Revisiting the education-HIV gradient in Africa: Cohort variation analysis of a temporal shift and speculation on causes

This study examines the changing relationship between education and HIV in Africa over time. While a negative relationship between education and morbidity is generally well-established, conflicting reports on the education-HIV gradient in Africa raised doubts about its consistency. Studies found positive, null, or negative gradients, suggesting that education might not influence HIV risk as expected. This paper argues that these conflicting findings could reflect a temporal shift in the education-HIV gradient, rather than errors in epidemiological data. Using Demographic and Health Surveys and a cohort analysis of 304,630 individuals from eight African countries, the study shows that, initially, education was linked to higher HIV infection rates for those born between 1960 and 1984, peaking in the 1965-69 cohort. However, from those born between 1985 and 1989 onward, education became protective, with gradients becoming stronger over time. This temporal shift is not sensitive to regional differences and methodologies and is corroborated by similar shifts for other risky behaviors. Additional hypothesized cohort variations for males and females, as well as countries with differing levels of HIV mitigation efforts, suggest potential drivers behind this shift. The dynamic ways in which education is associated with health adds complexity to the fundamental cause theory.