当前位置: 首页 > 期刊 > 《小儿科》 > 2005年第7期 > 正文
编号:11333193
Will I Be Alive in 2005 Adolescent Level of Involvement in Risk Behaviors and Belief in Near-Future Death
http://www.100md.com 《小儿科》
     University of Maryland/Sheppard Pratt Psychiatry Residency Program, Baltimore, Maryland

    Division of Adolescent and Young Adult Medicine, Children's National Medical Center, George Washington University School of Medicine and Health Sciences, Washington, DC

    Office of Research, Development, and Information, Centers for Medicare and Medicaid Services, Baltimore, Maryland

    ABSTRACT

    OBJECTIVE.: We examined the association between a belief in one's future mortality and various risk-taking behaviors among urban black adolescents. In particular, we investigated whether adolescents with higher levels of participation in various risk behaviors were more likely to believe in their future death as compared with adolescents with lesser levels of risk-taking behavior.

    METHODS.: Data obtained from April 1994 to March 1997 were analyzed for a total of 2694 adolescents, aged 12 to 21 years. The odds of believing that one would die within the next 2 years were calculated for various levels of participation in risk behaviors involving alcohol, drugs, and criminal or violent acts.

    RESULTS.: A total of 160 adolescents (7.1% of all boys and 5.4% of all girls) reported that they believed that they would die within the next 2 years. The adjusted odds of future death belief among adolescents who both actively engaged in and knew others who participated in all of the various risk behaviors, relative to adolescents who neither personally engaged in nor knew others who participated in any of the risk behaviors, was 3.22 (95% confidence interval [CI]: 2.01–5.17) vs 1.14 (95% CI: 0.67–1.95) for drug use and drug selling, 2.01 (95% CI: 1.38–2.92) vs 0.8 (95% CI: 0.39–1.62) for combined alcohol and drug use, and 5.60 (95% CI: 2.03–15.47) vs 1.61 (95% CI: 1.08–2.42) for violent physical behavior. In addition, residence in a foster home was significantly associated with death belief after adjustment for all other variables.

    CONCLUSIONS.: There is a significant relationship between certain risk behaviors and belief in near-future death. Moreover, higher levels of involvement in risk behaviors were associated with a stronger likelihood of belief in near-future mortality. Identification of adolescents who engage in certain risky behaviors, combined with a recognition of the degree to which the adolescent participates in the particular behavior(s), may be used to facilitate more rapid intervention among youths who either believe in their imminent demise or engage in behaviors that increase the likelihood of their untimely death.

    Key Words: adolescence attitude toward death risk-taking behavior substance abuse/use black youths

    Abbreviations: YRBSS, Youth Risk Behavior Surveillance System OR, odds ratio CI, confidence interval

    The hypothesis for this study originated both from medical care conversations with our adolescent patients and what we learned about the association of risk behaviors and future death belief through our previous research with the same population.1–8 In our clinical encounters, we became aware that many of our patients held a belief that their life would be cut short. They described vividly the conviction that their misery and threatening surroundings would not allow them to make things better for themselves. We could see them developing an attitude of hopelessness combined with increasing involvement in risky behaviors (eg, drug and alcohol use, violent physical and criminal acts).

    A number of studies indicate that risk behaviors in adolescents tend to co-occur, particularly among boys and older youths.9–12 On the basis of this growing consensus on the intercorrelation among the multiple adolescent health risk behaviors, we began to speculate that there may be an element or process of commonality that could tie together these diverse risk taking activities. This led us to the decision of exploring the possible association between the level of involvement in risk behaviors and the belief in a near-future death.

    Adolescents who engage in risk behaviors that have an adverse impact on their health may have a stronger belief in their imminent mortality than adolescents who do not pursue such behaviors. Conversely, a belief in one's imminent mortality may lead individuals to engage in behaviors that increase their risk of mortality. We sought to determine whether an association did, in fact, exist between participation in various risk behaviors and belief in near-future death among a population of black adolescents who resided in the Washington, DC, metropolitan area. We were interested more specifically in knowing whether the degree of the adolescent's involvement in a particular risk behavior (ranging from neither engaging in the behavior nor knowing anyone else who participates in the activity, to both active involvement and a relationship with other people who engage in it) was associated with a belief in the nearness of one's future death. To our knowledge, this is the first study to examine the relationship between belief in one's premature mortality and the level of participation in a particular risk behavior.

    METHODS

    Surveyed Population

    April 1994 through March 1997, 9562 adolescents (12–21 years of age) were treated in the Outpatient Clinic of Children's National Medical Center, Washington, DC. Of these, 3563 completed a survey that examined both direct and indirect exposure to a variety of risk behaviors, including use of drugs and alcohol and physical violence. The study was approved by the Institutional Review Board of Children's National Medical Center. Our survey was derived from the Youth Risk Behavior Surveillance System (YRBSS) survey developed by the Centers for Disease Control and Prevention13 to which we added 1 question relating to the imminence of death in the future.

    The YRBSS survey has a seventh-grade reading level and has been validated.13 Parental consent was waived for this study given the sensitive nature of the questions asked and that these questions would usually be asked anyway in the context of confidential care. Survey findings were blinded and anonymous. All participants were asked to provide information on standard demographic variables; complete the YRBSS survey; and respond to the question, "Which of the following applies to you: In the next 2 years (1) I am certain I will be alive, (2) It's very possible I will die, (3) It's unlikely I will die, or (4) I am certain I will be dead." Responses to this question formed the basis for construction of the dependent variable of belief in near-future death. A 2-year projection period was chosen because it was determined in a prestudy focus group that this was the longest time that adolescents believed that they could make a meaningful prediction for their "future."

    Because of the small number of nonblack individuals in the population (286 of 3563 [8%]), we restricted the analysis to black youths, yielding a subtotal of 3277 individuals. Of this number, 583 did not answer the question regarding future mortality and were excluded from the analysis, leaving a total of 2694 adolescents in the final analytic sample.

    Dependent Variable

    For analytic purposes, the survey respondents were categorized into 2 groups: adolescents who stated that their near-future death was certain or very possible were categorized as "death likely"; those who believed that, 2 years hence, they would certainly be alive or that their death was unlikely were classified as "death unlikely."

    Independent Variables

    Risk behavior variables were grouped into 3 general categories involving alcohol, drugs, and crime and violence. Each of these general categories contained 2 questions that examined exposure to each of these general types of risk behaviors. For alcohol, the questions concerned personal alcohol use and drinking and driving; for drugs, the questions concerned usage and selling; and for crime and violence, the questions addressed carrying a gun in the past month and injury by a weapon in the past year.

    We were interested in knowing whether the degree of one's involvement in a particular risk behavior was related to the individual's belief in near-future death. Accordingly, survey respondents were classified as to the nature of their participation in each of the risk behaviors: for active and passive participation, the behavior involved both the survey respondent and friends or relatives; active participation concerned the survey respondent only; passive participation meant that only friends or relatives engaged in the behavior; and nonengagement meant that neither the survey respondent nor his or her friends or relatives participated in the risk behavior.

    Coding of Independent Variables

    Alcohol use was coded on the basis of the participant's response to questions regarding the age at which he or she first consumed alcohol, the frequency of alcohol consumption, the percentage of friends who drank in the last month, and whether an excessive drinker resided in the adolescent's household. Exposure to drinking and driving was based on the age at which the respondent reported drinking and driving and the number of times that the respondent rode with a drunk driver in a particular month. Drug use was coded on the basis of the participant's response to questions regarding the age at which he or she first tried marijuana or other (unspecified) drugs and the percentage of friends that ever tried marijuana or other drugs. The level of engagement in drug selling was assessed through the survey respondent's replies to questions regarding the age at which he or she first sold drugs, whether a drug dealer resided in the same household, and whether a close friend was a drug dealer. The 2 indicators of crime and violence, gun carrying, and weapon injury were determined by using responses to questions regarding the number of days in the past month in which a gun was carried and the percentage of friends who carry a gun, and the number of times injured by a weapon in the past year and the number of times a friend was shot or stabbed in the past year.

    Living arrangements were categorized on the basis of the adolescent's residence with his or her mother only or mother and other relatives; father only or father and other relatives; mother and father or mother and father and other relatives; grandparent(s) only or grandparent(s) and other relatives; foster parents only; other relatives only; in a group home or shelter; or some other setting, including living with a boyfriend or girlfriend or living alone. Additional predictor variables included age (years), gender, education (last grade completed), and suicidal ideation/act (yes, no).

    Analysis

    Descriptive and Univariate Procedures

    Initial analysis involved calculation of descriptive statistics showing the number and percentage of adolescents by demographic and other characteristics, the distribution of adolescents who expressed the future death belief, and their level of participation in each of the previously described risk behaviors. All associations were tested for statistical significance through use of the 2 procedure. Next, the crude and age- and gender-adjusted odds of belief in near-future death were calculated for each of the risk behaviors through the use of logistic-regression modeling. All analyses were conducted by using SAS 8.02.14

    Multivariate Analysis

    The effect of individual predictor variables on future death belief, with adjustment for the effect of potential confounding variables, was assessed through the use of multivariate logistic-regression modeling. We elected to assess the role of level of involvement through categorization of adolescents into 4 distinct and homogeneous groups: both active and passive involvement in all 6 risk behaviors, active engagement only in all risk behaviors, passive engagement in all risk behaviors, and no involvement in any of the risk behaviors (referent group). Our assumption was that a natural gradient from high to low risk is inherent in this patterning. To accomplish this analysis, we combined the previously described 6 risk behaviors into 4 broad categories: alcohol, which examined the joint effect of both consumption and drunk driving; drugs, which represented the combined influence of both use and selling; drugs and alcohol combined; and violence, representing both gun carrying and weapon injury. The odds of belief in near-future death then were calculated for each of the 4 risk behaviors (alcohol consumption and drunk driving, drug use and selling, both alcohol and drug use, and violence) among individuals who consistently engaged in similar levels of participation for all 4 of the risk behaviors.

    RESULTS

    Descriptive

    Of the 2694 adolescents who responded to the survey, 1817 (67.4%) were female and 877 (32.6%) were male (Table 1). Girls were significantly older than boys (P = .001), with a correspondingly higher number of years of education (P = .001). Boys were more likely to express a belief in near-future death than girls (7.1% vs 5.4%, respectively), but this finding was not significant (P = .085) in 2 analyses. A history of suicidal ideation or acts was twice as common among girls (17.2%) than boys (8.5%; P = .001). Residence with one's mother or mother and other relatives was the single most common living arrangement among both boys (49.2%) and girls (50.3%). In contrast, only 4.2% of boys and 3.1% of girls reported living with their father only or father and other relatives. Residence within the traditional or extended nuclear family was more common among boys (17.8%) than girls (12.3%), respectively (P < .001). A routine physical examination was the single most frequent reason for treatment in the outpatient clinic among both boys (27%) and girls (15%).

    A significant positive association was noted between people who reported a history of suicidal ideation or acts and belief in their death within the next 2 years (P < .0001). No significant associations were noted among future death belief and age, gender, education, or living arrangements (Table 2).

    With the exception of drug use, all examined risk behaviors evidenced significant associations with level of participation and gender (Table 3). Boys exhibited higher levels of participation for drinking and driving (P = .02), drug selling (P = .001), and gun carrying (P = .001), but girls were more likely than boys to report alcohol use (P = .001).

    Crude and Adjusted Logistic-Regression Results

    Neither age nor gender played a major role in confounding the relationship between individual risk behaviors and future death belief. In both crude and adjusted logistic analyses, significant (P < .05) predictors of belief in near-future death were active and passive alcohol consumption; drinking and driving, regardless of level of participation; active and passive or passive-only drug use; drug selling regardless of level of participation; and active and passive or passive-only gun carrying or injury by a weapon (Table 4). Also, individuals who had either attempted suicide or reported suicidal ideation were significantly more likely to believe in their premature demise than those who did not engage in such behaviors. In addition, adolescents who resided within a foster home had an approximate threefold increased odds of believing in their future death as compared with adolescents who lived in a traditional nuclear or extended nuclear family (odds ratio [OR]: 3.07; 95% confidence interval [CI]: 1.32–7.16).

    Multivariate Logistic-Regression Results

    The number of adolescents who engaged in multiple risk behaviors is found in Table 5. Adolescents reported themselves as most likely to engage neither directly nor indirectly in any of the 4 combined categories of risk behavior. However, among youths who reported similar patterns of involvement in 1 of the 4 risk categories, both active and passive drug and alcohol use was most common (39%), as was passive-only gun carrying and weapon injury (27%). Very few adolescents reported consistent patterns of active-only participation across the 4 risk categories.

    Adolescents who both actively participated and knew others who were involved in each of the risk behavior combinations were significantly more likely to believe in their future death as compared with adolescents who engaged in lesser levels of participation in the risk behaviors: for drug use and selling combined, the odds of belief in future death was 3.22 (95% CI: 2.01–5.17; Table 6) for adolescents who both actively engaged in and knew others who were involved in the risk behavior as compared with an OR of 1.14 (95% CI: 0.67–1.95) for adolescents who only knew others who engaged in the behavior; for combined alcohol and drug use, the odds of death belief was 2.01 (95% CI: 1.38–2.92) compared with an OR of 0.8 (95% CI: 0.39–1.62); and for violent physical behavior, the respective values were an OR of 5.60 (95% CI: 2.03–15.47) as compared with an OR of 1.61 (95% CI: 1.08–2.42). In addition, residence in a foster home was significantly associated with death belief after adjustment for all other variables.

    DISCUSSION

    This study found a consistent pattern indicating higher odds of belief in near-future death with higher levels of involvement in various risk behaviors. Adolescents who both actively engaged in and knew others who engaged in all risk behaviors reported a greater likelihood of holding a belief in their own near-future demise as compared with adolescents with lesser levels of participation in all risk behaviors. The strongest association with near-future death belief was associated with adolescents who both carried a gun and reported having been injured by a weapon (OR: 9.83; 95% CI: 4.29–22.53), but active and passive participation in drug use and selling and alcohol and drug use were also associated with significantly increased odds of death belief relative to individuals who merely knew others who engaged in the respective risk behaviors (Table 6).

    Although it was not a primary research question, this study found a significant and independent association between residence in a foster care environment and belief in near-future death. Numerous studies have cited an association between behavioral difficulties during childhood and adolescence and an upbringing in a foster environment.15–19 However, it is impossible to attribute causality to the foster care environment itself or to characteristics that are brought into the foster environment from one's previous environment that may place an adolescent at increased risk.20 Given the increasing number of children who are placed in foster care,21,22 the results of this study provide additional evidence that children who reside in a foster environment are deserving of particular attention.

    A substantial proportion of adolescents, 17.8%, did not respond to the question addressing belief in near-future death. The 583 adolescents who did not respond to the future death question were significantly younger (P = .0047); had fewer years of education (P < .0001); and had different patterns of alcohol use (P < .0001) and drunk-driving behavior (P < .0002) as well as drug selling, gun carrying, and weapon injury (P < .004) relative to the 2694 people who did report on their potential for future mortality. It is possible that the observed differences in the above-mentioned risk factors among those who did or did not respond to the future mortality question may have biased this study's results. However, because a substantial number of people who did not answer the future death question also had missing values for level of involvement in the various risk behaviors, assessment of the direction of any potential bias was problematic. If nonresponders to the future death question were more likely to believe in their imminent demise and involve themselves in high-risk behaviors, then this study's results would be strengthened; if the converse situation were the case, then the study's results would be weakened. No significant associations were detected between those who did or did not answer the questions on future mortality and gender, suicidal act/ideation, living arrangements, or drug use. We suspect that the nonresponders, who were younger with fewer years of education and life experience, may have had more difficulty in relating the future death question to their own lives, hence the reason for their lack of response.

    Although the risk behavior data in this study are self-reported, a number of studies have found reasonable validity of adolescents' self-reported alcohol and drug use.23 Also, despite that only approximately one third of all patients who were seen in the outpatient clinic completed our survey, the responses of those adolescents who did provide data were generally similar to the responses of a nationally representative sample of adolescents who completed the 1997 and 1995 YRBSS survey administered by the Centers for Disease Control and Prevention.24,25 Our study population was restricted to black adolescents. Additional research is necessary to determine whether the associations that we observed between level of involvement in risk behavior and an inappropriately held death belief can be extended to other racial and ethnic groups.

    This study did not address possible determinants of risk-taking behavior among adolescents.26–28 Because of the cross-sectional nature of the data used in this study, no determination as to the direction of the causal relationship between future death belief and participation in risk behaviors could be made. The current study does show a congruence between engagement in risk-taking behavior or existence in an at-risk environment and the increased likelihood of an adolescent's belief in his or her premature death. This finding runs counter to studies indicating that adolescents minimize or misinterpret the degree of risk inherent in a particular activity, as compared with adults,29 and supports the research that they do understand the risks that they take.30

    Our findings are congruent with the recent research on high-risk behaviors in a survey of 2468 inner-city adolescents in Mobile, Alabama.31 Although a different set of questions were asked, a similar theme was detected: adolescents with a high degree of hopelessness had the highest rate of risk behaviors.

    It is possible that our results could be explained by considering that the adolescents who believed that they were going to die soon were depressed and or suicidal. Although our multivariate model (Table 6) did include suicidal ideation/act as a covariate, we had no information on depressive feelings among adolescents in our study population. Although an extensive literature exists on adolescent suicide and its risk factors,32–38 the present study offers a broader perspective, examining variations in premature death belief that were associated with the adolescent's own actions, the actions of others, or a combination of the 2.

    Our own interpretation of many adolescents' belief in a foreshortened life is that it is based on what they see and hear about violent death in their midst. We surmise that the combination of the emotional impact of learning of another teenager's death and an understandable lack of sophistication in interpreting epidemiologic data may result in an exaggerated assessment of the frequency of violent death and a distorted view of the likelihood of its occurrence. Thus, it is possible for those adolescents to think that a truncated life is a universal and unavoidable phenomenon. This would fit with the knowledge that lack of safety, security, and hope does not allow for long-term plans, when adolescents are convinced that "nothing is ever going to change"39 and may be the road that leads to the high-risk behaviors.

    The critical issue that emerged from this research is that the risk behavior and/or the inappropriately held belief need to be identified and dealt with in an expeditious manner. This feeling of having a foreshortened life and thoughts of hopelessness need to be addressed when dealing with adolescent morbidity and mortality. A 2-pronged approach might be required. On the one hand, there clearly is a need for societal restructuring to deal with lack of jobs, income inequality, and racism.40 On the other hand, a distorted cognitive interpretation of this hard reality can make matters worse and needs to be addressed with interventions such as skill building, mentoring, counseling, and enhancing connectedness.

    CONCLUSIONS

    Behavioral factors, typically involving risk taking, constitute the major determinants of morbidity and mortality among adolescents.41,42 Our study of a black adolescent urban patient population indicates a clear association between higher levels of risk behavior involvement and belief in future death. Although our data cannot be generalized or determine a causal relationship between future death belief and involvement with various risk behaviors studied, we propose that at the very least, this powerful and inappropriately held belief should be considered a marker for an increased risk for adolescent morbidity or mortality. Because the primary care physician is particularly well suited to play a critical role in identifying those who are at risk and those who are actively involved in risk-taking behaviors, we recommend that, during adolescent health care visits, a question assessing an adolescent's belief in a near-future death may serve as an indicator for associated risk behaviors. If the belief in a foreshortened life is present, then it is a sentinel finding that requires further exploration, both because the patient is suffering and because more in-depth questions about risk behaviors then are called for.

    FOOTNOTES

    Accepted Oct 14, 2004.

    The statements contained in this article are solely those of the authors and do not necessarily reflect the views or policies of the Centers for Medicare and Medicaid Services or the University of Maryland/Sheppard Pratt Psychiatry Residency Program.

    REFERENCES

    Pao M, Lyon M, D'Angelo LJ, et al. Psychiatric diagnoses in adolescents seropositive for the human immunodeficiency virus. Arch Pediatr Adolesc Med.2000; 154 :240 –244

    Guagliardo MF, Huang Z, D'Angelo LJ. Fathering pregnancies: marking health risk behaviors in urban adolescents. J Adolesc Health.1999; 24 :10 –15

    Brasseux C, D'Angelo LJ, Guagliardo M, Hicks J. The changing pattern of substance abuse in urban adolescents. Arch Pediatr Adolesc Med.1998; 152 :234 –237

    Lyon ME, Silber TJ, D'Angelo LJ. Difficult life circumstances in HIV infected adolescents: cause or effect AIDS Patient Care STDS.1997; 11 :29 –33

    Lyon ME, Richmond D, D'Angelo LJ. Is sexual abuse in childhood or adolescence a predisposing factor for HIV infection during adolescence Pediatr AIDS HIV Infect.1995; 6 :271 –275

    Cox JM, D'Angelo LJ, Silber T. Substance abuse and syphilis in urban adolescents: a new risk factor for an old disease. J Adolesc Health.1992; 13 :483 –486

    D'Angelo LJ, Getson PR, Luban NL, Gayle HD. Human immunodeficiency virus infection in urban adolescents: can we predict who is at risk Pediatrics.1991; 88 :982 –986

    Silber TJ, Iosefsohn M, Hicks JM, Getson PR, et al. Prevalence of PCP use among adolescent marijuana users. J Pediatr.1988; 112 :827 –829

    Flisher AJ, Kramer RA, Hoven CW, et al. Risk behavior in a community sample of children and adolescents. J Am Acad Child Adolesc Psychiatry.2000; 39 :881 –887

    Simantov E, Schoen C, Klein JD. Health-compromising behaviors: why do adolescents smoke or drink Arch Pediatr Adolesc Med.2000; 154 :1025 –1033

    Brener ND, Collins JL. Co-occurrence of health-risk behaviors among adolescents in the United States. J Adolesc Health.1998; 22 :209 –213

    Black MM, Ricardo IB. Drug use, drug trafficking, and weapon carrying among low-income African American, early adolescent boys. Pediatrics.1994; 93 :1065 –1072

    Kolbe LJ, Kann L, Collins JL. Overview of the youth risk behavior surveillance system. Public Health Rep.1993; 108 (suppl 1):2–19

    SAS [computer program]. Version 8.02. Cary, NC: SAS Institute Inc;2001

    Garwood MM, Close W. Identifying the psychological needs of foster children. Child Psychiatry Hum Dev.2001; 32 :125 –135

    Stewart SE, Manion IG, Davidson S, Cloutier P. Suicidal children and adolescents with first emergency room presentations: predictors of six-month outcome. J Am Acad Child Adolesc Psychiatry.2001; 40 :580 –587

    Rosenfeld AA, Pilowsky DJ, Fine P, et al. Foster care: an update. J Am Acad Child Adolesc Psychiatry.1997; 36 :448 –457

    Stein E, Evans B, Mazumdar R, Rae-Grant N. The mental health of children in foster care: a comparison with community and clinical samples. Can J Psychiatry.1996; 41 :385 –391

    Schor EL. Foster care. Pediatr Rev.1989; 10 :209 –216

    Runyan DK, Gould CL. Foster care for child maltreatment: impact on delinquent behavior. Pediatrics.1985; 75 :562 –568

    Orme JG, Buehler C. Foster family characteristics and behavioral and emotional problems of foster children: a narrative review. Fam Relations.2001; 50 :3 –15

    Simms MD, Dubowitz H, Szilagyi MA. Health care needs of children in the foster care system. Pediatrics.2000; 106 :909 –918

    Brener ND, Collins JL, Kann L, Warren CW, Williams BI. Reliability of the youth risk behavior survey questionnaire. Am J Epidemiol.1995; 141 :575 –580

    Kann L, Kinchen SA, Williams BI, et al. Youth risk behavior surveillance—United States 1997. J Sch Health.1998; 68 :355 –369

    Kann L, Warren CW, Harris WA, et al. Youth risk behavior surveillance—United States 1995. J Sch Health.1996; 66 :365 –377

    Shapiro R, Siegel AW, Scovill LC, Hays J. Risk-taking patterns of female adolescents: what they do and why. J Adolesc.1998; 21 :143 –159

    Quadrel MJ, Fischhoff B, Davis W. Adolescent (in)vulnerability. Am Psychol.1993; 48 :102 –116

    Tonkin RS. Adolescent risk-taking behavior. J Adolesc Health Care.1987; 8 :213 –220

    Cohn LD, Macfarlane S, Yanez C, Imai WK. Risk-perception: differences between adolescents and adults. Health Psychol.1995; 14 :217 –222

    Millstein SG, Halpern-Felsher BL. Judgments about risk and perceived invulnerability in adolescents and young adults. J Res Adolesc.2002; 12 :399 –422

    Bolland JM. Hopelessness and risk behavior among adolescents living in high-poverty inner-city neighborhoods. J Adolesc.2003; 26 :145 –158

    Vega WA, Gil AG, Zimmerman RS, Warheit GJ. Risk factors for suicidal behavior among Hispanic, African American, and non-Hispanic white boys in early adolescence. Ethn Dis.1993; 3 :229 –241

    Blum RW, Beuhring T, Shew ML, Bearinger LH, Sieving RE, Resnick MD. The effects of race/ethnicity, income, and family structure on adolescent risk behaviors. Am J Public Health.2000; 90 :1879 –1884

    Suicide among black youths—United States, 1980–1995. MMWR Morb Mortal Wkly Rep.1998; 47 :193 –196

    Michaud P-A, Blum RW, Ferron C. "Bet you I will!" Risk or experimental behavior during adolescence Arch Pediatr Adolesc Med.1998; 152 :224 –226

    Pfeffer CR. Childhood suicidal behavior. Psychiatr Clin N Am.1997; 20 :551 –562

    Shaffer D, Gould M, Hicks RC. Worsening suicide rate in black teenagers. Am J Psychiatry.1994; 151 :1810 –1812

    Woods ER, Lin YG, Middleman A, Beckford P, Chase L, DuRant RH. The association of suicide attempts in adolescents. Pediatrics.1997; 99 :791 –796

    Kazdin AE, French NH, Unis AS, Esrelt-Dawson K, Sherick RB. Hopelessness, depression and suicidal intent among psychiatrically disturbed in-patient children. J Consult Clin Psychol.1983; 51 :504 –510

    Duncan GJ, Brooks-Gunn J. Family poverty, welfare reform, and child development. Child Dev.2000; 71 :188 –196

    Sells CW, Blum RW. Morbidity and mortality among US adolescents: an overview of data and trends. Am J Public Health.1996; 86 :513 –519

    Irwin CE, Ryan SA. Problem behavior of adolescents. Pediatr Rev.1989; 10 :235 –246(Adela Valadez-Meltzer, MD)