OBJECTIVES

Migration is associated with HIV infection, but the relationship has been mostly assessed in cross-sectional studies. In a prospective study, we investigated whether out-migrants are a selection of high-risk individuals and whether rural-to-urban migration results in risky sexual behaviour for HIV incidence.

METHODS

A population cohort was enrolled in a stratified household census in four different community types in Manicaland, east Zimbabwe, between July 1998 and February 2000, and followed-up after 3 years. Out-migrants to the national capital (Harare), the provincial capital (Mutare) and other study areas were followed-up. A structured questionnaire was administered and an HIV test was conducted at each interview. HIV prevalence and sexual risk behaviour at baseline, and HIV incidence and sexual behaviour during follow-up were compared for out-migrants and residents.

RESULTS

At baseline, future migrants were significantly younger, better educated and more likely to be single than residents. For males, migration was highest from subsistence farming areas and roadside trading centres and lowest from estates. After adjusting for age, education, marital status and location, there were no differences in HIV prevalence and sexual risk behaviour between future migrants and residents at baseline, for either sex. No significant differences in HIV incidence or sexual behaviour during follow-up were detected between rural-to-urban out-migrants and residents.

CONCLUSIONS

Out-migrants from rural Zimbabwe did not have higher levels of HIV infection or sexual risk behaviour than residents either before or after they moved. These findings may be related to the mature stage of the HIV epidemic and the social and living conditions of migrants in Zimbabwean cities.