Quality oI whole milk powder

95
* Corresponding author. Mailing address: Departamento de Nutrio e Produo Animal (VNP FMVZ USP), Rua Duque de Caxias Norte, 225, Caixa.Postal
23, Pirassununga, SP, CEP 13630-000, Brasil. Fax: (5519) 561-6215, E-mail: carlosaIusp.br
Brazilian Journal of Microbiology (2000) 31:95-98
ISSN 1517-8382
EFFECT OF MICROBIOLOGICAL CHARACTERISTICS OF RAW MILK ON THE QUALITY OF
WHOLE MILK POWDER
Carlos Augusto Fernandes de Oliveira
1'
; Lucinia Mestieri
1
; Marcos Veiga Santos
1
; 1os Franchini Garcia Moreno
1
;
Aleksandrs Spers
1
; Pedro Manuel Leal Germano
2
1
Departamento de Nutrio e Produo Animal, Faculdade de Medicina Veterinaria e Zootecnia, Universidade de So Paulo,
Pirassununga, SP, Brasil.
2
Faculdade de Saude Publica, Universidade de So Paulo, So Paulo, SP, Brasil.
Submitted: June 07, 1999; Returned to authors Ior corrections: March 24, 2000; June 26, 2000
ABSTRACT
The relationship between microbiological characteristics oI raw milk and the quality oI high-heat whole
powder made Irom it was investigated. To this end, 16 lots oI milk standardized Ior Iat and non-Iat solids
were used. Powder average levels oI mesophiles and psychrotrophs were 7.8 x 10
1
and 4.0 x 10
1
CFU/g,
respectively, and these values correlated positively (P 0.05) with the corresponding raw milk counts (8.1 x
10
6
and 1.8 x 10
7
CFU/g, respectively). A signiIicant positive correlation (P 0.05) was also observed
between raw milk psychrotroph counts and titratable acidity oI milk powder (range: 0.15 - 0.19 as lactic
acid). Samples oI powder produced Irom raw milk with higher psychrotroph levels were Iound to have
higher insolubility indexes, yet this relationship was not statistically signiIicant. The importance oI milk
hygienic control as a basic instrument to warrant commercial milk powder quality is discussed.
Key words: milk powder, microbiological quality, physical and chemical quality
INTRODUCTION
Milk powder is generally considered a product oI good
microbiological quality. However, several Iactors may contribute
to changes in its physical and chemical properties which reduce
shelI-liIe and thus its commercial value (4). DiIIerent researchers
agree that the hygienic conditions under which raw milk is
produced are the main Iactor aIIecting powder quality (10, 12,
15). Storage temperature and transportation may also inIluence
the properties oI milk powder, especially its insolubility index
and acidity (5).
In Brazil, raw milk used by powder manuIacturers is required
to match at least the same characteristics oI Grade C Milk, Ior
which no threshold levels have been established on
microbiological parameters such as total bacterial count and
pathogenic bacteria (11). ThereIore, some dairy industries has
set their own standards on the quality oI raw milk purchased Ior
processing.
The regulations on microbiological quality oI milk powder
in Brazil have established maximum levels Ior mesophilic
bacteria, coliIorms (at 30 and 45C), Staphvlococcus coagulase
positive and Salmonella species (13). EIIorts oI dairy plants to
meet standard requirements have traditionally concentrated on
applying high temperatures in order to reduce initial bacterial
loads oI poor quality raw milk. This method, however, has been
associated with low powder quality, mainly with respect to oII-
Ilavors, color changes and reduced stability aIter reconstitution
(9, 14).
Brazil produces nearly 230.000 tons oI milk powder/year,
which account Ior almost 25 oI its total annual milk production
(3). However, little inIormation is available on Iactors that aIIect
the quality oI milk powder in the country. In view oI this, the
aim oI the present study was to investigate the possible
correlation between microbiological characteristics oI raw milk
and the quality oI the resulting milk powder manuIactured under
local conditions.
96
C.A.F. de Oliveira et al.
MATERIALS AND METHODS
Sampling procedure
Sampling procedures were carried out in a commercial dairy
plant located in the State oI Minas Gerais, Brazil, during June
to October, 1997. Raw milk previously stored at 4 to 7
o
C was
standardized in order to obtain a Iat/non-Iatty solids ratio oI
0.395. Heat treatment oI raw milk included pasteurization at
90C Ior 8 s and an uperization step at 108C Ior 2 s. Heat
treated milk was then concentrated to 48 total solids in a triple
eIIect evaporator (average temperature oI 50 + 4C). AIter
concentration, milk was spray dried by atomization and
simultaneous direct injection oI air at 270C.
Duplicate milk samples oI the 16 lots used were collected at
3 diIIerent stages during processing, namely: a) standardized
reIrigerated raw milk stored in silos; b) concentrated milk; and
c) milk powder, collected immediately aIter drying and cooling.
All samples were placed in sterilized glass tubes and transported
to the laboratory oI dairy plant under reIrigeration (in an isotherm
recipient with ice), Ior immediate evaluation.
Microbiological analyses
Microbiological analyses were perIormed on samples oI raw
milk, concentrate and powder, Iollowing the procedures oI the
American Public Health Association (2). Using a sterile blender,
25 g oI sample were homogenized with 225 ml oI 0.1 peptone
saline solution. Then, according to each sample, tenIold serial
dilutions up to 10
-6
were prepared using the same diluent and
subsequently plated onto standard plate count agar medium
(PCA) (Oxoid). Psychrotroph counts were determined aIter
incubation at 7C Ior 10 days. Mesophile and thermophile counts
were obtained aIter incubation at 32C Ior 48 h and 55C Ior 48
h, respectively.
Physical and chemical analyses
Powder samples were assayed Ior titratable acidity ( lactic
acid) and moisture content according to the Analytical Methods
oI the Instituto AdolIo Lutz (6). Additionally, a reconstitution
test was perIormed using the insolubility index method described
by the American Dry Milk Institute (1).
Statistical analyses
Average results oI duplicate samples oI the 16 lots were
submitted to statistical analyses in order to determine Pearson`s
Correlation CoeIIicient (r) between parameters. The number oI
colony Iorming units (CFU) was transIormed to log10. A Student
t test, with normal approximation, was used to compare
correlation values (7) at a 5 level oI signiIicance.
RESULTS AND DISCUSSION
The bacteriological quality oI raw milk, concentrate and milk
powder is shown in Table 1. The number oI decimal reductions
Ior mesophile and psychrotroph counts was 5 to 6 log cycles,
which meets the microbiological standard established by
Brazilian regulations (13) Ior aerobic mesophile counts
(maximum oI 3.0 x 10
4
CFU/g). The eIIectiveness oI heat
treatment in reducing bacterial load in milk powder was similar
to that previously observed during production oI spray-dried
milk powder (5, 8, 12). Little change in thermophile counts with
each processing step has been reported (5).
Table 1. Microbiological analysis oI raw milk, concentrate and powder. Results
are expressed as geometric means.
Parameter Raw milk Concentrate Powder
CFU/g
a
Mesophile 8.1x10
6
1,6x10
2
7.8x10
1
(8.2x10
4
-6.9x10
7
)
b
(1.0x10
1
-9.2x10
2
) (1-5.6x10
2
)
Psychrotroph 1.8x10
7
4,2x10
2
4.0x10
1
(5.0x10
4
-2.0x10
7
) (1.5x10
1
-6.4x10
3
) (1-2.2x10
2
)
Termophile 8.0x10
0
1.7x10
1
2.3x10
1
(1-1.0x10
3
) (1-9.5x10
2
) (1-6.4x10
2
)
a
Mean oI duplicate analyses oI 16 lots examined.
b
Range values.
Data Irom the physical and chemical analyses are presented
in Table 2. There was little variation in the moisture content and
titratable acidity oI powder, as also observed by GriIIiths et al.
(5) in skimmed milk powder aIter low-heat processing
(pasteurization at 74C Ior 16 s). Insolubility indexes varied
Irom 0.27 to 0.45 ml, hence being within the range oI Brazilian
standards (11) and also in accordance with the ADMI (1)
recommendations. However, iI we were to take into
consideration Spreer`s upper limit oI 0.3 ml Ior powder
manuIactured by a spray-drying process (14), then 11 out oI
our 16 lots analyzed would be above the recommended
insolubility index.
Table 2. Average values oI physical and chemical characteristics oI milk
powder.
Parameter Mean
a
Titratable acidity () 0.17 (0.01)
b
Moisture () 2.84 (0.13)
Insolubility index (ml) 0.36 (0.09)
a
Mean oI duplicate analyses oI 16 lots examined.
b
Standard deviations are between brackets.
The statistical analyses indicated a positive correlation
between mesophile and psychrotroph counts oI raw milk and
milk powder, as shown in Table 3. These Iindings agree with
reports by Lck (10), who, as was the case in the present study,
Quality oI whole milk powder
97
applied a high intensity heat treatment to reduce bacterial loads
in raw milk and Iound similar correlations. With respect to
thermophile counts, however, we Iound no signiIicant
relationships (r 0.03, n 16). This is consistent with earlier
Iindings by Kwee et al. (8), who likewise compared high-heat
raw milk and powder Ior thermophile as well as thermoduric
and spore-Iorming organisms (8).
Table 3. Correlations (r) between microbiological, physical and chemical
characteristics oI raw milk and powder.
Parameters Correlation (r)
Mesophiles 0.77*
Mesophiles x Titratable acidity 0.45
Mesophiles x Insolubility index 0.35
Mesophiles x Moisture 0.11
Psychrotrophs 0.69*
Psychrotrophs x Titratable acidity 0.64*
Psychrotrophs x Insolubility index 0.45
Psychrotrophs x Moisture 0.12
Termophiles 0.03
Termophiles x Titratable acidity -0.37
Termophiles x Insolubility index -0.38
Termophiles x Moisture 0.35
* signiIicant (P 0.05).
There was a slight increase in the titratable acidity oI milk
powder manuIactured Irom milk with higher mesophile and
psychrotroph counts, yet this chemical parameter correlated
signiIicantly only with the latter (r 0.64, n 16, P 0.05).
These results diIIer Irom powder quality data Irom other authors
(12); in their study, on the other hand, raw milk was submitted
to a single low intensity heat treatment during processing, unlike
the present investigation. Since the titratable acidity oI milk is
augmented by Iormation oI organic acids due to heat (9), the
high temperature used in our experiments may have contributed
to increased acidity values in the powder manuIactured Irom
raw milk with higher counts Ior psychrotroph organisms.
We Iound no correlation between moisture content oI the
powders and the microbiological parameters studied. In Iact,
moisture variations oI powders are more Irequently associated
with Iailures during the atomization procedure, and with changes
in the composition oI the standardized milks (14).
The main Iactors that increase insolubility indexes are high
temperature applied to raw milk and the load oI psychrotroph
bacteria (9, 14). In the present study, the correlation between
raw milk psychrotroph counts and milk powder insolubility was
positive but not statistically signiIicant (r 0.45, n 16). Similar
results were obtained by GriIIiths et al. (5) using raw milk with
psychrotroph counts Irom 3.1 x 10
5
to 4.3 x 10
8
CFU/ml.
Data Irom several authors indicate that psychrotroph numbers
exceeding 1 x 10
7
CFU/g may result in physical and chemical
changes, as well as deIects in the Iunctional properties oI milk
powder (4, 10, 15). This occurs mainly because oI the
appreciable amounts oI degrading enzymes secreted by the
psychrotroph Ilora (4). Thus, control oI these microorganisms
in milk is oI great importance to the dairy industry.
To conclude, the results oI the present study indicated a
positive correlation between mesophilic and psychrotroph counts
oI milk powder and the hygienic conditions oI raw milk used
Ior processing. Raw milk psychrotroph counts were also
negatively associated with titratable acidity levels in milk
powder. Our data demonstrate the importance oI reducing
microbiological loads oI raw milk in order to obtain a high
quality milk powder under our local conditions.
ACKNOWLEGDEMENTS
The authors thank the Fundao de Amparo a Pesquisa do
Estado de So Paulo (FAPESP, Brazil; grant n 96/1972-9) Ior
Iinancial support.
RESUMO
Efeito das caractersticas microbiolgicas do leite cru
sobre a qualidade do leite em p integral
Neste trabalho, avaliou-se a relao entre a qualidade do
leite em po integral e os principais parmetros microbiologicos
do leite empregado como materia-prima, utilizando-se 16 lotes
de leite padronizado quanto aos teores de gordura e solidos no
gordurosos. Os niveis medios de microrganismos mesoIilos e
psicrotroIicos do produto em po Ioram 7,8 x 10
1
e 4,0 x 10
1
UFC/g, respectivamente, sendo que estes valores estiveram
correlacionados signiIicativamente (P 0,05) com as contagens
de mesoIilos e psicrotroIicos do leite cru (niveis medios de 8,1
x 10
6
e 1,8 x 10
7
UFC/g, respectivamente). Observou-se,
tambem, uma correlao signiIicativa (P 0,05) entre a
contagem de psicrotroIicos do leite cru e o teor de acidez titulavel
do produto em po, cujos valores variaram de 0,15 a 0,19 (em
acido latico). As amostras de leite em po produzido a partir de
leite cru contendo niveis elevados de psicrotroIicos apresentaram
maiores indices de insolubilidade, porem, a correlao entre
estes parmetros no Ioi signiIicativa. Discute-se a importncia
do controle das condies higinicas do leite cru, como
instrumento Iundamental para garantir a qualidade do produto
em po oIerecido ao consumo.
Palavras-chave: leite em po, qualidade microbiologica,
qualidade Iisico-quimica.
98
C.A.F. de Oliveira et al.
REFERENCES
1. American Dry Milk Institute. Standards Ior grades oI dry milks. Bulletin
916. Chicago, ADMI, 1971.
2. American Public Health Association. Compendium of methods for the
microbiological examination of foods. 3
rd
ed. APHA, Washington, 1992,
1219p.
3. Brando, S.C.C. Tecnologia da produo de leite em po - 1
a
parte. Leite e
derivados, 17: 49-57, 1994.
4. Cousins, C.M.; Bramley, A.J. Microbiologia de la leche cruda. In:
Robinson, R.K. (ed). Microbiologia lactologica. V.1. Acribia, Zaragoza,
1987, p.109-150.
5. GriIIiths, M.W.; Phillips, J.D.; West, I.G.; Sweetsur, A.W.M.; Muir, D.D.
The quality oI skim-milk powder produced Irom raw milk stored at 2
degree. Food Microbiol., 5: 89-96, 1988.
6. Instituto AdolIo Lutz. Normas Analiticas do Instituto Adolfo Lut:. 3
a
ed.
V.1. IAL, So Paulo, 1985, 533p.
7. Johnson, R.; Bhattacharyya G. Statistics, principles and methods. John
Wiley and Sons, New York, 1987.
8. Kwee, W.S.; Dommett, T.W.; Giles, J.E.; Smith, R.A.D.; Roberts, R.
Microbiological parameters during powdered milk manuIacture. 2.
Relationships and predictability among counts. Aust. J. Dairv Technol.,
41: 6-8, 1986.
9. Lewis, M.J. Heat treatment oI milk. In: Robinson, R.K. (ed). Modern dairv
technologv. Chapman & Hall, London, 1993, p.1-60.
10. Lck, H. Control de la calidad de la industria lactologica. In: Robinson,
R.K. (ed). Microbiologia lactologica. V.2. Acribia, Zaragoza, 1987, p.255-
294.
11. Ministerio da Agricultura e do Abastecimento. Portaria n. 369, 04 set.
1997. Diario Oficial da Unio, Brasilia, 08 set. 1997, p.19699-19700.
12. Muir, D.D.; GriIIiths, M.W.; Phillips, J.D.; Sweetsur, A.W.M.; West, I.G.
EIIect oI the bacterial quality oI raw milk on the bacterial quality and
some other properties oI low-heat and high-heat dried milk. J. Soc. Dairv
Technol., 39: 115-118, 1986.
13. Secretaria de Vigilncia Sanitaria do ministerio da Saude. Portaria n. 451,
19 set 1997. Diario Oficial da Unio, Brasilia, 22 set.1997. p.21005-
21012.
14. Spreer, E. Lactologia industrial. 2
a
ed. Acribia, Zaragoza, 1991, 617p.
15. Woodhall, M. The application oI hazard analysis and critical control point
system to milk powder manuIacture. J. Soc. Dairv Technol., 42: 102-105,
1989.