Risk factors for HIV infection at enrollment in an urban male factory cohort in Harare, Zimbabwe.

Between March 1993 and March 1995, volunteers at 40 Harare factories were interviewed regarding sociodemographic characteristics and behavior; HIV serostatus was also determined. Among 2,691 men enrolled, HIV prevalence was 19.4%. Prevalence rose 2-fold with each year of age in young men (< 23 years). In a multivariate logistic model that included sociodemographic and behavioral variables, compared with those between 25 and 44 years, both younger men (OR = 0.51) and older men (OR = 0.49) were less likely to be HIV positive. In addition, marriage (OR = 2.01), history of sexually transmitted disease (STD) (genital ulcer, OR = 4.93, urethral discharge OR = 1.75), multiple partners (OR = 1.58), cash payment for sex (OR = 2.55) and condom use (OR = 1.35) were all independent risk factors for HIV infection. Home ownership conferred lower risk. Self-reported condom use was seen as a marker of correct personal risk assessment men who used condoms reported more risk behavior and had a higher prevalence of HIV, either because condom use was not consistent or because infection occurred prior to adoption of condoms. This study confirms established risk factors in a general population sample not selected for high risk of HIV infection. It suggests rapid acquisition of infection among young men and the importance of marriage (and separation of spouses) as correlates of HIV infection.