Every year approximately 30,000 people die by suicide in the United States, and one million worldwide. Approximately 650,000 people yearly receive emergency treatment after attempting suicide in the United States. It is the third leading cause of death among American youths and the eleventh for Americans of all ages. Over the last 100 years suicides have out-numbered homicides by at least 3 to 2. Almost 4 times as many Americans died by suicide than in the Vietnam War during the same time period. The rates of suicide are exceptionally high among certain populations: white males over 75 years of age, Native Americans, and certain professions (e.g., health professions, police). The rates among youth are rising.
For decades, the federal government of the United States has been concerned about high suicide rates. Thirty years after the first national effort was established at the National Institute of Mental Health in 1969, the Surgeon General of the United States issued a “Call to Action to Prevent Suicide.” Soon after, a National Strategy for Suicide Prevention (2001) presented a comprehensive assessment of future goals and objectives to combat suicide. Several federal agencies (the National Institute of Mental Health, the National Institute of Drug Abuse, the Veterans Administration, the Centers for Disease Control and Prevention, Substance Abuse and Mental Health Services Administration, and the National Institute on Alcohol Abuse and Alcoholism) joined together to ask the Institute of Medicine to convene the Committee on Pathophysiology and Prevention of Adolescent and Adult Suicide to examine the state of the science base,
gaps in our knowledge, strategies for prevention, and research designs for the study of suicide.
RISK AND PROTECTIVE FACTORS
Biological, psychological, social, and cultural factors all have a significant impact on the risk of suicide. The report reviews many of these risk factors individually, but the Committee emphasizes the need for an integrated understanding of their influence.
Over 90 percent of suicides in the United States are associated with mental illness and/or alcohol and substance abuse. Yet it is important to remember that as many as 10 percent of people who complete suicide do not have any known psychiatric diagnosis. This percentage appears higher still in non-Western societies such as China. It is also important to remember that over 95% of those with mental disorders do not complete suicide. The relationship between suicide and mental illness is a conundrum. Suicidality, although clearly overlapping the symptomatology of the associated disorders, does not appear to respond to treatment in exactly the same way. Depressive symptoms can be reduced by medicines without reduction in suicidality. And psychotherapy can reduce suicide without significant changes in affective symptoms.
Over 30 years of research confirms the relationship between hopelessness and suicide across diagnoses. Hopelessness can persist even when other symptoms of an associated disorder, such as depression, have abated. Impulsivity, especially among youth, is increasingly linked to suicidal behavior. Resiliency and coping skills, on the other hand, can reduce the risk of suicide. Research suggests that coping skills can be taught.
Biological changes are associated with completed and attempted suicide. Abnormal functioning of the hypothalamic-pituitary-adrenal (HPA) axis, a major component of adaptation to stress, has been shown to be promising for the prediction of future suicides, but not consistently for suicide attempts. Serotonergic function is reduced and noradrenergic function is altered in the central nervous systems of both suicide attempters and completers. Several lines of evidence, including adoption, twin, and family studies, point to a link between genetic inheritance and risk of suicide. Having a first-degree relative who completed suicide increases an individual’s risk of suicide 6-fold. The genetic liability may be linked to the heritability of mental illness and/or impulsive aggression. Since the heritability of liability to suicidal behavior appears to be on the order of 30–50%, interactions with social and cultural influences must also be significant.
Childhood trauma has emerged as a strong risk factor for suicidal behavior in adolescents and adults. The official rate of child victimization in 1999 was about 12 per 1000 and but the rates are even higher in surveys of children and parents. Independent of psychopathology and other known risk factors, child sexual abuse has been reported in 9–20 percent of suicide attempts in adults.
Social support is a protective factor. Those who enjoy close relationships cope better with various stresses, including bereavement, job loss, and illness, and enjoy better psychological and physical health. Divorced, separated and widowed persons are more likely to die by suicide. Being a parent, particularly for mothers, appears to decrease the risk of suicide. Discord within the family is correlated with increased suicide, while parental and family connectedness has a protective effect, especially among youth. Participation in religious activities is a protective factor for suicide, perhaps in part because of the social support it affords, but belief structures and spiritualism may also be important factors. Furthermore, epidemiological analyses reveal that occupation, employment status, and socioeconomic status affect the risk of suicide. These factors are not specific to the United States or even to the Western world, but global.
At both the individual and collective levels, the suicide rate has long been understood to correlate with cultural, social, political, and economic forces. Political coercion or violence can increase suicide. For example, in the former Soviet Union, areas experiencing sociopolitical oppression (the Baltic States) and forced social change (Russia) had higher suicide rates compared to other regions. Twenty years of civil war in Sri Lanka are associated with increased suicide rates, as well. Vast social changes associated with modernization and globalization are thought to break down traditional values and practices and result in increased suicide. Since the fall of the Soviet Union and its Eastern European satellite states much of East Europe has reported worsening health statistics, including adult mortality, alcohol and drug abuse, and some of the highest suicide rates in the world today. Here suicide is correlated with societal breakdown, severe economic dislocation, major cultural change and new political systems. Suicide carries a social and moral meaning in all societies. A society’s perception of suicide, or its stigma, can influence its rate, preventing suicide in those societies and social groups where it is frowned upon but increasing suicide where it is a culturally acceptable option in certain situations.
Despite the extensive knowledge that research has provided regarding these risk and protective factors, we are still far from being able to integrate these factors so as to understand how they work in concert to evoke suicidal behavior or to prevent it. We do not understand why sig-
nificant proportions (20–49 percent) of maltreated children do not display suicidal symptoms or why the majority of individuals affected with mental illness do not complete suicide. We also need to understand the large numbers of people who commit suicide in the absence of pathology, how suicide varies with social and cultural forces, and how it relates to individual, group, and contextual experiences.
Without a combination of a population-based approach and studies at the level of the individual patient within higher risk sub-groups, macrosocial trends cannot be related to biomedical measures. Most existing studies are retrospective or cross-sectional, involve few correlates, and do not address prediction of risk. Without specific data from well-defined and characterized populations whose community-level social descriptives are well known, normative behavior and abnormality cannot be estimated.
Suicidality can be treated. There is evidence that lithium treatment of bipolar disorder significantly reduces suicide rates. In fact, lithium may have specific anti-suicidal effects for people with this disorder since these effects may be separate from its antidepressant and antimanic effects. Rates are reduced only while the patients take lithium; after discontinuation of lithium, the rates begin to rise to levels similar to those seen prior to lithium treatment. Despite the encouraging evidence, the protective effects of lithium are not consistent across studies. Other psychiatric medications, including anti-psychotic (especially clozapine) and antidepressants, also show promise for the reduction of suicide. A correlation has been observed between an increase in prescription rates for antidepressants, in particular the serotonin re-uptake inhibitors (SSRIs), and a decline in suicide rates in a number of countries. However, randomized clinical trials with antidepressants have failed to reveal significant differences versus placebo, perhaps due to methodological limitations.
Medications alone are not sufficient for treating mental disorders or suicidality, nor are treatments equally effective across individuals and diagnoses. Psychotherapy provides a necessary therapeutic relationship that reduces the risk of suicide. Cognitive-behavioral approaches that include problem-solving training seem to reduce suicidal ideation and attempts more effectively than treatment as usual or nondirective therapy.
Patients are at much greater risk of suicide in the weeks immediately following discharge from the hospital. Discharged patients who committed suicide were 3.7 times more likely to have had their outpatient care reduced at their last session. On the other hand, patients who continued
care either through community services or with pharmacotherapy had lower suicide rates.
Psychological autopsy studies reveal that only 6–14 percent of depressed suicide victims were adequately treated and only 8–17 percent of all suicides were under treatment with prescription psychiatric medications. Yet significant opportunities to deliver adequate care exist since over 50–70 percent of those who complete suicide have contact with health services in the days to months before their death. However, suicide risk is difficult to assess. Individuals making serious suicide attempts may knowingly withhold their intentions. Currently, no psychological test, clinical technique, or biological marker is sufficiently sensitive and specific to accurately assess acute prediction of suicide in an individual.
There are significant barriers to receiving effective mental health treatment. About two-thirds of people with diagnosable mental disorders do not receive treatment. The stigma of mental illness deters people who need treatment from seeking it. The fragmented organization of mental health services and the cost of care are among the most frequently cited barriers to mental health treatment. Economic analyses of patterns of use of mental health services clearly indicate that use is sensitive to price: use falls as costs rise, while use increases with better insurance coverage. Physicians are reticent to talk to their patients about suicide; they often do not ask about intent or ideation, and patients often do not spontaneously report it. The goal of suicide treatment in specialty care is to develop and implement a treatment plan, which includes monitoring of medication efficacy and safety, as well as discharge planning. The details of treatment of suicidality, however, are not spelled out in any clinical guidelines. And many physicians are inadequately prepared to address suicide in their practices.
Primary care has become a critical setting for detection of the two most common risk factors for suicide: depression and alcoholism. According the American Medical Association, a diagnostic interview for depression is comparable in sensitivity and specificity to many radiologic and laboratory tests commonly used in medicine. Yet, currently only about 30–50 percent of adults with diagnosable depression are accurately diagnosed by primary care physicians. Treatment of depression in primary care is associated with reduced rates of completed suicide as shown by an ecological study on the Swedish Island of Gotland. Substance use disorders are especially important in suicide among young adults. Substance abuse and mood disorders frequently co-occur, with 51 percent of suicide attempters having both. In the primary care setting, numerous professional groups recommend routine screening for problem drinking in all patients.
A number of prevention programs have been explored to reduce the incidence of suicide and suicidal behaviors. At multiple levels (universal, selective, and indicated) interventions attempt to address risk factors and to enhance protective factors. Programs that integrate prevention at multiple levels are likely to be the most effective. The Air Force’s prevention program is one example of an integrated program that appears to have effectively reduced suicide rates in the community by removing barriers to treatment; increasing knowledge, attitudes, and competencies within the community; and increasing access to help and support with a consequent decrease in suicide rates. While there are several promising programs that have been implemented, long-term assessments and rigorous evaluation of their effectiveness are unavailable. Furthermore, some programs may work only in certain populations under certain circumstances; it is not yet evident whether these programs can be generalized to other populations or what characteristics of the programs are broadly applicable.
Reducing the availability or the lethality of a method (such as using blister packs for pills or enacting stricter gun control laws) results in a decline in suicide by that method; method substitution does not invariably occur. Education of the media regarding appropriate reporting of suicides can limit imitation effects and thereby reduce suicide rates.
Comprehensive school-based programs have shown some success in reducing suicidality. It is important to distinguish these programs from simple awareness interventions that can be detrimental. Screening programs, gatekeeper training programs, support/skills training groups, and school-based crisis response teams/plans can create a coordinated effort that identifies youth at risk for suicide and provides individual followup. If done appropriately, these interventions can provide the skills that allow youth to cope effectively with their life’s stresses.
Lack of longitudinal and prospective studies are a critical barrier to understanding and preventing suicide. Many prevention programs do not have the long-term funding that would allow them to assess reduction in the completion of suicide as an endpoint. The low-base rate of suicide, combined with the short duration of assessment and the relatively small populations under study, makes it difficult to acquire sufficient power for such trials. Some intervention analyses have increased their power by using alternate endpoints such as suicide attempts or ideation. Prevention studies sometimes use proximal outcomes like attitude or knowledge. There are statistical approaches used in other disciplines that can facilitate analyses of these low-rate events and would be worth-
while to apply to suicide. Refinements of Poisson regression models can be applied to low-base rate events and expand our understanding of suicide by sorting out the effects of age, race, and sex on clustered suicides (i.e., within counties) and improving our capability for assessment in prevention programs.
ADDITIONAL CHALLENGES IN SUICIDE RESEARCH
Because suicide has a low base-rate, studies need large populations to yield significant results. This problem is exacerbated by the poor reporting of suicides. Official suicide statistics capture completed suicides only and are fraught with inaccuracies. The numbers are inaccurate because of classification as death from undetermined causes and because of underreporting. It is estimated that most if not all of the undetermined cases are actually suicides. The quality of data on suicide attempts is even more tenuous than that of completed suicides.
Part of the reporting problem stems from the marked differences in the training and background of the person who by law certifies a death as a suicide. In the United States, the qualifications range from simply having an interest in the job (e.g., Indiana) to specialized training in forensic pathology (e.g., Oklahoma). Medico-legal officials may be elected, appointed or serve ex-officio (e.g., outside of the larger California counties, elected county sheriffs). Investigations may be centralized within a state (e.g., Rhode Island) or organized by each county (e.g., Utah). Internationally, differences in the organization and functioning of the officials monitoring suicide as a cause of death can produce artifactual differences even between similar countries such as England and Scotland. Most developing societies lack registries and expertly trained officials to record suicide. Further, there are cross-national differences in the underlying logic of classifications systems. In India, for example, the classification scheme focuses on social stresses and not psychopathology. Countries with religious sanctions against suicide tended to report suicide rates that were lower than for countries without sanctions. Furthermore, difference among countries may reflect the capacity of the emergency health care system to respond rather than differences in the intent of the individuals. The high rate of suicide among young Chinese women may result from their ready access to extremely toxic pesticides in the face of limited availability of emergency treatment. Each of these factors affects the nature, extent, and quality of the investigation and the classification of deaths as suicide.
Exclusion from Clinical Trials
Screening out clinical trial participants for suicidality has been the industry standard for trials of psychoactive medicines in an effort to reduce the risk of death. However, excluding those at risk for suicide precludes evaluation of treatments for this population. Suicide is, unfortunately, a medically expectable outcome of many mental illnesses. Death in a cancer clinical trial may be predictable or even inevitable, but trials do not exclude terminally ill patients. One might consider the outcome of suicide as a result of the mental illness, not the research or therapeutic intervention. The practice of excluding these patients from trials limits the opportunity for this population to benefit from such research. Excluding high-risk patients from these trials has been shown to weaken the sensitivity of intervention trials. Thus, suicide raises unique ethical issues. Furthermore, distinctive social and cultural contexts create different concerns for studying suicidal risk. These contexts must be taken into account, especially in global research.
Including participants at risk for suicide in clinical trials can be challenging. Before a person can consent to be part of a clinical trial, they must understand the purpose, the risks, and the possible benefits of the research. It has been argued that a suicidal person might understand the risks of a research protocol but not care about, or even welcome, the risks. Additional safeguards to ensure safety and ethical conduct of research can be provided, such as involving family members or other surrogates and ensuring that risks are not greater than under ordinary, usual, or standard care. In addition to the commonly used guidelines for protection of human subjects, a study design that affords greater protection to the participants can be used. For example, a risk-based allocation design, which assures that all of the sickest patients receive the experimental treatment, may be effective in research with suicidal patients.
To effectively study suicide requires a large population base, better reporting of completed and attempted suicide, careful control and analysis of risk and protective factors, common databases, and banks of biological tissues for analysis. Because of its low base rate, the difficulties in assessment, and the long-term nature of the risk and protective factors, the optimal approach to suicide is to use large populations with cultural and genetic diversity for long-duration, interdisciplinary studies.
Research centers have often been the mechanism used to address similar obstacles. Large research centers have the additional advantage of being able to provide training opportunities and thereby attract new re-
searchers to a difficult field. Furthermore, centers provide the opportunities for tissue banks and registries that are necessary resources to advance the field, and would enable ethnic, economic, and other social comparisons.
Suicide is responsible for about 30,000 deaths each year. Mental illness, the primary risk factor, afflicts over 80 million people in the United States; almost 15 million people have a serious mental illness (i.e., a mental disorder that leads to a functional impairment). For comparison, breast cancer claims the lives of about 40,000 women per year and between 10– 15 million people are living with the diagnosis of breast cancer. In 1998 over $400 million was allocated to research into the prevention, treatment, and cure of breast cancer. From estimates of the portfolios of SAMHSA, CDC, and NIMH, the funding for suicide was less than $40 million in 2000. The committee finds that this is disproportionately low, given the magnitude of the problem of suicide. A substantial investment of funds is needed to make meaningful progress.
In recognition of the current funding and research shortage in the area of suicide, the committee provides the following recommendation for establishing population research centers that will integrate the talents of experts from many disciplines and will draw upon a large population base with continuity over a long duration to improve prevention and treatment interventions. In addition, the recommendations aim to improve the monitoring of suicide, to increase the recognition and consequently the treatment of the primary risk factors in primary care, and to expand the efforts in prevention.
The National Institute of Mental Health (in collaboration with other agencies) should develop and support a national network of suicide research Population Laboratories devoted to interdisciplinary research on suicide and suicide prevention across the life cycle. The network of Population Laboratories should be administered by NIMH and funded through partnerships among federal agencies and private sources, including foundations. Very large study samples of at least 100,000 are necessary because of the relatively low frequency of suicide in the general public. A number of Population Laboratories (e.g., 5–10) are necessary to capture the data for numerous and complex interacting variables including the profound effects of demographics, region, culture, socioeconomic status, race, and ethnicity. Extending the efforts into the international arena where cultural differences are large may provide new information and can be fostered and guided by such global organizations as the World Health Organization and the World Bank and by the Fogarty International Center at NIH.
The network should be equipped to perform safe, high-quality, large-sample, multi-site studies on suicide and suicide prevention.
Each Laboratory would have a population base of approximately 100,000. At a base-rate of 10–12 suicides per 100,000 people, this population base of the network would significantly improve the available data for estimates of suicide incidence, capacity for longitudinal studies, development of brain repositories, access to representative samples for prevention and intervention studies, and studies of genetic risk for suicide. Several such laboratories would provide adequate data to assess the numerous and complex interacting variables including the profound effects of demographics, regions, culture, socioeconomic status, race, and ethnicity. Coordination and collaboration among centers should be encouraged to further enhance the breadth of the database.
The laboratories would cover an ethnically and socially diverse and representative population and would recruit higher risk individuals and subgroups in communities within the population laboratories for longitudinal and more detailed studies.
Treatment and prevention studies would be carried out in
high-risk patients recruited from within the population laboratories.
With these defined populations, the centers would conduct prospective studies—integrating biological, psychosocial, ethnographic, and ethical dimensions—that would be of great importance in advancing science and meeting public health needs. These studies would include such research initiatives as identified in the committee report including:
Evaluating means and effectiveness of promoting greater continuity of care, treatment adherence, and access to emergency services.
The effects of reducing hopelessness on suicide.
The potential of pharmacotherapies to reduce suicidal behavior.
The types and aspects of psychotherapy that are effective in reducing suicide.
The influence of HPA axis function on suicidality. Brain mapping studies on biological predictors of suicidal behavior.
The relationship between genetic markers and suicidal behavior and between suicide and aggression/impulsivity.
Prospective studies of populations at high risk for the onset of suicidal behavior.
Risk and protective effects of hospitalization, the relationships between length of stay and outcomes, and the factors post-hospital that account for the increased risk for suicide.
Outcomes of prophylactic/short-term versus maintenance/long-term treatment for suicidality.
Interactions of genetics and psychosocial, socio-political, and socioeconomic context.
Ethnographic research and other qualitative methods to obtain greater detail about the setting, conditions, process, and outcome of suicide.
Funding should be provided for the necessary infrastructure for these centers. This should include support for dedicated full-time staff at NIH to provide long-term (at least 10 years) continuity and consistency in these efforts. Furthermore, funding for centers should include support for the following:
Population cores to coordinate the social science, ethno-
graphic data and to maintain registries of deaths by suicide and suicide attempts.
Pathology cores to maintain the repositories for tissue samples from suicide victims.
Statistical cores to manage the databases on risk and protective factors including genetic markers and cultural contexts.
Clinical cores to recruit patients and to ensure their safe and ethical treatment.
Research efforts that encompass both program projects and individual projects. Centers should encourage collaborations across the centers and facilitate the sharing of data maintained by the cores.
In an effort to recruit excellent scientists to research in suicide, supported sites should develop training programs to provide local and distance mentoring, to attract new investigators from a wide variety of disciplines into the field, and to form research and research training partnerships with developed and developing countries.
National monitoring of suicide and suicidality should be improved. Steps toward improvement should include the following:
Funding agencies (including NIMH, NIA, NICHD, NIDA, NIAAA, CDC, SAMHSA, and DVA) should encourage that measures of suicidality (e.g., attempts) be included in all large and/or long-term studies of health behaviors, mental health interventions, and genetic studies of mental disorder. Funding agencies should issue program announcements for supplements to ongoing longitudinal studies to include the collection and analysis of these additional measures.
Suicidal patients should be included in clinical trials when appropriate safeguards are in place.
A national suicide attempt surveillance system should be de-
veloped and coordinated through the CDC. It might be developed as part of a broader injury reporting database. Modeled after Oregon’s program for the reporting of adolescent suicide attempts and the HIV/AIDS registry, pilot programs should be developed, tested, and implemented as soon as feasible. State participation should be encouraged by requiring reporting as a prerequisite for receiving funding for related programs.
Federal funding should be provided to support a surveillance system such as the National Violent Death Reporting System that includes data on mortality from suicide. The system should have sufficient funding to support a national effort. CDC would be the most appropriate agency to coordinate this database given their experience with HIV/AIDS surveillance. Efforts to create such registries in other countries should be encouraged and, where feasible, assisted.
Because primary care providers are often the first and only medical contact of suicidal patients, tools for recognition and screening of patients should be developed and disseminated. Furthermore, since over half of suicides occur in populations receiving treatment for mental disorders, it is critical to enhance the capacity of mental health professionals to recognize and address both chronic and acute suicide risk factors.
NIMH and other funding agencies should provide funds to clinical researchers to develop and evaluate screening tools that assess risk factors for suicide such as substance use, history of abuse and/or trauma, involvement with the criminal justice system, mental illness, psychological and personality traits such as impulsivity and hopelessness, abnormal neurobiology or genetic markers, employment problems, bereavement and other relationship stresses, etc. Funding agencies should issue program announcements to encourage efforts in this area.
Physicians should refer patients with multiple risk factors to consultation with a mental health professional. This should be standard in the same way finding high blood cholesterol levels
dictates further medical and behavioral interventions. This will only be effective if the issue of parity is addressed and insurance benefits are expanded adequately to cover mental health care.
Professional medical organizations should provide training to health care providers for assessment of suicide risk and provide them with existing tools. Mental health professional associations should encourage (or require, when appropriate) their memberships to increase their skills in suicide risk detection and intervention. National, state, county, and city public health organizations should build on their existing infrastructure to facilitate suicide screening especially in high-risk populations.
Medical and nursing schools should incorporate the study of suicidal behavior into their curricula or expand existing education.
NIMH and Agency for Health Care Research and Quality (AHRQ) should work with physician associations including American College of Physicians, American College of Family Physicians, American Academy of Pediatrics, American Society of Internal Medicine to implement these recommendations. In addition, through their health services research funds they should support efforts to improve approaches to identifying and treating those at risk.
Programs for suicide prevention should be developed, tested, expanded, and implemented through funding from appropriate agencies including NIMH, DVA, CDC, and SAMHSA.
Partnerships should be formed among federal, state, and local agencies to implement effective suicide prevention programs. Collaboration should be sought with professional organizations (including the American Psychiatric Association and the American Psychological Association) and non-profit organizations dedicated to the prevention of suicide (such as the American Foundation for the Prevention of Suicide or the American Association of Suicidology). NIMH and SAMHSA should work with the Department of
Education and the Administration on Aging to encourage national programs for youth and elderly populations.
Programs that have shown success within select populations should be expanded. For example, the Air Force program should be adopted by hierarchical organizations that employ groups with increased suicide rates, including police and rescue workers. Gatekeeper training programs and screening programs for youth and elderly should be implemented more broadly within work and educational settings to identify and intervene with those at suicide-risk. There should be a systemic identification of high suicide risk groups for targeted intervention.
Pilot programs for coping and resiliency training as part of the curriculum for school-aged children should be implemented, evaluated, and scaled up when feasible. Given the involvement of cumulative life-stressors in suicide and the existing efficacy data on these programs, it is expected that this intervention will reduce suicidality as well as other unwanted outcomes.
Restriction of access or reduced lethality of common means of suicide should be legislated (e.g., gun safety, barriers on bridges, altering the content of cooking gas, packaging of commonly used pills, and poison control).
Long-term public education campaigns and media training should be evaluated for their effectiveness both to change the public’s knowledge and attitudes and to reduce suicide and suicidal behaviors.