The publisher's final edited version of this article is available at Sex Transm Dis See other articles in PMC that cite the published article. Abstract Background Among adolescents and young adults, the extent that partner characteristics account for sexually transmitted diseases STDs in context of individual sexual activities and demographic characteristics is unclear. We used exact logistic regression to calculate odds ratios OR for several sexual partner characteristics age discordance, incarceration, STD diagnosis, other partners, alcohol problem, marijuana problem, and a calculated composite variable adjusting for demographics and individual sexual activities, including condom use.
In the United States, 9 million cases of sexually transmitted diseases STDs occurred among 15 to year-olds in The relative importance of partner characteristics and individual sexual activities e. Among urban STD clinic attendees ages 15 to 24 years, we pursued 2 objectives to improve understanding of the influence of partner characteristics on STD risk. First, we assessed the association between STD diagnosis and a composite measure of partner characteristics incorporating age discordance, incarceration, STD diagnosis, other partners, alcohol problem, and marijuana problem.
Second, we compared the association between STDs and partner characteristics to the association between STDs and individual sexual activities. STDs were assessed clinically and an interviewer administered a questionnaire about demographic variables, alcohol use, sexual activities, and sexual partner characteristics. The present analysis includes the black and white participants who reported having heterosexual sex.
Institutional Review Board approval was obtained for the study procedures from the University of Pittsburgh and for this analysis from the University of Florida. Sexual Partner Characteristics Based on the literature about partners and sexual risk, 3 — 5 , 9 , 12 — 14 we selected 6 measures of partner characteristics: Participants reported each characteristic for both their main partner and their most recent, not main partner.
Responses for each partner were considered equally. Using a predetermined method, we created a simple composite partner characteristic variable that could be incorporated into clinical practice and maximized the existing data. For each of the 6 partner characteristics described above, we assigned a value of 0 to the referent category and 1 to the risk category.
For each participant, we calculated the proportion of the responses in the risk category. This yielded a score ranging from 0 to 1 for each participant; 0 indicated all characteristics were in the referent category and 1 indicated all characteristics in the high risk category. We divided this score into 3 broad groups: For sensitivity analysis, we also created the composite without the measure of partner had a STD.
Individual Sexual Activities Based on the literature of individual sexual risk factors, 9 , 12 , 15 — 19 we selected 9 measures of individual sexual activities: Age at first intercourse, number of lifetime sex partners, and number of sex partners in the past year were collected as continuous variables and categorized by the overall median response for analysis.
We also created a composite of individual sexual activities. Because we detected moderate correlation Pearson correlation coefficient between 0.
We created the individual sexual activities composite variable with the 8 other individual sexual activities measures using the same methodology as we used to create the partner characteristic composite. In brief, we created a score from 0 to 1 by calculating the proportion of responses of increased risk among the individual sexual activities variables. Syphilis was tested among serum samples for both sexes.
Viral culture of genital herpes was performed for suspicious lesions and genital warts were diagnosed by clinical observation. Similar to our previous analysis from this study, 12 an individual diagnosed with any of the above diseases was defined as having a confirmed STD. Type of medical insurance was analyzed as any insurance Medicaid or private insurance versus no insurance.
For each variable, we used both race-adjusted and multivariate covariates included race, sex, age, and type of medical insurance exact logistic regression to estimate the association between partner characteristics and STD diagnosis.
Marital status was not included in the multivariate model because there was little variability in this measure: Common to models with several covariates, 21 , 22 we found that the exact permutation distribution for the sufficient statistic was computationally infeasible to calculate for our multivariate models; therefore, for multivariate models we used the LogXact network-based Monte Carlo sampling approach for conditional logistic regression to estimate unbiased exact confidence intervals.
All of the participants were sexually experienced. The median number of lifetime sex partners was 10 range: We found differences between women and men on reporting of partner characteristics and individual sexual activities Table 1. Women were more likely than men to report partners with risk characteristics including discordant ages, previously in jail, and had alcohol, or marijuana problems.
Men were more likely than women to report individual sexual activities with increased risk, except for condom use and sex under the influence of alcohol that were reported equally by men and women.