Why People Die By Suicide

by Thomas Joiner

fact that I am a surviving child of a dad who died by suicide adds

  credibility, I think, to the reassurance.

  Regarding the involvement of genes in behavior, twin samples are

  the most useful to study; they can help determine whether a genetic

  contribution to some trait or behavior exists. A usual strategy is to

  compare monozygotic, or identical, twin pairs, who share all of their

  Genetics, Neurobiology, and Mental Disorders ● 175

  genes, to dizygotic, or fraternal, twin pairs, who share on average half

  of their genes (as do any siblings who are not identical twins). If

  genes are involved, identical twins should share the trait or behavior

  more often than fraternal twins, because identical twins share all

  their genes, and fraternal twins share approximately half of their

  genes.

  A potential complication is that the family environment may also

  be more similar for identical twins than it is for fraternal twins, be-

  cause identical twins may be treated more similarly (e.g., dress alike)

  than fraternal twins. So, an even better strategy is to study twins

  separated at or near birth and reared apart from one another, as

  might happen when twins are adopted by different families. How-

  ever, because the confluence of twinship, adoption, and later suicide

  is rare, no twin adoption study on suicide has been conducted, to my

  knowledge. There have been informative nontwin adoption studies,

  however.

  Overall, twin studies have found that 13 to 19 percent of identical

  twin pairs were concordant for death by suicide as compared to less

  than 1 percent of the dizygotic (DZ) twin pairs, a significant differ-

  ence.2 Given that one twin has died by suicide, this means that the

  chances that the other will die by suicide are around 15 percent for

  identical twins and less than 1 percent for DZ twins. One percent is

  an elevated rate, it should be noted; the probability of any given indi-

  vidual in the United States dying by suicide is around .01 percent.

  Therefore, 15 percent is an extremely elevated rate.

  Researchers in Denmark have used an adoption register to

  study the genetic aspects of suicide. From a register of thousands of

  adoptions, they identified fifty-seven who eventually died by suicide.

  These fifty-seven were compared to fifty-seven matched adopted

  controls who had not died by suicide, specifically with regard to fam-

  ily history of suicide among their biological relatives. Over 4 percent

  of the biological relatives of the suicide group had themselves died by

  176 ● WHY PEOPLE DIE BY SUICIDE

  suicide, as compared to well under 1 percent of the biological rela-

  tives of the control group.3

  Further evidence of the role of genetics in suicide is shown through

  family studies. One early study examined the Old Order Amish over

  a hundred-year period. During this time there were twenty-six peo-

  ple who died by suicide, the majority of whom came from only four

  families. Interestingly, while these four families had a high genetic

  loading for depression in addition to suicide, other families had a

  similarly high loading for depression but no suicides. This suggests

  that the genetic component for suicide may be independent from the

  genetic component for depression.4 Other studies, too, have pointed

  to a unique genetic contribution to suicide, over and beyond the

  genetic contribution to mental disorders like depression.5 This is a

  key point about the relation of mental disorders to suicidality—

  mental disorders, though very important in understanding suicid-

  ality, do not fully explain it. Further, a simplistic view of the asso-

  ciation between mental disorders and suicidality does not explain

  why most people with mental disorders do not attempt or die by

  suicide.

  The twin and adoption studies converge to show that genes are in-

  volved in suicidal behavior. To return to the issue of the serotonin

  system, suicide’s “shoeing horn,” we are now honing in on specific

  genes that may confer risk for suicidal behavior. One gene that has

  received much attention is the serotonin transporter gene. As noted

  earlier, the neurochemical serotonin is important in mood, sleep,

  and appetite. The serotonin transporter maintains control over the

  availability of serotonin in the synapse, essentially by acting as its

  recycler—the transporter recycles serotonin back up into the neuron

  after serotonin is released into the synapse. SSRI drugs used to treat

  depression like Prozac, Zoloft, and Paxil exert their effects by shut-

  ting down or inhibiting the action of the transporter, one effect of

  which is to leave more serotonin “in play” in the synapse.

  Genetics, Neurobiology, and Mental Disorders ● 177

  A single gene is responsible for encoding, or for the “architectural

  plans,” of the transporter—the serotonin transporter gene. In hu-

  mans, this gene is located on chromosome 17. A region of this gene

  has been identified as having what is called a “polymorphism” in it. A

  polymorphism just means that something can take multiple forms.

  In the case of the serotonin transporter gene, there are two possible

  forms, depending on the presence or absence of an additional string

  of gene building blocks in the gene sequence. Each of the two varia-

  tions is referred to as an allele of that gene. If an allele has the inser-

  tion, it is called a long allele; if it does not, it is a short allele.6 Since all humans carry two copies of each gene, there are three possible combinations of the two alleles: two long alleles ( l/l), a long allele and a short allele ( l/s), or two short alleles ( s/s).

  There is some emerging consensus that those with the s/s genotype

  have more dysregulated serotonin systems and thus are more prone

  to attendant problems. A recent study that followed 103 suicide at-

  tempters over the course of a year found that the s/s genotype was more common in people with higher numbers of suicide attempts.7

  A postmortem study found that the s/s genotype was more common

  among suicide victims than among others, although this difference

  did not reach statistical significance.8 My colleagues and I reported

  that those with a significant family history of suicide were more likely

  to have the s/s genotype than were those without a family history.9

  This latter study is of personal interest to me, not only because it

  was my study, but also because I have the s/s genotype and have a significant family history of suicide, having lost my dad to suicide. I do

  not know my dad’s genotype, but he had to have had at least one s allele, because people get one allele from their mother and one from

  their father. Since I have two s alleles, my dad had to have had at least one himself, as must my mom, making both either s/s or s/l. It is not possible to know for sure which my dad was, though if I tested my

  sisters and both were, like me, also s/s, the likelihood increases that

  178 ● WHY PEOPLE DIE BY SUICIDE

  my dad was s/s. Given the research, my guess is that my dad did carry the s/s genotype.

  There are other serotonin-system genes besides the serotonin

  transporter. Perhaps the one that has received the most overa
ll atten-

  tion is the tryptophan hydroxylase (TPH) gene. Tryptophan is a

  precursor, or ingredient, of serotonin. TPH breaks down tryptophan

  and thus serves as a kind of braking system in the making of seroto-

  nin. This gene is located on chromosome 11, and two polymor-

  phisms in particular have been studied: A218C and A779C. The A and C represent different alleles (sort of like “long” and “short” on the serotonin transporter gene), and the numbers 218 and 779 represent locations on the chromosome.

  A meta-analysis (pooled results across studies) of the association

  between the A218C polymorphism and suicidal behavior found that

  presence of the 218A allele was significantly related to increased risk for suicide.10 Other studies have examined the A779C polymorphism, and its relationship to suicide is less clear.11

  One final gene that deserves mention is not a serotonin-system

  gene—the catechol-O-methyltransferase (COMT) gene has only re-

  cently been studied with regard to suicide. Somewhat similar to the

  relation of TPH to serotonin, COMT is responsible for breaking

  down neurochemicals like dopamine and norepinephrine, and thus

  can be viewed as a braking mechanism for these neurotransmitter

  systems. A gene on chromosome 22 codes for COMT activity and oc-

  curs in two variations, the H allele and the L allele, which trigger high or low COMT activity, respectively. As with most other research on

  candidate genes, results have been mixed. One study identified no

  difference in COMT genotype between patients at high risk for sui-

  cide and controls.12 However, other studies have suggested that varia-

  tions in the COMT gene are associated only with violent suicide.

  In one sample, the L allele was more frequent in violent suicide

  attempters versus nonviolent attempters and nonattempters. The

  Genetics, Neurobiology, and Mental Disorders ● 179

  nonviolent suicide attempters and nonattempters showed no differ-

  ence in COMT genotype.13 A similar study stratified the results by

  gender and found that the L allele was more frequent in males with a history of suicide attempts but not in females. Furthermore, males

  who carried the L allele were more likely to have made violent suicide attempts and more attempts overall, but this relationship did not

  hold for females.14

  In summary, twin, adoption, and family studies of suicidality have

  clearly shown that there is a genetic component to suicidal behavior.

  This genetic risk for suicidality appears to be partly independent of

  risk for mental illness. Several candidate genes for the transmission

  of suicide risk have been identified. The serotonin transporter gene,

  the TPH gene, and the COMT gene have all shown links to suicidal

  behavior, at least in some studies. At the same time, it is important to

  note that suicidal behavior is too complex to be accounted for by any

  one gene; the analysis of the effects of multiple genotypes in combi-

  nation may help to differentiate levels of genetic risk.

  Neurobiology

  Neurobiological variables also implicate the serotonin system in sui-

  cidal behavior. One of the most well-replicated findings involves 5-

  hydroxyindoleacetic acid (5-HIAA), which is the major metabolite

  of serotonin; that is, when the body breaks down serotonin, one of

  the main things it breaks it down into is 5-HIAA. Studies have found

  low levels of 5-HIAA in the spinal fluid of suicidal individuals. A

  meta-analysis (review of pooled studies) examining 5-HIAA, as

  well as metabolites of other neurotransmitters like dopamine and

  norepinephrine, found consistent evidence for lowered 5-HIAA in

  suicide attempters and completers but no evidence for consistent

  changes in other metabolites.15 This suggests that the serotonin sys-

  tem specifically is linked to suicidality, whereas other neurotransmit-

  180 ● WHY PEOPLE DIE BY SUICIDE

  ter systems may not be, at least not as strongly. A subsequent review

  came to similar conclusions and also indicated that low levels of 5-

  HIAA in suicide attempters is predictive of subsequent attempts.16

  Another approach to documenting abnormalities in the serotonin

  system is by administration of what is called a fenfluramine chal-

  lenge. Fenfluramine stimulates serotonin release. Results have gener-

  ally shown a decreased release of serotonin in suicide attempters ver-

  sus depressed patients and controls, indicating less serotonin activity

  despite the fenfluramine challenge. It is noteworthy that this applies

  to suicidal patients specifically, even as compared to depressed pa-

  tients.17 Those who attempted suicide by more lethal means show de-

  creased activity in an area of the brain called the prefrontal cortex, as

  compared to low-lethality attempters, and this was particularly ap-

  parent after the fenfluramine challenge.18 The prefrontal cortex may

  be involved in impulse control. These high-lethality attempters also

  show decreased serotonin release in response to the fenfluramine

  challenge, as compared to low-lethality attempters.19

  Still another method of evaluating the serotonin system’s role in

  suicide is by postmortem analysis of the brains of individuals who

  have died by suicide. This area of research is not as clearly defined

  as the 5-HIAA and fenfluramine literature, as some have found no

  difference in important serotonin-system parameters between those

  who died by suicide and those who died by other means.20 However,

  a postmortem study found decreased serotonin transporter binding

  in the prefrontal cortex of those who died by suicide.21 This means

  that the serotonin transporter was not working optimally in those

  who died by suicide; note also that the relevant brain area, the pre-

  frontal cortex, was the same as that identified in a previous study as

  important in suicide. In this study, as in others, results were specific

  to suicide as compared to major depression—emphasizing that the

  genetic vulnerability to suicide is distinct from the vulnerability to

  other conditions, even including depression.

  Genetics, Neurobiology, and Mental Disorders ● 181

  Another interesting angle to the association of the serotonin sys-

  tem to suicidality involves sleep. Serotonin appears to play a sig-

  nificant role both in suicide and in the regulation of sleep.22 The re-

  lease of serotonin is highest during waking states, reduced during

  slow-wave sleep, and lowest during REM sleep. Interestingly, seroto-

  nin-system dysfunction, particularly a reduction in the synthesis of

  serotonin, is believed to promote wakefulness.23

  Several studies have also demonstrated that disturbed sleep is

  related both to suicide attempts and to completed suicide.24 One of

  the first studies to examine sleep, depression, and suicide over time

  found that symptoms of global insomnia were more severe among

  those who later completed suicide within a thirteen-month period.25

  Depressed patients with self-reported repetitive and frightening

  dreams are more likely to be classified as suicidal compared to those

  without frequent nightmares.26 A similar relationship recently

  emerged in an impressive study conducted in Finland.
The study re-

  vealed a direct association between nightmare frequency at one point

  in time and completed suicides roughly fourteen years later. Com-

  pared to subjects reporting no nightmares, those reporting occa-

  sional nightmares were 57 percent more likely to die by suicide.

  Among those with frequent nightmares, the risk for suicide increased

  dramatically; individuals reporting frequent nightmares were 105

  percent more likely to die by suicide compared to individuals report-

  ing no frightening dreams.27

  My colleagues and I recently studied this issue at the FSU Psychol-

  ogy Clinic. Among a large sample of psychotherapy outpatients, we

  assessed the associations of sleep problems to suicidal symptoms.

  Our results indicated that insomnia, nightmare symptoms, and sleep-

  related breathing problems collectively predicted suicidal ideation,

  but that nightmare symptoms were uniquely associated with suicidal

  ideation, whereas insomnia and sleep-related breathing problems

  were not. Put differently, nightmare problems were clearly related to

  182 ● WHY PEOPLE DIE BY SUICIDE

  suicidality; on the other hand, the only reason insomnia and breath-

  ing problems appeared to be related to suicidality is because they

  were more common in those who had frequent nightmares—they

  had no independent influence on suicidality.28

  The specific association of nightmares to suicidality is interesting

  to consider in light of the framework developed in this book. People

  who have frequent nightmares, especially those in which they are

  subjugated or victimized, often have the thought, “I’m ineffective

  and powerless even in my sleep.” Insofar as ineffectiveness is a general

  quality of which perceived burdensomeness is a severe subset, night-

  mares may relate to suicidality partly as a function of general feelings

  of ineffectiveness. Also, those who are having nightmares often dis-

  turb the sleep of their partners, which could have implications for

  belongingness. In fact, this seemed to have been an issue for my dad,

  whose snoring (a sleep-related breathing symptom) was problem-

  atic. In our study at the FSU clinic, sleep-related breathing symptoms

  were not uniquely associated with suicidality; however, that study

  used a very rough measure of sleep-related breathing problems, and