Search tips
Search criteria 


Logo of bmjLink to Publisher's site
BMJ. 2002 September 7; 325(7363): 507–508.
PMCID: PMC1124043

Violence in society

Contribution of mental illness is low
Elizabeth Walsh, clinical lecturer
Thomas Fahy, professor

As increasing numbers of mentally ill patients have been treated and reside in the community, public concern about their potential for violence has increased. Fear and stigma of mentally ill people have been exaggerated by high profile and occasionally sensationalist reporting of rare, albeit tragic, violent acts.1

Are people with mental illness more violent than other people? An influential German study published in 1973 led to the belief that people with mental disorder were no more likely to be violent than the general population.2 This view remained unchallenged until the late 1980s. The best epidemiological data on violence and mental disorder come from the American ECA (epidemiologic catchment area) study.3 Self reported violence in the past year was measured among a representative community sample of 10 059 individuals. The prevalence of violence in people with no psychiatric disorder was 2%, and it was much higher in young men. Violence was reported in 8% of people with schizophrenia. People with alcohol (24%) or drug misuse or dependence disorders (34%) presented the highest risk. This study clearly shows that the increased risk of violence associated with mental disorder is limited to few, with only 10% of people with a mental disorder (in its broadest sense) admitting to assault in the previous year. Similarly, Australian men with schizophrenia have been shown to be four times more likely than the general population to be convicted for serious violence. To set these figures in perspective, however, in any given year only 0.2% of patients with schizophrenia received such a conviction.4 Both the American and Australian studies show the risk of violence in patients with psychosis and coexisting substance misuse to be considerably increased. Our interpretation of these and other impressive epidemiological studies is that patients with psychotic illness alone have a modest increase in risk for violent behaviour,5,6 but the greatest risk is associated with personality disorder,7,8 substance misuse,8 and in comorbid conditions where substance abuse is combined with severe mental illness.3,4

What proportion of societal violence is attributable to mental disorder? The answer to this question will vary according to the overall community rates of violence. In the United States, 16% of men aged 18-24 years and from low socioeconomic classes were violent in the ECA study, which presents a far greater risk than all people with schizophrenia in the sample. Variables such as male sex, young age, and lower socioeconomic status contribute a much higher proportion to societal violence than the modest amount attributable to mental illness.

Studies of unselected birth cohorts and epidemiological studies in the community allow us to consider the important public health issue of population attributable risk—the percentage of violence that can be ascribed to mental disorders in the population. If a person with mental illness is violent, however, it does not necessarily mean that this is due to the illness; it may be due to other variables that may be contributing to the increased risk of violence. This point is best illustrated by the findings of a study of a Danish birth cohort followed to age 44, which found that 7% of lifetime arrests in male participants for violence were attributable to psychotic disorder.5 Five per cent had coexisting substance misuse, however, meaning that only 2% of all arrests were attributable to a psychotic disorder alone, and it is plausible that some of these were attributable to other coexisting risk variables.

Overall, it seems that less than 10% of serious violence, including homicide, is attributable to psychosis.9 Additionally, strangers constitute only a small minority of the victims of violence committed by those with psychosis.10 The greater importance of personality disorder and substance misuse is highlighted by findings from the National Confidential Inquiry into Homicide and Suicide, which found that a third of all homicide offenders in the United Kingdom had a lifetime diagnosis of a mental disorder, the most common being personality disorder and substance misuse, and only 5% had schizophrenia.10

Could the closing of large psychiatric institutions over the last 30 years have meant that a greater proportion of societal violence is attributable to thosewith mental disorder? The evidence contradicts this seductive hypothesis. The contribution of mental disorder to homicide statistics in the United Kingdom seems to be falling rather than increasing.11 In Victoria County, Australia, violent acts (including homicide) committed by people with schizophrenia have risen since the shift to community care, but it has risen only to the same extent as in the general population.12 Even among patients who have already been seriously violent, reconviction rates have fallen over the past 20 years.13

Many health workers will encounter victims of violence in their day to day clinical work and will not need to be reminded of the impact of violence on their patients' wellbeing. It will be equally obvious to them that most of their patients are not at increased risk of violence compared with the readers of tabloid newspapers, members of parliament, mental health professionals, and other sections of the general population. The scientific literature supports these observations and refutes the stereotyping of all patients with severe mental illness as dangerous. In many mental health assessments it is appropriate to estimate the risk of violence to others as one of many dimensions of a comprehensive assessment. But it is inappropriate that mental health policy and legislation should be driven by preoccupation with risk of violence, rather than the delivery of effective treatments in the community.14


1. Ferriman A. The stigma of schizophrenia. BMJ. 2000;320:522. [PMC free article] [PubMed]
2. Hafner H, Boker W. Mentally disordered violent offenders. Soc Psychiatry. 1973;8:220–229.
3. Swanson JW, Holzer ChE, Ganju VK, Jono RT. Violence and psychiatric disorder in the community: evidence from the epidemiologic catchment area surveys. Hosp and Comm Psychiatry. 1990;41:761–770. [PubMed]
4. Wallace C, Mullen P, Burgess P, Palmer S, Ruschena D, Browne C. Serious criminal offending and mental disorder. Case linkage study. Br J Psychiatry. 1998;174:477–484. [PubMed]
5. Brennan PA, Mednick SA, Hodgins S. Major mental disorders and criminal violence in a Danish birth cohort. Arch Gen Psychiatry. 2000;57:494–500. [PubMed]
6. Hodgins S. Mental disorder, intellectual deficiency, and crime. Evidence from a birth cohort. Arch Gen Psychiatry. 1992;49:476–483. [PubMed]
7. Steadman H, Mulvey EP, Monahan J, Robbins PC, Applebaum PS, Grisso T, et al. Violence by people discharged from acute psychiatric inpatient facilities and by others in the same neighborhoods. Arch Gen Psychiatry. 1998;55:1–9. [PubMed]
8. Stueve A, Link BG. Violence and psychiatric disorders: results from an epidemiological study of young adults in Israel. Psychiatry Q. 1997;68:327–342. [PubMed]
9. Walsh E, Buchanan A, Fahy T. Violence and schizophrenia: examining the evidence. Br J Psychiatry. 2002;180:490–495. [PubMed]
10. Department of Health. Safety First. Five year report of the national confidential inquiry. London: Stationery Office; 2002. , May 2001.
11. Taylor PJ, Gunn J. Homicides by people with mental illness: myth and reality. Br J Psychiatry. 1999;174:9–14. [PubMed]
12. Mullen P, Burgess P, Wallace C, Palmer S, Ruschena D. Community care and criminal offending in schizophrenia. Lancet. 2000;355:614–617. [PubMed]
13. Buchanan A. Criminal conviction after discharge from special (high security) hospital. Incidence in the first 10 years. Br J Psychiatry. 1998;172:472–476. [PubMed]
14. Department of Health. Draft mental health bill. London: Stationery Office; 2002.

Articles from BMJ : British Medical Journal are provided here courtesy of BMJ Group