International Journal of Food Microbiology 116 (2007) 136 – 143

www.elsevier.com/locate/ijfoodmicro

Modelling the bacterial survival/death interface induced

by high pressure processing
Shigenobu Koseki ⁎, Kazutaka Yamamoto
National Food Research Institute, 2-1-12, Kannondai, Tsukuba, Ibaraki 305-8642, Japan
Received 28 April 2006; received in revised form 23 August 2006; accepted 19 December 2006

Abstract

The survival/death interface model was developed for prediction of inactivation of Listeria monocytogenes by high pressure processing (HPP).
The model was derived from data sets comprising 360 combinations of environmental factors such as pressure (200, 300 400, and 500 MPa),
pressure-holding time (1, 3, 5, 10, 20, 30 min), pH (3, 4, 5, 6, 7), and inoculum level (3, 5, 7 log10 CFU/ml). The determination of survival/death
of L. monocytogenes after HPP was confirmed by the presence/absence of colony forming ability on non-selective agar plates after 30 days of
incubation at 20 °C in broth to take into account recovery of HPP-induced injured cells. The developed linear logistic model with time
logarithmically transformed gave a degree of agreement between probabilities predicted by the fitted model and all observations as 99.3%
concordant. The model provided a good fit to the data as shown by performance statistics. The developed interface model in the present study
provided requisite process conditions for the target effect of HPP on L. monocytogenes. In addition to using the simple linear logistic model, a
polynomial logistic model was also fitted to the data where pressure-holding time was not logarithmically transformed. That model did not
produce a better fit to the data and resulted in some potentially misleading predictions. Optimization of HPP could be accomplished using the
model developed in this study. Furthermore, choice in processing factors allows for processing flexibility in HPP and specifies the process criteria
that are incorporated into the HACCP plan.
© 2007 Elsevier B.V. All rights reserved.

Keywords: High pressure processing; Inactivation; Logistic regression; Interface model; Recovery

1. Introduction ever, non-thermal microbial inactivation models that consider

Microbial inactivation using thermal or non-thermal proces-
sing is generally affected by various environmental factors such
as temperature, pH, water activity, and pressure. Knowledge of
the influence of the combination of various factors on the effect
of microbial inactivation is indispensable for the optimization of
process conditions. Numerous predictive models for microbial
inactivation have been developed as a tool for optimization of
the process. There have been numerous thermal microbial in-
activation kinetic models that take into account the effect of
plural factors such as temperature, pH, water activity, etc.
(Bigelow, 1921; Cerf et al., 1996; Chhabra et al., 1999; Chiruta
et al., 1997; Cole et al., 1993; Davey et al., 1995; Lou and
Nakai, 2001; Membré et al., 1997; Xiong et al., 1999). How-
⁎ Corresponding author. Tel.: +81 29 838 7152; fax: +81 29 838 8122.
E-mail address: koseki@affrc.go.jp (S. Koseki).
0168-1605/$ - see front matter © 2007 Elsevier B.V. All rights reserved.
doi:10.1016/j.ijfoodmicro.2006.12.031
plural factors have just been reported (Buchanan and Golden,
1995; Doona et al., 2005; Kilimann et al., 2005; Ross et al.,
2005).
A trend involving the minimum processing of food that
enables the item to retain its original flavour, taste, and texture
has been spreading worldwide in recent years. Various non-
thermal processing techniques have been studied with a view to
achieving this minimal processing (Novak et al., 2002). In
particular, high hydrostatic pressure processing (HPP) is one of
the emerging technologies that research and development have
been actively conducted. Predictive models of HPP-induced
microbial inactivation have been developed so that the process
conditions of HPP are optimized (Buzrul et al., 2005; Chen
et al., 2004, 2005; Kilimann et al., 2005; Reyns et al., 2000;
Rodriguez et al., 2004; Yamamoto et al., 2005). However, most
of these models are kinetic models that do not take into account
factors other than pressure and time.
 S. Koseki, K. Yamamoto / International Journal of Food Microbiology 116 (2007) 136–143
Various studies have been reported about the recovery of
pathogenic and spoilage bacteria damaged by HPP during
storage at several temperatures in broth, milk, buffer, and ground
pork within 6 h to 4 weeks (Bozoglu et al., 2004; Bull et al.,
2005; Chilton et al., 2001; Ellenberg and Hoover, 1999; Koseki
and Yamamoto, 2006). Bozoglu et al. (2004) showed that even
though injured cells were not detected immediately after HPP,
the injured cells recovered after repair within 1 to 15 days and
could potentially recover in food. Since the published predictive
models of HPP-induced microbial inactivation were developed
on the basis of survival cell count data just after HPP, these
predictive models did not take into account the recovery of
injured bacteria induced by HPP. Therefore, even though there
are many viable cells that are injured and do not have a colony
forming ability, the prediction of these models supposes bacterial
death by HPP. As a result, the effect of microbial inactivation
induced by HPP has been overestimated by these predictive
kinetic models. This inaccuracy could lead to an increased risk of
food poisoning or spoilage and could be a critical issue in terms
of the safety design of high pressure processed food.
In view of the above concern, a novel predictive model of
HPP-induced microbial inactivation is required so that the model
takes into account the effect of plural factors and the recovery of
HPP-injured cells. In order to develop such a model, we
attempted to apply a growth/no-growth interface model that
describes the limit condition of bacterial growth. The growth/no-
growth model is derived from a logistic regression procedure,
which uses a dichotomous dependent variable. In the present
study, we assessed the HPP-induced bacterial inactivation as a
dichotomy of cell survival or death that takes into account the
recovery of injured cells. The model developed in this study will
therefore present an appropriate outcome of HPP-induced
bacterial inactivation. Furthermore, this new model will allow
the determination of appropriate processing conditions based on
the combination of various factors. This modelling procedure
will be applied not only to HPP, but also to other processing
techniques regardless of their thermal or non-thermal basis. The
results of this study will contribute to the advancement of
predictive modelling for HPP-induced microbial inactivation.
2. Materials and methods
2.1. Cell preparation
The pathogenic bacterium Listeria monocytogenes ATCC
19117 was used in this study. This strain showed the greatest
baroresistance among the six strains of ATCC 19111, ATCC
19117, ATCC 19118, ATCC 13932, ATCC 15313, and ATCC
35152 (Koseki and Yamamoto, in press). The strain was main-
tained at − 85 °C in brain heart infusion (BHI) broth (Merck,
Darmstadt, Germany) containing 10% glycerol. A sterile dis-
posable plastic loop was used to transfer the frozen bacterial
cultures to 10 ml of BHI broth in a glass tube. The cultures were
incubated without agitation at 37 °C, with transfer using loop
inocula at three successive 24-h intervals. Cells were collected
by centrifugation (3000×g, 15 min, 20 °C), and the resulting
pellet was resuspended in 5 ml of each tested medium. The
137
media used in this study were 0.01 M phosphate-buffered saline
with a pH adjusted from 3 to 7 in unit increment using 6 M
hydrochloric acid (HCl). The inoculum levels of approximately
3, 5, and 7 log10 CFU/ml were adjusted by serial 10-fold
dilutions in each media.
2.2. High hydrostatic pressure processing
Cell suspensions in each medium (2 ml) were placed into a
sterile polyethylene bag and heat-sealed following the exclusion
of air bubbles. The bag was placed into a hydrostatic
pressurization unit, which contains a pressure chamber 3 cm
in diameter and 15 cm in height (HIGHPREX R7K-3-15,
Yamamoto Suiatsu Kogyosho, Osaka, Japan). HPP at constant
pressures from 200 MPa in 100 MPa increments to a maximum
of 500 MPa was carried out at 22 ± 2 °C for 1, 3, 5, 10, 20, and
30 min using water as the pressure medium. The apparatus was
equipped with a temperature controller. The temperature of the
medium (water in this study) in the pressure vessel was
measured by k-type thermo-couples during pressurization. The
come-up rate was about 250 MPa/min, and the pressure-release
time was less than 10 s. The HPP time reported in the
experiment does not include the come-up and pressure-release
times. Following treatment, samples were stored in crushed-ice
for up to 1 h until commencement of bacterial analysis. A total
of 360 combinations of pressure (200, 300, 400, 500 MPa),
pressure-holding time (1, 3, 5, 10, 20, 30 min), pH (3, 4, 5, 6, 7),
and inoculum level (3, 5, 7 log10 CFU/ml) in triplicate for each
combination were examined.
2.3. Determination of survival/death of L. monocytogenes after
HPP
After HPP, the presence or absence of viable L. mono-
cytogenes was determined by the following procedure. Each
pressurized medium (1 ml) was aseptically removed from each
plastic bag, added to BHI broth (9 ml), and incubated at 20 or
37 °C for 30 days. Samples were taken at 5-day intervals with a
sterile loop and streaked on duplicate non-selective agar plates
(tryptic soy agar, TSA, Merck). Plates were then incubated at
37 °C for 48 h. Survival of L. monocytogenes after HPP was
indicated by the presence of colonies on TSA plates. Moreover,
several typical colonies were taken from the plates using a
sterile loop, and the presence of L. monocytogenes confirmed
using a Listeria diagnosis kit (Singlepath® Listeria; Merck).
2.4. Model development
For each replicate response of L. monocytogenes, survival or
death was scored with values 1 or 0, respectively. Data were
fitted to a logistic regression model using SAS/STAT (Ver. 9.1
for Unix, SAS, Cary, NC, USA) based on the approach
described by Ratkowsky and Ross (1995). The model was of the
form shown in Eq. (1).
LogitðPÞ ¼ a0 þ a1 Press þ a2 log10 ðTimeÞ þ a3 pH þ a4 IC;
ð1Þ
138 S. Koseki, K. Yamamoto / International Journal of Food Microbiology 116 (2007) 136–143

Table 1 Table 2
Data illustrating recovery of L. monocytogenes in broth after HPP at pH 3 Parameter estimates of the logistic regression model for survival/death interface
of L. monocytogenes after HPP
Parameter Estimate Standard error Probability
a0: Intercept 12.9973 2.6307 b0.0001
a1: Pressure − 0.0775 0.0126 b0.0001
a2: log10(Time) − 9.1909 1.5958 b0.0001
a3: pH 2.3331 0.4350 b0.0001
a4: IC 1.6674 0.3238 b0.0001
Max-rescaled R2 = 0.9213; Hosmer–Lemeshow goodness-of-fit statistic =
1.3899 with df = 8 (P = 0.9944); percent concordant, 99.3.

relationship between effect of microbial inactivation and the

pressure-holding time is not linear (Buzrul and Alpas, 2004;
Buzrul et al., 2005; Chen et al., 2004, 2005; Yamamoto et al.,
2005).
Further, as a comparison to the newly developed model
(Eq. (1)), a polynomial logistic regression model was examined.
The model was of the form shown in Eq. (2).

LogitðPÞ ¼ a0 þ a1 Press þ a2 Time þ a3 pH þ a4 IC

þ a5 PressdTime þ a6 PressdpH þ a7 PressdIC
þ a8 TimedpH þ a9 TimedIC þ a10 pHdIC
þ a11 Press2 þ a12 Time2 þ a13 pH2 þ a14 IC2 ; ð2Þ

where the terms are defined as described above. The automatic

variable selection option with a stepwise selection method was
applied to determine the most significant effects (P b 0.05). The
predicted survival/death interfaces for P = 0.1, 0.5, and 0.9 were
calculated using Grapher® (Apple Computer, CA, USA).

3. Results

3.1. Recovery of L. monocytogenes after HPP

Even though viable L. monocytogenes cells were not

detected on TSA plates immediately after HPP, the colony
forming ability of L. monocytogenes recovered after incubation
at 20 °C for 10 to 30 days in some cases. An example of
this recovery is shown in Table 1. While viable cells of
L. monocytogenes (5 log10 CFU/ml) were not detected by HPP
at 300 MPa in the media with pH 3 for 3 and 5 min immediately
after treatment, viable cells were detected after incubation at

Table 3
Parameter estimates of the nonlinear logistic regression model for survival/death
interface of L. monocytogenes after HPP
Parameter Estimate Standard error Probability
a0: Intercept 24.3051 6.2423 b0.001
a1: Pressure − 0.0891 0.0157 b0.001
where, Logit (P) is an abbreviation of ln [P / (1 − P)], ln is the a2: Time − 0.8062 0.2039 b0.001
a3: pH − 2.9955 1.9271 0.1201
natural logarithm, P is the probability of survival (in the range
a4: IC 2.4490 0.5101 b0.001
of 0–1), ai are coefficients to be estimated, Press is the pressure a5: Time × IC − 0.0490 0.0205 0.0170
applied, Time is the time of pressure holding, pH is the pH of a6: Time2 0.0184 0.0051 0.0003
the medium, and IC is the inoculum level (log10 CFU/ml) of a7: pH2 0.5669 0.2091 0.0067
L. monocytogenes in the tested solution. We employed Max-rescaled R2 = 0.9292; Hosmer–Lemeshow goodness-of-fit statistic =
logarithm of the time because it has been reported that the 13.0381 (df = 8, P = 0.1105); Percent concordant, 99.4.
 S. Koseki, K. Yamamoto / International Journal of Food Microbiology 116 (2007) 136–143

Fig. 1. Survival/death interface for L. monocytogenes after high pressure processing (HPP). Model predictions from Eq. (3) were compared to data used to generate the model for the effect of pressure and logarithmically
transformed pressure-holding time at different inoculum levels (3, 5, and 7 log CFU/ml) on the potential for survival of L. monocytogenes after HPP in a media of different pH (3 to 7) at three levels of probability: P = 0.1
(upper curve), P = 0.5 (middle curve), and P = 0.9 (lower curve). The interface with P = 0.5 (50% probability of survival) is indicated by the solid lines; dotted and dashed lines represent predictions at P = 0.1
(more conservative) and P = 0.9 (less conservative), respectively.
139
140 S. Koseki, K. Yamamoto / International Journal of Food Microbiology 116 (2007) 136–143

20 °C for at least 10 days. Similar phenomena of recovery were
recorded for incubation at 20 °C under other treatment condi-
tions. However, recovery was not detected during incubation at
37 °C. Since no new recovery occurred after 20 days following
30 days of incubation at 20 °C, we defined the death of
L. monocytogenes by HPP on the basis of no detection of viable
cells after 30 days of incubation at 20 °C.
3.2. Model development
The survival/death response of L. monocytogenes was
monitored in 360 combinations with triplicates tested for each
condition. The initial inoculum levels were 3.0 ± 0.3 log CFU/
ml, 5.1 ± 0.2 log CFU/ml, and 7.0 ± 0.2 log CFU/ml, respec-
tively. Therefore, we used the initial inoculum level as 3, 5, and
7 log10 CFU/ml in the presented models. The inoculum levels of
3, 5, and 7 log10 CFU/ml represented the contamination levels
of low, mid, and high, respectively. In real food processing, a 5-
log reduction has been widely required for process criteria.
However, from a practical point of view, to ensure the balance
of nutritional quality and microbiological safety of foods,
optimization of process conditions is indispensable. Therefore,
three inoculum levels were chosen in order to apply to various
requirements of process condition.
All triplicate trials showed either all survival or all death. The
data were further modelled using a newly developed simple
logistic regression model (Eq. (1)). The parameters were
estimated using the SAS program PROC LOGISTIC, and the
resulting model was obtained:
and observed responses of L. monocytogenes are shown in
Fig. 1. Fig. 1 shows the interaction between applied pressures
and logarithmically transformed pressure-holding time at
different inoculum levels (3, 5, and 7 log10 CFU/ml) in the
media with different pH (3 to 7). The magnitude of the pressure
and initial cell concentration of L. monocytogenes influenced
the pressure-holding time required for inactivation. While
L. monocytogenes at 3 log10 CFU/ml was eliminated (50%
probability) by treatment at 400 MPa for 2.3 min, a longer
pressure-holding time at 400 MPa was required to eliminate
L. monocytogenes at 5 log10 CFU/ml (5.3 min) and 7 log10
CFU/ml (12.3 min). As the inoculum level and pH of media
increased, the time for inactivation was extended. The elimina-
tion (50% probability) of 3, 5, and 7 log10 CFU/ml of
L. monocytogenes in media of pH 6 were induced by HPP at
400 MPa for 1.3 min, 3.0 min, and 6.8 min, respectively. In
order to eliminate the higher inoculum of L. monocytogenes, a
longer pressure-holding time and higher pressure are required.
However, the reduced pH assisted the effect of bacterial

LogitðPÞ ¼ 12:9973−0:0775Press−9:1909log10 ðTimeÞ

þ 2:3331pH þ 1:6674IC: ð3Þ

Standard errors for the parameter estimates of Eq. (3) are

given in Table 2, which also shows performance statistics for
this model. This model provided a good fit to the data as shown
by the performance statistics (Hosmer and Lemeshow, 1989).
Further, the data were modelled using a polynomial logistic
regression model. The parameters were estimated using a
stepwise variable selection method, and the resulting model was
obtained as follows:

LogitðPÞ ¼ a0 þ a1 Press þ a2 Time þ a3 pH þ a4 IC

þ a5 Time  IC þ a6 Time2 þ a7 pH2 ð4Þ

Estimates for the parameters of Eq. (4) and their approximate

standard errors are given in Table 3, which also shows
performance statistics for this model. The model derived from
polynomial logistic regression showed a slightly poorer fit to
the data as indicated by the larger value of Hosmer–Lemeshow
goodness-of-fit statistics with respect to the simple logistic
regression model of Table 2.
Fig. 2. Comparison of survival/death interfaces for L. monocytogenes after high
pressure processing (HPP) derived from newly developed model (solid line) and
3.3. Model prediction polynomial logistic regression model (dashed line). (a) The interface between
pH of the tested solution and pressure-holding time with inoculum level of 7 log
Comparative representations of the predicted survival/death CFU/ml treated by 400 MPa. (b) The interface between inoculum level and
interfaces derived from the newly developed model (Eq. (3)) pressure-holding time at pH 4 treated by 300 MPa.
 S. Koseki, K. Yamamoto / International Journal of Food Microbiology 116 (2007) 136–143
inactivation of HPP. Even at a low applied pressure of 200 MPa,
L. monocytogenes with a low inoculum level of 3 log10 CFU/ml
in the media of pH 4 was eliminated by treatment for more than
19.4 min. HPP at 400 MPa eliminated the 7 log10 CFU/ml
inoculum of L. monocytogenes in the pH 4 media for more than
2.1 min at a 50% probability. The models developed in this
study appear to be of biological sense.
On the other hand, some of the interfaces derived from the
polynomial logistic model showed illogical shapes (Fig. 2). The
interaction between pH of the tested solution and pressure-
holding time with inoculum level of 7 log10 CFU/ml treated by
400 MPa did not make biological sense below a pH of 2.5
(Fig. 2a). The interface between inoculum level and pressure-
holding time at pH 4 treated with 300 MPa indicated a
contradiction (Fig. 2b). The interface predicted 0 min for
eliminating 2.2 log10 CFU/ml of L. monocytogenes at 300 MPa.
This result is not sensible biologically.
4. Discussion
A novel probabilistic predictive model for HPP-induced
L. monocytogenes inactivation was developed in this study. The
developed predictive model takes into account not only plural
factors affecting HPP, but also the recovery of injured cells
induced by HPP. The developed models allow the appropriate
prediction of the effects of HPP-induced bacterial inactivation.
Furthermore, the models could identify optimum process
conditions to achieve required process criteria.
No recovery of L. monocytogenes injured by HPP occurred
at 37 °C, which was around the optimum growth temperature.
Significant recovery appeared at 20 °C, which was lower than
the optimum growth temperature. This result is consistent with
other reports (Bozoglu et al., 2004; Bull et al., 2005; Koseki and
Yamamoto, 2006). We therefore adopted a 20 °C incubation for
determination of the survival/death of L. monocytogenes after
HPP. Since there were no changes in bacterial recovery from 20
to 30 days following incubation at 20 °C, the determination of
survival/death was made after 30 days of incubation. Other
reports have also determined the recovery of HPP-induced
injury cells within 4 weeks (Bozoglu et al., 2004; Bull et al.,
2005).
The number of studies on the microbial inactivation kinetics
of HPP has increased (Buzrul and Alpas, 2004; Buzrul et al.,
2005; Chen et al., 2004, 2005 Kilimann et al., 2005; Reyns et
al., 2000; Rodriguez et al., 2004; Yamamoto et al., 2005). The
patterns of the inactivation kinetics of microorganisms by HPP
are not those of first-order kinetics such as thermal inactivation,
but are generally nonlinear curves with tailing. In order to
describe the nonlinear kinetics of HPP-induced microbial
inactivation, appropriate functions such as log-logistic, Gom-
pertz, or Weibull were applied to the inactivation curve (Buzrul
and Alpas, 2004; Buzrul et al., 2005; Chen et al., 2004, 2005;
Yamamoto et al., 2005). These fitting procedures are effective
for an evaluation of HHP-induced microbial inactivation
curves. However, it should be noted that these procedures are
not predictive, only fitting. Inactivation kinetics in conditions
other than those examined cannot be predicted by a curve fitting
141
procedure. Furthermore, these curve-fitting procedures were
derived from survival cell count data obtained immediately after
HPP. The recovery of injured cells induced by HPP was not
taken into account. The injury induced by HPP may be
repairable and the cells can grow after repairing the site of injury
during storage. The recovery of food-borne pathogens such as
L. monocytogenes during storage is a critical issue for high
pressure processed foods because it can cause an overestimation
of the safety of such processed foods.
In order to resolve both the requirements of the novel
predictive model and considerations of injured cell recovery, we
have developed a probabilistic model of the survival/death
interface. The bacterial inactivation effect of HPP was assessed
by survival/death during incubation periods sufficient for the
recovery of injury (Bozoglu et al., 2004). The model was
developed as a probabilistic model of the occurrence of the
event using a logistic regression procedure. This kind of
probabilistic modelling procedure has been applied to predict
conditions of bacterial growth limits (Koutsoumanis et al.,
2004a,b; Le Marc et al., 2005; McKellar et al., 2002; Presser
et al., 1998; Ratkowsky and Ross, 1995; Ross et al., 2000;
Salter et al., 2000; Tienungoon et al., 2000). The growth/no-
growth interface has been used as an important model for
assessing microbial food safety (McMeekin et al., 2000).
However, there have been no reports on the application of this
procedure to microbial inactivation processes involving HPP or
other techniques.
Nonlinear logistic regression models have been used for
estimation of growth/no-growth interface. However, some
problems of use of the nonlinear logistic regression have been
reported (Ratkowsky, 2002). Furthermore, there is little or no
evidence presented in the published scientific literature to
support the notion of an interaction among the environmental
factors that control the spoilage process or enable the growth of
pathogenic microorganisms. Therefore, we employed only the
factors directly affecting bacterial inactivation of HPP such as
pressure, logarithm of pressure-holding time, pH, and inoculum
level. The use of logarithm of time is the main contribution of
this modelling procedure. The relationship between the effect of
microbial inactivation and the pressure-holding time is nonlinear
(Buzrul and Alpas, 2004; Buzrul et al., 2005; Chen et al., 2004,
2005; Yamamoto et al., 2005). The effect of pressure-holding
time decreases as the pressure-holding time prolongs. The
logarithm of time would reflect the effect of treatment time on
microbial inactivation of HPP. Moreover, the developed model
was simple compared with a polynomial logistic model that
included many variables. Besides, the goodness-of-fit statistics
of the polynomial model were not only worse than those of the
newly developed model, but that model resulted in an illogical
shape of the survival/death interface under some conditions.
This result illustrates the harmful effect of mechanical use of
polynomial modelling, as it may not produce biologically
meaningful results. Accordingly, the newly developed simple
model is both meaningful and appropriate for prediction of the
microbial survival/death interface induced by HPP.
The new model developed in the present study will be able to
provide requisite process conditions for the target effect of
142 S. Koseki, K. Yamamoto / International Journal of Food Microbiology 116 (2007) 136–143
processing. Optimization of HPP could be accomplished using
the developed model. Furthermore, the modelling procedures
presented in this study will be applied to not only HPP, but also
other microbial inactivation processing regardless of their
thermal or non-thermal basis. Choice in processing parameters
allows for processing flexibility and specifies the process
criteria that are incorporated into the HACCP plan. However,
model predictions will need to be validated in the target food
intended for commercial production.
In conclusion, the modelling procedure presented in this
study is applied to HPP and other microbial inactivation pro-
cesses. Although the main stream of the microbial inactivation
model has been a kinetic model, the survival/death interface
model can become an alternative microbial inactivation predic-
tive model. The model will contribute to process management
for safe food production in the future.
Acknowledgment
This work was supported by Grant of the Japan Society for
the Promotion of Science (JSPS) for Young Scientists. Further-
more, this work was funded by the Ministry of Agriculture,
Forestry and Fisheries (MAFF) of Japan through the project
“Integrated research on functionality and safety of food”. The
authors would like to thank Ms. Yasuko Mizuno for her
excellent technical assistance in the microbial experiments.
Also, authors would like to thank Dr. D. A. Ratkowsky for his
critical review and valuable suggestions.
References
Bigelow, W.D., 1921. The logarithmic nature of thermal death time curves.
J. Infect. Dis. 29, 528–536.
Bozoglu, F., Alpas, H., Kaletunc, G., 2004. Injury recovery of foodborne
pathogens in high hydrostatic pressure treated milk during storage. FEMS
Immunol. Med. Microbiol. 40, 243–247.
Buchanan, R.L., Golden, M.H., 1995. Model for the nonthermal inactivation of
Listeria-monocytogenes in a reduced oxygen environment. Food Microbiol.
12, 203–212.
Bull, M.K., Hayman, M.M., Stewart, C.M., Szabo, E.A., Knabel, S.J., 2005.
Effect of prior growth temperature, type of enrichment medium, and
temperature and time of storage on recovery of Listeria monocytogenes
following high pressure processing of milk. Int. J. Food Microbiol. 101,
53–61.
Buzrul, S., Alpas, H., 2004. Modeling the synergistic effect of high pressure and
heat on inactivation kinetics of Listeria innocua: a preliminary study. FEMS
Microbiol. Lett. 238, 29–36.
Buzrul, S., Alpas, H., Bozoghu, F., 2005. Use of Weibull frequency distribution
model to describe the inactivation of Alicyclobacillus acidoterrestris by high
pressure at different temperatures. Food Res. Int. 38, 151–157.
Cerf, O., Davey, K.R., Sadoudi, A.K., 1996. Thermal inactivation of bacteria—a
new predictive model for the combined effect of three environmental factors:
temperature, pH and water activity. Food Res. Int. 29, 219–226.
Chen, H.Q., Joerger, R.D., Kingsley, D.H., Hoover, D.G., 2004. Pres-
sure inactivation kinetics of phage lambda cI 857. J. Food Prot. 67,
505–511.
Chen, H.Q., Hoover, D.G., Kingsley, D.H., 2005. Temperature and treatment
time influence high hydrostatic pressure inactivation of feline calicivirus, a
norovirus surrogate. J. Food Prot. 68, 2389–2394.
Chhabra, A.T., Carter, W.H., Linton, R.H., Cousin, M.A., 1999. A predictive
model to determine the effects of pH, milkfat, and temperature on thermal
inactivation of Listeria monocytogenes. J. Food Prot. 62, 1143–1149.
Chilton, P., Isaacs, N.S., Manas, P., Mackey, B.M., 2001. Biosynthetic
requirements for the repair of membrane damage in pressure-treated
Escherichia coli. Int. J. Food Microbiol. 71, 101–104.
Chiruta, J., Davey, K.R., Thomas, C.J., 1997. Thermal inactivation kinetics of
three vegetative bacteria as influenced by combined temperature and pH in a
liquid medium. Food Bioprod. Process. 75, 174–180.
Cole, M.B., Davies, K.W., Munro, G., Holyoak, C.D., Kilsby, D.C., 1993. A
vitalistic model to describe the thermal inactivation of Listeria-mono-
cytogenes. J. Indust. Microbiol. 12, 232–239.
Davey, K.R., Hall, R.F., Thomas, C.J., 1995. Experimental and model studies of
the combined effect of temperature and pH on the thermal sterilization of
vegetative bacteria in liquid. Food Bioprod. Process 73, 127–132.
Doona, C.J., Feeherry, F.E., Ross, E.W., 2005. A quasi-chemical model for the
growth and death of microorganisms in foods by non-thermal and high-
pressure processing. Int. J. Food Microbiol. 100, 21–32.
Ellenberg, L., Hoover, D.G., 1999. Injury and survival of Aeromonas hydrophila
7965 and Yersinia enterocolitica 9610 from high hydrostatic pressure.
J. Food Saf. 19, 263–276.
Hosmer, D.W., Lemeshow, S., 1989. Applied logistic regression. Wiley and
Sons, New York.
Kilimann, K.V., Hartmann, C., Delgado, A., Vogel, R.F., Ganzle, M.G., 2005. A
fuzzy logic-based model for the multistage high-pressure inactivation of
Lactococcus lactis ssp. cremoris MG 1363. Int. J. Food Microbiol. 98,
89–105.
Koseki, S., Yamamoto, K., 2006. Recovery of Escherichia coli ATCC 25922 in
phosphate buffered saline after treatment with high hydrostatic pressure. Int.
J. Food Microbiol. 110, 108–111.
Koseki, S., Yamamoto, K., in press. Water activity of bacterial dispersion media
unable to account for the baroprotective effect of solute concentration on the
inactivation of Listeria monocytogenes by high hydrostatic pressure. Int. J.
Food Microbiol. (Available online 20 Dec 2006).
Koutsoumanis, K.P., Kendall, P.A., Sofos, J.N., 2004a. A comparative study on
growth limits of Listeria monocytogenes as affected by temperature, pH and
aw when grown in suspension or on a solid surface. Food Microbiol. 21,
415–422.
Koutsoumanis, K.P., Kendall, P.A., Sofos, J.N., 2004b. Modeling the boundaries
of growth of Salmonella typhimurium in broth as a function of temperature,
water activity, and pH. J. Food Prot. 67, 53–59.
Le Marc, Y., Pin, C., Baranyi, J., 2005. Methods to determine the growth domain
in a multidimensional environmental space. Int. J. Food Microbiol. 100,
3–12.
Lou, W.G., Nakai, S., 2001. Application of artificial neural networks for
predicting the thermal inactivation of bacteria: a combined effect of
temperature, pH and water activity. Food Res. Int. 34, 573–579.
McKellar, R.C., Lu, X., Delaquis, P.J., 2002. A probability model describing the
interface between survival and death of Escherichia coli O157: H7 in a
mayonnaise model system. Food Microbiol. 19, 235–247.
McMeekin, T.A., Presser, K., Ratkowsky, D., Ross, T., Salter, M., Tienungoon,
S., 2000. Quantifying the hurdle concept by modelling the bacterial growth/
no growth interface. Int. J. Food Microbiol. 55, 93–98.
Membré, J.M., Majchrzak, V., Jolly, I., 1997. Effects of temperature, pH,
glucose, and citric acid on the inactivation of Salmonella typhimurium in
reduced calorie mayonnaise. J. Food Prot. 60, 1497–1501.
Novak, J.S., Sapers, G.M., Juneja, V.K., 2002. Microbial Safety of Minimally
Processed Foods. CRC Press, Boca Raton, USA.
Presser, K.A., Ross, T., Ratkowsky, D.A., 1998. Modelling the growth limits
(growth no growth interface) of Escherichia coli as a function of
temperature, pH, lactic acid concentration, and water activity. Appl.
Environ. Microbiol. 64, 1773–1779.
Ratkowsky, D.A., 2002. Some examples of, and some problems with, the use of
nonlinear logistic regression in predictive food microbiology. Int. J. Food
Microbiol. 73, 119–125.
Ratkowsky, D.A., Ross, T., 1995. Modeling the bacterial-growth no growth
interface. Lett. Appl. Microbiol. 20, 29–33.
Reyns, K., Soontjens, C.C.F., Cornelis, K., Weemaes, C.A., Hendrickx, M.E.,
Michiels, C.W., 2000. Kinetic analysis and modelling of combined high-
pressure-temperature inactivation of the yeast Zygosaccharomyces bailii.
Int. J. Food Microbiol. 56, 199–210.
 S. Koseki, K. Yamamoto / International Journal of Food Microbiology 116 (2007) 136–143
Rodriguez, A.C., Larkin, J.W., Dunn, J., Patazca, E., Reddy, N.R., Alvarez-
Medina, M., Tetzloff, R., Fleischman, G.J., 2004. Model of the inactivation
of bacterial spores by moist heat and high pressure. J. Food Sci. 69,
E367–E373.
Ross, T., Dalgaard, P., Tienungoon, S., 2000. Predictive modelling of the growth
and survival of Listeria in fishery products. Int. J. Food Microbiol. 62,
231–245.
Ross, E.W., Taub, I.A., Doona, C.J., Feeherry, F.E., Kustin, K., 2005. The
mathematical properties of the quasi-chemical model for microorganism
growth-death kinetics in foods. Int. J. Food Microbiol. 99, 157–171.
Salter, M.A., Ratkowsky, D.A., Ross, T., McMeekin, T.A., 2000. Modelling the
combined temperature and salt (NaCl) limits for growth of a pathogenic
143
Escherichia coli strain using nonlinear logistic regression. Int. J. Food
Microbiol. 61, 159–167.
Tienungoon, S., Ratkowsky, D.A., McMeekin, T.A., Ross, T., 2000. Growth
limits of Listeria monocytogenes as a function of temperature, pH, NaCl,
and lactic acid. Appl. Environ. Microbiol. 66, 4979–4987.
Xiong, R., Xie, G., Edmondson, A.E., Sheard, M.A., 1999. A mathematical
model for bacterial inactivation. Int. J. Food Microbiol. 46, 45–55.
Yamamoto, K., Matsubara, M., Kawasaki, S., Bari, M.L., Kawamoto, S., 2005.
Modeling the pressure inactivation dynamics of Escherichia coli. Braz. J.
Med. Biol. Res. 38, 1253–1257.