Violence,victimization and mental health

Introduction

Violence perpetrated by people with mental illness is usually systematically misreported in media and wilfully misunderstood at large. As a public, we are sold a story with sharp distinctions; people with mental health conditions are one-sidedly represented as perpetrators, as opposed to actors in complex processes involving illness, experience and identity. While intricate, these issues are vitally relevant for understanding the lives of people with mental health problems. In this article, we defined victims as those harmed, injured or killed as a result of a crime, accident or other event, and we define perpetrators as a person who carries out a harmful, illegal or immoral act. It is therefore important to note that people who live with mental illness can fall into either, neither or both categories.

Theories of victimization

Violence, a fact of human societies, is imperceptibly complex. Experience tells us that there is a distinction between doing it and receiving it, although decades of research bear out a close correlation between the two.

Despite this complexity, researchers have often struggled to go beyond approaches focused on mental illness in perpetrators (Fazel, Lichtenstein, Grann, Goodwin, & Langstrom, 2010; Johnson et al., 2000; Swartz et al., 1998). In contrast, the broader non-medical study of victimization experiences has grown accustomed to this complexity over centuries. In the broader context of pre-scientific discussions of victimhood in the Bible and the Mesopotamian Codex of Hammurabi, Von Hentig (1948) has been considered the first empirical investigator to focus on victims, as opposed to perpetrators, of crime-criminal acts emerged from interactions between perpetrator and victim, rather than victims being passive recipients of criminality, which had been the consensus hitherto. For Von Hentig, the victim’s role could be to implicitly consent, cooperate, conspire, or provoke. The concept of ‘victim responsibility’ became an important but controversial research area, exemplified by titles, for example, Stephan Schafer’s ‘The victim and his criminal: a study in functional responsibility’ (Schäfer, 1968) and scholarship attributing responsibility for rape to victim behaviour (Amir, 1967).

Defining and measuring victimization

Framing assault as a straightforward-to-measure ‘exposure’ downplays factors influencing who is defined, and self-defines, as a victim (Dignan, 2004) – including weakness of the person in relation to perpetrator, comparatively virtuous actions of victims in relation to the act, the blamelessness of victims, the unrelatedness of victim to perpetrator, the ‘badness’ of perpetrators and a sociocultural position for victims which can sustain victimhood without threatening vested interests. Typical victimological data, for example, derived through national crime surveys, assess victimizations in cross-section, as static events rather than processes. Process-based understandings may however be important in understanding the influences victimization has on a person’s life and social circle.

Victims interact not just with perpetrators, but also service frameworks which contain, control and prevent crime. One-off measurements of victimization do not capture individuals who are exposed to repeated victimization (Genn, 1988) and criminologists have advocated wider use of repeat surveys (Gabor & Mata, 2004). Victimization data are influenced by the predictors of being able and willing to report/identify victimizations. Reporting victimization is correlated with higher social position and educational attainment; furthermore, more socially disadvantaged groups could be more prone to under-report assaults because they happen so frequently in these sub-groups of the population (Zedner, 2002). Simplistic representations of mental illness and violence undermine the real complexity of victimization, a construct with many facets, involving important links to social structure (the police, the law, the criminal justice system), experiences (the context of violent acts, the relationships involved, the types of hidden harm) and history (violence seen during war, and its relationship to the formation of new identities and the often-uneasy jettisoning of old ones).

Developing an agenda for violence research and mental health

Violence is a part of life for all people, including those living with mental health problems and their treatments (Dolan, Castle, & McGregor, 2012; Khalifeh & Dean, 2010; Short, Thomas, Luebbers, Mullen, & Ogloff, 2013). A broad range of domains have been identified for the effects of assault, and the health effects of violence are substantial (Resnick, Acierno, & Kilpatrick, 1997; Reza, Mercy, & Krug, 2001; Weaver & Clum, 1995). Intimate partner violence (IPV) is associated with suicide attempts in women (Devries et al., 2013; Devries et al., 2011), and there is incomplete evidence linking IPV to intimate partner homicide (Pilger & Watts, 2013). Increasing evidence indicates important health consequences of non-lethal IPV, for example, injury, chronic pain and depression (Campbell, 2002). Aside from direct effects, for example, through injuries and disability, violence is a plausible cause of morbidity through psychosocial insult. Information on the consequences of assault has been derived from survey data and clinical studies, the latter of which have focused on evaluating psychological consequences of violence (Zedner, 2002). Although injuries are intuitively the most likely consequence of assault, around 50% of those who reported assault in the 2002–2003 British Crime Survey (BCS) reported no injury. In all, 11% of assaults resulted in medical attention, and a third of instances involved wounding (Zedner, 2002). In the 1998 BCS, 44% of assault victims reported shock, nearly a third reported fear at the time, a quarter reported emotional upset and a fifth reported sleeping problems as a result. Anger was the commonest emotional response, present in 64% of the victims of violence. In total, 84% of victims according to the BCS reported being affected emotionally after the incident (Maguire, Kynch, & Britain, 2002). In comparison, putative effects of being victimized on family and other social relationships have received little research attention. Shapland, Willmore, and Duff (1985) suggest that 30% of crime victims reported financial loss as a result and that assault is associated with going out less frequently. Those lacking family/social support could be more severely affected by victimization by violence (Kaniasty & Norris, 1992; Maguire, 1980; Skogan, 1987).

As well as direct health effects (Resnick et al., 1997), assault could alter health-related behaviours, for example, involving access to care, physical activity levels (Ross, 1993), changes in smoking pattern and consumption of alcohol (Doherty, Robertson, Green, Fothergill, & Ensminger, 2012). Finally, it is possible that the experience of an assault results in higher levels of stress-induced damage to organ systems, activation of neuroendocrine/neuroinflammatory stress pathways, and ultimately, leads to the development of related diseases, for example, cardiovascular disease and cancers (Brunner & Marmot, 2005; Ford & Browning, 2014; McEwen, 1998). The allostatic load hypothesis suggests that the cumulative adversity results in disturbance in bodily responses to stress, which could in turn result in chronic inflammatory processes, for example, atheroma. In this respect, a causal association involving a combination of physical and psychological changes as mediators is plausible. However, these effects have seen relatively limited resourcing of interventions into the reduction of violence in communities, for example, through restrictions on alcohol (Cook & Moore, 1993) or educational programmes (Heise & Garcia-Moreno, 2002), despite the global identification of violence reduction as a public health imperative (Dias et al., 2005).

Suicide is a plausible result of violence exposure. Janoff-Bulman and Frieze (1983) suggest the experience of violent victimization is accompanied by changes in three assumptions about the world that the person is invulnerable, that life is meaningful and a positive view of the self. The latter two are consistent with cognitive models of depression (Beck, 2002). Individuals who report assault on any occasions over the past year could be a population cumulatively exposed to chronic violent adversity, which could be a more plausible cause of suicide. Chronic victimization is a plausible cause of learned helplessness (Walker, 1977), an important cognitive mechanism for depression, of which suicide is an important consequence.

There is an urgent need to develop better evidence on how much violent victimization there is in populations with established mental illness, what it is like, why it occurs and who is responsible. Qualitative representations of violent acts could be discovered and made consistent, through methods such as interviews, focus groups and process analysis. Processes are important in the study of victimization, as the experience of violence is framed within different institutional contexts. A person may be assaulted on the street, in the shower or on a psychiatric ward, each with potentially discrete ramifications for these engagements, and their implications for experience and mental health.

Rodway et al. (2014) demonstrate high rates of homicide victimization among individuals with severe mental illness (SMI) in comparison to the general population. Khalifeh et al. (2015) have recently demonstrated high levels of victimization in psychiatric outpatients in London, in comparison to controls. This excess was not explained by substance misuse, socioeconomic, ethnic or neighbourhood differences. Explaining the high rates of victimization in these populations is a complex problem. One explanation could relate to stigma, but others could relate to social cognitive differences among patients, implying a role for interventions to improve interactions and skills. Among all this research, several questions remain. What is the burden of inpatient victimization events on the natural history of psychotic disorders? Is there an effect of discrete nursing interventions on these outcomes? How does the complexity of victimization interplay with coping and help seeking for mental disorders?

In assessing the experiences of people with mental health problems in society, research has focused on stigma and discrimination, but less so on victimization experiences, despite their likely importance. Health researchers have seen it someone else’s problem, a task for social agencies, the police, the law and the vagaries of upbringing and background. It is also a like source of important biases – people who are victimized might not divulge their experiences to investigators, and the factors which influence reporting violence could vary by health status. Although the link between victimization and the later development of mental health problems is intuitive, clinicians do not see this manifest in their work in terms of interventions, assessments and day-to-day support. Perhaps it is because we don’t see it manifest in any way other than the mental health and distress of our patients. Or it could be because we don’t have substantive theories about how experiences of victimization events and their processing affect health.

One solution must be to develop thicker understandings of what victimization experiences are, perhaps at the expense of the breadth of information we gather in research studies. Qualitative representations of violent acts could be discovered and made consistent, through methods such as interviews, focus groups and process analysis. Processes are important in the study of victimization, as the experience of violence is framed within different institutional contexts. A person may be assaulted on the street, in the shower or on a psychiatric ward, each with potentially important implications. In particular, we can see that victimization is a useful conceptual tool in understanding relatively atheoretical conceptualizations of adversity. In doing so, we should more closely integrate victimological understandings and measurement and topics of interest with research into mental health.

Interventions should be evaluated for reducing the morbidity of violence exposure in people with severe mental illness. Among all of these risks, social skill training could be an important intervention targeting the way in which less socially able individuals become vulnerable to deliberate harm from others. Useful effects might be gained from training people to negotiate better with others, for example. These studies have not been done – addressing violent victimization in those with severe mental disorders must be considered an urgent key un-addressed research area to improve care quality for people with mental illness. Patient safety should not just be about avoiding iatrogenic harm, but also about trying to prevent all avoidable harms in patients, including those related to violence and its after-effects. This could also have important implications for how we perceive mental illness as a public.

The ethical hazards of psychiatric research into violence

Psychiatry research has generally enshrined the distinction between perpetrators and victims, and has typically assessed the role of mental illness in the former, rather than the latter. The reasons for this imbalance are several. Of course, epidemiological hypotheses and the investigations which assess them are constrained by the available data. Although it is recognized as a social ill in one regard, there is a competing viewpoint that considers violent victimization a part of life, politics and (inherent) human free will. School bullying is a part of life, muggings and assaults are random events. Such is the framework within which these accepted risk factors are viewed by many psychiatrists and clinicians. But this is palpably a representation of societal attitudes to violence, rather than a reflection of empirical evidence, which has accrued in other quarters to suggest violent victimization has discrete predictors (Brennan, Moore, & Shepherd, 2010; Miethe, Stafford, & Long, 1987; Zedner, 2002). Furthermore, psychiatric research has employed methodological approaches that rely on reductive theoretical frameworks, at the expense of approaches embedded in more ‘macro-’ understandings. Macro understandings usually invoke concepts and processes playing out across groups of individuals, rather than within individuals. The role of victimization within broader interpersonal circumstances is a key example of this and is thrown into relief by social science understandings of victimization.

Blogs, critical psychiatrists and a large body of academic opinion complain that psychiatric research has focused untowardly on the neurobiological bases of most neuropsychiatric disorders, at the expense of how these disorders play out in real life. However, such approaches do not exclude investigations that ask questions about how experiences of social life become imprinted not only in the biological brain, but also one’s cognitions. I hope that more attention is paid to the way in which research funders, grant calls and research training support evidence generation on victimization and mental illness, and, in particular, the largely unmeasured social world in which people live and experience mental illness. We must also begin to consider violent victimization as a more complex entity, in closer alliance with social science. In particular, victimization investigations must interface with victimological research considering the broader familial, cultural and societal structures which shape the experience of coming to deliberate harm from others in our society (Walklate, 2013). From an ethical standpoint, improving the safety of people with SMI and building resilience factors for violence must be considered crucial aims for public mental health (Gureje, 2015). Evidence, strategies and funding for these targets appear lacking and require urgent academic and quality improvement efforts.

Conclusion

Violence is a part of human life, but disproportionately affects those with mental illness; however, research and policy efforts to address this are lacking. Research could gather greater information on the nature and extent of violent experiences, with attention to psychological and biological correlates. From a policy perspective, settings where better collection of data and reporting of violence is necessary need to be identified and prioritized.