BACKGROUND: The reasons underpinning the large differences in the prevalence of resistance to macrolides in Streptococcus pneumoniae are imperfectly understood. We assessed if the volume of macrolides used in food-animals could play a role.
METHODS: Logistic regression was used to assess if the country-level prevalence of pneumococcal macrolide resistance was associated with country-level macrolide consumption in food animals and humans.
RESULTS: In both univariate and multivariate models, macrolide use in food-producing animals was significantly associated with pneumococcal macrolide resistance (coeff. = 339, 95% CI 21 - 658; P = 0.037).
CONCLUSIONS: In vitro and individual-level studies are required to verify or refute the hypothesis that macrolides used in food animals can promote the spread of macrolide resistance in humans.