Brucellosis risk in urban and agro-pastoral areas in Tanzania

Epidemiology of human and animal brucellosis may depend on ecological conditions. A cross-sectional study was conducted to compare prevalence and risk factors of bovine brucellosis, and risky behaviours for the human infection between urban and agro-pastoral areas in Morogoro region, Tanzania. Cattle blood sampling and interviews using a structured questionnaire were conducted with farmers. Rose-Bengal test was conducted for the cattle sera, and positive samples were confirmed with competitive ELISA. Farm-level sero-prevalences were 0.9% (1/106, 95% CI 0.0–5.9%) and 52.9% (9/17, 95% CI 28.5–76.1%) in urban and agro-pastoral areas, respectively. The animal-level-adjusted prevalences were 0.2% (1/667, 95% CI 0.0–1.1%) and 7.0% (28/673, 95% CI 5.7–8.4%) in those areas. The final farm-level model including both areas found two risk factors: history of abortion in the herd (P < 0.01) and cattle grazing (P = 0.07). The animal-level risk factors in agro-pastoral areas were age (P = 0.04) and history of abortion (P = 0.03). No agro-pastoral farmer knew about Brucella vaccine. Agro-pastoralists generally had poorer knowledge on brucellosis and practiced significantly more risky behaviours for human brucellosis such as drinking raw milk (17.6%, P < 0.01) and blood (35.3%, P < 0.01), and helping cattle birth (100%, P = 0.04) than urban farmers (0, 0 and 79.2%, respectively). Intervention programs through education including both human and animal health particularly targeting agro-pastoralists would be needed