Pediatric Annals

Risk and Prevention of Youth Suicide

David C Grossman, MD, MPH

Abstract

Injuries and death resulting from adolescent suicidal behavior present special problems for the pediatrician, in addition to the immediate physical injuries, the pediatrician must explore the behavioral and psychosocial dimensions of these selfinflicted injuries. The extraordinary psychologic distress experienced by the family and friends also must be addressed by the primary care provider. Perhaps the most important role of the pediatrician is to seek the prevention of these tragic injuries and deaths. This article discusses recent epidemiologie trends in youth suicide, new information regarding risk factors, and scientific evidence for the effectiveness of suggested preventive strategies.

INCIDENCE OF SUICIDE ATTEMPTS AND FATALITIES

Although completed adolescent suicide remains a rare event, its incidence increased markedly, particularly among white males, during the period from 1950 to 1980.1 During the 1980s, the rate remained at a plateau for young adults but continued to rise among 15- to 19-year-old youth. In 1987, the age-specific rate of completed suicide among 15- to 19-year-old youth was 10.3 per 100 000 with a male:female ratio of 4: 1.2 Because of the social stigma and lack of uniform criteria used by coroners, it is widely accepted that suicide rate are greatly underestimated.3-4

Suicide attempts are much more frequent than completions and occur most frequently among adolescents. In contrast to completed suicide, attempts by females greatly outnumber those by males. Hospitalizations for suicide attempts in the state of Washington between 1985 and 1987 were 30 times more frequent among 15- to 19-year-old females and three times higher among males the same age than suicide deaths.5 Surveys of high school students estimate a high prevalence of lifetime attempts. The national school-based Youth Risk Behavior Survey conducted by the Centers for Disease Control on a national sample of 9th through 12th grade students reported that 8.3% of these students had reported a previous attempt to kill themselves.6 Similar results were seen in large regional school-based surveys conducted in Minnesota7 and on the Navajo Indian reservation.8 In each of these surveys, only a fraction reported that their attempt required medical attention. Subsequent studies revealed that about half of these self-reported attempts represent self-inflicted injuries.9,10

WHOISATRISK:

The knowledge of risk factors for adolescent suicide enhances a clinicians ability to selectively inquire about past suicidal behavior and future risk. There is no suitable clinical screening instrument available that can reliably predict who will attempt suicide. However, recent epidemiologie studies have provided information regarding the risk factors that appear to be the most important in precipitating suicidal behavior. The most powerful risk factor is a history of previous suicide attempt(s). An individual who completes suicide is 20 to 30 times more likely to have had a previous attempt than are controls.11,12 Unfortunately, this risk factor alone has poor sensitivity for suicide completion because only 20% to 30% of suicide victims have a history of a previous suicide attempt10 and only 10% of these were hospitalized for their attempt David Shaffer, MD, personal communication). Other important risk factors include the presence of mental disorders, particularly an affective disorder such as major depression.13 Both clinical and population-based studies have demonstrated that individuals experiencing major depression have a 15- to 25-fold higher risk of suicidal behavior than the general population.14·15 Although this information was initially derived from adult samples, recent studies have confirmed this as a risk factor among adolescents also.12'16 Despite the high prevalence of depression, only about one third of suicide victims had been seen by a mental health practitioner prior to death.12

There has also been confirmation of an association of suicide and externalized behavior problems, eg, conduct disorder, that reflects an individual's difficulties in…

Injuries and death resulting from adolescent suicidal behavior present special problems for the pediatrician, in addition to the immediate physical injuries, the pediatrician must explore the behavioral and psychosocial dimensions of these selfinflicted injuries. The extraordinary psychologic distress experienced by the family and friends also must be addressed by the primary care provider. Perhaps the most important role of the pediatrician is to seek the prevention of these tragic injuries and deaths. This article discusses recent epidemiologie trends in youth suicide, new information regarding risk factors, and scientific evidence for the effectiveness of suggested preventive strategies.

INCIDENCE OF SUICIDE ATTEMPTS AND FATALITIES

Although completed adolescent suicide remains a rare event, its incidence increased markedly, particularly among white males, during the period from 1950 to 1980.1 During the 1980s, the rate remained at a plateau for young adults but continued to rise among 15- to 19-year-old youth. In 1987, the age-specific rate of completed suicide among 15- to 19-year-old youth was 10.3 per 100 000 with a male:female ratio of 4: 1.2 Because of the social stigma and lack of uniform criteria used by coroners, it is widely accepted that suicide rate are greatly underestimated.3-4

Suicide attempts are much more frequent than completions and occur most frequently among adolescents. In contrast to completed suicide, attempts by females greatly outnumber those by males. Hospitalizations for suicide attempts in the state of Washington between 1985 and 1987 were 30 times more frequent among 15- to 19-year-old females and three times higher among males the same age than suicide deaths.5 Surveys of high school students estimate a high prevalence of lifetime attempts. The national school-based Youth Risk Behavior Survey conducted by the Centers for Disease Control on a national sample of 9th through 12th grade students reported that 8.3% of these students had reported a previous attempt to kill themselves.6 Similar results were seen in large regional school-based surveys conducted in Minnesota7 and on the Navajo Indian reservation.8 In each of these surveys, only a fraction reported that their attempt required medical attention. Subsequent studies revealed that about half of these self-reported attempts represent self-inflicted injuries.9,10

WHOISATRISK:

The knowledge of risk factors for adolescent suicide enhances a clinicians ability to selectively inquire about past suicidal behavior and future risk. There is no suitable clinical screening instrument available that can reliably predict who will attempt suicide. However, recent epidemiologie studies have provided information regarding the risk factors that appear to be the most important in precipitating suicidal behavior. The most powerful risk factor is a history of previous suicide attempt(s). An individual who completes suicide is 20 to 30 times more likely to have had a previous attempt than are controls.11,12 Unfortunately, this risk factor alone has poor sensitivity for suicide completion because only 20% to 30% of suicide victims have a history of a previous suicide attempt10 and only 10% of these were hospitalized for their attempt David Shaffer, MD, personal communication). Other important risk factors include the presence of mental disorders, particularly an affective disorder such as major depression.13 Both clinical and population-based studies have demonstrated that individuals experiencing major depression have a 15- to 25-fold higher risk of suicidal behavior than the general population.14·15 Although this information was initially derived from adult samples, recent studies have confirmed this as a risk factor among adolescents also.12'16 Despite the high prevalence of depression, only about one third of suicide victims had been seen by a mental health practitioner prior to death.12

There has also been confirmation of an association of suicide and externalized behavior problems, eg, conduct disorder, that reflects an individual's difficulties in adapting to the social environment.11'17 Such individuals often experience difficulties with family, interpersonal relationships, school, employment, and the criminal justice system.

Alcohol and Illicit Drug Use

Alcohol and illicit drug use are often recognized to accompany the above-mentioned behavioral disorders and are also important risk factors for suicide.18,19 Both the chronic and acute use of alcohol have been shown to be important risk factors for both fatal and nonfatal attempts. The striking frequency of acute alcohol intoxication (46% of victims in one study) observed among young suicide victims is particularly alarming because it appears to substantially increase the chance that these youth will use a firearm, rather than a less lethal method, in the suicide attempt.20 Alcohol use appears to be a far more important risk factor than illicit drug use. Among drugs, only cocaine has been shown to be significantly associated with suicidal behavior, although this has been confirmed in only a few studies.14

Social Relationships

The absence or disruption of social ties are important risk factors among adolescents. Older adults who are single secondary to widowhood or divorce have long been known to be at higher risk; it now appears that the dissolution of nonmarital relationships among youth also brings greater risk. In one large uncontrolled study, the break-up of a strong relationship was often implicated as an acute precipitant of a suicide attempt.16 In a study conducted among Navajo, youth highly alienated from community and family were three times more likely to report a previous suicide attempt.8 Adolescent bereavement has been reported as a risk factor only in the context of loss by suicide of a family member or friend, although losses by other causes of death may be just as important.21 The potential "contagion" of suicide has long been suspected but only recently have epidemiologists been able to confirm the existence of unexpected suicide clusters of adolescent suicides in communities.22'24 In one study of US mortality data, the investigators were able to determine that, after accounting for chance, about 3.5% of suicides among 15- to 19-year-old adolescents occurred in clusters with wide variation (0.6% to 13.4%) between different states.22

While many of the individual risk factors discussed above have been confirmed individually, there is still a need to explore the temporal sequence of these behaviors relative to each other and the suicide attempt. Many of the risk factors are highly correlated (eg, substance abuse and conduct disorder) and need clarification of their individual causal relationships to suicide. For instance, is substance abuse really an independent risk factor for suicide or does it derive its association secondary to its association with depression?

Figure. Rate per 100000 of firearm suicide arnonge males aged 15 to 19 (bottom) and males aged 20-24 (top). Reprinted with permission from Boyd JH, Moscicki EK. Firearms and youth suicide. Am J Public Health. Copyright ® 1986, American Public Health Association.

Figure. Rate per 100000 of firearm suicide arnonge males aged 15 to 19 (bottom) and males aged 20-24 (top). Reprinted with permission from Boyd JH, Moscicki EK. Firearms and youth suicide. Am J Public Health. Copyright ® 1986, American Public Health Association.

An additional domain of risk that is relatively unexplored is the community environment in which an individual lives. Variation in rates of suicide among societies have been recognized since Emil Durkheim's work in the 19th century.25 Similar significant variations exist between counties and states in the United States today. The highest rates of youth suicide are among the rural Western states of Nevada, Wyoming, New Mexico, and Alaska.26 Can these high rates be completely explained by the demographic composition of each state and the prevalence of known individual risk factors? Since individual risk factors for suicide are clearly related to the integration of individuals in society, it is plausible that there also may be community risk factors that encompass a community's approach to education, economic policy, firearm regulation, or cultural and religious beliefs,

Firearm Ownership

One such variable that has been studied is community restriction of firearm ownership. There has been a recent increasing interest in the epidemiologie association of firearms to adolescent suicide.27 Firearms have been the increasingly dominant method of adolescent suicides for males (Figure) as well as females. Some contend that the increased rate of adolescent suicide appears to be largely explained by the increased availability and use of guns, particularly handguns, in such deaths.28,29

The community approach to handgun regulation was noted to be an important explanatory factor behind the variation in youth suicide rates between two cities of similar demographic, cultural, and economic backgrounds: Seattle and Vancouver, British Columbia. Firearms are much more strictly regulated in Vancouver compared to Seattle. Suicides among youth ages 15 to 24 were 40% more frequent in Seattle. Virtually all of the difference was attributable to the finding that handgun suicides in this age group were over nine times more common in Seattle.30 There was no compensatory increase in suicide by other methods by the Vancouver youth. The absence of the compensatory response among adolescents reflects the more impulsive nature of adolescent suicidal behavior; other equally lethal methods require more preparation and planning.

The results of ecologie analyses such as the SeattleVancouver study can be strengthened by confirmation that individual access to firearms also can be a risk factor for youth suicide. A recent well-designed casc'control study by Brent confirmed this association of household firearm access to adolescent suicide. In this study, guns were twice as likely to be found in the homes of suicide victims as in the homes of suicide attempters or never-suicidal psychiatric controls.31

PREVENTION

Pediatricians should regard suicide attempts as they do other injuries: as potentially preventable. The outcome evaluation of strategies to prevent adolescent suicide remains in early stages.32

The recent infusion of data from well-designed studies on risk factors have provided further impetus for preventive strategies to be promoted by pediatricians and other primary care providers at both the individual and community levels. However, few existing prevention strategies have been evaluated using rigorously designed studies.

In the Office

There is no simple, sensitive, and specific clinical screening instrument that will identify teens at risk of suicide. Several have been suggested but have not been validated in large studies.33 However, the absence of a screening instrument should not deter pediatrie providers from using the clinical encounter to identify individuals at risk by clinical case-finding because the uncovering of individual or combined risk factors for suicide are important domains that are also related to other injuries and health problems of adolescence.

Given the importance of a history of a previous suicide attempt, pediatricians should systematically incorporate inquiry about self-inflicted and nonintentional injuries in the past medical history in the evaluation of new adolescent patients. A routine inquiry about past mental health problems and substance abuse should also be instituted. Simple screening questions to identify individuals suffering from these problems should be incorporated into the visit. Individuals with a positive past history for these problems should be more closely evaluated for the presence of current suicidal ideation. If present, these individuals should be referred for more extensive evaluation. Furthermore, by inquiring about such problems, adolescents will understand that the primary care physician is a potential source of guidance if problems arise.

In the absence of community regulations to control access to firearms, physicians may elect to use the clinical encounter to ascertain the presence of household firearms and explain that they are a health hazard, particularly in homes where an individual is experiencing difficulty with depression or alcohol. Responsible family members should be asked to take action to remove the firearm altogether since locked storage and unloading of the weapon do not sufficiently reduce the risk of death from the gun.31

In the Community

Pediatricians have historically exercised important roles in injury control by influencing families in the office as well as serving as advocates for community prevention activities. It may be even more important for the prevention of suicide since less than half of the potential victims may have had a recent primary care visit prior to death.54 For this reason, pediatricians will find community-based strategies to prevent youth suicide of great interest.

Community-based prevention of suicide attempts has two dimensions: 1) interventions that address preconditional risk factors that place an individual at higher risk, and 2) interventions that target acute precipitating factors that lead to a suicide attempt.

Some community strategies targeting preconditional risk factors include suicide prevention classroom curricula, systematic school-based screening of youth at high risk for suicide attempts, and broad-based primary mental health prevention programs in school settings.

There has been a recent surge of school-based programs to prevent adolescent suicide.35 Most of these programs seek to enhance student knowledge and attitudes about suicide as well as to provide information on resources available to students in times of crisis. The use of short-term school-based curricula specifically designed to prevent suicide has recently been studied among adolescents attending high school. The results demonstrated that the program had little effect in changing attitudes of students who had made a previous suicide attempt, clearly the highest risk group. Individuals in this group were more likely to have a negative impression of the program and continue to endorse suicide attempts as an acceptable solution to difficult life circumstances.10 The impact of the program was evaluated by looking at previous attempters only; no comparable data were available on nonattempters.

School-based screening to identify youth at risk for suicide would need to depend on a highly sensitive and specific screening instrument that successfully identifies future suicide attempts. No such instrument for population-based screening currently exists. However, the potential to identify students suffering from mental disorders is much greater and may serve as an indirect approach. Despite potential success with such a method, numerous questions regarding the availability of treatment, cost, and ethical implications would need to be addressed.

There has also been long-standing interest in using the classroom to universally improve student mental health by using integrated school curricula to boost a child's social competence and his or her ability to effectively function with peers.36,37 Providing children and adolescents with a set of social problemsolving skills, including the generation of alternative solutions, that is reinforced by modeling behavior in the classroom, may provide them with a repertoire of behavioral strategies that will enhance their resilience to difficult emotional challenges. This strategy has received wide embrace among those planning violenceprevention programs for adolescents.38 Although few have been rigorously evaluated for behavioral outcomes, this type of program may hold greater promise than one designed to modify attitudes about suicide.

Community- based strategies that directly target the acute precipitating event primarily include the use of crisis centers and "hot lines" and the community regulation of firearms. As discussed earlier, increased community restriction of firearms may hold the most promising opportunity for the prevention of fatal suicide attempt. The notion of interpersonal outreach to an acutely suicidal individual originated in Britain with the advent of the Samaritan movement. In the United States, crisis centers advertise for clients with a variety of problems and usually are operated by not-for-profit institutions that refer clients to specific services. There have been few rigorous evaluations of the outcomes of such services. One national analysis compared suicide rates in counties having a new crisis center to counties that did not. There was a significant, albeit slight, reduction in the suicide rate noted among white females under 24 years of age residing in those counties with a recently added center.39 The persistence of this effect is unknown, and little additional positive evidence of effect exists.

SUMMARY

The rate of fatal suicide attempts among males rose dramatically between 1950 and 1980 and has continued to rise significantly during the 1980s among 15- to 19-year-old youth. Nonfatal suicide attempts, although far more frequent among females, appear to share similar risk factors with completed suicide. The most important preconditional risk factors for both fatal and nonfatal suicide attempts are a history of a previous attempt, major affective and conduct disorders, chronic substance abuse, and a history of personal losses, including experience of suicide attempts by peers and family members. Risk factors for the acute precipitation of an attempt include the acute disruption of an interpersonal relationship, alcohol intoxication, and access to firearms. Preventive efforts can be enhanced by pediatricians both in the office and the community. Clinical case finding of individuals thought to be at high risk will continue to identify some potential victims. Families with household firearms should be counseled regarding the risk of firearm injury and death, particularly in a home with an individual with other risk factors. Community campaigns to improve the mental health of youth and to reduce firearm ownership may hold future promise for the reduction of the unacceptable morbidity and mortality associated with adolescent suicide attempts.

REFERENCES

1 . Rosenberjj ML. Smith JC. Davidson LE, Conn JM. The emergence of youth suicide: an epidemiologie analysis and public health perspective. Anna Rev Public HeM. 1987;S:417-440.

2. National Center for Health Statistics. Vital Statistics of the United Staun, Vd 2: Mortality, Pan A, 1950-1988. Washington, DQ Government Printing Office; 1989.

3. Speechley M, Slavraky KM. The adequacy of suicide statisiics for use in epidemiology and public health. Can] Public HeM. 1991:82:38-42.

4. Centers for Disease Cuntrol. Operational criteria for determining suicide. MMWR. 1988:37:773-774.

5. Stanyk P. Suicide Attempts and Deaths m Washington State, I985-87. Olympia, Wash: Department of Social and Health Services; June 1989. Health Data Section.

6. Centers for Disease Control. Attempted suicide among high school students - United States, 1990. MMWR. 1991;40:633-635.

7. Blum R. Resnick M, Geer L, et al. The Minnesota Adolescent Health Survey: implications for physicians. Mirai Med. 1988;71: 143-146.

8. Gtossman DC, Milligan BC, Deyo RA. Rist factors for suicide attempts among Navajo adolescents. AmJ PMc Heohh. 199 1;8 1:870-874.

9. Meehan PJ, Lamb JA. Saltzman LE, O'Carroll PW. Attempted suicide among young adults: ptogreastowatd a progiess estimale of prevalence. Am J Psychiatry, 1991;149:4l44.

10. ShaJfcr D, Vleland V, Garland A, Rojas M. Underwood M, Busner C. Adolescent suicide attempters; response Io suicide prevention programs. JAMA. 1990:264:31513155.

11. Shaffer D. The epidemiology of teen suicide: an examination of risk factors. } Clin Psychiatry. 1988;49: 36-41.

1 2. Brent DA, Perper JA, Goldstein CE, et al. Risk factors for adolescent suicide. Arch Gen Psychiatry. 1988:45:581-588.

13. Moscicki EK, O'Carroll P, Rae DS, Locke BZ, Ruy A, Regier DA. Suicide attempts in the epidemiologie catchment area study. Yale J Bid Med. 1988;6!:259-268.

14. Petrunis KR, Samuels JF, Moscicki EK, Anthony JC. An epidemiologie investigation of potential risk factors for suicide attempts. Soc Psychiatry Psychiatr Epidemiol. 1990:15:193-199.

15. Monk M. Epidemiology of suicide. Epidemial Rev. l987;9;5l-69.

16. RicK CL, Fowler RC, Fogarry LA, Young D. San Diego suicide study. 111: relationships between diagnoses and Stressors. Arcfi Gen Psychiatry, i988:45:589-592.

17. Jofte RT, Offord DR, Boyle MH. Ontario Child Health Study: suicidal behavior in youth age 12-16 years. AmJ Psychiatry. 1988;145: 1420- 1423.

18. CrumleyFE.Substanceabuseandadolcsccntsuicidalbehavior.JAMA. 1990;263:305I3056.

19. Fowler RC, Rich CL, Young D. San Diego Suicide Study, II: substance abuse in ynung cases. Arch Gen Psychiatry. 1986:43:962-965.

20. Brent DA, Perper JA, Allman CJ. Alcohol, firearms, and suicide among youth. JAMA. 1987:257:3369.3372.

21. Shaffi M, ShahinC, WhittinghilljR, Derrick A. Psychological autopsy of completed suicide in children and adolescents. AmJ Psychiatry. 1985;142:1061-1064.

22. Gould MS, Wallenstein S, Kleinman M. Time-space clustering of teenage suicide. Am J Epidemial. 1990;13!:71-78.

23. Gould MS, Wallenstein S, Davidson L. Suicide clusters: a critical review. Suicidi Life Threat BeW. 1989;19:17-29.

24. Davidson LE, Rosenbetg ML, Mercy JA, Franklin J, Simmons JT. An epidemiologie study of risk factors in two teenage suicide clusters. JAMA. 1989:262:2687-2692.

25. DurkheimE. Suicide. New York, NY: Free Press; 1951.

26. Seiden RH. Death in the West - a regional analysis of the youthful suicide rate. West J Med. 1984:140:969-973.

27. Kellerman AL, Lee RK, Mercy JA, Banlon J. The epidemiologie basis for the prevention of firearm injuries. Anna Rev Public HeM. 1991; 1 2:17-40.

28. Boyd JH. The increasing rateof suicide by firearms. N EngiJ Med. 1983:308:872-874.

29. Boya JH, Moscicki EK. Fireatmsandyoutlisuicide. Am J Public Health. 1986;76:12401242.

30. Sloan JH, Rivara FP, Reay DT, Ferris JA, Kellermann AL. Firearm regulations and rates of suicide - a comparison of two metropolitan areas. N Engt } Med. 1990;322:369373.

31. Breni DA, Perper JA, Allman CJ, Moritz GM. Wartella ME, Zelenak JR The presence and accessibility of firearms in the homes of adolescent suicides - a case-controi study. JAMA. 1991:266:2989-2995.

32. Shaffer D1 Garland A, Gould M, Fisher P, Trautman P. Preventing teenage suicide: a critical review. J Am Acod Child Adolesc Psychiatry. 198S;27:675-687.

33. Slap GB, Vorters DF, Chaudhuri S, Centor RM. Risk factors for attempted suicide during adolescence, ftdiarrics. 19S9;84:762-772.

34- Lin EH, Von Korff M, Wagner EH. Identifying suicide potential in primary care. J Gen intern Med. 1989;4:l-6.

35. Garland A, Shaffer D, Whittle B. A national survey of school-based, adolescent suicide prevention programs. J Am Acod Child Adoieic Psychiatry 1989:28:911-934.

36. Pellegrini DS, Urbain ES. An evaluation of interpersonal cognitive problem solving training with children. } Child Psythd Psychiatry. 1985;26: 17-41.

37. Weissberg RP, Gesten EL, Rapkin BD. et al. The evaluation of a social problem solving program for suburban and inner-city third grade children. ) Consult CIm Psychology. 196 1;49:25 1-261.

38. Spivak H1 Hausman AJ, Prolhrow-Stith D. Practitioner's forum: public health and the primary prevention of adolescent violence - the Violence Prevention Project. Viofcnce and Victims. 1989:4:203-212.

39. Miller HL, Coombs DW, Leeper JD, Barton SN. An analysis of the effects of .suicide prevention facilities on suicide rates in the United Slates. Am J Public Health. 1984:74:340-343.

10.3928/0090-4481-19920701-10

Sign up to receive

Journal E-contents