Mental Illness and Violent Crime. Mental Illness and Violent Crime Series: NIJ Research Preview Published: October 1996 6 pages 8,898 bytes U.S. Department of Justice Office of Justice Programs National Institute of Justice Jeremy Travis, Director October 1996 National Institute of Justice Research Preview Mental Illness and Violent Crime A Summary of a Presentation by John Monahan, Ph.D., University of Virginia School of Law Throughout history, most societies have strongly believed that mental disorder is linked with violence. The earliest recorded account is found in a dialog written by Plato. William Shakespeare wrote of such a connection in The Taming of the Shrew and Henry IV. These perceptions have had important implications for public policy, such as involuntary hospitalization of individuals with mental disorders, and for people's general interaction with those hospitalized with mental illnesses. The primary issue can be defined by two questions: o Is there a fundamental link between mental disorder and violence? o If so, can people with mental disorders who will be violent be distinguished from those who will not be violent? Until 6 years ago, research on the connection between mental disorders and violent behavior was conducted only on institutionalized populations. Recent research that has also looked at the general population has found a statistically significant relationship between mental disorder and violence; but in absolute terms, the relationship is modest. One study found that 3 percent of the variance in violent behavior in the United States is attributable to mental disorder, and other studies have shown that people with mental illness are more likely to be victims than perpetrators of violence. Most significantly, the link of mental disorders to violent behavior is not based on a diagnosis of mental illness but on current psychotic symptoms. Research to date indicates static predictions of an individual's danger to others are not realistic. Recent epidemiological studies have instead sought to identify those psychotic symptoms and related factors that could predict the occurrence of violent behavior. One major study has linked symptoms of hostility and delusions, especially when combined with substance abuse, with a high probability of violent behavior. On the other hand, demographic factors, such as age and gender, appear to be more important for predicting violence among the general population than among those with mental disorders. Current risk assessment research The Risk Assessment study, sponsored by the Research Network on Mental Health and the Law of the MacArthur Foundation and the National Institute of Mental Health, is examining 1,000 mental patients released from acute care facilities in Massachusetts, Missouri, and Pennsylvania. They are male and female, between 18 and 40 years old, who have all types of diagnosed mental disorders except mental retardation. Data have been collected through interviews with the patients following their release and with a collateral individual (usually a family member), as well as through reviews of police and hospital records. Findings are expected to be available by mid-1997. The same research team has also conducted a second study on 500 Pittsburgh residents, which will determine violence rates of the general population and assess whether risk factors associated with violence are the same as or different from the patient sample. The researchers postulate that in a public health framework, the risk factors for violence can be classified into four categories: o Personal/dispositional factors: e.g., age, gender, ethnicity, control of anger, impulsiveness. o Developmental/historical factors: e.g., history of child abuse, work history, history of violence, hospitalization history for mental disorder. o Contextual factors: e.g., environmental stress, social support, weapons accessibility. o Clinical factors: e.g., delusions, hallucinations, substance abuse. Building on previous research In developing their approach, the researchers have built on the methods and findings of two pioneer studies made available in the 1990s that refocused the issue of the link between mental disorder and violent behavior to specific symptoms and factors. They also took into account a third study that explored clinicians' ability to predict violence among their mentally ill patients. Epidemiological Catchment Area (ECA) study. Jeffrey Swanson and colleagues at the University of Texas Medical School reanalyzed data collected for the National Institute of Mental Health's ECA study from the early 1980s. The study had involved interviews with 10,000 randomly chosen adults from five cities to determine the epidemiology of various psychiatric disorders. Swanson's team identified questions used to determine antisocial personalities and examined how responses to them related to mental disorders. Their analysis found that people with major mental disorders such as schizophrenia had statistically significant higher rates of violence than those who did not (11-13 percent -- the percentage varied depending on the particular diagnosed illness -- compared to 2 percent). However, the diagnosed illness most associated with violence was substance abuse: 25 percent for alcohol abuse and 35 percent for drug abuse. Threat Control Override Symptoms study. Bruce Link and Ann Stueve at Columbia University compared people released from mental hospitals in New York City's Washington Heights area with those who had never had mental health treatment. This study also had an epidemiological focus and included violence- related questions. The study found that 15 percent of those who had never been hospitalized for a mental disorder and 26 percent of former mental patients had been in a fight in the previous 5 years. When controlling for variables like gender, age, or ethnicity, the relationship was weaker, but former patients were still more likely than nonpatients to have been violent. Only when the study controlled for current psychotic symptoms were both patients and nonpatients just as likely to commit violence. The researchers found that three symptoms -- feeling that others wished one harm, that one's mind was dominated by forces beyond one's control, and that others' thoughts were being put into one's head -- could be associated with violent behavior. These were termed Threat Control Override Symptoms. Sixty percent of those who scored highest on measures of these symptoms got into fights, regardless of whether or not they had been hospitalized. Predicting violence. Charles Lidz, Edward Mulvey, and William Gardner at the University of Pittsburgh Medical Center obtained clinicians' predictions of the likelihood of violence among hundreds of patients from an acute hospital who returned to the community. The clinicians were moderately accurate at predicting violence among male patients but no better than chance at predicting violence among female patients. Clinical staff predicted that 45 percent of the males and 22 percent of the females would become violent. The data showed that 42 percent of males and 49 percent of females behaved violently within 6 months after release. Reasons that clinicians inaccurately predicted violence among female patients may be because men in the general population are 10 times more likely to be arrested for a violent crime than women, because women are usually not violent in public places, and because women are less likely than men to seriously injure their victims. Implications Findings of the MacArthur Risk Assessment study are expected to shed light on whether and to what extent the risk factors associated with violence among those with mental disorders are also associated with violence in the general population. ------------------------------------ This Research Preview is based on a presentation by John Monahan, Ph.D., professor of law and psychology and legal medicine at the University of Virginia School of Law. He is one of the researchers on the Risk Assessment study, conducted by the Research Network on Mental Health and the Law of the John D. and Catherine T. MacArthur Foundation. As part of NIJ's Research in Progress Seminar Series, Dr. Monahan discussed his study with an audience of researchers and criminal justice professionals and practitioners. A 60-minute VHS videotape, "Mental Illness and Violent Crime," is available for $19 ($24 in Canada and other countries). Please ask for NCJ 156925. ------------------------------------ Points of view in this document do not necessarily reflect the official position of the U.S. Department of Justice.