Because adolescents in Sub-Saharan Africa who have lost one or both parents are more likely than nonorphans to drop out of school, some researchers have suggested that differences in educational attainment and school attendance may explain excess sexual risk among orphans. However, in a cross-sectional study of female adolescents in urban Zimbabwe,1 disparities in schooling explained only a small portion of the excess sexual risk found among youth who had lost both parents, and did not explain high levels of sexual risk among those who had lost only one parent.

The data come from a randomized, household-based study conducted in 2004 in Highfield, a densely populated suburb of Harare. All 15–19-year-old women in the sampled households were invited to be interviewed about their living situations and sexual experiences (a small number of 14- and 20-year-olds completed interviews and were retained in the sample); consenting participants were also tested for HIV and HSV-2. About two-thirds of eligible young women completed an interview, and 97% of interviewees provided a sample for STI testing. These data were matched with survey data from parents on the adolescents' coresidents and dwellings. Most analyses focused on the 743 young women who were unmarried, since associations between orphan status and sexual risk were evident among the unmarried young women but not those who were or had been married; the only exception was the initial analysis of HIV and HSV-2 prevalence, which also included 110 married adolescents.

Participants were classified as paternal, maternal or double orphans (depending on whether they had lost their father or mother only, or both parents, respectively) or as nonorphans. Paternal orphans accounted for 26% of the sample, maternal orphans for 9%, double orphans for 15% and nonorphans for 49%; the remaining participants (about 2%) did not know if their parents were alive and were excluded from the study. The key educational variable took into account both attendance (whether participants were in or out of school) and attainment (whether they had completed secondary school). The researchers also collected information on possible correlates of sexual risk and orphanhood, including poverty (whether the adolescent ate fewer than three meals per day), residential mobility (whether she had lived in Highfield for less than a year) and church attendance (whether she attended church at least once per week).

Overall, 45% of married adolescents tested positive for HSV-2, HIV or both; prevalence did not vary by orphan status. However, the proportion of unmarried participants who tested positive for one or both infections varied greatly: While 6% of nonorphans tested positive for HIV or HSV-2, 20% of maternal orphans (odds ratio, 3.4), 14% of double orphans (2.3) and 11% of those who lost their father before age 12 (2.3) did so.

Sixty percent of unmarried adolescents were enrolled in school, and 93% had attended some secondary school. Double orphans had the lowest attendance of any group (47%) and were more likely than nonorphans to have left before completing secondary school, among both 14–17-year-olds (23% vs. 13%) and 18–20-year-olds (34% vs. 15%). The proportion of maternal orphans who had left before completing secondary school was elevated among 18–20-year-olds (38%) but not 14–17-year-olds (19%). Most respondents who had left school early (86%) said they had done so because they had been unable to pay for school fees, books or uniforms.

Twenty percent of unmarried adolescents reported having received assistance with school fees. Double and paternal orphans were more likely than nonorphans to have received such assistance (odds ratios, 2.8 and 2.3, respectively). Twenty-nine percent of maternal and double orphans reported that they had ever needed assistance with school fees, compared with 19% of nonorphans.

The prevalence of HIV or HSV-2 was higher among out-of-school adolescents (15–16%, depending on whether they had completed secondary school) than among those still attending secondary school (5%), even when potential confounders (age, poverty, residential mobility and church attendance) were taken into account (odds ratios, 3.2 for both types of out-of-school youth). Among 18–20-year-olds, only 2% of those in school were infected, compared with 26% of those who had left before completing secondary school (14.3). Infection with HIV or HSV-2 was not associated with poverty or residential mobility and was negatively associated with church attendance (0.3).

When the researchers adjusted for respondents' education status, the association between orphan status and sexual risk showed relatively little change: The odds of having HIV or HSV-2 increased slightly among maternal orphans (from 3.4 to 3.6) and paternal orphans (from 2.3 to 2.5), and declined by only 8% among double orphans (from 2.3 to 2.2). Potential confounders reduced the association among double orphans by another 11%, leaving most of the association between education and sexual risk unexplained.

These findings, the researchers note, suggest that "education might not play a strong role in orphans' HIV/HSV-2 infection in this particular setting," perhaps because of Zimbabwe's high rates of school completion. Alternatively, assistance with school expenses may have lessened the impact of orphanhood on education. They stress that, in the absence of a strong connection between orphans' schooling and sexual risk, or of data on whether leaving school early precedes or follows HIV or HSV-2 infection, "longitudinal research may help to understand the timing of events, like school drop-out and marriage, in relation to parent loss, and better trace the causal pathways of orphans' risk."—H. Ball


1. Birdthistle I et al., Is education the link between orphanhood and HIV/HSV-2 risk among female adolescents in urban Zimbabwe? Social Science & Medicine, 2009, 68(10):1810–1818.