Food Chemistry 159 (2014) 293301

Contents lists available at ScienceDirect

Food Chemistry
journal homepage: www.elsevier.com/locate/foodchem

Assessment of heat treatment of various types of milk

Lambros Sakkas, Alexandra Mouta, Ekaterini Moschopoulou, Golfo Moatsou
Laboratory of Dairy Research, Department of Food Science and Human Nutrition, Agricultural University of Athens, Iera Odos 75, 11855 Athens, Greece

a r t i c l e

i n f o

Article history:
Received 10 December 2013
Received in revised form 19 February 2014
Accepted 5 March 2014
Available online 14 March 2014
Keywords:
Heat treatment
b-Lactoglobulin
a-Lactalbumin
Lactulose
Furosine
Milk
Condensed milk
Milk powders

a b s t r a c t
Raw milk (RM), reconstituted condensed milk (CM) and three types of reconstituted milk powders (SMPs)
were heated indirectly at 80140 C for 4 s. Native b-lactoglobulin after 90 C treatment of RM was
1132 167 mg/L but no reliable quantities were estimated at temperatures >100 C, whereas
218 43 mg/L residual a-lactalbumin were found at 130 C. Average lactulose contents from 51 to
1549 mg/L were detected at P100 C; average furosine was 1.9 and 126.5 mg/L in raw and 140 C treated
milks respectively. The behaviour of heated CM was similar to that of heated RM except for higher
furosine concentration. Reconstituted SMPs contained high quantities of lactulose and furosine, the ratio
of which was lower than in similarly treated RM. Among the market milks analysed, the group of
high-pasteurised milks was highly variable; i.e. native b-lactoglobulin was 692831 mg/L, lactulose
0824 mg/L and furosine 3.368.8 mg/L.
2014 Elsevier Ltd. All rights reserved.

1. Introduction
The production of heat treated milk for human consumption
covers the spectrum from pasteurisation to in-container sterilisation with respect to the shelf life and the heat-induced changes
of milk. The efcacy and the effects of heat treatments are related
to the temperaturetime combinations, heating method utilised
and milk pre-treatment conditions (Walstra, Wouters, & Geurts,
2006). Between the two well-known categories of drinking milk,
i.e. (low) pasteurised and UHT milk, there is the peroxidase negative extended shelf life (ESL) milk, marketed in several countries
as high-pasteurised milk. It can be stored up to 60 d under refrigerated conditions, while its avour is expected to be similar to
pasteurised milk. Various processing and packaging technologies
are combined for the production of ESL milk, i.e. 130145 C for
<1 s by means of direct infusion method or combination of bactofugation, microltration, pasteurisation and specic packaging
technologies. The simplest approach is high-temperature pasteurised milk at P85 C for 20 s and usually at 115120 C for 25 s.
(Moatsou, 2013; Rysstad & Kolstad, 2006; Walstra & Wouters,
2006).
There are two types of effects of heating on milk. The rst
comprises the effects on components of special interest for the
keeping quality and technological and nutritional properties of
Corresponding author. Tel.: +30 210 5294630; fax: +30 210 5294672.
E-mail address: mg@aua.gr (G. Moatsou).
http://dx.doi.org/10.1016/j.foodchem.2014.03.020
0308-8146/ 2014 Elsevier Ltd. All rights reserved.

milk. Degradation of lactose to organic acids and formation of

lactulose, denaturation of whey proteins, destruction of some vitamins and enzymes, hydrolysis of proteins and lipids and disturbance of calcium/phosphorous equilibrium belong to this group.
The second type of effects related to cooked avour and loss of
nutritional value is due to new substances formed by Maillard
reaction, which continues during the storage of heated milks
(Elliott, Dhakal, Datta, & Deeth, 2003; Pellegrino, De Noni, &
Resmini, 1995a). Several approaches have been proposed for the
assessment of heating history of milk using indices that come from
the abovementioned mechanisms considered by Clayes, Van Loey,
and Hendrickx (2002b) as intrinsic time temperature integrators
(TTIs) for thermally processed milk. The most well-known are
the enzymatic assays of alkaline phosphatase and lactoperoxidase
that are mandatory in the control of market milks, whereas the
thermal behaviour of other indigenous enzymes has been studied
(Claeys et al., 2002b; Moatsou, 2013; Wilbey, 1996).
Susceptibility of whey proteins to heat denaturation resulting
from their high level of secondary and tertiary structure has been
studied extensively; ndings with respect to the present study
are summarised below. Their heat sensitivity is in the order
a-lactalbumin (a-la) < b-lactoglobulin (b-lg) < bovine serum
albumin (BSA) < immunoglobulins (Igs). Considering their initial
concentration in raw milk and their behaviour under heat treatments, the residual native b-lg and a-la in heated milks, estimated
as acid-soluble concentrations, can be indices of thermal treatment; the opposite is true for BSA and Igs. At temperatures

294

L. Sakkas et al. / Food Chemistry 159 (2014) 293301

>65 C, unfolding of b-lg, i.e. rst stage of denaturation, occurs.

According to Claeys, Ludikhuyze, Van Loey, and Hendrickx
(2001a), denaturation follows rst order kinetics with z-values of
7.9 C (D75C = 49.9 min) from 7080 C and z-values of 24.2 C
(D85C = 3.53 min) from 8390 C. Cystine disulphide bonds disrupt
and along with the free sulfhydryl group of the native molecule are
available to participate in thiol/disulde exchanges; they react
intra- or inter-molecularly mainly with j-casein on the micelle
surface and with proteins on milk fat globule membrane (MFGM).
The main result of this reaction is the formation of b-lg/j-casein
complexes, which are a characteristic feature of heat-treated milks.
At higher temperatures, i.e. between 70 and 96 C, denaturation of
a-la takes place that forms complexes with aggregates of b-lg.
Denatured a-la forms also complexes with as2-casein and with
MFGM at high temperatures (Considine, Patel, Anema, Singh, &
Creamer, 2007; Corredig & Dalgleish, 1996; Jeanson, Dupont,
Grattard, & Rolet-Rpcaud, 1999).
Several studies on residual native whey protein content of heattreated milk have been carried out (e.g., Corzo, Delgado, Troyano, &
Olano, 1994a; Corzo, Lpez-Fandio, Delgado, Ramos, & Olano,
1994b; Elliott, Datta, Amenu, & Deeth, 2005; Elliott et al., 2003;
Feinberg, Dupont, Efstathiou, Loupre, & Guyonnet, 2006; Jeanson
et al., 1999; Lorenzen et al., 2011; Mayer, Raba, Meier, & Schmid,
2010; Morales, Romero, & Jimnez-Prez, 2000; Villamiel, Arias,
Corzo, & Olano, 1999; Villamiel, Lpez-Fandio, Corzo, & Olano,
1997). According to them, residual acid soluble b-lactoglobulin
(b-lg) content of low pasteurised milks ranges from 1606 to
4140 mg/L, of ESL milk from 140 to 3680 mg/L, of UHT-direct
heating method (UHT-DM) milk from 150 to 1120 mg/L, of
UHT-indirect method (UHT-IM) milks is <170 mg/L and in sterilised milks is <10 mg/L. The respective values for native residual
a-lactalbumin (a-la) are from 850 to 1570 mg/L, from 740 to
910 mg/L, from 850 to 1000 mg/L, from 230 to 1130, <500
and < 50 mg/L. Acid-soluble native b-lg content is considered
steady during the storage of heated milks (Pellegrino, Resmini, &
Luf, 1995b), which is advantageous with respect to its use as a
heat-treatment index. However, Corzo et al. (1994b), Elliott et al.
(2005) and Feinberg et al. (2006) consider that formation of lactose
adducts through Maillard reaction or complexes with other proteins taking place during the storage of UHT milk results in changes
in the shape of b-lg HPLC peaks.
Isomerisation products, due to heating of lactose, i.e. lactulose
and epilactose, have been also proposed as heat treatment indices.
The latter is 10% of the isomerisation products and it is detected
only in sterilised milks (Lpez-Fandio & Olano, 1999); therefore
it is not adequate for the majority of heat treatments applied to
milk. On the other hand, lactulose (LCT) is considered as a very useful index. It is not a substance of raw milk, is not detected in pasteurised milk and results from the conversion of glucose to fructose
during heating. Its formation follows pseudo-zero order kinetics
with an Ea-value of 90.2 kJ/mol and K110C = 51.5 mg/L, min (Claeys,
Ludikhuyze, Van Loey, & Hendrickx, 2001b). It is proposed as an
index for UHT milks, considering that LCT content >600 mg/L
corresponds to sterilised milk, and as a tool for the discrimination
between UHT-DM and UHT-IM milks, being rather stable during
storage (Cattaneo, Masotti, & Pellegrino, 2008; Claeys et al.,
2002b; Elliott et al., 2005; Morales et al., 2000; Pellegrino et al.,
1995a, 1995b). However, Feinberg et al. (2006) report a signicant
increase of its concentration during 90 d of storage of UHT and
sterilised milk at room temperature but according to Pellegrino
et al. (1995a) further isomerisation occurs only at storage at 35 C.
According to the majority of the reports, it can be detected only in
UHT milks at concentrations of 50850 and 190830 mg/L for
direct and indirect heating respectively and in sterilized milks at
10801400 mg/L (e.g. Olano, Calvo, & Corzo, 1989; BirlouezAragon et al., 1998; Villamiel et al., 1999; Morales et al., 2000,

Elliott et al., 2003, 2005, Cattaneo et al., 2008), but it has been
found in some pasteurised and high temperature pasteurised milks
at up to 15 and 80 mg/L, respectively (Feinberg et al., 2006;
Marconi et al., 2004).
Furosine (FRS) detected at low quantities of 35 mg/100 g protein in raw milk is related to the rst stages of Maillard reaction. It
is the stable product produced by the acid hydrolysis of unstable
lactuloselysine, which is accumulated in heat-treated milks. During heated milk storage, its concentration increases depending on
storage temperature, due to the continuous formation of lactuloselysine through Maillard reaction (Corzo et al., 1994a, 1994b;
Pellegrino et al., 1995a; Van Renterghem & De Block, 1996; Van
o & Olano, 1999, Claeys et al., 2001b;
Boekel, 1998, Lpez-Fandin
Elliott et al., 2005; Feinberg et al. 2006). Pellegrino et al. (1995a)
observed that this mechanism is evident even at 4 C and estimated that 7 mg furosine per 100 g protein are produced every
10 days at 23 C. Moreover, they report that the kinetics of the
reaction during storage is expected to be practically the same in
the different types of heat-treated milk. In general, furosine is a
more convenient index than lactulose, since it covers nearly all
types of heat treatments. Its formation in milk follows pseudo-zero
order kinetics and it is characterised by an Ea value of 88.7 kJ/mol,
with K110C = 16.2 mg/100 g protein, min (Claeys et al., 2001b). An
upper limit of 8 mg/100 g proteins for (low) pasteurised milk and
20 and 250 mg/100 g proteins for high-temperature pasteurised
and UHT milks respectively has been proposed, as cited by Claeys
et al. (2002b) and Mayer et al. (2010). The furosine content of
raw and low pasteurised milk has been reported in the range
414 mg/100 g protein, in ESL milks 10260 mg/100 g protein, in
UHT-direct method milks 16485 mg/100 g protein, in UHTindirect method milks 40430 mg/100 g protein and in sterilised
milks 250440 mg/100 g protein (Corzo et al., 1994a; Corzo et al.,
1994b; Pellegrino et al., 1995a, 1995b, Van Renterghem & De Block,
1996, Birlouez-Aragon et al., 1998; Jeanson et al., 1999; Villamiel
et al., 1999, Elliott et al., 2003, 2005, Feinberg et al., 2006; Cattaneo
et al., 2008; Mayer et al., 2010; Lorenzen et al., 2011).
Other compounds related to the Maillard reaction proposed for
the distinction of UHT and sterilised milks are hydroxymethylfurfural (HMF) and carboxymethyllysine (CML), which were not
studied in the present work (Morales & Jimnez-Prez, 1999; Van
Boekel, 1998).
A great part of the abovementioned high variability of various
indices is due to the wide range of combinations of heating conditions and methods utilised for the production of each category of
drinking milk. Other technological factors are also related to this
variability. Cattaneo et al. (2008) concluded that plant equipment
and management conditions like milk recirculation or run time
before cleaning may contribute substantially to the nal heat damage of milk. Prolonged preheating of milk at temperatures <100 C
in the UHT process can result in high furosine values and low levels
of acid soluble b-lg that are not in accordance with lactulose content; the latter is not formed at temperatures <90 C (Pellegrino
et al., 1995a). Also, the storage of raw milk before processing
decreases the furosine and HMF initially after processing and
during storage of UHT milk (Elliott et al., 2003). Finally, milk gross
composition could affect the behaviour of some indices. Heatinduced interactions of b-lg with MFGM although limited can
reduce the amount of native b-lg in heated milks (Corredig &
Dalgleish, 1996). In this regard, Claeys, Van Loey, and Hendrickx
(2002a) found that D75C and z-values decreased with increasing
milk fat content indicating a faster b-lg denaturation clearly
depicted at temperatures >72 C. The same group (Claeys, Van
Loey, & Hendrickx, 2003) report that fat content does not inuence
formation kinetics of lactulose and HMF and that the signicant
differences observed for furosine kinetics in milks with different
fat contents were too small to be relevant. The conclusion of the

L. Sakkas et al. / Food Chemistry 159 (2014) 293301

various research groups mentioned in this section, is that a combination of indices is necessary for the evaluation of heat loading
history of milk. Also, heat treatment indicators have been proposed
by Resmini, Pellegrino, and Cattaneo (2003) as a tool for fraud
investigation of drinking milk. They suggest that when FRS is
>8.6 mg/100 g protein in peroxidase positive pasteurised milk,
either reconstituted milk powder (RMP) or high-temperature
treated milk has been added and they propose the ratio FRS/LCT
as an indicator for the presence of RMP in UHT or in-bottle
sterilised milk.
The objective of this study was the estimation of the concentration of residual major whey proteins and the detection of new
compounds, i.e. lactulose and furosine, in raw and heat-treated
milks. For this purpose, heat-treatment pilot-scale experiments
on raw milk, reconstituted milk and milk powders were carried
out. In the second part of the study, various types of drinking milks
marketed in Greece were analysed.

295

of milk were mixed with 5.5 mL of a solution consisted of

9.1% (w/v) zinc acetate dehydrate, 5.46% (w/v) phosphotungstic
acid tetracosahydrate and 5.81% (v/v) glacial acetic acid, to remove
fat and protein. After the adjustment of the volume to 50 mL with
ultrapure water and thorough mixing, the mixture was left at room
temperature for 60 min. After ltering through 0.45 lm lters
(Millex-HV PVDF; Millipore, Carrigtwohill, Ireland), the ltrate
was analysed using two Aminex HPX-87P (30  0.78 cm, BioRad,
Hercules, CA) columns in sequence combined with adequate
anion- and cation-exchange precolumns. Elution was carried out
at 75 C at 0.6 mL/min with ultrapure water. Samples (1030 lL)
were injected after ltration through 0.22-lm lters (Millex-HV
PVDF; Millipore). All samples were analysed in duplicate. The estimation of lactulose content of the samples was based on the height
of the relevant peak. For the construction of the standard curve,
samples of non-fat pasteurised milk supplemented with known
quantities of lactulose (SigmaAldrich 61,360), i.e. 0, 75, 150, 225
and 300 mg/L, were analysed.

2. Materials and methods

2.1. Heat treatments and milk samples
The pilot-scale heat treatments were carried out by means of
indirect process using an Armeld FT74 HTST/UHT Processing Unit
equipped with a tubular heat exchanger. The pump frequency was
set at 50 Hz, pressure was P4 Pa and the outlet temperature was
8.3 C. The holding time was calculated by ow rate according to
the instructions of the manufacturer. Raw milk (RM) was heated
at 80, 90, 100, 110, 120, 130 and 140 C for 4 s. Reconstituted condensed skim milk (CM) commonly used for industrial yogurt production was treated in the same processing unit at 80 and 130 C
for 4 s. Finally, three types of reconstituted milk powders, i.e. low
heat (LHP), medium heat (MHP) and high heat (HHP), were similarly heated at 80 and 130 C.
Forty samples from the above experiments and 70 market
drinking milk samples corresponding to different production dates
of 19 different product labels from the Greek market were analysed. In particular, this group consisted of seven UHT milks and
58 high-pasteurised (HP), including four low-lactose (LL) and two
microltered-pasteurised (MF + LP) milks. The two latter groups
can be considered as ESL milks according to their shelf life. In addition, four samples of thermisized milk (TM) taken from a dairy and
three full-fat (LP/FF) and three low-fat low (LP/LF) pasteurised
milks were analysed for comparison reasons.
2.2. Analyses
2.2.1. Residual acid soluble (native) whey proteins
Residual (native) whey proteins were estimated by means of
RP-HPLC analysis of the acid wheys of the samples. Five millilitres
of milk were mixed with an equal volume of water and 200 lL of
33% (w/v) CH3COOH were added. After 10 min, the pH was adjusted to 4.60 by the addition of 130 lL 3.33 l CH3COONa. This
mixture was centrifuged at 5000g for 10 min at 4 C, and the supernatants under the top fat layer were collected and stored at <-25 C
until analyses. Eighty microlitres of the acid wheys were analysed
by the RP-HPLC method reported by Moatsou, Hatzinaki, Kandarakis, and Anifantakis (2003). Residual native b-lactoglobulin (b-lg)
and the a-lactalbumin (a-la) contents were evaluated using standard curves, constructed by analysis of b-lg (L3908 SigmaAldrich)
and a-la (L5385 SigmaAldrich) standard dilutions (from 100 to
2500 mg/L). All samples were analysed in duplicate.
2.2.2. Lactulose (LCT)
Lactulose was estimated according to the HPLC method of IDF
147/ISO 11868 (2007), after some modications. Fifteen millilitres

2.2.3. Furosine (FRS)

Furosine content was estimated according to the IP-RP-HPLC
method described in IDF 193/ISO 18329 Standard (2004), Serrano,
Castillo, Muoz, and Hernndez (2002), and Mayer et al. (2010)
with some modications. Two millilitres of milk sample were
mixed in specic vials with 6 ml HCl (10.6 M) for 2 min under a
continuous nitrogen stream and the sealed vials remained at
110 C for 23 h for the partial conversion of e-lactuloselysine to
furosine. The hydrolysate was brought to room temperature and
ltered through Whatman 5951/2 paper lter. The ltrate was further processed by means of reversed-phase solid-phase extraction
cartridges (SPE) (Sep-Pak C18 WAT051910 cartridges; Waters Corporation, Milford, MA), i.e. 0.5 mL of ltrate were eluted with 3 ml
HCl (3 M). The eluate was dried using a rotary evaporator, the dried
extract was reconstituted with 3.5 mL of a mixture of 5% v/v acetonitrile/0.2% v/v formic acid and ltered through a 0.22-lm lter
(Millex PVDF; Millipore). Eighty microlitres of each sample were
injected onto a C18 EC 250/4 Nucleosil 3005 column (MachereyNagel, GmbH & Co. KG, Dren, Germany). The reference curve
was constructed after analysis of standard dilutions by estimating
the peak heights of FRS peak; peak heights were better correlated
with the actual concentration of the standard solutions than peak
areas. A stock solution of FRS powder was prepared (70.7% furosine; PolyPeptide Group, Strasbourg, France), i.e. 10 mg containing
7.07 mg furosine in 5 mL 0.1 M HCl, dried, reconstituted in acetonitrile/formic acid mixture and ltered as previously described. The
standards (from 1.12 to 5.6 mg FRS/L) were prepared by mixing
adequate quantities of stock solution and acetonitrile/formic acid
mixture.
3. Results and discussion
3.1. Experimental heat-treated milk samples
The assessment of heat treatment indices was based on the
chromatographic proles presented in Fig. 13 and the results of
the experimental milk analyses in Table 1. The following discussion is based on ndings of other researchers about experimental
heat-treated milks. The concentration of acid-soluble native b-lg
in non-treated raw milk (RM) in the present experiments was
higher than the average values 33003500 mg/L reported by
Claeys et al. (2002b) and within the limits reported by Mayer
et al. (2010). After treatment at 80 C for 4 s, about 33% of the
initial b-lg was denatured and its average concentration was
2716 mg/L close to the 30% denaturation found by Villamiel et al.
(1997) for treatments at 80 C for 15 s. The native b-lg contents

296

L. Sakkas et al. / Food Chemistry 159 (2014) 293301

Fig. 1. RP-HPLC prole of acid-soluble whey proteins of raw milk.

Fig. 2. Chromatographic determination of lactulose (LCT) in milk heated at 140 C for 4 s.

Fig. 3. RP-HPLC determination of furosine (FRS) in milk heated at 130 C for 4 s.

of 140 C/4 s samples of Table 1 was negligible as reported by

Elliott et al. (2003) and Feinberg et al. (2006) for UHT-indirect
method (UHT-IM) milks. However, Morales et al. (2000) found
322 16 mg/L native b-lg in UHT-IM (140 C/3 s) milks and
389 16 mg/L after severe treatment at 150 C for 4 s by means
of a direct heating system.

The acid-soluble native a-la content of the RM in Table 1 was

within the range 13101840 mg/L reported by Jeanson et al.
(1999). In the present study, considerable quantities of native
a-la were estimated after treatment of RM, of CM and of LH- and
HHSMPs even at 130 C, conrming its lower heat sensitivity
compared to b-lg. Treatments at 80 C did not affect its

297

L. Sakkas et al. / Food Chemistry 159 (2014) 293301

Table 1
Effect of indirect heat treatments of raw milk (RM), reconstituted condensed milk (RM) and reconstituted low-heat, medium-heat and high-heat skim milk powders (LHSMP,
MHSMP, HHSMP) at various temperatures for 4 s.

a-la b (mg/L)

LCT

4076 103
2716 181
1132 167
404 35
200 25
129 78
57 41
33

1445 99
1416 33
1264 15
1004 57
755 71
543 255
218 43
34 41

ND
ND
ND
51 18
121 22
198 20
551 50
1549 287

1.9 0.7
3.1 1.1
3.4 0.4
7.7 3.0
14.3 5.5
24.6 6.0
61.0 1.2
126.5 25.5

3421 225
2689 407
287 274

1333 146
1390 27
551 283

ND
ND
401 276

2292 422
1850 505
191 232

1450 123
1340 28
695 475

2622 200
2050 21
210 236
33
26 30
7 10

Treatment (C for 4 s)

b-lg

RM
Non-treated
80
90
100
110
120
130
140
CM
Non-treated
80
130
LHSMP
Non-treated
80
130
MHSMP
Non-treated
80
130
HHSMP
Non-treated
80 C
130 C

(mg/L)

b-lg/a-laa,b

LCT/FRSc,d

LCT/ FRS-Pc,e

5.2 1.8
8.9 3.9
9.5 0.1
21.5 8.5
40.9 15.8
68.8 14.9
173.7 11.9
363.2 80.5

2.83 0.20
1.92 0.17
0.90 0.14
0.40 0.03
0.26 0.02
0.26 0.04
0.25 0.14
0.14 0.09

6.9 1.0
8.9 2.1
8.2 1.2
9.0 0.6
12.3 0.2

2.5 1.0
3.1 0.8
2.9 0.3
3.2 0.1
4.3 0.2

4.6
4.4 0.2
35.4 16.2

13.0
12.1 0.6
104.3 46.1

2.58 0.18
1.94 0.33
0.45 0.27

10.7 2.93

3.6 1.06

106 21
134 28
525 358

123.2 3.4
158 24.8
135.7 0.9

337.1 6.1
439.1 70.7
373.7 5.1

1.58 0.29
1.38 0.35
0.21 0.19

0.75 0.12
0.84 0.04
3.88 2.66

0.3 0.05
0.3 0.01
1.4 0.98

1297 22
1226 41
488 194

48 19
38 9
450 331

72.7 22.5
62.0 6.8
96.3 2.3

211 74.1
178.1 26
286.1 5.3

2.02 0.18
1.67 0.04
0.36 0.34

0.69
0.6 0.08
4.63 3.32

0.2
0.2 0.02
1.6 1.19

46 26
101 12
25 10

449 18
464 17
829 264

145.4 16.3
159.2 24.9
142.1 9.4

489.4 65.2
539.1 82.2
484.3 26.6

0.07 0.07
0.24 0.27
0.22 0.31

3.01 0.27
2.94 0.35
5.78 1.48

0.9 0.1
0.9 0.1
1.7 0.45

(mg/L)

FRS

(mg/L)

FRS-P

(mg/100 g protein)

ND: not detected.

a
blactoglobulin;
b
a-lactalbumin;
c
lactulose;
d
furosine;
e
furosine expressed on total protein content of the samples estimated by an automated IR analyzer.

concentration and at 90 C, about 12% was denatured, similarly to

16% losses observed by Morales et al. (2000) after treatment at
85 C for 30 s. Moreover, 15% of the initial a-la concentration in
RM was not denatured at 130 C, whereas about 10% of the initial
b-lg was in acid-soluble form after treatment at 100 C. After treatment at 140 C practically no a-la remained in its native form, contrasting with 482 20 and 304356 mg/L reported for UHT-IM at
140 C for 3 s and at 138144 C for 26 s, respectively (Elliott
et al., 2005; Morales et al., 2000). At conditions similar to the present study, i.e. 139 C for 3 s by indirect method, Corzo et al. (1994b)
found about 250 mg/L residual a-la. Similarly to native b-lg, the
effect of direct heating method (DM) is less severe than that of
indirect (IM), being in UHT-DM milks from 2.5 to 6.5 times that of
UHT-IM (Corzo et al., 1994a; Elliott et al., 2005; Morales et al.,
2000).
The difference between the heat sensitivity of the major whey
proteins is depicted in the evolution of b-lg/a-la ratio (Table 1),
which was not constant. It was evident that contrary to a-la, the
residual b-lg can be a reliable index for raw milk heated up to
100 C under the conditions of the present experiments since its
quantity was too low thereafter. Nevertheless, the assessment of
our chromatographic proles showed that although a-la withstands rather intense treatments, its peak is not always clearly
integrated due to its shape, opposite to b-lg peaks. Our conclusion
is similar to the observation of Pellegrino et al. (1995b) for another
RP-HPLC method.
The present ndings (Table 1) conrm that LCT is a substance of
high-heated milks appropriate for UHT processes. It resulted from
indirect treatments at 100 C for 4 s and thereafter; at 130 C and
at 140 C its concentration was very high, i.e. 551 and 1549 mg/L.
The LCT values of the present study are in general higher than HPLC
determinations in milks processed indirectly on a pilot scale. Elliott
et al. (2003) report about 240 mg/L LCT for UHT-IM at 138 C/6 s,

Morales et al. (2000) about 250 mg/L for UHT-IM at 140 C/3 s
and Cattaneo et al. (2008) about 600 mg/L for UHT-IM at 145 C/3 s.
Furosine (FRS) is present at low quantities of 35 mg/100 g protein in raw milk. In the present RM samples (Table 1), average FRS
was 1.9 mg/L or 5.2 mg/100 g protein; protein content being determined by an automated IR milk analyser. After 90 C for 4 s the initial concentration was almost duplicated being dramatically
increased thereafter. These values were higher than those reported
in the literature for experimental milks, similarly to LCT; nevertheless, the FRS literature data vary greatly. Values from 56 to 220 mg/
100 g protein are reported for experimental UHT milks by Resmini,
Pellegrino, and Batteli (1990). Elliott et al. (2003) report 2551 mg
FRS/100 g protein in experimental UHT-DM milks (143 C/6 s)
throughout a 12 weeks storage period, whereas in UHT-IM milks
(138 C/6 s) it was 120195 mg/100 g. Similarly, Cattaneo et al.
(2008) found in UHT-DM (150 C/4 s) milks, from 51 to 92 mg
FRS per 100 g protein throughout 90 days storage, whereas in their
UHT-ID (150 C/4 s) counterparts FRS was from 174 to 227 mg/
100 g protein.
Based on the evolution of LCT/FRS and LCT/FRS-P ratios (Table 1)
the effect of treatment severity was more intense on lactulose
accumulation than furosine. This phenomenon can be attributed
to the degradation of Amadori compounds that takes place as Maillard reaction proceeds (Pellegrino et al., 1995a, 1995b).
Reconstituted non-treated condensed milk (CM) contained
great quantities of native whey proteins (Table 1), which were similar to those reported for raw milk, as previously mentioned. LCT
was not detected in reconstituted CM before treatment and after
treatment at 80 C and FRS content in non-treated CM was >5 mg
per 100 g protein. The changes in these heat indices were lower
than expected considering the heat treatments involved in the
manufacture of condensed milk. However, features of the
production lines like heating under reduced pressure and falling

298

L. Sakkas et al. / Food Chemistry 159 (2014) 293301

lm evaporation technology apparently protect the nal product

from heat damage (Walstra & Wouters, 2006). This nding is of
particular importance for products, in which CM is utilised, like yogurt, because heat denaturation of whey proteins and their subsequent complex with casein are crucial for the gel structure and the
rheological properties of the nal product. The high FRS content of
non-treated CM compared to the RM (Table 1), along with nondetectable LCT quantities, can be attributed to storage, since this
type of milk is handled as a reserve rather than a raw material,
in contrast to RM.
Residual native whey proteins in non-treated SMPs were much
lower compared to RM and CM; in fact traces of them were detected even before treatment of HHSMP. The latter was expected,
due to intense pre-heating applied before evaporation and spray
drying, e.g., at 90 C for 5 min or at 120 C for 1 min (Walstra &
Wouters, 2006). In particular, these products are classied as
low-heat when whey protein nitrogen index (WPN) P6 mg N/g,
medium-heat when 6 > WPN > 1.5 mg and high-heat when
WPN 6 1.5 mg N/g. Therefore, according to the WPN, the b-lg and
a-la contents of LH-SMP were expected to be lower than that of
MHSMP, opposite to the ndings in Table 1. LCT was detected in
all three types of untreated reconstituted SMPs. In non-treated
HHSMP, the LCT content was in accordance with the previously
mentioned pre-heating conditions; these values corresponded to
130 C RM treatments (Table 1). The non-treated reconstituted
SMPs contained considerable amounts of FRS (Table 1). Resmini
and Pellegrino (1993) report that 100400 mg FRS per 100 g of protein are typical for SMPs, and that >600 mg/100 g of protein can be
detected after a long storage period. It is attributed to low aw during spray drying that favours the evolution of Maillard reaction
during storage even if the treatment conditions were not favourable for LCT formation. Consequently, the LCT/FRS ratio of SMPs
treated at 130 C was much lower compared to similarly treated
RM and CM (Table 1).
3.2. Market milk samples
The results for the group of market milks (paragraph 2.1) are
presented in Table 2. Thermisized milk (TM) contained
3539 175 mg/L native b-lg close to the 3800 mg/L in thermalised milks analysed by Morales et al. (2000). Lower values,
3156 116 and 3350 115 mg/L, were found in LP/FF and LP/LF
milks, respectively, which can be also attributed to standardisation
of milk composition apart from possible heat denaturation; apparently the higher content of LF milk is due to its higher total protein
content. These values were within the range of values for LP market milks reported by Jeanson et al. (1999), Villamiel et al. (1999),
and Mayer et al. (2010), and close to those of Birlouez-Aragon et al.
(1998), and Feinberg et al. (2006). In the study of Lorenzen et al.
(2011), six LP milk samples taken from dairy factories contained
from 3302 to 4745 mg/L acid soluble b-lg, whereas in the study
of Mayer et al. (2010), a great variability was depicted in LP market
milks, i.e. the native b-lg content of 33 market samples ranged
from 2528 to 3799 mg/L. Considering the average b-lg content of
raw milk at 33003500 mg/L, the lowest value for pasteurised milk
can be 2600 mg/L, for high pasteurised milk 2000 mg/L and for
UHT milk 50 mg/L, as cited by Claeys et al. (2002b), and Mayer
et al. (2010). The latter group recommends an acid soluble b-lg
content of 1800 mg/L as an upper heating limit for ESL milk. The
UHT milks native b-lg content was, as expected, low in accordance
to the results of experimental milks of Table 1.
The native b-lg content of HP/ESL milks of Table 2 varied considerably. Bearing in mind the abovementioned limits, only 26
samples out of 63 fullled the limits of P1800 mg/L, that is 41%
of the total. Among them, the MF + LP milk was close to LP milks,
similar to the ndings of Lorenzen et al. (2011); on the contrary

28% of these milks had residual native b-lg 6500 mg/L, which is
consistent with UHT milk. Great variability for this type of market
milk from 790 to 2600 mg/L has been also reported by Villamiel
et al. (1999), and by Mayer et al. (2010), who report that the
acid-soluble b-lg content of 71 ESL milks from 17 brands was from
140 to 3679 mg/L and that 55% of ESL samples contained <500 mg/
L b-lg. Feinberg et al. (2006), and Birlouez-Aragon et al. (1998)
report about 1100 and 1630 mg/L native b-lg for market samples
of this category. Recently, Lorenzen et al. (2011) showed that
ESL-DM milks and ESL-IM native b-lg were within 15892968
and 341673 mg/L, respectively.
As expected from the results of the pilot-scale experiments
(Table 1), a-la was present in noticeable quantities in all market
milks of Table 2 including UHT. Similarly to b-lg, the residual
a-la content of HP/ESL milks was highly variable and overlapping
was observed with UHT category. Mean native a-la in TM was
1957 63 mg/L, in LP/FF milk 1802 36 in LP/LF 1873 37, in the
HP/ESL milks from 127 to 1840 mg/L and in UHT milks from 116
to 1358 mg/L. Morales et al. (2000), Corzo et al. (1994a) and Elliott
et al. (2005) report 125 to 574 mg/L native a-la for UHT samples
and 3 to 6-fold higher native a-la for directly compared to
indirectly heated milks. According to Jeanson et al. (1999), the
non-native a-la percentage in raw milk was 1.33% and in pasteurised milk 2.913.9%, increasing to 59.184% in UHT and reaching
97.199.3% in sterilised milk.
LCT was not detected in TM and LP milks and in the HP/ESL
milks with >1800 mg/L residual native b-lg. The absence of detectable LCT in the HP/LL samples contrasts with their very low b-lg
content; it is due to very low initial concentrations of lactose, indicating that LCT is not a proper index for LL milks. LCT content in the
majority of UHT milks was within the 100600 mg/L range
reported for this category (Claeys et al., 2002b). Samples of product
3 could be characterised as sterilised milk. Moreover, according to
the >100 mg/L limit, a considerable amount of HP/ESL samples
could be characterised as UHT, as happened also with their residual
acid-soluble b-lg. Nevertheless, Table 2 indicates that LCT was not
detected in samples with residual b-lg >1800 mg/L. There are several reports about the estimation of LCT content of market milks by
HPLC. Birlouez-Aragon et al. (1998) and Jeanson et al. (1999) did
not detect any LCT in pasteurised milks, in contrast to Feinberg
et al. (2006), who reported 15 12.1 mg/L LCT for milk heated at
74 C for 30 s. Birlouez-Aragon et al. (1998) did not detect LCT in
HP milks, whereas Feinberg et al. (2006) report 31 13.6 mg/L.
In UHT-IM LCT was about 400 mg/L (Feinberg et al., 2006),
460 mg/L (Elliott et al., 2005), 550 mg/L (Birlouez-Aragon et al.,
1998) or in the range 543650 mg/L (Cattaneo et al., 2008).
FRS upper limits cited by Claeys et al. (2002) and Mayer et al.
(2010) are 8 mg/100 g protein for pasteurisation, 20 mg/100 g
protein for high pasteurisation and 250 mg/100 g protein for UHT
process. Fourteen representative samples were analysed with
regard to their FRS content (Table 2). The FRS contents of analysed
TM, LP and MF + LP samples were higher than that of RM in Table 1,
but within the abovementioned limits for pasteurised milk and
close to the literature data. Birlouez-Aragon et al. (1998) found 7,
Jeanson et al. (1999) 310, Villamiel et al. (1999) 6.910, Feinberg
et al. (2006) 4 and Mayer et al. (2010) 8.113.3 mg/100 g protein.
A wide range for the FRS contents of UHT market milks
appears in the literature. Pellegrino et al. (1995a) estimated
120230 and 30110 mg/100 g protein furosine in UHT-IM and
UHT-DM milks close to the 35109 and 108240 mg/100 g
protein found by Van Renterghem and De Block (1996). Elliott
et al. (2005) reported 188 51.8 mg/100 g protein for UHT-DM
and 371 88.7 mg/100 g protein for UHT-IM milks. Cattaneo
et al. (2008) reported 42206 mg/100 g protein for market
UHT-DM milks and Mayer et al. (2010) 93.1485 mg/100 g
protein. On the contrary, Corzo et al. (1994a) reported low

299

L. Sakkas et al. / Food Chemistry 159 (2014) 293301

Table 2
Thermal treatment indices of market milks. ND: not detected. LP: low pasteurised. Peroxidase positive. HP: high-pasteurised peroxidase negative. UHT: ultra high temperature.
MF + LP: microltered and pasteurised. peroxidase positive. FF: full-fat. LF: low-fat. LL: low lactose.
Product label _Milk type

b-lg

(mg/L)

a-la c (mg/L)

LCT

1_UHT/FF

970
890
858

1306
1358
1376

129
152
123

2_UHT/FF

832
1046

1012
1159

132
119

3_UHT/FF

35
69

116
127

824
674

4_HP/FF

2514
2583
2280
2821
2617

1709
1701
1669
1861
1456

ND
ND
ND
ND
ND

4_HP/LF

2621
2831

1401
1464

ND
ND

5_HP/FF

2774
2657
2520

1663
1697
1655

ND
ND
ND

6_HP/FF

492
538
624

1073
1071
1195

60
66
64

6_HP/LF

358
376

865
997

117
100

7_HP/FF

1645
1680
1682

1474
1595
1607

ND
ND
7.8

7_HP/LF

1908
1822

1322
1415

ND
ND

8_MF + LP/FF

3166
3086
2989

1769
1724
1703

ND
ND
ND

8_MF + LP/FF

3555
3388

1489
1560

ND
ND

9_HP/FF

1963
1803
1755

1582
1618
1534

ND
ND
ND

10_HP/FF

1258
1609
1594

1491
1528
1547

ND
ND
ND

10_HP/LF

1882
2191

1362
1436

ND
ND

11_HP/FF

267
205
153

838
886
821

68
97
98

12_HP/FF

2510
2354
2379

1717
1680
1665

ND
ND
ND

13_HP/FF

256
253
219

873
890
857

136
132
109

14_HP/FF

417
304
258

1187
1118
1021

52
73
81

14_HP/LF

471
461

1076
1080

99
82

15_HP/FF

1759
1672
1693

1584
1553
1586

ND
ND
ND

16_HP/FF

3434
3168
1254
1153

1840
1790
1470
1332

ND
ND
ND
35

17_HP/FF

1071
1144

1033
1056

33
19

(mg/L)

FRS

(mg/L)

FRS-P f (mg/100 g protein)

15.8

53.3

5.3

16.6

3.7

11.3

9.0

24.2

5.0

15.7

2.4

6.4

4.3

13

7.1

20.9

3.3

10.4
(continued on next page)

300

L. Sakkas et al. / Food Chemistry 159 (2014) 293301

Table 2 (continued)
Product label _Milk type

a
b-f

b-lg

(mg/L)

a-la c (mg/L)

LCT

(mg/L)

FRS

(mg/L)

FRS-P f (mg/100 g protein)

1184

1143

33

17_HP/LF

1233
1290

1201
1219

76
52

3.8

11.8

18_HP/LL/LF

116
118

590
545

ND
ND

74.6

226

19_HP/LL/LF

374
275

882
875

ND
ND

23.3

72.8

According to product label.

as in Table 1;

amounts of FRS in UHT milks, i.e. 3953.4 and 15.717.8 for

UHT-DM and UHT-IM, respectively.
The average FRS content of the eleven samples of the category
HP was 52.9 73.9 mg/100 g protein or 17.0 23.8 mg/L. The
MF + LP milk was similar to LP and TM with respect also to this
index. The FRS along with LCT gures conrmed that very variable
heating conditions and methods were applied during the production of this category of milks. In ve samples with native b-lg
>1800 mg/L and no detectable quantities of LCT, FRS was lower
than the proposed limit of 20 mg/100 g protein. There are fewer
reports about ESL/HP milks in the literature compared to UHT milk.
Birlouez-Aragon et al. (1998) found an average FRS content of
11.4 2.35 mg/100 g protein in HP market milks; the data of
Villamiel et al. (1999) showed 10.131.4 mg FRS per 100 g protein.
The most informative study about ESL milk is that of Mayer et al.
(2010), who analysed 71 market ESL milks and found FRS contents
from 11 to 262.2 mg/100 g protein; the majority of them contained
>20 mg FRS per 100 g protein. They report that according to their
study an upper limit for FRS corresponding to 1800 mg/L acidsoluble b-Lg for ESL milks would be 40 mg/100 g protein.
4. Conclusion
Compilation of the experimental and market milk analyses
showed that non-detectable quantities of LCT and FRS <3.5 mg/L
characterise ESL/HP milks with low heat damage i.e. residual native acid-soluble b-lg about half that of raw milk. According to
the present pilot-scale experiments, these values are consistent
with treatments at temperatures from 80 C to 90 C for 4 s, by
means of indirect heating. The residual native b-lg content of
heated milks is a reliable index for the assessment of heat treatment of market milk up to the HP/ESL category, whereas residual
native a-la content can be utilised for more severe heat treatments.
Both these proteins are in abundance in many categories of heattreated milks and can be easily and simultaneously estimated in
the same sample preparation and chromatographic analysis.
Although FRS is an index appropriate for all types of heated milks,
it is inuenced by the storage duration and conditions and its estimation demands a laborious sample preparation. Finally, in our
opinion a yes or no answer for LCT existence in heat-treated
milks along with the proposed limits for b-lg can be reliable tools
for the evaluation of heating load history of unknown milk samples. Contradictory results can be claried rstly by considering
residual native a-la and secondly by a subsequent furosine
analysis.
Acknowledgments
We are grateful to the R&D department of Delta S.A. dairy company, in the premises of which, the experimental pilot-scale heat
treatments were carried out.

References
Birlouez-Aragon, I., Nicolas, M., Metais, A., Marchond, N., Grenier, J., & Calvo, D.
(1998). A rapid uorimetric method to estimate the heat treatment of liquid
milk. International Dairy Journal, 8, 771777.
Cattaneo, S., Masotti, F., & Pellegrino, L. (2008). Effects of overprocessing on heat
damage of UHT milk. European Food Research and Technology, 226, 10991106.
Claeys, W. L., Ludikhuyze, L. R., Van Loey, A. M., & Hendrickx, M. E. (2001a).
Inactivation kinetics of alkaline phosphatase and lactoperoxidase, and
denaturation kinetics of b-lactoglobulin in raw milk under isothermal and
dynamic temperature conditions. Journal of Dairy Research, 68, 95107.
Claeys, W. L., Ludikhuyze, L. R., Van Loey, A. M., & Hendrickx, M. E. (2001b).
Formation kinetics of hydroxymethylfourfoural, lactulose and furosine under
isothermal and dynamic temperature conditions. Journal of Dairy Research, 68,
287301.
Claeys, W. L., Van Loey, A. M., & Hendrickx, M. E. (2002a). Kinetics of alkaline
phosphatase and lactoperoxidase inactivation and of b-lactoglobulin
denaturation in milk with different fat content. Journal of Dairy Research, 69,
541553.
Claeys, W. L., Van Loey, A. M., & Hendrickx, M. E. (2002b). Intrinsic time temperature
integrators for heat treatment of milk. Trends in Food Science and Technology, 13,
293311.
Claeys, W. L., Van Loey, A. M., & Hendrickx, M. E. (2003). Kinetics of
hydroxymethylfourfoural, lactulose and furosine in thermal processing of
milk different fat content. Journal of Dairy Research, 70, 8590.
Considine, T., Patel, H. A., Anema, S. G., Singh, H., & Creamer, L. K. (2007).
Interactions of milk proteins during heat and high hydrostatic pressure
treatments-A Review. Innovative Food Science and Emerging Technologies, 8,
123.
Corredig, M., & Dalgleish, D. G. (1996). Effect of different heat treatments on the
strong binding interactions between whey proteins and milk fat globules in
whole milk. Journal of Dairy Research, 63, 441449.
Corzo, N., Delgado, T., Troyano, E., & Olano, A. (1994a). Ratio of lactulose to furosine
as indicator of quality of commercial milks. Journal of Food Protection, 57,
737739.
Corzo, N., Lpez-Fandio, R., Delgado, T., Ramos, M., & Olano, A. (1994b). Changes in
furosine and proteins of UHT-treated milks stored at high ambient
temperatures. Zeitschrift fr Lebensmittel Untersuchung und Forschung, 198,
302306.
Elliott, A. J., Datta, N., Amenu, B., & Deeth, H. C. (2005). Heat-induced and other
chemical changes in commercial UHT milks. Journal of Dairy Research, 72,
442446.
Elliott, A. J., Dhakal, A., Datta, N., & Deeth, H. C. (2003). Heat-induced changes in UHT
milks-Part 1. The Australian Journal of Dairy Technology, 58, 310.
Feinberg, M., Dupont, D., Efstathiou, T., Loupre, V., & Guyonnet, J. P. (2006).
Evaluation of tracers for the authentication of thermal treatments of milks. Food
Chemistry, 98, 188194.
IDF Standard 147/ISO 11868 (2007). Heat-treated milk. Determination of lactulose
content. Method using high-performance liquid chromatography. Brussels:
International Dairy Federation.
IDF Standard 193/ISO 18329 (2004). Milk and milk products. Determination of
furosine
content.
Ion-pair
reverse-phase
high-performance
liquid
chromatography method. Brussels: International Dairy Federation.
Jeanson, S., Dupont, D., Grattard, N., & Rolet-Rpcaud, O. (1999). Characterization
of the heat treatment undergone by milk using two inhibition ELISAs for
quantication of native and heat denatured a-lactalbumin. Journal of
Agricultural and Food Chemistry, 47, 22492254.
Lpez-Fandio, R., & Olano, A. (1999). Review: Selected indicators of the quality of
thermal processed milk. Food Science and Technology International, 5, 121137.
Lorenzen, P. C., Clawin-Rdecker, I., Einhoff, C., Hammer, P., Hartmann, R.,
Hoffmann, W., Martin, D., Molkentin, J., Walte, H. G., & Devrese, M. (2011). A
survey of the quality of extended shelf life (ESL) milk in relation to HTST and
UHT milk. International Journal of Dairy Technology, 64, 166178.
Marconi, E., Messia, M. C., Amine, A., Moscone, D., Vernazza, F., Stocchi, F., &
Palleschi, G. (2004). Heat-treated milk differentiation by a sensitive lactulose
assay. Food Chemistry, 84, 447450.

L. Sakkas et al. / Food Chemistry 159 (2014) 293301

Mayer, H. K., Raba, B., Meier, J., & Schmid, A. (2010). RP-HPLC analysis of furosine
and acid-soluble b-lactoglobulin to assess the heat load of extended shelf life
milk samples in Austria. Dairy Science & Technology, 90, 413428.
Moatsou, G., Hatzinaki, A., Kandarakis, I., & Anifantakis, E. (2003). Nitrogenous
fractions during the manufacture of whey protein concentrates from Feta
cheese whey. Food Chemistry, 81, 209217.
Moatsou, G. (2013). Sanitary procedures, heat treatments and packaging. In Y. W.
Park & G. F. W. Haenlein (Eds.), Milk and dairy products in human nutrition:
Production, composition and health (pp. 289309). West Sussex, U.K.: WileyBlackwell, John Wiley and Sons Ltd.
Morales, F. J., & Jimnez-Prez, S. (1999). HMF formation during heat-treatment of
milk-type products as related to milk fat content. Journal of Food Science, 64,
855859.
Morales, F. J., Romero, C., & Jimnez-Prez, S. (2000). Characterization of industrial
processed milk by analysis of heat-induced changes. International Journal of
Food Science and Technology, 35, 193200.
Olano, A., Calvo, M. M., & Corzo, N. (1989). Changes in the carbohydrate fraction of
milk during heating processes. Food Chemistry, 31, 259265.
Pellegrino, L., De Noni, I., & Resmini, P. (1995a). Coupling of lactulose and furosine
indices for quality evaluation of sterilized milk. International Dairy Journal, 5,
647659.
Pellegrino, L., Resmini, P. & Luf, W. (1995b) Assessment (indices) of heat treatment
of milk. In: Heat-induced changes in milk, Special Issue IDF 9501, 2nd ed., pp.
409453. Brussels: international Dairy Federation.
Resmini, P., Pellegrino, L., & Batteli, G. (1990). Accurate quantication of furosine in
milk and dairy products by a direct HPLC method. Italian Journal of Food Science,
3, 173183.

301

Resmini, P., Pellegrino, L., & Cattaneo, S. (2003). Furosine and other heat-treatment
indicators for detecting fraud in milk and milk products. Italian Journal of Food
Science, 15, 473484.
Rysstad, G., & Kolstad, J. (2006). Extended shelf life milk-advances in technology.
International Journal of Dairy Technology, 59, 8596.
Serrano, M. A., Castillo, G., Muoz, M. M., & Hernndez, A. (2002). Inuence of
hydrolysis, purication, and calibration method on furosine determination
using ion-pair reversed-phase high-performance liquid chromatography.
Journal of Chromatographic Science, 40, 8791.
Van Boekel, M. A. J. S. (1998). Effect of heating on Maillard reactions in milk. Food
Chemistry, 62, 403414.
Van Renterghem, R., & De Block, J. (1996). Furosine in consumption milk and milk
powders. International Dairy Journal, 6, 371382.
Villamiel, A., Arias, M., Corzo, N., & Olano, A. (1999). Use of different thermal indices
to assess the quality of pasteurized milks. Zeitschrift fr Lebensmittel
Untersuchung und Forschung A, 208, 169171.
Villamiel, M., Lpez-Fandio, R., Corzo, N., & Olano, A. (1997). Denaturation of blactoglobulin and native enzymes in the plate exchanger and holding tube
section during continuous ow pasteurization of milk. Food Chemistry, 58,
4952.
Walstra, P., Wouters, J. T. M. & Geurts, T. J. (2006). Dairy Science and Technology,
2nd ed., (pp. 297314, 513535) Boca Raton, FL: CRC Press Taylor and Francis
Group.
Wilbey, A. R. (1996). Estimating the degree of heat treatment given to milk.
International Journal of Dairy Technology, 49, 109112.