In a cluster randomized trial that assessed a cash transfer program in Malawi, young women who received monthly payments were less likely than others to become infected with HIV or herpes simplex virus type 2 (HSV-2)—but only if they had been attending school at the start of the trial.1 Among young women who had been enrolled in school at baseline, the prevalence of HIV at the 18-month follow-up was lower among those receiving payments than among those in an unpaid control group (1.2% vs. 3.0%); findings were similar for infection with HSV-2 (0.7% vs. 3.0%). However, among participants who had dropped out of school by baseline, the intervention and control groups had similar levels of HIV (8–10%) and HSV-2 (8% each) at follow-up.

The prevalence of HIV in Malawi is 9% among female and 2% among male 15– 24-year-olds. Prevalence is especially high in Zomba district (site of the current study), where rates of poverty are substantial and school enrollment is low. Because lack of education and economic dependence on men are believed to put women at increased risk for HIV infection, many countries have implemented cash transfer programs in an attempt to reduce poverty and promote education. However, to date, no randomized controlled studies have found a causal relationship between increased schooling or income levels and reduction in HIV risk.

In this study, investigators randomly selected households from 176 enumeration areas in urban and rural areas of Zomba, and interviewed 3,796 never-married females aged 13–22 who were enrolled in school or had dropped out; 1,706 of these young women underwent STI testing and were included in the current analysis. During the two-year trial, which began in 2008, young women in the intervention group received monthly payments of US$1–5, and their household received an additional US$4–10 per month. Payments to participants who were enrolled in school at baseline were either conditional (requiring at least 80% school attendance in the previous month) or unconditional (no attendance required); payments to participants who had dropped out of school at baseline were always conditional (i.e., dropouts had to resume their schooling and meet the attendance requirement). Young women assigned to the control group received no payments. To reduce possible interactions between women in the intervention group and those in the control group, all participants from a given area were assigned to the same treatment arm. Behavioral outcomes were assessed at baseline and 12 months, and infection with HIV or HSV-2 was determined at 18 months. Logistic regression was used to identify associations between outcomes and the different trial groups.

Baseline characteristics of the intervention and control groups were similar within each of the enrolled and dropout cohorts. Among females enrolled in school, the average age was 15, some 19–22% were sexually experienced and 3% had ever been pregnant. Young women who had dropped out of school by baseline had an average age of 17–18; two-thirds (68%) were sexually experienced and two-fifths (40–44%) had been pregnant.

Among participants who had been enrolled in school at baseline, the weighted HIV prevalence at the 18-month follow-up was 1.2% in the intervention group and 3.0% in the control group; in a regression analysis that adjusted for background characteristics, young women who had received cash transfers had reduced odds of being HIV-infected (odds ratio, 0.4). The prevalence of HSV-2 at follow-up was 0.7% and 3.0%, respectively, in the intervention and control groups, and those in the intervention group again had reduced odds of testing positive (0.2). At the 12-month follow-up, intervention participants were less likely than control participants to report having had intercourse at least once a week (3% vs. 7%) or a sexual partner who was 25 or older (0.5% vs. 3%) in the past year; in multivariate analyses, the odds of these risk behaviors were reduced among young women who had received cash transfers (0.5 and 0.2, respectively). Finally, a higher proportion of payment recipients than of controls remained enrolled in school during the first year of the intervention (90% vs. 84%).

Among young women who had dropped out of school at baseline, the cash transfer intervention was not effective in reducing STI levels: The intervention and control groups had similar levels of HIV (8–10%) and HSV-2 (8% each) at the 18-month follow-up. However, these intervention participants were far more likely than their control counterparts to have enrolled in school during the first year of the trial (57% vs. 12%; odds ratio, 8.8), and they were less likely to have married (17% vs. 29%; odds ratio, 0.5). As in the analysis of nondropouts, a smaller proportion of intervention participants than of control participants reported having had intercourse at least once a week (19% vs. 30%).

In general, among participants who were enrolled at baseline, outcomes in the conditional cash transfer group did not differ from those in the unconditional cash transfer group; the only exception was pregnancy, which was more common in the conditional group (4% vs. 1%). In regression analyses, young women in the conditional-payment group were more likely than those in the control group to have enrolled in school (odds ratio, 2.1), and less likely to be HIV-positive (0.3) or to have had a partner who was 25 or older (0.1). Participants who received unconditional payments, meanwhile, were less likely than control participants to be pregnant, to report having had intercourse at least once per week and to be infected with HSV-2 (0.1–0.4). Notably, the amount of the cash transfers did not affect the various outcomes.

The investigators note several study limitations, including the absence of baseline data on the prevalence of HIV and HSV-2 among study participants, and the short, two-year intervention period. Yet they believe that this simple intervention showed promise in the effort to reduce rates of risky sexual activity and, in turn, the likelihood of infection with either of these viruses. On the basis of their findings, the authors "suggest that financially empowering school-aged girls might have beneficial effects on their sexual and reproductive health [and] that cash transfer programmes could be attractive to policy makers in Sub-Saharan Africa."—J. Thomas

REFERENCE

1. Baird SJ et al., Effect of a cash transfer programme for schooling on prevalence of HIV and herpes simplex type 2 in Malawi: a cluster randomised trial, Lancet, 379(9823):1320–1329.