Mental illness and violence: Debunking myths, addressing realities

Many health service psychologists will, at some point in their careers, evaluate, treat, or study the relatively small number of people with serious mental illness who have committed or have the potential to commit violence toward others. Most often they see these individuals in psychiatric inpatient or forensic settings, but occasionally in private practice as well. Many more psychologists have also treated clients who have contemplated or even completed suicide, considered by some to be violence against the self.

To aid psychologists’ understanding of this challenging phenomenon, there is a growing body of research that is helping to tease apart why some people with serious mental illness are prone to aggression or violence while others are not, and how clinicians and others can help.

While perpetrating violence is relatively uncommon among those with serious mental illness, when it does occur, in many cases it is intertwined with other issues such as co-occurring substance use, adverse childhood experiences, and environmental factors, says Eric B. Elbogen, PhD, a psychologist and professor of psychiatry and behavioral science at the Duke University School of Medicine who studies violence and mental illness.

“If a person has a severe mental illness, [they] may have other risk factors for violent behavior,” he says. “So, it may not be mental illness that is driving the violence at all, but rather factors like having been abused as a child, being unemployed, or living in a high-crime neighborhood.”

It is important to learn about these issues not only to better treat these individuals and to aid their families and communities but to combat the misperception that most people with serious mental illness are violent, adds Jeffrey Swanson, PhD, a medical sociologist at the Duke University School of Medicine and a prominent researcher of the topic. For example, people often believe that people with mental illness are largely responsible for incidents of mass violence and that people with mental illness are responsible for a large share of community violence. Yet both views have been roundly debunked by research, says Swanson.

“We need to do some serious myth-busting around these ideas,” he says, “because people believe them, and they have real consequences.”

Overall, people with serious mental illness—which generally refers to those with major depressive disorder, bipolar disorder, schizophrenia, and schizoaffective disorder—are somewhat more likely than members of the general public to commit acts of violence, research shows. A study by Richard A. Van Dorn, PhD, of RTI International, and colleagues, for example, found that in a nationally representative community sample of 34,653 people from the National Epidemiologic Survey on Alcohol and Related Conditions, 2.9% of people with serious mental illness had committed violent acts between 2 and 4 years following the study’s baseline, compared with 0.8% of people with no serious mental illness or substance use disorder. However, 10% of people with both serious mental illness and substance use disorder committed such acts during that time (Social Psychiatry and Psychiatric Epidemiologyopens in new window, Vol. 47, No. 3, 2012).

Such research may seem to support the notion that people with mental illness pose a danger to society. But a closer look reveals that the situation is more complex, says Elbogen. In the most general sense, “there is a link between mental health and violence,” he says, “but at the same time, most people with severe mental illness are not violent.”

As important, a growing body of research shows that when people with serious mental illness commit violent or aggressive acts, other factors besides the illness itself are often at play, says Kimberly Brown, PhD, ABPP, an associate professor of clinical psychiatry and behavioral sciences at Vanderbilt University Medical Center and host of an APA 2020 convention workshop on the topic.

A big factor is co-occurring substance use. “If you have both a mental illness and a substance use diagnosis, the combination is synergistic and dangerous,” Brown says.

Other contextual factors likewise play a part in why people with mental illness may turn to violence or aggression, research finds. In the MacArthur Violence Risk Assessment Studyopens in new window—one of the most rigorous and widely cited studies on the topic to date—only two clinical symptoms were associated with violent acts among psychiatric inpatients 20 weeks after discharge: “command hallucinations,” or psychotic voices telling a person to harm others; and psychopathy (characterized by a lack of empathy, poor impulse control, and antisocial deviance), which is not typically considered a serious mental illness. Just as likely to play roles were a history of prior violence, a history of childhood physical abuse, having a father who abused substances or was a criminal, displaying antisocial behavior, and scoring high on anger measures.

One of the most striking findings from the original MacArthur Violence Risk Assessment Study is an environmental one: When the team compared discharged psychiatric patients without substance use disorder with people from their same neighborhoods, their rates of violence were about the same, says Paul Appelbaum, MD, Elizabeth K. Dollard Professor of Psychiatry, Medicine, and Law at Columbia University and a site principal investigator on the MacArthur Study. In other words, when neighborhoods are unsafe, poor, and high in crime, violence is an equally likely outcome whether a person has a mental illness or not.

In short, says Appelbaum, “a great deal of what is responsible for violence among people with mental illness may be the same factors that are responsible for violence among people without mental illness.”

Meanwhile, Elbogen and colleagues have been examining possible internal and situational forces that may drive people with serious mental illness to commit violence. Using data from the National Epidemiologic Survey on Alcohol and Related Conditions, they examined the potential role of three main domains of risk factors as described in the “I³ model” of aggression: “dispositional impellance,” or internal factors such as anger and perceived threats that may impel someone to commit violence; “situational impellance,” or contextual factors such as divorce or separation, financial problems, or victimization that may foster violence; and “disinhibition,” external factors such as drug or alcohol intoxication that make it easier for a person to commit violent acts. These factors mitigated the relationship between violence and mental illness; in fact, participants with serious mental illness were less likely to commit violent acts than those without serious mental illness once those factors were removed (Clinical Psychological Scienceopens in new window, Vol. 4, No. 5, 2016).

Despite significant advances in knowledge, however, research in the area is still mixed, says Brown. Some studies find larger links between violence and mental illness than others, and others find no relationship at all. Several reasons help to explain such differences, including that researchers use different population samples in their studies, as well as different definitions and variables to measure violence and mental illness. With violence, for instance, some studies include relatively minor acts of aggression, while others focus exclusively on serious acts of harm toward others. The same is true with mental illness, with some studies of schizophrenia and violence, for example, including only so-called positive symptoms like hallucinations, which are associated with high rates of violence. But other studies find that negative symptoms, such as apathy and loss of motivation, are associated with lower violence rates (Archives of General Psychiatryopens in new window, Vol. 64, No. 5, 2006).

The bottom line? “Diagnosis alone is never enough to tell you if someone is likely to be violent again in the future,” says Brown. Instead, a contextual approach is needed that considers symptoms, circumstances, and individual characteristics, among other factors, she says.

In an editorial in The American Journal of Psychiatryopens in new window (Vol. 176, No. 9, 2019), Appelbaum also points out that many of the current studies, reviews, and meta-analyses in the field rely on data originally collected for other studies, not for the specific purpose of studying the relationship between mental illness and violence. To help move the field forward, he suggests that researchers conduct more prospective studies that use violence as the primary outcome measure and have sufficient power to identify the range of risk-related variables and integrate them into sound theoretical models. Also helpful, he says, would be developing uniform research methods when studying the topic.

Sometimes, mental health symptoms in and of themselves can spur a person to aggression or violence. These include when the person suffers from:

Persecutory delusions and “command hallucinations.” For people with psychotic disorders such as schizophrenia, studies show that some of the conditions’ positive symptoms can provoke violence. These include persecutory delusions, such as when a patient thinks people are putting implants in their head or are targeting them with harmful laser beams, particularly when they are also feeling angry or irritable. Another such symptom is “command hallucinations”—when a patient reports hearing voices that order them to hurt someone.

Grandiosity, grandiose delusions, and mania. Grandiosity, a hallmark of the manic and hypomanic phases of bipolar disorder, can likewise play a role in violence and aggression, studies find. People can be overtaken by an exaggerated sense of their own power, which can stunt their ability to empathize with others and foster a sense of entitlement, including the right to take advantage of or exploit others, Brown explains.

Similarly, the high energy that often accompanies mania can lead to violence or aggression on its own terms. A study by Jillian K. Peterson, PhD, an associate professor at Hamline University in St. Paul, Minnesota, and colleagues, for instance, found that of 429 crimes committed by 143 offenders with mental illness, 3% related directly to depression, 4% to psychosis, and 10% to bipolar disorder, including impulsivity (Law and Human Behavioropens in new window, Vol. 38, No. 5, 2014).

Antisocial personality traits. Violence among people with serious mental illness often goes hand in hand with a youthful history of conduct disorder and a present diagnosis of antisocial personality disorder, characterized by disregard for others, deceitfulness, and manipulation of others for personal gain, research also shows. While these disorders are considered in the realm of personality disorders, it is likely that many more people would be placed in a category of people with mental illness and at high risk for violence if these disorders were included, Brown says.

Swanson and colleagues examined some of these factors in an analysis of data on 1,445 participants who took part in the NIMH Clinical Antipsychotic Trials of Intervention Effectiveness (CATIE) Study, conducted in the mid-2000s. People with schizophrenia, they found, were twice as likely to commit violent acts if they had a history of childhood conduct problems (28%) than if they did not (14%) (Law and Human Behavioropens in new window, Vol. 32, No. 3, 2008). And in a subset of the most violent patients in the MacArthur Study, University of California, Berkeley, psychology professor Jennifer Skeem, PhD, and colleagues found that only 12% had experienced psychotic symptoms before committing a violent act, while the subsample as a whole met criteria for antisocial personality disorder, had long histories of conduct problems, and had difficulty managing anger (Clinical Psychological Scienceopens in new window, Vol. 4, No. 1, 2016).

Clinicians need to consider such complexities when conducting treatment and assessments, say psychologists and others who work with patients with serious mental illness who may be at risk for violent behavior.

“Each patient needs to be assessed individually to identify factors that seem causally related to acts of violence that have occurred in the past, and to focus on intervening with those factors to reduce future risk,” says Appelbaum.

One basic intervention is making sure that people are following treatment protocols, says Shirley M. Glynn, PhD, a clinical and research psychologist at the Semel Institute for Neuroscience and Human Behavior at the University of California, Los Angeles.

“When I hear of somebody with a serious mental illness who has acted violently or aggressively, the first thing I ask [their clinician] is, ‘Where are they in terms of their treatment?’” she says. Often they have let treatment and medications lapse, “and if you can help them get back into treatment, a lot of their [aggressive or violent] behavior settles out.” Findings from the CATIE Study bear this out, showing that most patients with schizophrenia who took antipsychotics as prescribed were less likely to be violent than those who did not.

For people with mental illness who have been incarcerated for violent acts, good follow-up can prevent recidivism, adds clinical and forensic psychologist Joel Dvoskin, PhD, ABFP, an expert in mental health and criminal justice assessment for more than 4 decades who has served in numerous related leadership capacities for the states of New York, Nevada, Virginia, and Arizona.

“What you do after people are released matters,” says Dvoskin. He strongly encourages check-ins about how former inpatients or inmates are doing and whether they are following their medication regimens.

Also important is developing treatments that address family dynamics in relation to a patient’s violence, in part because research suggests that about 1 in 5 family members of people with serious mental illness is the target of such violence each year, says psychiatrist Lisa B. Dixon, MD, MPH, of the Columbia University Department of Psychiatry. In an article in a special issue of the Harvard Review of Psychiatry opens in new window (Vol. 29, No. 1, 2021) on violence and mental illness, Travis Labrum, PhD, of the University of Pittsburgh School of Social Work, Dixon, and colleagues reviewed 18 qualitative studies on factors correlated with violence toward family members, including a patient’s reliance on family members for financial and other support, limit-setting by family members, and the presence of criticism, hostility, and verbal aggression on both sides.

While research on family and caregiver interventions in this area is still limited, promising avenues include helping patients engage in recovery-oriented treatment and strengthening support services that could reduce reliance on family members, says Dixon. Also promising are evidence-based interventions that support the ability of family members to prevent and manage family conflict, including multifamily group treatment and the National Alliance on Mental Illness’s Family-to-Family program, a free, eight-session educational program for family, friends, and significant others of people with mental illness that guides caregivers in supporting their loved ones and themselves and handling crises when they arise. Meanwhile, strategies targeted to young people, like using internet-based treatments to help engage them, also hold promise.

Knowledge about medications is also improving, finds another review articleopens in new window in the 2021 Harvard Review of Psychiatry special issue by psychiatrists Leslie Citrome, MD, MPH, of New York Medical College in Valhalla, New York, and Jan Volavka, MD, PhD, a professor emeritus at NYU Langone Health. They find strong evidence that two common drugs given to people with schizophrenia, clozapine and olanzapine, can help reduce symptoms of hostility, but that other antipsychotic medications typically prescribed to people with schizophrenia are unlikely to make a difference.

Forensic assessment has likewise grown in sophistication, says Dvoskin. Currently, there are two types of assessment instruments: actuarial instruments, which calculate a person’s likelihood of committing future violence based on a decision-making tree, and structured professional judgment instruments, which use those same tools but give evaluators the ability to make more nuanced decisions based on a person’s particular circumstances.

While both methods are good predictors of future violence, Dvoskin prefers structured professional judgment because “it allows us to use clinical judgment to decide how serious a person’s risk is.” It also provides room to consider dynamic factors that may lead to violence, such as being laid off from work or experiencing domestic violence.

“I don’t think context is a get-out-of-jail-free card,” says Dvoskin, “but it’s certainly relevant to any risk assessment.”

Others are working to address environmental, situational, and contextual roots of violent or aggressive behavior in people with mental illness. Crisis intervention teamsopens in new window are an example: Now in hundreds of cities and towns around the country, these collaborations between police officers, mental health educators, and community advocates are helping resolve high-intensity police interactions in ways that minimize additional violence.

In another effort, Glynn and colleagues are conducting statewide trainings with mental health counselors designed to help people with first-time episodes of psychosis return to school or get and keep a job, which can prevent them from developing other kinds of problems.

In the policy arena, Swanson and others have been tackling the complicated issues of gun violence and mass violence as they relate to people with mental illness. Besides promulgating research showing that mass shootings account for less than 1% of firearm homicides and tend to be committed by those with issues besides diagnosable mental illnesses, Swanson and others are studying extreme risk protection order laws. In 19 states and the District of Columbia, these laws allow family members and police officers to seek a civil restraining order to temporarily remove guns from people who pose an imminent risk of harming themselves or others (Harvard Review of Psychiatryopens in new window, special issue, Vol. 29, No. 1, 2021).

When Swanson and colleagues evaluated the effects of one such policy in Connecticut, they estimated that for every 10 to 20 gun-removal actions, the policy prevented one suicide. The team also learned that most of these individuals were men and that the central concern was suicide threat (Law and Contemporary Problemsopens in new window, Vol. 80, No. 2, 2017).

The findings underscore the importance of paying at least as much attention to suicide risk as to violence committed toward others, Swanson adds, since research shows that 46% of those who die from suicide have a known mental illness, and 60% of deaths from firearms are suicides.

“It’s a big opportunity,” he says, “because if you can keep a gun out of somebody’s hand in a moment of despair, they’re likely to survive. And people who survive a suicide attempt are unlikely to die from suicide going forward.”