Sexual Behavior and Risk of Sexually Transmitted Diseases: Do Community Characteristics Moderate the Relationship between Individual Behaviors and STD Risk?

Abstract

This study examined individual demographic and behavioral characteristics associated with the risk of infection with two bacterial sexually transmitted diseases (STDs), Neisseria gonorrhoeae and Chlamydia trachomatis. Unlike other studies of STDs, this study also investigated how the neighborhood composition may influence individual behaviors and STD risk. Individual-level data were derived from a probability sample of 560 adults residing in households in Baltimore, Maryland in 1997-1998. The Baltimore STD and Behavior Study (BSBS) combined use of urine-based diagnostic tests for gonorrhea (GC) and chlamydia! (CT) infection with a population-based survey of health and STD-related risk behaviors. Neighborhood-level data on the geographic and social characteristics of Baltimore's neighborhoods were obtained from 1990 Census data and the Baltimore City Department of Planning (1993). Urine assays revealed that 8.3 percent of adults ages 18-35 were currently infected with gonorrhea and/or chlamydia. A self-reported history of GC/CT infection was assessed from BSBS interview data; 26.1 percent reported previous treatment for GC/CT. As predicted by the STD literature, bivariate and logistic regression analyses suggested that self-reported infections were significantly associated with individual and behavioral characteristics. However, the lack of STD-related risk behaviors among respondents currently infected was somewhat unexpected. Compared with uninfected participants, respondents with a current infection, for example, were less likely to report multiple sex partners, new partners, paid sex, or concurrent sexual relationships. Hierarchical logistic regression models indicated only the proportion of black residents within the neighborhood was positively and significantly associated with self-reported GC/CT infection after controlling for individual- and community-level characteristics. However, multilevel analyses did not detect an association between neighborhood characteristics and current infection. Neighborhood characteristics did not seem to matter when examining differences in the distribution of current infection in Baltimore. Unlike individuals with a self-reported infection, currently infected individuals were not more likely to report high-risk sexual behaviors or STD-related symptoms. One plausible explanation is that the partners of these 'low-risk' individuals may have engaged in 'high-risk' behaviors. These data urge further exploration of the social context of gonorrhea and chlamydiaI infection in conjunction with an investigation of the interactions between individuals and their sexual partners.