Cerebrum Article

Is Impulsive Aggression the Critical Ingredient?

Our understanding of suicide is changing for the better, but this alone may not be enough to save many young lives. If we equate the treatment of depression and other psychological disorders with prevention of suicide, we may leave out the essential (and likely genetic) aspect of suicide that turns psychopathology and social problems into the terrible act of self-murder. The first lesson of David Brent’s article is that self-destructive behavior that seems as complex as life itself, and as beyond our coping, may become intelligible when viewed as a specific brain disorder. So it may be with suicide by children, adolescents, and young adults, where a swirling constellation of familial, societal, and psychological factors begin to point in a common direction: toward the impulsively aggressive personality. The second lesson is that the role of genetic inheritance in most brain disorders will prove to be complex. Brent traces how the genetic roots of impulsive aggressiveness interact with those of psychiatric illness. In the end, it is more hopeful to think that youth suicide continues at an appalling rate not because we are helpless, but because we only think we understand it.

Published: July 1, 2001

I begin on a note of caution and with a paradox. The note of caution is that when we critically examine the explanations offered for youth suicide, we must do so with deep compassion. A parent, sibling, or friend in the wake of suicide by a child, adolescent, or young adult is enduring the virtually unendurable. To reach for an explanation, a way to come to terms, is human, healthy, and inevitable.

Now for the paradox. There is a new emphasis today on the role of mental illness in youth suicide. This is refreshing after the media myth, widespread in the 1980s, of young suicides as talented, misunderstood young people who were just under too much stress. The paradox is that, although psychiatric conditions like depression or substance abuse almost always accompany youth suicide, we may be sliding into another myth. This myth claims that if we understand mental illness, then we understand suicide and, therefore, can prevent it by identifying and treating the psychiatric disorders. But this may not prevent suicide in many people, particularly young people; we are discovering that suicide is more complex.

First, some background. Youth suicide is a serious public health problem in America, the third most common cause of death among 15-to 24-year-olds. Community surveys have revealed suicidal behavior among 3 to 8 percent of youth, making it a major source of morbidity, health care costs, and ultimate mortality. Suicide and suicidal behavior are often discussed together. The risk factors are similar; there is a high rate of completed suicide among those who have made earlier attempts; both conditions seem to run in families; and there are similar findings in the neurobiology of people who have attempted and completed suicide.

Suicide, especially among youths, has often been discussed from an environmental or sociological perspective. While these are valid viewpoints, they can obscure a crucial biological perspective on suicidal behavior. Better understanding of the biological and particularly the genetic processes that underlie youth suicide has urgent implications for preventing and treating youth at risk for suicide. An analogy can be made to understanding obesity or coronary artery disease. Environment and diet profoundly influence both of those diseases, but advances in prevention can come through understanding the biological processes that lead to illness when certain environmental conditions are present.

Risks—But for Whom?

The rate of youth suicide has increased markedly since 1950, but is this proof that suicidal behavior is primarily a social or an environmental phenomenon? In fact, there is evidence that increasing availability of both firearms and alcohol to youth during this period has contributed to the increase in the rise in their suicide rate. But note that the suicide rate, while unacceptably high, still encompasses a relatively small number of young people. If alcohol and guns are easily accessible, why are these environmental risk factors particularly potent for youth who kill themselves and not so for others also exposed to them? Two complex genetic biological factors may be interacting with risks in the environment: psychiatric disorders and a tendency to impulsive aggression.

Suicide and suicidal behavior almost always point to serious psychiatric disorder.1 Nine out of ten suicide completers and attempters have at least one major psychiatric illness. In half of these cases, two or more such illnesses are present, dramatically increasing the risk for suicidal behavior. The most common psychiatric conditions predisposing to suicide and suicide attempts are mood disorders, alcohol and substance abuse, and behavior disorders—again, often in combination. Also, about one-third of suicide victims have made previous attempts, which increases the likelihood that they will one day succeed. Often the psychiatric conditions have taken hold several years prior to the suicide.

The emphasis on the role of psychopathology in youth suicide is refreshing after media presentations in the 1980s of youth suicides as misunderstood young people under stress. Some prevention programs in schools even taught that a “myth” of suicide prevention was that suicide victims were mentally ill. Yet, if our attempts to understand the causes of youth suicide stop with a delineation of psychopathology, and attempts to treat that mental illness, we may make barely a dent in the number of youth suicides.

Treatment of depression does not automatically translate into prevention of suicidal behavior. Recently, scientists analyzed a large group of studies of depressed adults who were enrolled in clinical trials that compared medication with placebo (“dummy” treatment).2 Despite a 30 percent greater improvement in depressive symptoms in the adults who were medicated, the rates of attempted and completed suicides were similar in the two groups. Here is strong evidence that the treatment of depression may not be sufficient to prevent suicide. Additional studies show a strong effect on preventing repetitive suicidal behavior without showing an improvement in mood, while others, as noted, show a strong impact on mood without influencing suicidal behavior.3 Thus psychopathology may be a necessary, but not a sufficient explanation of suicide and suicidal behavior, and treatment of psychopathology may be necessary but not always sufficient to prevent suicide.

A Surprising Force in Suicide

That leaves us casting about for additional causes. While there are several co-factors for suicidal behavior (for example, the availability of lethal agents in the home or a history of sexual abuse), the single most significant predisposing factor appears to be a liability to impulsive aggression. Impulsive aggression is a tendency to respond with hostility or aggression when faced with stress or frustration. This is sometimes termed “reactive aggression,” as distinguished from “predatory aggression.” When a child is taunted and responds out of anger with physical force, that is reactive aggression. When a child waits for another child to come around the corner so that he can steal his lunch money, that is predatory aggression. Children with attention deficit hyperactivity disorder (ADHD) do have prominent levels of impulsivity, but this is not necessarily associated with reactive aggression.

There is mounting evidence for three facts about impulsive aggression: 1. It is related to suicidal risk; 2. it is strongly correlated with changes in brain levels of the chemical messenger serotonin; and 3. it is genetically transmitted. Suicidal behavior seems to be passed from generation to generation in families. Could this be because impulsive aggression is transmitted? Would a better understanding of impulsive aggression have implications for preventing suicide and treating those at risk for suicidal behavior?

Examining suicide across the life span makes clear that youth suicide is qualitatively different than suicide in persons older than 30.4 Younger victims are more often responding to interpersonal and legal difficulties, have more impulsive aggression and more problems with substance abuse, and are less frequently depressed. (That is, however, only relative to older individuals; depression is still an important contributor to suicide risk among the young.) All this suggests that the problem of impulsive aggression may be particularly salient in understanding youth suicide. Those who treat depressed suicidal young people can attest to the difficulties that can arise from addressing the depression but failing to address the impulsiveness. Their patients may partially recover from the depression but, when faced with stress, may try again to commit suicide.

There are important public health implications of the more impulsive nature of youthful suicidal behavior. For example, people under age 24 are most susceptible to media influences that increase suicidal behavior; they are more likely than older individuals to commit suicide around the same time and in the same community as other youth suicides. We call this a time-space cluster.1 The availability of firearms may be an especially prominent risk factor for completed suicide among younger victims, probably because suicide is a more impulsive act in the young.1, 3 This in turn suggests that for younger, more impulsive individuals, restricting of access to firearms and means of self-destruction— at a minimum securing them in households where youth live—may be a relatively important component of suicide prevention.

In any case, there is good evidence that impulsive aggression is much more common in those who complete or attempt suicide than in demographically similar individuals. Even where people have similar psychiatric risk factors, those with impulsive aggression are at greater risk of suicide. Impulsive aggression may be related to substance abuse, which can further disinhibit an already impulsive person, with disastrous consequences. And drinking alcohol makes the use of a gun in a completed suicide much more likely.3

On the biological level, studies have shown a remarkable convergence between impulsive aggression and suicidal behavior, with both behaviors apparently correlated with alterations in serotonin in the central nervous system.5 Serotonin levels have also been correlated with impulsive aggressive behavior in non-human primates. Postmortem studies in humans have found alterations in serotonin receptors in the brains of suicide victims; the brain region showing the most striking changes is the orbital prefrontal cortex, which is intimately involved in the exercise of restraint. In fact, one PET study of impulsive murderers showed diminished metabolic rate in the prefrontal cortex.7

Both impulsive aggression and serotonin levels are controlled by our genes. Studies in twins have shown that genes explain around 40 percent of impulsive aggression; studies of non-human primates show that around 40 percent of the variance in serotonin levels is influenced by genes. Several studies have related variations in genes to variations in serotonin transmission in the brain and also to suicide attempts and impulsive aggression.

To what extent is suicidal behavior familial (and perhaps genetic), and do suicide and impulsive aggression appear to run together in the same families? 

Making the Lethal Connections

So we know that impulsive aggression appears to be partly genetic and a logical culprit in suicide risk. But is it related to genetic risk for suicide? To what extent is suicidal behavior familial (and perhaps genetic), and do suicide and impulsive aggression appear to run together in the same families?

The evidence to support suicide running in families comes from adoption, twin, and family studies.6 Adoption studies look at how frequently suicide occurs in both the biological and adoptive relatives of adopted people who committed suicide, and then compare that with the frequency in biological and adoptive relatives of adopted people who did not commit suicide. Twin studies look at how frequently suicidal behavior occurs in both identical twins, and then compares this with the frequency in fraternal twins. If a condition is genetic, both identical twins should show suicidal behavior more frequently than both fraternal twins would, since identical twins completely share the same genes, while fraternal twins on average share only half their genes. Family studies look at the frequency of suicide or suicide attempts in the relatives of suicide victims or suicide attempters, compared to the frequency in the relatives of similar young people who do not exhibit suicidal behavior. An increased rate of suicidal behavior in the relatives of suicide victims compared to the rate in relatives of controls would point to family transmission of suicidal behavior, but would not differentiate between genetic and environmental causes of that transmission. For example, familial suicide rates might be elevated because of a true genetic predisposition to suicide, but it might be because of other shared family characteristics such as poverty, changing neighborhoods frequently, or discord. Let us look at some results of these three kinds of studies.

An adoption study conducted in Denmark shows strong evidence of some type of genetic effect for suicide. Adopted individuals who had committed suicide were matched with adopted individuals still living; then the rates of suicide were compared in their biological and adoptive relatives. The rate of suicide in the biological relatives of the suicide adoptees was six times higher than the rate of suicide in the biological relatives of the living adoptees, but there was no difference in the rates of suicide in the two sets of adoptive relatives. This strongly supports the theory of genetic causes for suicide, and also provides evidence against imitative behavior being an important component, since the adoptive relatives had no effect on the rate. What is less clear is what exactly is being transmitted. Is it an increased risk for a particular type or severity of psychiatric disorder, or some other trait?

Another adoption study by the same researchers examined the rate of suicide in the biological relatives of adoptees who had mood disorders. They found that biological relatives of adoptees with mood disorders had a 15-fold increased risk for suicide. What was curious is that the subgroup of adoptees whose relatives had the highest risk for suicide were those with borderline personality disorder (called “affect reaction” in Denmark). This is a condition that often occurs with mood disorders, but is not itself considered a mood disorder. Instead, it characterizes people with difficulties regulating their emotions and impulses—that is, with impulsive aggression. This observation suggests that familial transmission of suicidal behavior may be related to familial transmission of impulsive aggression. The suicides in this study were not further described, so we do not know whether the biological relatives of the adoptees who committed suicide were also impulsively aggressive.

Twin studies also point to a role for genes in suicidal behavior. Reviewing all reported cases where both twins committed suicide, a researcher showed that this occurred much more often in identical twins than in fraternal twins. Taking it a step further, this researcher showed that the rate of attempted suicide among those who had survived a twin’s suicide was higher in the identical twins. This is important. It illustrates that attempted and completed suicides fall along the same spectrum, and that examining both gives us the genetic component of suicidal behavior. The main weakness of these particular twin studies is that they are based on case reports in the literature, so are not necessarily representative of all twins.

Look next at a large, representative study of almost 3,000 Australian pairs of twins. Again, both identical twins were much more likely to attempt suicide than both fraternal twins. If a twin had made a suicide attempt, the other twin was at nearly an 18-fold increased risk for a suicide attempt. Serious suicide attempts were highly heritable, with 55 percent of the variance explained by genetic factors. There was no clear relationship between the timing of the suicide attempts by the sets of twins, suggesting that the cause was not imitation Even after controlling for other risk factors for suicidal behavior—mental illness, a history of abuse, personality problems, and exposure to stress—family history of suicide attempts persisted as a strong predictor of attempts in the remaining twin. This provides some of the most compelling evidence to date that familial transmission of suicidal behavior and of mental illness are separate strands. Unfortunately, this study once again shed no light on exactly what is being transmitted that predisposes a person to suicidal behavior.

Over the past two decades, about a dozen family studies have examined the rate of suicidal behavior in the relatives of either suicide attempters or completers and then compared it to the relatives of controls.1, 6 Several consistent findings emerged. First, the rate of suicide attempts is increased in the families of both suicide completers and suicide attempters. This further bolsters the view that what is being transmitted in families is a tendency toward suicidal behavior, rather than to suicide per se. Interestingly, suicidal thinking was not transmitted along with suicidal behavior, but instead was transmitted along with depression.

Second, the transmission of suicidal behavior cannot be explained by the transmission of psychiatric illness alone. The distinctness of the familial transmission of mood disorders and of suicide was best illustrated by a study of the Old Order Amish, which reported some genetic family trees that were loaded with mood disorders yet had not a single suicide, whereas others, also loaded with mood disorders, showed rates of suicide 100 times the expected rate. Several other studies showed that, while the rate of psychiatric disorder was increased in the relatives of suicide completers or attempters, their increased rate of suicidal behavior persisted even after controlling for the increased rate of psychiatric disorder.

Third, several studies support the initial observation in the Danish adoption study that the tendency to suicidal behavior is transmitted along with a tendency to impulsive aggression. Relatives of suicide attempters or completers who were more aggressive or who had made suicide attempts by violent methods were much more likely themselves to have engaged in suicidal behavior.6 This suggests that what families are transmitting is the tendency to convert suicidal thinking into action. An ongoing study comparing the risk of attempted suicide in the offspring of people with mood disorders who attempted suicide with a group of people with mood disorders who never attempted suicide shows patterns of familial transmission largely consistent with these conclusions.

I have dwelled at some length on these studies because they suggest not only the strong correlation of inherited impulsive aggression with suicide, but the interaction of multiple genes in complex brain disorders.

Other Possibilities

Are there alternative explanations of these observations? Could suicidal behavior run in families as a result of imitative behavior or of the accumulated impact of loss or psychiatric disability, or of shared family adversity, such as physical or sexual abuse?

Imitative suicidal behavior cannot be dismissed. Media exposure to fictional and nonfictional stories of suicide is followed by approximately a 10 percent increase in the suicide rate, usually for about two weeks 1, 7 The greater the publicity, the more marked the effect, particularly among adolescents and young adults. On the other hand, there is no indication of a higher rate of suicide or suicide attempts among friends and siblings of suicide victims. In fact, they seem to be inhibited from engaging in suicidal behavior by their exposure to the grief of the friends and families of the suicide victims. It may be that imitation is more likely to occur if the exposure is by hearsay, through the media, or by some other, less intimate contact. The absence of any imitative effect in adoption studies, or in the one twin study that examined this question, makes imitation less likely to explain familial clustering. On the other hand, exposure to a friend’s suicide attempt (rather than completed suicide) may increase the risk for suicide attempt in a young person, so that in theory, this mechanism could be at play. The studies published thus far do not support imitation as an explanation for familial transmission of suicidal behavior.

Could suicidal behavior be transmitted in families through the impairment or disability of psychiatrically-ill parents? Mental illness of parents, particularly depression and substance abuse, does increase the risk of youth suicide.1 Since the transmission of suicidal behavior persisted in several studies after controlling for rates of mental illness among relatives, however, parental disability alone is not a good explanation for the familial transmission of suicidal behavior.

Perhaps the most plausible alternative explanation for a family’s transmission of suicidal behavior is the experience of adversity shared by members of a family living in the same environment.1 Poor parent-child communication is a risk factor for completed suicide, even after controlling for parental psychopathology. In another study, parent-child discord was a risk factor for completed suicide, although after controlling for both parental and child psychopathology, it was not a significant contributor. Maltreatment, however—physical, sexual, or emotional abuse—is a potent risk factor for completed and attempted suicide,1 and contributes to suicidal risk even after controlling for other psychopathological risk factors. It appears that sexual abuse is a stronger risk factor for attempted suicide than physical abuse, and that the risk for suicide attempts is proportional to the severity of the abuse, with the greatest risk associated with more severe abuse, such as vaginal penetration.

Given that there are associations between parental depression, substance abuse, suicide attempt, and abusive behavior, the familial transmission of abuse and suicidal behavior may not be an either/or proposition. That is, genetic factors related to psychopathology and impulsive aggression may make a parent more likely to be abusive, thereby giving the child a double dose of risk through both the genetic endowment associated with the abuse and the trauma of the abuse itself. Depression appears to emerge in abused children as the result of interaction between a family history of depression and the stress of the abuse, 8 and there is no reason to believe that the same paradigm may not hold in the transmission of suicidal behavior as well.

These findings of the familial transmission of suicidal behavior are entirely consistent with the serotonin hypothesis of suicidal behavior, which says that alterations in central serotonin are associated with both attempted and completed suicide and with impulsive aggression; these alterations appear in part to be under genetic control.5, 6 While the results of studies to search for genes that may underlie suicidal behavior are not entirely consistent, they suggest that variations in serotonin-related genes (two of these are tryptophan hydroxylase, or TPH, and monoamine oxidase A, or MAOA) are related to changes in central serotonin metabolism, suicide attempts, and greater likelihood of impulsive aggression. What is particularly interesting about some of these gene studies is that the effects may be most prominent in men, who complete suicide much more commonly than for women do. Moreover, the serotonin hypothesis could even be considered consistent with a prominent role for abuse in the familial transmission of suicidal behavior, since it has been demonstrated in human and nonhuman primates that adverse early environment can interfere with central serotonin metabolism.9, 10

Because our ability to predict and prevent suicide is still limited, these research results make a compelling case for trying to understand the genetics of suicidal behavior and its relationship to impulsive aggression and family adversity. Genetic linkage strategies examining personality traits such as impulsive aggression may shed light on the genetic contribution to suicidal behavior. Moreover, examining changes in genetic expression, such as in postmortem samples of suicide victims, may also provide insight into which genes are involved in suicidal behavior. On a clinical level, the prominent role of impulsive aggression in youth suicidal behavior supports the idea that we should assess and target this behavior in the treatment of young people at risk for suicide.

By combining state-of-the-art treatments for both impulsive aggression and predisposing conditions like depression, we may ultimately be more successful in preventing youth suicide. For example, there are now studies suggesting that lithium may be helpful in decreasing aggression, independent of its effect on mood, so it may have an important role in decreasing aggression and lowering risk for suicide. A process called dialectical behavior therapy (DBT) has been shown in some studies to reduce the risk of suicidal behavior in impulsively aggressive women. Both of these promising treatments have yet to be evaluated in young people at high risk for suicide.

As we recognize our limited ability to detect and intervene with impulsive aggression in children, adolescents, and young adults, we may be able to consider other means of preventing suicide, such as restricting lethal agents (such as a loaded gun) from homes of vulnerable youth and avoiding sensationalistic publicity about suicide that is likely to trigger imitative acts among vulnerable youth. Currently, however, physicians frequently do not ask about firearms in the home, and when at-risk patients are counseled about removing firearms, the results are disappointing. Therefore, further work needs to be done about the best way to separate firearms from young people at risk. Work is ongoing with leaders in the media aimed at developing realistic journalistic standards that result in stories that are informative, but avoid unnecessary sensationalization.

The Talmud says that “he who saves a life, it is as if he has saved the entire universe.” This is particularly so in preventing suicide by young people, because the tragedy of youth suicide is that an impulsive act can end a lifetime of possibilities. Not only does suicide in a young person cruelly truncate what should be a normal lifespan, it leaves in its wake family and friends whose grief is often unconsolable. Although we still have much to learn, most teens who take their lives have treatable mental illnesses. By awakening public awareness of this problem, by improving the astuteness of detecting and properly treating the underlying problems, and by conducting innovative research about the causes of suicide, we can reverse the loss of a lifetime.


  1. Beautrais A. Risk factors for suicide and attempted suicide among young people. Australia and New Zealand Journal of Psychiatry. 2000;34:420-436.
  2. Khan A, Warner HA, Brown WA. Symptom reduction and suicide risk in patients treated with placebo in antidepressant clinical trials: An analysis of the food and drug administration database. Archives of General Psychiatry. 2000; 57:311-317.
  3. Brent DA. Assessment and treatment of the youthful suicidal patient. Annals of the New York Academy of Science. 2001; 932:106-131.
  4. Rich CL, Young D, Fowler RC. San Diego suicide study. I. Young vs old subjects. Archives of General Psychiatry. 1986;43:577-582.
  5. Mann JJ. The neurobiology of suicide. Nature Medicine. 1998;4:25-30.
  6. Mann JJ, Brent DA, Arango V. The neurobiology and genetics of suicide and attempted suicide: A focus on the serotonergic system. Neuropsychopharmacology. 2001; 24(5): 467-477.
  7. Gould MS, Shaffer D. The impact of suicide in television movies: Evidence of imitation. New England Journal of Medicine. 1986; 315:690-694.
  8. Kaufman J, Birmaher B, Brent D, et al. Psychopathology in the relatives of depressed-abused children. Child Abuse Neglect. 1998: 22:171-181.
  9. Pine D, Coplan J. Neuroendocrine response to D.1-fenfluramine challenge in boys: associations with aggressive behavior and adverse rearing. Archives of General Psychiatry. 1997; 54: 839-846.
  10. Kraemer GW, Ebert MH, Schmidt DE, et al. A longitudinal study of the effect of different social rearing conditions on cerebrospinal fluid norepinephrine and biogenic amine metabolites in rhesus monkeys. Neuropsychopharmacology. 1989; 2: 3:175-189.