Every additional year of schooling among young people in
Botswana was associated with an 8% reduction in the risk of HIV infection in
the years between 2004 and 2008, according to a study published last week in Lancet Global Health. The study provides
further strong evidence that improving retention in education protects against
HIV infection, especially for girls, say the authors.
Although some studies, conducted in South Africa, Zambia and
Uganda have shown that a greater duration of education is associated with a
lower likelihood of being HIV-positive, evidence from randomised studies of interventions
which keep young people in school is mixed regarding their effects. This may be
because the mechanisms by which education is linked to a reduced risk of
acquiring HIV are not direct. For example, being at school all day might not be
protective, and sex education acquired during school time might not be
protective, but on the other hand, having a higher level of education may
enable young people – especially girls – to gain jobs and livelihoods that will
place them at lower risk of HIV infection, and make them less vulnerable to
economic pressures for transactional sex.
The clearest evidence of a positive effect was found in a
study of cash transfers for school-age adolescent girls in Malawi. Even here,
the study found no difference between cash transfers that were unconditional or
those that were conditional on attending school – leading to speculation that
the effect could well be economic.
The lack of clarity on the effect of duration of schooling
led researchers at Harvard School of Public Health, the Botswana-Harvard AIDS
Institute and Boston University’s Center for Global Development to look at the
effect of a change in educational policy that encouraged young people to stay
in school on HIV prevalence in Botswana between national surveys in 2004 and
In 1996 Botswana shifted the incorporation of year 10 into
junior secondary schools. In order to qualify for most vocational programmes
young people must complete junior secondary school, so the effect of the change
was to encourage an average of one extra year of education for young people
Botswana also had good data on HIV prevalence among young
people affected by the change in policy, gathered through national AIDS Impact
Surveys conducted in 2004 and 2008. Using this data Jan-Walter De Neve and
colleagues analysed the effects of the policy change on educational attainment,
on years of schooling and on HIV status of young people affected by the policy
change compared to those not affected, after controlling for age.
AIDS Impact Survey data was available for 7018 respondents
(3965 women and 3053 men) with HIV antibody information who were born after 1975
and aged over 18 years at the time of sampling. In this study sample the reform
increased the average number of years of schooling by 0.79 years (P<0.0001).
Among this sample the baseline HIV prevalence was 25.5%, but the cumulative
risk of HIV infection was reduced by 8.1% for each additional year of schooling
caused by the policy change. (p=0.008) Each additional year of schooling after
the completion of nine years of schooling was associated with a reduction in
risk, but schooling prior to this point was not protective.
The effect was significant for both women and men, although
the effect was stronger in women (11.6% reduction in risk for each year of additional
The cost per HIV infection averted was $27,753, and the investigators
calculated that assuming anyone infected with HIV would eventually start
antiretroviral therapy with a lifetime cost of $12,400, the incremental
cost-effectiveness ratio of a year of secondary education would be $4387,
making it highly cost effective by World Health Organization standards. The
authors note that although circumcision and treatment as prevention would be cheaper
buys from an HIV prevention perspective, their analysis does not factor in the
wide economic benefits of additional education.
The authors speculate that the change in schooling duration
resulting from the Botswana policy change may have affected HIV risk because
years 10 to 12 are “a period where sexual behaviour patterns and labour market
opportunities are formed.” They caution that the effects might not be similar
in settings with a lower HIV prevalence, and that the results are a snapshot of
the effects of the reforms during the mid-2000s, the decade after the policy change.
Professor Jacob Bor of Boston University’s Center for Global
Development, a study investigator, said: "This study is among the first to
provide causal evidence that secondary education is an important causal
determinant of HIV infection. Our results suggest that schooling should be
considered alongside other proven interventions as part of a multi-pronged
'combination' HIV prevention strategy. Expanding the opportunities of young
people through secondary schooling will not only have economic benefits but
will also yield health benefits and should be a key priority for countries with
generalised HIV epidemics."