Changes in HIV prevalence among socioeconomic groups in eight African countries: analysis of sequential cross-sectional surveys

Abstract

BackgroundAcross sub-Saharan Africa, HIV prevalence was highest among higher socioeconomic groups during the 1990s. Population-based data from Tanzania from 2003 to 2007 suggested that this pattern is changing, with HIV prevalence falling among higher-educated groups but remaining stable among those with lower levels of education, suggesting lower incidence in higher-educated groups over this period. A multi-country analysis has not previously been undertaken. MethodsWe collated data on sociodemographic factors and HIV infection from 16 nationally representative surveys of adults (aged 15–49 years) from Burkina Faso, Ethiopia, Malawi, Lesotho, Kenya, Rwanda, Tanzania, and Zimbabwe. Surveys were conducted 4–6 years apart; sample sizes ranged from 5357 to 29 812 (188 315 individuals total). We used logistic regression to assess gender-stratified associations between highest educational level attained and HIV status in each survey. We adjusted for age and urban/rural setting, and tested for interactions between education level, urban/rural setting, and age. Our primary hypothesis is that higher education level becomes less of a risk for HIV, or protective, over time; we report p values for the statistical interaction between education and survey year. FindingsIn Ethiopia, there is evidence that the association between HIV prevalence and higher education level was weaker and/or more protective in the second survey in young women (p=0·14) and all men (p=0·11). Similar patterns were observed in Malawian urban women (p=0·03) and rural men (p=0·02), Rwandan older men (p=0·16), Tanzanian men (p=0·07), and rural Zimbabwean women (p=0·07). In Burkina Faso, the risk associated with higher education levels increased (p=0·01) in older men. We found no other changes in association between education and prevalent HIV. InterpretationChanges in the social epidemiology of HIV across sub-Saharan Africa appear heterogeneous. In most cases, where there is evidence of changing association between education and HIV, education has become less risky, or protective, over time. We continue to explore the hypothesis that patterns may depend upon identifiable characteristics of a country's HIV epidemic. FundingThe study was supported through the STRIVE consortium by UKaid from the Department for International Development. However, the views expressed do not necessarily reflect the department's official policies.