Comparison of Probabilistic and Deterministic Predictions of Time to Growth of Listeria monocytogenes as Affected by pH and Temperature in Food

Abstract

Stochastic models are useful for estimating the risk of foodborne illness and they can be integrated, besides other sources of variability, into microbial risk assessment. A stochastic approach to evaluate growth of two strains of Listeria monocytogenes influenced by different factors affecting microbial growth (pH and storage temperature) was performed. An individual-based approach of growth through optical density measurements was used. From results obtained, histograms of the lag phase were generated and distributions were fitted. Histograms presented increased variation when the factors applied were suboptimal for L. monocytogenes and they were combined. The extreme value distribution was ranked as the best one in most cases, whereas normal was the poorest fitting distribution. To evaluate the influence of pH and storage temperature on L. monocytogenes CECT 5672 in real food, commercial samples of courgette and carrot soup were inoculated with this pathogen. It was able to grow in both soups at storage temperatures from 4°C to 20°C. Using the distributions adjusted, predictions of time to growth (102 cfu/g) of L. monocytogenes were established by Monte Carlo simulation and they were compared with deterministic predictions and observations in foods.