Mitigation of enteric foodborne pathogens in raw and minimally processed pet food using food additives

Abstract

Raw meat-based diets (RMBDs) are minimally processed pet foods that are gaining in popularity, but they pose a public health risk because they can harbor enteric foodborne pathogens such as Salmonella enterica. They consist mainly of raw/uncooked animal sourced proteins (poultry, beef, veal, fish) which can be supplemented with cereals or tubers. However, these diets cannot be heat-processed, fermented, rendered, or extracted by their definition, and thus are potential vectors for the transmission of enteric foodborne pathogens. This study investigated the antimicrobial efficacy of encapsulated and raw glucono delta lactone (GDL), citric acid (CA), and lactic acid (LA) when challenged against S. enterica in a model RMBD for dogs. The results showed that both encapsulated and raw CA and LA had higher log reductions of S. enterica compared to GDL. Encapsulated CA and LA at 1.0% (w/w) exhibited higher (p<0.05) log reductions and preserved product quality compared to the raw acidulants (CA and LA) at 1.0%. Increasing the concentration of acidulants from 2.0 to 3.0% increased (p<0.05) the log reduction of the pathogens significantly, but resulted in discoloration, weeping, and syneresis of the RMBDs. In a subsequent study, combinations of encapsulated and raw acidulants were evaluated for their antimicrobial efficacy against Salmonella enterica. The results showed that combinations of raw and encapsulated acidulants resulted in significantly lower counts of S. enterica compared to when lactic and citric acids were used alone. Encapsulated acidulants (LE+CE) had a higher inactivation rate of the pathogen than when (LR+CR) was used as a treatment. The rate of inactivation of S. enterica increased over time. There was a significant interaction (p<0.05) between time and treatment in the rate of inactivation of S. enterica. The prediction models revealed a significant increase (p<0.05) in the rate of pathogen inactivation when inoculum levels increased from 6 to 9 log CFU/mL. There was a significant interaction between inoculum level and treatments on the rate of inactivation of the pathogens in the predictive models (p<0.05). Overall, the Weibull model performed better than the log linear models. The Weibull models were appropriate for modelling the survival kinetics of S. enterica in RMBDs treated with combinations of raw and encapsulated acidulants. In conclusion, combinations of encapsulated and raw acidulants can be used to pasteurize RMBDs and predictive models are useful microbial food safety tools in the prediction of the survival and inactivation of spoilage and pathogenic bacteria in RMBDs.