FROM THE CENTERS FOR DISEASE CONTROL AND PREVENTION Morbidity and Mortality Weekly Report

based Injury Statistics Query and Re- cide rates increased approximately 5.1% Methods of porting System (WISQARS™)1 and ex- annually (pϽ0.0001). Among persons amined by age group (i.e., persons aged aged 15-19 years, the suicide rate Among Persons 10-14 and 15-19 years) and method declined from 7.3 in 1992 to 4.1 in 2001; Aged 10-19 Years— (e.g., firearm, suffocation, and - the suffocation suicide rate increased ing) for each year and the 10-year pe- from 1.9 to 2.7. Rate regression analy- , riod. To analyze these data, codes were ses indicated that, during the study pe- 1992-2001 used from the International Classifica- riod, the average annual decrease in - tion of Diseases, Ninth Revision (ICD-9) arm suicide rates for this age group was Ͻ MMWR. 2004;53:471-474 and the International Classification of Dis- approximately 6.8% (p 0.0001), and eases and Related Health Problems, Tenth the average annual increase in suffoca- 1 table, 2 figures omitted Revision (ICD-10), which was imple- tion suicide rates was approximately mented in 1999. Although coding of 3.7% (pϽ0.0001). suicide IN 2001, SUICIDE WAS THE THIRD LEAD- mortality data changed in 1999, the two rates also decreased in both age groups, ing cause of death among persons aged revisions have near 100% agreement on at an average annual rate of 13.4% 10-19 years.1 The most common classification of by firearm, suf- among persons aged 10-14 years (Fig- method of suicide in this age group was focation, and poisoning3; thus, the sui- ure (1) and 8.0% among persons aged by firearm (49%), followed by suffoca- cide method was defined consistently 15-19 years (Figure 2). Because of the tion (mostly ) (38%) and poi- during the period analyzed. small number of suicides by poisoning, soning (7%).1 During 1992-2001, al- For each method of suicide, annual these decreases have had minimal im- though the overall suicide rate among suicide rates (per 100,000 population) pact on changes in the overall profile of persons aged 10-19 years declined from were calculated by age group and over- of youths. 6.2 to 4.6 per 100,000 population,1 all. Because and suffocation Among persons aged 10-14 years, methods of suicide changed substan- were the two most common methods of suffocation suicides began occurring tially. To characterize trends in sui- suicide among persons aged 10-19 years, with increasing frequency relative to cide methods among persons in this age accounting for approximately 90% of firearm suicides in the early- to mid- group, CDC analyzed data for persons suicide deaths in this age group, trends 1990s, eclipsing (i.e., ratio Ͼ1.0) fire- living in the United States during 1992- for these two methods (Figures 1 and (2) arm suicides by the late 1990s (Table). 2001. This report summarizes the re- were formally evaluated by using nega- In 2001, a total of 1.8 suffocation sui- sults of that analysis, which indicated tive binomial rate regression to ac- cides occurred for every firearm sui- a substantial decline in suicides by fire- count for overdispersion in the data. cide among youths aged 10-14 years. arm and an increase in suicides by suf- Supplemental information on suicides Among youths aged 15-19 years, an in- focation in persons aged 10-14 and by poisoning (the third most common crease in the frequency of suffocation 15-19 years. Beginning in 1997, among method) and suicide by all other meth- suicides relative to firearm suicides be- persons aged 10-14 years, suffocation ods (e.g., jumping from a height) is pro- gan in the mid-1990s; however, in 2001, surpassed firearms as the most com- vided for illustration purposes. Annual firearms remained the most common mon suicide method. The decline in ratios of suffocation suicides to firearm method of suicide in this age group, firearm suicides combined with the in- suicides were examined as an addi- with a ratio of 0.7 suffocation suicides crease in suicides by suffocation sug- tional means of assessing changes in the for every firearm suicide. gests that changes have occurred in sui- relative magnitude of these methods over cidal behavior among youths during the time (Table). Reported by: KM Lubell, PhD, MH Swahn, PhD, AE preceding decade. Public health offi- Among persons aged 10-14 years, the Crosby, MD, Div of Violence Prevention; SR Kegler, PhD, Office of Statistics and Programming, National cials should develop intervention strat- rate of firearm suicide decreased from Center for Injury Prevention and Control, CDC. egies that address the challenges posed 0.9 per 100,000 population in 1992 to by these changes, including programs 0.4 in 2001, whereas the rate of suffo- CDC Editorial Note: The findings in that integrate monitoring systems, etio- cation suicide increased from 0.5 in 1992 this report indicate that the overall sui- logic research, and comprehensive pre- to 0.8 in 2001. Rate regression analyses cide rate for persons aged 10-19 years vention activities. indicated that, during the study pe- in the United States declined during Annual data on suicides occurring in riod, firearm suicide rates decreased an 1992-2001 and that substantial changes the United States during 1992-20012 average of approximately 8.8% annu- occurred in the types of suicide meth- were obtained by using CDC’s Web- ally (pϽ0.0001), and suffocation sui- ods used among those persons aged

©2004 American Medical Association. All rights reserved. (Reprinted) JAMA, July 28, 2004—Vol 292, No. 4 427

Downloaded From: https://jamanetwork.com/ on 10/01/2021 FROM THE CENTERS FOR DISEASE CONTROL AND PREVENTION

10-14 and 15-19 years. Rates of sui- tistical simulations comparing the im- Department of Health and Human Services, CDC, Na- tional Center for Injury Prevention and Control, 2004. cide using firearms and poisoning de- pacts of different types of suicide- Available at http://www.cdc.gov/ncipc/wisqars. creased, whereas suicides by suffoca- prevention approaches suggest that 2. National Center for Health Statistics. Multiple cause- of-death public-use data files, 1992 through 2001. Hy- tion increased. By the end of the period, widely implemented population- attsville, Maryland: U.S. Department of Health and Hu- suffocation had surpassed firearms to based strategies will have a greater effect man Services, CDC, 2003. become the most common method of in reducing overall suicide rates than 3. Anderson RN, Minino AM, Fingerhut LA, Warner M, Heinen MA. Deaths: injuries, 2001. Natl Vital Stat suicide death among persons aged the more traditional strategies that con- Rep 2004;52:1-5. 10-14 years. centrate on persons at high risk.7 One 4. Cook PJ. The technology of personal violence. In: Tonry M, ed. Crime and Justice: An Annual Review The reasons for the changes in sui- of the most common population- of Research, vol. 14. Chicago, Illinois: University of Chi- cide methods are not fully under- based strategies focuses on restricting cago Press, 1991:1-71. 5. Gunnell D, Nowers M. Suicide by jumping. Acta stood. Increases in suffocation sui- access to highly lethal means (e.g., Psychiatrica Scandinavica 1997;96:1-6. cides and concomitant decreases in barriers and bubble-packaging 6. O’Carroll PW. A consideration of the validity and reliability of suicide mortality data. Suicide and Life- firearm suicides suggest that persons for over-the-counter medications) to Threatening Behav 1989;19:1-16. aged 10-19 years are choosing differ- diminish the likelihood of death from 7. Lewis G, Hawton K, Jones P. Strategies for pre- ent kinds of suicide methods than in the self-inflicted injuries.8 However, the pri- venting suicide. Br J Psychiatry 1997;171:351-4. 8. Beautrais AL. Methods of in New Zea- past. Data regarding how persons vate nature of suffocation, its wide- land: trends and implications for prevention. Austra- choose among various methods of sui- spread availability, and its high lethal- lian and New Zealand Journal of Psychiatry 2000;34: 413-9. cide suggest that some persons with- ity suggest that population-based 9. Goldsmith SK, Pellmar TC, Kleinman AM, Bunney out ready access to highly lethal meth- prevention efforts addressing access to WE, eds. Reducing Suicide: A National Imperative. ods might choose not to engage in a lethal means also must confront the un- Washington, DC: National Academies Press, 2002. suicidal act or, if they do engage in sui- derlying reasons for suicidality to avoid cidal behavior, are more likely to sur- the potential for method substitution. vive their injuries.4 However, certain Suicide is a multifaceted problem for Suicide Attempts subsets of suicidal persons might sub- which several risk factors exist, includ- stitute other methods.5 Substitution of ing societal beliefs regarding the ac- and Physical methods depends on both the avail- ceptability of suicidal behavior, social Fighting Among ability of alternatives and their accept- isolation and community disintegra- ability. Because the means for suffoca- tion, stressful life events, mental ill- High School tion (e.g., hanging) are widely available, ness, and substance abuse.9 When such Students— the escalating use of suffocation as a factors combine with widespread avail- method of suicide among persons aged ability of lethal means, higher rates United States, 2001 10-19 years implies that the acceptabil- of suicide tend to result.8 The actual ef- ity of suicide by suffocation has in- fectiveness of existing population- MMWR. 2004;53:474-476 creased substantially in this age group. based prevention approaches in reduc- 1 table omitted The findings in this report are sub- ing the number of suicide deaths is ject to at least two limitations. First, be- unclear. In addition to improving data- VIOLENCE IS A MAJOR CAUSE OF MORBID- cause U.S. mortality data include a lim- collection methods, public health offi- ity and mortality, particularly among ited number of variables, direct testing cials should emphasize the need to youths. In the United States, homicide of whether method substitution is oc- evaluate the effectiveness of current sui- and suicide are the second and third curring is not possible; however, no cide-prevention activities and exam- leading causes of death, respectively, for sizeable demographic changes in the ine how effective strategies can be persons aged 13-19 years.1 Although composition of youth suicide dece- implemented in diverse communities. suicide commonly is associated with dents have been documented that might CDC supports various efforts to evalu- anxiety, depression, and social with- explain the changes in methods.1 Sec- ate suicide-prevention programs for drawal, research suggests a link be- ond, official mortality data for suicide youth. Additional information is avail- tween violent behaviors directed at one- are known to reflect some degree of un- able at http://www.cdc.gov/ncipc. self (i.e., suicidal behaviors) and violent dercount; however, such undercounts behaviors directed at others among ado- do not appear to affect conclusions Acknowledgment lescents.2-6 Certain students who en- 6 about trends over time. The findings in this report are based in part on con- gage in extreme forms of violence, such The findings in this report demon- tributions by JL Annest, PhD, Office of Statistics and as school shootings, exhibit suicidal ide- Programming, National Center for Injury Prevention strate that rapid shifts in youth sui- and Control, CDC. ation or behavior before or during the cidal behavior can occur, underscor- attack.2,3 However, suicidal behavior ing the need for early effective suicide- REFERENCES also might be associated with involve- prevention efforts that focus on suicidal 1. CDC. Web-based Injury Statistics Query and Re- ment in less extreme forms of violent thoughts and pressures. Moreover, sta- porting System (WISQARS™). Atlanta, Georgia: U.S. behaviors, such as physical fighting,

428 JAMA, July 28, 2004—Vol 292, No. 4 (Reprinted) ©2004 American Medical Association. All rights reserved.

Downloaded From: https://jamanetwork.com/ on 10/01/2021