Why do some people kill themselves?

By Robert Pool

Finding out what makes certain people commit suicide and others not, could provide new tools for screening patients and providing help before it is too late (Image: Matt Carr/Taxi)

Finding out what makes certain people commit suicide and others not, could provide new tools for screening patients and providing help before it is too late

FOR a few months in late 2006 and early 2007, the woman who called herself kristi4 was one of the best-known members of the pro-anorexia community. As the administrator of a blog on LiveJournal.com, she dispensed advice, encouraged others and wrote candidly about her own struggles. Then, late one Friday night, after a series of entries describing what she was planning to do, kristi4 killed herself with an overdose of prescription sleeping pills, muscle relaxants and painkillers.

Her death was just one tragic data point in one of the most striking statistics in all of psychology. It has long been known that anorexia has the highest death rate of any mental illness: one out of every five people with anorexia eventually die of causes related to the disease. What has only now been recognised, however, is that a huge number of those deaths are from suicide rather than starvation. Someone who develops anorexia is 50 to 60 times more likely to kill themselves than people in the general population. No other group has a suicide rate anywhere near as high (Archives of General Psychiatry, vol 60, p 179).

Recently, psychologists have tried to explain why anorexia and suicide are so intimately connected, something which is helping to answer the wider question of why anyone would commit suicide. If this explanation holds up, it will give psychiatrists a new tool for screening patients and determining which of them are most likely to kill themselves, perhaps saving lives.

Suicide has always been a conundrum for psychologists and other researchers interested in human behaviour. Self-preservation is one of the strongest human instincts, so the drive to commit suicide must be even more powerful. But what causes it?

A century ago, both the sociologist Emile Durkheim and the psychoanalyst Sigmund Freud came up with sweeping explanations. Durkheim, not surprisingly, saw the roots of suicide in social factors, such as a failure to integrate into society, while Freud rooted his explanation in instinctual drives, particularly what he called the death instinct. More recent explanations have tended to focus on factors such as depression, hopelessness and emotional pain, but none of them have had much success in answering the fundamental question about suicide: why do some people kill themselves while others in seemingly identical circumstances do not?

Some progress has been made by crunching large amounts of data on suicide, says Harvard University psychologist Matthew Nock, who studies suicide and self-harm. Researchers have learned, for example, that suicide rates are rising and now account for 1.5 per cent of all deaths worldwide. Suicide is the second leading cause of death among people aged 15 to 24, after vehicle accidents. Women are more likely than men to attempt suicide, while men are much more likely to succeed.
1 million Approximate number of suicides worldwide each year

Most people who commit suicide have a mental disorder - anorexia, major depressive disorder, bipolar disorder, schizophrenia and borderline personality disorder are the most common, but an elevated suicide risk is part and parcel of many of the others, too. People who kill themselves also generally feel deeply depressed and hopeless at the time.
Every 40 seconds somebody dies by suicide

What the statistics do not tell us - and what psychologists most want to know - is exactly which people are most at risk. The vast majority of depressed, hopeless people do not commit suicide, so why do some do it?

In 2005, psychologist Thomas Joiner, a suicide specialist at Florida State University in Tallahassee whose own father committed suicide, set out to answer that question. By studying suicide statistics and paying particular attention to the groups with above average rates, Joiner believes he has found a common thread others have missed. "It was the first grand theory of suicide in quite a while," says Nock.

In essence, Joiner proposed that people who kill themselves must meet two sets of conditions on top of feeling depressed and hopeless. First, they must have a serious desire to die. This usually comes about when people feel they are an intolerable burden on others, while also feeling isolated from people who might provide a sense of belonging.

Second, and most important, people who succeed in killing themselves must be capable of doing the deed. This may sound obvious, but until Joiner pointed it out, no one had tried to figure out why some people are able to go through with it when most are not. No matter how seriously you want to die, Joiner says, it is not an easy thing to do. The self-preservation instinct is too strong.

There are two ways people who want to die develop the ability to override the self-preservation instinct, Joiner argues. One is by working up to it. In many cases a first suicide attempt is tentative, with shallow cuts or a mild overdose. It is only after multiple attempts that the actions are fatal.
20 Number of failed suicide attempts for each successful one

The other is to become accustomed to painful or scary experiences. Soldiers and police who have been shot at or seen their colleagues injured or killed are known to become inured to the idea of their own death. Both groups also have a higher-than-normal suicide rate. Similarly, doctors and surgeons who witness pain, injury and death are more likely to be able to contemplate it for themselves - the suicide rate for doctors is significantly higher than for the general population. Joiner describes this as a "steeliness" in the face of things that would intimidate most people.

Another group that displays steeliness are people with anorexia. Joiner had noted their heightened suicide rate in his original work, Why people die by suicide (Harvard University Press, 2005), but it wasn't until later that he grasped the importance of the connection.

That realisation began to dawn in 2006, during a seminar in which two of Joiner's graduate students, Jill Holm-Denoma and Tracy Witte, were listening to him describe the risk of suicide among people with anorexia. Witte observed that the high suicide rate had two possible explanations. Perhaps people with anorexia were no more likely to attempt suicide than people with other mental disorders, but the anorexia had so weakened their bodies that their suicide attempts were more likely to succeed. Alternatively, perhaps anorexia had so inured them to pain that they were more capable than others of doing what was necessary to kill themselves.

According to Joiner's hypothesis, the second explanation should be correct. So Holm-Denoma set out to test the prediction. She examined nine suicides chosen randomly, and what she found told a very clear story.

"These people would have died regardless of their body weight," she says. "We were just astounded by the lengths to which they went to make sure they were successful." Three jumped in front of trains. Two hung themselves. Two took large drug overdoses. One poisoned herself with sleeping pills and toilet bowl cleaner. And one locked herself in a gas station restroom and set fire to a trash can that produced enough carbon monoxide to asphyxiate her. Nine cases, of course, are not enough to prove the point, but the fact that all took such drastic measures to kill themselves says something (Journal of Affective Disorders, vol 107, p 231).

Anorexia offers a "perfect storm" of the factors laid out in Joiner's hypothesis, Holm-Denoma says. Social isolation is likely because people with anorexia avoid any interactions that might involve food - so that means not going out for a meal, no movies (the popcorn might be too tempting) and no stopping by a friend's house. The result is the "thwarted belongingness" that Joiner describes as a key factor in suicide.

Then there is the feeling that they have become an intolerable burden to their family and friends. One popular approach to treating anorexia in children, for example, involves having a parent oversee their child full-time.

Most importantly, anorexia means becoming inured to pain. Merciless starvation leads to intense and painful hunger pangs and major headaches. Osteoporosis is common, making fractures more likely, not to mention the chest pains caused by heart damage. Kristi4's blogs in the month leading up to her suicide show this perfect storm at work.

It is one of the strengths of Joiner's explanation, says Nock, that it makes testable predictions such as the one spotted by Witte. For example, it should be possible to develop psychological tests to measure how much of a burden people feel, or how thwarted, and then use them to predict who will commit suicide. It should also be possible to examine rates of suicide among various groups with the characteristics Joiner is talking about.

Those tests are slowly taking shape. Recent work by some of Joiner's students has shown that people who feel they are a burden and also experience thwarted belongingness are more likely to have suicidal thoughts (Journal of Consulting and Clinical Psychology, vol 76, p 72). A second study found that "painful and provocative events", such as shooting a gun or getting into a fight, tend to increase something Joiner calls "acquired capability" - a written test that measures someone's ability to hurt or kill themselves.

Meanwhile, University of Minnesota psychiatrist Scott Crow has studied suicide rates among people with bulimia and found that they, too, kill themselves at a much higher rate than the general population. Crow has found a four to six-fold increase in suicides in this group. Bulimia starves the body at some level, as indicated by various biochemical markers, so people with bulimia may well be inured to pain in much the same way as those with anorexia.
60 per cent increase in worldwide suicide rates since 1965

Even though the evidence is all pointing in the same direction, Joiner says many more tests will be needed before his ideas can be accepted as a general explanation for suicide. "It's a start," he says of the evidence assembled to date. "But we need something much more systematic."

Ultimately, he says, a better understanding of why people commit suicide should help clinicians better assess who is most at risk, and find new ways of preventing people from killing themselves. Long-term psychotherapy, for instance, could help chip away at a person's fearlessness and lessen the likelihood that they will commit suicide.

But as long as people steel themselves to pain, as long as they feel isolated and a burden to others, Joiner's theory predicts that suicide will be with us as well.

Robert Pool is a writer based in Tallahassee, Florida