UK Veterans in Prison: An Exploration of Current Symptoms and Contextual Issues

Adrian Needs, Georgina Hodgman and Emma Pollard


In the March 2011 issue of Military Times it was reported that the film Zulu has been voted the British Army’s “favourite war film of all time” (p. 24). The film vividly (though not always accurately) depicts the battle of Rorke’s Drift in 1879, an event where seemingly overwhelming odds were ultimately defeated by courage, discipline and effective organisation. As narrated in the film’s final moments, this single action led to the award of 11Victoria Crosses.

The subsequent lives of the men depicted are known much less widely. One of the VCs, Robert Jones, shot himself nearly 20 years later whilst another, William Jones, suffered from recurring nightmares in his later years and on one occasion took his granddaughter from her bed as he thought the Zulus were about to attack his Lancashire home (Bancroft, 1988). A letter written a few months after the battle by his company commander in the Royal Engineers described John Chard, who had commanded the defence, as “hopelessly slow and slack ……with such a start as he got, he stuck to the company doing nothing ……he placidly smokes his pipe and does nothing” (cited in Emery, 1977, p.241). A staff officer characterised the company commander of ‘B’ Company, 2/24th, Gonville Bromhead in similar terms: “……. the height of his enjoyment seemed to be to sit all day on a stone on the ground smoking a most uninviting looking pipe. The only thing that seemed equal to moving him in any way was any allusion to the defence of Rorke’s Drift. This used to have a sort of electrical effect on him, for up he would jump and off he would go, and not a word could be got out of him” (Greaves, 2002, p.180). Robert Jones still bore the pain of wounds received in the battle, William Jones had been destitute for long periods and a degree of professional jealousy can perhaps be detected in the contemporary accounts concerning the two officers. Nonetheless it is difficult to escape the conclusion that these soldiers, from an age sometimes regarded as more stoical than our own, had been profoundly affected by their experiences of the battle.


A degree of recognition of the possible effects of combat and other stressors can be traced as back at least as far as the American Civil War, with acceptance, reluctant in some quarters, becoming more firmly established in the industrial- scale conflicts of the early twentieth century (Regel & Joseph, 2010). Catalysed largely by the presenting problems of substantial numbers of individuals who had seen service in the Vietnam War, the appearance in 1980 within the Diagnostic and Statistical Manual (DSM) of the American Psychiatric Association of the category of “post- traumatic stress disorder” (PTSD) was a significant development. It has provided a foundation for a great deal of activity in research, theory and clinical practice that has extended to a wide variety of stressors, settings and populations. As ever, as new information and ideas become available new questions emerge. There have also come more fundamental criticisms and some issues relating to diagnosis and conceptualisation are regarded as problematic (McNally, 2010; Rosen & Lilienfeld, 2008; Hunt, 2010). This provides scope for disagreement over aspects such as prevalence, prevention and the adequacy of support available to sufferers.

The extent of deployment and casualties in the conflicts of the last decade in Iraq and Afghanistan has elevated such concerns with regard to current and former military personnel. Part of the controversy has centred on veterans, offending and the criminal justice system. The present paper is concerned with veterans in prison. Here there is still a shortage of basic descriptive information but no lack of controversy, with substantially different estimates of the number of veterans in prison in England and Wales. Estimates of the proportion of the prison population vary from 3.5% to 16.75%, , reflecting the use of different methods and samples (Brookes, Ashton & Holliss, 2010; Defence Analytical Services and Advice, 2010; National Association of Probation Officers: NAPO, 2008).

There is certainly no evidence to suggest a consistent connection, via PTSD, between military service and offending. Moore, Hopewell and Grossman (2009) argued that popular articles in the USA concerning murder charges against veterans who have seen service in Iraq or Afghanistan present a highly selective and misleading picture. A decade earlier, Burkett and Whitley (1998) challenged some of the myths and assumptions about the scale of violent offending and maladjustment by Vietnam veterans that have passed into popular culture. A report by the Howard League (2010) suggested that UK veterans are, as a whole, less likely to offend than their civilian counterparts despite factors associated with offending (such as being young, male, having low educational attainment, coming from areas of urban deprivation) being common amongst veterans. In the 29 interviews these researchers conducted with veterans in prison accounts making a link between exposures to combat and offending were conspicuous by their absence, though the extent of combat experience within this sample is uncertain.

A rather different picture to the latter is given by the cases summarised in recent NAPO briefing reports (NAPO, 2008, 2009). The representativeness of these cases is unclear, no comparative data are presented and the assertions made concerning the presence in many cases of PTSD are open to question. They do however highlight that patterns of domestic violence or public disorder can occur in veterans against a background of heavy drinking or drug abuse, depression, perceived lack of support (often existing alongside an unwillingness to seek help) and problems in readjusting to civilian life. In some cases long- standing problems, such as a history of abuse in childhood, were apparent. Similarly, Brookes et al’s (2010) article on veterans in a therapeutic community- based prison presented quotes from veterans that alluded to problems with emotional regulation and disconnectedness to others though the generality of such problems with regard to veterans in prison remains uncertain.

There is no evidence to suggest that military service, even deployment to a theatre of operations, invariably leads to diagnosable PTSD. Studies in the UK have argued that only a small minority of veterans can be characterised thus and prevalence is similar to that found in the general population (see Pinder et al, 2010). A consistent finding, however, is that other problems such as depression, anxiety and alcohol abuse are considerably more common amongst veterans than PTSD and there appears to be unequivocal evidence that the prevalence of PTSD increases with combat exposure and intensity (Hoge et al, 2004; Iversen et al, 2008). There is also evidence of fluctuation in symptoms over time with some cases showing remission and others a delayed onset, emerging more than three months after return from deployment (Milliken, Auchterlonie & Hoge, 2007). The question of why some individuals and not others meet the threshold for diagnosis has directed attention to situational and personal factors that are associated with susceptibility and recovery (Iversen et al, 2008). Related questions concern what the diagnosis actually refers to and the degree to which the conventional characterisation of the disorder is appropriate to the military context (Hunt, 2010).

Core symptoms of PTSD recognised in the Diagnostic and Statistical Manual (DSM- IV- TR) of the American Psychiatric Association (2000) are clustered into: re- experiencing aspects of the instigating trauma; avoidance and numbing; and hyperarousal. The support that this structure has received from research is mixed at best. This has indicated, for example, that combining avoidance and numbing obscures important differences between these sets of symptoms in terms of relationships to other variables and outcomes (Taylor, 2006; Elhai, Biehn, Naifeh & Frueh, 2011). Several researchers now favour a structure that incorporates this differentiation, giving four groups of symptoms (Rosen, Spitzer & McHugh, 2008). Buckley, Blanchard & Hickling (1998), who reported evidence to support the twinning of re- experiencing (‘intrusion’) with avoidance and hyperarousal with numbing, found that even some individual symptoms appeared to be misclassified, with hypervigilance and an exaggerated startle response related more to intrusion and avoidance than to the hyperarousal with which they are conventionally linked. One of the reasons such research is important is that it suggests hypotheses about functional relationships between variables. Thus from work such as that by Buckley et al  it appears that avoidance behaviours have a strategic role in preventing or regulating re- experiencing (Regel & Joseph, 2010).

Arguments for change have also been directed at the DSM- IV- TR criteria concerning the existence of an identifiable instigating stressor involving actual or threatened harm and a response centred on intense fear, helplessness or horror. Aside from the potential loss of precision and arguably, good sense in defining traumatic events in terms of responses (McNally, 2010, refers to claims of  people “traumatised” by television coverage or overhearing a crude joke) these are unlikely to be the only emotions involved. Peak moments of emotional distress often involve intense feelings of anger, sadness, betrayal or shame (Holmes Grey & Young, 2005). Such emotions can also colour appraisals after the event: in fact Brewin, Andrews and Rose (2000) noted that some individuals appear not to have experienced intense emotions during the event itself. Even more fundamentally, it has been argued that both the standard symptom clusters and the emphasis on a single event are of limited relevance to the multiple, complex and chronic stressors faced by personnel in combat deployment (Dyer et al, 2009; Hunt, 2010). Effects of combat appear to be cumulative (Jones & Wessely, 2005; Johnson & Thompson, 2008). Furthermore, limitations of standard criteria for PTSD have led to the recognition of ‘Complex PTSD’ or ‘Disorders of Extreme Stress, Not Otherwise Stated’ (DESNOS)  and six areas where functioning is likely to be altered: affect regulation, attention/ consciousness, self- perception, relationships, somatization and  meaning systems (Pelcovitz et al, 1994). This formulation shares several areas with the expanded set for PTSD proposed by Wilson (2004).

The more complex view of post- traumatic stress indicated by the above highlights possible shortcomings in the military context of a simple application of the DSM approach. It suggests, however, that there may be utility in examining the prevalence of individual symptoms reported by veterans. The configuration of symptoms, even if the range of potentially relevant symptoms is incomplete, might be expected to reflect something of the underlying processes involved. This approach is consistent with what Freeman and Freeman (2009) term a “trajectory” view, in which the focus is on variation and mediating factors rather than the presence or absence of a single diagnosed outcome. Such a view is in turn compatible with empirically- based theoretical models based on variations of the central premise that extreme and distressing experiences are often difficult to assimilate.

Such models share the view that the presence of symptoms indicates that the processing of the experiences is incomplete or impaired (Regel & Joseph, 2010). For example, the activation of poorly integrated memories may trigger emotions that characterised the original experience, often centred upon a sense of physical or psychological threat (Ehlers & Clark, 2000; Ehlers, Hackmann & Michael, 2004). Other formulations focus on the implications of events for an individual’s fundamental and sustaining beliefs about the world; some occurrences are difficult to reconcile with the existence of goodness, fairness, safety or self- worth (Janoff- Bulmann, 1989). Both processes are recognised in the dual representation model of Brewin, Dalgleish and Joseph (1996).

Information- processing models of trauma help account for individual symptoms and clusters of symptoms. As was suggested earlier, the significance of avoidance- related symptoms may lie in a need to regulate processing of deeply unsettling memories and experiences in order to prevent feelings of being overwhelmed and to avoid emotional exhaustion (Regel & Joseph, 2010). They also provide a rationale for providing support, circumstances or interventions that lead to the ultimate integration of traumatic experiences into the individual’s verbally accessible autobiographical memory as part of a more or less coherent personal narrative (Hunt, 2010; Regel & Joseph, 2010). Such a perspective also helps to interpret research findings on factors that increase or decrease the likelihood, severity and persistence of post- traumatic stress in veterans. Some of these indicate factors that may be associated with complications in the assimilation of intense, extreme and distressing experiences, whilst others highlight the importance of conditions that might be seen as supporting such assimilation.

It is to such studies that we now turn. King et al (1999) reported from their structural equation modelling of PTSD in Vietnam veterans that “war zone” stressors made the greatest contribution to prediction, followed by post- war factors then pre- war characteristics or experiences. Young age at deployment and early (typically childhood) exposure to trauma were central to the latter. Adverse experiences in childhood have been associated with later PTSD in other military (e.g. Cabrera, Hoge, Bliese, Castro & Messer, 2007; Iversen et al, 2008) and non- military samples (Brewin, Andrews & Valentine, 2000), though some variables such as family instability might have an indirect effect through, for example, the influence of attachment style on the availability of post- war social support. Early experiences of adversity may, for example, contribute to a propensity to “negative emotionality” that amplifies the impact of stressors (Meis, Erbes, Polusny & Compton, 2010), while prior abuse may leave a legacy of internalised shame (Wong & Cook, 1992) or dissociative tendencies that may impair the individual’s capacity to integrate intensely negative experiences (Bremner & Brett, 1997). Iversen et al (2008) also found in their UK sample of Iraq veterans the association of PTSD with young age at deployment, alongside low educational attainment and not being in a relationship.

Whether in a relationship or not, immediately prior to deployment individuals are likely to be faced with thoughts of home and of their own mortality (Reger & Moore, 2009) and for a small minority such concerns may trigger transient mental health problems (Turner et al, 2005) . Iversen et al (2008) reported that factors in theatre predictive of PTSD included deployment in a forward area in proximity to the enemy, potentially traumatic experiences appraised as involving threat to self and others and feeling unprepared for work in theatre. Perceived threats to self and others, a salient aspect of experience in theatre, can include attack, entrapment, injury or death (Kelly & Vogt, 2009). Other stressors strongly associated in the literature with trauma include death or injury of comrades, injury to self, handling dead bodies, failing to carry out a required action, having to carry out an action of which some might disapprove, repeated fire- fights and encountering atrocities or “extraordinarily abusive violence” (Friedman, 2006; Hoge et al, 2004; King et al, 1996, 1999; Leskela, Dierpink & Thuras, 2002; Sareen et al, 2007; Singer, 2004).

Neither should the influence of background stressors be ignored. In current conflicts potentially traumatic events are likely to occur in an environment that demands vigilance and having to make decisions with  far- reaching consequences in ambiguous situations, even in situations different to a conventional battlefield and often in conditions of heat, noise and exhaustion punctuated by periodic boredom; it has been suggested that over time such conditions may undermine the capacity to cope with potentially traumatic events (Cognoscenti, Vine, Papa & Litz, 2009; Freeman & Freeman, 2009; Reger & Moore, 2009).

There is some agreement that individual appraisals predict PTSD more accurately than the objective severity of stressors (Halligan, Michael, Clark & Ehlers, 2003; Iversen et al, 2008; Weiss et al, 2010). Background stress can influence the appraisals that are made, colouring them with themes such as anger, fatigue, frustration and loss until in some cases a “final straw” is reached (Freeman & Freeman, 2009). At a fundamental level the ability to maintain equilibrium and effectiveness can be seen in terms of ‘hardiness’ or resilience. This has been defined by Kobasa, Maddi and Courington (1981) as consisting of commitment (a sense of purpose and meaning), control (a sense of being able to influence events) and challenge (an awareness that change is a normal part of life). Research has indicated repeatedly that hardy individuals are less likely to suffer from adverse reactions after exposure to stressful or even traumatic circumstances (Kelly & Vogt, 2009). Crucially, hardiness may play a major role in determining whether potentially traumatic events in theatre result in PTSD (King et al, 1998). Consistent with this, finding positive meaning in experiences in theatre appears to reduce the risk of PTSD (Fontana & Rosenheck, 1998) and Iversen et al (2008) suggested that the relationship between PTSD and feeling ill- prepared may hinge upon appraisals of reduced control and autonomy. The latter authors also found that the risk of PTSD was increased by low morale.

One factor has been found to surpass hardiness as a mediator of the relationship between exposure to combat and PTSD. This is functional social support (King et al, 1998), though research suggests that these factors are actually highly related (King et al, 1999). Pre- trauma social support is associated with a low prevalence of PTSD and less severe symptoms, whilst post- trauma social support tends to lead to better outcomes (Brailey et al, 2007; Burnell, Coleman & Hunt, 2006; Guay, Billette & Marchand, 2006). The former authors, for example, found that level of connection, support and trust between members of a unit (i.e. cohesiveness) had a marked effect on lowering severity of symptoms. Functional social support refers to the quality of resources from others that provide “confirmation of social identity, instrumental aid, and readily available sources of advice and feedback to group members” (Iversen et al, 2008 p. 519). Iversen and colleagues also drew attention to important dimensions such as a sense of comradeship, of commitment to the group and its value systems. In addition to their suggestion regarding the role of social support in reducing the risk of trauma through increasing a sense of preparedness and control, these aspects seem to correspond quite closely to the characterisation of hardiness given above (and, for that matter,  to the formulation of morale in military units proposed by Manning, 1991). To this extent, resilience or hardiness is determined by the group (Wessely, 2006, cited in Regel & Joseph, 2010) and it is precisely this group ethos, at both unit and organisational level, that the military establishment traditionally has sought to inculcate and encourage (Christian, Stivers & Sammons, 2009; French, 2008).

Yet this powerful effect of the group may be double- edged. It might contribute to the vulnerability of current and former service personnel to ‘survival guilt’ when comrades are killed or wounded (Johnson & Thompson, 2008) or to the culture of heavy drinking that, in part, affirms the solidarity of the group (Fear et al, 2007). The question might also be raised of what happens when this intense level of comradeship, cohesion and mutual trust, forged by a range of shared and often extreme experiences, is no longer available. Finding a network that compares is difficult to achieve (Friedman, 2006). Individuals in this situation might be at risk of losing their sense of purpose in life along with their sense of identity, role, friendship and feeling understood (Burnell et al, 2006; Ormerod, 2009): as such they may lose something of the ‘hardiness’ that sustained them. If, in addition, they are unable or unwilling to talk to others the processing of memories that are difficult to assimilate may be impaired or interrupted (Guay et al, 2006) even if avoidance as a strategy is effective at least some of the time (Hunt, 2010).

These are not the only ways in which veterans may subsequently pay a price for what helped them to cope during military service. In order to meet the operational demands of service in theatre veterans are likely to be well- practised in avoidance, emotional numbing and alertness: service personnel tend not be break down in the middle of an emergency (Scurfield, 2009). There can be problems if such a style is imported into intimate or family relationships (Lyons, 2009). In addition, such a style might increase the likelihood that there is still some assimilation of experiences that needs to take place at precisely the time an individual is faced with the challenges of transition to a new life (Moore, Hopewell & Grossman, 2009).

It is in order to explore some of these issues that the current paper presents findings from a small- scale study on prevalence of PTSD and reported symptoms; a second, qualitative study used two focus groups as a forum in which veterans could discuss experiences before, during and after leaving the military.