Violence is an ever-growing public health concern, and adolescents are increasingly exposed to violence in their personal lives, homes, and communities. The most recent Youth Risk Behavior Surveillance System data (Centers for Disease Control and Prevention [CDC], 2010) reported widespread violence among adolescents as greater than 30% of students have reported being involved in a physical fight and approximately 20% have been bullied at school in the previous 12 months. Exposure to interpersonal, family, and community violence has been linked to adverse health outcomes including risky sexual behavior (Barroso, Peters, Kelder, Conroy, Murray, & Orpinas, 2008; Brady, 2006; Walton, Resko, Whiteside, Chermack, Zimmerman, & Cunningham, 2011). Behaviors that begin in adolescence may set the trajectory for lifetime behavioral norms and increase an individual's risk for sexually transmitted infections, unintended pregnancy, and the acquisition of HIV.
Exposure to different levels of violence throughout adolescence has been linked to an overall poorer self-rating of health (Bossarte et al., 2009). Boynton-Jarrett, Ryan, Berkman, and Wright (2008) analyzed data collected from years 1997–2004 of the National Longitudinal Survey of Youth to investigate the impact of violence on the self-rated health of adolescents. Among a sample of 8224 adolescents aged 12–18, exposure to violence had a negative effect on health rating. Adolescents with health ratings of poor/fair reported higher rates of not feeling safe or being threatened at school, being repeatedly bullied, or witnessing a shooting.
In addition to reduced levels of self-reported health, adolescents exposed to violence demonstrate higher levels of sexual risk taking. Berenson, Wiemann, and McCombs (2001) reported that among 517 sexually active females under the age of 18, those who witnessed violence in the community were two to three times more likely to have a partner with multiple other partners. When these adolescents had personally experienced violence, they were two to four times more likely to report multiple sexual partners, early initiation of intercourse, and sexually transmitted infections. Increased sexual risk was also investigated by Brady, Tschann, Pasch, Flores, and Ozer (2008). These researchers reported that there was an association between adolescents' exposure to violence and their risk behaviors. Among a sample of 302 Hispanic and non-Hispanic adolescents, those who were involved in personal or interpersonal violence during adolescence had an increased risk for sexual activity and substance use by age 19.
Adolescent females who presented to the emergency department for gynecologic exams were the subjects of a study by Whiteside, Walton, Stanley, Resko, Chermack, Zimmerman, et al. (2009). These researchers investigated the link between aggression in dating relationships and sexual risk behaviors. Among 148 adolescent females aged 14–18 there was a 49% rate of dating aggression within the previous year and 14% rate of pregnancy at the time of evaluation. Additionally, 40% of these females reported having more than one sexual partner within the past 12 months.
Walton, Resko, Whiteside, Chermack, Zimmerman, & Cunningham (2011) also explored the relationship between violence, substance use, and sexual risk among urban adolescents who utilized emergency department services. The results suggested a clustering of risk behaviors that were interwoven to increase overall risk. Among a sample of 1576 male and female adolescents aged 14–18, 60% were sexually active. Violence correlates of risky sexual behavior included weapon carriage and interpersonal peer violence in the presence of substance use. Substance use and binge drinking were independently related to multiple sexual risk behaviors including a greater number of sexual partners, lack of condom use, and substance use prior to sexual activity.
Although there is a link between violence and risky sexual behaviors in adolescents, less is known about gender differences in sexual behavior among male and female adolescents who have experienced different types (personal, interpersonal, community) violence. Additionally, this study aimed to explore whether the type of violence was correlated with sexual risk behaviors. Therefore, the purpose of this study was to use an ecological lens to explore gender differences in risky sexual behavior among urban adolescents exposed to violence at the personal, interpersonal, and community levels.
This study was a secondary analysis of data collected from a larger intervention trial targeting adolescents and drinking behaviors (Bernstein, Heeren, Edward, Dorfman, Bliss, Winter, et al., 2010). Participants between the ages of 14 and 21 were recruited from the pediatric emergency department of a large, inner-city teaching hospital. Research assistants asked adolescents who were waiting to be seen for care to complete a “health and safety needs survey.” Survey questions were drawn adapted from standard measures used by the CDC Youth Risk Behavior Surveillance (CDC, 2010). Over 7800 adolescents completed the survey. The first 2000 surveys were entered for analysis, after which a random 10% of surveys were entered. This method allowed for increased randomization and diversity of the sample. In total, 3742 surveys were available for analysis. Institutional Review Board approval was granted for both the original study and this secondary analysis.
The sample used for the secondary analysis included male and female adolescents who had complete data on violence and sexual risk variables. Of the 3742 adolescents with complete data, those who were currently sexually active (past 3 month sexual activity) were included in the analysis. This allowed for investigation into the relationship of violence on current sexual behavior. The final sample size for the secondary analysis was 2560.
An ecological model served as the framework to guide the analysis. Ecological models are increasingly recognized as accurately reflecting the complexities of health, especially those concerning issues of public health, because of the emphasis on examining multiple levels of influence on behavior (Institute of Medicine, 2000). A central and distinguishing assumption of an ecological approach is that there is equal attention on the individual and the individual's environment, both of which exert varying degrees of influence on health and health-related behaviors (Grzywacz & Fuqua, 2000). For this study, an ecological framework was useful to understand the dynamic influence between the series of interdependent environmental exposures to violence and adolescents' individual sexual behaviors.
To capture personal, interpersonal, and community effects of violence, questions relating to each adolescent's experience with the different levels of violence were identified from the original study questions. Personal violence was measured by assessing whether adolescents carried a weapon or were involved in a physical fight, as these are behaviors that contribute to personal involvement with violence (CDC, 2010). Interpersonal violence was measured by assessing threats or injury involving another individual. Community violence was assessed by measuring whether adolescents felt safe in their environment. These questions are outlined in Table 1.
Risky sexual behavior was defined in accordance with Silverman, McCauley, Decker, Miller, Reed, & Raj (2011). These researchers have previously defined more than one sex partner in the past 3 months as risky. Using this definition, the outcome measure (risky sexual behavior) was defined as having greater than one sexual partner in the past 3 months.
Descriptive statistics were computed for each of the study variables. The distribution of characteristics was analyzed with univariate statistics and inspected to assess normality, skewness of data, and to identify outliers. Differences between males and females on the background variables of age, race, education, and age of first intercourse were assessed using t-tests (continuous variables) or chi-square (categorical variables) analysis. Although the analysis did not reveal any significant differences, background variables typically found to be confounders in the literature (education, age of first intercourse, and race/ethnicity) were controlled for in the subsequent analysis. Violence variables (Table 1) known to be related to risky sexual behavior were included in the analysis.
The data were analyzed using logistic regression testing a model containing the variables of education, race/ethnicity, carried a weapon (past 30 days), physical fight (past 12 months), threatened or injured (at school), threatened or injured (off school), safe in intimate relationship, and safe in neighborhood as predictors of risky sexual behavior in the past 3 months for both males and females. Level of significance was .05. Statistical Package for the Social Sciences (SPSS v18) was used to perform all statistical analysis.
The sample of 2560 adolescents consisted of 1627 females and 933 males. The males had a mean age of 19 years and were slightly older than the females (mean age = 18.9). A majority of the sample (60.2%) identified themselves as Black or African American and 30.9% reported still being in high school at the time of the survey (Table 2). The mean age of first intercourse was higher in females (15.4 years) as compared to males (14.4) and more than 65% of the female adolescents and 85% of the male adolescents reported three or more lifetime sexual partners.
Violence and risky sexual behavior
The results of the bivariate analyses examining the relationship between the violence variables and risky sexual behavior are reported in Table 3. Greater than 27% (n= 712) of the sample reported having more than one sex partner in the past 3 months. Being bullied in the past 12 months and feeling safe at school were not significantly related to risky sexual behaviors, and therefore were not included in regression analysis.
Correlates of risky sexual behavior
Results of the gender-specific logistic regressions are presented in Tables 4 and 5. These analyses examined the gender differences in risky sexual behavior among adolescents exposed to varying levels of violence. These analyses revealed that race/ethnicity (p= .019) and being in a physical fight (p= .001) were individual correlates of risky sexual behavior for male adolescents. Specifically, males who identified as Black or African American were more likely to report having more than one sexual partner in the past 3 months. Age at first intercourse was not correlated with risky sexual behavior for adolescent males.
For adolescent females, age at first intercourse (p < .001), safety in intimate relationship (p= .020), and carrying a weapon in the past 30 days (p= .029), were all correlated with of risky sexual behavior. In contrast with male adolescents, race/ethnicity was not found to be related to risky sexual behavior among females. Older age at first intercourse for females was associated with having more than one sex partner in the past 3 months. Additionally, female adolescents who reported not feeling safe in their relationship were at greater risk of having greater than one sexual partner in the past 3 months. Experiences of community violence were not associated with risky sexual behavior for either male or female adolescents.
The study results must be viewed in terms of the limitations. This study was a secondary analysis of existing data and therefore analysis was limited to variables that were previously collected. Therefore, potentially influential variables may not have been available for this analysis. The original study utilized a cross-sectional design, and therefore causality cannot be inferred. Other limitations include the use of self-report survey methods. Because of social desirability, participants may answer survey questions in what they perceive is the best or most acceptable answer. Only adolescents who were sexually active were included in the analysis and the results may not apply to adolescents who are currently abstinent. Additionally, the original sample was drawn from a single academic medical center in one geographical area. Therefore, the results may not be generalizable to other populations or different locations.
Despite limitations, this study has important strengths. The research sample was large and diverse. Specifically, more than 80% of the adolescents reported belonging to a racial or ethnic minority and the sample was exclusively recruited from an inner city urban setting. These groups are traditionally more difficult to recruit and under-represented in research. The sample size and diversity of gender and background made it possible to examine relationships between subgroups. Additionally, the use of this data allowed for the investigation of a wide range of potential variables and included measures of individual-, interpersonal-, and community-level exposures to violence.
The purpose of this study was to use an ecological lens to explore gender differences in risky sexual behavior among urban adolescents exposed to violence. The results of this study suggest that adolescent involvement in violent behavior has an impact on sexual risk, adding to our knowledge of this relationship.
For both genders, risky sexual behavior was associated with personal experiences of violence. Adolescent females who reported carrying a weapon in the past 30 days and adolescent males who reported being in a physical fight were more likely to have more than one sex partner in the past 3 months. This is consistent with previous researchers who have linked perpetrating violence to a greater number of sexual partners in adolescence (Brady, Tschann, Pasch, Flores, & Ozer, 2008). Additionally, on the interpersonal level for adolescent women, not feeling safe in a relationship was associated with increased risky sexual behavior. This association was not seen in adolescent males and experiences of community violence were not associated with risky sexual behavior for either male or female adolescents.
The adolescents in this study reported multiple sexual risk behaviors that are consistent with national data on adolescent sexual activity, including age at first intercourse and number of sexual partners (Abma, Martinez, & Copen, 2010; CDC, 2010). The age at first intercourse was 14 for males and 15 for females. Additionally, 11% of the total sample reported having three or more lifetime sexual partners. Earlier age at first intercourse has been linked to a greater number of lifetime sexual partners, most likely because of a longer period of time in which to accumulate partners (Abma, Martinez, & Copen, 2010).
Racial differences existed in the relationship of violence to sexual risk. African American race was an individual correlate of having more than one sex partner in the previous 3 months for male adolescents. This relationship did not exist for females of any race and those adolescents who identified as Hispanic, Caucasian, or other. These results are similar to national data that supports the finding of Black males being the group that reports the highest number of sexual partners compared with other racial and ethnic groups (CDC, 2010). Although the reasons for this are unclear, this may reflect influences of peer behavior within this minority group.
From an ecologic perspective, environmental influences on health and behavior exist at varying levels, beginning with personal interactions and extending outward to include societal influences. Viewing the results through an ecological lens helps to understand how exposure to different levels of violence influences sexual risks and how to begin addressing those risks in the most effective way. The results of this study suggest that involvement with violence at the personal level may be a stronger predictor of risky sexual behaviors than experiencing violence at the community level. Contextually, experiences of personal violence involving physical fighting and the use of weapons places adolescents in situations of increased risk. This risk includes safety for themselves and also safety to others who may inadvertently become victims of violence. This association of personal violence and sexual risk among adolescent females has previously been explored by Berenson, Wiemann, & McCombs (2001) who reported that adolescent girls who experienced violence were more likely to have multiple sexual partners. Our study demonstrated that involvement with personal violence also increases sexual risk among male adolescents.
Suggestions for future research include considering more proximal interventions that target individual level behaviors to reduce both violence and risky sex. Interventions aimed at personal behavioral change that directly involve adolescent skill building may be more meaningful to individual adolescents than community-level interventions that are more distal. A focus on specific activities such as conflict resolution, effective communication, and negotiation skills may help adolescents recognize situations where violence may escalate. Allowing adolescents to practice these skills through role-playing activities may help reduce interactions that could result in violence. This same skill set may also provide adolescents with the ability to successfully reduce their risk during sexual encounters by encouraging partner communication and negotiation of safer sex behaviors.
Due to the limited amount of time nurse practitioners (NPs) have to spend with adolescents during clinical appointments, it is important that NPs understand how risk behaviors cluster so they can adequately screen adolescent clients. These results support national data and provide additional evidence of multiple risk behaviors that NPs need to be aware of when providing care to adolescent patients. Researchers have reported clustering of violence and risk behaviors (Muula, Rudatsikira, & Siziya, 2008; Walton, Resko, Whiteside, Chermack, Zimmerman, & Cunningham, 2011) and therefore clinicians need to be aware of the potential impact that violence has on other health risks, including sexual risk behaviors.
Adolescents who are exposed to violence should also be assessed for sexual risk behavior. This can be done by asking violence-screening questions that are direct and focused and woven into conversations about personal and sexual health and safety. Adolescents who have a positive violence screen, especially for involvement in personal violence, should have this addressed by clinicians in a way that reflects nonjudgmental concern for overall health. Asking direct questions about violence and sexual behaviors begins the dialogue about health promoting and risk reduction behaviors. All clinical interactions need to be viewed as an opportunity for providers to discuss threats to health and safety and discuss ways to improve health and reduce risk.
NPs can also play a role in reducing violence and sexual risk through a variety of community and school-based programs. NPs are often very involved in community programs and may be the first healthcare provider an adolescent interacts with at a health center, student health services, or an outreach program. They may also be the first to recognize an adolescent who is at risk for violence and risky sexual behavior. NPs who work in community or school settings can play a pivotal role in education and referral of at risk adolescents.
The results of this study indicate that a relationship may exist between exposure to personal and interpersonal violence and sexual risk among urban adolescents. Clinicians who interact with adolescent patients need to be aware of this link and view clinic visits as opportunities for risk reduction. An increased awareness of adolescent environment will allow clinicians to frame questions and discussions that will target specific risks in the most meaningful way for each individual adolescent.
Work on this article was supported by the National Institute for Nursing Research Grant (NINR) 1K23NR011175. The authors also wish to thank Dr. Judith Bernstein for access to the Reaching Adolescents for Prevention data set as well as her expert research consultation.
Abma, J. C., Martinez, G. M., & Copen, C. E. (2010). Teenagers in the United States: Sexual activity, contraceptive use, and childbearing, National Survey of Family Growth 2006–2008. Vital Health Statistics
Barroso, C. S., Peters, R. J., Kelder, S., Conroy, J., Murray, N., & Orpinas, P. (2008). Youth exposure to community violence
: Association with aggression, victimization, and risk behaviors. Journal of Aggression, Maltreatment, and Trauma
, 141–155. doi: 10.1080/10926770802374916.
Berenson, A. B., Wiemann, C. M., & McCombs, S. (2001). Exposure to violence
and associated health-risk behaviors among adolescent girls. Achieves of Pediatric Adolescent Medicine
Bernstein, J., Heeren, T., Edward, E., Dorfman, D., Bliss, C., Winter, M., & Bernstein, E. (2010). A brief motivational interview in a pediatric emergency department, plus 10-day telephone follow-up, increases attempts to quit drinking among youth and young adults who screen positive for problematic drinking. Academic Emergency Medicine
, 890–902. doi: 10.1111/j.1553-2712.2010.00818.x.
Bossarte, R. M., Swahn, M. H., & Breiding, M. (2009). Racial, ethnic, and sex differences in the associations between violence
and self-reported health among US high school students. Journal of School Health
, 74–81. doi: 10.1111/j.1746-1561.2008.00379.x.
Boynton-Jarrett, R., Ryan, L. M., Berkman, L. F., & Wright, R. J. (2008). Cumulative violence
exposure and self-rated health: Longitudinal study of adolescents
in the United States. Pediatrics
, 961–970. doi: 10.1542/peds.2007-3063.
Brady, S. S. (2006). Lifetime community violence
exposure and health risk behavior among young adults in college. Journal of Adolescent Health
, 610–613. doi: 10.1016/j.jadohealth.2006.03.007.
Brady, S. S., Tschann, J. M., Pasch, L. A., Flores, E., & Ozer, E. J. (2008). Violence
involvement, substance use, and sexual activity among Mexican-American and European-American adolescents
. Journal of Adolescent Health
Centers for Disease Control and Prevention. (2010). Youth risk behavior surveillance United States, 2009. Mortality and Morbidity Weekly Report
(No. SS-5), 1–142. Retrieved from http://www.cdc.gov/mmwr/pdf/ss/ss5905.pdf
Grzywacz, J. & Fuqua, J. (2000). The social ecology of health: Leverages points and linkages. Behavioral Medicine
Institute of Medicine. (2000). Promoting health: Intervention strategies from social and behavioral research
. Washington, DC: National Academics Press.
Muula, A. S., Rudatsikira, E., & Siziya, S. (2008). Correlates of weapon carrying among high school students in the United States. Annals of General Psychiatry
, 8–12. doi: 10.1186/1744-859X-7-8.
Silverman, J. G., McCauley, H. L., Decker, M. R., Miller, E., Reed, E., & Raj, A. (2011). Coercive forms of sexual risk and associated violence
perpetrated by male partners of female adolescents
. Perspectives on Sexual and Reproductive Health
(1), 60–65. doi: 10.1363/4306011.
Walton, M. A., Resko, S., Whiteside, L., Chermack, S. T., Zimmerman, M., & Cunningham, R. M. (2011). Sexual risk behaviors among teens at an urban emergency department: Relationship with violent behaviors and substance use. Journal of Adolescent Health
, 303–305. doi: 10.1016/j.jadohealth.2010.07.005.
Whiteside, L. K., Walton, M. A., Stanley, R., Resko, S. M., Chermack, S. T., Zimmerman, M. A., & Cunningham, R.M. (2009). Dating aggression and risk behaviors among teenage girls seeking gynecologic care. Academic Emergency Medicine
, 632–638. doi: 10.1111/j.1553-2712.2009.00426.x.