1809

Journal of Food Protection, Vol. 77, No. 10, 2014, Pages 1809–1813
doi:10.4315/0362-028X.JFP-14-074
Copyright G, International Association for Food Protection

Research Note

Peroxide Test Strips Detect Added Hydrogen Peroxide in Raw Milk

at Levels Affecting Bacterial Load
NICOLE H. MARTIN,1* ADAM FRIEDLANDER,1 ALLEN MOK,2 DAVID KENT,1 MARTIN WIEDMANN,1
AND KATHRYN J. BOOR1

1Milk Quality Improvement Program, Department of Food Science, Cornell University, Ithaca, New York 14853; and 2Major Products,
66 Industrial Avenue, Little Ferry, New Jersey 07643, USA

MS 14-074: Received 14 February 2014/Accepted 2 June 2014

ABSTRACT
Hydrogen peroxide (H2O2) has a long-established history of use as a preservative in milk worldwide. The use of H2O2 to
activate the inherent lactoperoxidase enzyme system has dramatically improved the quality of raw dairy products in areas in
which cooling is not widely available. In the United States, however, where refrigeration is widely available, the addition of H2O2
to milk is not permitted, with the exception of certain applications prior to cheesemaking and during the preparation of modified
whey. Due to the relatively quick deterioration of H2O2 in fluid milk, the detection of raw milk adulterated with the compound
can be challenging. In this study we evaluated (i) total aerobic bacterial counts and (ii) ability of peroxide test strips to detect
H2O2 in raw milk with various concentrations (0, 100, 300, 500, 700, and 900 ppm) of added H2O2, incubated at both 6 and 21uC
for 0, 24, and 48 h. Results showed that at both 6 and 21uC the H2O2 concentration and time had a significant effect on bacterial
loads in raw milk. Additionally, commercially available test strips were able to detect H2O2 in raw milk, with predicted
probability of .90%, immediately after addition and after 24 and 48 h for the higher concentrations used, offering a viable
method for detecting raw milk adulteration with H2O2.

Hydrogen peroxide (H2O2) has been shown to be an
effective bactericidal and sporicidal agent (2, 7, 10, 11, 13).
Use of liquid and vapor phase H2O2 is common in the food
industry and in the pharmaceutical and medical industries
(8) to reduce or eliminate bacterial contamination. Also, the
use of H2O2 to control microbial deterioration of food
products is common in the fruit and vegetable industry (1, 4,
6). Benefits of using H2O2 include its broad-spectrum
activity, also effective against pathogens, and its nontoxic
nature following degradation (10).
The global dairy industry also has a long history of
H2O2 use. Studies dating back to the 1880s have
demonstrated the preservative effects of H2O2 on raw milk
(14). In 1957, the Food and Agriculture Organization of the
United Nations and the World Health Organization (FAO/
WHO) Expert Panel on Milk Quality recognized the
potential for H2O2 as a processing alternative in developing
countries where the lack of infrastructure severely limits raw
milk quality (11). Concerns over the use of so-called ‘‘direct
addition’’ H2O2 as a milk preservative include the relatively
large concentrations of H2O2 used, the lack of regulatory
control, the potential for the destruction of certain inherent
vitamins (8), the alteration of casein and whey proteins,
which has been shown to lead to ‘‘soft curd’’ in cheese
making (12), and the potential for residual H2O2 to
* Author for correspondence. Tel: 607-255-2894; Fax: 607-254-4868;
E-mail: nhw6@cornell.edu.
inactivate starter cultures. General acceptance of ‘‘direct
addition’’ H2O2 application has, thus, been limited. The
only widely recommended method for preserving raw milk
in the absence of reliable refrigeration is activation of the
inherent lactoperoxidase enzyme system in raw milk
through the addition of small concentrations of both H2O2
(typically less than 10 ppm) and thiocyanate, which may not
be naturally present in consistently sufficient concentrations
(15 ppm; (15)) in raw milk. Alternatively, some lactic acid
bacteria are capable of producing H2O2 in quantities
sufficient to activate the lactoperoxidase system (5) and
have, therefore, been used in place of H2O2 to activate the
lactoperoxidase system. Because quantities of thiocyanate, a
necessary component of the lactoperoxidase system, vary
naturally in raw milk (15), activation of the lactoperoxidase
system either through the addition of H2O2 or H2O2-
producing lactic acid bacteria must be cautioned against as a
method of controlling bacterial growth in raw milk. In the
United States, the only approved use of H2O2 in the dairy
industry is in certain cheese and cheese by-product
applications at a maximum level of 0.05 and 0.04%,
respectively, and residual H2O2 is subsequently removed
(16); raw milk quality is controlled using good hygiene
during milking, proper equipment sanitation, and immediate
cooling. Despite the recommendation that direct addition of
H2O2 to raw milk should not be used as a raw milk
preservative, this practice is still used in some parts of the
world and on some farms. Thus, the objective of this study
1810 MARTIN ET AL. J. Food Prot., Vol. 77, No. 10

was to evaluate the effect of the direct addition of H2O2 on

bacterial growth in raw milk at refrigerated and ambient
temperatures at 24 and 48 h posttreatment and to evaluate
the ability of commercially available peroxide test strips to
detect residual H2O2 at each test point.

MATERIALS AND METHODS

Raw whole milk (,3.78 liters) was obtained on three separate
occasions from the Cornell Teaching and Research farm (Harford,
NY) between November 2011 and February 2012 and was
transported to the Milk Quality Improvement Program laboratory
(Cornell University, Ithaca, NY) at or below 6uC. For each of the
three independent trials, raw milk was thoroughly mixed according
to Standard Methods for the Examination of Dairy Products (9)
and was brought to the final incubation temperature (6 or 21uC)
prior to preparing 100 ml of raw milk aliquots with initial
concentrations of either 100, 300, 500, 700, or 900 ppm of H2O2 as
well as untreated raw milk controls (in 4-oz [118.3-ml] sterile
vials). All treated and control samples were spiral plated in
duplicate for enumeration of total bacterial count on standard plate
count agar (Difco, BD, Sparks, MD) and were tested for residual
H2O2 using commercially available peroxide test strips (EMD
Chemicals, Inc., Gibbstown, NJ) after 0, 24, and 48 h of incubation
at either 6 or 21uC. Peroxide test strips were used according to the
manufacturer’s instructions; briefly, the strip was fully immersed in
the treated and untreated milk samples for 1 s, and then, after
allowing residual milk to drip off the strip, the result was
determined after 30 s. Bacterial colonies on each plate were
enumerated after 48 h of incubation at 32uC. All bacterial count
data were log transformed prior to statistical analysis, which was
performed in R (R Foundation for Statistical Computing, http://
www.r-project.org/) using the following R packages: (i) brglm
(http://www.ucl.ac.uk/,ucakiko/software.html), (ii) plyr (http://
www.jstatsoft.org/v40/i01/), and (iii) ggplot2 (17).
FIGURE 1. Mean log bacterial counts from three independent
trials for raw milk treated with hydrogen peroxide (H2O2; 0, 100,
RESULTS AND DISCUSSION
300, 500, 700, or 900 ppm); held at (A) 6uC and (B) 21uC; and
H2O2 affects bacterial load in raw milk at refrig- tested at 24 and 48 h posttreatment. Error bars represent ¡1 SD
erated and ambient temperatures. Overall, our results from mean of data collected during three independent trials.
indicate that both H2O2 concentration (0, 100, 300, 500,
700, and 900 ppm) and incubation time after H2O2 addition
(0, 24, and 48 h), and their interaction, had a significant 0.12, respectively; Fig. 1A) from the control at 48 h. Using
effect (P , 0.05) on bacterial load in raw milk incubated at multiple linear regression analysis, we observed that, for
6 and 21uC after H2O2 addition. At refrigerated storage every additional hour (at refrigeration temperature) that raw
temperatures (6uC), bacterial counts in untreated raw milk milk is exposed to 900 ppm of H2O2, there is a reduction of
and raw milk treated with 100 and 300 ppm of H2O2 did not 0.04 log CFU/ml in the total bacterial count; at 500 ppm,
change significantly from 0 to 48 h (P ~ 0.48, 0.59, and every additional hour results in a reduction of only 0.02 log
0.69, respectively; Fig. 1A). Whereas bacterial counts CFU/ml (Fig. 2A and Table 1). For every 100 ppm of H2O2
decreased, on average, more than 1 log CFU/ml in milk increase, there was a decrease of 0.17 and 0.29 log CFU/ml
treated with higher H2O2 concentrations, the observed in the total bacterial count at 24 and 48 h, respectively,
decrease in bacterial numbers observed over 48 h in milk compared with the untreated control (Fig. 2A and Table 1).
treated with 500, 700, and 900 ppm and incubated at In other words, the longer raw milk is exposed to H2O2 at
refrigeration temperatures was not significant (P ~ 0.23, low temperatures, the more effect the H2O2 will have on the
0.16, and 0.13, respectively; Fig. 1A). Similarly, when bacterial load; and, similarly, the higher the concentration of
bacterial loads of the treated samples and the control were H2O2 raw milk is exposed to, the more dramatic the effect
compared after 48 h of refrigerated storage, the only on bacterial load over time.
treatment that was even marginally significantly different At ambient storage temperatures (21uC), total bacterial
(P ~ 0.09) from the control was 900 ppm; again, despite a counts in untreated raw milk and raw milk treated with 100
mean difference of more than 2.5 log CFU/ml. The and 300 ppm of H2O2 showed a significant increase over
remainder of the treatments (100, 300, 500, and 700 ppm) 48 h (P , 0.05 and Fig. 1B). No significant difference was
were not significantly different (P ~ 0.86, 0.91, 0.23, and found for raw milk samples treated with 500 and 700 ppm
J. Food Prot., Vol. 77, No. 10 EFFECTS OF HYDROGEN PEROXIDE ON RAW MILK BACTERIAL LOAD
FIGURE 2. Multiple linear regression analysis of mean log
bacterial count from three independent trials for raw milk treated
with H2O2 and held at (A) 6uC and (B) 21uC, as a function of time
and H2O2 concentration.
and held at ambient temperatures for 48 h (P ~ 0.90 and
0.26, respectively), despite the nearly 2-log decrease in
bacterial load in the sample treated with 700 ppm of H2O2
over the 48-h test period (Fig. 1B). Finally, the raw milk
sample treated with 900 ppm of H2O2 showed a significant
decrease in bacterial numbers (at 48 h versus 0 h; P , 0.05),
despite being held at ambient temperatures for 48 h
(Fig. 1B). When the bacterial loads of the treated samples
were compared with the control after 48 h of storage at
ambient temperatures, the 700- and 900-ppm treatments
were found to be significantly different from the control (P
, 0.05), whereas the remaining treatments (100, 300, and
500 ppm) were not significantly different (P ~ 1.0, 0.98,
and 0.29, respectively; Fig. 1B) from the control. Exploring
the interaction of time and H2O2 concentration at ambient
temperatures using multiple linear regression reveals effects
similar to those seen at refrigerated temperatures: the longer
raw milk is exposed to H2O2, the greater the effect on the
bacterial load (Fig. 2B). Conversely to results from samples
held at refrigeration temperatures, bacterial numbers in
samples exposed to 100, 300, and 500 ppm of H2O2 and
1811
TABLE 1. Estimate of effects of time and H2O2 concentration in
raw milk stored at both 6 and 21uC for 48 h
6uC 21uC
Reduction in Log bacterial
count/h at:
0 ppm 0.0029 0.1141
100 ppm 20.0021 0.0956
300 ppm 20.0122 0.0584
500 ppm 20.0222 0.0213
700 ppm 20.0323 20.0159
900 ppm 20.0423 20.0530
Reduction in Log bacterial
count/additional 100 ppm
of H2O2 at:
0h 20.0005 20.0012
24 h 20.0017 20.0056
48 h 20.0029 20.0101
held at ambient temperatures increased by 0.09, 0.05, and
0.02 log CFU/ml for each hour of storage, respectively
(Fig. 2B and Table 1). At higher concentrations, however,
there was a decrease in bacterial numbers for every hour of
ambient temperature storage of 0.01 and 0.05 log CFU/ml in
milk treated with 700 and 900 ppm, respectively (Fig. 2B
and Table 1). Further, for each additional 100 ppm of H2O2
used, there was a decrease in bacterial load of 0.56 and 1.00
log CFU/ml at 24 and 48 h, respectively, relative to the
untreated control (Fig. 2B and Table 1).
The effects of varying concentrations of H2O2 on the
bacterial load of raw milk held at refrigerated and ambient
temperatures are not unexpected. The bacteriostatic and
bactericidal activity of H2O2 has long been known (10).
Recommended concentrations for effective raw milk preser-
vation described in 1957 by the FAO/WHO Expert
Committee on Food Additives ranged from 0.10 to 0.40 g
of H2O2 per liter of raw milk (equivalent to 100 to 400 ppm)
(11). These recommendations have since been rescinded,
primarily because of the introduction of the use of activation
of the inherent lactoperoxidase system in raw milk, which
provides a sufficient extension of raw milk shelf life for
producers in developing countries, especially those in warm
climates, to market their product (3). Activation of the
inherent lactoperoxidase system requires only approximately
8.5 ppm of H2O2, 100 times less than required for the direct
addition method, and is widely considered safer, cheaper, and
more controllable than the latter method (3). Despite the
efficacy of the direct addition of H2O2 to raw milk as a
preservative, demonstrated by previous work (11) as well as
by the current study, the use of this method is prohibited in
the United States and is advised against by FAO/WHO (3).
Note that the use of the activation of the inherent
lactoperoxidase system should only be considered in the
most extreme cases, in which the lack of infrastructure
precludes adequate cooling; this preservation method is not a
substitution for cooling and/or pasteurization.
Commercially available test strips detect H2O2 in
raw milk. For milk stored at both 6 and 21uC, there was a
1812 MARTIN ET AL.
FIGURE 3. Predicted probability, based on a bias-reduced
logistic model, of a positive test result with a commercially
available peroxide test strip, on raw milk treated with varying
H2O2 concentrations and incubated at (A) 6uC and (B) 21uC for 0,
24, and 48 h. Numbers represent percent chance of detecting
hydrogen peroxide at corresponding concentration, time, and
temperature combination.
higher probability of detecting H2O2 at higher concentra-
tions and immediately after treatment than after 24 or 48 h
of storage (Fig. 3A and 3B). Specifically, based on a bias-
reduced logistic model of the data generated here, there
was a greater than 90% predicted probability of a positive
H2O2 test on raw milk (held at either temperature) that was
either (i) held for 0 h and treated at 300 ppm or above; (ii)
held for 24 h and treated at 600 ppm or above; or (iii) held
for 48 h and treated at 900 ppm (Fig. 3A and 3B). The
predicted probability of detecting H2O2 in raw milk after
48 h of storage at both temperatures was quite low at most
concentrations (Fig. 3A and 3B); this is not surprising
given the absence of continued, substantial decrease in
bacterial numbers, which indicates that the compound has
been degraded. This pattern of reduced bactericidal and
bacteriostatic activity over time is most readily observable
in the set of samples held at ambient temperature, due to
the faster growth rate of the inherent microbial populations
J. Food Prot., Vol. 77, No. 10
at 21uC (Fig. 1A and 1B). It is not known how often raw
milk is currently tested for the presence of added or
residual H2O2; however, overall, the use of peroxide test
strips provided a rapid and reliable method for detecting
added H2O2 at concentrations having the most impact on
bacterial numbers, making them a viable resource for
regulatory agencies and/or raw milk purchasers across the
globe.
In areas of the world where rapid on-farm cooling of
raw milk, by far the best practice for extending the keeping
quality of raw milk, cannot be easily implemented,
preservatives such as H2O2 have historically been used.
The addition of any chemical preservative to milk is widely
discouraged and is prohibited in most areas of the world
because of the difficulty in regulating the purity and
quantity of the compound used, the potential effects on
milk quality, and the negative impact on the image of
product wholesomeness. Despite the effectiveness of
bactericidal and bacteriostatic agents such as direct addition
H2O2, activation of the inherent lactoperoxidase system is
ultimately the only recommended method of maintaining
the keeping quality of raw milk for short periods of time in
locations with no access to cooling.
ACKNOWLEDGMENTS
This work was supported by the New York State Milk Promotion
Advisory Board through the New York State Department of Agriculture
and Markets–NYS dairy farmers committed to the production of high
quality dairy products. The authors acknowledge the staff and students of
the Milk Quality Improvement Program and the Dairy Foods Extension
Program, specifically, Rob Ralyea at Cornell University (Ithaca, NY) for
technical support.
REFERENCES
1. Aharoni, Y., A. Copel, and E. Fallik. 1994. The use of hydrogen
peroxide to control postharvest decay on ‘‘Galia’’ melons. Ann. Appl.
Biol. 125:189–193.
2. Baldry, M. G. C. 1983. The bactericidal, fungicidal and sporicidal
properties of hydrogen peroxide and peracetic acid. J. Appl.
Bacteriol. 54:417–423.
3. Barabas, J. 1995. An alternative method of milk treatment. World
Anim. Rev. 83:71–73.
4. Brennan, M., G. Le Port, and R. Gormley. 2000. Post-harvest
treatment with citric acid or hydrogen peroxide to extend the shelf life
of fresh sliced mushrooms. Lebensm-Wiss. Technol. 33:285–289.
5. De Vuyst, L., and E. J. Vandamme (ed.). 1994. Antimicrobial
potential of lactic acid bacteria, p. 91–142. In Bacteriocins of lactic
acid bacteria microbiology, genetics and applications. Springer
SciencezBusiness Media, New York.
6. Forney, C. F., R. E. Rij, R. Dennis-Arrue, and J. L. Smilanick. 1991.
Vapor phase hydrogen peroxide inhibits post-harvest decay of table
grapes. Hortic. Sci. 26:1512–1514.
7. Khadre, M. A., and A. E. Yousef. 2001. Sporicidal action of ozone
and hydrogen peroxide: a comparative study. Int. J. Food Microbiol.
71:131–138.
8. Krukovsky, V. N. 1949. Ascorbic acid oxidation in milk by
preformed hydrogen peroxide. J. Dairy Sci. 32:163–165.
9. Laird, D. T., A. Gambrel-Lenarz, F. M. Scher, T. E. Graham, and R.
Reddy. 2004. Microbiological count methods, p. 153–186. In H. M.
Wehr and J. F. Frank (ed.), Standard methods for the examination of
dairy products, 17th ed. American Public Health Association,
Washington, DC.
10. Linley, E., S. P. Denyer, G. McDonnell, C. Simons, and J. Y.
Maillard. 2012. Use of hydrogen peroxide as a biocide: new
J. Food Prot., Vol. 77, No. 10 EFFECTS OF HYDROGEN PEROXIDE ON RAW MILK BACTERIAL LOAD
consideration of its mechanisms of biocidal action. J. Antimicrob.
Chemother. 67:1589–1596.
11. Luck, H. 1962. The use of hydrogen peroxide in milk and dairy
products, p. 423–447. In Milk hygiene—hygiene in milk production,
processing and distribution. World Health Organization monograph
series no. 48. World Health Organization, Geneva.
12. Luck, H., and F. J. Joubert. 1955. Experiments on H2O2-treated milk.
Milchwissenschaft 10:160–165.
13. Mistry, V. V., and F. V. Kosikowski. 1985. Influence of potassium
sorbate and hydrogen peroxide on psychrotropic bacteria in milk. J.
Dairy Sci. 68:605–608.
1813
14. Roundy, Z. D. 1958. Treatment of milk for cheese with hydrogen
peroxide. J. Dairy Sci. 41:1460–1465.
15. Seifu, E., E. M. Buys, and E. F. Donkin. 2005. Significance of the
lactoperoxidase system in the dairy industry and its potential
applications: a review. Trends Food Sci. Technol. 16:137–154.
16. U.S. Department of Agriculture. 2013. Direct food substances
affirmed as generally recognized as safe. 21 CFR, part 184. Available
at: http://www.accessdata.fda.gov/scripts/cdrh/cfdocs/cfcfr/CFRSearch.
cfm?fr~184.1366. Accessed 27 May 2014.
17. Wickham, H. 2009. ggplot2: elegant graphics for data analysis.
Springer, New York.