Models for the Inactivation of Foodborne Pathogens in Salad Dressing from Challenge Studies

Donald W Schaffner,W Clifton Baldwin
DOI: https://doi.org/10.1016/j.jfp.2024.100384
Abstract:The Association for Dressings and Sauces' (ADS) members have conducted challenge studies on salad dressing products to assess pathogen survival. Data from 79 different challenge studies provided by ADS were used in this analysis. The acid-moisture ratio, pH, incubation temperature, and ingredient details were provided for each study. Linear regression models were used to predict the time to 3-log, 4-log, and 5-log reduction as a function of study parameters. A statistically based approach also was used to estimate the concentration of pathogens in ingredients based on testing history. This was combined with decline modeling to estimate pathogen concentration over time. The time-to-five log reduction for each of the target pathogens were highly skewed. A logarithmic transformation of time to 5 log reduction resulted in approximately normal distributions. Incubation temperature and formulation pH were highly significant (p < 1E-6), in predicting the number of days to a five-log reduction of Escherichia coli O157:H7, while the percentage of spices in the formulation is also quite significant (p = 0.01). Salmonella modeling showed that the most highly significant parameter was the percentage of water (p < 1E-8). Other parameters in order of descending significance include the percent fruit (p = 0.00032), incubation temperature (p = 0.00268), followed by percent sugar (p = 0.02161) and percent vegetables (p = 0.03149). The most significant parameter in predicting Listeria monocytogenes reduction was incubation temperature (p = 0.000687), followed by acid moisture ratio (p = 0.012423). The next two significant parameters in the Listeria model were percent lipid (p = 0.023772) and percent water (p = 0.025701). The least significant parameter that meets the minimum criteria for inclusion in the Listeria model (p < 0.05) was percent fruit (p = 0.047074). Our analysis will be useful in developing risk-based approaches to continue to assure the safety of commercially prepared salad dressings.
What problem does this paper attempt to address?