Skip Navigation LinksHome > July 31, 2009 - Volume 23 - Issue 12 > The effect of educational attainment and other factors on HI...
AIDS:
doi: 10.1097/QAD.0b013e32832d407e
Epidemiology and social

The effect of educational attainment and other factors on HIV risk in South African women: results from antenatal surveillance, 2000–2005

Johnson, Leigh Fa; Dorrington, Rob Ea; Bradshaw, Debbieb; Plessis, Hendrika duc; Makubalo, Lindiwed

Free Access
Article Outline
Collapse Box

Author Information

aCentre for Actuarial Research, University of Cape Town, South Africa

bBurden of Disease Research Unit, South African Medical Research Council, Cape Town, South Africa

cStatistics South Africa

dDepartment of Health, Pretoria, South Africa.

Received 13 November, 2008

Revised 15 April, 2009

Accepted 22 April, 2009

Correspondence to Dr Leigh Johnson, Centre for Actuarial Research, University of Cape Town, Private Bag, Rondebosch 7701, South Africa. Tel: +27 21 650 5761; fax: +27 21 650 5937; e-mail: Leigh.Johnson@uct.ac.za

Collapse Box

Abstract

Objectives: To assess the effect of educational attainment and other factors on the risk of HIV in pregnant South African women.

Design: Repeated cross-sectional surveys.

Methods: Pregnant women attending public antenatal clinics were tested for HIV annually between 2000 and 2005, and provided demographic information. Logistic regression models were applied separately to the data collected in each year, to identify factors associated with HIV infection. Data from all years were combined in a logistic regression model that tested for trends in HIV prevalence.

Results: Amongst women aged 15–24 years, HIV risk in those who had completed secondary education was significantly lower than in those who had only primary education, in all years except 2000. HIV risk increased by 8% per annum (odds ratio 1.08, 95% confidence interval 1.04–1.12) in young women with no secondary education but did not increase in young women with secondary education. In women aged 25–49 years, HIV risk increased over the 2000–2005 period, at all levels of educational attainment, and did not differ between women with completed secondary education and women with only primary education.

Conclusion: Together with other evidence, this study suggests that higher educational attainment did not protect against HIV in the early stages of the South African HIV/AIDS epidemic. In recent years, the risk of HIV infection in young South African women with completed secondary education has reduced significantly relative to that in young women with primary education, suggesting that HIV prevention strategies may have been more effective in more educated women.

Back to Top | Article Outline

Introduction

According to various estimates, there are close to five and a half million HIV-infected individuals living in South Africa, more than in any other country in the world [1–3]. The exceptional severity of the epidemic in the countries of Southern Africa necessitates a detailed understanding of the epidemiology of HIV/AIDS in these countries. In South Africa, national surveys of HIV prevalence in pregnant women attending public antenatal clinics have been conducted annually since 1990, and published reports have presented HIV prevalence levels by age, by province and most recently by health district [3–9]. These surveys have shown a rise in HIV prevalence from 1% in 1990 to 29% in 2006 [3]. There is substantial heterogeneity between provinces, with provincial HIV prevalence in the 2006 survey varying between 15 and 39% [3], and significant heterogeneity in HIV prevalence has also been found within provinces [10]. However, there has to date been no published multivariate analysis of the relationship between HIV infection and the various demographic variables recorded in the surveys.

Factors that are often thought to increase women's risk of HIV include low socioeconomic position relative to men and having substantially older male partners [11–15]. However, studies conducted in Africa have found that the relationship between these factors and the risk of HIV is complex. Higher educational attainment, which is often used as a proxy for higher socioeconomic status, has generally been found to be associated with increased risk of HIV in the early stages of the HIV/AIDS epidemic, both in men and women [16]. However, in more mature epidemics in which HIV/AIDS education programmes have been introduced, higher educational attainment appears to become protective against HIV [17–23]. Studies [24,25] have also suggested that the effect of having an older partner on a woman's risk of HIV infection may differ in relation to the woman's age, with young women tending to have their risk increased to a greater extent than older women.

This study aims to examine whether the relationship between HIV risk, educational attainment and partner age in South African women differs from that observed in other African studies. Two hypotheses are tested, using antenatal clinic survey data collected between 2000 and 2005. On the basis of observed age variation in the effect of educational attainment on HIV risk [18,19,26] and the effect of partner age difference on HIV risk [24,25], the first hypothesis tested is that the effects of these variables differ by age. The second hypothesis tested is that trends in HIV risk differ by educational attainment, as suggested by evidence from other African settings [16,17]. An ancillary objective of this study is to assess whether the observed interprovincial heterogeneity in HIV prevalence can be explained by differences in the demographic variables recorded in the antenatal surveys.

Back to Top | Article Outline

Methods

Sampling procedures and laboratory methods are described in detail in the survey protocol [27] and in the published survey reports [4–9]. Briefly, the survey was conducted during October of each year, by the South African Department of Health, amongst women attending public sector antenatal clinics for the first time in their current pregnancy. Clinics were sampled on a probability proportional to size (PPS) basis within each province, the measure of size being the average monthly number of antenatal visits to the clinic. The survey followed an unlinked, anonymous, cross-sectional design. The purpose of the survey was explained to all eligible women, and those who consented to participate in the survey provided demographic information that was recorded on a data collection form, which was linked to the HIV sample by a unique barcode. In addition to the blood sample that was routinely collected from all pregnant women for the purpose of syphilis screening at their first antenatal visit, women participating in the survey provided a second Vacutainer (Becton Dickinson and Company; Franklin Lakes, New Jersey, USA) of blood. Participants who wished to know their HIV status were referred for voluntary counselling and testing (VCT) within the clinic or at the nearest VCT site.

Blood samples were stored at 4°C before being transported to participating laboratories. In line with WHO recommendations for populations with a prevalence of more than 10% [28], a single enzyme-linked immunosorbent assay test (Abbott Axsym System for HIV-1/HIV-2, Abbott Park, Illinois, USA) was applied to each blood sample, and in the Western Cape province, where HIV prevalence is relatively low, confirmatory testing was also conducted. Testing for syphilis was conducted using a rapid plasma reagin (RPR) test. Quality assurance of HIV testing and RPR testing was conducted by the National Institute for Communicable Diseases and the Medical University of Southern Africa, respectively. Data were initially entered by participating laboratories and then reentered and checked by provincial coordinators.

Statistical analysis of the data was conducted using STATA 9.2 (StataCorp, College Station, Texas, USA), with survey commands that reflected the stratification by province, clustering by clinic and sample weights. The analysis was limited to the data collected from 2000 to 2005, as clinic identifiers could not be obtained prior to 2000, and clinic identifiers changed after 2005, following a substantial change in the clinics that were sampled. Three different statistical models were applied to the data. First, multivariate logistic regression analyses [29] were run separately for data collected in each year, not allowing for any interactions (Model 1). To test the hypothesis of age variation in the effect of educational attainment and partner age difference on HIV risk, the same multivariate regression was also applied separately to the 15–24 and 25–49 years age groups (Model 2). The second hypothesis, of differences in HIV prevalence trends by educational attainment, was tested by combining data from 2000 to 2005 (Model 3). As with Model 2, this multivariate analysis was conducted separately for the 15–24 and 25–49 years age groups. Tests for trend were conducted by including year as a continuous explanatory variable in the model, and allowing for interactions between year and educational attainment.

Back to Top | Article Outline

Results

A total of 99 153 women were tested for HIV from 2000 to 2005, with between 16 000 and 17 000 tested in each year. Sample sizes and levels of HIV prevalence in each year are shown in the online table. Over the 2000–2005 period, HIV prevalence rose from 24.5% [95% confidence interval (CI) 23.4–25.6] to 30.2% (95% CI 29.1–31.2), although HIV prevalence in women aged 15–24 years remained relatively stable. HIV prevalence levels increased at all levels of educational attainment. However, in women aged 15–24 years, HIV prevalence increased only in those women with no secondary education (Fig. 1).

Fig. 1
Fig. 1
Image Tools

In the first multivariate analysis (Table 1), it was found that women who had incomplete secondary education had a higher risk of HIV infection than women who had only primary education, although this was only significant at the 5% level in 3 years (2000, 2003 and 2004). Women who had completed their secondary education, however, were at a significantly reduced risk of HIV infection when compared with women who had only primary education, in all years except 2000. Women whose partners were 5 or more years older had a significantly increased risk of HIV infection in all years. Race was the most significant factor affecting HIV risk, with HIV risk in African women being substantially higher than that in women of other race groups. The risk of HIV infection varied significantly between the provinces, being consistently highest in KwaZulu-Natal and lowest in Limpopo. In the four provinces that were consistently associated with the highest HIV risk during the 2000–2002 period (Free State, Gauteng, KwaZulu-Natal and Mpumalanga), there appeared to be reductions in HIV risk relative to the Western Cape (baseline) over the 2000–2005 period. HIV risk also varied significantly in relation to age, with HIV risk being consistently highest in the 25–29 years age group. Other factors that were found to be significantly associated with the risk of HIV were second pregnancy and syphilis infection.

Table 1
Table 1
Image Tools

When the multivariate analysis was performed separately for the 15–24 and 25–49 years age groups (Model 2), the effect of educational attainment and partner age was found to differ considerably in the two age groups (Table 2). Amongst women aged 15–24 years, incomplete secondary education did not significantly affect the risk of HIV infection, but completed secondary education was associated with a significantly reduced risk of HIV infection in all years except 2000. Amongst women aged 25–49 years, however, incomplete secondary education was associated with a significantly increased risk of HIV infection in all years except 2001 and 2002, and the risk of infection in women with completed secondary education did not differ significantly from that amongst women with only primary education, except in 2001. Amongst young women, having a partner 5 or more years older significantly increased the risk of HIV infection, but amongst women aged 25 and older, having an older partner tended to reduce the risk of HIV, with this reduction reaching statistical significance in 2004 and 2005 [OR 0.90 (95% CI 0.80–1.00) and OR 0.88 (95% CI 0.78–0.98), respectively]. Other variables had similar effects to those estimated in Model 1 (results not shown).

Table 2
Table 2
Image Tools

When the data from all years were combined to test for trends (Model 3), trends in HIV risk were found to differ significantly in relation to age and educational attainment (Table 3). In the 15–24 years age group, there was no significant change in HIV risk over time amongst women with incomplete or completed secondary education, but amongst women who had no secondary education the risk of HIV infection increased by 8% per annum (OR 1.08 per annum, 95% CI 1.04–1.12). In the 25–49 years age group, the risk of HIV infection increased significantly over time, and the extent of this increase was not significantly affected by women's educational attainment.

Table 3
Table 3
Image Tools
Back to Top | Article Outline

Discussion

This analysis suggests that in recent years, higher educational attainment has become protective against HIV in young South African women, and there have been significant differences in trends in HIV risk amongst young women with different levels of education. This could be due to recent social marketing and school-based HIV/AIDS lifeskills programmes having more impact in educated women. Consistent with studies in Uganda [19,22] and Zambia [18], the protective effect of higher educational attainment appears to be limited largely to young women. The absence of a clear association between educational attainment and HIV risk at older ages, in our analysis, may be an indication that HIV prevention strategies have been more successful in young women than in older women. Alternatively, it may reflect the longer average duration of HIV infection in older women, most of whom would have been infected prior to the introduction of significant social marketing programmes.

Earlier data confirm the general lack of association between educational attainment and HIV risk in the early stages of the South African epidemic. Multivariate analysis of antenatal clinic data collected in 1998 and 1999 confirmed the patterns observed in 2000 (i.e. no protective effect of completed secondary education, either at young ages or at older ages), although these data were not included in the present analysis due to sample weights and clinic identifiers not being available. More recent South African studies [30,31] have found that higher educational attainment is associated with significantly reduced HIV incidence, suggesting that the relationship between educational attainment and HIV risk may have changed over time.

This analysis suggests that the effect of having an older partner also varies in relation to women's age. In young women with a partner 5 or more years older, the odds of HIV infection is 40 to 70% greater than that in young women with a partner less than 5 years older. However, in women aged 25 years and older, having an older partner does not increase HIV risk. This could be because in men aged 35 years and older, older age is associated with reduced HIV risk, and having an older male partner is therefore associated with reduced risk of HIV exposure.

Substantial interprovincial differences in HIV prevalence have been noted in previous reports, and this analysis demonstrates that these differences persist even after controlling for differences in demographic profiles between provinces. To some extent, these are due to differences in epidemic timing, as shown in the reductions in the odds of HIV in the four worst affected provinces, relative to the Western Cape, over the 2000–2005 period. Interprovincial differences in HIV prevalence could also be explained by differences in the prevalence of male circumcision [32] and differences in levels of urbanization [33], factors not controlled for in this analysis.

A key strength of this analysis is that it is based on what is probably the largest HIV prevalence data set in the world. The large sample size makes it possible to test for associations and interactions that would normally not be statistically significant with smaller samples. However, a major limitation of the study is that no information was captured in respect of women who refused to participate in the survey, except in the Western Cape. In the Western Cape, participation rates were close to 98% [10], suggesting that nonresponse bias is unlikely to distort the results substantially. An additional limitation is that the sample is drawn only from public health facilities, and hence there remains some uncertainty regarding the relationship between educational attainment and HIV risk in the general population, as highly educated women are more likely to attend private health facilities. Reductions in fertility due to HIV and differences in sexual behaviour between pregnant and nonpregnant women also limit the extent to which the findings can be generalized to the rest of the population.

The explanation for the growing gap between educated and less educated women, in terms of HIV risk, is not completely clear, but there are several possible policy implications. First, there is a need to reduce rates of school dropout and to strengthen existing school HIV/AIDS lifeskills programmes. Second, social marketing programmes may need to develop materials more appropriate for less educated individuals with low literacy, and it may be appropriate to introduce special campaigns in communities with low levels of education. Last, measures to improve the socioeconomic status of women may be appropriate in empowering women to reduce their exposure to HIV risk [34,35].

Back to Top | Article Outline

Acknowledgements

The South African Department of Health is thanked for providing the antenatal survey data.

L.M. was responsible for the survey design and organization. H.dP. and R.D. assisted in the processing and coding of the data. L.J. conducted the statistical analysis of the data and D.B. and R.D. assisted in the analysis. L.J. drafted the paper and all other authors participated in the writing of the paper.

There are no conflicts of interest.

Back to Top | Article Outline

References

1. UNAIDS. 2008 Report on the global AIDS epidemic. Geneva; 2008. http://www.unaids.org/en/KnowledgeCentre/HIVData/GlobalReport/2008/2008_Global_report.asp.

2. Dorrington RE, Johnson LF, Bradshaw D, Daniel T. The demographic impact of HIV/AIDS in South Africa: National and Provincial Indicators for 2006. Cape Town: Centre for Actuarial Research, South African Medical Research Council and Actuarial Society of South Africa; 2006. http://www.commerce.uct.ac.za/care.

3. Department of Health. National HIV and Syphilis Prevalence Survey: South Africa 2006. Pretoria; 2007. http://www.doh.gov.za/docs/reports-f.html.

4. Department of Health. National HIV and Syphilis Prevalence Survey, South Africa 2005. Pretoria: Directorate of Epidemiology and Surveillance; 2006.

5. Department of Health. National HIV and syphilis antenatal sero-prevalence survey in South Africa 2004; 2005. http://www.doh.gov.za/docs/reports-f.html.

6. Department of Health. National HIV and syphilis antenatal sero-prevalence survey in South Africa 2003; 2004. http://www.doh.gov.za/docs/index.html.

7. Department of Health. National HIV and syphilis antenatal sero-prevalence survey in South Africa 2002; 2003. http://www.doh.gov.za/docs/reports-f.html.

8. Department of Health. National HIV and syphilis sero-prevalence survey in South Africa 2001; 2002. http://www.doh.gov.za/docs/index.html.

9. Department of Health. National HIV and syphilis sero-prevalence survey of women attending public antenatal clinics in South Africa 2000; 2001. http://www.doh.gov.za/docs/index.html.

10. Shaikh N, Abdullah F, Lombard CJ, Smit L, Bradshaw D, Makubalo L. Masking through averages: intraprovincial heterogeneity in HIV prevalence within the Western Cape. S Afr Med J 2006; 96:538–543.

11. Pettifor AE, Measham DM, Rees HV, Padian NS. Sexual power and HIV risk, South Africa. Emerg Infect Dis 2004; 10:1996–2004.

12. Eaton L, Flisher A, Aarø L. Unsafe sexual behaviour in South African youth. Soc Sci Med 2003; 56:149–165.

13. Gregson S, Nyamukapa CA, Garnett GP, Mason PR, Zhuwau T, Carael M, et al. Sexual mixing patterns and sex-differentials in teenage exposure to HIV infection in rural Zimbabwe. Lancet 2002; 359:1896–1903.

14. Jewkes R, Vundule C, Maforah F, Jordaan E. Relationship dynamics and teenage pregnancy in South Africa. Soc Sci Med 2001; 52:733–744.

15. Katz I, Low-Beer D. Why has HIV stabilized in South Africa, yet not declined further? Age and sexual behavior patterns among youth. Sex Transm Dis 2008; 35:837–842.

16. Hargreaves JR, Glynn JR. Educational attainment and HIV-1 infection in developing countries: a systematic review. Trop Med Int Health 2002; 7:489–498.

17. Hargreaves JR, Bonell CP, Boler T, Boccia D, Birdthistle I, Fletcher A, et al. Systematic review exploring time trends in the association between educational attainment and risk of HIV infection in sub-Saharan Africa. AIDS 2008; 22:403–414.

18. Michelo C, Sandøy IF, Fylkesnes K. Marked HIV prevalence declines in higher educated young people: evidence from population-based surveys (1995–2003) in Zambia. AIDS 2006; 20:1031–1038.

19. de Walque D. How does the impact of an HIV/AIDS information campaign vary with educational attainment? Evidence from rural Uganda. World Bank; 2004. http://www.worldbank.org.

20. Mmbaga EJ, Hussain A, Leyna GH, Mnyika KS, Sam NE, Klepp KI. Prevalence and risk factors for HIV-1 infection in rural Kilimanjaro region of Tanzania: implications for prevention and treatment. BMC Public Health 2007; 7:58.

21. Mmbaga EJ, Leyna GH, Mnyika KS, Hussain A, Klepp KI. Education attainment and the risk of HIV-1 infections in rural Kilimanjaro Region of Tanzania, 1991–2005: a reversed association. Sex Transm Dis 2007; 34:947–953.

22. Kilian AH, Gregson S, Ndyanabangi B, Walusaga K, Kipp W, Sahlmuller G, et al. Reductions in risk behaviour provide the most consistent explanation for declining HIV-1 prevalence in Uganda. AIDS 1999; 13:391–398.

23. Gregson S, Garnett GP, Nyamukapa CA, Hallett TB, Lewis JJ, Mason PR, et al. HIV decline associated with behavior change in eastern Zimbabwe. Science 2006; 311:664–666.

24. Pettifor AE, Rees HV, Kleinschmidt I, Steffenson AE, Macphail C, Hlongwa-Madikizela L, et al. Young people's sexual health in South Africa: HIV prevalence and sexual behaviors from a nationally representative household survey. AIDS 2005; 19:1525–1534.

25. Kelly RJ, Gray RH, Sewankambo NK, Serwadda D, Wabwire-Mangen F, Lutalo T, et al. Age differences in sexual partners and risk of HIV-1 infection in rural Uganda. J Acquir Immun Defic Syndr 2003; 32:446–451.

26. Fylkesnes K, Musonda RM, Kasumba K, Ndhlovu Z, Mluanda F, Kaetano L, et al. The HIV epidemic in Zambia: socio-demographic prevalence patterns and indications of trends among childbearing women. AIDS 1997; 11:339–345.

27. Abdool Karim SS, Colvin M, Mullick S. The South African HIV/AIDS/STD Surveillance System Manual. Durban: The Centre for Epidemiological Research in South Africa, Medical Research Council; 1997.

28. UNAIDS/WHO. Guidelines for using HIV testing technologies in surveillance: selection, evaluation and implementation; 2001. http://data.unaids.org/Publications/IRC-pub02/JC602-HIVSurvGuidel_en.pdf.

29. Hosmer DW, Lemeshow S. Applied logistic regression. New York: John Wiley & Sons; 1989.

30. Hargreaves JR, Bonell CP, Morison LA, Kim JC, Phetla G, Porter JD, et al. Explaining continued high HIV prevalence in South Africa: socioeconomic factors, HIV incidence and sexual behaviour change among a rural cohort, 2001–2004. AIDS 2007; 21(Suppl 7):S39–S48.

31. Bärnighausen T, Hosegood V, Timæus IM, Newell ML. The socioeconomic determinants of HIV incidence: evidence from a longitudinal, population-based study in rural South Africa. AIDS 2007; 21(Suppl 7):S29–S38.

32. Connolly C, Simbayi LC, Shanmugam R, Nqeketo A. Male circumcision and its relationship to HIV infection in South Africa: results of a national survey in 2002. S Afr Med J 2008; 98:789–794.

33. Statistics South Africa. Census in brief; 1999. www.statssa.gov.za.

34. Pronyk PM, Hargreaves JR, Kim JC, Morison LA, Phetla G, Watts C, et al. Effect of a structural intervention for the prevention of intimate-partner violence and HIV in rural South Africa: a cluster randomised trial. Lancet 2006; 368:1973–1983.

35. Pronyk PM, Kim JC, Abramsky T, Phetla G, Hargreaves JR, Morison LA, et al. A combined microfinance and training intervention can reduce HIV risk behaviour in young female participants. AIDS 2008; 22:1659–1665.

Fig. 1
Fig. 1
Image Tools

Cited By:

This article has been cited 5 time(s).

Bmj Open
The relationship between age of coital debut and HIV seroprevalence among women in Durban, South Africa: a cohort study
Wand, H; Ramjee, G
Bmj Open, 2(1): -.
ARTN e000285
CrossRef
Bmj Open
HIV incidence and prevalence among cohorts of women with higher risk behaviour in Bloemfontein and Rustenburg, South Africa: a prospective study
Feldblum, PJ; Latka, MH; Lombaard, J; Chetty, C; Chen, PL; Sexton, C; Fischer, S
Bmj Open, 2(1): -.
ARTN e000626
CrossRef
Addiction
Factors associated with pathways toward concurrent sex work and injection drug use among female sex workers who inject drugs in northern Mexico
Morris, MD; Lemus, H; Wagner, KD; Martinez, G; Lozada, R; Gomez, RMG; Strathdee, SA
Addiction, 108(1): 161-170.
10.1111/j.1360-0443.2012.04016.x
CrossRef
Bmc Public Health
HIV risk perception and distribution of HIV risk among African, Caribbean and other Black people in a Canadian city: mixed methods results from the BLACCH study
Baidoobonso, S; Bauer, GR; Speechley, KN; Lawson, E
Bmc Public Health, 13(): -.
ARTN 184
CrossRef
Plos One
HIV Prevalence by Race Co-Varies Closely with Concurrency and Number of Sex Partners in South Africa
Kenyon, C; Buyze, J; Colebunders, R
Plos One, 8(5): -.
ARTN e64080
CrossRef
Back to Top | Article Outline
Keywords:

educational achievement; HIV seroprevalence; risk factors

© 2009 Lippincott Williams & Wilkins, Inc.

Login

Search for Similar Articles
You may search for similar articles that contain these same keywords or you may modify the keyword list to augment your search.