It’s time we put this myth to rest — violence and mental illness share about as much in common as violence and people who happen to be men. The key factor that determines whether someone with mental illness is at greater risk for serious violence in society is substance abuse.
Although we seek to find answers as to why people commit random acts of horrible violence, we should not focus on extraneous variables in a person, just because they’re convenient. Mental illness — by itself — is not the cause of violence in an individual.
And here’s the research to prove it.
Swanson and colleagues’ (1990) research over twenty years ago opened the door to this myth, finding that there was a link between mental illness and violence. However, I showed back in 2007 that a followup study led by the same researcher (Swanson et al., 2006) wasn’t as robust as researchers claimed.
One of the researchers on that study — Van Dorn — more recently claimed, “most researchers have concurred that a modest but statistically significant relationship exists between violence and SMI.”1 However, this is quickly followed by a footnote, noting: “Data from the MacArthur Violence Risk Assessment study showed that patients without alcohol or drug symptoms, as assessed by the MAST and DAST, were not significantly more violent than comparison group subjects without alcohol or drug symptoms.”
In fact, I’d argue “most” researchers do not believe such a relationship exists without the presence of a co-occurring substance abuse disorder, as Lurigio & Harris (2009) note in their evidence-based review on this issue.
Back in 2009, another important study on mental illness and violence was conducted by Elbogen & Johnson (2009). This work found that serious mental illness was statistically unrelated to community violence unless comorbid substance abuse or dependence was involved based upon a very large community dataset called NESARC. NESARC is a robust, two-wave survey that was conducted face-to-face with adults in the U.S. Wave 1 interviewed 43,093 people in 2001, and Wave 2 interviewed 34,653 people as a followup from the first wave. This is generally considered a gold-standard dataset that is representative of the U.S. population at the time.
Some researchers believe the relationship between mental illness, substance abuse, and violent behavior is more complicated than what Elbogen & Johnson found — but that “serious mental illness” is definitely a risk factor.2
So what do you do when someone’s research findings conflict with your existing beliefs? You reanalyze the data. Three of those researchers are Van Dorn, Volavka & Johnson (2012) . They reanalyzed the same dataset using a different set of research assumptions and analysis procedures. Some might call this a fishing expedition.
Can you guess what this research found?
Unsurprisingly, it again dredged up a link between serious mental illness — even without substance abuse — and violence.
But here’s what the researchers also found as a predictive risk factor for serious violence:
- Growing up in a unstable, antisocial household
- Parental history of physical abuse
- Parental history of neglect
- Parental history of both physical abuse and neglect
- Binge drinking
- Stressful life events
- Being male
Huh. How come none of these factors get much media attention?
In fact, Figure 2, entitled “Predicted probability of any violence between Waves 1 and 2 as a function of mental disorder, substance use disorder, and history of childhood events” is probably the most telling and relevant to this discussion:
SMI = serious mental illness, SU = substance use, MI = mental illness
See an interesting pattern there? It’s not mental illness that’s a good predictor of violence — it’s childhood abuse. Childhood abuse more than doubles your risk of violence alone. And while it shows that mental illness and substance abuse both increase this risk substantially on their own, the real multiplier is when you combine these two.3
Look at the blue bars for “SMI only” and “Other MI only.” They are equivalent to a person who has no mental illness but has suffered from child abuse or neglect.
Whether intended or not, what the researchers have clearly shown — once again — is that it’s not mental illness alone that contributes to a greater risk of violence. It is when mental illness is combined with substance abuse that matters. And as their data also show, it’s one more factor too: childhood abuse or neglect.
When you put those three factors together, you have a clinically significant risk for violence.
When you look at just one factor alone, it’s unlikely that risk of violence is of much clinical validity (although the data may show some statistical significance, as it does for being a man or binge drinking).
The upshot from this most recent research confirms what I’ve been harping on now for the past decade — the relationship between mental illness and violence is not a direct one. It is a complex one that is primarily mediated by substance use and abuse. Take away the substance abuse and you have a weak relationship that is likely no more predictive than the person’s age.
The latest research also demonstrates a number of risk factors we don’t look at often enough — a turbulent childhood household, and childhood abuse and/or neglect. Stressful life events and binge drinking also contribute to risk.
It is this profile — not a single characteristic — that suggests an increased risk factor for violence. And unless we are careful to consider the whole picture, policy makers risk scapegoating a significant group of people. As we will continue to reaffirm, people with mental illness are more likely to be the victims of violence, not the perpetrators of it.
Elbogen EB & Johnson SC. (2009) The intricate link between violence and mental disorder: results from the national epidemiologic survey on alcohol and related conditions. Arch Gen Psychiatry, 66, 152–161. doi:10.1001/archgenpsychiatry.2008.537
Lurigio, AJ & Harris, AJ. (2009). Mental illness, violence, and risk assessment: An evidence-based review. Victims & Offenders, 4, 341-347.
Swanson JW, Holzer CE, Ganju VK, Jono RT. (1990) Violence and psychiatric disorder in the community: evidence from the epidemiologic catchment area surveys. Hosp Community Psychiatr, 41, 761–770.
Swanson, J.W.; Swartz, M.S.; Van Dorn, R.A.; Elbogen, E.B; Wagner, H.R.; Rosenheck, R.A.; Stroup, T.S.; McEvoy, J.P. & Lieberman, J.A. (2006). A National Study of Violent Behavior in Persons With Schizophrenia. Arch Gen Psychiatry, 63, 490-499.
Van Dorn, R., Volavka, J., & Johnson, N. (2012). Mental disorder and violence: Is there a relationship beyond substance use? Social Psychiatry and Psychiatric Epidemiology, 47, 487-503.
- I would say it’s a bit ingenuous for a researcher to speak on behalf of all researchers. [↩]
- These researchers define “serious mental illness” as only being schizophrenia, bipolar disorder, or clinical depression. It’s an arbitrary definition, as anyone with a panic disorder, agoraphobia, social anxiety disorder, or a myriad of other disorders would argue. [↩]
- This graph, and the researchers’ findings in general, also demonstrate how this arbitrary distinction between “serious mental illness” and other mental illness is of little validity. [↩]