Previous Presentation: Presented in part at the Conference on Retroviruses and Opportunistic Infections; February 6, 2008; Boston, Massachusetts. Context: Studies of factors associated with acquiring human immunodeficiency virus (HIV) are often based on prevalence data that might not reflect recent infections. Objective: To determine demographic, biological, and behavioral factors for recent HIV infection in Uganda. Design and Setting: Nationally representative household survey of cross-sectional design conducted in Uganda from August 2004 through January 2005; data were analyzed until November 2007. Participants: There were 11 454 women and 9905 men aged 15 to 59 years who were eligible. Questionnaires were completed for 10 826 women (95%) and 8830 men (89%); of those interviewed, blood specimens were collected for 10 227 women (94%) and 8298 men (94%). Main Outcome Measure: Specimens seropositive for HIV were tested with the BED IgG capture-based enzyme immunosorbent assay to identify recent seroconversions (median, 155 days) using normalized optical density of 0.8 and adjustments. Results: Of the 1023 HIV infections with BED results, 172 (17%) tested as recent. In multivariate analysis, risk factors associated with recent HIV infection included female sex (adjusted odds ratio [aOR], 2.4; 95% confidence interval [CI], 1.1-5.2); current marital status (widowed vs never married, aOR, 6.1; 95% CI, 2.8-13.3; divorced vs never married, aOR, 3.0; 95% CI, 1.5-6.1); geographic region (north central Uganda vs central Uganda/Kampala, aOR, 2.6; 95% CI, 1.7-4.1); number of sex partners in past year (≥2 compared with none; aOR, 2.9; 95% CI, 1.6-5.5); herpes simplex virus type 2 infection (aOR, 3.9;95%CI, 2.6-5.8); report of a sexually transmitted disease in the past year (aOR, 1.7; 95% CI, 1.2-2.4); and being an uncircumcised man (aOR, 2.5; 95% CI, 1.1-5.3). Among married participants, recent HIV infection was associated with never using condoms with partners outside of marriage (aOR, 3.2; 95% CI, 1.7-6.1) compared with individuals having no outside partners. The risk of incident HIV infection for married individuals who used condoms with at least 1 outside partner was similar to that of those who did not have any partners outside of marriage (aOR, 1.0; 95% CI, 0.3-2.7). Conclusion: A survey of individuals in Uganda who were tested with an HIV assay used to establish recent infection identified risk factors, which offers opportunities for prevention initiatives. ©2008 American Medical Association. All rights reserved.
