Why People Die By Suicide

by Thomas Joiner

  so it would not surprise me if future research found a link between

  this variable and suicidality as well.

  The literature is clear. Of all neurotransmitter systems, the seroto-

  nin system is the most important with regard to suicide risk. Metab-

  olites of dopamine and norepinephrine are generally no different

  in those at risk for suicide than in others. However, one other brain

  system, the hypothalamic-pituitary-adrenal (HPA) axis, deserves

  mention.

  The HPA axis is the body’s main “stress reaction” system. Any

  stress lasting longer than a few minutes will stimulate release of a

  particular hormone by a structure in the brain called the hypothala-

  mus. The released hormone then acts on the pituitary gland, causing

  it to release still another hormone. This in turn causes the adrenal

  cortex to release cortisol, often called the “stress hormone.”

  Genetics, Neurobiology, and Mental Disorders ● 183

  When this sequence works well and normally, it prepares the body

  for “fight or flight” responses to stress. But if stress is chronic and se-

  vere, cortisol is always circulating. The problem with this is that

  cortisol signals the hypothalamus and the pituitary gland to stop

  producing their respective hormones. Chronic signaling makes the

  system unresponsive, leaves too much cortisol circulating, and thus

  impairs the person’s ability to respond to stress.

  One way to measure HPA-axis activity is by administration of a

  substance called dexamethasone. Dexamethasone is a synthetic ste-

  roid similar to cortisol, which, in normal people, suppresses release

  of one of the hormones that leads up to cortisol release. Therefore,

  giving dexamethasone should reduce this hormone and thus reduce

  cortisol levels, as long as the HPA system is working well.

  However, if the system has become dysregulated from chronic

  stress, it will not be sensitive to dexamethasone, and thus cortisol

  production will not be suppressed. Nonsuppression of cortisol in re-

  sponse to dexamethasone indicates an HPA system gone awry.

  Nonsuppression of cortisol in response to dexamethasone may be

  predictive of later death by suicide. One study followed a group of

  patients over fifteen years and found that those with nonsuppression

  of cortisol at baseline went on to have a fourteenfold greater risk of

  death by suicide than those who did suppress cortisol output in re-

  sponse to dexamethasone.29 These studies suggest that hyperactivity

  of the HPA system could be involved in suicidal behavior.

  It even is implicated in self-wounding behaviors in monkeys. Rhe-

  sus monkeys who frequently and severely wound themselves (usually

  through biting) showed the same kinds of responses to dexametha-

  sone as do humans whose HPA systems are dysregulated.30

  In summary, neurobiological research to date has clearly shown

  that there are serotonin-related differences in suicidal individuals as

  compared to others. These differences appear to be specific to seroto-

  nin and may not involve other neurotransmitters. Studies of spinal

  184 ● WHY PEOPLE DIE BY SUICIDE

  fluid and fenfluramine challenge responses indicate decreased sero-

  tonin-system function in suicide attempters and completers. Re-

  search has also raised the possibility of involvement of the HPA sys-

  tem in suicidality.

  Interestingly, many of the serotonin-system differences between

  suicidal people and others are also relevant to the personality vari-

  able of impulsivity. Serotonin-system problems may contribute to

  both suicidality and impulsivity. For example, differences between

  those who died by suicide and others regarding serotonin binding

  appear to be localized to an area of the prefrontal cortex, an area in-

  volved in impulse control.31 Disruptions in serotonin-system func-

  tioning may predispose people to an array of impulsive behaviors,

  which, in turn, may reduce fear of provocative experiences, including

  suicidality.

  Impulsivity and suicidality are not the only consequences of sero-

  tonin-system dysfunction. Another consequence involves tendencies

  toward negative emotion, depression, and anxiety. With regard to the

  theory proposed here, factors that increase negative emotionality

  may affect suicidality via impact on feelings of burdensomeness and

  failed belongingness.

  Impulsivity

  Impulsive personality characteristics are a well-documented risk fac-

  tor for serious suicidality. To get a sense of what “impulsive” means,

  consider these items from a measure of impulsivity: “Have people

  told you that you’re a daredevil type or that you take too many

  risks?” “Have you driven recklessly?” “Have you hurt yourself regu-

  larly, even if you didn’t mean to (e.g., falling, bruising)?” “Have you

  stolen material goods (such as clothes or jewelry) from a store or

  vendor?” “Have you impulsively spent money on clothes, jewelry, or

  other items?” Answering “yes” to most of these questions would indi-

  cate an impulsive personality style.

  Genetics, Neurobiology, and Mental Disorders ● 185

  Impulsivity can have serious negative consequences. Menninger

  wrote, “As for impulsiveness, a volume could be written about the

  disastrous consequences of this symptom. It has ruined many a busi-

  ness, many a marriage, and many a life.”32

  In a study of 529 mood-disordered patients, 36 participants died

  by suicide and 120 others attempted suicide during the fourteen

  years of the study. Impulsivity was among the variables that differen-

  tiated those who died by or attempted suicide from those who had

  no suicide attempt.33 In a study of suicide attempts in 295 women

  with bulimia nervosa, the binge-purge syndrome, over a quarter of

  the women had attempted suicide, often including severe and multi-

  ple attempts. Those who had attempted suicide differed from those

  who had not with regard to frequency of impulsive behaviors.34 These

  and other studies demonstrate that impulsivity is involved in suicidality.

  But how is it involved? The literature on suicide often implies that a principal mechanism underlying the relation of impulsivity to suicide is “spur-of-the-moment” suicide—that is, someone deciding all

  of a sudden, perhaps in response to a serious disappointment or con-

  flict, to die by suicide. I am very skeptical of this concept, and I doubt

  that true “spur-of-the-moment” suicides exist. Impulsivity is impli-

  cated not so much at the time of death, but beforehand, leading to

  experiences that allow people to get used to pain and provocation

  and engage opponent processes (e.g., impulsive people drink more

  and are injured in accidents more than others). Through repeated

  impulsive acts, suicidal and otherwise, impulsive people may become

  experienced, fearless, and competent regarding suicide and thus ca-

  pable of forming plans for their own demise.

  Musician Kurt Cobain’s suicide is a very clear example of a planned

  suicide in someone who was viewed as impulsive and had clearly ac-

  crued an array of pro
vocative experiences, including repeated self-in-

  jury. His impulsivity was involved in his death, but not in the sense

  of a “spur-of-the-moment” decision to die. Rather, his impulsivity

  led him to experiences that reduced his fear of death.

  186 ● WHY PEOPLE DIE BY SUICIDE

  Alvarez was also skeptical of the phenomenon of “spur-of-the-

  moment” suicides, which he termed “so-called ‘impetuous’ sui-

  cides.”35 He says of them that, if they survive, they “claim never to

  have considered the act until moments before their attempt. Once re-

  covered, they seem above all embarrassed, ashamed of what they

  have done, and unwilling to admit that they were genuinely sui-

  cidal . . . They deny the strength of their despair, transforming their

  unconscious but deliberate choice into an impulsive, meaningless

  mistake. They wanted to die without seeming to mean it.” To Alvarez,

  “impetuous suicides” are ersatz; they are really usual suicide attempts

  with a posthoc, shame-saving explanation of impulsivity.

  Menninger’s book Man against Himself is packed with newspaper

  clippings and clinical anecdotes about suicidal behavior, but not one

  is a “spur-of-the-moment suicide,” with one exception. The only ex-

  ception is a fictional one, Shakespeare’s Romeo.

  Interestingly, the serotonin system, as was implied above, is impli-

  cated as a basis for impulsive personality style. For example, com-

  pared to others, people with impulsive/erratic personality disorders,

  people with histories of aggression, arsonists and other violent of-

  fenders, and people who have murdered a relationship partner all

  have lower levels of serotonin metabolites in their spinal fluid. In vi-

  olent suicide attempters, those who were identified as having high

  impulsivity had significantly lower serotonin metabolites than non-

  impulsive attempters and controls.36 Plasma blood levels of serotonin

  metabolites were lower in impulsive suicide attempters than non-

  impulsive attempters and controls.37 Potentially lethal suicidal be-

  havior, impulsivity, and disruptions in the serotonin system appear

  to be inter-related.

  Fenfluramine challenge studies support this conclusion. As was

  noted earlier, fenfluramine stimulates serotonin release. A blunted

  response to fenfluramine challenge suggests less serotonin-system ac-

  tivity. Blunted response to the fenfluramine challenge is seen in im-

  pulsive people,38 as it is in people at high suicide risk. One study

  Genetics, Neurobiology, and Mental Disorders ● 187

  examined the relationship of suicide attempt lethality as well as im-

  pulsive behaviors to fenfluramine challenge. Individuals with high

  lethality attempts and impulsive personality characteristics showed

  the lowest responses, indicating the most underactive serotonin sys-

  tems.39

  My claim is that impulsivity is associated with suicidal behavior,

  but indirectly. According to the model developed in this book,

  impulsivity only relates to suicidal behavior because impulsivity fa-

  cilitates exposure to provocative and painful experiences. A similar

  statement could be made about the relation of serotonin-system

  problems to suicidality—the association exists only because seroto-

  nin-system problems tend to produce impulsivity (and negative

  emotion), which, in turn, increase the likelihood of provocative and

  painful experiences. The latter, according to my view, instills the ac-

  quired ability to enact lethal self-injury. Disruptions in serotonin-

  system functioning may predispose people to an array of impulsive

  behaviors, which, in turn, may reduce fear of provocative experi-

  ences. These experiences may lead to the acquired ability to enact le-

  thal self-injury and thus to increased risk for completed suicide.

  The definitive study to test this claim has not been conducted, to my

  knowledge. How would such a study look? I can think of two inter-

  esting strategies. The first would be to measure impulsivity, painful

  or provocative experiences, and suicidal behavior in a very large sam-

  ple of people (a large sample is needed because of the relative rarity

  of suicidal behavior; one could also study a smaller, high-risk group

  in whom suicidal behavior is more likely). I predict that there would be

  a significant association between impulsivity and suicidal behavior; but

  crucially, I predict that this association would be reduced or elimi-

  nated when painful or provocative experiences were accounted for.

  This is a simplification, but generally, this pattern of results would

  indicate that impulsivity is associated with suicidality only because it

  facilitates exposure to painful or provocative experiences.

  A second strategy would be to examine samples in which either

  188 ● WHY PEOPLE DIE BY SUICIDE

  impulsivity or painful or provocative experience was a constant.

  Imagine a sample selected so that everyone had the same level of

  impulsivity. I would predict an association between painful or pro-

  vocative experiences and suicidality in this sample, because, accord-

  ing to my theory, painful or provocative experiences confer risk to

  suicide by dampening fear of self-injury, and this is the case regard-

  less of the level of impulsivity. Now imagine a sample selected so that

  everyone had the same level of painful or provocative experiences. I

  would not predict an association between impulsivity and suicidal

  behavior in this sample, because, in my view, impulsivity only relates

  to suicidality through its relation to painful or provocative experi-

  ences. If the latter is held constant, impulsivity would have no “trac-

  tion” through which to predict suicidality. My students and I are cur-

  rently conducting experiments to test this claim.

  Of all personality dimensions, impulsivity has the most clearly

  documented association with suicidal behavior. I believe this has

  provided the misleading suggestion that the act of suicide itself is

  an impulsive decision. I don’t think so. Rather, there is a real and im-

  portant association between impulsivity and suicidality, and it exists

  because impulsivity leads people to habituate to pain and provoca-

  tion. They thus acquire the ability to enact lethal self-injury, and are

  thereby at increased risk for suicide, if the desire for death is in place.

  Impulsivity could relate to suicidality through increasing the desire

  for death as well. As Menninger noted, impulsivity can ruin lives. Ac-

  cordingly, it would not be surprising if it tended to increase feelings

  of burdensomeness and failed belongingness.

  Childhood Adversity

  There is now little doubt of an association between childhood mal-

  treatment and later suicidality—a real association not explained

  away by other variables. Other variables are important to consider. It

  Genetics, Neurobiology, and Mental Disorders ● 189

  could be, for example, that the same genes that predispose a parent

  to be abusive predispose the child to be suicidal. In fact, excellent

  candidates for such genes would be those that underlie i
mpulsivity—

  an impulsive parent is more likely to abuse, and an impulsive child is

  more likely to attempt suicide. Under this scenario, there is no real

  association between child abuse and later suicidality. The real mech-

  anism is genes and personality simultaneously raising risk for abuse

  by the parent and suicidality in the child.

  This scenario appears to have been ruled out. There seems to be a

  direct link between childhood adversity and later suicidality, a link

  not explained by other variables. For example, as was noted earlier,

  feelings of expendability (including burdensomeness) have been em-

  pirically linked to suicide; it would not be at all surprising if the

  experience of childhood abuse and neglect were a main source of

  feeling expendable. A study of Eskimos in the Bering Strait region

  showed that the majority of a sample of suicide attempters had lost

  a parent during childhood.40 Similar results were reported among

  patients with borderline personality disorder. Patients with border-

  line personality disorder who had died by suicide experienced more

  childhood losses such as death of a parent as compared to living con-

  trol patients with borderline personality disorder.41

  As noted in Chapter 3, neglectful parenting is an independent risk

  factor for adolescent suicidal ideation and attempts. This is true even

  after adjusting for other powerful variables like the presence of psy-

  chiatric disorder.42 Childhood physical abuse differentiates adoles-

  cents who died by suicide from matched controls.43 A study of over

  3,000 female adolescent twins found that childhood physical abuse

  was one of the factors most associated with a history of attempting

  suicide.44 Childhood physical abuse was also associated with lifetime

  suicide attempts in a study of people with alcohol-use disorders.45

  A very persuasive study on this topic followed 776 randomly se-

  lected children from a mean age of five years to adulthood in 1975,

  190 ● WHY PEOPLE DIE BY SUICIDE

  1983, 1986, and 1992 during a seventeen-year period. More than 95

  percent of the sample was retained throughout the entire study pe-

  riod, a considerable achievement. The researchers ascertained the oc-

  currence of abuse through official records as well as by participants’

  recall of abuse incidents. Results showed that childhood abuse con-