Patterns of Condom Use over Time in Sexual Partnerships

Sexual and contraceptive behaviors differ at different stages of a relationship with a sexual partner.  For example, condoms are typically used early in a relationship and with more casual sexual partners, whereas hormonal methods become more common later in relationships. However, most studies of this topic rely on retrospective reporting on changes in condom use.  Thus, little is known about the precise timing of contraceptive-switching behaviors, or about processes related to changes in condom use over time in a sexual partnership.  We examine these issues using intensively collected weekly diary data and time-varying effect models.    

Method. Data are from a longitudinal study of women who were recruited from public records when they were 18-20 years old and provided weekly reports of their sexual behaviors and romantic relationships for up to 2.5 years (N=1,003, 35% African American, Mage=18.7).  We focused on sexual partnerships that began after the start of the study.  Outcomes include occurrence of sexual intercourse and condom use, and these behaviors were predicted by sexual and condom attitudes; time was measured by weeks from the start of a specific partnership (i.e. time from event).

Results. Preliminary results show that odds of sexual intercourse increase sharply over the first 20 weeks of a relationship; after this time, the prevalence of intercourse held steady, with about 65% of women having sex with their partner each week.  Among weeks where sexual activity occurred, rates of condom use declined steadily over time.  More than half of women used condoms at the start of a partnership, whereas only about one-quarter used condoms after two years in a partnership.  Baseline attitudes predicted behaviors, but only early in a relationship.  For example, individuals who perceived more barriers toward condom use were less likely to use condoms for the first ten weeks in a partnership, but after about 5 weeks there was no association between perceived barriers and condom use.  Additional analyses will examine relationship predictors of condom use over time.  

Conclusions. Results demonstrate how TVEM can be used to examine processes that occur over time in relationship to an event, such as the start of a relationship with a partner.  Consistent with research suggesting that condom use is seen as indicating a lack of closeness and trust with a partner, condom use decreased over time in a partnership.  Participants’ prior attitudes only predicted condom use for a short period of time early in a partnership.  Results show particular periods which may be particularly suited for interventions, as well as the necessity of including relationship factors in sexual health interventions.