ABSTRACT

Recent Staphylococcus aureus outbreaks linked to meat and poultry products underscore the importance of understanding the growth kinetics of S. aureus in these products at different temperatures. Raw pork, ham, and sausage (each 10 ± 0.3 g) were inoculated with a three-strain cocktail of S. aureus, resulting in an initial level of ca. 3 log CFU/g. Samples were stored isothermally at 10, 15, 20, 25, 30, 35, and 40°C, and S. aureus was enumerated at appropriate time intervals. The square root model was developed using experimental data collected from S. aureus grown on all samples (where data from raw pork, ham, and sausage were combined) so as to describe the growth rate of S. aureus as a function of temperature. The model was then compared with models for S. aureus growth on each individual sample in the experiments (raw pork, ham, or sausage) and the S. aureus ComBase models, as well as models for the growth of different types of pathogens (S. aureus, Escherichia coli O157:H7, Clostridium perfringens, Salmonella serovars, and Salmonella Typhimurium) on various types of meat and poultry products. The results show that the S. aureus model developed here based on the pooled data from all three pork products seems suitable for the prediction of S. aureus growth on different pork products under isothermal conditions from 10 to 25°C, as well as for S. aureus growth on different meat and poultry products at higher temperatures between 20 and 35°C. Regardless of some high deviations observed at temperatures between 25 and 40°C, the developed model still seems suitable to predict the growth of other pathogens on different types of meat and poultry products over the temperature ranges used here, especially for E. coli O157:H7 and Salmonella Typhimurium. The developed model, therefore, may be useful for estimating the effects of storage temperature on the behavior of pathogens in different meat and poultry products and for microbial risk assessments evaluating meat safety.

Pork is the most highly consumed meat in the world (8). In the European Union, people eat more pork than any other meat (15). In China, the consumption of pork has also continued to increase with economic development. Reports from 1989 to 1993 show that the daily consumption of pork increased from 0.18 to 0.21 pounds per person (9). According to another report, from August 2013 to August 2014, the United States exported over 5.5 million pounds of pork, making the U.S. swine industry an important form of trade (21). For these reasons, ensuring the safety of the pork supply chain is crucial.

Although pork has been less associated with foodborne illness than other meat sources, it remains of significant concern due to the large consumption (8). Salmonella serovars and Staphylococcus aureus are among the top pathogens causing foodborne illness and death annually and have been well documented to be present in pigs or pork products. These facts make pork a potential contributor to foodborne illness (1). S. aureus contamination itself is typically associated with the handling of meat products after processing. Moreover, the S. aureus strains causing foodborne illness are of human origin, not from the raw meat or livestock (12). For instance, in 2005, an outbreak of S. aureus in southeast Kansas occurred at a catered event. Smoked sausage was implicated as the source of infection, which was likely the result of contaminated equipment or humans, combined with improper cooling, reheating, or holding of the product (10).

If contaminated meat or poultry or their products are stored at elevated temperatures, pathogens can grow to reach high concentrations within 24 h, without overt visual signs of spoilage. In recent years, there have been S. aureus outbreaks linked to meat and poultry dishes in the United States (3), and several other foodborne diseases and illness outbreaks related to meat and poultry products. For instance, the consumption of contaminated meat, poultry, and associated products accounted for 12.2 to 18.3% of salmonellosis cases in 2004 and 2005. Meat was also found to be a substantial (11.2 to 25.0%) source of bacterial toxins produced by microorganisms like Clostridium perfringens (15). Furthermore, the pathogen-commodity pairs most commonly responsible for outbreaks were Salmonella and poultry (145 outbreaks), C. perfringens and poultry (3,452 illnesses), and C. perfringens and beef (2,963 illnesses) (6). Modeling the growth of various pathogens that can influence or contribute to the risk of multiple types of meats and meat products is therefore necessary.

Various growth models of S. aureus, Salmonella serovars, Escherichia coli O157:H7, and C. perfringens on meat, poultry, and their products have been developed in recent years (4, 5, 11, 13, 14, 16, 19). All of these models have delivered good predictions compared with observations made under isothermal conditions, but they were developed and validated for use in just one type of meat or meat product. In addition, to date, predictive models developed using microbiological growth data generated in one type of meat or meat product are not yet applicable to estimate bacterial growth in different types of meats and meat products. It is therefore important to study the relationship between the S. aureus growth model developed in the current study and the published growth models established for other pathogens in various types of meats and meat products so as to investigate the possibility that the S. aureus model developed here may be applicable for the prediction of different pathogens grown in various types of meats and meat products.

The purpose of this study, therefore, was to develop a mathematical model capable of predicting the growth of S. aureus on raw pork, ham, and sausage as a function of temperature and to associate the observed growth rates with those generated from ComBase. The model was also compared with existing growth models for S. aureus, Salmonella, E. coli O157:H7, and C. perfringens on meat, poultry, and their products (4, 5, 11, 13, 14, 16, 19).

MATERIALS AND METHODS

Culture preparation.

Three strains of S. aureus (ATCC 12598, ATCC 12600, and ATCC 25923) were obtained from the Korean National Institute of Health (Seoul, Republic of Korea). Prior to use, each strain was grown separately in tryptic soy broth (TSB; Difco, BD, Sparks, MD) at 37°C with two consecutive transfers after a 24-h period for a total 48 h of incubation. All working cultures grown in TSB were centrifuged separately at 4,000 × g for 10 min at 4°C, and the supernatants were discarded. The cell pellets were washed twice with 0.1% sterile buffered peptone water (Difco), pH 7.1, and resuspended in 10 ml of the same solution to obtain a final cell concentration of ca. 5 log CFU/ml. The three strains were then combined to make a cocktail with approximately equal numbers in the final population (ca. 5 log CFU/ml). The bacterial population in each culture cocktail was confirmed by plating 0.1-ml portions of appropriately diluted culture on tryptic soy agar (Difco) and incubating them at 37°C for 24 h.

Sample preparation.

Boneless pork loin, fresh ham, and sausage were purchased from a local supermarket (Lotte Mart, Chuncheon, Korea) and transported to the laboratory under temperature-controlled conditions in a sterile plastic container. Each sample (raw pork, ham, or sausage) was then separately cut into pieces using a sterile knife. Aliquots with a weight of 10 ± 0.3 g were used for the storage tests.

Inoculation and storage.

Each sample (10 ± 0.3 g) was spot inoculated by pipetting 0.1 ml of each culture cocktail (ca. 5 log CFU/ml) onto the surface to obtain an initial level of ca. 3 log CFU/g. The inoculated samples were then air dried in the laminar flow hood for 1 h at 23 ± 2°C to allow attachment of the bacteria. After drying, the samples were separately transferred into sterile stomacher bags (Whirl-Pak, Nasco, Janesville, WI) and stored isothermally at one of seven temperatures (10, 15, 20, 25, 30, 35, or 40°C).

Bacterial enumeration.

During storage, subsamples were analyzed at 6- to 12-h intervals for the samples stored at 10°C and at 1-h intervals for those stored at 40°C. Generally, samples stored at lower temperatures had longer sampling intervals, while shorter intervals were selected for samples at higher temperatures. At each sampling point, samples were mixed with 90 ml of 0.1% sterile buffered peptone water and homogenized for 2 min in a Seward stomacher (400 Circulator, Seward, London, UK). After homogenization, 1-ml aliquots of each sample were serially diluted in 9 ml of 0.1% sterile buffered peptone water, and 0.1-ml amounts of the diluents were spread plated onto Baird Parker agar (Difco) supplemented with 50 ml of egg yolk Tellurite emulsion. All plates were then incubated at 37°C for 24 h. Each experiment consisted of three independent trials, with duplicate analyses in each trial.

Model development.

A model was developed using the experimental data collected from S. aureus grown on all samples, combining data from raw pork, ham, and sausage. Bacterial counts from Baird Parker agar medium were used in the model fitting. DMFit Excel Add-in software (Institute of Food Research, Norwich, UK) was used to model the growth from all experimental observations, fitting the data to the Baranyi and Roberts model (2), and the growth of S. aureus was expressed as a function of time. Ratkowski's square root model (17, 18) was then used to describe the maximum growth rate (μmax, log CFU/h) of S. aureus as a function of temperature as follows:

formula

where c is the regression coefficient, T is the temperature (°C), and T0 is the theoretical minimum temperature for bacterial growth.

Model comparison.

The developed growth model here was compared with the growth models for S. aureus grown on each individual sample (raw pork, ham, or sausage) and the S. aureus ComBase models. The use of the ComBase models requires the water activity (aw) and pH values of the samples to be measured. The average pH and aw values of raw pork, ham, and sausage were 5.77 ± 0.12 and 0.97 ± 0.01, respectively.

The growth model developed was also compared with several existing growth models for pathogens on various types of meat, poultry, and their products, as shown in greater detail in Table 1. The square root growth rate (log CFU per gram per hour) data from all the models were used for comparison so as to investigate the relationship between them (7). Comparisons were only made in the range of temperatures examined here (from 10 to 40°C).

TABLE 1.

Pathogen growth models used for comparison with the Staphylococcus aureus growth model developed here

Pathogen growth models used for comparison with the Staphylococcus aureus growth model developed here
Pathogen growth models used for comparison with the Staphylococcus aureus growth model developed here

RESULTS AND DISCUSSION

The growth curves of S. aureus on raw pork, ham, and sausage stored under various isothermal conditions (10, 15, 20, 25, 30, 35, and 40°C) are shown in Figure 1. The growth curves were produced using the primary growth model of Baranyi and Roberts (2), which showed a high correlation coefficient (R2 > 0.94), except for those obtained from ham and sausage stored at 10°C (R2 > 0.64 and R2 > 0.47, respectively). The μmax of S. aureus on each pork sample and at each storage temperature were calculated. The results show that the μmax increased as the storage temperature increased. No naturally occurring bacterial colonies were evident when uninoculated samples were plated on selective medium (data not shown). The square root of μmax (with combined raw pork, ham, and sausage data) was fitted to the Ratkowsky equation [μmax (log CFU/h) = −0.150 × (T − 0.173)] with satisfactory results (R2 > 0.986).

FIGURE 1.

Growth curves of S. aureus on raw pork (A), ham (B), and sausage (C) stored under various isothermal conditions (10 to 40°C). Data are the mean results ± standard deviations.

FIGURE 1.

Growth curves of S. aureus on raw pork (A), ham (B), and sausage (C) stored under various isothermal conditions (10 to 40°C). Data are the mean results ± standard deviations.

Figure 2 shows a comparison of the S. aureus models developed from each individual pork product (raw pork, ham, or sausage) examined here, the models of S. aureus growth on bratwurst, raw pork, and ready-to-eat (RTE) pork established by Borneman et al. (4), Ingham et al. (11), and Min and Yoon (16), respectively, and the S. aureus growth model developed here using a combination of the data from the three pork products. The y axis shows the square root of the growth rate (SQRT) in log CFU per gram per hour, while the x axis shows the temperature in degrees centigrade. The SQRT in all models was linear with temperature. Wide spread of the SQRT data occurred at the higher temperatures between 25 and 40°C. Similar growth rates of S. aureus were observed for all the models from 10 to 25°C. Despite the putative differences between the different pork products, the S. aureus growth rates were similar on all pork products, except on raw pork examined here and bratwurst in the study by Borneman et al. (4). Interestingly, the S. aureus growth rates obtained by the model based on the pooled data from all three pork products here were quite similar to those predicted by the ComBase, where they were developed using the average pH and water activity values of raw pork, ham, and sausage. The results show that the S. aureus growth model developed here using the data from all three pork products seems to have the capability of predicting S. aureus growth on different pork products stored at isothermal temperatures from 10 to 25°C. Besides that, we have also successfully validated the ComBase model for its suitability for the prediction of S. aureus in pork and pork products.

FIGURE 2.

Comparison of S. aureus growth rates measured for the model developed here based on data pooled from all three pork products (—), individual raw pork (▪), ham (•), and sausage (▴) samples, ComBase growth rate predictions (♦), and the growth rate predictions in the literature for raw pork (□), RTE pork (○), and bratwurst (▾).

FIGURE 2.

Comparison of S. aureus growth rates measured for the model developed here based on data pooled from all three pork products (—), individual raw pork (▪), ham (•), and sausage (▴) samples, ComBase growth rate predictions (♦), and the growth rate predictions in the literature for raw pork (□), RTE pork (○), and bratwurst (▾).

A comparison of the S. aureus models developed for different meat and poultry products is shown in Figure 3. The SQRT data points of the comparison models tend to be equally distributed no matter the temperature. At lower temperatures between 10 and 20°C, the SQRT data obtained from chicken and turkey breast samples (5) were much lower than those of the model developed here, whereas at higher temperatures between 20 and 35°C, the SQRT data obtained from ground turkey and beef (4) were similar to those of the model developed here. As the temperature rose (between 35 to 40°C), however, the spread between predictions increased, particularly for the SQRT data obtained from ground turkey, which had a greater deviation than the SQRT from ground beef. The results reveal that the S. aureus model based on the pooled data from all three pork products also seems capable of predicting the growth of S. aureus in different meat and poultry products, while the predictions are likely to be more accurate if the products are stored at higher temperatures between 20 and 35°C.

FIGURE 3.

S. aureus growth rates measured for the model based on data pooled from all three pork products (—) compared with the growth rate predictions in the literature for ground beef (▪), ground turkey (□), turkey breast (•), and chicken breast (○).

FIGURE 3.

S. aureus growth rates measured for the model based on data pooled from all three pork products (—) compared with the growth rate predictions in the literature for ground beef (▪), ground turkey (□), turkey breast (•), and chicken breast (○).

The S. aureus growth model developed here was also compared with other existing pathogen growth models, including Salmonella Typhimurium on RTE pork (16) and E. coli O157:H7 and Salmonella serovars on bratwurst (4) and raw pork (11), as shown in Figure 4. The results show that the S. aureus model developed here was nearly identical with the model of E. coli O157:H7 and Salmonella serovars grown on bratwurst (4). The SQRT data of the three models were similar, except for those at high temperatures close to 40°C, in which the SQRT data of the other models were much lower than those of our model. On the other hand, wide spreads between the SQRT data of E. coli O157:H7 and Salmonella serovars on raw pork compared with those of the S. aureus model developed here were observed. In addition, the Salmonella Typhimurium growth model on RTE pork established by Min and Yoon (16) shows slightly greater SQRT predictions than the S. aureus growth model. Overall, most of the models were fairly similar, given that different strains, laboratories, and methods were used in their creation. Thus, the S. aureus growth model developed here is likely suitable for predicting the growth of other pathogens (especially E. coli O157:H7 and Salmonella Typhimurium) in pork and pork products over the temperature range of interest examined here.

FIGURE 4.

S. aureus growth rates measured for the model based on pooled data from all three pork products (—) compared with the growth rate predictions in the literature for Salmonella Typhimurium on RTE pork (▴), E. coli O157:H7 on raw pork (▪) and bratwurst (□), and Salmonella serovars on raw pork (•) and bratwurst (○).

FIGURE 4.

S. aureus growth rates measured for the model based on pooled data from all three pork products (—) compared with the growth rate predictions in the literature for Salmonella Typhimurium on RTE pork (▴), E. coli O157:H7 on raw pork (▪) and bratwurst (□), and Salmonella serovars on raw pork (•) and bratwurst (○).

Figure 5 shows a comparison of the data for growth models of various pathogens in different meat and poultry products. The S. aureus growth model developed here was compared with the models for Salmonella serovars in ground beef and turkey (4), E. coli O157:H7 in ground beef (4) and beef (19), and C. perfringens in cooked, uncured beef (13) and cured chicken (14). The spread between the SQRT predictions increased with rising temperature, especially at the temperatures between 25 and 40°C. C. perfringens growth in cooked, cured beef had the widest spread, followed by the models of C. perfringens in cooked, cured chicken and of Salmonella serovars in ground turkey. In contrast, the SQRT predictions of the Salmonella serovars and E. coli O157:H7 models in ground beef were fairly similar to those of the S. aureus model developed here, especially at the temperatures between 10 and 35°C. The results also indicated that the S. aureus growth model seems to have limitations when used to predict the growth of other pathogens in various meat and poultry products, especially at temperatures between 25 and 40°C. Natural variability and additional factors affecting bacterial growth not included in the model design, such as the presence of contaminating microorganisms, the addition of preservatives, and modified atmosphere packaging, etc., may also account for the lack of similarity between the developed models (20).

FIGURE 5.

S. aureus growth rates measured for the model based on pooled data from all three pork products (—) compared with the growth rate predictions in the literature for Salmonella serovars on ground beef (▪) and turkey (□), E. coli O157:H7 on beef in two models (▴ and ▾), and C. perfringens on cooked, uncured beef (•) and cured chicken (○).

FIGURE 5.

S. aureus growth rates measured for the model based on pooled data from all three pork products (—) compared with the growth rate predictions in the literature for Salmonella serovars on ground beef (▪) and turkey (□), E. coli O157:H7 on beef in two models (▴ and ▾), and C. perfringens on cooked, uncured beef (•) and cured chicken (○).

In conclusion, a mathematical model describing the SQRT of S. aureus on all three pork products (where data from raw pork, ham, and sausage were combined) as a function of temperature was developed. The SQRT predictions for S. aureus in the developed model were compared with those from the ComBase predictor and other pathogen growth models obtained from several selected sources in the literature. The results show that the S. aureus growth model based on the pooled data from all three pork products seems suitable for the prediction of S. aureus growth on different pork products under isothermal conditions from 10 to 25°C and of S. aureus growth on different meat and poultry products under higher temperatures from 20 to 35°C. Regardless of some high deviations observed at the temperatures between 25 and 40°C, the developed model is still likely suitable for predicting the growth of other pathogens on meat and poultry products over the temperature range of the experiments conducted here, especially for E. coli O157:H7 and Salmonella Typhimurium. This study, therefore, may provide a fast and cost-effective alternative to laboratory studies to estimate the effects of storage temperature on pathogen behavior in different meat and poultry products and may also be used in subsequent quantitative microbial risk assessments evaluating meat safety.

ACKNOWLEDGMENTS

This project was financially supported by a grant from the Animal, Plant & Fisheries Quarantine and Inspection Agency project no. Z-FS03-2011-12-01.

REFERENCES

REFERENCES
1.
Baer
,
A. A.
,
M. J.
Miller
, and
A. C.
Dilger
.
2013
.
Pathogens of interest to the pork industry: a review of research on interventions to assure food safety
.
Compr. Rev. Food Sci. Food Saf
.
12
:
183
217
.
2.
Baranyi
,
J.
, and
T. A.
Roberts
.
1994
.
A dynamic approach to predicting bacterial growth in food
.
Int. J. Food Microbiol
.
23
:
277
294
.
3.
Bennett
,
S. D.
,
K. A.
Walsh
, and
L. H.
Gould
.
2013
.
Foodborne disease outbreaks caused by Bacillus cereus, Clostridium perfringens, and Staphylococcus aureus—United States, 1998–2008
.
Clin. Infect. Dis
.
57
:
425
433
.
4.
Borneman
,
D. L.
,
S. C.
Ingham
, and
C.
Ané
.
2009
.
Mathematical approaches to estimating lag-phase duration and growth rate for predicting growth of Salmonella serovars, Escherichia coli O157:H7, and Staphylococcus aureus in raw beef, bratwurst, and poultry
.
J. Food Prot
.
72
:
1190
1200
.
5.
Castillejo-Rodríguez
,
A. M.
,
R. M.
García Gimeno
,
G.
Zurera Cosano
,
E.
Barco Alcalá
, and
M. R.
Rodríguez Pérez
.
2002
.
Assessment of mathematical models for predicting Staphylococcus aureus growth in cooked meat products
.
J. Food Prot
.
65
:
659
665
.
6.
Centers for Disease Control and Prevention
.
2013
.
Surveillance for foodborne disease outbreaks—United States, 1998–2008
.
Morb. Mortal. Wkly. Rep
.
62
:
1
34
.
7.
Danyluk
,
M. D.
,
L. M.
Friedrich
, and
D. W.
Schaffner
.
2014
.
Modeling the growth of Listeria monocytogenes on cut cantaloupe, honeydew and watermelon
.
Food Microbiol
.
38
:
52
55
.
8.
Delgado
,
C. L.
,
M. W.
Rosegrant
, and
S.
Meijer
.
2001
.
Livestock to 2020: the revolution continues
.
Available at: http://ageconsearch.umn.edu/bitstream/14560/1/cp01de01.pdf. Accessed 3 November 2014
.
9.
Guo
,
X.
,
T. A.
Mrox
,
B. M.
Popkin
, and
F.
Zhai
.
2000
.
Structural change in the impact of income on food consumption in China, 1989–1993
.
Econ. Dev. Cult. Change
48
:
737
760
.
10.
Huang
,
A.
,
C. B.
Ocfemia
, and
D. C.
Hunt
.
2006
.
Staphylococcal food poisoning outbreak in southeast Kansas. Kansas Department of Health and Environment, Topeka, KS
. .
11.
Ingham
,
S. C.
,
M. A.
Fanslau
,
G. M.
Burnham
,
B. H.
Ingham
,
J. P.
Norback
, and
D. W.
Schaffner
.
2007
.
Predicting pathogen growth during short-term temperature abuse of raw pork, beef, and poultry products: use of an isothermal-based predictive tool
.
J. Food Prot
.
70
:
1445
1456
.
12.
International Commission on Microbiological Specifications for Foods
.
2005
.
Meat and meat products
,
p
.
1
88
.
In
Microorganisms in foods 6, Microbial ecology of food commodities
, 2nd ed.
Kluwer Academic/Plenum Publishers
,
New York
.
13.
Juneja
,
V. K.
, and
H. M.
Marks
.
2002
.
Predictive model for growth of Clostridium perfringens during cooling of cooked cured chicken
.
Food Microbiol
.
19
:
313
327
.
14.
Juneja
,
V. K.
,
H. M.
Marks
, and
H.
Thippareddi
.
2008
.
Predictive model for growth of Clostridium perfringens during cooling of cooked uncured beef
.
Food Microbiol
.
25
:
42
55
.
15.
Mataragas
,
M.
,
P.
Skandamis
, and
E. H.
Drosinos
.
2008
.
Risk profiles of pork and poultry meat and risk ratings of various pathogen/product combinations
.
Int. J. Food Microbiol
.
126
:
1
12
.
16.
Min
,
K. J.
, and
K. S.
Yoon
.
2010
.
Development and validation of a predictive model for foodborne pathogens in ready-to-eat pork as a function of temperature and a mixture of potassium lactate and sodium diacetate
.
J. Food Prot
.
73
:
1626
1632
.
17.
Ratkowsky
,
D. A.
,
R. K.
Lowry
,
T. A.
McMeekin
,
A. N.
Stokes
, and
R. E.
Chandler
.
1983
.
Model for bacterial culture growth rate throughout the entire biokinetic temperature range
.
J. Bacteriol
.
154
:
1222
1226
.
18.
Ratkowsky
,
D. A.
,
J.
Olley
,
T. A.
McMeekin
, and
A.
Ball
.
1982
.
Relationship between temperature and growth rate of bacterial cultures
.
J. Bacteriol
.
149
:
1
5
.
19.
Tamplin
,
M. L.
,
G.
Paoli
,
B. S.
Marmer
, and
J.
Phillips
.
2005
.
Models of the behavior of Escherichia coli O157:H7 in raw sterile ground beef stored at 5 to 46°C
.
Int. J. Food Microbiol
.
100
:
335
344
.
20.
te Giffel
,
M. C.
, and
M. H.
Zwietering
.
1999
.
Validation of predictive models describing the growth of Listeria monocytogenes
.
Int. J. Food Microbiol
.
46
:
135
149
.
21.
U.S. Department of Agriculture, Economic Research Service
.
2014
.
Meat and livestock monthly U.S. trade—all commodities
. .

Author notes

† Present address: Food Analysis Center, Korea Food Research Institute, Anyangpangyo, Bundang, Seongnam, Gyeonggi 463-746, Republic of Korea.