|
| |
Food microbiology is the study of both beneficial and pathogenic microorganisms in raw and processed foods. One area of food microbiology, predictive microbiology, uses mathematical models to define the growth kinetics of food microorganisms and predict microbial behavior over a range of conditions.
Predictive microbiology is used to assess the risks of food processing, distribution, storage and food handling; and, to implement control measures in order to protect the microbiological quality of foods, important for both food safety and product quality. | |
Four Phases of Bacterial Growth | | Microbial load in a food source depends upon the initial level of bacterial contamination as well as environmental conditions (temperature, pH, water activity, preservatives, antimicrobials and the composition of the atmosphere) which influence growth, inactivation and survival in the food. Predictive models have been made for each scenario of bacterial behavior in food, but more models have been developed for growth than for inactivation or survival.
Bacterial growth can exhibit at least four different phases: lag phase, growth phase, stationary phase and death phase.
Lag Phase
During the lag phase, cells increase in size but not in number because they are adapting to a new environment, and, synthesis and repair are taking place. The length of the lag phase depends on the current environment as well as the previous physiological state of the cells. Cells that are from a very different environment or are damaged from their previous physiological state may require more time to adjust. In some foods a lag phase does not exist which results in cells that are ready for immediate growth.
Growth Phase
During the growth phase, cells grow exponentially and at a constant rate. The maximum slope of the curve is the specific growth rate of the organism. Cell growth is dependent upon the current environment (nutrients, temperature, pH, etc.), but is not dependent upon the previous physiological state. In the field of predictive microbiology, growth rate is commonly expressed as the change in cell number per time interval.
Stationary Phase
The stationary phase occurs at the maximum population density, the point at which the maximum number of bacterial cells can exist in an environment. This typically represents the carrying capacity of the environment. However, environmental factors such as pH, preservatives, antimicrobials, native microflora and atmospheric composition as well as depletion of growth-limiting nutrients can affect the maximum population density.
Death Phase
The death phase occurs when cells are being inactivated or killed because conditions no longer support growth or survival. Some environmental factors such as temperature can cause acute inactivation. Others may cause mild inactivation as with growth in the presence of organic acids.
Predictive microbiology has primarily concentrated on the growth phase because estimating the growth rate of pathogens is important to risk assessment and quality control; however, each phase is important to understanding bacterial growth. For example, variations in lag phase can result in a very different growth curve even though the growth rates may be identical. The lag and growth phases are more relevant to modeling pathogen growth in foods because food spoilage occurs before the stationary and death phases are reached. However, the stationary and death phases give additional information about bacterial growth and inactivation in relationship to environmental factors encountered in foods.
Back to top | |
Development of Predictive Models | |
Developing a useful model begins with a good experimental design. The following variables must be considered when planning a predictive microbiology experiment:
Bacterial Strain(s)
Single strain models are strain specific and may not represent a worse-case scenario of bacterial growth for a food pathogen or within a food source. On the other hand, a multiple strain cocktail allows for the most negative outcome to be determined because the strain with the highest growth rate will predominate. The strain source is important because pathogens is lated directly from foods are preferred.
Test Matrix
Microorganisms sometimes grow differently in food matrices than in microbiological media. Experiments planned for model development should use food as the growth medium. In the past, the majority of published microbial models used microbiological media. Today researchers can compare models using food as the growth medium to past models to learn about the food matrix and how it affects pathogen growth.
Inoculum Preparation
Typically cells developed in growth media and taken from late growth phase are used to inoculate the test matrix. What is known about the previous environment for the inoculum should be a consideration in experimental design, and inoculum preparation should be consistent from experiment to experiment. Prior environments that are relevant to food safety issues are not commonly used in experiments though this could be a factor in relating models to real-world food situations.
Environmental Conditions
The experimental design should consider what environmental conditions are relevant to the food source of interest. For each condition, it is necessary to define the potential range to be encountered in the food so that the model will be able to provide predictions within that range of values. It is crucial that experiments cover several test values over the array of values defined. Growth of the inoculum in environments more pertinent to the food source is essential to detect environmental effects on pathogen growth.
Microbial Flora
The presence or absence of native microflora (spoilage organisms) is an important characteristic of the test matrix. Numerous reports document competition between added pathogen and native microflora in retail foods and a resulting decrease in maximum population density for the inhibited strain.
Back to top | |
Measures of Model Performance | | Models must undergo validation before they are used to aid in food safety decisions. Validation involves comparing model predictions to experimental observations not used in model development. Two primary tools for measuring model performance are the bias and accuracy factor. Both of these do not identify where prediction errors exist in a model.
Bias Factor
Bias factor is a multiplicative factor that compares model predictions and is used to determine whether the model over- or under-predicts the response time of bacterial growth. A bias factor greater than 1.0 indicates that a growth model is fail-dangerous. Conversely, a bias factor less than 1.0 generally indicates that a growth model is fail-safe. Perfect agreement between predictions and observations would lead to a bias factor of 1.0.
Accuracy Factor
It is the sum of absolute differences between predictions and observations, and it measures the overall model error.
Several benefits of mathematical models to predict pathogen growth, survival and inactivation in foods are: ability to account for changes in microbial load in food as a result of environment and handling; use of predictive microbiology in management of foodborne hazards; and, preparation of Hazard Analysis Critical Control Point (HACCP) plans.
Back to top | |
USDA/ARS Predictive Microbiology Research | |
Programs to develop models for the growth of foodborne microbial pathogens have been initiated in the United States, the United Kingdom, and other nations. These programs have resulted in a range of models that have been incorporated into software packages available online.
For more than 15 years, the USDA ARS Eastern Regional Research Center (ERRC) in Wyndmoor, Pennsylvania has worked on the development of mathematical models for foodborne bacterial pathogens. The USDA ARS Pathogen Modeling Program (PMP) is a software program developed in this laboratory which is utilized by the food industry to estimate pathogen behavior in food. Each year, the PMP is downloaded from the Internet by thousands of users in world-wide. Users are able to input specific information about their food products, and receive predictions of pathogen behavior via easy-to-read graphical output.
Research objectives in the Pennsylvania laboratory include:
- Validating predictions of the PMP in raw and processed foods
- Developing models of pathogen behavior in the presence of naturally-occurring microflora found in foods
- Developing models of pathogen behavior over a series of food processing operations
- Modeling the virulence of pathogens during growth and survival in foods
Other ARS research objectives include:
- Model farm-to-table foodborne pathogen transmission
- Model fluctuating microbial populations and their aperiodic outbreaks
- Control foodborne disease agents to enhance food safety
- Create computer integrated food manufacturing systems
- Develop predictive models to improve the microbiological safety of raw and processed meat
- Model pathogen growth, survival, and inactivation in ground meat and ready-to-eat products
- Achieve lethality performance standards for fully-cooked meat products
- Improve processing and packaging techniques for snack and other processed foods
- Model and validate continuous flow microwave heating of liquid foods
- Improve food safety through more realistic models of spore germination
- Create a relational database of predictive microbiology information
Back to top | |
Predictive Microbiology Research in Foreign Countries | | The Computational Microbiology Team at the Institute of Food Research (IFR) in the United Kingdom has several laboratories involved in predictive microbiology research. The institute has developed a web-based tool, ComBase, that is maintained by the ComBase Consortium, a collaboration between the Food Standards Agency and the IFR, UK, and the USDA ARS ERRC. ComBase is a relational database with thousands of predictive microbiology data sets from across the world used in model development and validation.
Current research interests and projects at IFR include:
- Dynamic modeling of bacterial growth as a function of the environment
- Stochastic modeling of the kinetics of individual cells
- Probability of growth and growth domain of pathogens
- Predictive microbiology tools for quantitative microbial risk assessment
The Danish Institute for Fisheries Research Microbiology Group is involved in the study of seafood spoilage and predictive microbiology. Since spoilage of seafood is most often due to microbial activity, this laboratory has developed the Seafood Spoilage Predictor (SSP) software package to predict the shelf-life of seafood. The software uses research data from multiple temperatures to produce dynamic models of seafood spoilage. These dynamic models can be used to:
- Determine the shelf-life of seafood by establishing the microbiological and chemical indices of spoilage
- Predict the shelf-life of seafood by developing microbial spoilage models
- Extend the shelf-life of seafood by targeting inhibition of selected spoilage organisms
The Bacanova Project is funded by the European Commission Frame V Programme. The purpose of the project is to develop a method based on stochastic mathematical modeling of lag times of individual cells to improve the microbial safety and quality of food. The project has three main research objectives:
- Optimize the effect of processing methods with respect to microbial food safety and quality
- Predict the probability of bacterial survival, lag and growth in food
- Utilize the information on the variability of individual cells to develop predictive microbiology models
The Centre for Food Safety and Quality at the University of Tasmania, Australia is involved in predictive microbiology research such as the physiology of temperature, osmotic stress and pH stress on foodborne pathogens. In 2001, the Centre conducted the following research projects:
- Development of predictive models for inactivation of bacterial pathogens in meat and meat products
- Development of predictive models for inactivation of Escherichia coli in uncooked comminuted fermented meat
- Development of predictive models for microbiological changes during food chilling
Back to top | |
Glossary |
- accuracy factor: The sum of absolute differences between predictions made by a model and observations found during an experiment conducted for validation of the model.
- antimicrobial: An agent that kills microorganisms or inhibits their growth.
- bacterial strain: Population of bacterial cells all descended from a single pure isolate.
- bias factor: A measurement of whether a model gives an over- or under-prediction.
- death phase: The decrease in viable microorganisms that occurs when conditions no longer support growth or survival.
- exponential growth: A period of sustained growth of a microorganism in which the cell number constantly doubles within a fixed time period.
- food microbiology: The study of microorganisms in food.
- growth curve: A graph displaying the behavior of a bacterial population over time.
- growth rate: The change in bacterial numbers over time, typically expressed as the change in cell number per time interval.
- Hazard Analysis Critical Control Point (HACCP): A system used to identify, evaluate and control hazards by identifying potentially unsafe links in the food processing chain.
- inoculum: A medium containing microorganisms to be introduced into fresh media or food source in an experiment.
- lag phase: The period of time prior to microbial growth in which there is no growth because cells are adapting to a new environment.
- mathematical model: Simplified mathematical expressions for describing the numerous processes that affect bacterial growth in foods.
- maximum population density: The point at which the maximum number of bacterial cells can exist in an environment.
- microbial load: Total number of living microorganisms in a given volume or mass of microbiological media or food.
- microorganism: A living organism too small to be seen with the naked eye.
- native microflora: The microorganisms that are normally found within a food source (often referred to as spoilage organisms).
- pathogen: A disease-causing microorganism.
- predictive microbiology: An area of food microbiology that uses mathematical models to define growth kinetics of microorganisms in food.
- primary model: A model that describes changes in microbial numbers in response to time.
- secondary model: A model that predicts changes in primary model parameters based on environmental conditions.
- spoilage organisms: Microorganisms naturally found within a food source that cause food spoilage.
- stationary phase: The phase of microbial behavior where bacterial population growth reaches maximum carrying capacity, and growth begins to slow and eventually cease.
- tertiary model: Computer software routines that turn the primary and secondary models into "user-friendly" programs in the forms of application software.
- validation: The determination of the degree of authenticity of a model.
Back to top | |
Resources |
- The Centre for Food Safety and Quality
University of Tasmania
- Encyclopedia of Food Microbiology: Predictive Microbiology and Food Safety
2000. Editors: Richard K. Robinson, Carl A. Batt and Pradip D. Patel. Publisher: Academic Press.
- Modeling Microbial Responses in Foods
2003. Editors: Robin C. McKellar and Xuewen Lu. Publisher: CRC Press.
- Predictive Microbiology - Quantitative Microbial Ecology (PDF Format)
Institute of Food Research, UK. March 2004.
- Research Activities of the Microbiology Group
Danish Institute of Fisheries Research
- Understanding the Microbiology of Safe, Minimally Processed Food
Institute of Food Research, UK
Back to top |
|
