HIV seroincidence and correlates of seroconversion in a cohort of male factory workers in Harare, Zimbabwe

Objectives: Given that health promotion messages on transmission of HIV and other sexually transmitted diseases (STD) have been widely publicized in Zimbabwe and elsewhere in the late 1980s, it is vital to analyse which risk events still expose individuals to infection. A cohort was established with the objectives of estimating HIV seroincidence, behavioural and biological determinants of infection, and ultimately, evaluating the impact of AIDS prevention interventions in the workplace. Methods: HIV seroincidence was estimated in a prospectively followed cohort of male factory workers recruited in Harare, Zimbabwe during the period prior to a workplace AIDS prevention intervention. Correlates of HIV seroconversion were identified using Cox regression analysis. Results: There were 51 seroconversions among 1607 HIV-negative participants following 1738 person-years (PY) of observation [seroincidence, 2.93 per 100 PY; 95% confidence interval (CI), 2.18-3.86]. The prevalence of HIV in the cohort was 19.1%. HIV seroincidence was significantly increased among men who were younger [hazard ratio (HR) per year, 0.96; 95% CI, 0.93-0.99], were single (HR, 3.29; 95% CI, 1.56-6.96), were married but resided separately from their wives (HR, 2.18; 95% CI, 0.99-4.80), reported having any STD (HR, 3.00; 95% CI, 1.53-5.86), reported having a genital ulcer (HR, 4.87; 95% CI, 2.18-10.91), and reported paying for sex (HR, 2.01; 95% CI, 1.06-3.77). Seroincidence also increased with the number of sex partners reported in the year preceding enrolment (HR per partner, 1.10; 95% CI, 1.01-1.21). In multiple Cox regression analysis, three independent associations with HIV seroconversion were reporting a genital ulcer (adjusted HR, 3.55; 95% CI, 1.52-8.29), number of sex partners (adjusted HR, 1.10; 95% Cl, 1.01-1.21:), and being married but residing separately from one's wife (adjusted HR, 2.21; 95% CI, 1.00-4.89). Conclusions: Innovative and sustained workplace-based interventions are needed to address the high risk of HIV infection in this economically productive population. The predictors of HIV seroconversion described in this study underscore the need for public health efforts to simultaneously address the biological, socioeconomic and behavioural factors that continue to place individuals at risk of HIV in general populations of Africa.