JUVENILE OFFENDERS ARE AT increased risk for sexually transmitted diseases (STDs), including human immunodeficiency virus (HIV) infection. Although HIV infection is infrequently discovered in youth detention facilities and training schools, significant rates of STD among incarcerated adolescents are cause for concern and indicative of HIV risk. A high prevalence of 2 bacterial STDs, chlamydia and gonorrhea, has been documented in adolescents entering juvenile detention facilities.1,2 When compared with school-based populations, the median positivity for both chlamydia and gonorrhea is consistently higher for young women in juvenile correctional facilities (females: 14.8% vs. 9.6% and 5.6% vs. 1.6%).3 The asymptomatic nature of chlamydia and gonorrhea is cause for concern regarding the long-term reproductive health consequences for this population and the potential for further transmission of STDs in the community when they are released.
Development of effective sexual risk reduction intervention and STD prevention programs that clearly identify factors associated with the acquisition of STDs among juvenile detainees are urgently needed.4 Although a number of predictors of STD infections in youth have been identified, much of the research on adolescent STD risk has focused almost exclusively on the characteristics of individuals5,6 or on individual sexual risk behaviors. Admittedly, demographic characteristics such as female sex, minority race (other than Asian), and older age are consistently associated with testing positive for sexually transmitted infections.2,7–9 However, these are physiological attributes that cannot be modified by any prevention or intervention strategies. It is also well documented that sexual behaviors such as inconsistent condom use and having multiple sex partners are associated with adolescent STD infections.10–13 Unfortunately, not all efforts to change these adolescent risky sexual behaviors have been successful. These shortcomings suggest a need for consideration of additional factors to guide intervention studies.14 Identification of the psychologic and social influences on adolescent sexual behavior has recently been advocated as necessary to increase the effectiveness of interventions for adolescents.6,15
Identifying the influences on adolescent sexual behavior among juvenile offenders presents an even greater challenge. In addition to the increased intensity of behaviors that are commonly associated with the acquisition of STDs, the social and psychologic factors (eg, drug and alcohol use, physical and sexual abuse, and family structure) that could influence sexual risk-taking are very different for juvenile offender and nonoffender populations. Studies of incarcerated adolescents have revealed that marijuana and alcohol use are associated with HIV risk behavior16,17 and that there is a higher rate of substance abuse among juvenile offenders than among the general adolescent population.18
The prevalence of neglect and abuse (physical and sexual) among delinquents is substantially greater than that in the general adolescent population,19 and a youth’s history of physical and sexual abuse has been linked to several sexual risk-taking behaviors and outcomes.20,21 It is theorized that adverse psychologic and interpersonal effects of abuse can create apathy and reduce motivation for personal health and safety.22 Sexual abuse has been associated with risk factors such as the number of sexual partners,23,24 condom use,25 and STD infection.26,27
Parents and family life are also important influences on adolescent sexual behavior. Family structure (number of biologic parents in household), parental education and income, parental monitoring of adolescent behavior, parent–adolescent relationship quality, and parent–adolescent communication are all linked to STD infection.7,28,29 Living in a 2 biologic parent home appears to be the most beneficial family arrangement.30 For children and youth who do not live in an “intact” home, we hypothesize that living with relatives and having more than 1 adult caregiver can also reduce the probability of STD exposure. Relatives are much more likely to be emotionally invested in the youth than nonrelatives. Multiple adult caregivers can share supervision and monitoring of behavior and thus may be more effective at social control than a single parent/caregiver. Given the association of lower parental monitoring with greater delinquency, substance use, and sexual risk-taking,31,32 this family process variable appears to be a pertinent social factor for modifying juvenile offender risk for STD infection. There is little research, however, that examines the influence of family structure and process variables on STD risk of juvenile offenders.
The present research assesses the prevalence, multiple correlates, and gender differences in chlamydia and gonorrhea infections among adolescents incarcerated in a youth detention center in the southern region of the United States. We examine a number of demographic, psychologic, behavioral, and family factors that may affect the probability of infection with an STD. We argue that, although biologic factors, ie, age, physical development, and gender, and sexual risk behaviors, are more proximal correlates of STD transmission, psychosocial factors also influence sexual behavior and STD exposure. Thus, by identifying psychologic and family factors relevant for juvenile offenders, we hope to inform STD risk reduction interventions for this population.
Materials and Methods
The study was conducted in a juvenile detention facility in a Southern city that accepts males and females between the ages of 10 and 18. As part of the booking admissions process, all youth submitted urine samples, which were screened for Chlamydia trachomatis, Neisseria gonorrhoeae, and 7 illicit drugs. Between April 2002 and May 2003, 1816 of the juveniles booked into the detention center were screened. Only those juveniles who were aged 13 and older and who had been recently detained (within 3 days of booking into the facility) were approached to complete the survey. Over 800 (N = 816) youth who met inclusion criteria were approached and 763 (94%) assented to participate and completed the self-administered survey. Six hundred ninety detainees gave permission to link their STD test results to their survey responses. The majority of the juveniles who participated in the study were black (89%) and male (67%). Our sample was not significantly different for the population entering the facility on the basis of race and sex (90.5% black and 70.4% male).
Self-administered questionnaires are recommended to increase candid responses when soliciting sensitive information.33 Because many juvenile offenders have difficulty reading, self-report data were collected using audio computer-administered interview (A-CASI) technology. Use of this survey method has been shown to result in higher reporting rates for sensitive sexual behaviors by adolescents when compared with face-to-face interviews.34 Juveniles were taken from holding cells by project staff to a private room within the juvenile detention facility and away from the presence of any detention center personnel. A research associate explained the study to participants and asked them to complete the survey. It was made clear that participation was voluntary and that assent to participate could be withdrawn at any time. Survey content was communicated using both audio headphones and the computer screen. The juveniles used the computer keypad to enter their own responses. Mississippi State University Institutional Review Board and the federal Office of Human Research Protections Panel on Prisoners approved the study protocol.
The outcome of interest is STD status, ie, testing positive for either chlamydia or gonorrhea.
Demographic variables included the youth’s sex (gender), self-reported race, and age in years. These factors reflect biologic attributes that are not modifiable by the behavioral interventions and which are related to the probability of exposure to STD infection. 0
Psychologic variables included reports of past sexual abuse and alcohol and other drug (AOD) beliefs. Any lifetime report of unwanted or forced fondling, oral, vaginal, or anal sex was considered sexual abuse and was measured using a categorical yes/no response format. A 20-item measure of AOD beliefs was developed by taking items from the Alcohol Expectancy Questionnaire–Adolescent35 and by modifying items by substituting the word “marijuana” for “alcohol.” The response choices are true/false and respondents received 1 point for each true answer. Using principal component factor analysis, 9 of the alcohol and other drug expectations items loaded on a belief factor that we call enhances social/sexual relations. Some examples from this scale are “People feel sexier after a few alcoholic drinks” and “A few drinks makes it easier to talk with people.” Scores on the enhances social/sexual relations scale range from 0 to 9. Higher scores indicate that the respondent believes that alcohol and/or marijuana enhances social interactions and sexual experiences (Cronbach’s α = 0.78). A second scale, loss of control, consists of 8 items, including “When drinking alcohol, people do not feel in control of their behavior. They are apt to do something that they do not want to do” and “People do stupid, strange, or silly things when they smoke marijuana.” Scores on this scale range from 0 to 8 with higher scores indicating that the respondent believes that alcohol and/or marijuana reduces inhibition and causes a loss of control over behavior (Cronbach’s α = 0.76).
Substance use and sexual behaviors constituted the potential behavioral predictors of STDs. Youth alcohol and other drug use were assessed by self-reported use and by drug screening. Youth reported the frequency of alcohol and other drug use in the 3-month period before incarceration. We also tested urine for the presence of amphetamines, barbiturates, benzodiazepines, cocaine, opiates, phencyclidine, and marijuana. A history of STD infection has been shown to be a strong predictor of subsequent infection in research with other samples.28,36,37 Therefore, respondents were asked if they had ever been told they had an STD, responding to the item in a yes/no format. The number of lifetime sex partners was a count of all male and female sex partners. Because this measure was highly skewed with a few individuals reported improbably high numbers of partners relative to their young ages (mean, 14.7; median, 5), we transformed the variable using the natural log plus one (logged scores range from 0–6.73). We also collected information on sexual behaviors for the 3 months before incarceration, including sex with an intravenously drug user, sex with someone who has many other partners, casual sex, the joint occurrence of sex and alcohol use, the joint occurrence of sex and other drug use, pressured or unwanted sex, and trading money or drugs for sex.38
The final behavioral measure is risk reduction strategy. Study participants reporting any lifetime sexual activity (683 of 763, 89.5%) were categorized according to their recent sexual activity and condom use. Those youths who denied oral, vaginal, or anal sex during the 3 months before incarceration were considered temporarily sexually abstinent. Those who reported recent sexual activity and who reported 100% condom use were categorized as consistent condom users, whereas those reporting less than 100% condom use were classified as inconsistent condom users.
A measure of family structure was derived from 1 question asking about the juvenile’s living arrangements before incarceration. Although 71.4% of the study participants lived with 1 biologic parent, there were a variety of other familial and nonrelative households. Adolescents living with both biologic parents (the reference category) were compared with those who were living in a stepfamily (1 biologic parent and a stepparent), living in an extended family (1 biologic parent and other adult relatives or 2 adult relatives such as grandparents), living in a single (biologic) parent family, and those living with others (which could include living with a single adult relative but not a biologic parent).
Family process variables included parental communication, involvement, and supervision. Parental negative communication to the adolescent was measured by a single question: “How often do your parents yell, shout or scream at you?” with 5 response choices ranging from “never” coded 4 to “always” coded 0 such that higher scores indicated less frequent negative communication. Parental involvement was measured by “How often have your parents taken part or shared in your school activities?” Respondents had 5 response choices ranging from “not at all” to “very much” with higher scores indicating greater parental involvement. Supervision and monitoring of the adolescent was measured by the mean of 4 items: How much do your parents or guardians know about where you go at night; … about what you do with your free time; … about who your friends are; and when you go out at night, do they set curfews (tell you when you must get home). Response choices ranged from “doesn’t know at all/never” coded as 0 to “always knows/always” coded as 3. Cronbach’s α for the supervision and monitoring scale was 0.70. Following the example of DiClemente and colleagues,32 adolescents scoring 2.5 or higher were classified as exposed to more parental monitoring; the remainder as having less parental monitoring.
To address issues of multicollinearity and model instability that are inherent when a large number of potential predictors are available for analyses, a model-building strategy was used. Following Hosmer and Lemeshow,39 we explored bivariate relationships and retained variables for further analyses if the P value was less than or equal to 0.25. Significance for the multivariable regression models was defined by a P value less than or equal to 0.05. We also computed a pseudo-R2, analogous to the coefficient of determination from Ordinary Least Squares regression to measure the proportion of variance in STD status explained by all predictor variables.40 In the final logistic regression analysis, we entered predictor variables into the model in sets of demographic, psychologic, behavioral, and family variables to determine the contribution of each set of variables to the total variance. Self-report data and STD test results were available on 690 study participants. Almost all of the study participants reported that they have been sexually active (89.5%) and most (64.5%) reported engaging in sexual intercourse in the 3 months before incarceration. Youth who denied any lifetime sexual activity were excluded from bivariate and logistic regression analyses examining predictors of STD status. Thus, logistic regression analyses were conducted for 618 sexually active study participants and separately for males (n = 400) and females (n = 218).
Prevalence of Chlamydia, Gonorrhea, and Illicit Drug Use
STD and drug use screening was conducted by urinalysis. STD test results available on 1789 detainees revealed that 231 or 12.9% of the adolescents tested positive for chlamydia (24.7% of females and 8.1% of males) and 57 or 3.2% tested positive for gonorrhea (7.3% of females and 1.5% of males). One third (598 youths) tested positive for 1 of 7 illicit drugs. Marijuana was the most frequently detected drug (36% prevalence among those screened) followed by cocaine (2.8%).
Among the study participants with self-report information and STD test results (n = 690), rates of infection varied by reports of sexual activity. Interestingly, 7.1% of participants who denied ever engaging in sexual intercourse tested positive. Of those participants who at some point have been sexually active, but who denied sexual activity in the 3 months before incarceration, 12.7% tested positive. Twenty-one percent of those reporting recent sexual activity tested positive.
Bivariate analyses reveal 13 variables are associated with the presence of an STD (P ≤0.25): sex (gender), age, race, sexual abuse, alcohol use, history of STD infection, the joint occurrence of sexual activity and alcohol use, number of lifetime partners, risk reduction strategy, family structure, parental supervision and monitoring, parental involvement and communication. The AOD Beliefs scales were associated with STD status when examined by gender and thus were retained for further analyses. Table 1 presents the percentage testing positive for either chlamydia or gonorrhea for each independent variable. Females were more likely to test positive for chlamydia or gonorrhea than males. Prevalence of infection generally increases with age. Black study participants are more than twice as likely to test positive as all other racial groups combined (52 of the 57 in the other category self-identified as white, 1 as a Native American and the remaining 4 as other; no participant identified as Asian or Hispanic). Participants who reported sexual abuse, the joint occurrence of sexual activity and alcohol use before incarceration, and a history of STD infection were more likely to test positive than those who did not make these reports. The only substance use measure found to be associated with STD status was frequency of alcohol use. As alcohol use increases, so does the prevalence of STD infection. The AOD Beliefs scales and the number of lifetime partners are continuous measures but were each divided at the median value for ease of presentation in the table. As the number of lifetime sexual partners increases, so does the likelihood of infection. Higher scores on the AOD Beliefs–Loss of Control scale was associated with lower STD infection for males, whereas higher scores on the AOD Beliefs–Enhances Sex scale was associated with lower STD infection for females. The sexually abstinent juveniles (previously sexually active, but no reported sex in the 3 months before incar-ceration) were about half as likely to test positive as study participants who reported recent sexual activity. The rates of infection were very similar for juveniles reporting consistent (20.2%) and inconsistent condom use (23.1%).
In the bivariate analysis of family structure, study participants living with both biologic parents have the lowest rate of STD infection (9.9%), whereas those living in stepfamilies have the highest (23.8%). More supervision and monitoring by adult caregivers was associated with a slightly lower STD infection rate (15.7%), but not significant, than less supervision and monitoring (20%, P = 0.22). Lower parental involvement in school activities (not at all) and more negative communication (almost always and always yelling at the youth) were also associated with higher rates of STD infection.
Logistic Regression Analyses
Logistic regression analyses results are displayed in Table 2. Crude odds ratios are reported for bivariate relationships between STD status and each independent variable. Adjusted odds ratios reflect the change in the odds for that variable when all other variables in the equation are controlled. Several variables such as sexual abuse, alcohol use, history of STD, risk reduction strategy, parental involvement, and parental communication that were significant at the conventional level (P ≤0.05) in the bivariate analyses are no longer significant in the full logistic regression model. Adjusting for all other variables in the model, females are just over 4 times more likely than males to test positive for either chlamydia or gonorrhea. Black study participants are 3.4 times as likely to test positive as juveniles of other races. Those reporting 1 or more joint occurrences of sexual activity while under the influence of alcohol are almost 2 times more likely to test positive than study participants who did not engage in this behavior. In addition, for every 1-unit increase in the logged number of lifetime sex partners, the odds of testing positive increased by 37%.
The only family factor that remained significant in the multivariable logistic regression model was family structure. In the bivariate regression, living in an extended family household was not significantly different from living with 2 biologic parents. However, youth living in all other types of arrangements, ie, stepfamily, single parent, or living with other adults, were over 2 times more likely to test positive than youth living with 2 biologic parents. When demographic, psychologic, behavioral, and family process variable are held constant, the only family structure significantly associated with increased risk of STD infection was the stepfamily. Youth living in stepfamilies are 2.5 times more likely than youth living in a 2-biologic parent home to test positive for either chlamydia or gonorrhea.
The total explained variance for the multivariable regression model was 20.8%. The 3 demographic variables, sex (gender), race and age, explained 52% of the variance. The behavioral factors, ie, alcohol use, sexual activity under the influence of alcohol, number of lifetime sex partners, and risk reduction strategy, accounted for approximately one third (32.2%) of the explained variance. Psychologic factors and family factors accounted for 8.6% and 7.2%, respectively.
Predictors of STD status were also examined separately by gender. Only statistically significant predictors are displayed in Table 3. As can be seen from this table, the predictors of infection are very different for males and females. For males, the strongest predictor was having a history of an STD. Older age was also a factor for males, because a 1-year increase in age increased the odds of testing positive for chlamydia or gonorrhea by 43%. Race is a predictor for females. Compared with females of other races, black females were almost 4 times more likely to test positive.
Sexual behaviors associated with STD infection are different for boys and girls. For females, sexual activity in conjunction with alcohol use and the number of sexual partners are risk factors. For males, it is condom use. Compared with young men who had been sexually active in the past, but reported no sexual activity in the 3-month period just before incarceration, males who reported consistent condom use were almost 4 times as likely to test positive. Surprisingly, adolescent males who reported inconsistent condom use had lower odds for testing positive than those reporting 100% condom use.
Lastly, AOD beliefs are related to STD status but vary by gender. Adjusting for all other variables in the model, females who scored higher on the AOD Beliefs–Enhances Sex scale were less likely to test positive for chlamydia or gonorrhea than females with lower scale scores. The opposite was true for young men. Higher scores on the AOD Beliefs–Enhances Sex scale were marginally associated with greater likelihood of testing positive for male participants. Adjusting for all other variables in the model, AOD Beliefs–Loss of Control scale remained predictive for males. Males with higher scores on this scale were less likely to test positive than young men with lower scores. These opposing gender-specific AOD Beliefs account for the lack of significant effects for the total sample.
This study assessed the prevalence of 2 common STDs, chlamydia and gonorrhea, and assessed predictors of STD status in a high-risk group of youth, incarcerated juveniles. Potential predictors of STD infection such as demographic and behavioral factors were identified from the empiric literature. Our findings are consistent with previous studies of high-risk youth. Who gets infected is determined primarily by demography and the individual’s sexual risk behavior. In addition, we identified psychologic and family factors associated with a host of adolescent well-being outcomes. We found that the psychosocial variables examined do contribute to a more comprehensive view of STD risk. Moreover, analyzing the data by gender revealed significant predictors for girls or for boys that were masked when the sexes are combined.
We examined a number of family factors associated with adolescent well-being and health outcomes. The results showed that family structure, parental involvement in school activities, and communication of the parent/guardian with the adolescent were each separately related to risk for STD infection. Controlling for demographic and behavioral factors, we found that household composition predicts STD status. Although 2 adult guardians are certainly better than 1, we found that living in a stepfamily was a greater risk for adolescent STD infection than living with a single parent. We also discovered that living with 2 related adults such as grandparents was just as protective as living with 2 biologic parents. Our measure of family structure may serve as a proxy for the quality of relationships between adolescents and their adult caregivers.
Although there is general consensus on the relationship between adolescents’ alcohol/drug use and sexual risk behavior, the definitive nature of this association has been the subject of debate.41 Consistent with other research, our study confirms that the link between alcohol use and sexual intercourse is a significant predictor of STD status. This study, however, helps to delineate that relationship by substantiating that alcohol in the environment where sexual activity takes place is an important risk factor for female juvenile offenders. Gender-specific beliefs about the effects of alcohol and marijuana provide clues about the sexual decision-making process. For these young women, believing that substance use can facilitate socializing with peers and sexual activity is a protective factor against STD. Perhaps these beliefs help young women to recognize that men may use alcohol and other drugs to encourage sexual activity. On the other hand, believing that alcohol and other drugs can reduce inhibitions and lessen self-control is a protective factor for these young men. Perhaps males who believe alcohol makes them less in control of their actions are more apt to take precautions against this perceived loss of control than others who do not hold such beliefs. Because the literature is divided on this issue, longitudinal studies and ethnographic studies that address sexual decision-making about substance abuse might prove helpful in clarifying this relationship.
The results of this study are potentially biased by the convenience nature of the sample and restricted access to STD testing results. Of all the participants who were admitted to the youth correctional facility, almost half were released before project staff could recruit them for the study. Although 91.3% of those who were approached agreed to be in the study, it is possible that the most risky youths opted out of the study or that the least risky youth were among those who were quickly released because they had less serious charges pending. Although STD results were obtained for the majority of the participants, those respondents who completed the survey had the option of not allowing us access to the results. Of those consenting to be in the study, 690 authorized the release of their STD/drug use test results. Twenty-six study participants refused to release urine test results, and another 47 were not tested. Still, this limitation may be meaningful if the most risky youth refused to release the results of their tests. However, a comparison of youth with complete data with those who were not tested and those who refused to release their urinalysis results showed no statistically significant differences for any predictor variable (all P >0.05). In any case, those who were approached and were interviewed had risks consistent with other published studies of incarcerated youth. Finally, our sample is representative of the race and gender distribution of adolescents entering the facility, but is not representative for age because we excluded 18% of the population who were under the age of 13.
Another limitation of the study involves the veracity of reports of condom use and sexual activity. These measures were used to exclude some study participants from analyses and to place others into risk-reduction strategy categories. Youth who denied ever engaging in sex were excluded from analyses predicting STD status, yet 7% of them tested positive for either chlamydia or gonorrhea. Twelve percent of sexually active youth who reported no sex in the 3 months before incarceration, the abstainers, tested positive. Of the juveniles who reported current sexual activity, 55% (238 of 433) reported consistent condom use and 20% of them tested positive.
The inconsistency between self-report behavior and STD test results is vexing. Despite a data collection protocol designed to minimize social desirability bias and a recall period selected to maximize accuracy,42 we suspect that sexual abstinence and condom use may have been exaggerated. Other studies found substantially lower rates of condom use among juvenile offenders.43–45 A possible explanation is that our study participants were not using condoms correctly contributing to breakage and reduced protection from infection. Another explanation for the inconsistency between self-report data and STD test results is that the youths could have acquired the infection some time before the 3-month reporting period.
The gender differences that emerged in this study have implications for STD risk reduction intervention. If different factors predict acquisition of an STD for girls and for boys, then this may suggest that individual-level interventions may be more effective when they are organized for sex-specific groups rather than in gender-neutral interventions. One further finding from this study is that STD/HIV prevention interventions also should include elements of alcohol risk education, demonstrations of correct condom use for sexually active youth, and family communication.
Finally, we recommend standard STD screening for all youth admitted to a youth detention facility. Given the high rates of STD in this population (roughly 10 times higher than the nonincarcerated adolescent population in the same catchment area as the detention center) and the increased risk for acquisition if an STD-infected adolescent is exposed to HIV, detention facilities offer a window in time when it is possible to test and treat very high-risk youths.
1. Katz AR. Prevalence of chlamydial and gonorrheal infections among females in a juvenile detention facility. J Community Health 2004; 29:265–269.
2. Kelly PJ, Bair RM, Baillargeon J, et al. Risk behaviors and the prevalence of chlamydia in a juvenile detention facility. Clin Pediatr 2000; 39:521–527.
3. Centers for Disease Control and Prevention. HIV/AIDS Surveillance Report 2001. September 2002.
4. Institute of Medicine (IOM). The Hidden Epidemic: Confronting Sexually Transmitted Diseases. Washington, DC: National Academy Press, 1997.
5. DiClemente RJ, Wingood GM. Expanding the scope of HIV prevention for adolescents: Beyond individual-level interventions. J Adolesc Health 2000; 26:377–378.
6. Kotchick BA, Shaffer A, Forehand R, et al. Adolescent sexual risk behavior: A multi-system perspective. Clin Psychol Rev 2001; 21:493–519.
7. Crosby R, Wingood GM, DiClemente RJ, et al. Family-related correlates of sexually transmitted disease and barriers to care: A pilot study of pregnant African American adolescents. Fam Community Health 2002; 25:16–27.
8. Kent CK, Branzuela A, Fischer L, et al. Chlamydia and gonorrhea screening in San Francisco high schools. Sex Transm Dis 2002; 29:373–375.
9. Risser JMH, Risser WL, Gefter LR, et al. Implementation of a screening program for chlamydial infection in incarcerated adolescents. Sex Transm Dis 2001; 28:43–46.
10. Crosby R, DiClemente RJ, Wingood GM, et al. Predictors of infection with Trichomonas vaginalis
: A prospective study of low income African-American adolescent females. Sex Transm Infect 2002; 78:360–364.
11. DiClemente RJ, Wingood GM, Sionean C, et al. Association of adolescents’ history of sexually transmitted disease (STD) and their current high-risk behavior and STD status: A case for intensifying clinic-based prevention efforts. Sex Transm Dis 2002; 9:503–509.
12. Fortenberry JD, Brizendine EJ, Katz B, et al. Subsequent sexually transmitted infections among adolescent women with genital infection due to Chlamydia trachomatis, Neisseria gonorrhoeae
, or Trichomonas vaginalis
. Sex Transm Dis 1999; 26:26–32.
13. Tyler KA, Whitbeck LB, Hoyt ER, et al. Predictors of self-reported sexually transmitted diseases among homeless and runaway adolescents. J Sex Res 2002; 37:369–377.
14. DiClemente RJ. Looking forward: Future directions for HIV prevention research. In: Peterson JL, DiClemente RJ, eds. Handbook of HIV Prevention. New York: Kluwer Academic/Plenum, 2000:311–324.
15. Pedlow CT, Carey MP. Developmentally appropriate sexual risk reduction interventions for adolescents: Rationale, review of interventions and recommendations for research and practice. Ann Behav Med 2004; 27:172–184.
16. Kingree JB, Braithwaite R, Woodring T. Unprotected sex as a function of alcohol and marijuana use among adolescent detainees. J Adolesc Health 2000; 27:179–185.
17. Otto-Salaj LL, Gore-Felton C, McGarvey E, Canterbury RJ. Psychiatric functioning and substance use: Factors associated with HIV risk among incarcerated adolescents. Child Psychiatry Hum Dev 2002; 33:91–106.
18. DiClemente RJ, Lanier M, Horan P, Lodico M. Comparison of AIDS knowledge, attitudes, and behavior among incarcerated and a public school sample in San Francisco. Am J Public Health 1991; 81:628–630.
19. Weibush R, Freitag R, Baird C. Preventing delinquency through improved child protection services. Juvenile Justice Bulletin. Washington, DC: Office of Juvenile Justice and Delinquency Prevention, July 2001.
20. Roosa MW, Tein JY, Reinholtz C, Angelini PJ. The relationship of childhood sexual abuse to teenage pregnancy. J Marriage Fam 1997; 59:119–130.
21. Wyatt GE, Myers HF, Williams JK, et al. Does a history of trauma contribute to HIV risk for women of color? Implications for prevention and policy. Am J Public Health 2002; 92:660–665.
22. Rosenfeld S, Lewis D. The hidden effects of childhood sexual abuse on adolescent and young adult HIV prevention: Rethinking AIDS education, program development, and policy. AIDS Public Policy J 1993; 8:181–186.
23. Luster T, Small SA. Sexual abuse history and number of sex partners among female adolescents. Fam Plann Perspect 1997; 29:204–211.
24. Acoca L, Dedel K. No Place to Hide: Understanding and Meeting the Needs of Girls in the California Juvenile Justice System. San Francisco, CA: National Council on Crime and Delinquency, 1998.
25. Silverman JG, Amaro H. The relationship between sexual abuse and sexual risk among high school students: Findings from the 1997 Massachusetts Youth Risk Behavior survey. Matern Child Health J 2000; 4:125–134.
26. Vermund SV, Alexander-Rodriguez T, Macleod S, Kelley KF, Alexander-Rodriguez T. History of sexual abuse in incarcerated adolescents with gonorrhea or syphilis. J Adolesc Health Care 1990; 11:449–452.
27. Zierler S, Feingold L, Laufer D, et al. Adult survivors of childhood sexual abuse and subsequent risk of HIV infection. Am J Public Health 1991; 81:572–575.
28. Crosby R, Leichliter JS, Brackbill R. Longitudinal prediction of sexually transmitted diseases among adolescents: Results from a national study. Am J Prev Med 2000; 18:312–317.
29. Sionean C, DiClemente RJ, Wingood GM, et al. Socioeconomic status and self-reported gonorrhea among African American female adolescents. Sex Transm Dis 2001; 28:236–239.
30. McLanahan, SS, Sandefur G. Growing Up With a Single Parent: What Hurts, What Helps. Cambridge, MA: Harvard University Press, 1994.
31. Patterson GR, Stouthamer-Loeber M. The correlation of family management practices and delinquency. Child Dev 1984; 55:1299–1307.
32. DiClemente RJ, Wingood GM, Crosby R, et al. Parental monitoring: Association with adolescents’ risk behaviors. Pediatrics 2001; 107:1363–1368.
33. Sonenstein F. Using self report to measure program impact. Child Youth Serv Rev 1997; 19:567–587.
34. Schroder KEE, Carey MP, Vanable PA. Methodological challenges in research on sexual risk behavior: II. Accuracy of self-reports. Ann Behav Med 2003; 26:104–123.
35. Brown SA, Christiansen BA, Goldman MS. The Alcohol Expectancy Questionnaire: An instrument for the assessment of adolescent and adult alcohol expectancies. J Stud Alcohol 1987; 48:483–491.
36. Tubman JG, Langer LM, Calderon DM. Coerced sexual experiences among adolescent substance abusers: A potential pathway to increased vulnerability to HIV exposure. Child Adolesc Soc Work J 2001; 18:281–303.
37. Wiesenfeld HC, Lowry DLB, Heine RP, et al. Self-collection of vaginal swabs for the detection of chlamydia, gonorrhea, and trichomoniasis: Opportunity to encourage sexually transmitted disease testing among adolescents. Sex Transm Dis 2001; 28:321–325.
38. Robertson AA, Levin ML. AIDS knowledge, condom attitudes and risk-taking sexual behavior of substance-abusing juvenile offenders on probation or parole. AIDS Educ Prev 1999; 11:450–461.
39. Hosmer DW, Lemeshow S. Applied Logistic Regression. New York: John Wiley and Sons, 1989.
40. Menard S. Applied Logistic Regression Analysis. Thousand Oaks, CA: Sage Publications, 1995.
41. Fortenberry JD. Adolescent substance use and sexually transmitted diseases risk: A review. J Adolesc Health 1995; 16:304–308.
42. Kauth MR, St. Lawrence JS, Kelly JA. Reliability of retrospective assessments of sexual HIV risk behavior: A comparison of biweekly, 3-month, and 12-monthself reports. AIDS Educ Prev 1991; 3:207–2124.
43. Gillmore MR, Morrison DM, Lowery C, et al. Beliefs about condoms and their association with intentions to use condoms among youths in detention. J Adolesc Health 1994; 15:28–237.
44. Magura S, Shapiro JL, Kang SY. Condom use among criminally-involved adolescents. AIDS Care 1994; 6:595–601.
45. Rickman RL, Lodici M, DiClemente RJ, et al. Sexual communication is associated with condom use by sexually active incarcerated adolescents. J Adolesc Health 1994; 15:383–388.