Hargreaves, J. and Boler, T. (2006) Girl power: the impact of girls’ education on HIV and sexual behaviour. Johannesburg, South Africa: ActionAid International.Link to Source
The authors include a total of 45 studies from various countries in eastern, southern and central Africa, which measure the associations between education and HIV-related outcomes. Fifteen investigate the relationship between education and HIV rates, seven investigate the link between female educational attainment and age of sexual debut, 13 investigate the link between girls’ education on condom use, and nine assess the link between education and other characteristics of sexual relationships. As noted by the authors, the evidence is mainly based on designs in which conclusively attributing causality is not possible. The authors report the following findings:
- Overall, the study results suggest that education is associated with a decrease in HIV rates and a reduction of risky sexual behaviour (such as early sexual debut, number of sexual/casual partners and unprotected sex).
- The authors report the evidence suggests a reversal of the trend for education to increase HIV risk, as suggested by studies conducted after 1995 in eastern, southern and central Africa (probably as more information about HIV/AIDS became included in the curricula). They emphasise that education is associated with a reduction of HIV rates and risky sexual behaviour (early sexual debut, number of sexual/casual partners and unprotected sex) for girls and boys alike since 1995.
- Secondary education appears to be more strongly associated than primary education with reduced HIV vulnerability in women. The authors argue that one underlying mechanism of the link between education and HIV vulnerability is gender inequality, and they argue that education empowers women, particularly within sexual relationships.
- The relationship between education and other characteristics of sexual relationships (such as levels of coercive sex, transactional sex, age difference between partners, relationships with commercial sex workers) is difficult to assess, owing to the low number of studies concerned with the respective outcome variables.
AIDS was diagnosed for the first time in 1981. In those early days the epidemic mainly affected men, but particularly in developing countries an increased trend towards high infection rates among young females has been observed. The reasons for the feminisation of the epidemic are manifold, but the major factor can be broadly ascribed to ‘gender inequality’. One way of addressing this inequality is to empower girls and young women through access to education. An earlier systematic review concluded that in Africa higher educational attainment tends to be associated with a greater risk of HIV infection. This systematic review aims to investigate whether education can reduce HIV vulnerability and improve sexual behaviour in girls in eastern, southern and central Africa.
The objectives of this systematic review are to assess the evidence on the impact of girls’ education on sexual behaviour and HIV vulnerability, to distinguish the impacts of primary versus secondary education on these outcomes, and to analyse possible mechanisms underlying the link between education and HIV.
The authors included peer-reviewed empirical studies published in academic journals between 1990 and 2006 which reported on the association between indicators of education and indicators of sexual behaviour and/or prevalence or incidence of HIV infection. To be included, studies had to adjust for possible confounding factors such as age, gender and setting that could affect the results. The authors note that “most of the studies were cross-sectional (i.e., they only collected data at one point in time). It is therefore not possible to determine causality”. (p. 21).
Included studies had to be conducted in eastern, southern or central Africa, had to focus on women or contain women in the sample (men-only studies were excluded), and had to have a broadly representative sample size of at least 500 individuals.
The authors conducted a search of the published literature, using the major medical research databases Embase and PubMed. They also screened the reference lists of the most relevant articles.The authors did not assess study quality and they applied a narrative-synthesis approach, including elements of vote counting. They present a theoretical model of impact.
The systematic review reports criteria for study inclusion and information about the included studies. However, the review has major limitations. Most importantly, the review is limited to evidence from association studies, but the findings are interpreted as evidence of impact and causality. This is despite the authors’ own acknowledgement that “most of the studies were cross-sectional (i.e., they only collected data at one point in time). It is therefore not possible to determine causality” (p. 21). Nor do the authors assess the quality of the included studies in attributing outcomes to the intervention. This is especially problematic, because the review authors draw quite strong policy conclusions which go far beyond the scope and informative value of the review. The review fails to specify the included interventions, and the search underlying the review is not sufficiently comprehensive with respect to the databases searched and the explicit exclusion of unpublished and grey literature. The authors acknowledge the limitations of the existing evidence in terms of its geographical focus on eastern, southern and central Africa, noting that their findings might not be generalisable to other contexts.