The Noonday Demon

by Solomon, Andrew

  Suicide is astonishingly common and is disguised and trumped up more even than depression. It is, indeed, a vast public health crisis that makes us so uncomfortable that we divert our eyes from it. Every seventeen minutes, someone in the United States commits suicide. Suicide ranks number three among causes of death for Americans under the age of twenty-one, and it is number two for college students. In 1995 (for example), more young people died of suicide than of AIDS, cancer, stroke, pneumonia, influenza, birth defects, and heart disease combined. From 1987 to 1996, more men under the age of thirty-five died of suicide than of AIDS. Nearly half a million Americans are taken to hospitals every year because of suicide attempts. Suicide was, according to the World Health Organization, responsible for almost 2 percent of deaths worldwide in 1998, which puts it ahead of war and way ahead of homicide. And the rate of suicide is climbing steadily. One recent study conducted in Sweden showed the likelihood that a young man in the study’s catchment area would commit suicide has increased by 260 percent since the 1950s. Half of those with manic-depression will make a suicide attempt; one in five people with major depression will do the same. Someone in a first depressive episode is particularly likely to attempt suicide; a person who has lived through a few cycles has in general learned to live through cycles. Previous attempts at suicide are the strongest factor in predicting suicide: about a third of the people who kill themselves have tried before; 1 percent of those attempting suicide will complete a suicide within a year; 10 percent will kill themselves within ten years. There are approximately sixteen suicide attempts to every completed suicide.

  I have seen in a single document both the assertion that depressed people are five hundred times as likely to commit suicide as their nondepressed counterparts and the statistic that depressed patients have a suicide rate twenty-five times the societal norm. I have read, elsewhere, that depression increases the likelihood of suicide twofold. Who knows? Such rates depend in large part on how one defines that slippery demon depression. For what appear to be public health reasons, the National Institute of Mental Health grandly if unscientifically averred for a long time that “almost all people who kill themselves have a diagnosable mental or substance abuse disorder”; they recently downgraded the “almost all” to “90 percent.” This notion helps people who have made unsuccessful suicide attempts, and those left grieved by the suicide of someone to whom they were close, to expel some of the guilt that might otherwise fetter them. Comforting though this is, and useful though it may be in drawing attention to the high level of illness-associated suicide, it is a gross exaggeration, not corroborated by anyone I’ve met who has treated suicidal patients.

  Suicide statistics are even more chaotic than depression statistics. People most often commit suicides on Mondays; suicides are most prevalent between later morning and noon; the preferred season for suicide is springtime. Women have a high rate of suicide during the first and last weeks of the menstrual cycle (a phenomenon for which there may be hormonal explanations) and a low rate during pregnancy and in the first year after birth (a phenomenon that makes obvious evolutionary sense, but for which we have so far no definitive chemical explanation). One school of suicide researchers loves comparative statistics and uses them as though correlation implied causality. Some of these correlations come close to absurdity: one can calculate the average body weight of those who commit suicide, or the average length of their hair, but what exactly would that prove and what use would it be?

  Émile Durkheim, the great nineteenth-century sociologist, pulled suicide out of the realm of morality and placed it in the more rational domain of social science. Suicides are subject to categorization, and Durkheim argued that there are four significant types. Egoistic suicide is committed by people who are inadequately integrated into the society they inhabit. Apathy and indifference motivate them to sever permanently their relationship to the world. Altruistic suicide comes of being overly integrated into one’s society; Durkheim’s category would include, for example, Patrick Henry’s devotion to the idea “Give me liberty or give me death!” Those who commit altruistic suicide are energetic, passionate, and determined. Anomic suicide is the consequence of irritation and disgust. “In modern societies,” Durkheim writes, “social existence is no longer ruled by custom and tradition, and individuals are increasingly placed into circumstances of competition with one another. As they come to demand more from life, not specifically more of something but simply more than they have at any given time, so they are more inclined to suffer from a disproportion between their aspirations and their satisfactions, and the resultant dissatisfaction is conducive to the growth of the suicidal impulse.” As Charles Bukowski once wrote, “We demand more of life than there is”—and our inevitable disappointment may be occasion enough to end life. Or as de Tocqueville wrote of American idealism in particular, “the incomplete joys of this world will never satisfy the human heart.” Fatalistic suicide is committed by people whose lives are genuinely miserable in a way beyond change—the suicide of a slave, for example, would be fatalistic in Durkheim’s taxonomy.

  Durkheim’s categories are no longer used for clinical purposes, but they have defined much modern thinking about suicide. Contrary to beliefs of his time, Durkheim proposed that though suicide is an individual act, its sources are societal. Any single suicide is the result of psychopathology, but the relatively consistent appearance of psychopathological suicidality seems to be tied to social constructs. In each society there is a different context for the act, but it may be the case that a certain percentage of the population in every society kill themselves. The values and customs of a society determine which causes will lead to the act in which place. People who believe that they are operating on the basis of unique trauma are often, in fact, simply manifesting a tendency in their society that drives people to death.

  Though many meaningless statistics clutter suicide studies, some tendencies can usefully be identified. Members of families in which there has been a suicide are far more likely than others to kill themselves. This is in part simply because family suicides make the unthinkable thinkable. It is also because the pain of living when someone you love has annihilated himself can be almost intolerable. A mother whose son had hanged himself said to me, “I feel as though my fingers are being caught in a slamming door and I’ve been stopped permanently in midscream.” It is also because, at a presumably genetic level, suicide runs in families. Adoption studies show that biological relatives of a suicidal person are more frequently suicidal than are that person’s adoptive relatives. Identical twins tend to share suicidality, even if they are separated at birth and have no knowledge of one another; nonidentical twins do not. It cannot be a selective advantage to have single-function “suicide genes,” but the combination of genes that cause depression and violence and impulsivity and aggression may provide a genetic map that is both somewhat predictive of suicidal behavior and advantageous in particular situations.

  Suicide breeds suicide in social communities as well. The contagion of suicide is incontrovertible. If one person commits suicide, a group of friends or peers will frequently follow; this is especially true among teenagers. Locations for suicide are used over and over again, carrying the curse of those who have died: the Golden Gate Bridge in San Francisco, Mount Mihara in Japan, particular stretches of railway lines, the Empire State Building. Suicide epidemics have recently occurred in Plano, Texas; Leominster, Massachusetts; Bucks County, Pennsylvania; Fairfax County, Virginia; and in a number of other apparently “normal” communities in the United States. Public accounts of suicides also inspire suicidal behavior. When Goethe’s The Sorrows of Young Werther was published in the early nineteenth century, copycat suicides in the mode of Goethe’s protagonist’s were committed across Europe. Whenever a major suicide story breaks in the media, the suicide rate goes up. In the period immediately following Marilyn Monroe’s suicide, for example, the rate of suicide in the United States increased 12 percent. If you are hungry and see a
restaurant, you’re likely to go in. If you’re suicidal and read about a suicide, you’re likely to take the final step. It seems clear that a reduction in reporting on suicide would lower the suicide rate. At the moment, evidence suggests that even the best-intentioned suicide-prevention programs often introduce the idea of suicide to a vulnerable population; it seems possible that they actually increase the suicide rate. They are helpful, however, insofar as they make people aware that suicide is often the result of mental illness and that mental illness is treatable.

  Contrary to popular myth, those who talk about suicide are the most likely to kill themselves. Those who attempt tend to attempt again; in fact, the best predictor of an actual suicide is an attempt. No one makes much use of this fact. Maria Oquendo’s 1999 study of treatments points out that though “a history of attempting suicide could be used by clinicians as a marker of a propensity for future attempts, patients with such a history were not treated more intensely than those without. It remains unclear whether those patients at substantial risk for suicide in association with major depression because of their history of suicidal acts are either not being recognized as at risk or are not receiving adequate somatic treatment despite the clinician’s recognition of their heightened vulnerability.”

  Though sweeping existential arguments are engaging, the reality of suicide is not fine and pure and philosophical, but messy and appalling and physical. I have heard it said that severe depression is “a living death anyway.” A living death is not pretty, but unlike a dead death, it offers scope for amelioration. Suicide’s finality makes it a problem beyond any other discussed in this book, and the capacity of antidepressants to avert suicide needs urgently to be measured so that the appropriate medications can be deployed. Industry researchers find suicidality hard to monitor, especially as culminating self-abnegation does not ordinarily take place during the twelve-week span of a “long-term” controlled study. None of the SSRIs, the most popular class of antidepressant drugs in the world, have been monitored for their capacity to prevent suicide. Among other drugs, lithium has been most rigorously tested—the rate of suicide among bipolar patients who discontinue lithium treatment increases sixteenfold. Some drugs that alleviate depression may increase the motivation for suicide because they augment motivation in general; drugs can set off mechanisms of self-destruction as they mitigate depression’s torpor. It is important to distinguish between this enabling and actual causality. I do not believe that people commit suicide as a direct result of medication unless the suicidality has been strong in them for some time. Careful interviews should be conducted with patients before they are prescribed activating antidepressants. ECT can immediately allay urgent or delusional suicidal impulses. One study shows the suicide rate nine times higher for patients with severe illness on medications than for similarly ill patients treated with ECT.

  Freud, at about the same time as Durkheim, proposed that suicide is frequently a murderous impulse toward someone else that is carried out on oneself. The psychologist Edwin Shneidman has more recently said that suicide is “murder in the one hundred eightieth degree.” Freud posited that a “death instinct” is always in uncertain balance with the life instinct. This fascination does clearly exist and is certainly responsible for suicides. “The two basic instincts operate against each other or combine with each other,” Freud wrote. “Thus the act of eating is a destruction of the object with the final aim of incorporating it, and the sexual act is an act of aggression with the purpose of the most intimate union. The concurrent and mutually opposing action of the two basic instincts gives rise to the whole variation of the phenomena of life.” Suicide, here, is the necessary counterpoint to the will to live. Karl Menninger, who has written extensively on suicide, said that suicide requires the coincidence of “the wish to kill, the wish to be killed, and the wish to die.” G. K. Chesterton, following in this mode, wrote:

  The man who kills a man kills a man

  The man who kills himself kills all men.

  As far as he is concerned, he wipes out the world.

  As we confront chronic stresses for which we are ill equipped, we rely on and overuse neurotransmitters. The rush of neurotransmitters we induce in sudden stress cannot be kept up during more sustained stress. For this reason, people who experience chronic stress tend to sap their transmitters. Suicidal depression appears to have some distinguishing neurobiological characteristics, which may cause suicidal behavior or may simply reflect suicidal tendencies. Actual suicide attempts are usually brought on by external stresses, which frequently include use of alcohol, acute medical illness, and negative life events. How prone someone is to suicide is determined by personality, genetics, childhood and rearing, alcoholism or substance abuse, chronic illness, and cholesterol level. Most of our information on the suicidal brain comes from postmortem studies. Suicides have low levels of serotonin at certain key locations in the brain. They have excessive numbers of serotonin receptors, which may reflect the brain’s attempt to make up for the low levels of serotonin. The level of serotonin seems to be especially low in the areas associated with inhibition, and this deficiency appears to create a powerful freedom to act impulsively on emotion. People given to unbridled aggression often have low serotonin in the same area. Impulsive murderers and arsonists have lower serotonin levels than do most people—lower than those of nonimpulsive murderers or other criminals. Animal experiments show that primates with low serotonin are more likely to take risks and be aggressive than are their counterparts. Stress can cause both leaching of neurotransmitters and excess production of enzymes that destroy them. Levels of noradrenaline and norepinephrine appear to be reduced in postmortem suicidal brains, though the results are less consistent than the serotonin results. Enzymes that break down norepinephrine appear to be present in excessive quantities, and chemicals necessary for adrenaline to function are present in low quantities. What all of this means in functional terms is that people with low levels of essential neurotransmitters in key areas are at a high risk for suicide. This is the result consistently found by John Mann, a leading suicide researcher now working at Columbia University. He has used three different measures of serotonin levels in suicidal patients. Marie Åsberg, at the Karolinska Hospital in Sweden, has extrapolated clinical implications from such material. In a pioneering study, she kept track of patients who had previous suicide attempts and whose levels of serotonin appeared to be low; 22 percent of them had killed themselves within a year. Subsequent work has confirmed that while only 15 percent of depressives kill themselves, 22 percent of low-serotonin depressives will commit suicide.

  Given that stress leaches serotonin and low serotonin increases aggressivity and high aggressivity leads to suicide, it is no surprise that stressed depression is the sort most likely to lead to suicide. Stress leads to aggressivity because aggressivity is frequently the best way to deal with short-term threats that induce stress. Aggressivity, however, is nonspecific, and though useful in combating an attacker, it may likewise be turned against oneself. It seems likely that aggression is a basic instinct, while depression and suicidality are more sophisticated cognitive impulses that developed later on. In evolutionary terms, the desirable trait of learning self-protective behavior is inextricably mingled with the undesirable trait of learning self-destructive behavior. The capacity for suicide is a burden that comes with the consciousness that distinguishes us from other animals.

  Genetics may determine low levels of serotonin, and the gene that sets levels of the enzyme tryptophan hydroxylase is now clearly associated with high rates of suicide. Genes not only for mental disease but also for impulsivity, aggression, and violence may open them up to heavy risk. Animal experiments on monkeys brought up without mothers show that a deprived upbringing lowers serotonin levels in specific areas. It seems likely that early abuse may permanently lower serotonin levels and so increase the likelihood of suicide (quite apart from the problem of cognitive depression caused by that abuse). Substance abuse may furth
er lower serotonin levels—and so, interestingly, does low cholesterol. Neurological damage to the fetus, caused by alcohol or cocaine use, may predispose children to mood disorders that lead to suicide; lack of maternal attention may deprive them of early developmental stability; diet may work adversely on their brain. Men have lower serotonin levels than women. So a stressed male with a genetic predisposition to low serotonin who has had a deprived upbringing, abuses substances, and has low cholesterol would fit the profile of a likely suicide. Drugs that increase serotonin levels in such individuals would be good agents to prevent suicide. Brain scans to detect levels of serotonin activity in the relevant parts of the brain—technology that does not exist now but may exist soon—might be used to estimate the likelihood of someone’s attempting suicide. Better brain-imaging techniques may eventually allow us to check out the brains of depressed people and assess who is likely to attempt suicide. We have a long way to go. “For scientists to minimize the complexity of the chemical interactions within the brain or at the synapses,” Kay Jamison writes in her masterful book on suicide, “would be a damning mistake, a late-twentieth-century equivalent of earlier, primitive views that deranged minds were caused by satanic spells or an excess of vapors.”