The Influence of Anger and Imagery on the Maintenance and Treatment of PTSD

Abstract

Research on post-traumatic stress disorder (PTSD) has burgeoned since its introduction in DSM-III in 1980. PTSD is conceptualised as a disorder of recovery and has been regarded as intrusion-driven, disordered, anxiety. However, recently there has been a call for explanatory theories of PTSD that better capture the complexity of the condition. Problematic anger is now recognised as an important aspect of PTSD in most sufferers. It is a key predictor of the development, maintenance and severity of PTSD and may be the principal impediment to successful treatment. Nevertheless, the psychological mechanisms underlying the relationship between PTSD and anger are not well understood. This chapter reviews evidence that imagery is an important mechanism within this relationship and is fundamental to the experience of traumatic stress reactions. Imagery is directly related to the prevalence of intrusions in PTSD and is highly correlated with posttraumatic anger. Further, visual imagery with angry content has profound psycho-physiological effects, magnifies the intensity of experienced anger and, ultimately, mediates the experience of PTSD itself. This review elucidates the linkages between angry imagery and PTSD symptomatology and offers propositions for adapting imagery-based PTSD interventions to the treatment of anger-affected PTSD.

1. Introduction

Interest in PTSD as a clinical phenomenon burgeoned following its formal introduction in the third edition of the Diagnostic and Statistical Manual of Mental Disorders in 1980 [1]. Today, the PTSD literature is vast and “internationalised” [2]. Based on available estimates [3] and publication trends [4] more than 20,000 articles on PTSD have thusfar been produced.

Post DSM-III, research has driven many advances in conceptualising how PTSD develops, how it is experienced, prolonged and effectively treated. That work has resulted in detailed explication of the disorder’s phenomenology, extensive theorising about its aetiology and clinically important descriptive and explanatory models of PTSD associated with best practice treatment.

Despite this progress, there remain many uncertainties, controversies, prominent points of disagreement and critical knowledge gaps about PTSD. For example, there is ongoing debate over which experiences may be considered traumatic [5] and the nature of the stressor(s) required to meet diagnostic criteria. Moreover, DSM 5 [6] and ICD 11 [7] differ in how they derfine the construct of complex PTSD. There is also uncertainty about how to best treat PTSD. This is most evident when PTSD is complicated by psychiatric comorbidities such as anxiety, substance abuse and depressive disorders that require decisions to be made about treatment sequencing. Practitioners have expressed concern about sub-optimal implementation and translation of evidence-based psychological treatments for PTSD [8]. This has led to calls for further examination of transdiagnostic protocols [9] and personalised approaches for treating PTSD to enhance treatment outcome efficacy [10].

Another significant problem relates to the influence of dysfunctional anger in PTSD. Dysfunctional anger there is associated with increased morbidity and mortality. For example, through physical ill-health via conditions like cardio-vascular [11] and hypertensive disease, especially stroke [12]. The ruminative processes and negative self, other and world-appraisals associated with dysfunctional anger can impair the capacity to reason, and diminish problem-solving and goal-setting ability [13, 14]. Anger can also be a sign of unsuccessful attempts to deal with the experience of a traumatic event.

Sufferers of enduring PTSD often experience unrelenting, intense and distressing anger and significant impairment in interpersonal functioning. Those experiences are associated with degraded interpersonal relatedness [15], disparagement of others, and failure in relationship and family functioning [16]. When anger is extreme this can produce social alienation [17]. This leaves a trail of hurt and damage that most disconcertingly has a strong nexus to aggression [18, 19], violence [20] and suicidality [21].

Current and ex-serving military personnel are troublingly prone to experiencing problematic anger associated with their combat service and are considered the angriest veterans yet encountered [22, 23]. However, while dysfunctional anger is best documented in military personnel and veterans with PTSD [24, 25], it has been identified as a problem across a range of PTSD-affected populations. Such populations include those occupationally at risk for PTSD, like first responders [26], crime victims and perpetrators, those injured in workplace and road traffic accidents [27] and survivors of disasters [28], terrorism, atrocities, torture and political oppression [26, 29, 30].

PTSD has been characterised as a disorder of recovery requiring efficacious application of better psychological theories and treatments if better treatment outcomes are to be achieved [31, 32]. Highly pertinent to this objective, anger is not only strongly related to the maintenance of PTSD but plays a critical causal role. Upper-end estimates suggest 40% of PTSD score variance can be attributed to anger [33] and that reductions in anger lead not only to reductions in PTSD symptoms but caseness [34, 35]. Anger is also associated with poorer prognosis, high treatment dropout and is, perhaps, the principal impediment to the successful treatment of PTSD. This has been recognised in comprehensive reviews of anger [9] and anger in PTSD [34] and treatment studies, such as those focused on combat veterans [35].

Anger has emerged as an important aspect of the experience of PTSD that warrants further attention. However, there remains a need for nuanced explanatory models which capture the psychological mechanisms underlying the relationship between anger and PTSD. Accordingly, this chapter seeks to make clear why and how those phenomena are joined. Specifically, it seeks to elucidate the role of imagery as a psychological mechanism underlying the relationship of anger and PTSD in an integrated (visual-linguistic) cognitive model. These core objectives are pursued by reviewing the dominant theories of anger and anger in PTSD, and imagery. The limits of existing explanations of the influence of anger on PTSD and the role of imagery in that relationship are considered. A model of anger in PTSD is proposed that emphasises the role of imagery. This provides a platform for developing a fuller description and theoretical account of the aetiology and maintenance of PTSD. This platform permits identification of conditions required for better adapting imagery-based PTSD interventions to the treatment of anger-affected PTSD. It also suggests directions for future research on the role of anger and anger-Influenced imagery in PTSD.

2. The relationship between PTSD and anger

Anger is strongly associated with other dysphoric and aversive emotions. When dysfunctional it is closely associated with psychiatric conditions. Research shows that individuals with anxiety disorders experience greater anger severity, aggression, hostility and anger-related impairment, compared to those without them [9]. Anger is also often prominent in chronic pain conditions (particularly complex regional pain syndromes and fibromyalgia) [36] and psychotic spectrum disorders [9]. It is further associated with personality disorders, especially borderline, narcissistic, paranoid and antisocial types [37].

Anger’s strongest link is, however, with PTSD. Sufferers of PTSD experience greater anger compared to individuals with subthreshold PTSD [38] and those without it [9]. They also have greater difficulty with anger control, compared to individuals suffering social or generalised anxiety or panic and obsessive-compulsive disorders [9, 39]. This has been identified in clinical samples like combat veterans [24] and non-clinical student samples [38].

The anger-PTSD association reflects several influences. Initially, there is the effect of anger’s base-rate prevalence. Although it is not well understood clinically, anger, along with anxiety and depression, forms the “big three” [40], “unholy trinity” or “FAD” (fear-anger-depression) [41] of negative affect, but is considered the predominant emotion in treatment contexts [42]. Moreover, negative affects rarely occur alone. Typically, they share a content overlap and are recursively interdependent in their dysphoric effects. Consequently, beyond the direct impact of its high base-rate, anger also has a significant indirect association with PTSD through its strong relationship with other primary emotions [43].

Anger is more intertwined with other negative affects than any other emotion and comorbid aversive emotions are the norm in PTSD. Prominent, high impact emotions in PTSD that share an intimate relationship with anger include horror and disgust [44] and “responsibility emotions”, such as guilt and shame [45, 46]. Shame-proneness is thought to be related to anger arousal, resentment and irritability and associated with indirect/non-expression of anger [47]. The angriest people have strong underlying feelings of guilt and shame [48]. Tournier observed that “Irritation-aggressiveness: this is the law of unconscious and repressed guilt [… and …] those affected by guilt may be understandably irritable and at times, explode in anger or rage” (page 150) [49].

Considered one of the “moral emotions” [50], anger is a response to perceived failure to meet responsibility [51] and social norms [52] and is concerned with norm enforcement [53]. Exposure to potentially traumatising events (PTEs), and particularly extreme and malevolent interpersonal PTEs, almost inevitably raises issues of causality and intense meaning-related issues around responsibility and failure by commission or omission. This is provided for in hypothetical constructs that seek to explain trauma-related responses associated with anger. For example, the core symptoms of posttraumatic embitterment disorder [54] include: negativity, helplessness, blame of self and others, non-specific somatic symptoms, phobic avoidance of persons or situations related to the PTE, intrusions, phantasies of revenge and aggression. Additionally, in betrayal-based moral injury [55], anger is directed at another person who is perceived as having perpetrated a betrayal resulting in the injury or worse of associates.

Importantly, it is the relationship anger has with PTSD that is critical. This was highlighted almost 30 years ago by Lasko, Gervis, Kuhne, Orr and Pitman [56], who observed “increased aggression in war veterans is more appropriately regarded as a property of PTSD, rather than a direct consequence of military combat” (page 373). That observation was elaborated upon in the first meta-analysis undertaken in relation to anger and PTSD by Orth & Wieland in 2006. They concluded that “anger and hostility are substantially related to PTSD among samples who have experienced all possible types of traumatic events, not only in individuals with combat-related PTSD” (page 704) [57].

3. Anger in PTSD and the role of imagery: Insights, theories and models

Imagery has emerged as a fundamental cognitive process which is intimately involved in all emotional responses and an important feature of the experience of many forms of psychopathology [58]. It is also integral to interventions aimed at ameliorating negative emotional states and appears to be an important mechanism where anger is associated with PTSD [34].

3.1 Imagery in the stress and trauma literature

Mental imagery is the quasi-perceptual, subjectively-influenced, cognitive representation and recollection of perceptual experience in working memory in the absence of the originating stimulus [59, 60]. Although all five senses can generate imagery, its predominant form is visual [61]. Importantly, while all senses can be involved, the visual domain is most typically engaged in the application of imaginal treatments of PTSD. Referred to by a variety of names-including visual imagination [62], pictures in the mind [63] and seeing in the mind’s eye [64]—visual imagery is less visually acute than perception. However, it preserves the perceptible properties of the stimulus and ultimately gives rise to the subjective experience of perception [60]. Henceforth, the term imagery refers to visual imagery. As imagery is more affectively-valenced than thought [58, 65], it imbues emotional memory with an intensity consistent with the original experience [66]. Thus mental images can be experienced as “realer than real” [62] and even influence the ability to experience emotion [67].

Imagery has an especially important relationship with negative emotion [68, 69] and psychopathology. Its role is well-documented in psychotic, dissociative, mood, substance-related, psychosomatic and anxiety disorders [58, 70, 71]. Imagery’s strongest connection to negative, disordered emotion is however, among the anxiety disorders, notably specific phobia, social anxiety and stress disorders [72]. Imagery is implicated in various forms of specific phobia, including the imagery of the feared stimuli central to snake, spider and vomiting phobias [73, 74]. It is also a key factor in social anxiety disorder [75] especially where it pertains to negative imagery of the self [76]. Its relationship to PTSD is axiomatic, intrusive imagery being accepted as a core symptom, risk factor and severity-moderator [77].

The role of imagery in PTSD has been investigated overwhelmingly by reference to the affect of fear. This dates to the late nineteenth century, when Pierre Janet proposed that posttraumatic syndromes driven by intrusive recollections were experienced as disordered anxiety [78, 79], An anxiety-based conception of trauma response is evident in the early posttraumatic stress case descriptions and constructs (e.g., nostalgia and railway spine). It is also emphasised in the early-to-mid twentieth century constructs of shell shock, combat exhaustion and compensation neuroses [80]. Subsequently, the diagnostic and classificatory systems of psychiatry—in ICD-9 [81] and DSM-III [1], formalised this anxiety-based definition of PTSD and ICD 10 [82], and DSM-IV [83] in turn, reiterated it. Highly effective, anxiety-focused treatments have, consequently, been designed and implemented consistent with this anxiety-based conceptualisation of PTSD. Those treatments are summarised and discussed shortly for their implications.

Notwithstanding the advances made in PTSDs treatment, the historical focus on fear as its primary emotion may have impeded a fuller understanding of the disorder [84]. This is demonstrated in the growing recognition that PTSD is associated with emotions other than anxiety [85]. Fifty percent or more of PTSD’s affective experience is estimated to relate to anger directly or in conjunction with disgust, guilt and shame [31, 86].

The relocation of PTSD to the Trauma and Stressor Related Disorders section of DSM 5 [6] from its previous location in the Anxiety Disorders section of DSM-III [1] and V [83] indicates the disorder’s ongoing conceptual evolution. There is, of course, no barrier to exploring PTSD as strongly-anger related because of an association with imagery: the worth of that will be proven by the data.

These classificatory changes, however, remain contentious [87]. Proponents of major theories of PTSD such as the emotional processing model of PTSD understand it as an anxiety disorder [80, 88, 89] and the view that PTSD is an anxiety disorder is tenacious [80]. Irrespective of this, the insistence that the distress of all who experience PTSD is, without exception, anxious, has obscured evidence that the distress of a substantial group of sufferers is not primarily anxious in nature.

Research of anger in PTSD related to imagery has received less attention than that focused on anxiety in PTSD. Nonetheless, various studies suggest imagery and anger share an important relationship in the presence of PTSD. Higher levels of anger are associated with greater responsivity to imagery, while visual intrusions not only compound, but elicit, anger [57, 90]. Two reviews have explored the underlying relationship between imagery and anger as an aspect of negative emotion or in its own right. The first, by Holmes and Mathews [58] identified three overarching explanatory themes for the impact of imagery on negative emotion (anger included). The first centred on imagery’s direct effect on the brain’s emotional systems (whereby, imagery stimulates and is stimulated by, emotional and physiological arousal). The second noted its similar impact to real events (imagining an act engages the same neurological motor and sensory programs involved in carrying it out). The final theme was the capacity of imagery to reactivate feeling states such that attempts to avoid or suppress imagery result in unwanted and unintentional increases in its frequency and intensity. The second review by McHugh and colleagues [34] identified neurological, psychopathological and affective lines of evidence for the relationship of anger and imagery. They also formulated an imagery-informed model for better understanding the relationship of anger to PTSD, while detailing imagery’s powerful effect on anger in PTSD.

Research has further identified imagery’s role in eliciting angry mood and physiological reactivity [43]. Angry imagery has a particularly powerful effect on human physiological responses [91] and can generate responses greater than those derived from anger-provoking events themselves [92]. This link is evidenced by imagery’s role in successful treatments for problematic anger across a range of populations [13, 93, 94].

Anger research shows that imagery impacts anger irrespective of imagery ability or repetition [95]. Consistent with this, high-intensity, emotional distress in PTSD promotes the experience of intrusions [59]. Posttraumatic intrusion repetition in turn increases imagery capacity [96] and vividness [97], in a circular affect-imagery relationship. Anger is not exempt from this and imagery’s association with anger and anger in PTSD may not be dose-dependent. Thus, any level of imagery may be potent in its capacity to escalate anger. Finally, in PTSD, there is evidence that visual intrusions are both a cause and consequence of posttraumatic anger [57, 90, 98] and PTSD sufferers with high imagery control have fewer intrusions and less anger than those with low imagery control [99].

3.2 Theoretical models of anger and anger in PTSD

Several theoretical perspectives on the aetiology and maintenance of anger are potentially applicable to better understanding anger in PTSD. These theories are detailed in Table 1.

Theory	Content
Social learning theory [100]	Understands behavioural modelling is crucial to new learning. It emphasises the individual propensity to imitate behaviour observed in significant others, especially parents. According to it, three regulatory systems are thought to control behaviour: (1) contingencies, (2) feedback and (3) cognitive function. It proposes anger can serve as a habit to deal with distress and is more likely to occur when there is affective distress and some contingency exists for angry problem alleviating (e.g., as in anxiousness and anger) and the result is a circular problem of aversion, distress and anger.
Information processing theory [101, 102]	Asserts that the manner in which individuals perceive phenomena plays a role in maintaining emotions. Problem anger is often be associated with threat identification and unless corrective information is applied, misinterpretation ensues. Past events, thoughts, feelings and behavioural responses and meanings, are all stored in a “memory network”.
Appraisal theory [103, 104]	Proposes that it is the perception of an event, not the event itself which is the key determinant of angry affect. It further emphasises that appraisals are not only necessary, but sufficient causal factors for the experience of anger.
Primary/secondary emotion taxonomy [43]	Suggests overcontrol of (primary^) anger around concern about the consequences of its expression and uncontrolled expression of (secondary) anger are both potent causes of dysfunction.
Contextual model of anger [105]	Emphasises the experience of anger reflects the role of: Situational factors; like, disrespectful treatment, unfairness/injustice, frustration/interruption, annoying traits in others and irritations. Distal factors; such as embeddedness (via issues that may be personal., familial or social), interrelatedness (with other emotions and past experiences) transformationalism (from isolated instances of anger to chronic anger problemsand severe acts of aggression) and Ambient factors; for example, those relating to the environment (incl. weather).
Neo-Associationist Theory [106]	Proposes anger involves a constellation of inter-related physiological, motoric and cognitive “responses”. These are associated with the inclination to defend against or attack a target.

Table 1.

Major explanatory theories of anger.

Of these theories, many are foundational psychological theories. Among them, social learning theory [100] understands behavioural modelling as crucial to the development of learned anger through the propensity of individuals to imitate behaviour observed in significant others (e.g., parents). In this theory three regulatory systems control behaviour—contingencies, feedback loops and cognitive function. It is proposed that anger can exist as a means for dealing with distress and is more likely to occur during affective distress and when some contingency exists for anger alleviation. The result is a circular problem of aversion, distress and anger.

Social learning theory fits well with the conditioning model of PTSD proposed by Keane and colleagues [107]. A social-interactionist learning theory, it was derived from conditioning theories of pathological anxiety, such as classic Pavlovian fear conditioning and Mowrer’s two factor model [31]. Consistent with such theories, it posits that unconditioned stimuli (e.g., traumatic events associated with military experiences) automatically evoke unconditioned emotional (fear) responses. The intensity of this response generates avoidant protective responses. Warzone and combat-situations are quintessential stress environments. In combat life and death contingencies motivate highly-charged anger and emotional information processing that can become distorted and predictive of later anger in PTSD [108]. These reactions are often facilitated by pre-combat military training that mobilises the supposed “strength” of anger to avoid the dysphoric “weakness” of anxiety. This training dehumanises enemy combatants and operationalises the military imperative to negate and eliminate their threat. This renders military personnel more likely to respond to (objective) stress and trauma with anger, thereby precluding or impeding the development of other salient emotions such as anxiety and remorse [107]. The programmatic nature of pre-combat training makes anger in PTSD difficult to de-operationalise and anger can become associated with a multitude of seemingly trivial day-to-day occurrences. These are not directly associated with the original traumatic experience but are subjectively interpreted as if they were and are associated with extreme levels of distress. What is not as readily understood is the significant function anger can play in the avoidance of such affects. This is specifically accommodated in Greenberg and Paivio’s [43] primary-secondary emotion taxonomy and Beck’s model of anxiety [109] which both emphasise the tendency to replace incapacitating distress with anger’s action-orientation. Anger thus becomes a costly camouflage for other primary emotions.

Two more anger theories with significant potential utility in explaining anger in PTSD, are the information processing theory of anger [101] and the appraisal theory of anger [103, 104]. The information processing theory asserts that the perception of potential cues to anger is critical in maintaining angry affect. It proposes that problem anger is associated with threat identification and unless corrective information is applied, misinterpretation ensues. It also proposes that past events, thoughts, feelings and behavioural responses and meanings, are all stored in a latent “memory network” ready for activation. The appraisal theory of anger proposes event perception, and not the actual event, is the key determinant of anger, and that appraisals are not only necessary, but sufficient, causal factors for the experience of anger. This is particularly the case for perceptions of responsibility, culpability and entitlement [51, 110]. Neither of these theories have been explicitly investigated for their specific utility in explaining anger in PTSD populations. Nonetheless, they are consistent with established PTSD theories. Thus, the potential utility of the information processing theory of anger is suggested by both the information processing theory [111] and the conditioning model [107] of PTSD and the threat identification and misinterpretation they identify as occurring after exposure to military PTEs. The potential utility of the appraisal theory of anger is suggested by the warning signal hypothesis [112, 113] and the cognitive vulnerabilities model of PTSD [114]. Euphemistic of notions of neuroticism, the latter emphasises the impact of a negative attributional style for past and current-events and looming cognitive style for future-events on PTSD. This functions as a danger schema for predicting future threat and is strongly connected to perceptions about trauma and PTSD symptoms.

Two anger models have been explicitly investigated in relation to anger in PTSD. The first theory, the primary, secondary and instrumental emotion taxonomy developed by Greenberg and Paivio [43] holds that primary emotions, like fear and shame, are fundamental, direct and initial reactions to events and situations. Secondary emotions, by definition, are responses to thoughts or feelings, rather than the situation (e.g., anger in response to hurt, fear or guilt). Their theory posits that anger can be experienced as preferential to underlying, aversive, dysphoric states.

This taxonomy has been applied to research on veterans [115] and female crime victims [18]. The data support the view that anger deflects sufferers from intrusion-activated fear to a state less associated with feelings of vulnerability. This is consistent with the assertion by Riggs, Dancu, Gershuny, Greenberg and Foa that, in PTSD, anger and dissociation are both processes of disengagement or avoidance of the traumatic memory and fear network [18]. It is also consistent with Forbes and colleagues [22] finding that angry veterans with PTSD believe they are misunderstood and maltreated and tend to blame others for their mixed-emotion distress. These processes, are avoidant in nature. Although they may afford temporary relief from anxious distress through the pseudo-positivity of anger, they inhibit habituation. This prevents disconfirmatory or safety-related cues being incorporated into the trauma memory network to modify its associations and interpretations [18, 115, 116].

The neo-associationist memory network theory [106] extends the insights of the information processing and appraisal theories of anger. It proposes that anger involves a constellation of inter-related physiological, motoric and cognitive “responses”. Associated with the inclination to defend against or attack a target, research on veterans with PTSD shows that associative networks connect negative affect with anger-related feelings, thoughts, intrusive memories and aggressive behavioural inclinations [117].

In addition to the PTSD-focused research on anger theories, studies of survivor mode [35, 118] have been undertaken in relation to anger in PTSD. This is a dysregulation model of anger in PTSD and not an anger theory per se. It views anger as governed by higher-order cognitive perceptual processes and emotional functions. It is the only PTSD specific, anger model designed and researched with PTSD in mind. It emphasises the importance of anger-related schemata in interpreting the self, others and the world. Such schemata reset anger activation-inhibition patterns toward a cognitive set revolving around mis-perceived threat. This invokes an unrecognised, peremptory, all-consuming threat-anger action program that is enacted automatically in response to the merest or ambiguous, cues.

Studies of these three perspectives are few and their propositions around the mechanisms likely to contribute to anger in PTSD are not well-established. Taken together, their propositions emphasise anger’s threat perception and appraisal-tendencies and its interconnectedness with other emotions (especially anxiety, guilt, shame and disgust). They also underscore that anger is linked to meaning making around responsibility and culpability associated with the conduct of the individuals or others. This is particularly true for behaviours with questionable morality, injustice or malevolence, diminishment and the self-focused expectations and behaviours of others.

Such a theoretical synthesis is supported by the testimony of individuals with PTSD involving prominent anger. McHugh [4] reported on the accounts of posttraumatic anger in a sample of 500 treatment seeking current and ex-serving military personal and first responders with PTSD. The sample was comprised primarily of police but also included ambulance officers, fire services personnel and other emergency services workers. The content of their intrusions and their recollections, cognitions and associated negative emotions, states of being and action tendencies are described in Table 2.

Danger	Emotions/states	Contrariness	Injustice	Diminishment	Expectations
Perceived:	Feelings/sensations of:	Perceived:	Perceived:	Perceived:	Unmet/misplaced:
Risk (physical and psychological) to self/significant others; threat of attack to self/significant others; other’s failure to see risk or danger to self or significant others; vulnerability (physical and psychological^); recklessness; impulsivity; carelessness; mistakes; other’s (pre)caution failure; bullying; intimidation; coercion; malevolence	(Dis)tress; worry, fear; insecurity; tension; jealousy, envy; hatred; annoyance; vengefulness; guilt and shame; embarrassment; disgust; loss, grief and despair; depression; helplessness; hopelessness; powerlessness; sadness and sorrow; fatigue; remorse, regret; disappointment; suspicion & paranoia; perfectionism	Disagreement; challenging; individualism; “non-team”; behaviour; norm exception; defiance; disloyalty; “passivity”; indifference; ambivalence; disengagement; selfishness; rejection; deceit; disloyalty; disobedience; dishonesty; “stupidity”; deviance; perversity; demanding	Unfairness; wrongdoing; moral transgression; malfeasance; plotting; culpability; persecution; discrimination; revenge; overreaction; insensitivity; “sloth”	Dismissal; diminishment; disrespected; devalued; shamed; humiliation; blame; betrayal; exploitation; duplicity; criticism; rejection; neglect; abandonment; intolerance	Rights; wants; desires; needs; entitlements; standards; norms; (esp. of others^)

Table 2.

Perceptions, cognitions, emotions and constructs associated with anger in PTSD.

That testimony demonstrates the plethora of phenomena that can underlie the experience of anger in PTSD. It also conveys both the mental busyness of those with, enduring, angry PTSD and the powerful avoidant role anger plays in distracting from contemplation of crucial underlying issues.

In summary, theories and explanatory models of anger and anger in PTSD provide potentially important clinical understandings. Research suggest that frameworks such as the neo-associationist and primary/secondary taxonomy have direct clinical relevance to anger in PTSD. To date, however, research of anger has been limited in comparison to other affects such as anxiety and depression [52, 119] and there has been little attention given to anger in PTSD. Furthermore, empirical research focusing on the involvement of imagery in the facilitation, exacerbation and prolongation of anger in PTSD has thus far been negligible. Consequently, despite the involvement of imagery in emotional, cognitive and memory processes in PTSD, there is a dearth of theories and explanatory models of the relationship of imagery and anger in PTSD. There is therefore a clear need for multi-representational descriptive and explanatory theories and models.

3.3 The sequential impacts of anger and imagery on PTSD: a summary understanding

Drawing on the theoretical models of anger and PTSD described in the preceding sections, this section outlines a conceptual understanding of imagery’s role in the sequential processing of anger in PTSD. Figure 1 lays out a hypothetical sequential processing map of how anger develops in response to trauma-related cues and triggers.

Figure 1.

Sequential processing map of the hypothetical development of in response to trauma-related cues and triggers.

As shown in Figure 1, PTSD is considered to involve disordered information processing along the lines of the information processing and appraisal literatures. Such disordered processing is strongly associated with the experience of dysphoric and aversive emotional states. Typically, this involves anxiety, emotions of responsibility (e.g., guilt and shame) and emotions of repugnance (e.g., horror or disgust). As argued by Greenberg and Paivio [43], in this context anger emerges as a primary, secondary or substituted affect associated with such primary emotions.

Anger is particularly evident where the PTE has involved human culpability and, above all, where there is malevolence or morally dubious behaviour. That is provided for in trauma typologies [4, 34, 120]. The impact and chronicity of PTSD is thus intimately associated with the event perpetrator’s intention, the event’s meaning and its moral status [121, 122, 123]. Under such circumstances, information processing demands are significant, and cognition becomes laboured and prone to be “affected by affect”. Where such demands apply, consistent with Beck’s cognitive model [109] imagery is likely to be invoked in a vicious circle where processing becomes automated due to the effect of priming or the encoding of information and schematic processing via mental maps. These maps interpret tasks and the environment in which they occur out of conscious memory. Such schematic pattern analysis and interpretive maps are likely to drive and respond to priming (Figure 1). This part of the framework is consistent with appraisal-focused theories of PTSD.

The encoding of information in memory, increases the capacity to recall the event and related information from cuing, intentionally or otherwise [124]. This robustly exacerbates angry responses to post-priming trivial events [125, 126] and repetitive anger episodes lead to automaticity and secondary episodes that occur within 10–20 min of the first [126]. Anger episode duration increases with repetition and intensity. Primed by via repeated provocation (imagined or real) anger escalates in a non-linear fashion and significantly outlasts event duration (Figure 1). Thus provoked, anger can escalate even when the provocation remains constant (e.g., via a constant low-level annoying sound) and the sequence of escalation proceeds via reproaches, insults and threats [126]. Priming, perceived or actual provocation and externally-located causation are associated with rumination. That rumination may be verbal or imaginal in nature [34].

Other important anger moderating and mediating factors include disruption of anger due to fear, the presence/absence of distraction possibilities and apology [126, 127]. The duration and, concomitantly, rate of anger decline is influenced by the inability to control and cease rumination. Importantly, when anger occurs the form of revenge fantasies, rumination can have a half-life of more than a couple of weeks [128].

Because of the content of the intrusions and recollections involved in PTSD (Table 3) or the process repetition of the memory, poorly controlled verbal or imagery-based cognition increases the intensity of PTSD symptoms and associated anger. Appraisal theories of PTSD and anger suggest anger is likely to be prolonged where rumination is poorly controlled and potentially distracting factors and apology are absent. That perspective aligns with the cognitive vulnerabilities model of PTSD [114] and multi-representational theories of PTSD; such as the Dual Representation Theory of PTSD [129] and Schematic, Propositional, Analogue, Associative, and Representational Systems model of PTSD [132].

Cognitive Therapy for PTSD [129]	In CT-PTSD, the individual is assisted to modify unhelpful responses which result from biased or distorted thoughts and memories associated with or arising from their traumatic experience (s). It seeks to modify excessively negative appraisals, correct autobiographical memory disturbance and address cognitive distortions and problematic behaviours, as well as any subsequent maladaptive or unhelpful beliefs individuals may develop about themselves, others and the world. It does so on the basis that PTSD remains persistent when individuals process the trauma in a way that leads to a sense of serious, current stress and danger which leads to involuntary re-experiencing of aspects of the trauma.
Prolonged exposure [88]	In PE, the individual is supported to gradually confront their memories of their traumatic experience and situational reminders of that experience that are otherwise avoided. This involves assisting them to change the way they think and feel about the traumatic experience and develop more helpful ways of coping, through education about common reactions to trauma, breathing retraining, behavioural exposure (to feared situations that clients avoided due to trauma-related fear), imaginal and processing (discussion of thoughts and feelings related to the exposure exercises)
Eye movement desensitisation and reprocessing [130]	In EMDR, the individual is required to focus on particular images, thoughts and bodily sensations related to the traumatic experience while simultaneously being sensorily stimulated. Most commonly, this is done by having the client move their eyes back and forth across their visual field (e.g., by tracking the movement of the therapist’s finger). This process may be repeated many times. It is proposed that this dual attention protocol facilitates the processing of the traumatic memory into existing knowledge and memory networks and assists the individual to process the trauma. Although the precise operational mechanisms involved are not known, over time, EMDR has increasingly included treatment components comparable with CBT interventions.
Cognitive processing therapy [131]	In CPT, the individual is assisted to identify unhelpful thoughts and beliefs (“stuck points”) and subsequently challenge and replace them with rational alternatives. This is done via an adaptation of standard cognitive therapy approaches. CPT has a smaller exposure component than PE that is typically restricted to writing an account of the traumatic experience. CPT also helps to address associated problems such as depression, guilt and anger.

Table 3.

Evidence-based treatments for PTSD.

In this sequential model, anger can be ignited by the recollection of what happened, who caused or permitted it to happen and what occurred afterward. Anger can also emanate from the ruminative replaying of unfinished business in dreams or involuntary or, often unwitting, voluntary daytime recollections [133]. Shown in Figure 1, irrespective of the source or intentionality, the resulting distress intensifies avoidance.

The pressure to avoid is associated with deliberate and unwittingly attempts to suppress intrusions. This is especially the case for aversive negative-emotion-influenced autobiographical imagery. Such imagery is typically experienced with a greater sense of reality (e.g., as measured by vividness), compared to non-emotional or semantic imagery [58]. Attempted suppression of intrusive phenomena is, however, almost always ineffective. This is well accounted for by Ironic Process Theory/the Zeigarnik Effect [134, 135, 136]. Efforts to suppress mental content, images included, can paradoxically lead to increased (re)occurrence of that specific content. Thus, rather than its intended outcome, avoidance leads to perverse, counter-intentional and unwitting rebounds [137] and even increases intrusions and imagery [138].

In an apparent case of the “greater the emotion, the greater the PSTD”, symptoms are magnified when the dysphoric emotions associated with PTSD reach sufficient intensity thresholds. Anger can be either or both an initiator or potentiator of to this symptom escalation. Imagery is never far from such increases and the relationships among PTSD, anger and imagery are anything but coincidental [4].

4. A guideline for the evidence based psychological treatment of anger in PTSD

Treatments for PTSD with a high level of supporting evidence (i.e., randomised controlled trials with substantial sample sizes, systematic reviews and meta-analysis) are listed in Table 3. These treatments comprise: trauma-focused cognitive therapy, prolonged exposure, eye movement desensitisation and reprocessing and cognitive processing therapy. Collectively, the interventions fall under the rubric of trauma-focused cognitive-behavioural therapy, meaning they directly focus on recollections, cues and triggers of the PTE and its associated cognitions and emotions [139].

All these interventions require practitioners to pay deliberate, detailed attention to the provision of psychoeducation to assist clients to develop and maintain a personal model of recovery. They also require clinicians to help clients to develop skills to manage the maladaptive cognitions and affects which arise in the context of PTSD. Ultimately, they necessitate clinicians assisting clients to reprocess memories of causal traumatising events that they most likely, will have been avoiding. Finally, treatment aids the (re)development of functional abilities relating to everyday events based on the restoration of homeostatic mechanisms, habituation to emotion and new learning [88].

As well as exposure, imagery underpins various PTSD treatment interventions, including imagery rescripting and reprocessing therapy [140] and the treatment of post-traumatic nightmares via imagery rehearsal [141]. Imaginal exposure is, however, by far the most researched of these intervention techniques and its efficacy has been demonstrated across a broad sweep of trauma-exposed populations. Over 50 randomised controlled trials with substantial sample sizes [142, 143], multiple meta-analyses [144] and systematic reviews [145] support the use of imaginal exposure. It has deep historical roots in the treatment of pathological anxiety and has demonstrated treatment superiority across the range of anxiety disorders. It considered the most important advance in the psychological treatment of anxious distress in the last 70 years [146].

The utility of exposure in the treatment of PTSD characterised by fear, of course, does not explain whether, how or why it will ameliorate anger in PTSD. Effectively treating anger in PTSD, requires understanding of its aetiology and maintaining factors and its best practice treatment. This review has detailed the former; the latter is, however, yet to be established. To begin to address this gap the following prototypical guidelines are suggested for the effective application of evidence-supported imagery interventions to PTSD where anger is the predominant emotion.

4.1 Know and observe the limits of imagery-based interventions in the context of PTSD

Whether in the psychological or physical treatment realm, no intervention is, or can be, effective without exception. Thus while exposure is the gold standard treatment for PTSD [147, 148] and the archetypal example of imagery-focused approach to PTSD treatment, not all PTSD presentations respond well to it. For example, a study of active US military personnel with PTSD (N = 326) randomly allocated to massed exposure (10 daily treatment sessions delivered over 2 weeks) or spaced exposure (10 treatment sessions delivered over 8 weeks) treatments identified classes of responders. These classes comprised rapid responders, steep linear responders, gradual responders, non-responders and symptom exacerbation [88]. Similar responder typologies emerged in a study of individuals with PTSD associated with multiple interpersonal traumata and previous attempts at treatment (N = 73) [149]. Participants received an intensive phase of treatment (12 daily 90-minute sessions over 4 days) followed by a booster phase (4 weekly 90-minute booster sessions). While 71% were classed as treatment responders, cluster analysis demonstrated four treatment response trajectories. These were: fast responders (13%), slow responders (26%), partial responders (32%), and non-responders (29%).

The reasons for variations in treatment response to exposure have been ongoingly reviewed. Research has identified the impact on response to exposure treatment of symptom profiles. For example, those associated with PTSD’s numbing symptom cluster [150], residual sleep problems [151] and emotional dysregulation [152] and ruminative and absorption processes [34]. Other studies have demonstrated that specific brain regions are involved in a differential response to or discriminated between responders and non-responders to exposure treatment. These include the bilateral superior frontal gyrus and pre-supplementary motor area [153] and pre-treatment hippocampal volumes [154]—and pre-treatment hormone levels, as measured in cortisol [155].

Importantly, a variety of imagery-related characteristics can mediate or moderate the efficacy of exposure in the treatment of PTSD. Although PTSD, anger and imagery share a connection, exposure and other imagery-based treatments may be unsuited to the certain presentations of anger in PTSD due to the appraisals and the presence of other negative affects [4]. This is because of the effect of the event types and the associated appraisals. To illustrate, PTSD-related anxiety classically revolves around issues of danger and risk. In contrast, in PTSD-associated anger (as noted in section 2, page 4), those appraisals, their underlying assumptions and the emotions that arise extend well beyond issues of danger and risk. As depicted in Table 3, those issues relate to wrong and injustice, diminishment, annoyance around expectations, disgust, guilt and shame and other non-anxiety-based affects. Such issues and associated anger are highly likely to be present in the face of horrific, macabre or morally-questionable behaviours associated with PTEs.

Given this, it is possible that the application of exposure, to problematic anger may not produce image decay and emotional habituation—the active mechanism involved in the remediation of the anxiety associated with PTSD. Instead, this may induce or exacerbate the frequency, intensity and duration of angry affect. This is because anger control is not about habituation and holding an angry image in the mind’s eye until it decays. Rather, it is related to image control. Importantly, PTSD sufferers with high imagery control are known to have fewer intrusions and less anger than those with low imagery control [99]. While what constitutes imagery control in anger is opaque, imagery elimination and suppression are unviable alternatives to the unmitigated experience of imagery. Arguably, imagery control is characterised by the ability to down-regulate, disconnect from or relinquish imagery.

At different times, imagery may have little effect on anger in PTSD or have a large, singular and direct effect on anger. Arguably, any such effect may also be indirect and multiply determined in its interaction with other cognitive mechanisms, like thought-based appraisals. Furthermore, the motivations for angry responding may also possess a utilitarian social value (e.g., in relation to social justice and protection of the weak). Consequently, activating/invoking angry mentation via imaginal exposure may not produce the sought-after reduction in angry affect. This is greatest in PTSD, where anger interferes with the development of the treatment alliance that is so important in the successful treatment of PTSD [156] and, is perhaps, the principal impediment to the successful treatment of PTSD. Comprehensive reviews [9, 34] and treatment studies [35] having recognised its impact and metanalyses demonstrate its interference in exposure tasks [39]. Anger’s impact on treatment is particularly notable in combat veterans [157], but is also likely to affect populations subject to malevolent interpersonal trauma (e.g., childhood and pernicious adult sexual assault [158]).

As Meichenbaum [13] incisively observed, intense anger is not easy to work with and clients with angry presentations may become more so during treatment and direct anger and aggression toward clinicians. He further noted that such angry clients are often highly impatient, easily frustrated, unrealistic about treatment goals, typically noncompliant with treatment and treatment resistant. In the face of such anger, clinician effectiveness is subject to significant challenge. Consequently, in worst case scenarios, clinicians are liable to be rendered impassive in the face of significant client anger. This may partially account for the all-too-often failure of clinicians to follow evidence based practice [8, 159] and to divert from protocols when faced with difficult-to-treat anger [160]. This is of concern as it is likely to produce sub-optimal implementation and translation of evidence-based psychological treatments of PTSD.

4.2 Recognise the influence of individual differences and circumstances on imagery

To be effective, exposure must be applied while cognisant of individual differences in the capacity to image in a person-appropriate manner. Group characteristics, such as ethnicity influence the capacity to experience imagery. Notably, imagery is greater among the peoples of East Asia [161] and indigenous cultures, like those of Australia, which utilise imagery in practical day-to-day tasks [162]. There are also gender differences in the capacity to utilise imagery [163]. Women have a superior ability to generate and maintain images [164] and experience more vivid images than men [165, 166]. They also react more strongly to unpleasant affective images, while men react more strongly to pleasant affective images [167]. Finally, the experience of imagery reduces with age [164, 168, 169, 170] and there is unequivocal evidence that imagery capacity degrades with age [171, 172].

The efficacy of exposure is also influenced by innate differences in imagery capacity. There are various abilities or traits related to the capacity to image. These include absorption, thinness of reality-imagination boundaries and imagery vividness [173]. Some types of imagery are personality based [174]. The style of imagery-based information processing is also affected by cognitive style and appraisal tendencies. This involves the balance between reflective thinking and thought-based information processing, known as verbalising style [175, 176]. Individuals with a higher capacity for visual imagery experience more visual and other sensory details when remembering or imagining past and future events [177].

Many abilities and personality traits influence the capacity to image. Imagery vividness has been associated with angry personality-based obsessionality [178] and individuals with high trait anger have been shown to have greater reactivity to angry imagery in the absence of enhanced imagery ability [95]. Imagery absorption is another dispositional trait that is highly correlated with the tendency to image and the intensity of the imagery experience [179, 180]. Under conditions of significant stress, absorption can become an imagery-based coping mechanism.

Greater imagery control is associated with greater internal locus of control [181], while extraversion and introversion are associated with imagery fluency [182]. The latter is consistent with the long-standing interpretation that extraverts are verbalisers and introverts are imagers [175, 176]. Importantly, this appears to be mediated by stress levels and Stricklin and Penk [183] found that, among incarcerated female offenders, extroverts reported more vivid imagery than introverts under high-distress. In contrast, introverts reported more vivid imagery than extroverts under low-distress.

The suitability of imagery to the remediation of anger is PTSD is also likely to vary according to situational circumstances. This is well illustrated from workplace injuries, where compensation and treatment claims involving psychological injuries are more difficult to administer and likely to become complicated where their aftermath possesses certain characteristics. These include a workplace climate which fails to promote workplace health and wellbeing or where the risk of injury is poorly managed or increased by inappropriate work practices. It can also involve the failure of the insurer to promptly approve best practice treatment for the injuries and worker perception(s) that the employer and/or insurer do not adequately care for them or where the event is associated with the injury involved horror, disgust human malevolence or culpable negligence [147]. In such circumstances, a sense of injustice motivates angry psychological distress in workers. Anger is associated with what happened and who allowed it to happen and/or failed to respond to their needs after an injury they neither expected nor caused. This powerful sense of wrong and the restorative justice required to address can instigate blazing and righteous anger and consequent revenge fantasises in those who perceive themselves unduly treated in the workplace [128]. This can lead to the targeting of health service providers and compensatory health and legal systems by individuals with traumatic injuries [184].

The Contextual Model of Anger (see Table 1) heavily emphasises the importance of the interaction of several contributing factors. First is the anger derived from situational factors involving the experience of perceived or objective disrespectful conduct, unfairness, injustice, being wronged, thwarting of goal attainment and annoying behaviour of others. The second anger causing factor relates to ambient factors. For example, those relating to the environment, such as noise odour and temperature. Over time, through repetitious exposure and associative tendencies these factors become compounded in a third, distal factor. This involves embeddedness (via issues that may be personal, familial or social), interrelatedness (with other emotions and past experiences) and transformationalism (from isolated instances of anger to chronic anger problems and severe acts of aggression).

Novaco proposes that anger involves regulatory deficits in three psychological domains These include a cognitive domain (justification, attentional focus, rumination, hostile attitude and suspicion); an arousal domain (intensity, duration, somatic activation and irritability) and a behavioural domain (impulsive reactions, verbal aggression, physical confrontation and indirect expression of anger). The more deficits, characteristics and domains activated the greater the anger. Thus, in a study of veterans with PTSD and high levels of anger in multi-year anger treatment trial, Chemtob and others [35] described individuals who displayed high intensity regulatory deficits in all three domains of anger as “ball of rage patients”.

4.3 Apply evidence-supported and anger-specific treatments to anger in PTSD

There is strong evidence of the general effectiveness of anger treatment. Bushman and colleagues have written extensively about this and have emphasised the utility of anger regulation interventions, such as cognitive therapy (especially distancing, behavioural distancing and language moderation), skills training and relaxation training [185, 186]. Meta-analyses by Tafrate [187], Edmondson and Conger [188], Beck and Fernandez [189], DiGiuseppe and Tafrate [190], Del Vecchio and O’Leary [191] and Henwood and others [192] and Lee and DiGiuseppe [193] report effect sizes as measured by Cohen’s d [194] ranging from 0.64 to 1.16. Notably, Glancy and Saini [195] in their systematic review of meta-analyses of psychological treatments of anger and aggression, observed there are strong effect sizes for “classical” CBT approaches (skills training and problem solving) but lesser effect sizes for “variants”, like acceptance and commitment therapy. They found multi-component interventions to be most effective (d = .93), followed by skills training (d = .85) and cognitive interventions alone (d = .83). They reported a trend for manualised treatments to produce larger effects than non-manualised treatments (d = .85 v. 76) The number of sessions also has a positive and significant relationship to the magnitude of the effect size, although the modal number of sessions is 8.5 and range of sessions provided is from 3 to 40. They further noted that such outcomes been established for clients of diverse backgrounds. This included forensic clients, violent recidivists, batterers, adults with intellectual and learning disabilities or serious mental illness, aggressive drivers and military personnel and veterans.

In comparison to the treatment of anger per se, the development of treatment strategies for anger in PTSD is in its formative stages and few interventions for anger and aggression in PTSD have been elaborated [196]. The CBT approaches identified in these reviews and meta-analyses have implicit applicability to anger in PTSD, provided they are nuanced for the impact of traumatisation. Research on PTSD associated anger demonstrates that cognitive therapy and skills training are effective in treatment of dysfunctional anger. These appear most effective when delivered individually or by group and face-to-face or remotely (e.g., by teleconferencing) [197].

Two studies have examined treatment approaches with solid face validity with promising results. The first by Mackintosh and colleagues [22], investigated the roles of anger regulation skills (i.e., relaxation training) and therapeutic alliance in reducing anger symptoms in contemporary (N = 109) US veterans. It identified that gains in calming skills predicted significantly larger reductions in anger symptoms. This finding has intuitive merit, for it is not possible that, apart from situations involving Schadenfreude, an individual can be simultaneously calm and angry. It also fits well with the above-identified effectiveness literature on the treatment of anger.

The other study used self-instruction training (SIT) which is an intervention with intuitive validity in the treatment of anger in PTSD [13, 198]. The aim of SIT is to enhance coping in face of adverse events by the use of pre-rehearsed self-talk instruction. Cash and colleagues [199], in an Australian contemporary serving combat personnel population, established a cognitive skills training set centred around self-instruction training. They used SIT targeted at negating the operation of schema modes [200], and reported impressive anger and PTSD symptom reduction effects (they reported effect sizes of 1.6 as measured by Cohen’s d) pre to post-treatment. This efficacy of SIT for anger in PTSD fits with the long-demonstrated history of the efficacy of SIT, including for individuals with overlapping comorbidities, like anxiety, mood and substance abuse disorders [13]. It is also supported by the reviews and metanalyses cited at the start of this sub-section of the review.

Although such studies do not explicitly argue in favour of the use of imagery, it is clearly possible for it be involved in their implementation. Imagery may also be utilised in other treatment strategies for anger regulation, such as cognitive therapy, and treatment methods, such distancing and behavioural rehearsal. The best practice use of imagery in treatments of anger and anger in PTSD is yet to be determined. For example, the style of SIT deployed by Cash and colleagues emphasised a coping approach that sought to tolerate challenging situations. Their approach stands in contrast to the imaginal reliving of hierarchically organised provocative situations and experiences in stress-related anger studies by Novaco on populations as diverse as incarcerated individuals with intellectual disabilities and police [201]. The relative merits of these approaches-with their competing emphases on coping versus habituation-for reducing or exacerbating anger in PTSD, require explication. This is also the case for other treatments of anger involving imagery. This may require existing anger treatments to be finessed for the impact of trauma when applying them to anger in PTSD.

4.4 Locate anger work within a phased PTSD treatment model

All treatments for PTSD are recommended for delivery via a staged sequence. This has been agreed for some time among theoreticians and researchers and long argued from diverse theoretical paradigms, including eclectic, psychodynamic and integrative perspectives (see Table 4).

Eclectic [202]	Psychodynamic [203]	Psychodynamic [204, 205, 206]	Integrative approaches [207]
Encounter/Education Phase	Psycho-physiological Assessment Phase	Initial Exploration and development of therapeutic relationship Phase	Education and Affect Regulation Phase
Skill building & client empowerment Phase	Behaviour therapy Phase	Working through Phase and event re-appraisal Phase	Self-harming/limiting behaviour reduction via cognitive interventions Phase
Exploring trauma & its impact	Dynamic therapy	Development of new adaptive actions	Trauma focussed work
Evaluation & integration Phase	Existential therapy	Working Through (Marmar/Lindy) Practice of adaptive actions until automatic Horowitz	Self-awareness and self-acceptance Phase
Termination Phase	Termination Phase	Termination & Loss Phase

Table 4.

Staged treatment models and theoretical treatment orientations.

A stage intervention sequence requires clinicians to provide an explanatory model of posttraumatic stress and how it best treated. A treatment roadmap is also needed that outlines the shared and individual responsibilities of clinician and client. Such an approach, provides a psychologically reassuring treatment structure for clients and enables them to build a personal model of recovery. A comprehensive staged model of treatment has been articulated by Keane and Kaloupek [208]. It involves six stages of treatment: (1) emotional and behavioural stabilisation, (2) education and information, (3) arousal management, (4) exposure treatment, (5), cognitive restructuring and (6) relapse prevention and maintenance. A space for exposure is provided in its fourth stage. This is preceded by an arousal management treatment stage, where any anger-focused work required may be given focus.

It has long been recognised that, where posttraumatic anger is intense, more treatment will typically be needed before the implementation of exposure treatment [209]. Anger is particularly associated with enduring PTSD resulting from events characterised by culpability, issues of existential meaning (or in its simple form, moral injury [210]. It is also associated with intense primary (e.g., shame and guilt) responsibility and/or repugnance-related (e.g., horror and disgust) emotions. In prolonged PTSD anger is further connected to strong overarching concerns with atonement or revenge that manifest as psychopathology [116]. This almost inevitably requires treatment of greater frequency and duration.

4.5 An algorithm for the use of imagery-focused treatment in PTSD

Taking the treatment requirements, and the literature on which they are based, into account, a phased decision sequence can be proposed for imagery-focused treatment of PTSD with or without anger. When the primary presenting emotion is:

• anxiety experienced as worry or fear—use imaginal and behavioural exposure augmented by trauma-informed anger-work for any irritability that may arise in conduct of that exposure work

• anger secondary to anxiety or a sense of danger—use behavioural and imaginal exposure preceded by trauma-anger-work focused on coping in the moment not reliving past experiences and addressing the avoidant, emotion substitution and camouflaging impact of the presenting anger and

• a non-fear-based affect-and especially where there are issues of responsibility (guilt and shame), repugnance (horror, disgust or repulsion), wrong or injustice or a sense of diminishment, and expectations that run contrary to a “team ethic” or the needs of others—use trauma-informed anger-work that takes account of thought, appraisal, language and imagery inputs and utilises evidence-based anger interventions.

5. Directions for future research

Over a decade ago, Shalev noted that the US Institute of Medicine judged the scientific evidence for the treatment of PTSD as below the level expected for such a common, disabling disorder. He observed significant progress was being made in the disorder’s treatment, but was limited by an apparent treatment-ceiling-effect and a need for more efficacious application of better psychological theories [32]. Consistent with this view, the need to broaden the focus of enquiry in PTSD beyond anxiety-based models was simultaneously identified in comprehensive reviews [31, 211].

As shown in this review, anger may be the predominant emotion for a majority of PTSD presentations [31] and it has vast costs that heavily impact on individuals, partners, families and communities. Yet, research on anger remains surprisingly sparse, with the most recent estimates suggesting it may equal as little 0.6 % of all PTSD publications [4]. There is an unequivocal need to increase research on anger in PTSD. This section suggests directions for future work.

An initial objective of research on PTSD must be to explicitly recognise the importance of anger to PTSD. Anger has been described as forgotten [212], unrecognised [9] and misunderstood [213] and the rate of research of anger in PTSD is lower than the level that might be reasonably expected on account of wide-ranging deleterious impacts. Explicit acknowledgement of anger’s importance for prolonging PTSD, and as a factor influencing treatment outcome, will facilitate research on enhancing treatment protocols for treatment resistant clients. This proposition is supported by the work of various PTSD researchers who have illuminated the problem of anger, including Pitman and others [209], Elbogen, Johnson and Beckham [20], Forbes, McHugh and Chemtob [214], Morland and others [23, 197], Rona and colleagues [108] and Worthen and associate [215].

A second important research objective is to better understand the nature of anger and its relationship to imagery in PTSD. Despite its importance, the phenomenology of dysfunctional anger and imagery in anger in PTSD is not well described. Clearer identification of imaginal and linguistic cognitive processes and their relationship to anger-related PTSD in an integrated (visuo-linguistic) cognitive model of anger in PTSD would be another important step in research on anger in PTSD.

Research that identifies how anger in PTSD interacts with the characteristics of the PTE will be important in identifying differences in maintaining factors of PTSD after specific PTEs. It will also be useful to understanding the relationship between anger and cumulative trauma (e.g., due to vocation and occupation), the impact of perpetration versus experience of PTEs and the respective impact of exposure to PTEs versus how individuals are cared for post event.

The minting of PTSD in DSM-III created a research impetus that led to critical advances in knowledge of the disorder. That research impetus continues and DSM 5 has identified a dissociative PTSD subtype, sub-syndromal/prodromal PTSD and even a PTSD genotype [31, 211]. There is potentially significant value in exploring the possibility of an angry PTSD subtype to clearly identify individuals for whom anger is the primary emotional and evolve treatments to assist them.

A third area for further inquiry relates to the linkages between anger and other trauma-related emotions. This review has shown that anger can occur in PTSD as a primary or secondary emotion connected to anxiety, responsibility-related emotions (especially shame and guilt) [43] and repugnance-related affects (like disgust and horror) [4, 44]. Understanding how anger is linked to other emotions and to symptom maintenance will aid the further refinement of treatment interventions for anger in PTSD. A full comprehension of the nature of anger in PTSD also requires a deeper understanding of the many cognitive processes associated with anger in PTSD. The application of imagery-based interventions to anger in the context of PTSD may be differentially efficacious depending on the cognitions, appraisals and affective causal pathways involved.

Based on the central argument of this review, a specific set of imagery-related research objectives could be pursued. As part of this, it will be important to establish what it is about visual imagery that promotes or hinders the efficacy of imaginal exposure in PTSD in the presence of anger. This exploration of imagery as a mechanism underlying anger’s relationship to PTSD and the treatment of the disorder, will aid the development of a fuller account of the aetiology and maintenance of PTSD and offer new possibilities for enhanced treatment outcomes. Research is needed on psychological phenomena such as control, voluntary and involuntary experience of intrusions, the impact of content and process-related imagery and distress, cognitive style as expressed in the balance between imagery and linguistic-cognition and the association between personal style and imagery.

The dearth of theories and explanatory models about the role of imagery in anger in PTSD and any other mechanisms which may underlie their relationship are compelling grounds for testing and developing models and theories about different aspects of imagery in its interaction with anger in PTSD. So that imagery is integrated into descriptive and explanatory models with other important influences, like cognitive style and personality, it is important that such theories operate at multi-representational levels of explanation, are well operationalised and easily testable. Finally, it is important that focus be applied to the role of anger in PTSD as an explanation for the non-response to proven imagery-based treatments, like exposure. As part of this, any the means by imagery may reduce the treatment interference of anger must be investigated.

6. Conclusion

This review has elucidated the linkages between anger in PTSD and advanced various propositions underscoring the role of imagery as an underlying mechanism in their relationship. Based on that exposition, guidelines and an algorithm for the efficacious treatment of anger in PTSD have been proposed. That algorithm and those guidelines underscore the importance of appreciating the limits of imagery-based interventions in the context of PTSD and recognising the influence of individual differences and circumstances on imagery in PTSD. They also stress the need to apply evidence-supported anger-specific treatments to anger in PTSD and locate anger-work within a phased PTSD treatment model. This represents the first articulation of such guidance and it is, accordingly, best understood as prototypical in nature.

PTSD is a disorder of recovery and there are significant, positive treatment outcomes associated with evidence-based, gold standard psychological interventions like exposure treatment. This particularly the case where anxiety is the predominant emotion experienced by those with PTSD. Many individuals, however, are likely to be troubled by enduring PTSD characterised by posttraumatic anger. A significant minority of this group are either slow or fail to benefit from the receipt of first rank PTSD treatments, like imaginal exposure. This group is the logical target for increased conceptual and empirical research and descriptive and explanatory theoretical models of anger in PTSD and its treatment are much required. It is also important, given the imagery-based connections between anger and PTSD identified in this review, that such theoretical models attend to the role of imagery. Poor imagery control is the law of poor anger control. This is particularly so in anger in PTSD.

To better account for and treat anger in PTSD, the theoretical models developed need to be multi-representational in nature and attend to affective, physiological, behavioural and account for both linguistic and imaginal cognitive processes. Such an approach is entirely consistent with the observation of Aaron T Beck that “effective cognitive therapy depends greatly on moving beyond purely verbal exchanges to encouraging patients and therapists to resort to their auditory or visual imagery capacities” (page 107) [216].