The impact of incarceration on social norms contributing to heterosexually acquired HIV infection among adults in Baltimore City, Maryland

Background. Significant racial and ethnic disparities in HIV and sexually transmitted diseases (STD) exist in the United States. Many studies cite high incarceration rates and low sex ratios as a cause of racial disparities in HIV/STD rates, but there is little empirical evidence to support this hypothesis. This study examines the role of social norms and perceptions in linking incarceration and sexual risk behaviors among African Americans in Baltimore.;Objectives. This study aims to: (1) develop a scale to measure social norms about engaging in concurrent sexual partnerships; (2) determine the proportion of heterosexual partnerships that are spatially and racially assortative; and (3) assess whether perceptions about partnerships (perception of partner availability, gender-based power in relationships, and norms about concurrent partnerships) mediate the relationship between incarceration and concurrency behavior.;Methods. Male and female adults were recruited from high-poverty, high-HIV prevalence areas in Baltimore as part of a national cross-sectional study of heterosexuals at high risk of HIV infection. A standardized questionnaire collected information on sexual activity, drug use, incarceration status, and other behavioral and partner characteristics. We measured the reliability of the new scale to measure social norms about concurrency. Racial, spatial, and age assortativity were calculated for sexual partnerships. We examined associations between concurrency behavior and concurrency norms, spatial and racial assortativity, and incarceration rates using logit regressions with generalized estimating equations to adjust for clustered outcomes.;Results. A total of 332 individuals were enrolled in the study. Concurrency was common among men and women. Respondents perceived male concurrency behavior as more normative than female concurrency behavior. Scale scores were significantly positively associated with concurrency among women. Almost half of partnerships were spatially assortative. Participants who lived in high HIV-prevalence areas were more likely to choose spatially assortativc partners than residents of lower prevalence areas after adjusting for partnership type, gender, and number of partners. Neighborhood incarceration rates were significantly associated with concurrency behavior among males, but not females. Sex ratios, concurrency norms, and attitudes about gender-based power did not mediate the relationship between incarceration rate and concurrency among men.;Conclusions. This study was a quick, multi-level, and cost-efficient examination of factors affecting high-risk patterns of sexual partnering. The data were not consistent with the proposed mechanism for the relationship between incarceration rates and concurrency. Alternative theories to explain this relationship should be explored.