Wealth and wealth inequality in adult HIV prevalence

Abstract

# Background Studies investigating a correlation between economic status of household (wealth or income) and human immunodeficiency virus (HIV) prevalence have shown relationships ranging from insignificant to strong-positive. Using data from over 400,000 Demographic and Health Surveys (DHS) respondents in 29 African low- and middle-income countries receiving assistance from the US President’s Emergency Plan for AIDS Relief (PEPFAR), this study was conducted on HIV prevalence as it relates to wealth and wealth inequality. # Methods Individual-level Demographic and Health Surveys (DHS) data was obtained for 2010-2016 from 29 PEPFAR recipient countries through the DHS data portal. These data link survey responses with the HIV test results from biomarker data. To adjust for sample selection bias in the observational data, an inverse probability reweighting (IPW) technique was used to create sample weights and estimate model parameters. # Results After reweighting, the wealth effects on prevalence are smaller, often statistically insignificant or inconsistent, indicating that higher levels of wealth are not associated with higher HIV prevalence. Low wealth (being poor) is associated with a higher risk for some segments of the population. The growing interest in the literature of "wealth" as a driver of the maturing HIV epidemic may be an artifact of the longer survival of wealthy people in the DHS observational data. The extent of wealth inequality in the region of residence is consistently a positive driver of higher prevalence. In all three wealth categories (tertiles of high, medium, low), more wealth inequality in a region is consistently associated with a higher HIV prevalence, other things remaining the same. # Conclusions The findings of the study suggest that supply-oriented policies (e.g. distribution of providers and programs) to integrate PEPFAR activities into country health systems targeted to areas where wealth inequality is high and where pockets of low-wealth populations are found would have the greatest impact.