What is the Relationship between Violence and Mental Disorder?

Recently I got a letter asking: What is the relationship between mental disorder and violence? "Oh, boy," you think, "There's an easy one for you." After all, we are about to inaugurate a Centre for the Study of Violence and Mental Disorder with McMaster University. Members of the Research Department have published at least a hundred scholarly works on violence. Here's what our research tells us now.

What is Violence?

How would we know if a person was violent, whether or not he had a mental disorder? To think we could give a sensible answer to our original question, we have to be confident that we can measure violence. We need indirect measures, however, because violence rarely occurs where researchers can see it. One indirect measure is asking participants if they have done anything violent recently and checking with other people who know them. Also, of course, many studies use official police records. These seem like good measures of violence, but they might have problems.

When we speak about violence and mental disorder, we probably mean serious violence -- murder, sexual assault, causing bodily harm, and so on. Presently, we do not know whether the violence measured by self reports is the same kind of violence that shows up on police blotters. We do know that police records do not reflect all violent crimes.

Bias can be a problem too. Maybe mentally disordered people, though actually no more violent than others, are more likely to admit to violence. Perhaps mentally disordered persons are easier for the police to catch. If the authorities and citizens overall were less tolerant of violent behaviour by mentally disordered people, we could not necessarily trust any differences we found. If any of these things, called biases, were true, answering our original question would be much harder.

What is Mental Disorder?

Unfortunately, violence is part of the definition of some mental disorders (e.g., conduct disorder, manic episode, pedophilia, sexual sadism, and intermittent explosive disorder). To discover links between such disorders and violence is about as useful as "discovering" a link between stealing and dishonesty. To study links between violence and mental disorder, we must ignore such disorders or change the definitions so that violence does not count as a symptom or defining property. The standard catalogue of mental disorders, the Diagnostic and Statistical Manual of Mental Disorders (4th edition, often called DSM-IV) lists a bewildering array of mental disorders -- so many that it's impossible to know now how each might be associated with violence. Because DSM-IV is mostly not a scientific classification system, it is very difficult for research to address "mental disorder" as defined by it.

Some diagnoses can be considered in the question of violence and mental disorder, however. These include psychopathy (characterized by callous use of others, extreme selfishness, irresponsibility, early behaviour problems, and criminality, all closely related to antisocial personality disorder), substance abuse disorders, major mood disorders, and schizophrenia. These are the diagnoses most people mean when they ask about violence and mental disorder. The last two are often called psychoses, or major mental disorders. For the rest of the many psychiatric diagnoses listed in DSM-IV, the association with violence is moot.

Alright Already, Enough Qualification; What's the Bottom Line?

1. Psychopathy is strongly linked with violence. Violent offenders are more likely to be psychopaths and psychopaths are more likely to commit violent offenses. The childhood precursors of psychopathy are strongly predictive of a lifelong pattern of persistent violent behaviour.
2. Substance abuse, especially alcohol abuse, is related to violence. Alcohol induced violent conduct and drunk driving are a greater risk to public safety than psychotic disorders. Thus, as far as violence is concerned, psychopathy and substance abuse are much more serious than the "major mental disorders."
3. Apart from these first two points, psychiatric diagnosis tells us nothing about how violent a person might be. After psychopathy perhaps, the best single way to know about someone's likelihood of violent behaviour is to know about past violence. This is just as true for people who have psychotic disorders as those who do not.
4. Among forensic patients, released prisoners, and discharged mental patients, a psychotic diagnosis does not raise the risk of violence. Our research (and that of others) shows that, other things being equal, schizophrenic offenders are less dangerous than other offenders.
5. Among people suffering from psychotic disorders who have a history of violence, some symptoms might show when violence will happen. These symptoms are feeling threatened combined with feeling under the control of outside forces. Increased hostility has also been reported to indicate increased risk of violence. These seem like sensible findings, but it is unclear how significant the increased risk is. It seems sensible too that treatment to alleviate these symptoms would lower risk, but there is no research on that. There is almost no evidence that any other psychotic symptoms are associated with violence.
6. In the population at large, having those few symptoms mentioned above increases the risk of violence a little. That increased risk is much smaller, however, than the risk associated with alcohol abuse, for example.

Conclusions and Comments

The public perception, promoted by some movies, of mental patients as raving, homicidal lunatics is false. Psychopaths and persons who abuse alcohol are a clear risk to the safety of others. Some mental patients (and some people who have never been mental patients) have a history of violence, and it is the history of violence that should guide our thinking about their dangerousness. Although they are psychiatric diagnoses, many mental health professionals do not regard psychopathy and substance abuse as mental disorders. Our research shows that forensic clinicians sometimes, therefore, mistakenly neglect the real dangerousness of psychopaths because they exhibit few psychiatric symptoms. Instead, secure beds and intensive supervision are reserved for patients who are actually less dangerous. Civil commitment probably contributes little to public safety because it is used only with persons who have psychotic disorders. It might enhance public safety more if it were used to lower the risks posed by psychopaths and people who abuse alcohol.

Finally, our recent findings and theories also lead us to the idea that psychopathy is not, in fact, a disorder. That is, although psychopaths are very different from everyone else in their emotions and conduct, these differences do not reflect a malfunction or defect. According to our theory, psychopaths are not damaged -- they function as designed by nature. It's just that nature has designed them differently. If we are right about this, psychopaths may not require therapy in the sense that mental health professionals usually think of it. Our research shows that some psychotherapy makes psychopaths more dangerous. Of course, whether psychopathy is a disorder or not, forensic clinicians still must deal with psychopaths (and this would be even more common if civil commitment were applied to psychopathy). In dealing with psychopaths, clinicians should never take at face value clients' reports of feelings, intentions, and behaviour -- such reports should be independently checked out. No matter how sincere the patient appears, first-hand monitoring is essential. Clinicians should ensure that sessions concentrate on behaviour, and not feelings, insight, or building self esteem.