1697

Journal of Food Protection, Vol. 80, No. 10, 2017, Pages 1697–1704
doi:10.4315/0362-028X.JFP-17-096
Copyright Ó, International Association for Food Protection

Research Paper

Effects of Nitrite and Erythorbate on Clostridium perfringens

Growth during Extended Cooling of Cured Ham
KATIE J. OSTERBAUER,1 AMANDA M. KING,2 DENNIS L. SEMAN,2 ANDREW L. MILKOWSKI,2 KATHLEEN A. GLASS,1
AND JEFFREY J. SINDELAR2*

1Food Research Institute, University of Wisconsin–Madison, 1550 Linden Drive, Madison, Wisconsin 53706; and 2Department of Animal Sciences,
University of Wisconsin–Madison, 1805 Linden Drive, Madison, Wisconsin 53706, USA

MS 17-096: Received 7 March 2017/Accepted 22 May 2017/Published Online 8 September 2017

ABSTRACT
To control the growth of Clostridium perfringens in cured meat products, the meat and poultry industries commonly follow
stabilization parameters outlined in Appendix B, ‘‘Compliance Guidelines for Cooling Heat-Treated Meat and Poultry Products
(Stabilization)’’ (U.S. Department of Agriculture, Food Safety and Inspection Service [USDA-FSIS], 1999) to achieve cooling
(54.4 to 4.48C) within 15 h after cooking. In this study, extended cooling times and their impact on C. perfringens growth were
examined. Phase 1 experiments consisted of cured ham with 200 mg/kg ingoing sodium nitrite and 547 mg/kg sodium
erythorbate following five bilinear cooling profiles: a control (following Appendix B guidelines: stage A cooling [54.4 to 26.78C]
for 5 h, stage B cooling [26.7 to 4.48C] for 10 h), extended stage A cooling for 7.5 or 10 h, and extended stage B cooling for 12.5
or 15 h. A positive growth control with 0 mg/kg nitrite added (uncured) was also included. No growth was observed in any
treatment samples except the uncured control (4.31-log increase within 5 h; stage A). Phase 2 and 3 experiments were designed to
investigate the effects of various nitrite and erythorbate concentrations and followed a 10-h stage A and 15-h stage B bilinear
cooling profile. Phase 2 examined the effects of nitrite concentrations of 0, 50, 75, 100, 150, and 200 mg/kg at a constant
concentration of erythorbate (547 mg/kg). Results revealed changes in C. perfringens populations for each treatment of 6.75,
3.59, 2.43, 0.38, 0.48, and 0.50 log CFU/g, respectively. Phase 3 examined the effects of various nitrite and erythorbate
concentrations at 100 mg/kg nitrite with 0 mg/kg erythorbate, 100 with 250, 100 with 375, 100 with 547, 150 with 250, and 200
with 250, respectively. The changes in C. perfringens populations for each treatment were 4.99, 2.87, 2.50, 1.47, 0.89, and 0.60
log CFU/g, respectively. Variability in C. perfringens growth for the 100 mg/kg nitrite with 547 mg/kg erythorbate treatment was
observed between phases 2 and 3 and may have been due to variations in treatment pH and NaCl concentrations. This study
revealed the importance of nitrite and erythorbate for preventing growth of C. perfringens during a much longer (25 h) cooling
period than currently specified in the USDA-FSIS Appendix B.
Key words: Clostridium perfringens; Cooling; Sodium erythorbate; Sodium nitrite

Clostridium perfringens is a gram-positive, spore-
forming anaerobe that is found ubiquitously in the
environment, including soil, water, and animal intestines.
This pathogen affects nearly 1 million people each year,
making it the third leading cause of foodborne illness
according to the Centers for Disease Control and Prevention
(3, 6, 14, 16). Illness caused by this pathogen is generally
self-limiting, and signs include abdominal pain, cramping,
and diarrhea caused by toxicoinfection with C. perfringens
enterotoxin. Toxicoinfection can occur when a large dose of
vegetative cells (106) is ingested, and those cells sporulate
in response to the conditions of the gastrointestinal tract,
releasing enterotoxin during sporulation (20).
C. perfringens is a concern in meat products because of
its widespread environmental prevalence and its growth
ecology. Although vegetative cells are inactivated by
* Author for correspondence. Tel: 608-262-0555; Fax: 608-265-3110;
E-mail: jsindelar@wisc.edu.
thermal cooking processes used in the production of
ready-to-eat (RTE) meat and poultry products, cooking
temperatures of 70 to 808C also cause the heat-resistant
spores to germinate (4). Food safety challenges can then
arise based on the time that cooked product temperatures are
within the 12 to 508C growth temperature range of C.
perfringens, such as during extended chilling. Within the
optimal growth temperature range of 43 to 478C, C.
perfringens can grow rapidly, with generation times of less
than 10 min (5). Thus, there is significant potential for high
numbers of vegetative C. perfringens cells to be present in
RTE products when temperatures in production are not
appropriately controlled (17).
Various intrinsic factors can also influence C. perfrin-
gens growth, such as NaCl concentration, pH, and water
activity (aw). Therefore, in addition to controlled cooling,
increasing NaCl concentrations, decreasing pH, and lower-
ing aw also decrease proliferation of C. perfringens (7).
Sodium nitrite, an ingredient commonly used in processed
1698 OSTERBAUER ET AL.
TABLE 1. Experimental formulations in boneless ham for phases
1, 2, and 3
Ingredienta Phase 1 Phase 2 Phase 3
Water-ice (50:50 mix) (%) 25.0 25.0 25.0
Salt (NaCl) (%) 2.00 2.25 2.25
Sugar (%) 1.65 1.65 1.65
Sodium tripolyphosphate (%) 0.35 0.35 0.35
Sodium erythorbate (mg/kg)b 547 0–547 0–547
Sodium nitrite (mg/kg)b 200 0–200 100–200
a
All nonmeat ingredients added based on a meat block only
weight.
b
Curing ingredients added on a meat block basis and at various
concentrations per treatment (see Table 2 for treatment details).
meat products to impart a cured color and flavor, control
oxidation, and limit growth of Clostridium botulinum, also
reduces growth of C. perfringens, especially when used in
conjunction with other curing ingredients (8). Sodium
erythorbate and sodium ascorbate, known as cure acceler-
ators, act as reducing agents and increase the rate of nitrite
conversion to nitric oxide, thus increasing development of
the cured meat color (15). Cure accelerators also can
enhance the efficacy of nitrite for controlling C. perfringens
growth (11, 12, 15).
Concentrations of nitrite and cure accelerators (e.g.,
erythorbate and ascorbate) in commercial meat products can
vary widely. The U.S. Department of Agriculture (USDA)
regulates maximum concentrations of both ingredients,
allowing sodium or potassium nitrite at 156 mg/kg in
comminuted products (meat weight basis), nitrite at 200 mg/
kg in massaged, pumped, or immersion cured products, and
erythorbate or ascorbate at a maximum of 547 mg/kg (18).
Most conventionally cured items contain near the maximum
regulatory concentrations of both ingredients; however, a
growing number of RTE products made with alternative
sources of nitrite and ascorbate, namely celery and cherry
powders, often have lower concentrations of formulated nitrite
and cure accelerators than do conventional curing products
(10). Regardless of the source of nitrite or cure accelerator, the
concentration of nitrite or cure accelerator compound
determines their antimicrobial impact on C. perfringens (11).
Taking into account the impacts of both temperature
control and nitrite concentration on C. perfringens out-
growth, the USDA Food Safety Inspection Service (FSIS)
has produced guidance for the meat industry to meet
performance standards for safe cooling of RTE meat and
poultry products (Appendix B) (19). These time and
temperature standards are considered a safe harbor for meat
product stabilization after thermal treatment and are
designed to limit outgrowth of C. perfringens to ,1 log
CFU/g. To minimize the time product spends in the optimal
C. perfringens growth range, the cooling parameters are
bilinear: stage A refers to the time that the internal
temperature is between 54.4 and 26.78C, and stage B refers
to the time between 26.7 and 4.48C. The time allotted in
each stage differs based on whether the product is
formulated with a minimum of 100 mg/kg nitrite. Product
without nitrite should be held for a maximum 1.5 h in stage
A and 5 h in stage B, whereas product with the minimum
J. Food Prot., Vol. 80, No. 10
concentration of ingoing nitrite can be held for up to 5 h in
stage A and 10 h in stage B (19).
Although meat processors design chilling processes to
comply with these performance standards, exceptions
(cooling deviations) can occur. When target time and
temperature standards are not met in the cooling of products,
an assessment of product safety must occur to determine
final disposition. This assessment can be done using
designated process authorities, who utilize detailed temper-
ature data from the deviated product and any of various
available bacterial growth modeling programs to assess the
potential for C. perfringens outgrowth. Although various
models contain time, temperature, and formulation factors
for predicting C. perfringens growth, these models do not
cover every possible scenario (13).
The objective of this study was to examine the effects of
changes in cooling time and temperature and nitrite and
erythorbate concentrations on the potential for C. perfrin-
gens outgrowth during extended chilling beyond the FSIS
performance standards. The study was conducted in three
phases. In phase 1, extensions of either the stage A or the
stage B portions of the bilinear cooling profile were
evaluated with ham formulated with 200 mg/kg nitrite and
547 mg/kg erythorbate. Phases 2 and 3 were designed to
determine the minimum concentrations of nitrite and
erythorbate required to prevent C. perfringens outgrowth
under a worst-case cooling scenario (25 h total) with similar
bilinear cooling profiles.
MATERIALS AND METHODS
Meat preparation and inoculation. Fresh ham muscles were
obtained from a commercial supplier by the University of
Wisconsin–Madison Meat Science and Muscle Biology Laboratory
(Madison, WI), ground through a 3.18-mm grinder plate, and
stored at 48C until use (within 3 days).
For phase 1, a single large batch of meat plus nonmeat
ingredients (NaCl, sodium tripolyphosphate, sucrose, and water-ice
mixture) was formulated and mixed. This master batch was
subsequently divided into two smaller batches: one batch, denoted
the control, was used directly and contained no curing ingredients,
and the second batch was blended with 200 mg/kg nitrite and 547
mg/kg erythorbate. The ingredients and their proportions are
shown in Table 1. All nonmeat ingredients with half of the water-
ice mixture, except nitrite and erythorbate, were added as a brine.
The meat and brine were mixed together for 2 min in a Hobart
mixer (model AS 200, Hobart Corp., Troy, OH). The meat was
subsequently mixed with three strains of C. perfringens spores
(ATCC 12915, ATCC 12916, and ATCC 13124) in equal levels.
The mixture was adjusted to yield an average initial C. perfringens
level of 2 to 3 log CFU/g of product by applying a 1% (v/w)
inoculum based on weight of the product using the procedure
described by Kennedy et al. (9) and mixing for 2 min. The nitrite
and erythorbate were dissolved in the remaining half of the water-
ice mixture and then added to the batch and mixed for 2 min. After
inoculation and cooking, both batches of meat were divided into
the five chilling treatment groups (Fig. 1A).
Phases 2 and 3 were designed to elucidate the effect of
different concentrations of nitrite and erythorbate on C. perfrin-
gens growth potential. In phase 2, the concentrations of added
nitrite incrementally ranged from 0 to 200 mg/kg, but all
treatments included the same concentration of erythorbate (547
J. Food Prot., Vol. 80, No. 10 CONTROLLING C. PERFRINGENS GROWTH DURING COOLING OF HAM 1699

FIGURE 1. Plot of cooling profiles (A)

and growth curves (B) of C. perfringens in
phase 1 for extended cooling (20 h) in
hams. Each cooling treatment was com-
posed of two time and temperature profiles:
stage A (cooling from 54.4 to 26.78 C) and
stage B (cooling from 26.7 to 4.48 C). Cured
hams were formulated with 200 mg/kg
sodium nitrite and 547 mg/kg sodium
erythorbate on a meat weight basis, an
average 75.6% moisture, pH 6.2, and 1.8%
NaCl. The mean (6SD) C. perfringens
inocula were 2.67 6 0.27 log CFU/g for
cured ham and 3.36 6 0.08 log CFU/g for
uncured control ham.

mg/kg). The phase 3 experiment was designed to evaluate the going nitrite (100 mg/kg), and the other two treatments had
impact various erythorbate concentrations on C. perfringens various concentrations of both nitrite and erythorbate (Tables 1
growth. Thus, four of the six treatments had various concentrations and 2). All curing ingredients were added on a meat-weight-only
of erythorbate (0 to 547 mg/kg) with a fixed concentration of in- basis. A breakdown of phase 2 and phase 3 treatments is shown in
Table 2. Phase 2 and phase 3 treatments were produced the same
TABLE 2. Sodium nitrite and sodium erythorbate concentrations
formulated for ham in phases 2 and 3 way as were treatments in phase 1 and included a control without
nitrite or erythorbate.
Phase 2 Phase 3 All inoculated treatment meat mixtures for phases 1, 2, and 3
Nitrite Erythorbate Nitrite Erythorbate
were divided into 50-g samples, vacuum sealed in oxygen- and
Treatment (mg/kg) (mg/kg) (mg/kg) (mg/kg) moisture-impermeable bags (3 mil high barrier pouches; oxygen
transmission rate 50 to 70 cm3/m2, 24 h at 258C and 60% relative
1 0 0 100 0 humidity; water transmission rate 6 to 7.5 g/m2, 24 h at 258C and
2 50 547 100 250 90% relative humidity; UltraSource, Kansas City, MO) using a
3 75 547 100 375 vacuum packaging machine (Multivac AGW, Sepp Haggemuller
4 100 547 100 547
KG, Wolfertschewenden, Germany), and pressed to a uniform
5 150 547 150 250
flattened size approximately 3.18 mm thick. Additional uninocu-
6 200 547 200 250
lated samples for phase 1 and inoculated samples for phases 2 and
1700 OSTERBAUER ET AL. J. Food Prot., Vol. 80, No. 10

TABLE 3. Stage A and B cooling rates and sampling intervals for phases 1, 2, and 3
Stage A Stage B
Phase cooling (h) cooling (h) Sampling intervals (h)

Phase 1 control 5 10 0 5 10 15
Phase 1 treatments 5 12.5 0 5 10 15 17.5
5 15 0 5 10 15 17.5 20
7.5 10 0 5 7.5 12.5 17.5
10 10 0 5 7.5 10 15 20
Phases 2 and 3 10 15 0 5 7.5 10 25

3 were vacuum sealed and stored at 48C for proximate analysis.
Samples were also stored at 808C for residual nitrite analysis.
Thermocouple probes (traceable thermometer and type K
probe, Thermo Fisher Scientific, Waltham, MA) were inserted into
individual packaged meat samples (five for phase 1 and three for
phases 2 and 3) using a rubber septum (Chembond round patches,
Tru-Flate, Plews & Edelmann, Dixon, IL) to monitor temperatures
in real time during cooking and every 30 min during cooling. Data
loggers (123iButton, Maxim Integrated, San Jose, CA) were also
placed into meat samples (10 for phase 1 and 3 for phases 2 and 3)
to record internal meat temperatures every 5 min throughout the
process.
Product packages were attached with small binder clips to
removable incubator racks. Samples were randomly distributed on
the racks and stored at 48C overnight before cooking and chilling
(approximately 12 to 15 h).
Cooking, cooling, and sampling. Prepared packages were
cooked (3 to 5 min) by immersion into a 758C water bath until the
internal meat temperature reached 728C. These cooking tempera-
tures were sufficient to heat activate the spores and kill vegetative
cells. Cooked samples were immediately transferred into a
programmable air incubator (model BOD50A16 incubator, Revco,
Thermo Electron Corp., Asheville, NC; model UP550 program
controller, Yokogawa Electric, Tokyo, Japan) where samples were
allowed to equilibrate for 5 to 10 min to 54.48C before starting the
test. The program was set to cool the samples according to the
design of the scenarios described below.
Biphasic cooling consisted of dividing the cooling profiles
into two segments: stage A (cooling from 54.4 to 26.68C) and stage
B (cooling from 26.68C to 4.48C). Cooling scenarios were chosen
to extend the time for either stage A (control for 5 h or extended for
7.5 or 10 h) or stage B (control for 10 h or extended for 12.5 or 15
h) for phase 1 (Fig. 1A; maximum 20 h of cooling) and for both
stages for phases 2 and 3 (10 h for stage A and 15 h for stage B, for
a total of 25 h of cooling).
Triplicate inoculated samples from each formulation were
removed from the incubator and placed in an ice water bath for C.
perfringens enumeration as described below. Treatments were
assayed at zero time after cooking, at the end of each cooling stage,
and at appropriate intermediate intervals (Table 3).
From each 50-g sample, a representative 25-g sample was
homogenized in Whirl-Pak bags (3 ml, high barrier; Nasco, Fort
Atkinson, WI) with 50 mL of Butterfield’s buffer (pH 7.1 6 0.1) in
a lab blender (Stomacher 400, A.J. Seward, London, UK) for 1 to 2
min at 230 rpm. Serial dilutions were plated onto tryptose-sulfite-
cycloserine (TSC; Oxoid, Basingstoke, UK) agar plates with a TSC
overlay. Plates were anaerobically incubated (AnaeroPack System
7.0 L jar with Pack-Anaero Anaerobic Gas Generating System,
Mitsubishi Gas Chemical Co., Tokyo, Japan) at 358C for 24 h,
when colonies were counted and recorded. Phase 1 was replicated
twice, and phases 2 and 3 were replicated three times using
different batches of ham and different spore levels. The lower
detection limit by direct plating was 1.48 log CFU/g.
Proximate analysis. Triplicate samples for each treatment
were analyzed for pH, aw, NaCl, moisture, and nitrite. The pH of
each sample was obtained by homogenizing a representative 10-g
meat sample into 90 mL of deionized water in the Stomacher 400,
and pH of the slurry was measured (Accumet Basic pH meter and
Orion 8104 combination electrode, Thermo Fisher Scientific). The
aw was measured using an aw meter (AquaLab 4TE, Decagon,
Pullman, WA). NaCl concentration was determined by AgNO3
potentiometric titration as the percentage of chloride ions (model
DL22 food and beverage analyzer, Mettler, Columbus, OH).
Procedures from the Association of Official Analytical Chemists
(1) were used to determine moisture content (5 h at 1008C, vacuum
oven method 950.46) and residual nitrite (colorimetric method
973.31). Residual sodium nitrite was tested in all treatments
containing nitrite.
Statistical analysis. Two replications for phase 1 and three
replications for phases 2 and 3 were completed. Growth means and
standard deviations (SDs) for each time point were log transformed
using Excel (Microsoft, Redmond, WA). Log increases over time
for all treatments in phases 1, 2, and 3 were analyzed using least
square means techniques and the GLM procedure of Minitab
17.3.1 (2013 and 2016, Minitab, Inc., State College, PA). Main
effects of nitrite and erythorbate and their interactions were
considered significant at P , 0.05. Growth was defined as a .1-
log increase over the initial inoculum level (20).

TABLE 4. Proximate analyses of samples in phases 1, 2, and 3a

Phase pH NaCl (%) Water activity Moisture (%) Residual nitrite (mg/kg)b

1 6.21 6 0.04 1.80 6 0.17 0.980 6 0.001 75.63 6 0.01 120.0 6 8.2
2 6.15 6 0.09 2.10 6 0.16 0.976 6 0.002 75.44 6 1.27 (see Table 5)
3 6.08 6 0.21 1.75 6 0.22 0.973 6 0.003 75.65 6 0.50 (see Table 5)
a
Values are mean 6 SD.
b
Residual nitrite for phase 1 was averaged over the six treatments in that phase. Residual nitrite values for phases 2 and 3 are displayed by
treatment in Table 5.
J. Food Prot., Vol. 80, No. 10 CONTROLLING C. PERFRINGENS GROWTH DURING COOLING OF HAM 1701

TABLE 5. Residual nitrite calculated for phases 2 and 3a (after cooking) were 3.31 6 0.08 and 2.67 6 0.27 log CFU/
g for uncured and cured phase 1 ham, respectively (Fig. 1B).
Phase 2 Phase 3
Phase 1 consisted of cured ham with 200 mg/kg ingoing
Residual nitrite Residual nitrite nitrite and 547 mg/kg erythorbate, and treatments followed
Formulationb (mg/kg) Formulationc (mg/kg) five different cooling curves: the FSIS guideline (5-h stage A
and 10-h stage B), extended stage A to 7.5 or 10 h, and
0 100 with 0 73.7 6 12.6 B
extended stage B to 12.5 or 15 h (Fig. 1A). No growth
50 32.8 6 3.9 A 100 with 250 75.7 6 38.5 B
(defined as ,1-log increase) was observed in any cured
75 48.5 6 2.8 B 100 with 375 70.8 6 12.6 B
100 63.0 6 5.1 C 100 with 547 60.8 6 13.1 B
treatment samples for phase 1 through 20 h regardless of the
150 102.1 6 14.1 D 150 with 250 100.9 6 20.2 A cooling profile. In contrast, the uncured control supported a
200 131.9 6 10.7 E 200 with 250 149.4 6 42.8 A 4.31-log increase at the end of the 5-h stage A (54.4 to
26.68C) and an additional 1.2-log increase at the end of the
a
Within a column, means followed by different letters are 10-h stage B (26.6 to 4.48C) (Fig. 1B). The results were not
significantly different (P , 0.05). surprising, because the inhibition of C. perfringens by nitrite
b
Given as mg/kg nitrite.
c has been well described. For example, Redondo-Solano et
Given as mg/kg nitrite with mg/kg erythorbate.
al. (14) reported a decline in recoverable C. perfringens
when 200 mg/kg nitrite and 547 mg/kg erythorbate were
RESULTS AND DISCUSSION utilized in tandem with a 15-h exponential cooling curve
Proximate analysis and temperature monitoring. from 54.4 to 4.48C. These data also suggest that the number
Proximate analyses for phases 1, 2, and 3 are presented in of hours for cooling may be extended in processed meat
Table 4, and residual nitrite concentrations for phases 2 and products (using ham as a model) without additional concern
3 are presented in Table 5. Because phase 1 treatments were for growth of C. perfringens. In the present study, ham
all conducted with the same formulation, no testing between formulated with 75.6% moisture, 1.8% NaCl, and pH 6.21
treatments was conducted, and the residual nitrite concen- with the maximum concentrations of nitrite (200 mg/kg) and
tration is listed in Table 4. erythorbate (547 mg/kg) inhibited growth of C. perfringens
Proximate analysis values for phase 2 for pH (6.15 6 during 20 h of cooling, even when stage A cooling (optimal
0.09), NaCl (2.10% 6 0.16%), aw (0.976 6 0.002), and growth range) was extended to 10 h.
moisture (75.44% 6 1.27%) were also consistent between
Effect of nitrite concentration during 25 h of
treatments and were similar to those values observed in
cooling. In phase 1 treatments, the addition of 200 mg/kg
phase 1 treatments. As expected, residual nitrite for phase 2
nitrite inhibited growth of C. perfringens when either stage
was dependent on ingoing nitrite concentrations. Residual
A or stage B cooling was extended (total of 15 to 20 h).
nitrite was 63 to 66% of the ingoing formulated nitrite. Such
Phase 2 was conducted to determine the effect of various
changes in nitrite are typical of cured processed meat
nitrite concentrations (50 to 200 mg/kg) with a single fixed
products (2) in which up to 50% of ingoing nitrite is lost
concentration of erythorbate (547 mg/kg) on growth during
during processing.
a 25-h bilinear cooling process (both stage A and B cooling
Phase 3 treatments had results similar to those in phase
extended). Postcooking populations of C. perfringens were
2 for NaCl (1.75% 6 0.22%), aw (0.973 6 0.003), and 2.05 6 0.31 log CFU/g regardless of cure addition. In this
moisture (75.65% 6 0.50%). The pH range (6.08 6 0.21) set of experiments, the ability of C perfringens to grow in
was greater between replicate experiments than that in cured ham during the chilling curve described was
previous trials in phase 2 (5.75 to 6.3 in phase 3). Residual dependent on the concentration of ingoing nitrite (Fig. 2).
nitrite concentrations were 61 to 149 mg/kg and were again Compared with inoculation levels, populations of C.
dependent on ingoing nitrite concentrations. However, in perfringens increased 4.3, 2.7, and 0.6 log CFU/g within 5
ham treatments with lower concentrations of erythorbate h for the 0, 50, and 75 mg/kg nitrite treatments, respectively,
residual nitrite concentrations were higher than the ingoing but no growth was observed for the treatments formulated
concentrations. Recoveries ranged from 60.8% for the 100 with 100 mg/kg nitrite during the same cooling time. At
mg/kg nitrite with 547 mg/kg erythorbate treatment to the end of stage A cooling (10 h), compared with inoculation
75.7% for the 100 mg/kg nitrite with 250 mg/kg erythorbate levels, populations of C. perfringens increased 6.6, 4.0, 2.3,
treatment. Similar trends were reported in pork sausage, but and 0.5 log CFU/g for the 0, 50, 75, and 100 mg/kg nitrite
the recovery values were of a lower magnitude (10). treatments, respectively. No additional growth was observed
For product cooling overall, temperature measurements for any treatment during stage B cooling (26.7 to 4.48C).
were consistent between iButtons and thermocouples and These results are consistent with findings of other research-
matched closely the target temperatures for cooling (data not ers who observed dose-dependent inhibition of C. perfrin-
shown). Temperatures during cooling were consistent gens at high nitrite concentrations and 547 mg/kg
among replicate samples within each trial and between trials erythorbate (11, 14).
(data not shown).
Effect of nitrite and erythorbate combinations
Effect of extended stage A and stage B cooling. during 25 h of cooling. In phase 2, 100 mg/kg nitrite was
Mean (6SD) populations of C. perfringens at zero time sufficient for extended cooling when used with a fixed
1702 OSTERBAUER ET AL.
FIGURE 2. Mean populations (log CFU/g) of C. perfringens in phase 2 in ham samples with various nitrite concentrations during a 25-h
two-stage bilinear cooling profile. All formulations included 547 mg/kg sodium erythorbate, 75.4% moisture, pH 6.15, and 2.2% salt.
Values are the mean of three trials; error bars represent the SD (n ¼ 9). C. perfringens inoculum for all treatments was 2.05 6 0.31 log
CFU/g.
concentration of erythorbate (547 mg/kg). To better
understand the relationship between curing ingredients and
C. perfringens control, phase 3 was conducted to examine
various nitrite with erythorbate concentrations, respectively,
of 100 mg/kg with 0 mg/kg, 100 with 250, 100 with 375,
100 with 547, 150 with 250, and 200 with 250 following the
same 25-h cooling curve as phase 2 (stage A, 54.4 to 26.78C
in 10 h; stage B, 26.7 to 4.48C in 15 h).
The mean (6SD) C. perfringens inoculum for all
treatments was 2.39 6 0.27 log CFU/g. Changes in C.
perfringens populations from inoculated levels after 25 h of
bilinear cooling for treatments with nitrite with erythorbate
concentrations, respectively, of 100 mg/kg with 0 mg/kg,
100 with 250, 100 with 375, 100 with 547, 150 with 250,
and 200 with 250 were 4.9, 2.8, 2.3, 1.6, 0.7, and 0.60 log
CFU/g (Fig. 3). Similar to phase 2, mean populations of C.
perfringens did not significantly increase (P . 0.05) after
the first stage of cooling (Fig. 3; data not shown). As
expected, increasing the concentration of erythorbate in
combination with nitrite resulted in greater inhibition of C.
perfringens during cooling. With 100 mg/kg nitrite and no
erythorbate, C. perfringens populations increased in a
manner similar to that of the uncured (no nitrite added)
control in phase 2 (6.75 log CFU/g), indicating that
erythorbate is an important component in cured products
as a cure accelerator and antimicrobial agent for limiting
growth of C. perfringens. As increasing concentrations of
erythorbate were added to the 100 mg/kg nitrite treatments,
total growth decreased, further illustrating the combined
effect of erythorbate and nitrite (Fig. 3).
Although the treatment containing 100 mg/kg nitrite
with 547 mg/kg erythorbate in phase 2 inhibited growth of
J. Food Prot., Vol. 80, No. 10
C. perfringens to ,1 log CFU/g over 25 h (i.e., 0.5-log
increase; Fig. 2), 1.6-log growth was observed for this same
treatment combination when included and repeated in phase
3 experiments (Fig. 3). A closer analysis of growth for the
three individual trials of this repeated treatment in phase 3
revealed variation in overall population changes (Fig. 4).
Reviewing proximate analyses for all replicates of this
treatment indicated variation in pH and NaCl concentration,
suggesting a possible explanation for the observed C.
perfringens growth disparity. Although moisture levels were
consistent among all samples (75.49% 6 0.87%), the
differences in pH and NaCl concentrations were identified
as potential factors affecting growth. Although the three
individual replicate phase 3 trials had a mean (6SD) pH of
6.08 6 0.21 and 1.75% 6 0.22% NaCl when analyzing
individually, the pH values for the three individual replicates
were 5.75 to 6.30 and NaCl concentrations were 1.32 to
2.13%. When reviewing microbiological data for the
individual replicates, C. perfringens populations increased
2.7, 1.5, and 0.2 log CFU/g from inoculation levels for the
treatments combinations of pH 6.30 with 1.3% NaCl, pH
6.15 with 1.8% NaCl, and pH 5.75 with 1.5% NaCl,
respectively, compared with a 0.5-log increase for the mean
for pH 6.15 with 2.15% NaCl. The differences in growth
observed among the replicates of this treatment reveal that
pH and NaCl concentrations should also be considered when
evaluating the safety of cooling deviations when products
are formulated with 100 mg/kg nitrite and 547 mg/kg
erythorbate.
NaCl and pH are factors known to affect C. perfringens
(7) but are typically not included in critical control limits
during production of perishable RTE meats. The results
J. Food Prot., Vol. 80, No. 10 CONTROLLING C. PERFRINGENS GROWTH DURING COOLING OF HAM 1703

FIGURE 3. Mean populations (log CFU/g) of C. perfringens in phase 3 ham samples with varying sodium nitrite concentrations (100 to 200
mg/kg) and varying sodium erythorbate concentrations (0 to 547 mg/kg) examined over a 25-h two-stage bilinear cooling profile. Values are
the mean of three trials; error bars represent the SD (n ¼ 9). C. perfringens inoculum for all treatments was 2.39 6 0.27 log CFU/g.

from these experiments suggest that pH and NaCl may affect
the growth potential of C. perfringens in ham (or other cured
meat products) formulated with 100 mg/kg nitrite and 547
mg/kg erythorbate. These analytical differences reflect
commercial production based on differences in incoming
ingredients and processing rather than on controlled
experimental variation, and the 25-h cooling regime is
greatly exaggerated from the FSIS 15-h cooling require-
ments for cured meat products.
When 150 or 200 mg/kg nitrite was used with lower
concentrations of erythorbate (250 mg/kg), growth of C.
perfringens was limited to ,1 log CFU/g, suggesting that

FIGURE 4. Comparison of C. perfringens growth for trials for the treatment with 100 mg/kg nitrite and 547 mg/kg erythorbate (phase 3:
three individual trials, pH-salt combination identified in figure; phase 2: average of three independent trials with 6.15 pH and 2.2% NaCl)
over the 25-h cooling time. Note the influence on growth as the meat pH increases and salt concentration decreases.
1704 OSTERBAUER ET AL.
physiochemical properties (e.g., NaCl and pH) have less of
an impact on C. perfringens at nitrite and erythorbate
concentrations 150 and 250 mg/kg, respectively (Fig. 3).
Additional research is needed to better understand this
relationship. With these nitrite concentrations (150 or 200
mg/kg), erythorbate concentrations could be lowered from
the regulatory maximum of 547 mg/kg and without affecting
microbial inhibition, even during cooling deviations.
Consequently, these data can be used as support of safety
for (i) extended cooling processes or (ii) cooling deviations
for product formulations similar to those described herein.
Although this study utilized boneless ham for the experi-
mental growth medium, growth or control of C. perfringens
is not likely dependent on product or meat species.
Therefore, similar results could be expected during the
extended cooling of other meat and poultry products.
ACKNOWLEDGMENTS
We thank Amanda Skarlupka, Max Golden, Brandon Wanless, Ming
Mu, and Nicole Baker (Food Research Institute Laboratory, University of
Wisconsin–Madison) for their technical support in conducting experiments.
Research was supported by the University of Wisconsin–Madison Meat
Science and Muscle Biology Laboratory and the Food Research Institute.
REFERENCES
1. Association of Official Analytical Chemists. 1990. Official methods
of analysis, 15th ed. Association of Official Analytical Chemists,
Washington, DC.
2. Cassens, R. C. 1979. Reactions of nitrite in meat. Food Technol.
33:46–57.
3. Centers for Disease Control and Prevention. 2014. Surveillance for
foodborne disease outbreaks. United States, 2012, annual report.
Centers for Disease Control and Prevention, Atlanta.
4. Craven, S. E., and L. C. Blankenship. 1985. Activation and injury of
Clostridium perfringens spores by alcohols. Appl. Environ. Microbiol.
50:249–256.
5. International Commission on Microbiological Specifications for
Foods. 1996. Microorganisms in foods. 5. Characteristics of microbial
pathogens. Blackie Academic, London.
6. Juneja, V. K., L. Huang, and H. H. Thippareddi. 2006. Predictive
model for growth of Clostridium perfringens in cooked cured pork.
Int. J. Food Microbiol. 110:85–92. doi:10.1016/ijfoodmicro.2006.01.
038.
7. Juneja, V. K., B. S. Marmer, J. G. Phillips, and S. A. Palumbo. 1996.
Interactive effects of temperature, initial pH, sodium chloride, and
sodium pyrophosphate on the growth kinetics of Clostridium
perfringens. J. Food Prot. 59:963–968.
8. Juneja, V. K., J. S. Novak, and R. J. Labbe. 2010. Clostridium
perfringens, p. 53–70. In V. K. Juneja, J. S. Novak, and R. J. Labbe
(ed.), Pathogens and toxins in foods: challenges and interventions, 1st
ed. ASM Press, Washington, DC.
J. Food Prot., Vol. 80, No. 10
9. Kennedy, K. M., A. L. Milkowski, and K. A. Glass. 2013. Inhibition
of Clostridium perfringens growth by potassium lactate during an
extended cooling of cooked uncured ground turkey breasts. J. Food
Prot. 76:1972–1976.
10. King, A. M., K. A. Glass, A. L. Milkowski, D. L. Seman, and J. J.
Sindelar. 2016. Modeling the impact of ingoing sodium nitrite,
sodium ascorbate, and residual nitrite concentrations on growth
parameters of Listeria monocytogenes in cooked, cured pork sausage.
J. Food Prot. 79:184–193.
11. King, A. M., K. A. Glass, A. L. Milkowski, and J. J. Sindelar. 2015.
Comparison of the effect of curing ingredients derived from purified
and natural sources on inhibition of Clostridium perfringens
outgrowth during cooling of deli-style turkey breast. J. Food Prot.
78:1527–1535.
12. King, A. M., K. A. Glass, A. L. Milkowski, and J. J. Sindelar. 2015.
Impact of clean-label antimicrobials and nitrite derived from natural
sources on the outgrowth of Clostridium perfringens during cooling
of deli-style turkey breast. J. Food Prot. 78:946–953.
13. Mohr, T. B., V. K. Juneja, H. H. Thippareddi, D. W. Schaffner, P. A.
Bronstein, M. Silverman, and L. V. Cook, Jr. 2015. Assessing the
performance of Clostridium perfringens cooling models for cooked,
uncured meat and poultry products. J. Food Prot. 78:1512–1526.
doi:10.4315/0362-028X.JFP-15-015.
14. Redondo-Solano, M., C. Valenzuela-Martinez, D. A. Cassada, D. D.
Snow, V. K. Juneja, D. E. Burson, and H. H. Thippareddi. 2013.
Effect of meat ingredients (sodium nitrite and erythorbate) and
processing (vacuum storage and packaging atmosphere) on germina-
tion and outgrowth of Clostridium perfringens spores in ham during
abusive cooling. Food Microbiol. 35:108–115. doi:10.1016/j.fm.
2013.02.008.
15. Sauter, E. A., J. D. Kemp, and B. E. Langlois. 2006. Effect of nitrite
and erythorbate on recovery of Clostridium perfringens spores in
cured pork. J. Food Sci. 42:1678–1679.
16. Scallan, E., P. M. Griffin, F. J. Augulo, R. V. Tauxe, and R. M.
Hoekstra. 2011. Foodborne illness acquired in the United States—
unspecified agents. Emerg. Infect. Dis. 17:16–22. doi:10.3201/eid170.
P21101.
17. Schroder, D. J., and F. F. Busta. 1971. Growth of Clostridium
perfringens in meat loaf with and without added soybean protein. J.
Mill Food Technol. 34:215–217.
18. U.S. Department of Agriculture, Food Safety and Inspection Service.
1995. Processing inspectors’ calculations handbook. Food Safety and
Inspection Service directive 7620.3. U.S. Government Printing Office,
Washington, DC.
19. U.S. Department of Agriculture, Food Safety and Inspection Service.
1999. Performance standards for the production of certain meat and
poultry products, appendix B. Compliance guidelines for cooling
heat-treated meat and poultry products (stabilization). U.S. Depart-
ment of Agriculture, Food Safety and Inspection Service, Wash-
ington, DC. Available at: https://www.fsis.usda.gov/OPPDE/rdad/
FRPubs/95-033F/95-033F_Appendix%20B.htm. Accessed 3 March
2017.
20. U.S. Food and Drug Administration. 2012. Clostridium perfringens,
p. 83–86. In Bad bug book: foodborne pathogenic microorganisms
and natural toxins, 2nd ed. U.S. Food and Drug Administration, Silver
Spring, MD.