioral measures and coding systems. Even disorders. Although past research studies can with these revisions, the main component in help guide the treatment of anger, limitations attempting to decrease one's anger is a sincere exist in the literature. The most obvious is the motivation and desire to change for the better. small number of treatment outcome studies on

34 References Greer, S., & Morris, T. (1975). Psychological attributes of women who develop breast Beck, A. (1999). Prisoners of hate: The cogni- cancer: A controlled study. Journal of Psycho- tive basis of anger, hostility, and violence. somatic Research, 19, 147-153. New York, N.Y.: HarperCollins Publishers, Kassinove, H., & Sukhodolsky, D. G. (1995). Inc. Anger disorder: Basic science and practice Beck, A. T. (1976). Cognitive therapy and the issues. In H. Kassinove (Ed.), Anger disor- emotional disorders. New York: International ders: Definition, diagnosis, and treatment (pp Universities Press. 1-26). Washington, DC: Taylor & Francis. Campbell, D. T., & Stanley, J. C. (1963). Experi- Gorman, B. S., Primavera, L. H., & Allison, D. mental and quasi-experimental designs for B. (1995). POWPAL: A program for estimat- research on teaching. In N. L. Gage (Ed.), ing effect sizes, statistical power, and sample Handbook of research on teaching (pp. 171- sizes. Educational and Psychological Mea- 246). Chicago: Rand-McNally. surement, 55, 773-776. Chambless, D. L., & Hollon, S. D. (1998). Richardson, F. C. & Suinn, R. M. (1972). The Defining empirically supported therapies. mathematics anxiety rating scale: Psychomet- Journal of Counseling and Clinical Psychol- ric data. Journal of Counseling Psychology, ogy, 66(1), 7-18. 19, 551-554. Deffenbacher, J. L., & Stark, R. S. (1992). ZUMASTAT. (n.d.). Retrieved October 3, 2005, Relaxation and cognitive-relaxation treat- from http://www.zumastat.com/Home.htm ments of general anger. Journal of Counseling Psychology, 39(2), 158-167.

35 Table 1

Effectiveness of Treatments by Measure

Source Measure Effect Size (d)

Reilly & Shopshire (2000) Anger Control 0.59 Trait-Anger -0.81 Total Mood Disturbance -0.61 Violence -1.62

Taylor et al. (2002) Anger Intensity -1.92 WARS Anger Index N/A

Lindsay et al. (2004) Dundee Provocation Inventory -1.43 Anger Provocation Role-play -2.36 Self-report Diaries -2.35 Overall Effect (Provocation Invent) -1.67

Ireland (2004) Anger Management Assess. Questionnaire -1.28 Wing Behavior Checklist (WBC) -0.80 Bums et al. (2002) Novaco Anger Scale -2.24 Spielberger State-Trait Anger Expression -1.56 Modified Overt Aggression Scale -0.67

Nickel et al. (2004) State-Anger -3.64 Trait-Anger -2.81 Anger-in -0.59 Anger-out -2.80 Anger Control 2.91

Davidson et al. (2000) (Cognitive Behavioral Hostility Treatment) Self-reported hostility ("Barefoot's Ho") -0.54 CAB-V (anger verbal behavior) -0.51 DBP (Diastolic blood pressure) -0.74

Lanza et al. (2002) PPG — Psychodynamic group psychotherapy OAS N/A STAXI-Trait STAXI-State STAXI-Control

CBG — Cognitive Behavior Group OAS 0.03

36 STAXI-Trait -0.27 STAXI-State -0.34 STAXI-Control 0.32

Wilson et al. (2000) Relapse Prevention Treatment 0.88 Roy-Byrne et al. (2004) Brief Symptom Inventory -1.62 Peritraumatic Emotional Distress Inventory -1.22

Romano and De Luca (2005) Multidimensional Anger Inventory -0.87 Polaschek and Dixon (2001) Violent re-offending rate 1.31

Berry, S. (2003) Probability of Reconviction -3.88 Number of violent offences -0.74

Rodriguez et al. (2005) Impulsivity -2.98

Evershed et al. (2003) Dialectical Behavior Therapy (DBT) -1.34

Donohue and Van Hasselt (1999) Beck Depression Inventory -0.63 Abuse -1.31 Loneliness -1.14 Distress -1.13 Rigidity -0.01 Unhappiness -0.94 Communication 1.17 Willner et al. (2002) Anger -0.50 Provocation -0.44

Taylor et al. (2004) State Anger -0.75 Trait Anger -0.63 Anger-In -0.30 Anger-Out -0.52 Anger Control 0.89

Williamson et al. (2003) Anger-Out -0.41 Anger Control 0.20 Readiness for positive change 0.36

Mammen et al. (2004) Anger attacks -0.41

McMurran & Cusens (2003) STAXI Anger Control (in and out) N/A Impulsivity N/A Social Problem Solving N/A Alcohol-related Aggression N/A

37 Devilly (2002) Depression -0.94 Anxiety -0.52 Stress -0.97 Novaco Anger Inventory -0.50 Alcohol Intake -0.14

Note. Negative effect sizes indicate a decrease in that particular symptom. Positive effect sizes usually indicate a healthy increase, such as in control of anger.

38