Why People Die By Suicide

by Thomas Joiner

  present or not as extreme). Thus their activities may be less likely

  to produce habituation and engage opponent processes. Among

  women with bulimia, however, I would predict that those with purg-

  ing behaviors (a provocative experience) would endorse more sui-

  cidal symptoms than those with nonpurging behaviors (e.g., exces-

  sive exercise, a relatively less provocative experience). In fact, a

  history of suicide attempt is more prevalent in purging bulimic

  women than in other bulimic women.72

  As is the case for women with anorexia, women with bulimia have

  198 ● WHY PEOPLE DIE BY SUICIDE

  decreased pain sensitivity. In fact, their decreased sensitivity to pain

  may persist even after their eating disorder resolves. One study com-

  pared bulimic women who had recovered from bulimia at least a year

  ago to fifteen healthy volunteer women.73 All women in the study re-

  ceived two pain evaluations. The first was a thermal pain stimulation

  test, which evaluates heat tolerance; the second was the submaximal

  effort tourniquet test, which assesses tolerance to pain induced by in-

  flation of a blood pressure cuff. In general, recovered bulimic pa-

  tients showed higher pain tolerance on both tests as compared to

  controls. To my knowledge, no study has examined the difference be-

  tween purging and nonpurging women with bulimia regarding pain

  tolerance. Since purging bulimics have been through more provoca-

  tion than nonpurging bulimics, I would predict higher pain toler-

  ance in the former group. If true, their higher pain tolerance may

  play a role in their higher suicidality.

  It is potentially important that high pain tolerance remains in

  women with bulimia, even well after they recovered. High pain toler-

  ance in particular and the acquired ability to enact serious self-injury

  in general may be slow to fade. Once in place, these psychological

  features likely endure for quite some time. As will be pointed out in

  the next chapter, this has implications for prevention and treatment

  of suicidal behavior. The acquired ability to enact lethal self-injury

  may be resistant to change, more so than other aspects of the model

  (like perceived burdensomeness and low belongingness). These latter

  qualities thus may be more fruitful targets for treatment and preven-

  tion programs.

  It is interesting to recall that, in general, women have low rates of

  completed suicide. Women who undergo an array of provocative ex-

  periences, however, may be exceptions to the general rule. Patients

  with borderline personality disorder, anorexia nervosa, and, perhaps

  to a lesser degree, bulimia nervosa may represent examples of such

  women.

  Genetics, Neurobiology, and Mental Disorders ● 199

  Of course, mood disorders deserve consideration in any discus-

  sion of suicide. The rates of death by suicide in mood disorders are

  substantial, and this is true for major depression, bipolar I disorder

  (with clear manic and depressive phases), and bipolar II disorder

  (hypomanic and pronounced, recurrent depressive phases).74 Viewed

  through the lens of the model proposed here, high suicide rates

  in mood disorders may be a function of the ability to enact lethal

  self-injury, which is acquired through repeated past experience with

  suicidality and through various provocative experiences associated

  with manic symptoms. Indeed, manic episodes frequently land peo-

  ple in jail, fights, or accidents. Moreover, mood disorders often

  include acute feelings of ineffectiveness and social isolation, a promi-

  nent symptom and associated feature, respectively, of major depres-

  sion. Therefore, those suffering from mood disorders are vulnerable

  on all three of the dimensions emphasized in my model—acquired

  ability for lethality, perceived burdensomeness, and failed belong-

  ingness.

  There is a form of major depression called the atypical subtype.

  This subtype’s symptoms include oversleeping, overeating, and ex-

  treme interpersonal rejection sensitivity. The subtype is labeled

  “atypical” because the symptoms of oversleeping and overeating are

  unusual among depressed people; usually, depressed people lose

  their appetite and have insomnia. With regard to the rejection sensi-

  tivity symptom, it includes reactions to perceived criticisms or

  rebuffs that are so intense that it is difficult to maintain long-term re-

  lationships. New relationships are avoided for fear of potential rejec-

  tion. Belongingness will therefore be a long-standing and vexed issue

  for people with the atypical subtype of depression. There is mixed

  evidence as to whether people with the atypical subtype experience

  higher risk for suicidal behavior than do other depressed people.

  One study found that people with the atypical subtype had more sui-

  cidal ideas and suicide attempts than other depressed people; atypi-

  200 ● WHY PEOPLE DIE BY SUICIDE

  cal depressions also had earlier age of onset than other depressions.75

  Early age of onset is one marker of severity of a disorder, and this

  alone could explain why those with the atypical subtype had more

  suicidality than others. But if it is established that atypicality is asso-

  ciated with suicidal symptoms in a real way, my model would predict

  that this occurs, in part, because people with this syndrome struggle

  so intensely with rejection sensitivity and thus low belongingness.

  Interestingly, rates of suicide are lower for people suffering from

  dysthymia (a low-grade but very chronic form of depression)76 than

  from other depressions. Again viewed from the present perspective,

  this stands to reason, in that the feelings of ineffectiveness and social

  isolation in dysthymia may not reach the level of severity necessary

  to fully instill the desire for death.

  Antisocial personality disorder is interesting to consider in light of

  the model proposed here. The disorder is characterized in the cur-

  rent psychiatric nomenclature as a long-standing pattern of aggres-

  sive behavior and reckless and impulsive disregard for others and for

  rules and norms. However, recent research, informed by classic work

  by the psychiatrist Hervey Cleckley,77 suggests that there are two dif-

  ferent kinds of antisocial personality. One type is characterized by

  emotional detachment (i.e., low anxiety; fake or shallow emotions;

  immunity to guilt and shame; callousness; and incapacity for love,

  intimacy, and loyalty). The other type is characterized by impulsive,

  reckless, and under-controlled behaviors.

  Cleckley reserved the term “psychopath” for those with the cardi-

  nal feature of emotional detachment. Research has demonstrated

  that the two types of antisociality are separable.78 One factor is cur-

  rently emphasized by DSM and prioritizes antisocial behavior. The

  other factor was formerly emphasized in DSM to some degree and

  corresponds to Cleckley’s emphasis on “emotional detachment.” Ac-

  cording to this research, there are two kinds of people with antisocial

  personality—those who are emotion
ally detached (and who are also

  Genetics, Neurobiology, and Mental Disorders ● 201

  prone to poor behavioral control, in part because of their emotional

  detachment), and those who are primarily impulsive, aggressive, and

  irresponsible but who are not emotionally detached (and actually

  may be especially emotionally reactive).

  My colleagues and I predicted that this latter type of individual

  would be prone to suicidal behavior (due to the combination of

  impulsivity and emotional reactivity), but that emotionally detached,

  “Cleckley psychopaths” would not be, due in part to low emo-

  tional reactivity. Our study of 313 inmates supported this predic-

  tion: “antisocial behavior” was associated with history of suicide at-

  tempts; “emotional detachment” was not, and in fact, was negatively

  associated with suicide history, although to a nonsignificant degree.

  Moreover, we found that the link between “antisocial behavior” and

  suicidality occurred in part because antisocial characters were prone

  to the combination of negative emotionality and impulsivity.79

  According to the model of suicidality described here, emotionally

  detached antisocial personalities may not be prone to suicide, be-

  cause their callousness and incapacity for intimacy and loyalty would

  insulate them from perceived burdensomeness and disconnection

  from others. By contrast, antisocial personalities characterized by

  under-controlled behaviors would be at higher risk, because their

  recklessness gives them an opportunity to habituate to pain and in-

  jury, and because their negative emotionality increases the likelihood

  of a sense of burdensomeness and low belongingness.

  Virtually everyone who dies by suicide experienced one or more

  mental disorders at the time of their death. Certain disorders are

  more associated with suicidal behaviors than others, and it is impor-

  tant to recall that relatively few people with a mental disorder die by

  suicide. My model explains these facts by arguing that some mental

  disorders are more likely than others to lay down the ability to enact

  suicide and to instill perceived burdensomeness and failed belong-

  ingness. Those with one of these suicide-related disorders who do

  202 ● WHY PEOPLE DIE BY SUICIDE

  not die by suicide have managed to avoid perceived burdensomeness,

  low belongingness, or acquiring the ability to seriously harm them-

  selves, despite their mental disorder. Certain mental disorders sub-

  stantially increase the likelihood but do not guarantee that the three

  conditions will be present that I propose are required for serious

  suicidality.

  At the moment of conception, a baby’s future is not fully plotted,

  but some of its general trajectories can be discerned. Genes influence

  neurobiology, including the serotonin system. Genes also influence

  personality traits like impulsivity, and this influence may occur

  mostly through genes’ impact on the serotonin system. Genetics,

  neurobiology, and personality all interact in complex ways with an

  individual’s life experience. Early adverse experience, including

  childhood abuse and neglect, heightens the risk for later problems,

  especially in vulnerable people. One set of such problems is men-

  tal disorders, which, in addition to the agony and impairment they

  cause, clearly confer risk to suicidal behavior. Genes, neurobiology,

  impulsivity, childhood adversity, and mental disorders are inter-

  connected strands that converge and can influence whether people

  acquire the ability for lethal self-injury, feel a burden on others, and

  fail to feel that they belong. This lethal endpoint is the culmination

  of processes started at conception and furthered, biologically and

  through experience, over a person’s lifetime.

  RISK ASSESSMENT,

  CRISIS INTER VENTION,

  TREATMENT,

  AND PREVENTION

  6

  Time and again, psychopathology theorists and researchers go to

  great lengths to develop theories and models of psychopathology, but

  then when it comes time to talk about applications like assessment,

  treatment, and prevention, there is a great disconnect between the-

  ory and application. I think this occurs in part because applications

  are often developed without theory in the clinic—on the fly, as it

  were. This is not all bad, because many treatments that are discon-

  nected from theory are very good—and, it must be added, some

  treatments that are awash in theory are not very good at all.

  Some examples of good treatments are interesting to consider.

  A first is called Interpersonal Psychotherapy (IPT) and it was devel-

  oped in the 1970s by the late psychiatrist Gerald Klerman and

  colleagues. IPT is a down-to-earth, here-and-now kind of psycho-

  therapy originally developed for depression but now used for other

  conditions too. Its central idea is that if a major interpersonal issue

  connected to symptom onset is worked out—say, a grief problem or

  a hostile standoff in a marriage—then that is bound to help relieve

  symptoms. IPT also recommends the sensible strategy of staying fo-

  203

  204 ● WHY PEOPLE DIE BY SUICIDE

  cused on one interpersonal issue, trusting that progress made on it

  will generalize to improve other areas too.

  That IPT relieves symptoms is beyond doubt; randomized, con-

  trolled clinical trials have attested to the fact. In an intriguing study,

  IPT was assessed in rural Uganda.1 Thirty Ugandan villages were

  studied. In each village, men or women who were self-identified

  and viewed by other villagers to have symptoms of depression were

  interviewed. In the local language, there is no single term to de-

  scribe depression. Instead the interviewers asked for persons with

  Yo’kwekyawa or Okwekubazida, two depression-like syndromes well known to villagers. These two syndromes together include all the

  major depression symptom criteria in the DSM-IV. Approximately

  eight per village of the most depressed people were selected for par-

  ticipation, totaling around 250. Eight of fifteen male villagers and

  seven of fifteen female villagers were randomly assigned to the ther-

  apy and the remainder to a control group. People in the control vil-

  lages did not receive the therapy; however, people in both control

  and intervention villages were free to seek whatever other interven-

  tions they wished throughout the study. The intervention villages

  received the depression therapy in group meetings for weekly ninety-

  minute sessions for sixteen weeks. Groups were led by a local per-

  son, of the same sex as the group, who had received brief training in

  the therapy. During each session, the group leader reviewed each

  participant’s depressive symptoms, and participants described recent

  events and linked the events to his or her mood. The group leader

  then facilitated supportive statements and suggestions for change

  from other group members. The therapy was very effective. Among

  those who received the treatment, rates of severe depression went

  from around
90 percent before treatment to around 6 percent after

  treatment; by contrast, among those in the control groups, rates of

  severe depression went from around 90 percent before treatment to

  around 55 percent after treatment.

  Assessment, Intervention, Treatment, and Prevention ● 205

  This and other studies show that IPT is effective. But it is remark-

  ably theory-free. As IPT was being developed in the 1970s and there-

  after, a scientific and theoretical literature on the interpersonal as-

  pects of depression was developing too.2 Strangely, these two strands

  of work rarely if ever intersected. IPT’s relative lack of theory has not

  hamstrung it; it works, and additionally, no theoretical errors or

  obfuscations were introduced as part of the treatment description.

  A second example of a good treatment being disconnected from

  theory is the Cognitive Behavioral Analysis System of Psychotherapy

  (CBASP).3 This treatment relies heavily on past work by people like

  Aaron T. Beck and Albert Ellis on cognitive therapy, as well as on the

  field of applied behavior analysis. The gist of CBASP is, in ways, sim-

  ilar to IPT. The idea is to repeatedly focus on specific, discrete situa-

  tions and then to mould one’s thoughts and behaviors so that those

  situations tend to produce one’s goals. Like IPT, the idea is down-to-

  earth, and like IPT, impressive clinical trial data support the treat-

  ment’s effectiveness.4

  The theory behind the treatment, however, is both flawed and

  largely irrelevant to the treatment. The theory makes unfortunate

  and unsubstantiated claims about the nature of depressed people—

  for example, that the chronically depressed individual is “a cognitive-

  emotionally retarded adult child who brings a negative ‘snapshot’

  view of the world to the session. The chronic patient functions, at

  least in the social-interpersonal arena, with the structural mindset of

  a 4–6-year-old preoperational child.”5 I find this a ludicrous claim,

  and would feel even more strongly, I’m sure, if I were a chronically

  depressed adult. Moreover, the claim is not necessary or even very

  relevant to the treatment, which, far from being ludicrous, has been

  shown to be effective and useful.

  The cognitive theorizing and treatment recommendations of Beck