Meta-Regression models describing the effects of essential oils and added lactic acid bacteria on pathogen inactivation in cheese

Biopreservatives such as plant-based antimicrobials and bacteriocinogenic starter cultures have been proposed as hurdles to increase microbiological safety of a variety of products, including cheese, and numerous studies have reported their pathogen inhibitory properties. For that reason, the objective of this meta-analysis was to summarise the inactivation of Listeria monocytogenes (LM), Staphylococcus aureus (SA) and Salmonella spp. (SS) in cheese attained by added lactic acid bacteria (LAB) and essential oils (EOs); and to compare the inhibitory effectiveness by application mode and specific antimicrobial. After systematic review, 1810 observations on log reduction data and study characteristics were extracted from 53 studies. Comparing among the factual methods of application of antimicrobials (in milk, cheese surface and incorporated in films), meta-regression models pointed out that addition of EOs to milk renders, as a whole, the lowest inhibitory effect against LM, SA and SS in the finished product; whereas for added LAB, incorporation in milk prompts a faster inactivation of LM than onto cheese surface. Lemon balm, sage and basil EOs showed the best inhibitory outcomes against LM and SA; whereas clove, oregano and bay EOs presented the highest bactericidal effect against SS. For a given increase in EO concentration, the application on cheese surface provides the greatest inhibitory effect against LM and SS, while EO-embedded films lead to a more rapid inactivation during maturation/storage. The experimental practice of inoculating the antimicrobial in cheese mixture should no longer be employed in challenge studies, since the meta-regression models have demonstrated that this application method biases the results, overestimating or underestimating the inhibitory effects of EOs or added LAB, respectively. This meta-analysis has also emphasised the need to further investigate the relationship between pathogen's inoculum size and their concentrations in time.