In this issue of Perspectives on Sexual and Reproductive Health, Marín and colleagues examine sixth-, seventh- and eighth-grade characteristics as predictors of sexual activity in ninth grade among roughly 2,500 students in California (page 76). The study provides empirical evidence of the independent contribution of nonsexual romantic relationships in the seventh grade to the onset of sexual intercourse by the ninth grade for both males and females. In addition, it shows that among females, seventh graders in serious relationships with older teenagers—uniquely defined as those two or more years older—have an increased likelihood of sex in the ninth grade. Finally, the study demonstrates that seventh graders of both genders who have had serious romantic relationships were already significantly different in the sixth grade from those who have not: They had peers who were more accepting of sexual activity, they had experienced more unwanted sexual advances and situations that could lead to sex (i.e., where parental monitoring is limited) and, for females, they had undergone earlier menarche. The findings suggest important aspects of the pathways to early sexual intercourse that have not typically been addressed in either school- or parent-based sex education programs.

Current Sex Education Programs

School-based sex education programs differ dramatically in content, depending on whether they provide abstinence-only or comprehensive sex education. Abstinence-only programs emphasize the importance of postponing vaginal intercourse until marriage; the contraceptive information provided focuses primarily on the failure rates of various methods. Comprehensive sex education programs more often focus on the delay of sexual activity and protection for those who are sexually active, and include training in sexual negotiation and communication skills and information about obtaining contraceptive and reproductive health services. Because the vast majority of younger adolescents at the middle school level are not yet sexually active, both abstinence and comprehensive programs aimed at this group tend to focus on puberty, pregnancy and HIV information, and assertiveness and refusal skills.1 Although a few comprehensive sex education programs have been effective in encouraging teenagers to delay intercourse or use effective contraceptives if they are sexually active, only a handful of school-based programs have been able to demonstrate a significant impact on preventing pregnancy or STDs.2

Parent-based efforts are designed to help parents understand adolescent development, build communication skills, increase comfort in discussing sexual behaviors and encourage open communication with their teenagers.3 Encouraging parental involvement and improving parents' communication with their young adolescents seem to have universal appeal, even though such efforts have had mixed results. Limited studies have connected parent-child relationship and communication variables to improved sexual outcomes.4 However, evaluations of most parent-child programs focus on the communication between parent and child, not the sexual behaviors of the child, as the outcome. Although these evaluations have demonstrated improvements in communication frequency and parental comfort in discussing issues related to sex, none have demonstrated that these improvements have changed long-term sexual risk behaviors.5

In their current form, both youth and parent programs focus on vaginal intercourse almost exclusively.6 This is due, in part, to a primary emphasis on pregnancy prevention, and may reflect that only a few studies examine behaviors other than intercourse in the adolescent population.7 Other sexual behaviors, both ones that precede intercourse and ones that adolescents consider alternatives to intercourse, may be more characteristic of the early romantic relationships to which the Marín study refers.

Notably absent from virtually all types of programs is a clear and distinct discussion of sexuality within the context of love and intimacy, the risks of early romantic relationships and the risks implicit in relationships with older partners.8 With few exceptions (e.g., discussions of date rape and violence, and a recent push for abstinence-only programs to offer "preparation for marriage"), school-based curricula do not emphasize healthy and developmentally appropriate dating relationships. In providing evidence of the independent effect of early nonsexual relationships on subsequent early onset of sexual intercourse, Marín's findings suggest that such a focus could be one of the missing pieces in both youth and parent interventions. In particular, interventions that target young adolescents may be improved by the discussion of sexual behaviors that usually precede vaginal intercourse, as well as the impact of having boyfriends and girlfriends in middle school. Few parent programs specifically address parental values with respect to dating in the early grades, provide direct information about the relationship between early dating and later sexual behavior, or help parents recognize and reconcile the pressures for teenagers to date in order to be happy or popular among their peers; discussion of these issues could also improve the behavioral outcomes of parent education programs.

The Impact of Age Differences

One focus of the Marín study is the issue of age difference in the seventh-grade relationship; the researchers considered a two-year difference the criterion for same-age versus older boyfriends or girlfriends, hypothesizing that at the middle school level, even a small age difference could be important. (Other studies have generally used differences of at least three years between partners.9) Marín and her team found that the two-year age gap in the seventh-grade relationship predicts onset of intercourse by the ninth grade. This is critical because the smaller age difference may be less likely to raise the parental concerns that a much larger difference would. As a result, parents may not feel the need to monitor this type of relationship as carefully as they otherwise might or may neglect to discuss it with the young teenager. The Marín study finds that even these "more acceptable" nonsexual relationships have implications for later sexual behaviors. This is a message that young teenagers and their parents need to hear.

When relationships between younger and older partners involve sexual intercourse, it is commonly assumed that these are between partners with large age differences and, consequently, large differences in power. However, in the majority of cases, sex between younger teenagers and older individuals is reported to be voluntary.10 This finding may, in part, reflect that the overwhelming majority of these encounters (86%) are between younger teenagers and older teenagers, and not between young teenagers and adults, as is often assumed.

Relationships between young teenagers and older teenagers are generally not seen as problematic by teenagers themselves. One in three 12–14-year-olds agree that it is "okay for teenagers aged 14 or younger to date someone who is three or more years older"; this perception, however, diminishes as adolescents get older.11 Thus, the youngest teenagers are the least able to project or consider the risks of early relationships with older teenagers. If three-year differences are seen as appropriate and if most sex between young people three or more years apart is not forced, then the sexual behavior that is occurring between partners with a smaller age gap is even more likely to be consensual. The authors also note that in certain ethnic groups—particularly Hispanics, who made up a large portion of the study sample—relationships between younger females and older males may not be outside the norm.

Implications for Educational Programs

Parent-based approaches could be an effective strategy in the repertoire of programs to delay sexual intercourse and reduce teenage pregnancy and STDs. However, simply encouraging parents to talk with their children about sexual risk-taking without providing specific guidance about the dynamics of adolescent sexuality is probably of limited value.12 We know that it is critical to provide adolescents with consistent messages;13 however, they are often exposed to multiple and conflicting messages, including those from parents who admonish them about sexual involvement while being proud of and encouraging these early relationships. As demonstrated by the Marín study, parents need to be aware that these early romantic relationships carry their own risks, even when they are not yet sexual in nature. Alerting parents to the relationship between early nonsexual relationships and subsequent early initiation of sexual intercourse, the impact of age differences between partners, the importance of supervision and monitoring even in seemingly "innocent" situations and the effect of early menarche are good examples of the specific guidance that they need.

In demonstrating the influence of Hispanic background as a predictor of ninth-grade sexual activity, Marín adds to the body of literature on the relationship between culture and teenage sexual behavior. Interaction between ethnicity and parenting styles affects teenage sexual behavior,14 suggesting that enhancing parent education programs may be particularly valuable for certain ethnic groups. Further exploration of specific cultural beliefs and norms with respect to early dating and age differences between partners is warranted. Meanwhile, programs that engage parents in discussion of age-appropriate behaviors, and that provide specific information about the implications of certain practices, may assist Hispanic parents during these discussions with their young teenagers.

Can school-based prevention programs mitigate the relationship between early romantic relationships and early sexual activity? Notably, this study took place within the context of an intensive middle school intervention to reduce sexual risk behaviors.15 Although the authors state that the analyses were not designed to assess the intervention's impact, program participation was not a significant predictor of ninth-grade sexual activity. It is important to note that this intervention, like most, did not specifically address dating relationship issues.

The article makes a strong case for a simple and powerful message: For those young people having sex by the ninth grade, the seeds of sexual risk-taking are sown early in adolescence. Furthermore, early romantic relationships, opportunities for sex and, for females, early puberty significantly increase the likelihood of sexual intercourse by the ninth grade. Directly addressing these issues in youth and parent interventions has the potential to reduce the prevalence of early romantic relationships and provide some protection from early onset of sexual intercourse, even among youth who are already involved in such relationships.