Being objective about violence

A recent study of mental illness and crime in Sweden seems to be causing some confusion. The study results are being cited by what appear to be opposing views in the debate about the link between mental illness and violence. It is nonetheless a very informative study that can help all of us better understand this issue if we look at the data objectively.

One the one hand, the researchers point out that many studies have focused on the relative risk of violence in mental illness – the risk of violence for a person with mental illness as compared to the risk of violence for a non-mentally ill person. They note that focusing on relative risk “gives an incomplete picture” because it does not reflect the “proportion of violent crimes that can be attributed” to the mentally ill. In other words, because the mentally ill comprise such a small percentage of the population (1.4% in Sweden during the study period), their contribution to overall violent crime is relatively small. The researchers calculated a “population-attributable risk fraction” of 5.2% meaning that the mentally ill were responsible for only 1 in 20 violent crimes in Sweden. A simpler way of looking at it, which results in a slightly higher number, is that there were 21,119 crimes committed by mentally ill out of a total of 324,383 violent crimes in Sweden during the study period, meaning 6.5% of crimes were committed by people with mental illnesses. Thus, it is correct to try to defuse the stigmatization of people with severe mental illnesses by saying that they are responsible for a small fraction of violent crime. It is also true that only a small percentage of the mentally ill are violent. The study found that only 6.6% of the mentally ill (6,510 out of 98,082) were convicted of a violent crime.

On the other hand, the researchers found that patients with severe mental illnesses were nearly 4 times more likely to have committed at least one violent crime as compared to the general public (6.6% of mentally ill patients had a violence conviction compared with 1.8% of the general population). The study confirms findings in the U.S. that the relative risk of violence is higher for severely mentally ill patients (specifically those with diagnoses involving psychoses) than for the general public. The CATIE violence study found that patients with schizophrenia were 10 times more likely to engage in violent behavior than the general public (19.1% vs. 2% in the general population). In other words, it is more likely that a person with mental illness will be violent than a person in the general public. It is also interesting to note that patients with schizophrenia were nearly twice as likely to be violent (the number of violent crimes committed was 328 per 1,000 patients with schizophrenia and 173 per 1,000 patients with other psychoses).

Because there is a greater risk that people with mental illness may be violent, it is important for families and caregivers to recognize and have a better understanding of what factors contribute to this risk. For example, a recent CATIE study (see Figure pg. 496) reported that schizophrenia patients experiencing certain positive symptoms (hallucinations, paranoid delusions, and grandiosity) were 9 times more likely to have an episode of serious violent behavior than patients who had negative symptoms (apathy, social withdrawal, poverty of thoughts, blunting of emotions, slowness of movement, and lack of drive). This information is also important in terms of public policy because it supports the need for more timely and effective emergency psychiatric interventions when the most severely mentally ill become symptomatic.

Additional data from the study illustrates why it is important to understand the link between violence and mental illness. In a separate analysis by type of offense, researchers found that patients with severe mental illnesses, who accounted for only 1.4% of the general population, were responsible for 18% of homicides and attempted homicides in Sweden. That is not a statistic that can be ignored.