Journal of Dairy Research (1988), 55, 125-142 Printed in Great Britain 125

REVIEW ARTICLE

Storage stability and some nutritional aspects of milk powders and

ultra high temperature products at high ambient temperatures
BY EDMUND RENNER
Dairy Science Section, Justus Liebig University, D-6300 Oiessen, FRG

{Received 23 March 1987 and accepted for publication 13 August 1987)

CONTENTS
PAGE PAGE
Introduction 125 Maillard reaction 136
Milk powders 125 Flavour quality 136
Biochemical reactions 125 Nutritional aspects 138
Nutritional aspects 128 Proteins 138
Ultra high temperature products 129 Vitamins 139
Proteolytic activity 129 Conclusions 141
Lipolytic activity 134 References 141

INTRODUCTION

There are two main methods of achieving a long shelf life for milk and dairy
products even when stored at ambient temperature: inactivation of all micro-
organisms by a sterilization procedure and by subsequent aseptic packaging, or
reducing the water content to such an extent that microbiological reactions can no
longer take place. However, other biochemical reactions of enzymic origin as well as
those of non-enzymic origin still occur, which lead to changes during storage at
ambient temperature.
It must also be considered that 'ambient temperature' means something quite
different in different parts of the world and that the rate of the biochemical reactions
depends largely on the storage temperature. In tropical areas, it is particularly
important to apply procedures which extend the keeping quality of milk products
and the same shelf life may not be achieved as in areas with a temperate climate.
This review discusses the effects of storage at elevated ambient temperatures on
the shelf life, storage stability, and nutritional and sensory values of milk powders
and ultra high temperature (UHT) milk products.

MILK POWDERS
Biochemical reactions
In skim milk powder, the water content is reduced to ~ 4% and although it is
not a sterile product, this low water activity (Aw < 06) does not permit microbial
growth and microbial activity. The same is true of enzymic activity. It is generally
accepted that growth of bacteria and most yeasts virtually ceases below Aw 09, while
moulds do not grow at Aw < 07 (Driessen, 1983). Thus, the storage temperature
alone is not the primary factor which will induce changes in the product, but the
126 E. RENNER

1-2 r
1-1
s 10
b 0-9
-08
c 0-7

x 0-5
0-4
0-3

6-5

5-5

4-5

3-5

2-5

0 3 6 9 12
Storage time, months
Fig. 1. Moisture and hydroxymethylfurfural (HMF) content of milk powder under different storage
conditions (De Vilder, 1982). a, 20 C, 72% RH; b, 30 C, 55% RH; c, 30 C, 95% RH.

storage temperature in combination with an increased moisture content. Therefore,

for the storage of milk powder, it is important to avoid the penetration of moisture
into the product by storing it at a low relative air humidity or in a container which
does not permit the penetration of moisture.
One important reaction which occurs during the storage of milk and dairy
products at ambient temperature is the Maillard reaction, where reducing sugars
such as lactose react with proteins and breakdown products of proteins such as
peptides. One of the first reaction products detected in the Maillard reaction is
hydroxymethylfurfural (HMF), the concentration of which increases with storage
time and storage temperature. Many of the products of the Maillard reaction are
aroma substances (aldehydes, reductones, other furfurals etc.) which have an
appetizing odour, but high levels of these compounds may cause undesirable odours
and flavours. The e-amino groups of lysine are mainly involved and because the
compounds formed during the reaction such as fructoselysine, furosine etc. are
resistant to enzymes, the content of available lysine is reduced, but usually only to
a small extent (Renner, 1983). Another result of the reaction is the production of
substances which lead to colour changes, but only in a more advanced stage of the
reaction.
In Fig. 1, the results of work by De Vilder (1982) are presented, where skim milk
 Storage of milk powders and UHT products 127
0-6T30 n680

0-5- -25

0-4+20 6-75

. 0-3-15 I
a

w 0-2-1-10 " 6-70

0-1--5

J
0-Lo 6-65
0 2 4 6 8 10 12
Storage time, months
Fig. 2. Changes in heat stability, pH and solubility index of high-heat milk powder stored at 30 (J
(Kieseker & Clarke, 1984). , Heat stability; . pH; A, solubility index.

powder, packed in industrial milk-powder bags, was stored under different conditions
of temperature and relative humidity (RH): at 20 C with 72 % RH and at 30 C with
55 and 95% RH, the last representing extreme conditions. Under these conditions
the moisture content of the powder increased to > 6%, and a clear degradation in
quality occurred at the end of the storage periods, indicated by the strongly
increased HMF values. This inferior quality was expressed also by a high insolubility
index of the powder. The industrial milk-powder bags used in this work did not
provide enough protection against water penetration if the RH of the surrounding
air was high (95%).
Browning reactions of the Maillard type occur during the storage of powdered
dairy products at ambient temperature if the moisture content is > 5%. The HMF
content is further increased in milk powder with added iron or added vitamin A:
after storage for 6 months the HMF content of the control sample was 28-4, with
added iron, 30-9 and with added vitamin A, 39-0 mmol/m 3 (Caric et al. 1984).
In another study (Kieseker & Clarke, 1984) the storage stability of skim milk
powder was investigated over the temperature range of 10-40 C. At a moisture
content below 4% only minor changes occurred in powders stored at up to 20 C. At
higher storage temperatures there was a reduction in heat stability and an increase
in the inherent viscosity of the reconstituted powder, which was accompanied by
changes in solubility indices and pH values (Fig. 2). The initial solubility indices for
powders, which were classified as low- and medium-heat powders according to the
degree of whey protein denaturation, were < 0-1 ml. This value did not change with
the low-heat powder, irrespective of the storage temperature, while small increases
were noted with the medium-heat powder held at higher storage temperatures. The
high-heat powder had an initial solubility index of 0-2 ml, which was virtually
unchanged after storage at 10 and 20 C for 12 months. The solubility index of the
same powder held at 30 C increased slowly to 0-4 ml over the same period and
powder held at 40 C reached a figure of 1-0 ml in 6-8 months. Use of such powders
may result in sedimentation in some recombined products.
In other experiments it was confirmed that the storage of milk powder at
128 E. RENNER

80

70

60

|50
To
40
x
I

30

20

10
10 20 30 40 50 60 70 80 90
Storage time, weeks
Fig. 3. Formation of hexanal during the storage of whole milk powder at +25 C (Hall et al. 1985).

temperatures between 18 and +22-5 C up to 8 months had no influence on the

solubility index, and little change occurred in the acceptability of the recombined
product (Wade et al. 1983, 1984).
The greatest threat to the storage stability of products with a high fat content is
oxidation during storage. Fat oxidation in whole milk powder stored in darkness at
+ 25 C has an initial linear phase which gradually changes to an exponential phase
which can be seen from the development of the oxidation product hexanal which is
effective as an unfavourable aroma compound (Fig. 3). The break point at which the
oxidation changes from the first to the second phase was found to be at 37 weeks
(Hall et al. 1985). Oxidation can be considerably reduced by cold storage and by the
addition of antioxidants.
It has been observed that the sensory quality of skim milk powder is significantly
affected by storage temperature during long term storage. Powder held at 32 C for
6 months began to develop off flavours and by 24 months was considered
unacceptable when evaluated by a trained sensory panel. Milk powder samples
stored at 21 C were rated unacceptable only after storage for 4 years. Storage at
10 C resulted in minimal flavour changes in 52 months. Unacceptability of samples
was due to an oxidized/stale flavour (Driscoll et al. 1985).

Nutritional aspects
The quality of the milk protein is changed very little when milk powder is
properly stored at a moderate temperature and low RH. Storage experiments in
which milk powder was kept at 25 and 37 C for one year showed that the N
digestibility, the protein efficiency ratio (PER) and the net protein utilization (NPU)
were not affected. The content of available lysine was decreased by about 8%. The
loss of available lysine during storage increased as the moisture content of the milk
powder increased. The Maillard reaction is thought to be the main cause of the
protein damage which occurs when powders are stored for a long time. This process
can be monitored by measuring the concentration of HMF in milk powder. To
prevent losses in the biological protein value, milk powders therefore should not be
stored for extended periods of time (Renner, 1983).
To examine changes in the protein value, skim milk powders were stored at 30
 Storage of milk powders and UHT products 129
Table 1. Content of available lysine in skim milk powders of different water activity
(Aw) after storage at 40 C for 1 month {original milk powder = 100)
Aw Available lysine
Dry 98
0-23 98
0-40 99
057 95
0-80 71
Open* 88
Data from Okamoto & Hayashi (1985).
* Stored open to the air at room temperature.

and 40 C for 1 month under various Aw (0-23-0-82). The samples of Aw > 0-40 turned
to a yellowish brown colour and to a dark brown colour at Aw 0-80. At Aw < 0-23 no
change in colour could be seen. These effects were independent of the storage
temperature. The solubility of the coloured samples decreased and some samples
formed solid particles when the powder was dispersed in water. The Maillard reaction
occurs to its greatest extent in milk powder when Aw ~ 0-75 (Okamoto & Hayashi,
1985).
The content of available lysine decreased when skim milk powder was stored at
40 C for 1 month at Aw > 0-40. At Aw = 0-80 the decrease was 29% (Table 1).
Samples stored at a low water activity or under dry conditions showed no loss of
available lysine. Storage at 40 C at Aw 0-57 and 0-80 considerably reduced the
amount of methionine (by 6 and 19% respectively). Losses of other essential amino
acids were observed only after storage at 40 C and Aw = 0-80 for 1 month, the loss
of tryptophan and arginine being about 10 %, but others such as proline and tyrosine
decreased only slightly. Under these rather extreme storage conditions the
proteolytic digestibility was significantly reduced (Okamoto & Hayashi, 1985).
The vitamin content can also be affected by storage conditions, although vitamin
losses are relatively low. In one case a 33% reduction in the content of vitamin B 6
was detected in skim milk powder after storage for 40 months (Kirchgessner &
Kosters, 1977), in another case losses of vitamin Bx and C were 10 % after 2 years
of storage (Renner, 1983). The amount of vitamin C lost depends on the O2 and water
vapour permeability of the packaging material. Milk powder should also be protected
against light to keep losses of the light sensitive vitamins, in particular riboflavin, as
low as possible (Renner, 1983).
The vitamin A stability in skim milk powder was measured over a period of 16
weeks at 21, 26 and 32 C. Samples kept in the dark lost 20% of the vitamin activity
when stored at 21 C, 27% at 26 C and 38% at 32 C. (DeBoer et al. 1984).
In other work, vitamin-fortified milk powder was stored at ambient temperature
for 2 years. Most of the vitamins studied, namely thiamin, riboflavin, niacinamide,
pyridoxine, ascorbic acid and a-tocopherol acetate survived the storage period
without loss. From these results it was concluded that these vitamins will also
survive when milk powder is stored at higher temperatures for a shorter period, i.e.
for 6 months. Only vitamin A was substantially degraded (Woollard & Edmiston,
1983).
ULTRA HIGH TEMPERATURE PRODUCTS
Proteolytic activity
During storage of UHT-treated milk, the main cause of degradation is the
activity of lipases and proteinases, both of native and bacterial origin, which survive
5 DAR55
130 E. RENNER
ultra heat treatment and may cause major changes in the ultra heat treated product
during storage, depending on the storage temperature.
Proteolytic activity during storage of UHT milk and UHT products causes age-
thinning (transparency), age-thickening (gelation and coagulation), a bitter taste
and an increase in the content of non-protein N (NPN). These changes may be caused
by native or bacterial proteinases, and physicochemical effects may be involved.
However, De Koning et al. (1985) attributed the process of gelation in unconcentrated
and in concentrated UHT milk to different mechanisms. In unconcentrated UHT
milk gelation is generally preceded by proteolysis, whereas in concentrated milk
gelation is caused by a non-enzymic physicochemical process.
The native milk proteinase retains some activity after UHT-treatment. The
keeping quality of UHT milk thus may be reduced by the residual activity of native
milk proteinase. However, the enzymic activity was considerably reduced and the
keeping quality of UHT-sterilized milk considerably improved after preheating the
milk before sterilization for 60 min at 55 C. In this case, during 11 weeks of storage
at 20 C, gelation did not occur, bitterness and transparency of the milk were
retarded, and proteolysis was decreased (Driessen, 1983).
Proteolytic enzymes of bacterial origin occur in milk owing to the growth of
psychrotrophic bacteria which are able to grow even at chill temperatures. It was
observed that storing raw milk for more than 72 h at chill temperatures might result
in the formation of thermoresistant bacterial proteinases which significantly limit
the shelf life of the milk after UHT processing (Mottar, 1984 a). A psychrotroph
population of < 1 x 104/ml may be sufficient to produce ^ 10 units of heat stable
proteinases/ml milk. Attempts were made to evaluate the quality of raw milk for
manufacturing UHT products by determining some specific protein breakdown
components (using high pressure liquid chromatography to obtain information on
the activity of Gram-negative psychrotrophic bacteria (Mottar et al. 1985). Two of
these specific protein breakdown components underwent changes during the
refrigerated storage of raw milk which were related to the bacterial count.
Consequently it may be possible to establish relatively rapidly whether a raw milk
is suitable for processing into UHT milk with regard to the shelf life of the end
product. Certain levels were proposed concerning the maximum quantity of the
components (Mottar et al. 1985). It is not known whether such a procedure can be
applied for the evaluation of milk powder to be used for recombined UHT
products.
The proteinases of Pseudomonas, Flavobacterium and Achromobacter species are
very heat stable and are, therefore, only partly inactivated by ultra heat treatment.
The thermoresistance is expressed by the D-value which indicates how long the milk
must be heated at a given temperature to destroy 90% of the enzymic activity.
These D-values were calculated as 5-6 min at 140 C (Mottar, 19846) and 30 s to
several minutes at 150 C (Renner, 1983). Such a heat treatment, necessary for the
complete inactivation of the proteinases, cannot be applied to milk as it would cause
considerable damage to the product.
The hypothesis of enzymic causes for gelation of UHT milk was quantitatively
proved by Guthy et al. (1985) who measured the volume of aggregating particles
during storage of UHT milk. The exponential growth of the particles shortly before
aggregation after a distinct lag phase in all milk samples clearly indicated the
activity of one or more enzymes which survived UHT treatment (Fig. 4). Gelation
occurred after a storage period of 10 weeks when the UHT milk was stored at 35 C,
after 16 weeks at a storage temperature of 20 C, and after 18 weeks at the
 Storage of milk powders and UHT products 131

0 4 8 12 16 20
Storage time, weeks
Fig. 4. Gelation of ultra high temperature milk as indicated by the volume increase of the casein
particles at different storage temperatures (Guthy et al. 1985). O, 35 C; , 20 C; A, 4 C.

refrigeration temperature of 4 C. With a non-enzymic origin of the microstructural

changes, the kinetics would have yielded a linear increase with no lag phase and no
rapid increase shortly before gelation.
Other examinations by electron microscopy showed that the gel of UHT milk was
formed by thread-like structures while casein micelles were almost fully disintegrated
(De Koning et al. 1985).
In a study where UHT milk was held at temperatures between 2 and 50 C, the
rate of increase in proteolysis was found to be temperature dependent (Kocak &
Zadow, 1985a). However, age-gelation did not occur in samples stored at 40 or 50
C. When gelation was denned as milk viscosity > 10 mPa. s, UHT milk stored at
ambient temperature (20-26 C) gelled after 120-150 d, whereas the onset of age
gelation was considerably retarded at temperatures of ^ 10 C (Fig. 5). The level of
proteolysis in samples stored over 200 d was highest in samples stored at 30 C.
Gelled samples did not show a common level of proteolysis.
When proteinases from psychrotrophic bacteria were added to UHT milk, there
was a correlation between the level of proteinase activity and gelation time (Mitchell
& Ewings, 1985). A threshold value for the enzyme activity of 30 x 10~3 U/ml was
determined for a shelf life in excess of 5 months at 23 C. It was demonstrated that
bacterial proteinases could cause age-gelation of UHT milk and that the onset of
gelation was directly related to the level of proteinase activity in the milk. However,
the proteinase levels in milk are so low that they cannot be measured by current
assay procedures. An objective, precise and sensitive method is still required for the
measurement of proteinase activity in UHT milk.
Other authors (Richardson & Te Whaiti, 1978) recommended that the activity of
heat stable bacterial proteinase in UHT milk should be < 6 x 10"3 U/ml, as UHT
milk stored at 30 C is unlikely to have a shelf life of > 3 months when containing
this level of proteinase. An increase in proteolytic activity will occur at higher
5-2
132 E. RENNER

12

10

2
CO

a.
E
o
12
1
>10

2
50 100 150 200 250 300
Storage time, d
Fig. 5. Viscosity of ultra high temperature whole milk during storage at different temperatures (Kocak
& Zadow, 1985a). A, 2C; Q, 10 C; O, 20 C; A, 25 C; , 30 C; , 40 C.

storage temperature. The storage temperature taken over any 6 monthly period will
average out closer to a temperature of about 23 C.
In UHT concentrates (UHT cream, UHT coffee cream, evaporated milk) age-
gelation occurs after a fairly short period of storage at ambient temperature - no
information is available on the storage stability at higher temperatures. Age-gelation
of such UHT products can be retarded either by the addition of polyphosphate salt
or by suitable heat treatment. A product shelf life in the order of 6 months can be
achieved at ambient temperature of 20 C (Muir, 1984).
However, the addition of polyphosphates for the control of age-gelation does not
always have the desired effect, as the different polyphosphate salts differ widely in
their ability to control age-gelation and to increase the shelf life of UHT milk (Kocak
& Zadow, 19856).
The formation of sediment in recombined UHT milk is affected by the quality of
the water used for recombination. Whilst hard water has little effect, high saline
water causes a marked increase in sediment (Newstead & Simpson, 1986).
Typical indicators for casein micelle cleavage by proteolytic enzymes splitting
the casein are the decrease of casein N and the increase of NPN. These chemical
changes are linked with microstructural changes and an increase of sediment
formation and viscosity. During storage the casein N content decreases at a rate
dependent on the storage temperature, the decrease being more pronounced in
directly heated UHT milk as shown by Guthy et al. (1985) (Fig. 6), which is explained
 Storage of milk powders and UHT products 133
90r

80

70

.60

50

40

30

10 20 30 40 50 60
Storage time, weeks
Fig. 6. Casein N content in directly and indirectly heated ultra high temperature milk during storage
at different temperatures (Guthy et al. 1985). Indirect heat treatment: A, 4 C; , 20 C; , 35 C.
Direct heat treatment: A, 4 C; Q, 20 C; O, 35 C.

Table 2. Proteolytic activity in ultra heat treated milk during storage at different
temperatures (as % increase of non-protein nitrogen related to total N)
Storage temp. C Proteolytic activity, %
4 10
20 3-2
30 5-3
40 6-9
Data from Malatje (1986).

by the reduced chemical damage caused by direct ultra heat treatment and therefore
a lower inactivation of proteolytic enzymes. McKellar et al. (1984) found that the
extent of proteolysis was closely correlated with length of storage at ambient
temperature. A better correlation was found for direct than for indirect UHT milk.
Thus, proteolysis can serve as a predictor of shelf life for UHT milk sterilized by the
direct procedure.
The NPN content of milk increased during storage at a rate dependent on the
storage temperature (Malatje, 1986). At 20 C a l-6-fold increase after ~ 1 year was
observed and 2- and 2-4-fold increases at 30 and 40 C. The optimum temperature for
proteinase production by psychrotrophic microorganisms is ~ 45 C. At room
134 E. RBNNER
temperature this activity was reduced to 30 % and at refrigerator temperatures to
16%. It is possible to calculate the proteolytic activity from the difference between
the relative amounts of NPN and total N before and after the storage period. Malatje
(1986) observed a dramatic increase in proteolytic activity with increasing storage
temperature of UHT milk (Table 2).
It is thought that proteolysis which is accompanied by an increase of NPN and
the formation of para-K-casein is caused by bacterial enzymes, while the native milk
proteinase induces an increase of non-casein N and the formation of y-caseins. This
is considered as a clear difference with diagnostic value for the detection of the cause
of certain defects of practical significance (Driessen, 1983). UHT milk samples that
exhibited a bitter taste before gelation occurred exhibited increases in NPN content
from 0-03 to 0-06%, but there was no common level of NPN above which the level
of bitterness could be considered as unacceptable (Mitchell & Ewings, 1985).
Storage of UHT milk at a temperature > 20 C showed a general decrease in the
pH value: at 30 C decreasing to about pH 62 after 1 year and at 40 C to about
pH 6-0. However, the onset of age-gelation could not be related to the level or extent
of decrease in pH (Kocak & Zadow, 1985a).
Kroll & Klostermeyer (1984) found that the activity of bacterial proteinases
could be reduced by preheating the milk before UHT treatment. The average rates
of inactivation following heat treatment in milk at 55-60 C for 20 min were
45-50%. This can be regarded as the most effective temperature range for the
inactivation of the proteinases in milk.
The proteolytic activity in UHT milk prepared from recombined powder cannot
be compared with that of UHT milk prepared from fresh raw milk as there is little
information available on the inactivation of bacterial proteinases during the
manufacturing process of milk powder, where the milk is subject to a temperature of
4080 C when it is concentrated and to a temperature of 80-100 C when it is spray
dried, but only for a relatively short period. It cannot be presumed that the
proteinases in milk powder are completely inactivated as age-gelation can also be
observed in recombined UHT milk.

Lipolytic activity
When lipolytic activity occurs during the storage of UHT milk the content of free
fatty acids (FFA) increases inducing the development of off flavours (Renner, 1984).
When UHT milk was kept refrigerated the content of PFA remained unchanged
during a storage period of 8 weeks, but there was an increase (+ 0-4%) when it was
held at 20 C which was recognized as a sensory difference; at 38 C the content of
FFA was further increased ( + 0-9%) (Fig. 7).
A linear increase in the content of FFA in UHT milk up to a storage period of 6
months was also observed by Malatje (1986). When stored at 4 C for 6 months there
was only a small increase (1-4-fold), which would not impair the organoleptic quality,
but a 2-0- and 2-5-fold increase at 30 and 40 C respectively had a very unfavourable
effect.
It was found that lipolysis due to native milk lipase could be reduced sufficiently
by normal pasteurization. For instance no milk lipase activity was measured after
heating for 10 s at 85 C (Driessen, 1983). However, bacterial lipases, like bacterial
proteinases, are very heat-resistant: extracellular Pseudomonas lipases have D-values
of 1-5 min at 150 C.
Exposure of extracellular lipase of strains of Ps.fluorescens to temperatures above
50 C resulted in a two-stage loss of lipolytic activity (Driessen, 1983). The initial
 Storage of milk powders and UHT products 135
0-70

0-60

Z 0-50

0-40

0-30

0 2 4 6 8
Storage time, weeks
Fig. 7. Content of free fatty acids (FFA) in ultra high temperature milk at different storage
temperatures (Renner, 1984). , 38 C; , 20 C; O, 4 C.

50

5 4
CO

0L
0 1 2 3
Storage time, weeks
Fig. 8. Fat acidity of ultra high temperature milk with added culture of Pseudomonasfluorescensat
different temperatures (Driessen, 1983).

inactivation can be seen after heating to 50-60 C for ~ 1 h, the second one needs a
more intensified heat treatment, i.e. 130 C for 5-10 min or 150 C for ~ 5 min. It is
reported by Bucky et al. (1987) that a novel heat treatment (140 C for 5 s followed
by 60 C for 5 min) reduced the rate of production of volatile FFA to < 10% of the
rates achieved after the normal UHT treatment.
At high storage temperatures, a marked hydrolysis of milk fat can occur. When
the BDI value (a measure of fat acidity, Bureau of Dairy Industries), exceeds 1*5,
UHT milk is judged to be rancid. Driessen (1983) stored milk with lipolytic activity
enhanced by the addition of Ps. fluorescens at different temperatures (Fig. 8). The
threshold BDI value of 1*5 was exceeded after storage for 3 weeks at 20 C.
136 E. RENNER

30

s20
o
a.

10

12 16 20 24 28 32
Storage time, weeks
Fig. 9. Changes in hydroxymethylfurfural (HMF) values during storage of ultra high temperature milk
(Fink & Kessler, 19866). Heated directly: T, 35 C; O, 4 and 20 C. Heated indirectly: V, 35 C;
D, 4 and 20 C.

Maillard reaction
During, the storage of UHT milk the Maillard reaction will proceed to an
appreciable extent only when the storage temperature is above 20 C. In a study by
Fink & Kessler (1986 a, b) the HMF values of UHT milk were 3-5 /tmol/1 immediately
after processing. During storage the HMF content remained constant over a period
of 6 months when the storage temperatures were between 4 and 20 C. At 35 C there
was a slight increase in HMF concentration, slightly greater in indirectly heated
UHT milk than in directly heated milk. The flavour threshold value of HMF is about
16 /^mol/l so that the products of the Maillard reaction may lead to a sensory change
when the UHT milk is held at 35 C or above (Fig. 9).
Similar results were obtained by Malatje (1986) who found that after storing
UHT milk at 30 C for 6 months the HMF concentration approached the flavour
threshold value and at 40 C the threshold value was exceeded after ~ 10 weeks.

Flavour quality
The shelf life of UHT milk is also limited by the period in which it is of acceptable
organoleptic quality. Some factors which contribute to the organoleptic quality of
UHT milk have already been discussed: occurrence of a bitter taste, due to
proteolysis; concentration of FFA, due to lipolysis; the Maillard reaction.
The rate of development of stale flavour is affected by the concentration of
dissolved O2. However, the most likely products of lipid oxidation (acetaldehyde,
propanal and others) did not appear to be principal contributors to the staling
flavour in UHT milk during storage (Wadsworth & Bassette, 1985). Furthermore,
the stale flavour development did not parallel changes in thiobarbituric acid (TBA)
values. Acid degree values (ADV) increased at 32 C during prolonged storage, but
changes in these values also did not parallel development of the stale flavour. The
decrease in flavour acceptability paralleled an increase in stale intensity, which can
be clearly observed especially when UHT milk is stored at 32 C. In samples stored
 Storage of milk powders and UHT products 137
120 r

100

% 80
E
a.
a.
I 60

g 40

20

0L
6 8 10 12 14
Storage time, weeks
Fig. 10. Effect of storage temperature on the free -SH groups in ultra high temperature milk (Fink
& Kessler, 1986a).

at 7 C the flavour acceptability decreased only slightly and scores were still relatively
high after 16 weeks storage.
When UHT milk was stored at refrigeration temperatures (2-5 C) and at room
temperature (25 C) (Bassette & Jeon, 1983) it was observed that the increase in the
concentration of volatile compounds at room temperature, primarily of aliphatic
aldehydes, was closely related to the rapid decrease in product acceptability and to
the increase in the intensity of stale flavour. Relatively little change occurred in the
concentration of these aldehydes in milk stored at refrigeration temperatures.
Fink & Kessler (1986a) observed that increasing storage temperature accelerated
the oxidation of free -SH groups (Fig. 10), which correlates with the cooked flavour
of UHT milk. The concentration of O2 in UHT milk which was indirectly heated
without degassing ranged from 5 to 7 mg/1 which was sufficient to oxidize completely
the free -SH groups. Therefore, in UHT milk stored at 20 C cooked flavour
completely disappeared after about 2 weeks.
The overall acceptability of UHT milk stored at different temperatures was
described by Malatje (1986) (Fig. 11). A tasting panel found a significant change in
the flavour quality of UHT milk (expressed as sensory difference compared to the
sample stored at 4 C) after storage for 1 week at 40 C, after 2 weeks at 30 C and
only almost 6 months at 20 C. The UHT milk was rejected as not acceptable after
7 weeks at 30 C or after 5 weeks at 40 C.
Other workers confirmed that indirectly heat treated non-degassed UHT milk
stored at 20 C still had an acceptable flavour after 4 months. Although the quality
of the milk deteriorated slightly with increasing storage time it was still judged by
the tasting panel to be acceptable (Fink & Kessler, 1986a). When directly heated
UHT milk, which has a very low O2 content, was examined there was no significant
preference for either of the two samples stored at 4 and 20 C. After 6 months a slight
cooked flavour could be detected in both samples (Fink & Kessler, 1984).
138 E. RENNEB

I
CO

5
o

0Li 1 2 3 4 5 6 7 8 9
Storage time, weeks
Fig. 11. Sensory differences of ultra high temperature milk stored at different temperatures compared
to samples kept in the refrigerator (Malatje, 1986). , 20 C; A, 30 C; , 40 "C.

Nutritional aspects
Proteins. During the storage of UHT milk the e-amino group of lysine in milk
proteins may react extensively with lactose by the Maillard reaction before the milk
develops marked off flavour, discoloration or instability. Lactuloselysine and
fructoselysine have been identified as products of the Maillard reaction in UHT milk
stored at 4-37 C for 6 months to 3 years. The formation of these products appeared
to involve 10-30% of lysine residues in the stored milk samples. The content of
available lysine, related to the concentration in raw milk, was 98% after UHT
processing, 95% after storage for 6 months at 4 C and 87 % at 37 C (Moller, 1981).
That means that only after extended storage at high temperatures will significant
amounts of the lysine become unavailable. This temperature of storage is not
dissimilar to that of ambient conditions in some areas of the tropics, and therefore
losses of available lysine could reach such levels (Zadow, 1984).
However, in other experiments the loss of available lysine was lower. When UHT
milk was stored in the refrigerator the concentration of available lysine remained
unchanged during a storage period of 6 months; the decrease was only 6 % when kept
at room temperature of 20 C, 10% at 30 C, and 15% at 37 C (Malatje, 1986).
A correlation between HMF content and available lysine was found at storage
temperatures above 20 C (Malatje, 1986) (Fig. 12). From the correlation coefficient
of r = 0-92 (P < 001) at a storage temperature of 30 C it can be concluded that
under these conditions the loss of available lysine can be estimated from the HMF
values.
Renner et al. (1986) have shown that when UHT milk with hydrolysed lactose was
stored, a more intensive Maillard reaction occurred than in the non-hydrolysed
control milk samples, as can be seen from the increased HMF values (Fig. 13).
However, the content of available lysine was not reduced after storage for 6 months
when the UHT milk was stored in the refrigerator or at room temperature, but when
stored at 38 C the values decreased by 14 % in the lactose hydrolysed UHT milk and
by 10% in the control milk samples (Fig. 14).
 Storage of milk powders and UHT products 139
28

24

o
^20
o
o

u_ 16

12

0 10 20 30 40 50 60
Storage time, weeks
Fig. 12. Correlation between the hydroxymethylfurfural (HMF) content and concentration of
available lysine in ultra high temperature milk stored at 30 C (Malatje, 1986). r = 092***.
, available lysine; # , HMF.

350-

300

_250
o
E
5 200

150

100

50

0L
0 1 2 3 4 5 6
Storage time, weeks
Fig. 13. Effect of storage temperature on hydroxymethylfurfural (HMF) content of lactose hydrolysed
and control ultra high temperature milks (Renner et al. 1986). Hydrolysed: O, 38 C; , 20 C;
A, 4 C. Control: , 20 and 38 C.

Vitamins. Increasing storage temperatures accelerated the destruction of ascorbic

acid. In indirectly heated non-degassed UHT milk the concentration of 0 2 was
sufficient to oxidize completely the ascorbic acid. The ascorbic acid in UHT milk
decomposed after 10 weeks at a storage temperature of 4 C, after 2-3 weeks at 20
C, and after only 1 week at 35 C (Fink & Kessler, 1986a).
Other vitamins which are heat sensitive such as vitamin Bj, B 6 , B12 and folic acid
140 E. RENNER

80

c7'8
o
Q.
"76

17-2

<70

6'8

0 1 2 3 4 5 6
Storage time, months
Fig. 14. Content of available lysine of lactose hydrolysed and control ultra high temperature milk
depending on the storage temperature (Renner et al. 1986). O, Hydrolysed, 38 C; , control, 38 C;
0 , hydrolysed and control, 4 and 20 C.

1601-

-120

8 .2-9% fat

0-15% fat

> 40

I I I I I I
0 8 10 12 14 16 18 20 22
Storage time, d
Fig. 15. Degradation of vitamin A in ultra high temperature milk containing either 015 or 2-9% fat,
during storage at 26 C (Lau et al. 1986).

might also be affected by the storage temperature. When the concentration of

thiamine in UHT milk was examined, at storage temperatures of 4 and 20 C, no
significant degradation of the vitamin Bj could be observed during a storage period
of 6 months. When stored at 35 C, a steady decrease in the vitamin Bj content
occurred, so that after 30 weeks it had declined by 50% (Fink, 1984).
The stability during storage at 26 C of vitamin A in UHT milks containing either
0-15 or 2-92% fat was studied over 3 weeks (Fig. 15). Milks were fortified with
 Storage of milk powders and UHT products 141
synthetic retinyl palmitate to a final concentration of ~ 120 /ig retinol equivalent/
100 ml milk. The degradation rates for vitamin A were found to vary inversely with
the fat content in the milk (Lau et al. 1986).

CONCLUSIONS

Many reactions which occur in milk powder as well as in UHT milk are
temperature dependent. Therefore, the shelf life, the storage stability, the flavour
quality and the nutritional value are impaired when these products are stored at high
temperatures over a long period. Furthermore, it is important to avoid the
penetration of moisture during storage of milk powder as microbial growth and
activity do not occur at a low Aw (< 0-7).
It is difficult to recommend a limit for the maximum storage temperature as the
effects of the individual reactions are affected by the storage temperature to a
varying extent.
In milk powders held at 30 C changes occur in heat stability, solubility index and
pH value of skim milk powder; at 32 C off flavours develop after 6 months. However,
the protein value of milk powder is impaired significantly only under rather extreme
storage conditions (40 C).
With UHT milk gelation occurs after 10 weeks when stored at 35 C, while a
marked proteolytic activity can be seen already at a storage temperature of 20 C.
Storage at 30 C has an unfavourable effect on the organoleptic quality of UHT
milk.
At storage temperatures above 20 C there is a negative correlation between HMF
content and available lysine.
Therefore, in general, it seems reasonable to recommend that the storage
temperature does not exceed 25 C.

REFERENCES
BASSETTE, R. & JEON, J. J. 1983 Journal of Food Protection 46 950-953
BUCKY, A. R., HAYES, P. R. & ROBINSON, D. S. 1987 Journal of Dairy Research 54 275-282
CARIC, M., GAVARIC, D. & MILANOVIC, S. 1984 In Challenges to contemporary dairy analytical techniques
(seminar, Reading University) pp. 303-311. London: RSC (Royal Society of Chemistry Special Publication
no. 49)
DEBOER, M., DEMAN, L. & DEMAN, J. M. 1984 Journal of Dairy Science 67 2188-2191
D E KONINO, P. J., KAPER, J., ROLLEMA, H. S. & DRIESSEN, F. M. 1985 Netherlands Milk and Dairy Journal
39 71-87
D E VILDKR, J. 1982 Revue de VAgriculture 35 2051-2063
DRIESSBN, F. M. 1983 Thesis, (NIZO-Verslagen No. V236), Wageningen
DRISCOLL, N. R., BRENNAND, C. P. & HENDRICKS, D. G. 1985 Journal of Dairy Science 68 1931-1935
FINK, R. 1984 Thesis, Technical University Miinchen
FrNK, R. & KESSLER, H. G. 1984 Deutsche M olkerei-Zeitung 105 1444-1454
FINK, R. & KESSLER, H. G. 1986a Milchwissenschaft 41 90-94
FINK, R. & KESSLER, H. G. 19866 Milchwissenschaft 41 638-641
GUTHY, K., YOUN-HO-HONO & KLOSTERMEYER, H. 1985 Food Microstruclure Congress, Las Vegas, USA
HALL, G., ANDERSSON, J., LINGNERT, H. & OLOPSSON, B. 1985 Journal of Food Quality 7 153-190
KIESEKER, F. G. & CLARKE, P. T. 1984 Australian Journal of Dairy Technology 39, 74-77
KIRCHGESSNER, M. & KOSTERS, W. W. 1977 Zeitschrift fur Lebensmittel-Untersuchung und -Forschung 164
15-16
KOCAK, H. R. & ZADOW, J. G. 1985a Australian Journal of Dairy Technology 40 14-21
KOCAK, H. R. & ZADOW, J. G. 19856 Australian Journal of Dairy Technology 40 65-68
KROLL, S. & KLOSTERMEYER, H. 1984 Zeitschrift fur Lebensmittel-Untersuchung und -Forschung 179
288-295
LAU, B. L. T., KAKUDA, Y. & AENOTT, D. R. 1986 Journal of Dairy Science 69 2052-2059
142 E. RENNER,
MCKELLAR, R. C , FROEHLICH, D. A., BUTLER, G., CHOLETTE, H. & CAMPBELL, C. 1984 Canadian Institute of
Food Science and Technology Journal 17 14-17
MALATJE, S. M. 1986 Thesis, University of Giessen
MITCHELL, G. E. & EWINGS, K. N. 1985 New Zealand Journal of Dairy Science and Technology 20 65-78
MOLLBR, A. B. 1981 Progress in Food and Nutrition Science 5 357-368
MOTTAR, J. 1984a Lait 64 29-^5
MOTTAR, J. 1984& Lait 64 356-367
MOTTAR, J., VAN RBNTERGHEM, R. & D E VILDER, J. 1985 Milchwissenschaft 40 717-721
MUIK, D. D. 1984 Journal of the Society of Dairy Technology 37 135-141
NEWSTEAD, D. F. & SIMPSON, I. D. 1986 New Zealand Journal of Dairy Science and Technology 21 133-138
OKAMOTO, M. & HAYASHI, R. 1985 Agricultural and Biological Chemistry 49 1683-1687
RENNER, E. 1983 Milk and Dairy Products in Human Nutrition. Mtinehen: Volkswirtschaftlicher Verlag
RENNER, E. 1984 Deutsche Molkerei-Zeilung 10S 1010-1015, 1043-1047
RENNER, E., FETTER, CH. & RENZ-SCHAUEN, A. 1986 Meijeritieteellinen Aikalcauskirja 44 12-22
RICHARDSON, B. C. & T E WHAITI, I. E. 1978 New Zealand Journal of Dairy Science and Technology 13
172-176
WADE, V. N., AL-TAHIRI, R. A. & CRAWFORD, R. J. M. 1983 In Production Preservation and Processing of Food,
pp. 6061 (Eds J. V. McLoughlin and B. M. McKenna) Dublin: Boole Press {Research in Food Science and
Nutrition 1; International Congress of Food Science and Technology 6)
WADE, V. N., AL-TAHIRI, R. A., CRAWFORD, R. J. M. & TAMIME, A. Y. 1984 Egyptian Journal of Dairy Science
12 1-13
WADSWORTH, K. D. & BASSETTE, R. 1985 Journal of Food Protection 48 487-493
WOOLLARD, D. C. & EDMISTON, A. D. 1983 New Zealand Journal of Dairy Science and Technology 18 21-26
ZADOW, J. G. 1984 Australian Journal of Dairy Technology 39, 104-108