I'm taking a class on occupational safety and health, and last night I was reading a chapter in the textbook about workplace violence and how to prevent it. As I was reading along, I found some rather troubling statements about workplace violence risk factors that I was pretty sure weren't correct. From the textbook
Individual Factors Associated with Violence Psychotic behavior. Individuals who incessantly talk to themselves, express fears concerning conspiracies against them, say that they hear voices, or become increas- ingly disheveled over time may be violence prone. Depression. People who suffer from depression are prone to hurt either them- selves or someone else. An employee who becomes increasingly withdrawn or overly stressed may be suffering from depression. (Occupational Safety and Health: For Technologists, Engineers, and Managers, 7th Edition. Prentice Hall/CourseSmart, 01/11/2010. p. 284).
Something smelled fishy... So I investigated the textbook's claims about mental illness and violence, and came up with some resources. I'll share them with you folks--originally I wanted to write a proper paper on the subject, but I'm short on time and I wanted to be sure to write this down while I was still motivated. (ADHD! Yay! Got to catch the inspiration when you can get it.)
This is probably the original study that they were basing their ideas on. You can see it's from way back in 1990, and reports increased risk of violence in mentally ill subjects. It does however reveal that mental illness and substance abuse are closely related, and that substance abuse disorders increase the risk of violence.
This one's a literature review, from 2003. No direct connection between mental illness and violence is found; however, interestingly, mentally ill people are revealed to be at risk of being the victims of violence--about 2.5 times more likely. Whether this is a direct relationship or an indirect one is still unknown.
More recent work from 2009, multivariate analysis--shows that mental illness and violence are not directly related. Mental illness does not independently predict violent behavior; mentally ill people without other risk factors have a rate of violence similar to the general population. One of those other risk factors--substance abuse--explains increased violence much better than mental illness; and because mentally ill people tend to be more likely to have addictions, the mental illness was assumed to be the true cause when in reality, mentally ill people who don't have addictions aren't at increased risk of committing violent acts. It was a spurious correlation all along.
(Substance abuse and violence, of course, is another topic altogether. I'm not sure that's a causal relationship either, but it's more closely related than mental illness, anyway. Could just be that substance abuse is more common in people who have high stress and less ability to inhibit impulses--both of which we know from experiment are directly related to violence.)
Other links that aren't from journals, but still pretty reputable and often have links to journal articlesPsychCentral's review of the problem MSNBC news story, analyzing a possible schizophrenia connection in a shooting New England Journal of Medicine opinion piece Fact sheet from the US department of health and human services WebMD covers the topic
I think maybe the textbook must have been depending on that first study, the one without the multivariate analysis, as well as on "common sense" that turned out to be incorrect. I have found that often times, textbooks are somewhat out of date--by the time the information gets from the journals to the textbooks, some of it will have been invalidated by more recent work.
This kind of misconception is really problematic for the people with mental illness--it makes it much harder for them to get hired; and since unemployment leads to stress and stress makes mental illness worse, that's a real vicious cycle.