BACKGROUND: One hundred years ago, Albert Calmette developed an avirulent strain of Mycobacterium bovis, but there is no evidence that his BCG strain was more immunogenic than wild-type M. bovis. Geographic variations in BCG efficacy remain ill-understood. We hypothesized that exposure to M. bovis through unpasteurized milk might protect against Mycobacterium tuberculosis and Mycobacterium leprae.
METHODS: After excluding high-income countries (with universal milk pasteurization) and microstates, an ecological study comprising 113 countries was conducted. National data were obtained from United Nations agencies and international organizations about milk production per capita (1980-1999) as a proxy for exposure to wild-type M. bovis, TB (2000-2019) and leprosy (2005-2019) incidence, HIV prevalence (2000-2019), human development index (2010), global hunger index (2010), neonatal BCG coverage (1980-1999), urbanization (2000) and temperature (1990-2020). Multiple linear regression analyses were performed using log-transformed variables.
RESULTS: For TB, the association differed by region. An inverse association with milk production was seen in regions outside, but not within, sub-Saharan Africa, after adjustment for confounders. The incidence of leprosy was inversely associated with milk production when combining all countries, but the association was stronger in sub-Saharan Africa.
CONCLUSIONS: Exposure to wild-type M. bovis through unpasteurized milk may provide cross-protection against M. tuberculosis and M. leprae and contribute to geographic disparities in BCG efficacy. This needs to be confirmed by individual-level studies.
a1878-3503