The effects of household wealth on HIV prevalence in Manicaland, Zimbabwe – a prospective household census and population-based open cohort study

INTRODUCTION: Intensified poverty arising from economic decline and crisis may have contributed to reductions in HIV prevalence in Zimbabwe. OBJECTIVES: To assess the impact of the economic decline on household wealth and prevalent HIV infection using data from a population-based open cohort. METHODS: Household wealth was estimated using data from a prospective household census in Manicaland Province (1998 to 2011). Temporal trends in summed asset ownership indices for sellable, non-sellable and all assets combined were compared for households in four socio-economic strata (small towns, agricultural estates, roadside settlements and subsistence farming areas). Multivariate logistic random-effects models were used to measure differences in individual-level associations between prevalent HIV infection and place of residence, absolute wealth group and occupation. RESULTS: Household mean asset scores remained similar at around 0.37 (on a scale of 0 to 1) up to 2007 but decreased to below 0.35 thereafter. Sellable assets fell substantially from 2004 while non-sellable assets continued increasing until 2008. Small-town households had the highest wealth scores but the gap to other locations decreased over time, especially for sellable assets. Concurrently, adult HIV prevalence fell from 22.3 to 14.3%. HIV prevalence was highest in better-off locations (small towns) but differed little by household wealth or occupation. Initially, HIV prevalence was elevated in women from poorer households and lower in men in professional occupations. However, most recently (2009 to 2011), men and women in the poorest households had lower HIV prevalence and men in professional occupations had similar prevalence to unemployed men. CONCLUSIONS: The economic crisis drove more households into extreme poverty. However, HIV prevalence fell in all socio-economic locations and sub-groups, and there was limited evidence that increased poverty contributed to HIV prevalence decline.