Background Young women are known to be particularly at risk for HIV infection. Identifying characteristics associated with HIV-acquisition in younger and older women could assist with developing effective interventions which target the determinants of HIV.

Methods The MDP301 study was an international, multi-centre, randomised controlled trial to assess the efficacy of 0.5% PRO2000/05 microbicide gel. 2508 HIV-negative women, ≥18 years, were enrolled in Soweto and Orange Farm and followed up for 12 months. Associations between baseline demographic, behavioural and clinical risk factors and HIV acquisition were assessed using univariate Poisson regression.

Results Data on 2451 women were analysed. 110 seroconversions were observed over 2356.5 woman-years (wy). Overall, HIV incidence was 46.7/1000 wy (95% CI 38.7 to 56.3). Younger women (18–24 years) were more likely to acquire HIV (IRR1.4, 95% CI 1.0 to 2.0) than older women (≥25 years). Difficulty accessing money for medical treatment (IRR 1.5, 95% CI 0.06 to 2.0, p=0.019) and being HSV2 seropositive at baseline (IRR1.7, 95% CI 0.99 to 2.8, p=0.055) were associated with an increased risk of seroconversion in younger women. In older women, decreasing partner contribution to the household economy (IRR 1.3, 95% CI 1 to 1.7) and Trichomonas vaginalis infection at enrolment (IRR 2.3, 95% CI 1.1 to 4.9) were associated with increased risk of HIV acquisition. Older women who reported condom use at last sex act had a greater risk of acquiring HIV (IRR 2.8, 95% CI 1.3 to 5.9), which could reflect high risk sexual behaviour in this group. Reporting multiple partners was associated with increased risk of HIV acquisition in both younger (2.6, 1.3 to 5.3) and older women (2.8, 1.3 to 6.1).

Conclusions STIs remain an important marker of ongoing HIV risk. Interventions that address the economic context of women's HIV risk will be important to evaluate. Additional investigation using multivariate analysis may elucidate these findings further.