Abstract
In Indonesia, BCG vaccine protection against Mycobacterium tuberculosis infection decreased with increasing exposure to the pathogen. We aimed to validate these findings in Africa. Poisson regression was used to estimate BCG protection, stratified by pathogen exposure using an exposure score, against enzyme-linked immunospot assay conversion at 3 months in 220 Gambian case contacts. Although the interaction between BCG and exposure was not significant (P = .13), BCG protection was strongest in the lowest-exposure tertile (relative risk, 0.35 [95% confidence interval, .15–.82; P = .02] vs 0.50 [.30–.83; P = .008] and 0.71 (.45–1.13; P = .1] for the middle and highest-exposure tertiles, respectively. These results are consistent with those from Indonesia.