Schizophrenia violence

Media has a particular fascination with schizophrenia violence, and these cases are often thrust into the media limelight. Violent crime, on its own, is often falsely sensationalised in the tabloid and mainstream media as well as in Hollywood, and then spiced up with mental illness. This has resulted in a very skewed perception of mental illness by the public. A 2007 study in Australia found that the mentally disordered were portrayed as 10 times more likely to be a violent criminal than non mentally disordered television characters. A another survey demonstrated that as television viewing increased so did the belief among viewers that locating mental health services in residential neighbourhoods would endanger the residents. Viewers who watch television news were less likely to support living next to someone who was mentally ill. In the US in the NESARC study, 75% of people viewed people with severe mental illness as dangerous.

A number of studies have been released recently that do show an association between schizophrenia and violent crime, but we need to have a closer look at these results.

The National Epidemiologic Survey on Alcohol and Related Conditions (NESARC), was undertaken between 2001 and 2005 and published in 2009. It had 34,600 respondents and examined the relationship between severe mental illness and violence. It found that severe mental illness was associated with increased rates of violence (assault, sexual assault, arson) but this was only significantly so, where there was co-occurring substance abuse/dependence. Violence was better predicted by historical factors such as juvenile detention, physical abuse, past violence and parental arrest. Clinical risk factors for violence included substance abuse and perception by patients that they were under threat. Other factors for violence included age, sex, income, recent divorce, and unemployment.

The NESARC study in summary found that the 8 strongest predictors of violence, in order of strength of prediction, were younger age, history of a violent act, male sex, history of juvenile detention, divorce or separation in the past year, history of physical abuse, parental criminal history, and unemployment for the past year. Co-occurring mental illness and substance use was ninth on the list, followed by victimization in the past year.

Thus a person with a severe mental illness without substance abuse or history of violence, had the same chance of being violent as anyone in the general population.

Another review published in 2009 involved 20 studies across 11 countries and more than 18,400 patients. It found that compared to the general population, the risk for violent outcomes was increased in schizophrenia populations. The main findings were that the risk for violence was increased in individuals with schizophrenia and other psychoses. Secondly, that comorbid substance abuse dramatically increased this risk, on average about 4 times higher compared to individuals without substance abuse. And a third finding was that the risk for violence was the same in individuals abusing substances compared to individuals with schizophrenia abusing substances. In other words, from this, having schizophrenia did not increase the risk of violent outcomes.

One further findig was that the risk of homocide was very small at 0.3%, the same as that for substance abusers.

A Swedish study involving 8000 patients with schizophrenia, looked at the risk of violent crime after the diagnosis, and compared it to the general population. Patients with schizophrenia were found to have higher rates of violent offences, but this risk was mostly confined to patients with comorbid substance abuse. The risk increase was small in schizophrenia patients without substance abuse. Evidence seems to be converging that there although schizophrenia is associated with violence, this association is multifactorial. Besides substance abuse disorder, there are a range of other factors. A history of longstanding antisocial behaviour and attitude, antisocial lifestyle and peers, poor compliance, poor parental models and domestic criminal influences and chaotic social circumstances.

Other studies have found similar predictors for reoffending comparing mentally disordered offenders to non mentally disordered offenders. The major predictors were criminal history. Clinical factors such as psychotic symptoms, had the smallest strength as a predictor.

What is the take home message with schizophrenia violence? What's emerging from all this research is that most people with schizophrenia are not violent, but some are. Overall the incidence of violence is low and serious mental illness increases the risk 3 times. In comparison having an alcohol and drug use disorder increases the risk 9 times. Having both serious mental illness and an alcohol and drug use disorder increased the risk 13 times.

Fazel et al, Plos 2009,6(8)
Fazel et al JAMA. 2009;301(19):2016-2023
Elbogen et al Arch Gen Psych 2009;66(2):152-161
Swanson JAMA 2010;304(5):563-4.
Mullen Adv Psych Treat. 2006(12):239-248