A critical evaluation of whether mental illness has an inherent link to violent behaviour This essay will aim to explore whether mental illness has an inherent link to violent behaviour. Specifically it will critically evaluate the literature surrounding this contention. A definition of both mental illness and Violence will be offered before outlining the conflicting understanding regarding the inherent link. The essay will conclude that the issue of an inherent link between mental illness and violence is a complex one.
That when controlling for substance use and other factors such as gender, age, race/ethnicity, individual and neighbourhood socio-economic status (SES), physical and sexual abuse, stressful life events, impaired social support the influence of mental illness upon violent conduct is minimal. However, many of these factors influence both mental illness and violence irrespective of the presence of both, making it difficult to tease apart the contribution of any factor in the resultant expression of violent behaviour. In addition, offence and victim characteristics appear to be different for mentally ill offenders than non-mentally ill offenders.
For example where violence takes place in those with active psychotic symptoms the likelihood is that they will offend against family and friends rather than the public at large. This directly contradicts the unhelpful public perception that there is an increased risk of general violence by those experiencing mental illness. Silver et al (2008) add to this by suggesting that the violent acts committed by the mentally ill are greater in severity than those committed by non-mentally ill individuals. Violence has been defined by Glasser (1998) as ‘a bodily response with the intended infliction of bodily harm on another person’.
Mental Illness/disorder includes those ‘major’ disorders of affect and thought which form a subgroup of Axis I disorders in the Diagnostic and Statistical manual of mental disorder (4th Edition: DSM-IV) (American Psychiatric Association 1994) The inclusion of Personality Disorder, (Psychopathic Disorder Classification) traditionally caused controversy, however as it’s treatment and management has been included in the recent revision of the Mental Health Act (1983) it will be included for discussion here and difficulties in its inclusion in research on mental illness and violence will be highlighted.
Research into the link between Mental Illness and Violence began in 1990 (Stuart 2003). Initially, it was thought that those with mental illness were no more likely to be violent than those without mental illness (Phelan & Link 1998) However, Taylor & Gunn (1999) analysed the records of 1241 men remanded in Briton prison and found that in the group charged with homicide just over a third were psychiatrically ‘abnormal’. 11% of these were Schizophrenic and patients who had been violent towards others without resulting fatality exceeded the researcher’s expectations by 22 times.
However, the relationship between mental illness and violence is not clear; there are a variety of factors influencing these data, for example which types of disorders and differential degrees of violence and shared aetiological factors between mental illness and violent conduct. Indeed Silver et al (2008) state that ‘the main problem in studying the relationship between mental illness and violence is the strong possibility that the relationship is spurious, that many of the life circumstances and experiences that affect the likelihood of violence also affect the likelihood of mental illness. ’
To demonstrate this point it is useful to acknowledge the observations of Swanson et al (1990) they conducted a survey to assess the violence rate in people with and without mental illness. They found that major mental disorders correlated with at least a fivefold increase in rates of violence over a 1 year period, compared with rates among individuals with no disorder. Despite this finding it is interesting to note the possible influence of certain diagnoses, such as Conduct Disorder and Antisocial Personality Disorder (APD) which have a direct association with crime and violence (Singh 2007).
Violent acts are part of their diagnostic criteria and APD comprises between 33% – 80% of the population of chronic criminal offenders (Hare 1980). This raises the important question of whether such diagnoses are confounding studies aiming to assess the link between violence and Axis I disorders. When focussing on Axis I disorders such as those ‘major’ disorders of affect and thought Link & Steuve (1994) found the presence of certain psychotic symptoms such as thinking others are intending to do them harm, thoughts that others could control their thoughts, and thers could put thoughts into their head – characterised by the authors as TCO (threat/control/override) are good predictors of violence in psychotic patients. Swanson et al, (1996) replicated these findings and found that respondents with TCO symptoms were twice as likely to indulge in violent acts compared with patients who had other psychotic symptoms. Swanson et al (2006) have gained further evidence of this influence. In their study of 1410 schizophrenic patients they found a reported 6 month prevalence of simple violence (15. 5%) with 3. 6% showing serious violent behaviour.
They concluded that positive psychotic symptoms like persecutory ideation, and history of childhood conduct problems, and victimisation increased the risk of violence. This again demonstrates the varying influence of discreet characteristics of symptomatology experiences in influencing the risk of violent conduct, all mental illness cannot be judged equally in their influence upon the expression of violent behaviour. In addition the variety of influencing factors, such as childhood experience and victimisation also adds to the debate as they are identified risk factors for violence alone.
However these findings could be viewed differently when observing the findings by Appelbaum et al (2000) who found that if symptoms of substance abuse are excluded, no significant difference exists in the prevalence of violence between discharged psychiatric patients and community controls. Interestingly, Steadman et al (2000) report no significant relation between psychosis (with TCO symptoms) and violence which again, leaves the link between mental illness and violence unclear.
Similar to Appelbaum et al, Swanson et al (1990) reported substance abuse as a more significant variable for risk of violence than any major mental disorders and Fulwiler et al (1997) also found that substance abuse either alone or in combination with mental disorder significantly increased the chances of violent behaviour. The MacArthur study also found no increased in rate of violence in mentally ill patients compared to the general population if they were not using illicit drugs or alcohol.
And again, Wallace et al (2004) looked at the conviction rate of patients with schizophrenia for over 25 years and compared it with a community sample, concluding that substance abuse in schizophrenic patients increases their risk of criminal conviction by sixteen fold. Therefore it appears that substance use has a major impact upon violence, whether those with mental illness are prone to use substances, therefore biasing data is unclear, or whether the substance use impacts on the development of mental illnesses and then to violent behaviour.
Whichever of these possibilities is most accurate is beside the point. One major point is that historically, society has perceived people with mental illness as being more violent and dangerous than those without metal illness. This has always been a part of the stigma associated with the mentally ill and indeed, has lead to the more concerning aspect of victimisation of the mentally ill and their propensity to be the victims, rather than perpetrators of violence (Marley & Bulia 2001).
The phenomenon of deinstitutionalisation has increased the rate of homelessness in psychiatric patients which in turn has made people with mental illness more vulnerable to victimisation. Silver (2002) conducted a study on 270 acute inpatient psychiatric patients and found that 15. 2% of them had been physically assaulted, forced to have sex, threatened or attacked with a weapon within the preceding 10 weeks. Additionally, Brekke et al (2001) found that 38% of their sample of Schizophrenic patients had been victimised in the last 3 years and that in 91% of those cases the incident was violent in nature.
They found that the incidents of violent crimes in mentally ill patients to be four times higher than national crime victimisation survey rates. So far the link between mental illness and violence remains unclear, it is understood that some psychotic symptoms increase the propensity for violence and that where substance use is apparent this risk increases dramatically; in addition, the extent to which the mentally ill are the victims of violence cannot be underestimated.
Where individuals with mental illness have committed violent offences 86% occurred within the context of family and friends (Steadman et al, 2000) to highlight the minimal risk of violence by those with mental illness they point out that members of the Pittsburgh general (non patient) public who were violent were more likely to target strangers (22%) than were Pittsburgh’s patient population (11%). Monahan (1993) and Rabkin (1972) both state that irrespective of the presence or non presence of mental illness it s those with a criminal history prior to the index offence that present the greatest risk for recividism of violent offending. As outlined above understanding a link between mental illness and violence can be contradictory and also the possible factors influencing this link are wide and varied. Despite the vast body of research into the area, the association between mental illness and violence remains inconclusive, this is thought to be because of limitations of the research.
Cross-sectional epidemiological studies tend to analyse past violence and current or lifetime psychiatric diagnoses, tending to mean that the question of whether mental illness causes violence, that it precedes violence becomes unclear as data before that time is ignored. Also, longitudinal studies rely on participants who are already in institutional settings, meaning the sample isn’t representative of the general population and the extent of the influence of mental illness on violence is not captured.
Silver et al (2008) have attempted to overcome such difficulties, in addition to taking into account a wide number of control variable to focus their investigation. They claim that past research has tended to focus on a small number of the risk factors shared by both mental illness and violence. They have attempted to include all known variables such as, gender, age, race, SES, prior victimisation, alcohol/drug use and prior violent behaviour in order to tease apart the individual impact of mental illness upon violent conduct.
In addition to this main aim of the research Silver et al (2008) hypothesised that inmates with mental health problems commit violent crimes that are more deviant or anti-normative than those commited by inmates without mental health problems. They suggested a ‘Deviant Hypothesis’ that is ‘offenders with mental health problems are more likely to engage in violent behaviour that is either unusual or particularly serious or both’. They posed that mental illness is not as likely to be a factor in “ordinary crime”.
This follows on from the augmentation principle in attribution theory (Corrigan et al 2003) according to the augmentation principle, when a behaviour occurs despite strong inhibiting external forces, observers tend to assume that a strong internal force (mental illness) is responsible (Kelley, 1973) In this way, silver et al have gone beyond asking whether mental illness predicts violence but seek to find which specific types of iolence mental illness becomes a risk factor for – more serious offences such as homicide rather than physical assault. The found that among incarcerated offenders, mental health problems were more strongly associated with assaultative violence and sexual offences than other types of crimes, in addition, they found that mental illness played a greater role in homicide than in physical assault, supporting the deviance hypothesis, although they do state that the effect is relatively weak.
They concluded that mental illness is a causal factor in deviant behaviour, some of which involves violence, and that the more deviant the behaviour the greater the effect of mental illness. This somewhat addresses the question of a link between mental illness and violence, but still leaves questions to be answered concerning the specific influence of mental illness on violent conduct. In conclusion, the majority of mentally ill individuals who are receiving appropriate treatment, do not carry more risk for violence then the general population.
However, it is understood that major determinants of violence, such as socio-demographic and socio-economic factors, when combined with the presence of mental illness increase the risk of violence. Early identification and treatment of substance use problems, especially if it occurs concurrent with mental illness symptomatology can significantly reduce the risk of violence. In addition, Past history of violence is crucial in predicting potential violence in the mentally ill.