Abstract: Using Tvem to Examine Time-from-Event Processes: The Example of Sexual Behavior over Time in a Relationship (Society for Prevention Research 23rd Annual Meeting)

250 Using Tvem to Examine Time-from-Event Processes: The Example of Sexual Behavior over Time in a Relationship

Schedule:
Thursday, May 28, 2015
Regency D (Hyatt Regency Washington)
* noted as presenting author
Sara Vasilenko, PhD, Research Associate, The Pennsylvania State University, State College, PA
Introduction. Time-varying effect modeling (TVEM) is well-suited to examining how processes involved in health-risk behaviors may differ over time from an observed event or life transition. For example, sexual behaviors change over the course of a relationship with a romantic or sexual partner, but little work has examined the particular predictors or processes that play a differential role at different times in a relationship. This study uses intensively collected weekly journal data analyzed using TVEM to examine how attitudes, intentions, and relationship factors differentially predict sexual behavior in a given week during the first year of a relationship with a new sexual partner.

Method. Data are from 597 young women (60% white, 34% black; M age = 19.9 at the start of first new partnership, SD = .9) who reported at least one new sexual partnership (N = 1.9 partnerships) during a 2.5 year longitudinal study. Participants completed baseline measures of demographics, attitudes, intentions, and peer norms about sexual behavior and contraceptive use, followed by up to 130 weeks of intensive weekly journals about their past-week relationships, sexual and contraceptive behaviors.

Results. About 35% of relationships included sexual intercourse in the first week of a partnership; this percentage rose steadily until 30 weeks, where it plateaued around 60%. During weeks of sexual activity, about 65% used a condom in the first week, with only about 30% using a condom at 50 weeks. Stronger intentions to have intercourse were associated with engaging in intercourse at all points in the first year of a relationship; however, intentions to use condoms only predicted condom use for a short period during the first seven weeks of a partnership.

Conclusion. This study suggests heterogeneity in young women’s sexual behavior, in that many young women engage in sexual intercourse early in a relationship, but many still do not engage in weekly intercourse after a year in a relationship. Women’s intentions are more consistent predictors of couples’ sexual behavior compared to condom use behaviors. Findings suggest that when applied to sexual behaviors, models of health behaviors need to incorporate partner intentions and relationship stage. TVEM can help uncover when to intervene and what factors to target at different stages of a relationship, providing information that can be useful in creating adaptive interventions for sexual risk behaviors.