 |
|
The Relationship Between Early Age of Onset of Initial Substance Use and Engaging in Multiple Health Risk Behaviors Among Young Adolescents
Robert H. DuRant, PhD;
Jeffrey A. Smith, PhD;
Shelley R. Kreiter, MD;
Daniel P. Krowchuk, MD
Arch Pediatr Adolesc Med. 1999;153:286-291.
ABSTRACT
| |
Background Previous research based on problem-behavior theory has found that early age of onset of substance use is associated with engaging in multiple health risk behaviors among high school students. It is unknown whether these relationships begin during early adolescence.
Objective To examine the relationships between early age of onset of cigarette, alcohol, marijuana, and cocaine use and engaging in multiple risk behaviors among middle school students.
Methods A modified version of the Centers for Disease Control and Prevention Youth Risk Behavior Survey was administered to 2227 sixth through eighth grade students attending 53 randomly selected middle schools in North Carolina. A Health Risk Behavior Scale was constructed from 16 behaviors, including indicators of violence and weapon carrying; current substance use; nonuse of helmets when biking, in-line skating or skateboarding; not wearing a seat belt; riding with a driver who had been drinking; and suicide plans. Among this sample of middle school students, the scale had a mean (SD) of 4.1 (2.7) (range=0-15), and had a high internal reliability coefficient ( =0.74). The independent variables included first time use of cigarettes, alcohol, marijuana, and cocaine at age 11 years or earlier; actual age of onset of each substance; race and ethnicity; family composition; sex; school grade; academic ranking; and older age for school grade. These data were analyzed with analysis of variance, Spearman r, and multiple linear regression.
Results All the independent variables were found to be associated (P<.005) with the Health Risk Behavior Scale during the bivariate analyses. When each of these significant variables were entered into a multiple regression model, having smoked at age 11 years or younger accounted for 21.9% of the variation in the Health Risk Behavior Scale. Male sex, early marijuana or cocaine use, older age, lower academic rank, white race, and living in a 1-parent family explained an additional 19.1% of variation in the model (adjusted R2=0.41, P<.001). When the actual ages of onset of the use of substances were analyzed, in order of magnitude; age of onset of smoking; male sex; age of onset of alcohol and marijuana use; age; lower academic ranking; age of onset of cocaine use; white race; and lower academic rating accounted for 52.8% (P<.001) of the variation in the Health Risk Behavior Scale.
Conclusion Even when considering sociodemographic factors, early age of onset of cigarette use was the strongest correlate of the number of health risk behaviors in which these young adolescents had engaged. Early onset of use of other substances was also associated with a clustering of health risk behaviors among this sample of middle school students. The findings suggest that screening for early experimentation with tobacco and other substance use will help identify young adolescents at increased risk for engaging in multiple health risk behaviors.
INTRODUCTION
PREVIOUS STUDIES have found that health risk and problem behaviors tend to cluster among adolescents, with the strongest associations occurring between substance use and other health risk behaviors.1-16 Problem-Behavior Theory provides an explanation for why substance use behaviors cluster with one another and with other health risk behaviors.17-18 First, the social ecology of adolescent life provides socially organized opportunities to learn risk and problem behaviors and normative expectations that they be performed together.17 Second, health risk and problem behaviors cluster because different risk behaviors serve the same social and/or psychological developmental functions for adolescents, such as affirming individuation from parents, trying to achieve adult status, and seeking acceptance from peers. Third, health risk behaviors cluster because they are the manifestations of similar underlying factors.18 In addition, Jessor et al17 have posited that some adolescents develop a "risk behavior syndrome" that is caused by a general latent variable of unconventionality combined with exposure to risk factors from 5 risk domains. The degree of clustering is dependent on the adolescent's exposure to multiple risk domains from 5 broad areas: biological and genetics, social environment, perceived environment, personality, and behavior. Jessor also argues that an early age of onset of health risk behaviors is associated with an increased likelihood that adolescents will engage in multiple risk behaviors as they progress through adolescence.18 In support of this hypothesis, several studies have found that an early age of onset of substance use is associated with engaging in other health risk behaviors during middle and late adolescence5, 7, 9, 15-16; however, it is not known if early age of onset of substance use is associated with engaging in multiple health risk behaviors during early adolescence.
The purpose of our study is to examine the relationships between early age of onset of cigarette, alcohol, marijuana, and cocaine use and engaging in multiple health risk behaviors among middle school students.
PARTICIPANTS, MATERIALS, AND METHODS
SAMPLE
The Centers for Disease Control and Prevention developed the Youth Risk Behavior Survey (YRBS) to assess the prevalence of health risk behaviors among 9th through 12th grade students across the United States. The National YRBS has been approved by the Centers for Disease Control and Prevention Institutional Review Board. The Centers for Disease Control and Prevention considers the state YRBS as surveillance, thus not requiring Institutional Review Board approval. Secondary data analyses of state YRBS data do not need Institutional Review Board approval because the survey is anonymous. In the spring of 1995, a modified and shortened version of the YRBS was administered to 2231 randomly selected sixth though eighth grade students attending 53 randomly selected public middle schools in North Carolina. Within schools, classrooms are randomly selected and all students in the class who were present on the day the YRBS was administered were surveyed. This represented the 463 public middle schools and 261,309 students in the state. The response rate for schools was 74% and the response rate for students was 86%. Student participation was voluntary, and students were assured of anonymity during administration of the survey. Passive parental consent was used for the North Carolina YRBS. Demographic data on age, grade, sex, family type, greater than 1 year older for school grade, and self-report of the academic level of the student are presented in Table 1.
|
|
|
|
Table 1. Demographic and Descriptive Variables of Middle School Students in North Carolina
|
|
|
Using an approach that we have described previously,15 we constructed a Health Risk Behavior Scale (HRBS) from 16 health risk behaviors. The percentage of students engaging in each of these 16 behaviors is presented in Table 2. An item analysis was conducted to determine the effect of removing any item from the scale on the internal consistency or reliability (Cronbach ) of the scale. All items contributed to the internal reliability of the scale. The scale had a Cronbach of 0.74, indicating good internal reliability. When the 16 items were divided into 2 scales similar to the clustering of behaviors described by Basen-Engquish et al19 the levels dropped to an unacceptable level. Complete data were available on 2075 students.
|
|
|
|
Table 2. Number of Students Engaging in Each Behavior for the Variables Included in the Health Risk Behavior Scale*
|
|
|
The YRBS also measured the age of onset of smoking a whole cigarette, the first drink of alcohol for other than religious reasons, marijuana use, and any form of cocaine use. Each question was measured on 8-point scales ranging from never having used the substance to "15 years or older." The scales were recoded so that "9 years or younger" was coded as a "1" and nonuse was coded as an "8." The age distributions of these 4 variables are presented in Table 3.
|
|
|
|
Table 3. Age of Onset of Cigarette, Alcohol, Marijuana, and Cocaine Use
|
|
|
STATISTICAL ANALYSIS
Bivariate associations between the HRBS and age of onset of substance use, age, school grade, and type of student were tested using Spearman  (r) correlation coefficients. Age of onset of substance use variables were also recoded as age of onset 11 years or younger vs 12 years or older or never used. Age 11 years was chosen because 91.4% of the students were age 12 years or older at the time of the survey. Bivariate analyses of early age of onset of substance use and the HRBS were conducted with analysis of variance tests. Variables significantly (P .01) associated with the HRBS and demographic variables were then entered into a stepwise, multiple linear regression model with a forward inclusion of the independent variables. Two regression models were constructed; 1 including early age of onset of substance use and 1 including actual age of onset of substance use.
All analyses were conducted on weighted data. Weighting procedures were to used to correct for the sampling scheme in the data collection. Westat Inc (Rockville, Md) performed the data cleaning and the computation of the weights under contract by the Centers for Disease Control and Prevention. Weighting compensated for nonresponse and reflected the likelihood of sampling each student so the sample represents all middle school students in North Carolina. The weight used for estimation is given by the following: W indicates W1*W2*f1*f2*f3; W1, inverse of the probability of school selection; W2, inverse of the probability of classroom selection; f1, a school-level nonresponse of adjustment factor calculated by school size; f2, a student-level nonresponse adjustment factor calculated by school size; and f3, a poststratification adjustment factor by sex and grade.
RESULTS
Of the 16 variables in the HRBS, middle school students in this sample reported engaging in an average of 4 health risk behaviors (Table 2 ). The percentage of students who reported engaging in substance use at age 11 years or younger was 45.9% for alcohol, 25% for cigarettes, 5.8% for marijuana, and 1.7% for cocaine (Table 3). Students reporting cigarette, alcohol, marijuana, or cocaine use at age 11 years or younger also reported engaging in a significantly greater number of health risk behaviors than students whose age of onset of these substances was 12 years or older or students who reported never using these substances (Table 4). Higher mean HRBS scores were also found for students whose ages were greater than 1 year older than expected for school grade, not living in a 2-parent home, white, non-Hispanic race, male sex, and in the seventh or eighth grades (Table 4).
|
|
|
|
Table 4. Analyses of Variance Tests of the Health Risk Behavior Scale*
|
|
|
When each of these variables were entered into a multiple regression model, early age of onset of cigarette use accounted for 21.9% of the variation in the HRBS. Male sex accounted for an additional 8.2% of the variation in the scale when entered as the second variable in the model. Early onset of marijuana use accounted for 5.3% of the variation in the HRBS as the third variable in the model. Age, student type, early onset of cocaine use, minority ethnicity, and living in a 2-parent home accounted for an additional 5.6% of variation in the model, with 41% of the variation in the HRBS being explained by these 8 variables (Table 5).
|
|
|
|
Table 5. Regression Analyses of the Effects of the Early Age of Onset (11 Years of Age) of Substance Use on the Health Risk Behavior Scale
|
|
|
Age of onset of cigarette, alcohol, and marijuana use were moderately inversely correlated with the HRBS (Table 6). Age of onset of cocaine use was weakly inversely correlated with the number of health risk behaviors in which these students reported engaging. When analyzed with linear regression, age of onset of cigarettes use accounted for 32.3% of the variance in the HRBS (Table 7). Male sex and age of onset of alcohol use accounted for 7.6% and 7% of additional variance, respectively, in the scale as the second and third variables in the model. Age of onset of the initial marijuana use accounted for an additional 3.9% of variance as the fourth variable in the model. These 4 variables and age, age of onset of cocaine use, minority ethnicity, and student type accounted for a total of 52.8% of the variance in the HRBS (Table 7).
|
|
|
|
Table 6. Spearman Correlation Coefficients Between the Health Risk Behavior Scale and Age of Onset of Substance Use and Demographic Variables
|
|
|
|
|
|
|
Table 7. Regression Analysis of the Associations Between Age of Onset of Substance Use and the Health Risk Behavior Scale
|
|
|
COMMENT
Research has clearly documented that many adolescents in this country are at significant risk for health-compromising problems resulting from early initiation of health risk behaviors. The middle school students in this study reported engaging in 4 health risk behaviors and there was a clustering of health risk behaviors. Similarly, studies of older adolescents have found that health risk behaviors cluster1-21 and, in some adolescents, the clustering is sufficiently strong to create a "risk behavior syndrome."17-18
Some investigators have questioned the unidimensional clustering of both health risk and problem behaviors and have proposed that the clustering of health risk behaviors may be multidimensional.19 In an analysis of the Texas high school YRBS, Basen-Engquist et al19 found 4 different dimensions to the clustering of health risk behaviors. The first cluster consisted of perceived poor academic performance, no participation in sports, and low fruit and vegetable consumption. The second cluster included injection drug, cocaine, crack, and anabolic steroid use, suicidal behavior, fighting, and carrying a weapon. The third cluster included not participating in aerobic activity and eating a high-fat diet. Unsupervised swimming, using smokeless tobacco, riding a motorcycle without a helmet, and unsafe dieting comprised the fourth cluster. The HRBS, developed for our study, contained many of the behaviors in the second and fourth dimensions of the study by Basen-Engquist et al19; however, an item analysis performed on the HRBS indicated that dividing the 16-item scale into 2 separate scales, similar to the aforementioned dimensions 2 and 4, sharply reduced the internal reliability coefficients. These 16 risk behaviors composed a single dimension for the middle school students in our study.
Although previous research has found strong empirical evidence for a relationship between substance use and other health risk behaviors, the precise sequence of the relationship remains uncertain.7, 22-23 Whether tobacco, alcohol, and other substance use is antecedent to or the consequence of other health risk behaviors, or if there is a reciprocal relationship, continues to be debated.7, 24-27 We found that early age of onset of tobacco, alcohol, and other substance use was associated with engaging in other health risk behaviors during early adolescence. The variable with the strongest association with the number of health risk behaviors that these young adolescents engaged in was onset of cigarette use at 11 years or younger. In addition, students whose age of onset of alcohol and marijuana use was at 11 years or younger also engaged in a greater number of health risk behaviors than has been reported previously.1-5,9-16
We also found that actual age of initiation of cigarette, alcohol, marijuana, and cocaine use were all associated with the number of risk behaviors in which these middle school students reported engaging. When these 4 variables were combined with male sex, age, white race, and self-assessment of academic level, 52.8% of the variation was accounted for in the HRBS. Although these results are consistent with previous studies of older adolescents,5, 10, 15-16 the amount of variation in the scale accounted for by our model is substantially higher than reported in other studies.
These findings should not be interpreted as implying that early onset of substance use causes other risk behaviors among young adolescents; however, it is likely that the same social, psychological, biological, and environmental casual factors responsible for early age of onset of substance use also increase the risk of young adolescents engaging in other health risk behaviors. Similar to the relationship between exposure to violence in the home and in the community and the use of violence by adolescents,28 these social environments provide both modeling and opportunities to engage in multiple health risk behaviors.17-18 Early age of onset of substance use may also contribute more directly to engaging in other health risk behaviors. For example, an adolescent who begins to use marijuana regularly may be rejected by conventional peer groups and increase his or her association with peer groups that approve of marijuana use. Social learning from the new group could result in engaging in other health risk behaviors and a weakening of bonds to conventional groups.20
A limitation of our study is that the YRBS is a cross-sectional survey and causality cannot be implied. We also do not know the validity of the students' responses.7 We adjusted our P value from .05 to .01 to control for the cluster sampling of the YRBS; however, most of the bivariate relationships had P values of .0001 or less.
The findings from this study have several implications for primary health care providers. Screening for substance use is important even as young as 9 years. If tobacco, alcohol, marijuana, or cocaine use are discovered in preadolescents or young adolescents, patients are at an increased risk of engaging in multiple other behaviors that can compromise their health and that these patients are at risk of developing a risk behavior syndrome. Appropriate intervention should not only focus on helping patients discontinue their use of the substance, but also address those factors in children's lives that are associated with the substance use. This often requires the help of mental health care professionals.
The findings from this study also have implications for public health professionals and health educators who are developing and implementing prevention programs. Due to the strength of the associations among the early age of onset of substance use variables and other health risk behaviors among these middle school students, prevention programs should be initiated during elementary school and continued at least through middle school.20, 29-31 Also, programs should focus on the clustering of health risk behaviors, address multiple risk factors, and incorporate multiple protective factors.17-18,20 Focusing on behaviors, such as violence or weapon carrying, without also addressing other related behaviors, such as substance use, is likely to be less effective than health education or prevention programs that are comprehensive.
Tobacco, alcohol, and other drug use and engaging in other health compromising behaviors remain important public health problems in this country.20 Research is still needed to better understand the causes and prevention of health risk behaviors, particularly of young adolescents. Pediatricians and other health care professionals have a responsibility to advocate for children and adolescents at the local, state, and national levels concerning the prevention of early age of onset of substance use and its association with engaging in other behaviors that threaten the health of our youth. Physicians and other primary health care providers can be instrumental in screening their pediatric and adolescent patients for early substance use and other health risk behaviors and help or refer them to appropriate services that will assist in the discontinuation of behaviors that will compromise their health.
AUTHOR INFORMATION
Accepted for publication August 11, 1998.
Presented in part at the Society for Pediatric Research, New Orleans, La, May 3, 1998.
Corresponding author: Robert H. DuRant, PhD, Department of Pediatrics, Wake Forest University School of Medicine, Medical Center Boulevard, Winston-Salem, NC 25157-1081 (e-mail: rdurant{at}wfubmc.edu).
From the Department of Pediatrics, Brenner Children's Hospital and the Brenner Center for Child and Adolescent Health, (Drs DuRant, Kreiter, and Krowchuk), and the Department of Psychiatry (Dr Smith), Wake Forest University School of Medicine, Winston-Salem, NC.
REFERENCES
| |
1. DuRant RH, Rickert VI, Ashworth, CS, Newman C, Stevens G. Multiple substance use associated with anabolic steroid use among adolescents. N Engl J Med. 1993;329:922-926.
2. DuRant RH, Escobedo LG, Heath GW. The relationship between anabolic steroid use, strength training, and multiple drug use among adolescents in the United States. Pediatrics. 1995;96:23-28.
FREE FULL TEXT
3. Resnick MD, Bearman PS, Blum RW, et al. Protecting adolescents from harm. JAMA. 1997;278:823-832.
ABSTRACT
4. Middleman AB, Faulkner AH, Woods ER, Emans SJ, DuRant RH. High-risk behaviors among high school students in Massachusetts who use anabolic steroids. Pediatrics. 1995;96:268-272.
FREE FULL TEXT
5. Spingarn RW, DuRant RH. Male adolescents who cause pregnancy associated health risk and problem behaviors. Pediatrics. 1996;98:262-268.
FREE FULL TEXT
6. Brener ND, Collins JL. Co-occurrence of health-risk behaviors among adolescents in the United States. J Adolesc Health. 1998;22:209-213.
FULL TEXT
|
ISI
| PUBMED
7. Warren CW, Karrn L, Small ML, Santelli JS, Collins JL, Kalbe LJ. Age of initiating selected health-risk behaviors among high school students in the United States. J Adolesc Health. 1997;21:225-231.
FULL TEXT
|
ISI
| PUBMED
8. DuRant RH, Treiber F, Goodman E, Woods ER. Intentions to use violence to resolve conflict in hypothetical situations among young adolescents. Pediatrics. 1996;98:1104-1108.
FREE FULL TEXT
9. Escobedo LG, Reddy M, DuRant RH. Relationship of cigarette smoking to health risk and problem behaviors among US adolescents. Arch Pediatr Adolesc Med. 1997;151:66-71.
ABSTRACT
10. Shrier LA, Emans SJ, Woods ER, DuRant RH. The association of sexual risk behaviors and problem drug behaviors in high school students. J Adolesc Health. 1997;20:377-383.
FULL TEXT
|
ISI
| PUBMED
11. DuRant RH, Kahn J, Beckford PH, Woods ER. The association of weapon carrying and fighting on school property and other health risk and problem behaviors among high school students. Arch Pediatr Adolesc Med. 1997;151:360-366.
ABSTRACT
12. Woods ER, Lin VG, Middleman A, Beckford P, Chase L, DuRant RH. The association of suicide attempts with other risk behaviors in adolescents in Massachusetts. Pediatrics. 1997;99:791-796.
FREE FULL TEXT
13. DuRant RH, Knight J, Goodman E. Factors associated with aggressive and delinquent behaviors among patients attending an adolescent medicine clinic. J Adolesc Health. 1997;21:303-308.
FULL TEXT
|
ISI
| PUBMED
14. Shrier LA, Pierce JD, Emans SJ, DuRant RH. Gender differences in risk behaviors associated with reported forced sex. Arch Pediatr Adolesc Med. 1998;152:57-63.
FREE FULL TEXT
15. Garofolo R, Wolf C, Kessle S, Palfrey J, DuRant RH. The association between health risk behaviors and sexual orientation among a school-based sample of adolescents. Pediatrics. 1998;101:895-902.
FREE FULL TEXT
16. DuRant RH, Krowchuk DP, Sinal SH. The relationship between number of male sexual partners of male adolescents, victimization, use of violence, and drug use at school, and other violence and victimization. J Pediatr. 1998;132:113-118.
17. Jessor R. Risk behavior in adolescence: a psychosocial framework for understanding and action. J Adolesc Health. 1992;21:597-605.
18. Jessor R, Donovan JE, Costa FM. Beyond Adolescence: Problem Behavior and Young Adult Development. New York, NY: Cambridge University Press; 1991.
19. Basen-Engquist K, Edmundson EW, Parcel GS. Structure of health risk behavior among high school students. J Consult Clin Psychol. 1996;64:764-775.
FULL TEXT
|
ISI
| PUBMED
20. DuRant RH, Buchanan C. Tobacco, alcohol and other substance use among children and adolescents. In: Ripped JM, ed. Textbook of Medicine, Exercise, Nutrition, and Health. Cambridge, Mass: Blackwell Science. In press.
21. Osgood DW, Johnson LD, O'Malley PM, Bachman JG. The generality of deviance in late adolescence and early adulthood. Am Soc Rev. 1988;53:81-93.
FULL TEXT
22. Greenwood PW. Substance abuse problems among high-risk youth and potential interventions. Crime Delinquency. 1992;38:444-458.
ABSTRACT
23. Harrison L, Gfroerer J. The intersection of drug use and criminal behavior. Crime Delinquency. 1992;38:422-443.
ABSTRACT
24. Ellliott DS, Huizinga D, Ageton SS. Explaining Delinquency and Drug Use. Beverly Hills, Calif: Sage Publications; 1985.
25. Gottfredson MR, Hirschi T. A General Theory of Crime. Palo Alto, Calif: Stanford University Press; 1990.
26. Huizinga D, Loeber R, Thornberry T. Urban Delinquency and Substance Abuse: Technical Reports. Vols 1 and 2. Washington, DC: Dept of Justice; 1991.
27. McCord J. Problem behaviors. In: Feldman SS, Elliott GR, eds. At the Threshold: The Developing Adolescent. Cambridge, Mass: Harvard University Press, 1990;414-430.
28. DuRant RH, Cadenbread C, Pendergrast RA, Slavens A, Linder CW. Factors associated with the use of violence among black adolescents. Am J Public Health. 1994;84:612-617.
FREE FULL TEXT
29. Botvin GJ, Baker E, Busenbury L, Botvin EM, Diaz T. Long-term follow-up results of a randomized drug abuse prevention trial in a white middle-class population. JAMA. 1995;273:1106-1112.
ABSTRACT
30. Botvin GJ, Schinke SP, Epstein JA, Diaz T, Botvin EM. Effectiveness of culturally focused and generic skills training approaches to alcohol and drug abuse prevention among minority adolescents. Psychol Addict Behav. 1995;9:183-194.
31. Botvin GJ, Schinke SP, Epstein JA, Diaz T. Effectiveness of culturally focused and generic skills training approaches to alcohol and drug abuse prevention among minority youth. Psychol Addict Behav. 1994;8:116-127.
FULL TEXT
THIS ARTICLE HAS BEEN CITED BY OTHER ARTICLES
|
Transnational Comparisons of Adolescent Contraceptive Use: What Can We Learn From These Comparisons?
Santelli et al.
Arch Pediatr Adolesc Med 2008;162:92-94.
FULL TEXT
School-Based Drug Prevention Among At-Risk Adolescents: Effects of ALERT Plus
Longshore et al.
Health Educ Behav 2007;34:651-668.
ABSTRACT
Enabling Parents Who Smoke to Prevent Their Children From Initiating Smoking: Results From a 3-Year Intervention Evaluation
Jackson and Dickinson
Arch Pediatr Adolesc Med 2006;160:56-62.
ABSTRACT
| FULL TEXT
Tobacco, Alcohol, and Other Drugs: The Role of the Pediatrician in Prevention, Identification, and Management of Substance Abuse
Kulig and and the Committee on Substance Abuse
Pediatrics 2005;115:816-821.
ABSTRACT
| FULL TEXT
Cigarette Consumption During Childhood and Persistence of Smoking Through Adolescence
Jackson and Dickinson
Arch Pediatr Adolesc Med 2004;158:1050-1056.
ABSTRACT
| FULL TEXT
Smoking, Adolescents and Health: Isfahan Healthy Heart Programme-Heart Health Promotion from Childhood
Kelishadi et al.
Asia Pac J Public Health 2004;16:15-22.
ABSTRACT
New Inroads in Preventing Adolescent Drug Use: Results From a Large-Scale Trial of Project ALERT in Middle Schools
Ellickson et al.
Am. J. Public Health 2003;93:1830-1836.
ABSTRACT
| FULL TEXT
Television Viewing and Initiation of Smoking Among Youth
Gidwani et al.
Pediatrics 2002;110:505-508.
ABSTRACT
| FULL TEXT
Risk Factors for Dating Violence in Adolescents
McFee et al.
JAMA 2001;286:2813-2813.
FULL TEXT
The Role of Interpretation Processes and Parental Discussion in the Media's Effects on Adolescents' Use of Alcohol
Austin et al.
Pediatrics 2000;105:343-349.
ABSTRACT
| FULL TEXT
|
