If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
Department of Food Science and Technology, University of Nebraska, Lincoln 68588Department of Biological Systems Engineering, University of Nebraska, Lincoln 68583Department of Food Science and Technology, University of Arkansas, Fayetteville 72704
While the increase in thermal resistance of microorganisms at reduced water activity is demonstrated for low-moisture food products, the effect of storage time on the thermal resistance of microorganisms in low-moisture foods is not well established. As low-moisture foods are stored for long periods and are used as ingredients, cross-contamination can occur at any time period before the lethality step. Therefore, this study was designed to investigate the effect of storage time (30, 60, and 90 d) on the thermal resistance of Salmonella and Enterococcus faecium NRRL B-2354 in milk powders at a low water activity of 0.10 (conservative level). In this study, 2 milk powders, whole milk powder (WMP) and nonfat dry milk (NFDM), were inoculated with a 5-serotype Salmonella cocktail or E. faecium and equilibrated to a water activity of 0.10. The thermal resistance of Salmonella and E. faecium in WMP and NFDM were determined at different storage times (30, 60, and 90 d) at 85°C. The storage time had no effect on the thermal inactivation kinetics of Salmonella within 90 d of storage at 85°C. In the second part of this study, isothermal treatments were also conducted at higher temperatures (90 and 95°C) to evaluate the suitability of E. faecium as a surrogate for Salmonella in milk powders. The D-values of Salmonella in WMP with 30 d of storage at 85, 90, and 95°C were 7.98, 3.35, and 1.68 min. The corresponding values for E. faecium were 16.96, 7.90, and 4.16 min. Higher D-values of E. faecium indicates that it is a conservative surrogate. Similar results were found for NFDM. In general, D-values of both microorganisms are slightly higher in NFDM than WMP. Two primary models (log-linear and Weibull) were compared for their goodness-of-fit. The Weibull model was found to be more appropriate than the log-linear model. This study provides valuable information for establishing process validation for the pasteurization of milk powders.
). Most of the foodborne pathogens require water activity >0.9 for proper growth; thus, they were historically not a concern in low water activity foods. However, the survival of Salmonella in low water activity foods, such as almond (
Salmonella Montevideo infections associated with salami products made with contaminated imported black and red pepper — United States, July 2009-April 2010.
), have resulted in several foodborne outbreaks in the United States, indicating that low water activity food can no longer be considered as inherently microbiologically safe.
Milk is a nutrient-rich liquid that is often spray-dried into powder forms for a longer shelf-life. Milk powders are commonly used as ingredients in many food products including ready-to-eat foods, such as confectionery, drink mixes, seasoning, nutritional bars, and dry blend infant formula. However, no pathogen lethality step is applied to eliminate foodborne pathogens in milk powders after spray drying (
Salmonella survival during spray drying and subsequent handling of skimmilk powder. III. Effects of storage temperature on Salmonella and dried milk properties.
), which could potentially result in serious food safety issues. For instance, 37 babies in France were known to have fallen sick due to the presence of Salmonella Agona in powdered milk in 2018 (
After spray drying of milk powders, there is a risk of cross-contamination through poor postprocessing practices, such as poor environmental sanitation, inappropriate storage conditions, or mishandling during transportation (
). Before further processing of milk powders, especially for ready-to-eat foods, an effective pasteurization process should be integrated to minimize the risk of contamination of final products.
, validation studies are required to scientifically demonstrate that processing technologies effectively limit the hazards they are designed to address. Also, due to the difficulty of directly introducing the target pathogen into the processing plant, the use of a surrogate to replace the pathogen is recommended by
) have previously demonstrated that Enterococcus faecium NRRL B-2354 is a suitable surrogate for Salmonella during thermal processing of low water activity foods.
In commercial practice, ingredients such as milk powders are usually stored at 23°C for a long period before usage. It has been reported that high storage temperature (45°C) could greatly affect lipid oxidation in whole milk powder (WMP), and the maximum level of free radicals was detected after 47 d of storage at 45°C (
, the typical storage time for WMP is 6 to 9 mo at room conditions (<27°C, <65% relative humidity). Flavor deterioration of WMP has been reported to occur after 3 mo of storage (
). However, the effect of storage time on the thermal inactivation kinetics of Salmonella in milk powders is currently unknown. Therefore, it is important to understand the effect of storage time on thermal inactivation kinetics for process validation studies. It is also important to evaluate the suitability of surrogate at different storage times and temperatures.
The overall goal of this study is to evaluate the effect of storage time on thermal inactivation kinetics of Salmonella and E. faecium NRRL B-2354 in milk powders. The specific objectives were to (1) determine the effect of storage time (30, 60, and 90 d) on the thermal inactivation kinetics of Salmonella and E. faecium NRRL B-2354 in milk powders at 85°C and water activity of 0.10, and (2) evaluate E. faecium NRRL B-2354 as a surrogate for Salmonella at different temperatures (85, 90, and 95°C) after 30 d of storage of milk powders at a water activity of 0.10.
MATERIALS AND METHODS
Milk Powder Products
Three batches of grade A pasteurized WMP and nonfat dry milk (NFDM) from different production lots were received from Mars Inc. (McLean, VA). The milk powders were stored inside polyethylene bags and held at ambient temperature (23 ± 2°C) before inoculation. On receiving the samples, the proximate analysis and background microorganism tests were conducted as described in
. A water activity meter (4TE, METER Group, Pullman, WA) was used to measure the water activity of milk powders at 25°C.
Inoculum Preparation
The details of the bacterial serotypes used in this study are described in Table 1. A Salmonella cocktail, consisting of 5 serotypes of Salmonella enterica ssp. enterica that were previously isolated from different foodborne outbreaks, was used to inoculate both powder samples. Enterococcus faecium NRRL B-2354 was used to inoculate both powder samples for surrogate tests.
Salmonella Montevideo infections associated with salami products made with contaminated imported black and red pepper — United States, July 2009-April 2010.
, which yields a stable bacterial population before isothermal treatment and consistent D-value. The frozen stock of each bacteria was thawed at 37°C for 5 min, then was subjected to 2 sequential transfers (24 h at 37°C) in 10 mL of trypticase soy broth (TSB; Becton, Dickinson and Company, Franklin Lakes, NJ) with 0.6% (wt/vol) yeast extract (YE; Becton, Dickinson and Company). A loop (~10 µL) of the incubated culture was streaked onto tryptic soy agar (TSA; Becton, Dickinson and Company) supplemented with 0.6% (wt/wt) YE and incubated at 37°C for 24 h to prepare the working stock plate. A 10-µL loop was used to harvest one isolated colony from a working stock plate and was then transferred to 10 mL of TSBYE and incubated for 24 h at 37°C. The incubated culture (0.1 mL) was then spread-plated onto a TSAYE plate and incubated for 24 h at 37°C. The grown bacterial lawn was then harvested with 3 mL of 0.1% buffered peptone water (Becton, Dickinson and Company) with an L-shaped spreader. By mixing the equal amount of lawn from each Salmonella serotypes, the Salmonella cocktail inoculum was prepared with the bacterial population of approximately 1,010 cfu/mL. Similarly, E. faecium NRRL B-2354 inoculum was prepared with the bacterial population of approximately 109 cfu/mL.
Milk Powder Inoculation and Equilibration
The WMP or NFDM from each batch was weighed (100 g) and transferred to a resealable 1-gallon polyethylene bag. A sterile 15-mL centrifuge tube (Thermo Fisher Scientific, Rochester, NY) with a finger-operated spray head contained a 10-mL aliquot of the prepared Salmonella cocktail or E. faecium NRRL B-2354 inoculum was used to spray-inoculate both milk powder samples. The inoculated sample was then mixed by hand for 5 min to manually detach powders that were stuck to the inner lining of the bag. Subsequently, the sample was transferred into a sterile Whirl-Pak bag (2.04 L, Nasco, Fort Atkinson, WI) and then placed in a paddle mixer (9000471, Neutec Group Inc., Farmingdale, NY) and mixed for 20 min to homogenize the inoculum distribution and minimize the clumps. The mixed inoculated sample (5 mm thickness) was then placed on a sterile aluminum tray (230 × 300 × 15 mm) and transferred into a custom-made relative humidity chamber (
), a low water activity of 0.10 was selected, which could represent the worst-case scenario where Salmonella exhibits high thermal resistance. This low water activity (0.10) represents a conservative study, as Salmonella resistance is higher at lower water activity. After 5 d of equilibration in the chamber, the inoculated sample was packed into heat-sealable aluminized pouches (7.62 cm × 7.62 cm × 0.10 cm, IMPAK Corporation, Los Angeles, CA) in quantities of approximately 2 g and then sealed for storage at 23°C for 30, 60, and 90 d. The whole packing process was conducted inside the relative humidity chamber to ensure a minimal change in the water activity.
Determination of Thermal Resistance at Different Storage Times
Stored inoculated milk samples were sampled on 30, 60, and 90 d to evaluate the effect of storage time on the heat resistance of Salmonella and E. faecium. Isothermal treatment was conducted at 85°C by using the custom-designed thermal death time (TDT) sandwich (
) for both WMP and NFDM to obtain the isothermal death curve of Salmonella and E. faecium at 85°C. The TDT sandwich has 2 resistive heating elements that could apply dry heat to samples. Each packed inoculated sample was placed in between the 2 heating elements in one TDT sandwich unit for isothermal treatment of one time and temperature combination.
Before conducting the isothermal treatment, the come-up time was measured in triplicates by a type T thermocouple (5TC-TT-T-40–36, Omega Engineering Inc., Norwalk, CT) with an accuracy of ±0.5°C inserted to the inside center of an aluminized pouch. The come-up time was the average time required for the sample to be heated to 84.5°C by the TDT sandwiches plus twice the standard deviation, which would ensure that at least 95% of TDT sandwiches would reach at least 84.5°C at time zero in the inactivation curve.
Six equally spaced time points including time zero were used for the isothermal treatment of inoculated samples to obtain thermal death curves for Salmonella or E. faecium in both WMP and NFDM. Immediately after each treatment, the treated pouch was transferred into an ice-water bath (0°C) for at least 1 min before it was enumerated for bacterial survivors.
Surrogate Evaluation at Different Temperatures
To evaluate the suitability of E. faecium as a surrogate for Salmonella at different temperatures, isothermal treatments at the higher temperatures (90 and 95°C) were conducted for Salmonella or E. faecium in inoculated milk powders. Based on the results of the storage study as described in the previous section, a single storage time was used to determine the thermal inactivation kinetics for both bacteria at higher temperatures. Inoculated samples were isothermally treated by the TDT sandwiches as described in the previous section to obtain thermal inactivation curves for Salmonella or E. faecium in both WMP and NFDM at 90 and 95°C. All isothermal treatments for inoculated WMP and NFDM were performed in 3 independent replicates by using inoculated samples from each batch.
Bacterial Enumeration
Enumeration was performed as previously described (
). Briefly, the Salmonella or E. faecium survivors in the thermally treated sample were enumerated by serially 10-fold diluting samples in 0.1% buffered peptone water and were then spread-plated onto TSAYE supplemented with 0.05% (wt/vol) ammonium iron citrate (Sigma-Aldrich Co., St. Louis, MO), and 0.03% (wt/vol) sodium thiosulfate (mTSA; Fisher Chemical, Fair Lawn, NJ) for Salmonella or TSAYE supplemented with 0.05% (wt/vol) ammonium iron citrate, and 0.025% (wt/vol) esculin hydrate (eTSA; Acros Organics, Morris Plains, NJ) for E. faecium and incubated for 24 h at 37°C (
). Colonies with a black center were counted as Salmonella on incubated mTSA plates and black colonies were counted as E. faecium on incubated eTSA plates, respectively. The untreated WMP or NFDM samples were enumerated as a control to monitor the bacterial population change during the 90 d of storage.
Curve Fitting of The Survivor Models
The Salmonella or E. faecium survivors in both milk powders were fitted with 2 primary inactivation models, the log-linear model [1] and the Weibull-type model [2] as described in
where N and N0 (cfu/g) are the Salmonella or E. faecium survivors at time t and 0; t is the instantaneous time of the thermal treatment (min); D is the decimal reduction time, which is the time required to reduce the microbial population by 10-fold; δ (min) is the scale parameter; and β is the shape parameter of the Weibull inactivation curve, where β > 1 indicates a concave downward trend, β < 1 indicates a concave upward trend of the inactivation curve, and β = 1 indicates a linear trend.
The z-value is defined as the increase in temperature required to achieve a 90% reduction of the decimal reduction time D and was determined by the following equation:
[3]
where
are the D-values at temperature T1 and T2, respectively.
The bacterial survival data for each isothermal treatment were fitted to Equations [1] and [2] using the Levenberg–Marquardt algorithm for nonlinear least squares regression as implemented in the curve_fit function of the Python SciPy library (
). The statistical significance of differences between different storage times and different milk powder products were analyzed using Fisher's least significant difference test as implemented in the LSD_test function of the R agricolae library (
The log-linear and Weibull models were also globally fitted across all 3 storage times. To evaluate the appropriateness of model fitting, model selection analyses were performed using root mean squared error (RMSE) and corrected Akaike information criteria (AICc;
where log10(N)observed,i is the observed log population, log10(N)predicted,i is the predicted log population from the model, and n is the total number of observations, SS is the sum of squares of the residuals, and K is the number of parameters being estimated plus 1. Lower RMSE and AICc values indicate that the model prediction is more likely to be correct for the experimental data. According to
, the percent likelihood of a given model is more correct, which can be estimated using the following equation:
[6]
RESULTS AND DISCUSSION
Survival of Salmonella and E. faecium in WMP and NFDM
The bacterial populations of Salmonella and E. faecium in the prepared inoculum were 10.84 and 9.92 log cfu/mL, respectively. Immediately after inoculation, Salmonella and E. faecium populations were determined to be 8.13 and 8.86 log cfu/g in WMP and 8.25 and 8.92 log cfu/g in NFDM, respectively. After 24 h equilibration in the relative humidity chamber, the inoculated samples were equilibrated to the water activity level of 0.10 ± 0.01, and the Salmonella and E. faecium population were dropped to 7.75 and 8.61 log cfu/g in WMP and 7.86 and 8.78 log cfu/g in NFDM, respectively. The standard deviations of 3 subsamples from inoculated WMP and NFDM were determined to be 0.18 and 0.15 log cfu/g, respectively, which indicated a good homogeneity of the inoculated samples.
The survival results of Salmonella and E. faecium in WMP and NFDM are shown in Figure 1. During 90 d of storage, the reductions of Salmonella in WMP and NFDM were observed to be 0.79 and 0.86 log cfu/g, whereas the reductions of E. faecium in WMP and NFDM were observed to be 0.10 and 0.16 log cfu/g (Figure 1). Most of the decrease in the initial population of Salmonella was observed in the first 30 d of storage. The results in this study showed that Salmonella in milk powder experienced a rapid death rate in the first few weeks of storage followed by a slower death rate are in agreement with several published studies (
Salmonella survival during spray drying and subsequent handling of skimmilk powder. III. Effects of storage temperature on Salmonella and dried milk properties.
). A similar reduction in Salmonella population has been observed during the postinoculation process of low water activity foods, such as black pepper (
). Salmonella cells need time to adapt to a low water activity environment. Thus, a higher death rate was observed during the adaption period and then the death rate leveled off in the remaining storage time.
Figure 1Viability of Salmonella and Enterococcus faecium population in whole milk powder (WMP) and nonfat dry milk (NFDM) during storage period (23°C). Error bars indicate 1 SD of 3 subsamples.
, the population of Salmonella in skim milk powder dropped by 4.23, 4.69, and 5.58 log cfu/g after 2 mo of storage at the water activity of 0.33, 0.53, and 0.81, respectively. As the water activity of milk powder is this study was much lower at 0.10, a much lower reduction of 0.86 log cfu/g of the Salmonella population in NFDM after 90 d of storage was observed. Enhanced survivability of Salmonella was observed at lower water activity level during the storage of low water activity foods, as desiccated conditions have been reported to increase survivability (
Survival of Salmonella Tennessee, Salmonella Typhimurium DT104, and Enterococcus faecium in peanut paste formulations at two different levels of water activity and fat.
) have been shown to survive well in milk powders at a reduced water activity level. Therefore, the long-term survival of foodborne pathogens in milk powders should be of concern to the dairy industry and it is important to have a kill step to enhance the safety of milk powders.
Effect of Storage Time on the Thermal Resistance of Salmonella and E. faecium
The effect of storage time on the thermal resistance of Salmonella and E. faecium was investigated at 85°C and at the water activity of 0.1. The low water activity of 0.1 was selected in this study to represent the worst-case scenario, as increased thermal resistance of Salmonella has been reported at reduced water activity levels in both WMP and NFDM (
). The isothermal treatment temperature of 85°C was chosen because there was a minimum quality impact on sensory properties of milk powders at this temperature.
The model parameters of the log-linear model and the Weibull model for both WMP and NFDM were compared across the 3 storage times in Table 2. For both WMP and NFDM, the D-values of Salmonella were not significantly (P < 0.05) different among the storage times. Similarly, there were no significant (P < 0.05) differences of D-values among the storage times for the Weibull model parameters of both WMP and NFDM. The fitted models were plotted against the inactivation data in Figure 2 for WMP and Figure 3 for NFDM. The proximity (closeness) and similar curvatures (slope and shape) of the curves for different storage times also reveal the similarity in model parameters.
Table 2Effect of storage time on model parameter estimates for the log-linear and Weibull models at 85°C
Figure 2Experimental and predicted survival of Salmonella and Enterococcus faecium in whole milk powder stored for 30 d (●, dashed lines), 60 d (+, dotted lines), and 90 d (♦, solid lines) at 85°C. N and N0 (cfu/g) are the Salmonella or E. faecium survivors at time t and 0.
Figure 3Experimental and predicted survival of Salmonella and Enterococcus faecium in nonfat dry milk stored for 30 d (●, dashed lines), 60 d (+, dotted lines), and 90 d (♦, solid lines) at 85°C. N and N0 (cfu/g) are the Salmonella or E. faecium survivors at time t and 0.
also reported that the storage time had no effect on the heat resistance of Salmonella in almonds, which is in agreement with this study. Similarly, the thermal resistance of L. monocytogenes adapted to NFDM (
) was shown to remain constant during prolonged storage.
Therefore, a conclusion can be drawn that when viable cells of Salmonella reach stable status in the low water activity environment, their heat resistance would not significantly change with increasing storage time up to 90 d. Although Salmonella can survive in low water activity foods for extended periods, this study suggests that their resistance to thermal treatment does not change significantly during the storage of 90 d at 85°C. While the effect of storage period on thermal resistance at other pasteurization temperatures needs to be evaluated, this result implies that the pasteurization protocol does not need to consider the prior storage period up to 90 d at 85°C. This may increase the flexibility of the application of process control if the results from the effect of storage period hold at other pasteurization temperatures.
In Table 2, it has been shown that D and δ values of E. faecium were significantly higher (P < 0.05) than Salmonella at the same storage condition in both WMP and NFDM. This demonstrated that E. faecium consistently exhibited greater thermal resistance (P < 0.05) than Salmonella during the storage of WMP and NFDM.
Thermal Inactivation Kinetics of Salmonella and E. faecium in WMP and NFDM
Because storage time (90 d) had no effect on the heat resistance of Salmonella, the isothermal treatments at 90 and 95°C were conducted by using the inoculated samples at d 30 of storage time to acquire thermal inactivation kinetics of Salmonella and E. faecium. The bacterial inactivation data were fitted to both log-linear and Weibull models in Figure 4. The model parameters for both WMP and NFDM are summarized in Table 3. For both Salmonella and E. faecium, D and δ values were found to significantly (P < 0.05) decrease with increased temperature, which indicated that a higher microbial reduction could be achieved at a higher temperature for the same treatment time. Significantly higher (P < 0.05) D and δ values of E. faecium were also observed at 85, 90, and 95°C than Salmonella. Similarly,
conducted isothermal treatments for NFDM with storage time of 48 h after inoculation and found that the decline of Salmonella exceeded that of E. faecium during isothermal treatment of NFDM at all 3 temperatures (85, 90, and 95°C). These results suggest that E. faecium is a suitable nonpathogenic surrogate for Salmonella in the thermal processing of milk powder at different process temperatures with 30 d of storage.
Figure 4Log-linear and Weibull models fitted to Salmonella and Enterococcus faecium inactivation data of 90 and 95°C. Salmonella-90°C (●, blue lines), Salmonella-95°C (★, green lines), E. faecium-90°C (+, red dotted lines), and E. faecium-95°C (♦, black lines). N and N0 (cfu/g) are the Salmonella or E. faecium survivors at time t and 0.
Figure 5 shows the effect of temperature on the D-values of Salmonella and E. faecium in dairy powders. The z-values of Salmonella and E. faecium were determined to be 14.77 and 16.39°C for WMP and 12.82 and 16.58°C for NFDM, respectively. The z-values of Salmonella and E. faecium found in this study were very close to the z-values of Salmonella and E. faecium (16.4 and 16.5°C) reported by
. The determined microbial inactivation kinetics of Salmonella and E. faecium could be a practical tool for the food industry to identify appropriate thermal processing conditions for pasteurization of milk powders.
Figure 5Thermal resistance constants (z-values) of Salmonella and Enterococcus faecium in whole milk powder (WMP) and nonfat dry milk (NFDM).
Table 2, Table 3 show that D-values of both microorganisms in NFDM were slightly higher than WMP. Although the initial water activity of both milk powders was equilibrated 0.10 at 25°C, the water activity at 85°C could be different because of differences in the food matrix. As water activity at high temperature plays an important role in thermal inactivation of low-moisture foods, the water activity at 85°C for both milk powders may need to be determined to explain the slight differences in D-values. Fat has been shown to have a protective effect on high-moisture foods; however, the protective effect of fat in low-moisture foods is not known and need to be investigated in future.
Bacterial Inactivation Model Comparison
Storage time (90 d) did not have an effect on the thermal resistance of bacteria in WMP and NFDM. Therefore, the thermal inactivation data at different storage times were combined and globally fitted to both log-linear and Weibull models for model comparison. The AICc and RMSE analysis (Table 4) indicated that the Weibull model was more likely correct (probability >0.72 for Salmonella and >0.99 for E. faecium) for all cases. The visualization of the models fitting on the combined data set was shown in Figure 6.
Table 4Model parameter estimates for the log-linear and Weibull models globally fitted across all storage times at 85°C
D = decimal reduction time; δ = scale parameter for Weibull model; β = shape parameter for Weibull model; RMSE = root mean square error; AICc = corrected Akaike information criteria; WMP = whole milk powder; NFDM = nonfat dry milk. Microbial populations from 3 storage periods and triplicates were simultaneously used to fit model to estimate model parameters along with the SE of estimate reported in parentheses.
Sample
Bacteria
Model
D or δ (min)
β
RMSE (log10 cfu/g)
AICc
Probability
WMP
Salmonella
Log-linear
7.93 (0.32)
—
0.21
−165.80
0.04
Weibull
7.07 (0.33)
0.89 (0.04)
0.20
−172.15
0.96
Enterococcus faecium
Log-linear
17.16 (0.78)
—
0.33
−116.79
~0
Weibull
23.05 (0.62)
1.31 (0.04)
0.23
−158.52
~1
NFDM
Salmonella
Log-linear
11.00 (0.30)
—
0.32
−119.55
0.28
Weibull
11.9 (0.34)
1.07 (0.03)
0.32
−121.42
0.72
E. faecium
Log-linear
19.95 (1.42)
—
0.28
−133.69
~0
Weibull
25.53 (0.83)
1.29 (0.05)
0.21
−166.09
~1
1 D = decimal reduction time; δ = scale parameter for Weibull model; β = shape parameter for Weibull model; RMSE = root mean square error; AICc = corrected Akaike information criteria; WMP = whole milk powder; NFDM = nonfat dry milk. Microbial populations from 3 storage periods and triplicates were simultaneously used to fit model to estimate model parameters along with the SE of estimate reported in parentheses.
Figure 6Global log-linear and Weibull models fitted to Salmonella and Enterococcus faecium inactivation data combined across all storage times at 85°C at whole milk powder (WMP)-Salmonella (★, dash-dot line), WMP-E. faecium (●, dashed lines), nonfat dry milk (NFDM)-Salmonella (+, dotted lines), and NFDM-E. faecium (♦, solid line). N and N0 (cfu/g) are the Salmonella or E. faecium survivors at time t and 0.
Several studies have reported that the Weibull model is a better fit than the log-linear model in low water activity food, such as Salmonella in whey protein powder (
reported that the log-linear model provided a better prediction than the Weibull model for Salmonella inactivation in wheat flour with minimal difference in AICc values for both models at most water activity levels. In terms of model fitting, the Weibull model was shown to be more suitable than the log-linear model for describing the thermal inactivation of Salmonella and E. faecium in WMP and NFDM.
CONCLUSIONS
In this study, both Salmonella and E. faecium showed good survivability in WMP and NFDM at low water activity with <0.9 log cfu/g reduction of Salmonella and <0.2 log cfu/g reduction of E. faecium after 90 d of storage. The thermal resistance of Salmonella and E. faecium did not change significantly during the 90 d storage of WMP and NFDM. This finding simplifies the process validation study in the industry. Enterococcus faecium was a suitable surrogate for Salmonella at different temperatures (85, 90, and 95°C). The TDT comparison between Salmonella and E. faecium in milk powders provides an important foundation for the dairy industry for evaluating different thermal treatment technologies that can achieve a suitable reduction of target pathogens. Use of E. faecium, which is a conservative surrogate for process validation, may result in over-processing beyond what is needed for Salmonella reduction, which could affect product quality. Therefore, identifying the kill ratio between Salmonella and E. faecium could contribute to efficient process validations. Based on microbial inactivation kinetics of Salmonella and E. faecium, the food industry could establish an appropriate thermal process validation for pasteurization of milk powders.
ACKNOWLEDGMENTS
Funding and the milk powder samples for this project were provided by Mars Wrigley (Chicago, IL). The authors have not stated any conflicts of interest.
REFERENCES
Abd S.J.
McCarthy K.L.
Harris L.J.
Impact of storage time and temperature on thermal inactivation of Salmonella Enteritidis PT 30 on oil-roasted almonds.
Salmonella Montevideo infections associated with salami products made with contaminated imported black and red pepper — United States, July 2009-April 2010.
Survival of Salmonella Tennessee, Salmonella Typhimurium DT104, and Enterococcus faecium in peanut paste formulations at two different levels of water activity and fat.
Salmonella survival during spray drying and subsequent handling of skimmilk powder. III. Effects of storage temperature on Salmonella and dried milk properties.
30960-7&id=gr1.jpg){kind=link}
30960-7&id=gr2.jpg){kind=link}
30960-7&id=gr3.jpg){kind=link}
30960-7&id=gr4.jpg){kind=link}
30960-7&id=gr5.jpg){kind=link}
30960-7&id=gr6.jpg){kind=link}