Understanding the factors related to sexual behaviors in the adolescent population is important as early initiation of sexual intercourse and other risky sexual behaviors may result in unintended pregnancy, human immunodeficiency virus (HIV) infection, and other sexually transmitted infections (STIs) (Coker at al., 1994; Kaestle, Halpern, Miller, & Ford, 2005; McNeely & Blanchard, 2009). Sexually active adolescents aged 15–19 years and young adults aged 20–24 years are at higher risk than older adults of acquiring sexually transmitted diseases (Centers for Disease Control and Prevention [CDC], 2014a). Untreated STIs can result in negative heath, quality of life, and psychosocial consequences (Gottlieb et al., 2014). Unintended pregnancy can also have negative consequences for both mothers and children. Teens who become pregnant are less likely to complete high school or college and more likely to live in poverty than teens who do not get pregnant (Amato & Maynard, 2007; Hofferth, Reid, & Mott, 2001). Children born to teen mothers have poorer lifetime educational, behavioral, and health outcomes than do children born to older mothers (Office of Adolescent Health, 2016). Though teen pregnancy rates in the United States are on the decline (Curtin, Abma, Ventura, & Henshaw, 2013), the U.S. teen birth rate (births per 1,000 women aged 15–19) is still higher than all other industrialized countries (Blum, 2013). Risk and protective factors for STIs, HIV, and unintended pregnancy among adolescents occur at all levels of the social ecology, and this evidence suggests that a comprehensive, socioecological approach to prevention may be the most effective (DiClemente, Salazar, Crosby, & Rosenthal, 2005).

Previous literature has found significant associations between TDVV and several sexual risk behaviors, such as ever having sexual intercourse, anal intercourse, earlier age at first intercourse, current sexual activity, having multiple partners, using alcohol or drugs before sexual intercourse, and lack of condom use (Black, Noonan, Legg, Eaton, & Breiding, 2006; Decker et al., 2014a; Eaton, Davis, Barrios, Brener, & Noonan, 2007; Himelein, Vogel, & Wachowiak, 1994; Howard & Wang, 2003; Lormand et al., 2013; Roberts, Auinger, & Lein, 2005; Silverman, Raj, & Clements, 2004; Silverman et al., 2001; Vagi, Olsen, Basile, & Vivolo-Kantor, 2015). For example, Vagi and colleagues (2015) used data from the 2013 YRBS and categorized students by combined physical and sexual TDVV. The authors found that male students who experienced sexual TDVV only or both physical and sexual TDVV were more likely than students with no TDVV experience to have had sexual intercourse with four or more partners and be currently sexually active (Vagi et al., 2015).

A study of Massachusetts YRBS data examined the association between TDVV and sexual risk behaviors among female students (Silverman et al., 2001). Adjusted analyses demonstrated that students experiencing only sexual (and not physical) TDVV were more likely to have had sexual intercourse before age 15 (1997 and 1999 data) and to have had sexual intercourse with at least three partners (1999 data). Students experiencing both physical and sexual TDVV were more likely to have had sexual intercourse before age 15 and had sexual intercourse with at least three partners (1997 and 1999 data; Silverman et al., 2001). Unadjusted analyses also showed some associations between sexual TDVV and substance use before last intercourse and no condom use during last sexual intercourse (Silverman et al., 2001).

In the wider TDV and sexual violence literature, multiple factors have been proposed to mediate the relationship between violence and sexual risk behaviors. While a full discussion of these factors is beyond the scope of this paper, we mention several here. As noted above, using alcohol before sexual intercourse is a risk factor for TDVV (Silverman et al., 2001). Alcohol use has also been linked to earlier age of onset of sexual activity (Doran & Waldron, 2017) and at least one study found it to be one of the main reasons for unwanted sexual activity reported by both male and female high school students (Jackson, Cram, & Seymour, 2000). Social norms, including peer attitudes and behaviors that are supportive of sexual violence, have also been found to be risk factors for sexual violence perpetration among adolescent and young men (Jewkes et al., 2006). Because peer relationships are particularly influential in adolescence and sexual violence has been reported to take place at parties where many adolescents may gather (Jackson et al., 2000), alcohol use before sex and peer attitudes that support TDV may play an important role in the relationship between TDVV and sexual risk behaviors, including lack of condom use.

TDVV studies that have examined associations with health-risk behaviors have generally reported results for physical or overall TDVV and have not focused on sexual TDVV. Studies on sexual violence, in general, have not separated out such violence that occurs in a dating relationship. In addition, little published literature includes measures of TDVV frequency. By examining the frequency of TDVV, we can get a better sense of the burden of TDVV and the potential consequences of the varying levels of victimization, which could have important implications for screening and intervention programs. In 2015, Vagi and colleagues conducted the first study using the revised physical TDVV item and the new sexual TDVV item in the 2013 YRBS. The purpose of this paper was to report the prevalence and frequency of TDVV among male and female students and also to consider associations with the revised/new items and selected health-risk behaviors. The authors included only two sexual risk behaviors in their analysis (having four or more sexual partners and current sexual activity) and categorized TDV as physical TDVV only, sexual TDVV only, both physical and sexual TDVV, or any of the two TDVV forms, but did not consider the association between frequency of these behaviors and health outcomes. One of the stated next steps from this study was to examine how the frequency of TDVV behaviors may impact negative health outcomes.

Thus, this study builds on the work from Vagi and colleagues (2015) in several ways. First, we examine the association between sexual TDVV frequency and the sexual risk behaviors, which is the first study to do so with the new sexual TDVV item in YRBS. Second, we examine a broader range of sexual risk behaviors. Finally, we also include more recent data (the 2015 YRBS administration).

We hypothesize that both the dichotomized sexual TDVV variable as well as the frequency of sexual TDVV will be associated with the selected sexual risk behaviors. However, we also hypothesize that it’s not only the exposure (yes/no) to sexual TDVV victimization that increases risk, but that those students who experience more acts of sexual TDVV in dating relationships will be at higher risk for engaging in sexual risk behaviors than those experiencing fewer acts of sexual TDVV.

Demographics and prevalence of study variables in the overall analytic sample are described in Table 1. The sex and grade distributions of the analytic sample were approximately even. The sample consisted of 55.5% white students, 14.7% black students, and 21.5% Hispanic students. Among students who had dated or gone out with someone during the 12 months before the survey, 4.2% reported sexual TDVV one time, 3.3% reported sexual TDVV two or three times, 1.0% reported sexual TDVV four or five times, and 2.0% reported sexual TDVV six or more times.

The prevalence of sexual TDVV in the overall analytic sample and by student demographics is presented in Table 2. Overall, 10.5% of students who dated during the past 12 months reported sexual TDVV. A significant difference in sexual TDVV prevalence was observed by sex and race/ethnicity. Female students were more likely than male students to have experienced sexual TDVV (15.0% vs. 5.8%). While the overall Chi-square test showed a significant difference in sexual TDVV by race/ethnicity (p=0.01), pairwise testing results indicated no difference between white, black, and Hispanic students. No significant difference was observed by grade.

The associations between any sexual TDVV and sexual risk behaviors were stratified by sex and are shown in Table 3. Among male students, those who experienced any sexual TDVV were more likely than those not experiencing any sexual TDVV to have had sexual intercourse before age 13 (PR: 2.93; 95% CI: 2.42, 3.53), have 4 or more lifetime sexual partners (2.12; 1.84, 2.46), be currently sexually active (1.55; 1.40, 1.73), use alcohol or drugs before last sexual intercourse (2.16; 1.78, 2.62), not use a condom during last sexual intercourse (1.35; 1.14, 1.60), and use no method of pregnancy prevention during last sexual intercourse (1.83; 1.34, 2.49). These students also were less likely to have not been tested for HIV (0.85; 0.79, 0.91). The same significant results were observed for female students who experienced any sexual TDVV with one exception: no significant association was identified for no condom use during last sexual intercourse (1.09; 0.99, 1.21). Overall, among students who had dated or gone out with anyone during the past 12 months, those who experienced any sexual TDVV were more likely to engage in sexual risk behaviors, but more likely to have been HIV tested.

Table 4 presents the associations between sexual TDVV frequency and sexual risk behaviors by sex. No clear dose-response association was observed for either sex. However, the strongest associations were generally observed among students reporting sexual TDVV four or five times or six or more times. For students experiencing sexual TDVV six or more times, significant associations were observed for sexual intercourse before age 13 (male: 4.58; 3.49, 6.02 and female: 6.35; 4.29, 9.41), having four or more lifetime sexual partners (male: 3.19; 2.63, 3.86 and female: 2.67; 2.01, 3.55), being currently sexually active (male: 1.93; 1.71, 2.19 and female: 1.42; 1.19, 1.69), using alcohol or drugs before last sexual intercourse (male: 2.49; 2.01, 3.08 and female: 2.15; 1.53, 3.03), not using a condom during last sexual intercourse (male: 1.58; 1.33, 1.88 and female: 1.26; 1.02, 1.55), using no method of pregnancy prevention during last sexual intercourse (male: 2.31; 1.53, 3.48 and female: 2.36; 1.65, 3.36), and not having been tested for HIV (male: 0.80; 0.71, 0.91 and female: 0.86; 0.75, 0.98). Among both male and female students who had dated or gone out with anyone during the past 12 months, only those experiencing sexual TDVV six or more times were significantly more likely to engage in all sexual risk behaviors, but more likely to have been HIV tested.

The purpose of this study was to assess the relationship between sexual TDVV and sexual risk behaviors among high school students in the United States using data from a nationally representative, cross-sectional, school-based survey. In models adjusting for race/ethnicity and grade, this study found consistent associations between any sexual TDVV and sexual risk behaviors by sex with one exception. Among female students, lack of condom use during last intercourse was not associated with sexual TDVV. This was a surprising finding as previous literature has indicated that being in an abusive relationship is associated with coercive condom dynamics, fear of condom negotiation, and unprotected sex (Decker et al., 2014a; Wingood, DiClemente, McCree, Harrington, & Davies, 2001; Wu, El-Bassel, Witte, Gilbert, & Chang, 2003). Though largely similar by sex, the sexual TDVV frequency findings showed only one sexual risk behavior (alcohol/drug use before last sexual intercourse) was significantly associated with a sexual TDVV frequency of one time for female students, while the association was significant for five sexual risk behaviors (sexual intercourse before age 13, four or more lifetime partners, current sexual activity, alcohol/drug use before last sexual intercourse, and no HIV testing) among male students. The results suggest that though sexual TDVV is more common among female students, it is an important factor associated with sexual risk behaviors among males. Therefore, it is important for any prevention and intervention programs to include a focus on male students.

There are multiple contributions of this particular study’s findings to the current literature base. First, other than Vagi and colleagues (2015), this is the only study with a nationally representative sample to assess the relationship between sexual TDVV and several sexual risk behaviors. Vagi et al. (2015) also found significant associations between sexual TDVV (either alone or with physical TDVV) and the two sexual risk behaviors they included, with some differences by student sex. Second, our study includes a more detailed measure of TDVV than Vagi and colleagues (2015). The national YRBS questionnaire did not include an item on sexual TDVV until 2013. Further, the existing physical TDVV question did not include the response option “I did not date or go out with anyone during the past 12 months;” this was included for both TDVV questions when the sexual TDVV question was added to the questionnaire in 2013. The ability to remove students who had not dated adds precision to the TDVV measurement and also allows for a more accurate estimate of the relationship between TDVV and sexual risk behaviors.

In addition, starting with the 2013 YRBS questionnaire, both the physical and sexual TDVV questions included responses to assess the specific number of times that TDVV occurred. Previously, the TDVV question only had a “yes” and “no” response option. To our knowledge, this is the first analysis that examined the association between sexual TDVV frequency and a range of sexual risk behaviors; this analysis was not included in the Vagi et al. (2015) study. The use of frequency response options allows for the detection of meaningful differences and emphasizes a sense of burden, particularly for those youth who experience violence more frequently. Specifically in our study, the strongest associations were generally observed among students of both sexes reporting sexual TDVV at least four times in the past 12 months. In particular, both male and female students reported alcohol or drug use before last intercourse and males reported not using any pregnancy protection, though the association was significant only for females who reported sexual TDVV more than six times in the past 12 months. In addition, the association with no condom use was significant only for both males and females who reported sexual TDVV more than six times in the past 12 months. Thus, students who experience TDVV more frequently may represent a higher-risk group for exposure to other sexual risk behaviors. Though our methodology and analyses cannot answer questions about directionality of the relationship, it is possible that students engaging in these risky sexual behaviors, such as having multiple sex partners, are putting themselves at risk for repeated victimization. However, further research on this topic including studies that have more students in various frequency categories and additional studies that use population-based samples, are needed to replicate these findings.

Our results highlight the importance of addressing TDVV and sexual risk behaviors using a comprehensive, cross-cutting approach that promotes healthy dating relationships. Several evidence-based and promising school- and family-based prevention strategies exist that address both TDV perpetration and victimization as well as other risk behaviors including sexual risk behaviors. Healthy relationship interventions implemented in schools, such as Safe Dates (Foshee et al., 1998) or Dating Matters^® (Tharp et al., 2011; Tharp, 2012), focus on social-emotional learning components including clear communication, negotiating conflict, emotion regulation, and healthy coping. Though these programs have been implemented primarily to decrease risk for TDV, the content included in these programs has the potential to also impact sexual risk behaviors, thus allowing communities to prevent multiple risk behaviors with a single program. Comprehensive school-based programs may also be particularly important avenues for prevention given the role that social norms and peer attitudes toward violence play in TDV perpetration and sexual risk behaviors.

Schools also can provide support for victims through early identification, counseling services, and referrals to outside support services. Healthcare providers and other health professionals, within and outside the school environment, can support the identification and management of adolescent violence issues (Cutter-Wilson & Richmond, 2011; Duke & Borowsky, 2014; Moore, Sargenton, Ferranti, & Gonzalez-Guarda, 2015). This may include the development of protocols for handling TDV cases (Cutter-Wilson & Richmond, 2011; Duke & Borowsky, 2014), but it also may be particularly important to conduct screenings among sexually active adolescents. Though a limited amount of evidence-based programs exist for TDV, the field stands to benefit from lessons learned in the prevention of HIV/STI and teen pregnancy given the associations between TDVV and sexual risk behaviors in this study.

This study has at least three limitations. First, YRBS data are self-reported and students may under- or over-report behaviors. However, a psychometric study has shown that the national YRBS questions have good test-retest reliability (Brener et al., 2002). Second, the data are cross-sectional; therefore, causality and directionality cannot be concluded. We cannot conclude that sexual TDVV causes students to take on more sexual risks. Similar factors not measured by the YRBS may place students at risk for both TDVV and sexual risk behaviors. Third, data are only collected on adolescents who attend school and, therefore, are not representative of all persons in this age group. Nationwide, in 2012, of persons aged 16–17 years, approximately 3% were not enrolled in a high school program and had not completed high school (Stark & Noel, 2015).

Sexual TDVV is significantly associated with sexual risk behaviors among female and male students. Therefore, it is appropriate and valuable for counseling, psychological, and social services addressing these types of violence to target both sexes. Additionally, HIV, STI, and pregnancy prevention efforts may consider incorporating messages and tools to address violence prevention and skills for developing healthy relationships. The prevalence of TDV can be even more prevalent in the college environment as the most common age of first intimate partner violence victimization is 18–24 years of age (Breiding et al., 2014); therefore, research that will further understanding of these behaviors and help prevent them in adolescence and young adulthood is important.