Contraceptive behavior by an individual varies both across and within relationships, and the nature and characteristics of relationships can influence use and the type of method used. This study utilizes the retrospective sexual relationship histories of young adults available in the most recent wave (2001-2002) of the National Longitudinal Study of Adolescent Health (Add Health) to explore relationship-specific contraceptive method choice. Hierarchical generalized linear modeling is used to estimate the effects of both relationship-level and individual-level characteristics on the type of contraceptive method used at last sex. Four mutually exclusive categories for method use were constructed: condom only, hormonal method only, dual method (condom plus hormonal method), and no method. A number of relationship characteristics are significantly associated with contraceptive use even when controlling for individual characteristics. Moreover, the effects of relationship characteristics differ depending on the type of method examined. Significant between-individual variation in the type of contraceptive method used remains. Not only are the characteristics of the individuals forming the relationships of importance in determining contraceptive practices, but so too are the unique features of the relationship itself. Further investigation of other aspects of the relational context is warranted.

The relational patterns and behaviors that are experienced during adolescence and young adulthood are influential for reproductive health outcomes and set the stage for future family formation choices and behaviors. In order to better understand the extent to which the relational context influences behavior within relationships, it is imperative that researchers first explore these relationships in a more comprehensive manner than has been done previously. Using the retrospective sexual relationship histories of young adults from the most recent wave (2001-2002) of the National Longitudinal Study of Adolescent Health (Add Health), this paper provides a detailed description of young adults’ nonmarital sexual relationships and explores young adults’ relationship histories. Findings suggest that the characteristics of the nonmarital sexual unions that are formed during the early life course are quite diverse. The type of relationship (defined as cohabiting, exclusively dating, frequently but not exclusively dating, dating once in a while, and having sex only) and other features of these sexual relationships (e.g., duration, age difference between partners) differ depending on the gender, race/ethnicity, and age of the young adult. Additionally, some patterns of associations across racial/ethnic groups and age categories differ for women and men. Moreover, the features of these sexual relationships vary across the different relationship types. Finally, a number of individual-level sociodemographic characteristics are associated with having experienced certain types of relationship histories. In particular, there are enduring effects of family background and religious denomination during adolescence that differ for women and men.

Context: Adolescents have among the highest rates of STD, but until recently it has been difficult to characterize the multiple social and behavioral factors that affect STD risk because of measurement and methodological issues.

Methods: Data from Wave I of the National Longitudinal Study of Adolescent Health (Add Health) are used to estimate effects of school, neighborhood, family, and individual characteristics on acquiring an STD. For sexually active teens, we also estimate the extent to which age at first intercourse affects STD acquisition. Additionally, we investigate the determinants of STD occurrence between Waves I and II of Add Health, and model the effects of prior STD acquisition. Random intercept logistic regression and random intercept piecewise exponential hazard regression are used to account for possible clustering in the Add Health data.

Results: Overall, 7.0 percent of sexually active teens reported ever having an STD as of Wave I, and 6.7 percent reported having an STD between Waves I and II. Among all Wave I teens—uninitiated as well as sexually active—age, gender, race/ethnicity, family background characteristics, neighborhood and school characteristics affect STD acquisition. Among sexually active Wave I teens, those with a younger age at first intercourse are at greater risk of experiencing an STD. Other factors contribute, but to a lesser degree. For example, family structure becomes nonpredictive. For acquisition of an STD between Waves I and II, females, blacks, teens with lower levels of mother’s education, and those who had an STD previously are at higher risk.

Conclusions: The findings support our hypothesis that multiple social and behavioral factors influence lifetime history of STD. Also, age at first intercourse and STD history are valid indicators of risk-related behavior.

Young adults experience the highest rates of sexually transmitted diseases (STDs) in the United States and gender and racial/ethnic differentials in STD prevalence persist. Rich data are available to investigate an array of sociodemographic and behavioral risk factors that might explain variation in STD infection. These data also include biomarker test results enabling an investigation of current infection among a population-based sample of youth. We use data from Wave III of the National Longitudinal Study of Adolescent Health (Add Health) to examine sociodemographic and behavioral risk correlates of STDs, using nested binary logistic regression techniques. We consider whether respondents who refused or were unable to provide urine samples differ from those who provided urine samples. We also consider whether respondents whose specimens were not used, or whose specimens were used but for whom no definitive test result was obtained, differ from individuals for whom definitive test results were obtained. Finally, we explore sociodemographic and behavioral correlates of positive test results for those for whom definitive test results were obtained. All analyses are weighted and account for the complex study design of Add Health. Current STD infection is associated with a number of sociodemographic characteristics, including age, gender, and race/ethnicity, the effects of which differ depending on the specific pathogen. The effects of the sociodemographic characteristics remain even after accounting for differences in sexual risk behaviors. Future research should explore the contribution of partner characteristics and behaviors as well as sexual networks in explaining variation in STD infection.