Objective: To examine the role played by mobile residents in the spread of HIV through rural Uganda.
Design: Travel history and sexual network data were collected from a random sample of 1627 residents aged 15–49 years in Rakai District, Uganda during 1994.
Methods: Travelers and non-travelers are compared with respect to socio-demographic attributes, risk exposure, knowledge, attitudes and use of condoms using descriptive statistics and multivariate logistic regression. A demographic profile of travelers' partners is developed using information from a local network survey module.
Results: The population is highly mobile, with over 70% reporting travel to a potentially higher risk destination in the past year. Travelers are somewhat more likely to have higher levels of sexual risk behavior, but the risk appears to be offset by significantly greater knowledge, acceptance, and use of condoms. In multivariate analysis, the sexual risk differential for travelers is explained by occupational exposure and higher socio-economic status. The differential in condom acceptance, by contrast, appears to be associated with travel itself. Condom use with non-spousal partners is three times higher among travelers than non-travelers (P < 0.001), and travel remains a significant predictor after controlling for age, education, residence, occupation and multiple partners. Travelers are more likely to use condoms with both their local and non-local partners. Partners of male travelers are likely to be younger and better educated than those of male non-travelers.
Conclusions: The mobile population in this rural region appears willing to adopt risk reduction measures appropriate to their exposure. This suggests that targeting condom promotion programs to travelers and their partners is likely to be effective in reducing the spatial diffusion of HIV, and may be an efficient method for spreading behavioral change into rural areas.
From the Departments of Sociology and Statistics, The Pennsylvania State University, University Park, Pennsylvania, the aCenter for Population and Family Health, Columbia University, New York, New York, USA, bDepartment of Medicine and Clinical Epidemiology Unit, Makerere Medical School, Entebbe, Uganda and the cDepartment of Sociology, University of California at Berkeley, Berkeley, California, USA.
Sponsorship: Supported by NIH grant number 1 R29HD 34957-01 and the Rockfeller Foundation.
Requests for reprints to: M. Morris, Departments of Sociology and Statistics, The Pennsylvania State University, University Park, Pennsylvania, USA 16802.
Received: 22 January 1998;
revised: 14 July 1999; accepted: 22 December 1999.