Inhibition of Clostridium botulinum in Model Reduced-Sodium Pasteurized Prepared Cheese Products

The 1986 Food Research Institute-Tanaka et al. model predicts the safety of shelf-stable process cheese spread formulations using the parameters of moisture, pH, NaCl, and disodium phosphate (DSP) to inhibit toxin production by Clostridium botulinum. Although this model is very reliable for predicting safety for standard-of-identity spreads, the effects of additional factors have not been considered. The objective of this study was to create a predictive model to include the interactive effect of moisture, pH, fat, sorbic acid, and potassium-based replacements for NaCl and DSP to reflect modern reduced-sodium recipes. Eighty formulations were identified using a central composite design targeting seven factors: 50 to 60% moisture, pH 5.4 to 6.2, 0 to 0.2% sorbic acid, 10 to 30% fat, 1.7 to 2.4% NaCl, 0.8 to 1.6% DSP, and 0 to 50% potassium replacement for sodium salts. Samples were inoculated with proteolytic C. botulinum spores at 3 log spores per g, hot filled into sterile vials, and stored anaerobically at 27°C. Samples were assayed at 0, 1, 2, 3, 4, 8.5, 17.5, 26, and 40 weeks for the presence of botulinum toxin using the mouse bioassay. A parametric survival model was fit to the censored time-to-toxin data. All linear, quadratic, and pairwise effects were considered for model fit. As hypothesized, the effects of pH, sorbate, moisture, DSP, and NaCl were highly significant (P < 0.001). Fat concentration and potassium replacement effects were significant at P < 0.021 and P < 0.057, respectively. The model consistently predicted the safety failure of the toxic samples, but it also predicted failure for some samples that were not toxic. This model is an adjunct to existing models by adding the factors of potassium salts, fat, and sorbic acid to predict the botulinal safety of prepared process cheese products but is not intended to be a substitute for formulation evaluation by a competent process authority.