'Kusvika taparadzaniswa nerufu' (Until death do us part).

A cross-sectional study of 7 667 non-virgins between 15 and 54 years of age was carried out to assess the protective effect of marriage against HIV acquisition in a rural population in Zimbabwe, whilst taking into account gender-differentials in risk factors for seroconversion. Persons in stable first marriages and long-term consensual cohabiting unions had higher odds of HIV infection than never-married people but a lower risk than those who had been divorced or widowed, even after adjusting for known confounders and significant risk factors for infection. Partner-related risk factors appear to play a more pivotal role in determining HIV prevalence in females than for males, for whom personal sexual behaviour risk factors are more dominant.