Objective: To study demographic, social, behavioral, and biological variables as risk factors for HIV infection among men and women in Kenya.

Methods: Data from the cross-sectional, population-based 2003 Kenya Demographic and Health Survey were used. During the course of survey fieldwork, 3273 women aged 15 to 49 years and 2941 men aged 15 to 54 years gave consent to have a few drops of blood taken for anonymous testing. HIV serostatus data for men and women were analyzed for their relationships to key characteristics using bivariate and multivariate techniques to determine factors associated with being HIV-positive.

Results: National HIV prevalence in Kenya was found to be 6.7%. In the analysis of the study sample, uncircumcised men were 4 times more likely to be HIV-positive than those who were not. Compared with nonpolygynously married women, widowed women (odds ratio [OR] = 10.9), divorced women (OR = 2.3), and women who were 1 of 3 or more wives (OR = 3.4) were all at higher risk for being HIV-positive. Both men and women from Nyanza province were at a significantly higher risk for infection with HIV (OR = 2.9 and 2.3, respectively) than were the men and women from Nairobi. Men aged 35 to 44 years had the highest risk of being HIV-positive, whereas the ages of highest risk for women were 25 to 29 years. Increased wealth was positively related to risk for HIV: the wealthiest women were 2.6 times more likely than the poorest women to be HIV-positive. A key finding was that both men and women who considered themselves to be at low risk for contracting HIV were, in fact, the most likely to be HIV-positive.

Conclusions: This analysis demonstrates that HIV is a multidimensional epidemic, with demographic, residential, social, biological, and behavioral factors all exerting influence on individual probability of becoming infected with HIV. Although all of these factors contribute to the risk profile for a given individual, the results suggest that differences in biological factors such as circumcision and sexually transmitted infections may be more important in assessing risk for HIV than differences in sexual behavior.