Evaluation of Water Content as a Convenient Metric in Thermal Inactivation Modeling for Low-moisture Foods

Introduction: FSMA Preventive Controls Rules require validation of pathogen reduction steps in the food industry, which increases the importance of microbial modeling for process validations. Water activity (a_w) has been the most commonly used metric when evaluating and modeling the effect of water on Salmonella inactivation in low-moisture foods. However, because of the nature of these products and processes, as well as the definition of a_w (a function of temperature), it may not be the most suitable metric for real-world applications.

Purpose: The objective was to quantitatively compare the correlation of moisture content (%mc) and a_w with D-values, to evaluate the utility of %mc as a metric in pathogen inactivation models and industrially-relevant process validation protocols.

Methods: D_80°C values for Salmonella Enteritidis PT30 were calculated by linear regression of isothermal inactivation data from multiple related studies in our laboratory (wheat flour, almonds, dates; 0.25, 0.45, 0.65 a_w; 70-90°C, triplicate). Water activity and %mc were measured and/or calculated from moisture isotherms (sorption and desorption, accordingly). Linearity of inactivation curves was confirmed, and correlation coefficients were estimated between logD and a_w and %mc.

Results: The D_80°C values for the different products exhibited a log-linear trend with a_w, as well as with %mc. The correlation coefficients varied less than 5% when comparing a_w and %mc vs. logD (e.g., -0.96 and -0.95, respectively, for wheat flour). The results suggest that %mc may be a suitable, or even preferable, metric for the effect of water on inactivation process.

Significance: When comparing the utility of a_w vs. %mc for inactivation modeling or process validation, %mc has the advantage of being measurable (potentially real-time in dynamic processes), and this study shows a consistent correlation with logD. This is critically important to both monitoring and modeling inactivation processes in low-moisture foods.