Male Circumcision and HIV Infection in Rural Malawi

Michelle Poulin, University of Pennsylvania
Adamson Muula, University of Malawi

Observational studies in various settings have found consistent, significant, and negative associations between male circumcision and HIV infection, and circumcision’s protective effect has been confirmed recently by randomized control trials in South Africa, Uganda, and Kenya. Yet population-based data from husbands and wives (N=2240) from the 2004 wave of the Malawi Diffusion and Ideational Change Project (MDICP) present an interesting anomaly. When comparing HIV prevalence rates across the MDICP’s three study sites in the country’s three main administrative regions, we find that the region home to the highest HIV prevalence (8.6%) is also where most husbands (80%) are circumcised. Individual-level analyses using logit/probit models within this region, southern Balaka, do show a negative relationship between circumcision and infection status, but our data further find that certain behaviors, and possibly STI status, supersede the protective effect of male circumcision. These findings are expected to be important considerations for prevention policies.