Studies have explored the use of various nonlinear regression techniques to better describe shoulder and/or tailing effects in survivor curves. Researchers have compiled and developed a number of diverse models for describing microbial inactivation and presented goodness of fit analysis to compare them. However, varying physiological states of microorganisms could affect the measured response in a particular population and add uncertainty to results from predictive models. The objective of this study was to determine if the shape and magnitude of the survivor curve are possibly the result of the physiological state, relative to growth conditions, of microbial cells at the time of heat exposure. Inactivation tests were performed using Escherichia coli strain K-12 in triplicate for three growth conditions: statically grown cells, chemostat-grown cells, and chemostat-grown cells with buffered (pH 6.5) feed media. Chemostat cells were significantly less heat resistant than the static or buffered chemostat cells at 58°C. Regression analysis was performed using the GInaFiT freeware tool for Microsoft Excel. A nonlinear Weibull model, capable of fitting tailing effects, was effective for describing both the static and buffered chemostat cells. The log-linear response best described inactivation of the nonbuffered chemostat cells. Results showed differences in the inactivation response of microbial cells depending on their physiological state. The use of any model should take into consideration the proper use of regression tools for accuracy and include a comprehensive understanding of the growth and inactivation conditions used to generate thermal inactivation data.
All Science Journal Classification (ASJC) codes
- Food Science