Abstract
There is no clear evidence to guide mental health professionals in assessing and treating angry clients. Recent reviews have considered cognitive and behavioral approaches to the treatment of anger, but little is known about the potential effectiveness of other treatment modalities. A meta-analytic review was conducted to examine the effects of treating dimensions of anger by using various psychological treatments found in the scientific literature. The final analysis included 96 studies and 139 treatment effects. The nine types of psychological treatments included cognitive, cognitive behavior therapy, exposure, psychodynamic, psychoeducational, relaxation-based, skills-based, stress inoculation, and multicomponent. The overall weighted standardized mean difference across all treatments was 0.76 (95% confidence interval [CI], 0.67β0.85, Q = 403.13, df 138, p < .001, I2 = 65.76), which suggests that psychological treatments are generally effective in treating anger. The results also suggest a considerable degree of variability in the effect sizes of specific treatments for anger. The results show that at least some of the variability may be explained by the number of treatment sessions offered to participants, the use of manuals to guide delivery of the treatment, the use of fidelity checks, the setting of the research, and whether the study was published or unpublished. This review builds on previous evidence of the effectiveness of psychological treatments of maladaptive anger, and it provides the basis for developing evidence-based guidelines for specific populations with anger problems.
Little attention has been given to maladaptive anger over the years in comparison to other emotional disorders. There is no clear consensus regarding the best ways to define, assess, and treat various dimensions of anger,1 as it is often camouflaged2 within definitions of violence, hostility, and aggression. G. Stanley Hall3 first lamented more than a century ago that there was a lack of literature about anger and that there was βno comprehensive memoir on this very important and interesting subjectβ (Ref. 3, p 516). He envisioned that mental health professionals would move toward creating a more precise definition and understanding of anger to help clarify the ambiguity surrounding this phenomenon.
Anger is an internal state involving various degrees of, and interactions among, physiological, affective, cognitive, and verbal components.4 It is considered to be a common human experience expressed on a continuum of healthy-disturbed, adaptive-maladaptive, constructive-destructive, and pragmatic-problematic. It has been correlated with different forms of aggressive behaviors, including assault,5,6 violence,7β9 and property damage.10 It has also been connected with increased health risks, including coronary heart disease, heart attacks, high blood pressure, and high cholesterol.11
Despite the prevalence of and the problems created by maladaptive anger, anger disorders are not currently recognized by the Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition, Text Revision (DSM-IV-TR).12 Anger is sometimes viewed as a residual of other diagnosable mental health problems, such as borderline personality disorder, antisocial personality disorder, and conduct disorder.13,14 Others view anger as a cluster of symptoms that are distinguished by the severity, frequency, and nature of anger episodes and patterns.15 Some have identified several ways that people express anger, including direct expression, reciprocal communication, thinking before responding, taking a self time-out, physical assault of people, physical assault of objects, negative verbal expression, keeping it in, controlling it, taking corrective action, diffusion or distraction, passive-aggressive sabotage, and relationship victimization.16,17 The diversity of these expressions causes further confusion when attempting to isolate the core characteristics of anger and to distinguish it and disentangle it from other emotional and behavioral disorders.
Anger is generally conceptualized as a multifaceted construct.18 Different measures have been developed to distinguish its various dimensions,4 which generally include levels of anger, anger control (ability to control angry feelings), person-specific anger (feelings of anger toward another), anger expression (linked to aggressive violence), anger state (distinct episodes of anger), anger trait (frequency of angry episodes), and angry driving. Anger is often defined psychometrically (e.g., by cutoff scores) rather than by a theoretical model.19 Therefore, most treatments have been developed to address elevated levels of maladaptive anger.
Although there is no clear conceptual framework for distinguishing functional and dysfunctional anger, treatments for anger continue to be utilized across many mental health disciplines.14 Since it is both common and debilitating,19 it is incumbent on psychiatrists and other mental health professionals to be informed of the current evidence regarding the potential efficacy of various treatment modalities.
Glancy and Saini1 completed a comprehensive review of the literature regarding the psychological treatments of anger and aggression, including five meta-analytic reviews,15,19,20β22 and they concluded that there is no consensus in the literature regarding the best ways to treat and reduce anger and aggression. Most reviews focused on treatments containing components of cognitive and behavioral therapies or a combination of the two. Other modalities, such as psychodynamic and psychoeducational have not been included in previous reviews. Although there is good evidence that treating anger generally works,23 more information is needed to guide forensic psychiatrists and other mental health professionals in making evidence-based decisions when choosing from the various treatment modalities. The purpose of this meta-analytic review was to complete a systematic and exhaustive search of all relevant studies, to include a greater variety of psychological approaches to the treatment of anger.
This meta-analysis supports the potential integration of psychological and pharmacological treatments of maladaptive anger. Glancy and Knott24β26 completed a three-part series on the use of pharmacology to treat anger and aggression. Based on their analysis, they introduced an algorithm to provide clinicians with an evidence-based model for treating anger and aggression with medication. They strongly suggest that pharmacologic agents are most effective when used with adjunctive psychosocial therapy. Very few studies have combined pharmacological and psychological interventions in treating anger. There also is no consensus among mental health professionals and researchers on the most effective ways to treat anger by using psychological interventions, and so the present meta-analysis is a good first step toward integrating psychological and pharmacological treatments.
The Role of the Forensic Psychiatrist
Within the current demand for evidence-based practice (EBP), there is an increased onus on forensic psychiatrists to become more sophisticated in the areas of risk assessment and management27 and more capable of distinguishing effective treatments from inadequate and harmful approaches. Accurate scientific evidence about the various effects on anger of different psychological treatments and the influences of moderator variables on effectiveness and relevance of treatments on various and specific populations have significant consequences for the forensic psychiatrist practicing within an EBP environment.
Evidence-based guidance on how best to treat and manage anger is important for forensic psychiatrists, especially because working with angry clients is as common as working with those who are anxious or depressed.28 Forensic psychiatrists should be informed about anger because their work often involves individuals with angry relational problems, families in high conflict, medical patients struggling with hypertension, offenders with histories of violence, and people with trauma and substance abuse histories.29 DiGiuseppe and Tafrate30 observed that mental health professionals are generally less comfortable working with angry clients than with those who are experiencing anxiety or depression. Anger directed at the mental health professional is considered to be a major stressful situation, second only to threats of suicide.31β33 This discomfort may be due to a lack of knowledge regarding how best to intervene to ensure the safety of both the client and the professional.30 Despite the difficulties of working with angry clients, there remains less guidance in the literature for working with clients who are experiencing episodes of anger than for treating those with depression and anxiety. Kassinove and Sukhodolsky,34 for example, found 10 articles on depression and 7 on anxiety for every 1 on anger. The lack of attention given to anger1,19 leaves substantial room for growth.
In expanding the current state of evidence regarding anger, it is imperative to move beyond simple or narrowly focused approaches4 and address it in a comprehensive way, to consider its various dimensions and the many treatments that have been used in an attempt to remedy its problematic forms. Psychiatrists working with angry clients should consider the range of options available. With sufficient specification about the characteristics of effective treatments, psychiatrists are in a better position to consider the overall fit between treatments and clients. Evidence is needed, therefore, to guide treatment selection so that connections can be made between client characteristics and factors in various settings.
Current Evidence on the Treatment of Anger and Aggression
As mentioned earlier, there have been five meta-analytic reviews of the treatment of anger.15,19,20β22 In the first, Tafrate22 explored the effects of cognitive relaxation-based, skills training, and multicomponent treatments in 17 studies, with effect sizes ranging from 1.82 to 1.16. Bowman-Edmondson and Cohen-Conger21 considered the results of the cognitive, cognitive-relaxation, social skills, and relaxation therapies reported in 18 studies and found effect sizes ranging from 0.64 to 0.80. Beck and Fernandez20 expanded their inclusion criteria to include unpublished doctoral dissertations and single-group designs, which resulted in the inclusion of 50 studies. They found that cognitive behavioral therapy had a 76 percent success rate in reducing anger scores. DiGiuseppe and Tafrate19 aggregated 230 effect sizes from 57 published and unpublished studies, and found an overall mean effect size of 0.71, with no significant main effect for the different treatment models. Del Vecchio and O'Leary15 narrowed their meta-analytic review to include studies that primarily addressed anger, thereby excluding studies that were predominantly about aggression or hostility. Based on their 27 studies, the mean weighted effect size ranged from 0.61 to 0.90.
The reviews demonstrate overall mean effect sizes ranging from 0.64 to 1.16, which is considered to be moderate to strong. Based on these results, there is evidence that the psychological treatment of anger is effective. The range of the overall effect sizes, however, suggests that other factors may influence the variability of effects among the reviews. These factors may include the retrieval process of included studies, the inclusion of diverse populations, different treatment modalities, and the effects of various outcomes for anger and the influence of moderator variables. Exploration of these differences can contribute to a more comprehensive understanding of both the assessment and treatment of anger (see also Glancy and Saini1).
Retrieval of Included Studies
Other meta-analytic reviews have not provided sufficient details about their information retrieval strategies. Inclusion of information regarding search terms, databases searched, term limiters and expanders used, and other methods of retrieving potential studies provides a transparent process so that the information retrieval strategies can be scrutinized based on the sensitivity and specificity of the searches. A comprehensive search for all potential studies is important, given that meta-analyses with fewer than 50 studies tend to report higher effect sizes.35 In addition, Bowman-Edmondson and Cohen-Conger,21 and Tafrate22 did not include unpublished studies. It is important to include unpublished studies in a meta-analysis, because failing to do so can introduce a bias toward favorable outcomes,36 as other reviews suggest that unpublished studies are likely to have lower overall effects.37,38
Separating the Influence of Diverse Populations
Some reviews included children, adolescents, and adults in the same analysis, and thus failed to consider age differences and the role of development in treating problems related to anger.19 Combining children and adults in the same review can camouflage potential differences by treating these different populations the same. For example, by pooling the effects in their review, Beck and Fernandez20 risked the assumption that the effects are similar for different populations including abusive parents, violent and resistant juvenile offenders, inmates in detention facilities, and aggressive school children. Until further testing is completed, in considering the potential variability between adults and children, it is best to complete separate analyses and to treat these groups as if they were different.
Within adult populations, the meta-analytic reviews provide evidence that treating anger is effective across diverse groups including persistently violent male prisoners, adults with intellectual and learning disabilities, forensic patients, angry parents, female batterers, mental health patients, undergraduate students, incarcerated male juveniles, male batterers, aggressive drivers, faculty members, Vietnam War combat veterans, and patients with schizophrenia. These results must be viewed with caution, given that these broad groups are not homogenous and their influence on the variability of effect sizes remains unknown. Further analysis is particularly needed in this area, given that there is an overrepresentation of undergraduate student volunteers, thereby limiting the generalizability of the findings to other populations.
Toward a Comprehensive List of Treatment Modalities
A range of treatments has been considered, but most interventions have been based on cognitive, behavioral, or cognitive-behavioral models, and there may not be sufficient variability in the treatment approaches to produce noticeable differences.1 Other treatment modalities, such as psychodynamic and psychoeducational have not been included in previous reviews, despite some preliminary evidence suggesting the potential benefits of these treatments in reducing anger in various populations.39β42
Anger as the Primary Outcome
Because anger is often disguised by other negative behaviors, such as aggression, hostility, and violence, it is important to explore it separately and as an independent primary outcome. With the exception of Del Vecchio and O'Leary,15 reviews have mixed outcomes that include both anger and aggression, and little attention has been directed toward distinguishing these to explore possible differences. As they point out, the anger construct is considered distinct from the concepts of hostility, aggression, and violence and therefore merits separate analysis.
Considering Differences in Effect Sizes for Various Anger Outcomes
Bowman-Edmondson and Cohen-Conger21 provided the first meta-analysis to demonstrate that differences in effect sizes could be attributed to the use of specific treatment modalities for specific dimensions of anger (e.g., anger control, anger expression). They found that changing the expression of anger was best achieved by relaxation treatment (1.19). Whereas relaxation treatment had the largest effect size (0.79) for self-reported anger, behavior and social skills training had the largest effect size (1.13) for the observation and assessment of angry behavior. To change physiological anger, relaxation-based therapies had the largest effect size (1.21) compared with cognitive-relaxation (0.76), cognitive (0.57), and social skills training (0.58). Although the design of the study precluded statistical analysis of these findings based on a small number of included studies, it was an important step toward a more robust analysis of the potential moderator variables that may influence the range of effect sizes. DiGiuseppe and Tafrate19 and Del Vecchio and O'Leary15 also clustered effect sizes according to the type of outcome measure used for each intervention; however, judgments made regarding the similarity and dissimilarity between effect sizes were not determined statistically, because in many cases, the effect sizes were derived from fewer than five studies, which would result in an inaccurate statistical comparison.
The Influence of Moderator Variables
Not all reviews considered the potential influence of moderator variables, and most had too few studies for meaningful statistical analysis. The moderator analysis by Del Vecchio and O'Leary15 revealed an overall positive relationship between session length and treatment outcomes. DiGiuseppe and Tafrate19 found significant positive relationships between the use of manuals and fidelity checks and treatment outcomes. They also found positive effects for individual treatment formats, publication status, and type of participant. The number of sessions, participantsβ sex, and allocation to groups all failed to predict the effect size of treatment outcomes. Further research is needed to explore the influences of moderator variables that are specific to anger treatment outcomes.
Rationale for the Current Review
Despite preliminary evidence supporting psychological treatments of anger, there is still no clear consensus among mental health professionals and researchers on the best way to treat angry clients. Kobayashi and Norcross state, βWithout a consensus on the identified phenomenon, we will continue to disagree on the proper psychotherapy of anger disordersβ (Ref. 33, p 277). To develop evidence-based guidelines, it is important to consider both the absolute effects of treating anger and the relative effects of each treatment type. Moderator variables can provide additional information for mental health professionals when setting up and conducting treatments to reduce anger. Factors such as setting, location, type of participants, characteristics of participants, number of sessions, and use of manuals can augment the applicability and relevance of treatment, thereby increasing the likelihood that the treatment will be effective.
Methods
A meta-analysis is a quantitative procedure for evaluating treatment effectiveness by the calculation of effect sizes derived from individual studies for the purpose of integrating the findings.43,44 The larger the effect size, the stronger the effect or the greater the degree of effectiveness of the treatment.45
Search Strategy for Study Selection
The information retrieval strategy included a search of 12 electronic databases (Cochrane Central Register of Controlled Trials, MEDLINE, PsychInfo, EMBASE, DARE, ASSIA, ERIC, 95% CINAHL, IBSS, Social Work Abstracts, Social Sciences Abstracts, and Social Service Abstracts), references of previous reviews, and unpublished studies (dissertations and gray literature).
The search terms used in OVID (and modified according to the electronic database) included: (anger control or anger or angry or aggression or hostility or anger-related disorders or aggressive driving behavior) and (anger management or treatment or intervention or counseling or cognitive behavior therapy or psychotherapeutic techniques or psychological psychotherap* or psychodynamic or group psychotherap* or exposure therapy* or eye movement desensitization therapy or relaxation therapy or therapeutic group or mental health services or treatment effectiveness evaluation) not (child* or youth).
Criteria for Study Selection
Studies were eligible for inclusion in the review if the population included men and/or women over the age of 18 years. Studies that included children and youth were excluded from the review, to avoid additional heterogeneity between studies. Because the main objective of this review was to compare psychological treatments of anger, pharmacological interventions were excluded. (For a comprehensive evidence-based review of published literature on the psychopharmacology of anger and aggression, see the three-part review by Glancy and Knott,24β26 mentioned earlier in the article). Outcome measures included the different constructs of anger (general, control, person specific, expression, state, and trait) and angry driving (see Table 2). Two reviewers independently categorized each outcome into a predetermined list; all disagreements were resolved by consensus.
Studies were eligible if they included experimental, two-group quasi-experimental designs or within-group designs with at least pre- and post-treatment scores. The comparison group could include no treatment, minimal treatment, other treatment, or a wait list comparison. Methods of constructing the comparison groups varied in the use of statistical controls to reduce the threat of selection bias, and these differences were coded to explore their influences on results during data synthesis. Included studies had to report an effect size for the intervention or contain sufficient data to calculate an effect size.
Data Analysis
Comprehensive Meta-Analysis software, version 246 was used for all statistical analyses. This software can produce Cohen's d, Hedge's g, Q values, confidence intervals, fixed effects, random effects, and heterogeneity testing results. An examination of the internal and external validity of this meta-analysis was accomplished using multiple linear regression of the continuous and dummy categorical moderator variables in SPSS.
The standardized mean difference effect size statistic47,48 was used to record intervention effects. The standardized mean difference, Cohen's d, the most widely known effect size formulation for meta-analytic purposes, was used to measure effect sizes in cases in which both the means and standard deviations were reported.48 Cohen's d reflects the differences between the post-treatment means of the treatment group and the control group, divided by the pooled standard deviation, adjusted for sample size or, in the case of a study that did not use a control group, d reflects the difference between the pre- and post-treatment scores, divided by a pooled standard deviation. Thus, d represents differences in means expressed in standard deviation units. According to Cohen,49 effect sizes of 0.20, 0.50, and 0.80 refer to, respectively, small, moderate, and large effects. When possible, d was calculated directly from the means and standard deviations, because it is the most precise method. If this method was not possible, d was calculated from F- or t values.
A random-effects model was used for pooling results if significant statistical heterogeneity was present and if there were substantial between-study variations. Statistical heterogeneity in the outcome measures were assessed using the Q statistic and the associated p-value for each analysis and the I2 statistic.50 A significant Q statistic suggests heterogeneity within a set of studies and the need for moderator analyses. The I2 statistic determines the percentage of variability that is due to heterogeneity, where a value greater than 50 percent suggests moderate heterogeneity.
Typically, studies reported results on multiple outcome constructs (e.g., general anger, anger expressions, and anger control). All effect sizes that assessed anger that could be extracted from a study were coded but were analyzed using the overall mean of anger scores or sequential placement of outcomes to ensure statistical independence.48 Within each data file, extreme values were tested and corrected as recommended by Lipsey and Wilson.48 Correcting for extreme values in quantitative reviews is consistent with the purpose of meta-analyses, specifically to βarrive at a reasonable summary of the quantitative findings of a body of research studiesβ (Ref. 48, p 107).
Results
Studies Included in the Review
Based on the information retrieval strategy, 4,438 titles were retrieved. Of these, 879 were excluded because they were duplicates, 550 were excluded because they focused on children or youth, and 898 were excluded because they focused on pharmacology. The remaining 2,111 titles were screened by two independent raters. Cohen's ΞΊ formula was used to calculate inter-rater reliability between the two raters during the initial screen. Based on the initial screen of 2,111 titles, the raters included 232 titles with a Cohen's ΞΊ of 0.81. Based on the second screen of full text articles, 96 studies passed to the third phase for data extraction and inclusion in the meta-analytic review.
Characteristics and Frequencies of the Selected Studies
A breakdown of the characteristics of the studies selected for the meta-analysis is presented in Table 1. These characteristics have the potential to influence the effect sizes obtained from the studies.
Most studies were completed in the United States, involved fewer than 50 participants, and did not distinguish between males and females in their analyses. Although one-third of the studies included college or university settings, there was good representation from community treatment programs, correctional facilities, and general hospital settings. Participants included students in college or university, health care patients, incarcerated offenders, and mental health clients. Most treatment modalities included a group format, and many administered treatment with the help of manuals and fidelity checks. Although 76 percent of studies used a random assignment to the treatment and comparison group, there were differences in the types of comparison groups used. Many comparison groups received alternative, other, or minimal treatment during the period of the studies. Fewer received no treatment (including 26% who did not receive treatment because they were assigned to a wait list).
Mean Effects of Anger Outcomes
The 96 eligible studies generated 139 standardized mean difference effect sizes for the treatment of anger. Most studies generated more than one effect size for the anger outcome including anger control, anger expression, anger situation, anger symptom, angry driving, general anger, state anger, and trait anger. Table 2 shows the anger outcomes and examples of instruments used to measure them.
To create a set of independent effect sizes for analysis, a combination of procedures was used. First, each anger outcome was treated and analyzed separately. When studies reported results on different types of anger, an overall mean of anger scores for each treatment effect was calculated. In addition, analysis was performed using the sequential placement of outcomes so that one outcome from each study was used in the analysis. A crude analysis is reported in Table 3, but should be considered with caution as the overall effect is not independent.
The overall effect size across all types of dependent variables was 0.76 (CI = 0.67β0.85) indicating that participants in the treatment groups had significantly lower anger scores than the comparison groups after participating in the various treatment models. Analysis of the homogeneity of variance of effect size values was significant (Q = 403.1, df 138, p < .001), suggesting the presence of heterogeneity and the need to consider moderator variables. The I2 statistic of 65.75 suggests more than moderate heterogeneity. Significant homogeneity of variance of effect sizes was also noted for specific outcomes of anger including general anger, anger situation, anger symptom, state anger, and trait anger.
Type of Treatment for Anger at Post-test
Table 4 reports the effect sizes by treatment type and the homogeneity of variance of effect sizes. A Kruskal-Wallis H-test for one-way analysis of variance (H = 18.00, p < .05) indicates that the values of effect sizes are different across treatment groups. To inspect differences between two groups, Mann-Whitney U-tests were used and revealed significantly larger effect sizes for multicomponent compared with cognitive behavioral therapy (U = 149.00, p < .05), psychoeducational (U = 5.00, p < .05), relaxation-based (U = 167.50, p < .05), and stress inoculation (U = 33.00, p < .05). Larger effect sizes were found for cognitive compared with cognitive behavioral (U = 66.00, p < .05) and psychoeducational (U = 66.00, p < .05) therapies, but psychodynamic therapy had larger effects than did cognitive (U = 4.50, p < .05). No other differences between groups were significant.
Binomial Effect Size
Based on the recommendations of Rosenthal56 and Cohen,47 the binomial effect size display (BESD) was computed to compare those who participated in treatment with those who were included in the comparison groups. BESD is reported in Table 5.
Anger Outcome Within Each Treatment Type
The effect sizes for anger control were moderate across all treatment groups, ranging from 0.26 to 0.83. Anger expression (e.g., anger outward) had relatively small effect sizes for all treatments, ranging between 0.18 and 0.61. For anger situation (e.g., situations that may provoke anger), exposure therapy had significantly larger effect sizes with an average of 1.09 and 95% CI between 0.71 and 1.47. For those with a tendency to suppress anger in (e.g., anger inwards), moderately strong effect sizes were found for multicomponent therapies (0.74), and moderate effects for cognitive therapies and cognitive behavioral therapies. Cognitive and multicomponent therapies also produced large overall effect sizes for anger symptoms outcomes (0.94). Relaxation-based and stress inoculation therapies were the only two treatments found that explored angry driving. Both of these treatments produced large effect sizes, ranging from 1.27 to 1.89. For general anger, only psychoeducational therapy produced below moderate effect sizes, as all other treatments produced effect sizes above 0.76. Psychodynamic therapy had the largest effect size for general anger based on only two studies (1.43). For state anger (e.g., how angry the respondent is feeling), relaxation-based (0.97), and cognitive (0.83) both had large effects. Finally, for trait anger (e.g., anger response patterns), multicomponent therapies (1.046) had the largest effect sizes based on six studies, followed by cognitive (0.98) and relaxation-based (0.65) therapies.
Analysis of Moderator Effects
Analysis of the homogeneity of variance of effect sizes was significant (Q = 403.1, df 138, p < .001), suggesting the presence of heterogeneity and the need to consider moderator variables. The I2 statistic of 65.75 suggested above-moderate heterogeneity. Analysis of moderator effects was explored to consider the variability in the effect sizes across the different studies. Results show that published studies (0.837, 95% CI = 0.78β0.89) had larger effect sizes than unpublished studies (0.54, 95% CI = 0.43β0.66) as indicated by the independent-samples t test (t = 2.05, p < .05). There were no differences between within- and between-group designs or other study design details, including the type of comparison, sample size, and location of the study.
The setting of the research had a significant influence on the overall effect sizes at post-test (F = 2.784, p < .05). Colleges and universities (0.86, 95% CI = 0.77β0.95) and community treatment programs (0.88, 95% CI = 0.80β0.95) had larger effects than did correctional facilities (0.058, 95% CI = 0.44β0.72), psychiatric facilities (0.50, 95% CI = 0.33β0.68), and general hospitals (0.65, 95% CI = 0.52β0.79). Manuals (0.812, 95% CI = 0.074β0.877) and fidelity checks (0.85, 95% CI = 0.78β0.92) both produced larger effect sizes than did treatments that did not involve the use of manuals (0.76, 95% CI = 0.69, 0.83) and fidelity checks (0.73, 95% CI = 0.67, 0.79), but these differences were just over the statistically significant cutoff of p = .05 (p = .059 and p = .090, respectively).
The mean number of treatment sessions was 8.5 (SD 3.72), with the number of sessions ranging from 3 to 40. To determine the relationship of the number of treatment sessions on the overall mean effect of anger at post-test, a meta-regression was completed. For the slope, the z-value was 2.189 (p < .05). Figure 1 depicts the line of the slope, which demonstrates that the number of sessions affects the overall magnitude of the effect sizes.
Publication Bias
The existence of publication bias was explored by a funnel plot.48 In addition, a fail-safe N for the average effect size was computed in CMA. This meta-analysis incorporated data from 139 treatment effects, which yielded a z-value of 29.197 and corresponding two-tailed p = .000. The fail-safe N is 30,709, meaning that an additional 30,709 null studies would have to be located and included for the combined two-tailed p to exceed .05.
Follow-up Studies
Fifty-nine treatment effects were calculated from follow-up data and are displayed in Figure 2. The most common follow-up period was between 4 and 8 weeks (n = 54). At 12 to 16 weeks, only seven follow-ups were completed. At the one-year mark, 10 treatment effects were recorded. The Q statistic indicated a nonsignificant test result for overall effect sizes of follow-up data and for each identified period. This finding suggests that the variability of the effects found at post-test were not maintained at follow-up. The change is most likely due to the decreased number of studies involved in the follow-up data compared with post-test data. The results show a slight decrease in the overall effect sizes between 4 and 16 weeks (0.59) with an increase at the one-year point similar to the overall effect size recorded at post-test (0.76). It is important to note that the studies included at the one-year follow-up were homogenous. All 10 studies included undergraduate students in either cognitive behavioral therapy or skills-based training. More research is needed before any assertion can be made about this interesting finding.
Conclusions
There is strong evidence that psychological treatment of anger is moderately successful at reducing anger problems across various dependent variables. The results of the meta-analysis support previous reviews. The accumulation of results from these reviews suggests a consistent message that psychological treatments generally work with various populations to redress maladaptive anger. The magnitude of the gains found in this study is comparable with those reported in other meta-analytic reviews completed in the past 20 years.15,19,20β22 An overall mean effect of 0.76 was found for the studies, which was robust enough to be unaffected by unpublished null results. The findings are similar to the mean effects of 0.71 found by DiGiuseppe and Tafrate19 and within the range of 0.61 to 0.90 found by Del Vecchio and O'Leary.15
The overall effect sizes were generally maintained at 4 to 8 weeks, 12 to 16 weeks. and one year. Similar to the results in the study by Bowman-Edmondson and Cohen-Conger,21 some effects on anger actually improved more at follow-up than at post-test. DiGiuseppe and Tafrate30 suggest that effects maintained over time may tend to incorporate multiple interventions into one protocol. The effects maintained in this review were based mostly on undergraduate students, and so caution must be used in generalizing beyond the limits of these studies. More research is needed to explore whether these effects are maintained over time in this population and to determine whether similar maintenance of effects could be achieved in other populations.
Although other outcomes related to anger have been shown to have positive gains after treatment,19,21 it was important for this review to clarify the anger outcome by separating dimensions of anger from other similar constructs, such as aggression. As our understanding of anger improves, it is important to carefully consider the various dimensions of anger and to consider symptom-and-treatment modality matching by targeting specific elements of anger without further complicating this matching process by including other constructs.
This review suggests that findings from meta-analyses are worth considering as a part of what works in the treatment of anger by offering comparative information on how well different interventions work. Unlike previous reviews, significant differences were found for different treatment modalities. The strong showing of multicomponent therapies is consistent with the results in Tafrate,22 who reported an effect size of 1.00 for multicomponent therapies, and with the focus of DiGiuseppe17 on a multi-theoretical and comprehensive package of anger treatments. In 2001, DiGiuseppe and Tafrate30 presented a comprehensive treatment model for working with a wide variety of clients with anger problems. There is evidence that suggests the merit of pursuing a comprehensive approach. Future work to coalesce treatments should follow systematic procedures so that relative influences of various treatment components can be identified, isolated, and assessed for the influence on the variability of effect sizes. It is imperative to determine the parts of therapeutic programming that work and to vet out factors that are found to be ineffective.
The results of the moderator variables on the use of manuals and fidelity checklists are consistent with those found by DiGiuseppe and Tafrate.19 In this study, manuals and fidelity checklists were used in more than half of all studies, and their use produced increased effect sizes. DiGiuseppe and Tafrate19 suggest that therapists who participate in studies may have higher compliance rates when using manuals than those not involved in research. The findings support a more broad use of both manuals and fidelity checks.
The evidence also supports eight sessions as an adequate amount of treatment to demonstrate positive results to reduce anger problems. As treatments increase in the number of sessions beyond the average mean of eight, so too does the attrition rate, creating a higher number of withdrawals from the study.1 Beck and Fernandez20 suggested that in planning treatment for anger, cost-effective strategies should be considered alongside outcome efficacy. The results of this review clearly show that treating angry clients beyond eight sessions has a limited influence on the overall effects of treatment.
The strength of this review is the extent of the literature that was included in the information retrieval process. The strategy included over 4,000 titles, screened down to 96 studies by following a transparent and comprehensive level of screening protocol. In addition, a specific effort was made to include unpublished studies. Despite this comprehensive approach to information retrieval, the majority of the studies were completed in the United States and one-third were completed by college or university students. Even so, the comprehensive search across a wide range of literature helped to contextualize the research, and a process emerged for in-depth analysis. Another strength of this review is the inclusion of additional treatment modalities to provide a broader perspective regarding the relative merits of various treatments of anger.
Interpreting the results of meta-analytic reviews and comparing the relative magnitude of effects for treatment is not without its limitations. Studies included in the review were identified based on the inclusion criteria of whether the study reported on a treatment modality for anger. The criteria did not include a discernable assessment based on psychiatric diagnosis or psychometric measures, so generalizing results to psychiatric populations and those diagnosed with anger-related problems should be made with caution. Second, the review used broad definitions of anger. As a result, there was considerable statistical heterogeneity in the results of outcome categories, which indicates that caution should be used in combining the effects and interpreting the findings. Finally, the differentiation of treatment categories had good inter-rater reliability based on the categories created for this review; however, no other tests were performed to ensure the validity of these categories. Although differences in treatment allocation would not influence the overall results, differences found among treatment classifications must be considered with caution. As the evidence continues to accumulate and improve, future reviews should include additional moderator variables to explore the quality of research designs, and more attention should be directed toward critically appraising the evidence beyond allocation, sample sizes, and types of comparisons.
Implications for Developing Evidence-Based Guidelines
Treatments of anger have been used around the world for many years. There is a growing body of evidence that treatments are effective at improving a variety of anger-related problems. These positive effects are found within diverse populations, settings, and locations. The effect of treatment of anger is relatively smaller than that of treatments for anxiety and depression.19 Mental health professionals conducting at least eight sessions of treatment, using manuals to guide treatment, and using fidelity checks to ensure consistency are likely to improve outcome scores. Treatment of anger should therefore be supported as a unique approach and should not remain in the shadow of treatments for aggression, hostility, and violence.
Overall, treating anger with psychological treatments is beneficial. Some prudence is required, as there is some variation of treatment effect according to sample setting. When implementing a treatment program, mental health professionals should routinely evaluate and critically monitor to ensure that the client is receiving the perceived benefits and to verify that the approach is consistent with the client's expectations, values, and judgments, to avoid imposing evidence-based guidelines rather than following the procedures for evidence-based practice. The meta-analysis described in this review adds weight to the growing body of evidence. However, the variation of the impact of treatments, combined with the differences among the treatments themselves, makes it difficult to offer specific guidelines for choosing specific treatments without further research and analysis.
This meta-analysis supports the potential integration of psychological treatments and pharmacological agents to treat maladaptive anger. Glancy and Knott26 found several pharmacological agents to be efficacious in managing maladaptive anger. Careful assessment of the patient to detect underlying major psychiatric conditions is the first necessity, as these conditions may impede positive gains from psychological treatment alone. In that case, psychiatrists would follow the algorithm presented by Glancy and Knott to address mental health status with appropriate pharmacological agents. Further research is needed to determine whether pharmacological algorithms can be complemented with the combination of psychological interventions according to mental health status. Although this meta-analysis included only psychological treatments for anger, future studies including pharmacological approaches, with and without psychological interventions, are needed for consideration in the devolvement of evidence-based guidelines for the treatment of anger.
Appendix
Studies Included in the Meta-analysis
Achmon J, Granek M, Golomb M, et al: Behavioural treatment of essential hypertension: a comparison between cognitive therapy and biofeedback of heart rate. Psychosom Med 51:152β64, 1989
Acton R, During S: Preliminary results of aggression management training for aggressive parents. J Interpers Viol 7:410β17, 1992
Beck R, Fernandez E: Cognitive-behavioural self-regulation of the frequency, duration, and intensity of anger. J Psychopathol Behav Assess 20:217β29, 1998
Bennett P, Wallace L, Carroll D, et al: Treating type A behaviours and mild hypertension in middle-aged men. J Psychosom Res 35:209β23, 1991
Bolanos CR: An evaluation of a social skills training curriculum for anger management in a chronic hospitalized population. Dissertation Abstracts International, Section B: The Sciences and Engineering 59:223, 1998. Available by subscription at http://www.proquest.com
Boyle SW: A study comparing two methods of intervention to control high levels of general anger. Dissertation Abstracts International 52:2068, 1991. Available by subscription at http://www.proquest.com
Bradbury KE, Clarke I: Cognitive behavioural therapy for anger management: effectiveness in adult mental health services. Behav Cognit Psychother 35:201β8, 2007
Briscoe YB: A cognitive behavioral anger management intervention for women with histories of substance abuse. Dissertation Abstracts International, Section B: The Sciences and Engineering 62(11-B):5358, 2002
Cain JA: Anger control: a comparison of cognitive-behavioural and relaxation training, with post-treatment follow-up. Doctoral dissertation, United States International University San Diego, CA. Dissertation Abstracts International 48:1804β5, 1987. Available by subscription at http://www.proquest.com
Cary M, Dua J: Cognitive-behavioural and systematic desensitization procedures in reducing stress and anger in caregivers for the disabled. Int J Stress Manag 6:75β87, 1999
Chan H, Lu R, Tseng C, et al: Effectiveness of the anger-control program in reducing anger expression in patients with schizophrenia. Arch Psychiatr Nurs 17:88β95, 2003
Chang H, Saunders D: Predictors of attrition in two types of group programs for men who batter. J Fam Violence 17:273β92, 2002
Chemtob CM, Novaco RW, Hamada RS, et al: Cognitive-behavioural treatment for severe anger in posttraumatic stress disorder. J Consult Clin Psychol 65:184β9, 1997
Coon DW, Thompson L, Steffen A, et al: Anger and depression management: psychoeducational skills training intervention for women caregivers of a relative with dementia. Gerontologist 43:678β89, 2003
Dahlen ER, Deffenbacher JL: A partial component analysis of Beck's cognitive therapy for the treatment of general anger. J Cognit Psychother 14:77β95, 2000
Davison GC, Williams ME, Nezami E, et al: Relaxation, reduction in angry articulated thoughts, and improvements in borderline hypertension and heart rate. J Behav Med 14:453β68, 1991
Deffenbacher JL: Cognitive behavioural conceptualization and treatment of anger. J Clin Psychol 55:295β309, 1999
Deffenbacher JL: Cognitive-relaxation and social skills treatments of anger: a year later. J Counsel Psychol 35:234β6, 1988
Deffenbacher J, Dahlen E, Lynch R, et al: An application of Beck's cognitive therapy to general anger reduction. Cognit Ther Res 24:689β97, 2000
Deffenbacher J, Demm P, Brandom A: High general anger: correlates and treatment. Behav Res Ther 24:481β9, 1986
Deffenbacher JL, Filetti LB, Lynch RS, et al: Cognitive-behavioural treatment of high anger drivers. Behav Res Ther 40:895β910, 2000
Deffenbacher J, Lynch R, Oetting E, et al: Anger reduction in early adolescents. J Counsel Psychol 43:149β57, 1996
Deffenbacher J, Oetting E, Huff M, et al: Fifteen-month follow-up of social skills and cognitive-relaxation approaches to general anger reduction. J Counsel Psychol 42:400β5, 1995
Deffenbacher JL, Stark RS: Relaxation and cognitive-relaxation treatments of general anger. J Counsel Psychol 39:158β67, 1992
Deffenbacher JL, Story DA, Stark RS, et al: Cognitive-relaxation and social skills interventions in the treatment of general anger. J Counsel Psychol 34:171β6, 1987
Deffenbacher J, Thwaites G, Wallace T, et al: Social skills and cognitive-relaxation approaches to general anger reduction. J Counsel Psychol 41:386β96, 1994
Deffenbacher JL, McNamara K, Stark RS, et al: A comparison of cognitive-behavioural and process-oriented group counselling for general anger reduction. J Counsel Devel 69:167β72, 1990
Deffenbacher JL, McNamara K, Stark RS, et al: A combination of cognitive, relaxation, and behavioural coping skills in the reduction of general anger. J Coll Stud Devel 31:351β8, 1990
Deffenbacher JL, Huff ME, Lynch RS, et al: Characteristics and treatment of high-anger drivers. J Counsel Psychol 47:5β17, 2000
Diaz LA: A comparison of cognitive restructuring and systematic desensitization techniques for anger reduction with an inmate population. Dissertation Abstracts International, Section B: The Sciences and Engineering 61:1078, 2000. Available by subscription at http://www.proquest.com
Dua J, Swinden M: Effectiveness of negative-thought-reduction, meditation, and placebo training treatment in reducing anger. Scand J Psychol 33:135β46, 1992
Eamon K, Munchua M, Reddon J: Effectiveness of anger management program for women inmates. J Offend Rehabil 34:45β60, 2001
Erwin B, Heimberg RG, Schneier FR, et al: Anger experience and expression in social anxiety disorder: pretreatment profile and predictors of attrition and response to cognitive-behavioural treatment. Behav Ther 34:331β50, 2003
Evershed S, Tennant A, Boomer D, et al: Practice-based outcomes of dialectical behaviour therapy (DBT) targeting anger and violence, with male forensic patients: a pragmatic and non-contemporaneous comparison. Crim Behav Ment 13:198β213, 2003
Fane RB: The effect of structured and unstructured group therapy on the anger and attitudes of domestic violence perpetrators. Dissertation Abstracts International, Section B: The Sciences and Engineering 59:4-B, 1998. Available by subscription at http://www.proquest.com
Fava M, Rosenbaum JF, Pava JA, et al: Anger attacks in unipolar depression. Part 1: clinical correlates and response to fluoxetine treatment. Am J Psychiatry 150:1158β63, 1993
Fehrenbach P, Thelen M: Assertive-skills training for inappropriately aggressive college males: effects on assertive and aggressive behaviours. J Behav Ther Exp Psychiatry 12:213β17, 1981
Forbes MR: The effects of prosocial skills training on anger management of aggressive adult inmates. Dissertation Abstracts International, Section B: The Sciences and Engineering 52:2-B, 1991. Available by subscription at http://www.proquest.com
Galovski T. Blanchard E: The effectiveness of a brief psychological intervention on court-referred and self-referred aggressive drivers. Behav Res Ther 41:1055β67, 2002
Gerzina MA, Drummond P: A multimodal cognitive-behavioural approach to anger reduction in an occupational sample. J Occup Org Psychol 73:181β94, 2000
Gidron YK, Davidson K, Bata I: The short-term effects of a hostility-reduction intervention on male coronary heart disease patients. Source Health Psychol 18:416β20, 1999
Gildea TJ: Anger management in the treatment of mild hypertension. Dissertation Abstracts International 49:4595, 1989. Available by subscription at http://www.proquest.com
Gonzalez-Prendes AA: A study of the effects of anger-control group counseling on attributional styles and levels of trait anger in women recovering from alcohol and or drug addiction. PhD dissertation, Wayne State University, Detroit, MI, 2002. Retrieved July 12, 2007, from Dissertations and Theses: Full-text database. Publication No. AAT 3071782. Available by subscription at http://www.proquest.com
Gonzalez-Prendes AA: Cognitive-behavioural treatment of men and anger: three single case studies. Cognit Behav Pract 14:185β97, 2007
Grodnitzky GR, Tafrate RC: Imaginal exposure for anger reduction in adult outpatients: a pilot study. J Behav Ther Exp Psychiatry 31:259β79, 2000
Haaga DA, Davison GC, Williams ME, et al: Mode-specific impact of relaxation training for hypertensive men with type-A behaviour pattern. Behav Ther 25:209β23, 1994
Hagiliassis N, Gulbenkoglu H, DiMarco M, et al: The anger management project: a group intervention for people with physical and multiple disabilities. J Intellect Dev Disabil 30:86β96, 2005
Hanusa DR: A comparison of two group treatment conditions in reducing domestic violence. Dissertation Abstracts International, Section A: Humanities and Social Sciences 54:4589, 1993. Available by subscription at http://www.proquest.com
Harris AH, Luskin FM, Norman SB, et al: Effects of a group forgiveness intervention on forgiveness, perceived stress, and trait-anger. J Clin Psychol 62:715β33, 2006
Hart KE: Anxiety management training and anger control for type A individuals. J Behav Ther Exp Psychiatry 15:133β9, 1984
Hazaleus S, Deffenbacher J: Relaxation and cognitive treatments of anger. J Consult Clin Psychol 54:222β6, 1986
Howells K, Day A, Williamson P, et al: Brief anger management programs with offenders: outcomes and predictors of change. J Forensic Psychiatry Psychol 16:296β311, 2005
Kennedy SM: Anger management training with adult prisoners. Dissertation Abstracts International 52:6087, 1992. Available by subscription at http://www.proquest.com
King NN, Lancaster WG, Nettleton N, et al: Cognitive-behavioural anger management training for adults with mild intellectual disability. Scand J Behav Ther 28:19β22, 1999
Kolko DJ: Clinical monitoring of treatment course in child physical abuse: psychometric characteristics and treatment comparisons. Child Abuse Neglect 20:23β43, 1996
Lanza M, Anderson J, Bosvert C, et al: Assaultive behaviour intervention in the veterans administration: psychodynamic group psychotherapy compared to cognitive behaviour therapy. Perspect Psychiatr Care 38:89β97, 2002
Lindsay WR, Allan R, Parry C, et al: Anger and aggression in people with intellectual disabilities: treatment and follow-up of consecutive referrals and a waiting list comparison. Clin Psychol Psychother 11:255β64, 2004
Linehan MM, Heard HL, Armstrong HE: Naturalistic follow-up of a behavioural treatment for chronically parasuicidal borderline patients. Arch Gen Psychol 50:971β4, 1993
Medd J, Tate RL: Evaluation of an anger management therapy programme following acquired brain injury: a preliminary study. Neuropsychol Rehabil 10:185β201, 2000
Mendes de Leon CF, Powell LH, Kaplan BH: Change in coronary-prone behaviours in the recurrent coronary prevention project. Psychosom Med 53:407β19, 1991
Moon JR, Eisle RM: Anger: an experimental comparison of three behavioural treatments. Behav Ther 14:493β505, 1983
Murphy CM, O'Farrell TJ: Factors associated with marital aggression in male alcoholics. J Fam Psychol 8:321β35, 1994
Nakano K: Effects of two self-control procedures on modifying type A behaviour. J Clin Psychol 46:652β7, 1990
Novaco RW: Treatment of chronic anger through cognitive and relaxation controls. J Consult Clin Psychol 44:681, 1976
O'Donnell CR, Worell L: Motor and cognitive relaxation in the desensitization of anger. Behav Res Ther 11:473β81, 1973
Pennebaker D, Lesen T: South Metro Men's Respite: an innovative service for violent men. Australas Psychiatry 10:20β3, 2002
Porzelius JE: The effects of relaxation training on psychophysiologic and psychological reactivity to stress in males at risk for developing essential hypertension. Doctoral dissertation, Rush University, College of Nursing, Chicago, IL. Retrieved July 12, 2007, from Dissertations and Theses: Full Text Database. Publication No. AAT 8822753, 1988. Available by subscription at http://www.proquest.com
Richards H, Kaplan M, Kafami D: Progress in treatment and experienced and expressed anger among incarcerated men. J Offend Rehabil 30:35β58, 2000
Rimm DC, Hill G, Brown A, et al: Group-assertive training in treatment of expression of inappropriate anger. Psychol Rep 34:791β8, 1973
Rokach A: Anger and aggression control training: replicating attach with interaction. Psychotherapy 24:353β62, 1987
Rose J, West C, Clifford D: Group interventions for anger in people with intellectual disabilities. Res Dev Disabil 21:171β81, 2000
Salwan JF: Management of stress-related anger in vocational rehabilitation clients: comparison of cognitive-behavioural therapy and relaxation coping techniques. Dissertation Abstracts International 47:2631, 1986. Available by subscription at http://www.proquest.com
Sarason I, Johnson G, Berberich JP, et al: Helping police officers to cope with stress: a cognitive-behavioural approach. Am J Community Psychol 7:593β603, 1979
Schmitz B, Presley S: An anger control program for students with behavioural problems. Iowa J School Soc Work 9:32β53, 1997
Schmitz JK: The effects of a structured treatment group for women with difficulty managing anger. MSSW dissertation, The University of Texas at Arlington, Arlington, TX, Retrieved July 12, 2007. Dissertations & Theses: Full Text Database. Publication No. AAT 1342510, 1990. Available by subscription at http://www.proquest.com
Schmitz MJ: An outcome study to determine the clinical effectiveness of an anger management program in an adult, rural Minnesota sample. Dissertation Abstracts International, Section B: The Sciences and Engineering 66:4500, 2006. Available by subscription at http://www.proquest.com
Siddle R, Jones F, Awenat F: Group cognitive behaviour therapy for anger: a pilot study. Behav Cognit Psychother 31:69β83, 2003
Smith LL, Beckner BM: An anger management workshop for inmates in a medium security facility. J Offend Rehabil 19:103β11, 1993
Stapleton JA, Taylor S, Asmundson G: Effects of three PTSD treatments on anger and guilt: exposure therapy, eye movement desensitization and reprocessing, and relaxation training. J Traum Stress 19:19β28, 2006
Stermac L: Anger control treatment for forensic patients. J Interpers Viol 1:446β67, 1987
Tang M: Clinical outcome and client satisfaction of an anger management group program. Can J Occup Ther 68:228β36, 2001
Taylor JL, Novaco RW, Gillmer B, et al: Cognitive-behavioural treatment of anger intensity among offenders with intellectual disabilities. J Appl Res Intellect Disabil 15:151β65, 2002
Taylor JL, DuQueno L, Novaco R: Piloting a Ward anger rating scale for older adults with mental health problems. Behav Cognit Psychother 32:467β79, 2004
Taylor JL, Novaco R, Gillmer W, et al: Individual cognitive-behavioural anger treatment for people with mild-borderline intellectual disabilities and histories of aggression: a controlled trial. Br J Clin Psychol 44:367β82, 2005
Thomas MK: Assessment of the effectiveness of anger management treatment in Vietnam veterans with posttraumatic stress disorder. PhD dissertation, University of Arkansas, Fayetteville, AR. Retrieved July 12, 2007, from Dissertations and Theses: Full Text Database. Publication No. AAT 3122391, 2003. Available by subscription at http://www.proquest.com
Thomas NB: The effectiveness of cognitive-behavioural anger management group in reducing measured attitudes of aggression. PsyD dissertation, George Fox University, Newberg, OR. Retrieved July 12, 2007, from Dissertations and Theses: Full Text Database. Publication No. AAT 9304475, 1990. Available by subscription at http://www.proquest.com
Thurman CW: Effectiveness of cognitive-behavioural treatment in reducing Type A behaviour among university faculty. J Counsel Psychol 32:74β83, 1985
Thurman CW: Effectiveness of cognitive-behavioural treatments in reducing type A behaviour among university faculty: one year later. J Counsel Psychol 32:445β8, 1985
Timmons PC: A cognitive-behavioural treatment for anger: comparison of group counseling and computer guidance. Dissertation Abstracts International, Section B: The Sciences and Engineering 55:2019, 1994. Available by subscription at http://www.proquest.com
Timmons P, Oehlert M, Sumerall S, et al: Stress inoculation training for maladaptive anger: comparison of group counseling versus computer guidance. Comp Hum Behav 13:51β64, 1997
Trampus JW: Analysis of a psychoeducational intervention in the affective domain (anger). Dissertation Abstracts International, Section A: Humanities and Social Sciences 59:3715, 1999. Available by subscription at http://www.proquest.com
Vannoy SD, Hoyt WT: Evaluation of an anger therapy intervention for incarcerated adult males. J Offend Rehabil 39:39β57, 2004
Walley JC: Imaginal exposure and response prevention for anger and aggressive behaviour. Dissertation Abstracts International, Section B: The Sciences and Engineering 63:2080, 2000. Available by subscription at http://www.proquest.com
Whiteman M, Fanshel D, Grundy JF: Cognitive-behavioural interventions aimed at anger of parents at risk of child abuse. Soc Work 2:469β74, 1987
Willner P, Jones J, Tams R, et al: A randomized controlled trial of the efficacy of a cognitive-behavioural anger management group for clients with learning disabilities. J Appl Res Intellect Disabil 15:224β35, 2002
Wlazelek BG: The effects of stress inoculation for anger control and relaxation training on anger and related measures with nurses. Dissertation Abstracts International 51:2640, 1990. Available by subscription at http://www.proquest.com
Acknowledgments
The author wishes to thank Graham Glancy, MB, ChB, for useful feedback and constructive contributions to this review.
- American Academy of Psychiatry and the Law