Relationship between socioeconomic status and HIV infection: findings from a survey in the Free State and Western Cape Provinces of South Africa

Abstract

Background: Studies have shown a mixed association between socioeconomic status (SES) and prevalent HIV infection across and within settings in sub-Saharan Africa. In general, the relationship between years of formal education and HIV infection changed from a positive to a negative association with maturity of the HIV epidemic. Our objective was to determine the association between SES and HIV in women of reproductive age in the Free State (FSP) and Western Cape Provinces (WCP) of South Africa (SA).

Study design: Cross-sectional.

Setting: SA.

Methods: We conducted secondary analysis on 1906 women of reproductive age from a 2007 to 2008 survey that evaluated effectiveness of Prevention of Mother-to-Child HIV Transmission Programmes. SES was measured by household wealth quintiles, years of formal education and employment status. Our analysis principally used logistic regression for survey data.

Results: There was a significant negative trend between prevalent HIV infection and wealth quintile in WCP (P<0.001) and FSP (P=0.025). In adjusted analysis, every additional year of formal education was associated with a 10% (adjusted OR (aOR) 0.90 (95% CI 0.85 to 0.96)) significant reduction in risk of prevalent HIV infection in WCP but no significant association was observed in FSP (aOR 0.99; 95% CI 0.89 to 1.11). There was no significant association between employment and prevalent HIV in each province: (aOR 1.54; 95% CI 0.84 to 2.84) in WCP and (aOR 0.96; 95% CI 0.71 to 1.30) in FSP.

Conclusion: The association between HIV infection and SES differed by province and by measure of SES and underscores the disproportionately higher burden of prevalent HIV infection among poorer and lowly educated women. Our findings suggest the need for re-evaluation of whether current HIV prevention efforts meet needs of the least educated (in WCP) and the poorest women (both WCP and FSP), and point to the need to investigate additional or tailored strategies for these women.

Keywords: education; hiv or aids; socioeconomic status.

Figure 1

HIV prevalence by age. Data were derived from the Prevention of Mother to Child Transmission of HIV, Effectiveness in Africa and, Research and Linkages to HIV Care survey (PEARL study). There were only three study participants aged 45–49 years and one aged ≥50 years.

Figure 2

HIV prevalence by household wealth quintiles. Data were derived from the Prevention of Mother to Child Transmission of HIV, Effectiveness in Africa and, Research and Linkages to HIV Care survey (PEARL study). FSP, study participants from the Free State Province; WCP, study participants from the Western Cape Province. Both=data for both WCP and FSP. Household wealth quintiles were generated separately for each province using data for the specified province only. The wider CI for WCP than FSP is due to clustering, as described in text. Furthermore, note that WCP had larger probability weights, used in weighting of the survey-derived data than FSP, as described in text.