PDF Archive

Easily share your PDF documents with your contacts, on the Web and Social Networks.

Share a file Manage my documents Convert Recover PDF Search Help Contact

file .pdf

Original filename: file.pdf

This PDF 1.7 document has been generated by PDFium, and has been sent on pdf-archive.com on 05/04/2015 at 03:27, from IP address 50.142.x.x. The current document download page has been viewed 641 times.
File size: 662 KB (15 pages).
Privacy: public file

Download original PDF file

Document preview

Trends in Food Science & Technology 17 (2006) 423–437


Does homogenization
affect the human
health properties of
cow’s milk?
Marie-Caroline Michalski*,1
and Caroline Januel


UMR INRA 1253, Science et Technologie du Lait et
de l’Œuf, Agrocampus Rennes, 65 rue de
Saint-Brieuc, 35042 Rennes Cedex, France
During the processing of marketed milk, homogenization
reduces fat droplet size and alters interface composition by
adsorption of casein micelles mainly, and whey proteins.
The structural consequences depend on the sequence of the
homogenization and heat treatments. Regarding human
health, homogenized milk seems more digestible than
untreated milk. Homogenization favors milk allergy and
intolerance in animals but no difference appears between
homogenized and untreated milk in allergic children and
lactose-intolerant or milk-hypersensitive adults. Controversies appear regarding the atherogenic or beneficial bioactivity of some casein peptides and milk fat globule membrane
proteins, which might be enhanced by homogenization. In
children prone to type I diabetes, early cow’s milk
consumption would be a risk but no link was observed in
the general population and the effect of homogenization has
not been studied. In the current context of obesity and
allergy outbreaks, the impact of homogenization and other
technological processes on the health properties of milk
remains to be clarified.

* Corresponding author.
Now present at: INRA UMR 1235/INSERM U 449, Me´canismes
Mole´ culaires du Diabe` te, Faculte´ de Me´ decine R. Lae¨ nnec,
Universite´ Claude Bernard Lyon I, 8 rue Guillaume Paradin,
69372 Lyon cedex 08, France. Tel.: C33 47 877 1046; fax: C33
47 877 8762; e-mail: marie-caroline.michalski@sante.univ-lyon1.fr
0924-2244/$ - see front matter q 2006 Elsevier Ltd. All rights reserved.

Milk is a complex biological fluid composed of water,
fat, proteins (mainly casein micelles and whey proteins),
carbohydrates (mainly lactose) and quantitatively minor
though bioactive components: minerals, vitamins and
enzymes (Jensen, 1995). Cow’s milk is a nutritive food
regarding human health and a functional food on a
technological viewpoint. Fat is present in milk in the form
of fat globules in suspension in the aqueous phase. Milk fat
is composed mainly of triacylglycerols and some diglycerols, complex lipids and unsaponifiable lipophilic compounds (Table 1). The milk fat globules are surrounded by a
native biological membrane (Fig. 1) composed mainly of
phospholipids, proteins and enzymes, cholesterol, glycoproteins, vitamins: the milk fat globule membrane (MFGM;
Mather, 2000). Saturated fatty acids represent 60–70%
(w/w) of total milk fatty acids and unsaturated fatty acids
30–35% (mainly mono-unsaturated). Moreover, each fatty
acid has a preferential position on the triglycerol backbone
so that thousands different triacylglycerols are found in milk
fat (Jensen, 2002). The role of milk fat in contributing to
health and disease is quite controversial (Berner, 1993a).
While saturated fatty acid and cholesterol contents are
suspected to take part to the risk of coronary heart disease,
some milk lipids such as conjugated linoleic acids (CLA),
sphingomyelin and butyric acid would present anticarcinogenic properties (Parodi, 1997).
Regarding proteins, four casein classes coexist (as1, as2,
b, k) with similar composition, rich in glutamic acid, leucin,
serine, lysine and prolin. They are all phosphoproteins of
150–200 amino acids, differing in the number of phosphoseryl groups, the presence of cystein, carbohydrates and
content in some amino acids. Caseins, that do not present
secondary structure, are organized as so-called micelles via
hydrogen bonds, hydrophobic, electrostatic and disulfide
bonds, and different salts are involved (calcium, phosphorus, magnesium, citrate). The k-casein is located at the
surface of the micelle, which is held together by calcium
phosphate bonds. The soluble proteins (whey proteins)
represent only 15–22% of milk proteins but form a complex
group consisting of albumins, immunoglobulins and
proteose–peptones. They are globular proteins (presenting
secondary—a-helix and b-pleated sheet—to quaternary
structure) whose high lysine, tryptophan and cysteine
content provide a great nutritional value. They contain
less proline than the caseins and are compact molecules.

M.-C. Michalski, C. Januel / Trends in Food Science & Technology 17 (2006) 423–437


Table 1. Gross composition of milk lipids (adapted from Walstra, Geurts, Noomen, Jellama, & van Boekel, 1999; Jensen, 2002)

Neutral glycerides

Content in total
fat (%, w/w)

Fraction in
globule core (%)

Fraction in
MFGMa (%)

Fraction in skim
phase (%)





Free fatty acids





Phospholipids (incl. sphingomyelin)




Cholesteryl ester





CarotenoidsCvitamin A






Native milk fat globule membrane.

They are more heat sensitive than caseins (the latter are
stable until 100 8C): heat treatments cause their denaturation, aggregation and insolubilization. The denaturation and
insolubilization rates depend on the protein class and
physico-chemical conditions. b-Lactoglobulin (b-Lg) is
the most abundant whey protein (51%) and can bind and
transport small hydrophobic molecules such as retinol
(vitamin A precursor). a-Lactalbumin (a-La) represents
22% of whey proteins and is involved in lactose synthesis
(Jensen, 1995; Whitney, 1988). Individual milk proteins
have a wide range of beneficial health and functional effects
via bioactive peptides, such as anti-carcinogenic effects,
enhancement of certain physiological functions (Meisel,
2005; Silva & Maltaca, 2005), improved iron bioavailability
(Bouhallab & Bougle, 2004) or prevention of dental caries
(Aimutis, 2004). On the other hand, milk proteins are also
important food allergens (Host, 2002) and are suspected to
be involved in some diabetes cases (Schrezenmeir & Jagla,
2000). In this respect, the issue of milk as a healthy or
deleterious foodstuff remains controversial in many aspects.
Marketed milk is manufactured from raw milk obtained
from sane cows, subjected to various processing steps in
order to collect and preserve milk along the supply chain:
machine milking, cooling, cold storage, homogenization,
heat treatment, packaging and storage. Each one of
these processing steps induces changes in the intrinsic
quality of milk (Table 2). This article will focus on
homogenization, which results in the most profound
changes in the physical structure of milk and might result
in altered health properties. Homogenization is defined as
the process of subdividing the relatively large polydisperse
oil globules of a coarse oil-in-water emulsion into a large
number of smaller globules of narrow size range. Milk is
pressurized in order to destroy milk fat globules into fine
lipid droplets, thereby preventing the cream separation
(Mulder & Walstra, 1974). Though, homogenization does
not kill microorganisms, so that intensive heat treatments
(pasteurization or UHT, ultra high temperature process) are
necessary in order to preserve product microbiological

quality (Hui, 1993, Chapter 5). Pasteurization consists in
heating milk at 72 8C for 15 s then cooling it immediately. It
is sometimes necessary to heat milk up to 85 8C during 20 s.
The UHT process allows shortening of the heating step:
140–150 8C during a few seconds. Extended shelf-life of
chilled product can also be achieved by bacteria removal
using microfiltration (Saboya & Maubois, 2000). Heat
treatments are used to preserve milk easier, but they
enhance the impact of homogenization on the organization
of milk, and possibly on its quality and health properties.
Firstly, the different homogenization processes will be
summarized. The consequences of homogenization and heat
treatment sequences on milk structure will be described.
Finally, we will evaluate the possible effects of homogenization on the health properties of milk through literature
evidence or suggestions and highlight research area that
should be further explored.
Principles of homogenization
Throughout the Western world, commercial milk is
homogenized and heat-treated. Homogenization has been
set-up by August Gaulin at the early 20th century; it consists
in forcing pressurized milk (8–20 MPa) between a valve
needle and seat (Gaulin-type homogenizer), resulting in a
dramatical reduction in fat globule size due to shear stress,
inertial forces and cavitation. A second stage processed at
lower pressure dissociates aggregates formed at the first
stage (Pouliot, Paquin, Robin, & Giasson, 1991). Homogenization is usually operated at 60 8C, though pressure and
temperature conditions vary according to apparatus and
valve type (Paquin, 1999; Wilbey, 2002). Homogenization
efficiency increases with temperature from 42 to 72 8C and
stabilizes around 72–77 8C (Hui, 1993, Chpater 5).
According to Stoke’s law, the smaller milk fat globule
size dramatically decreases the cream separation rate that is
due to the density difference between milk fat and
the aqueous phase. To some extent, it also prevents
coalescence; the milk emulsion is thus more stable and
shelf-life increases.

M.-C. Michalski, C. Januel / Trends in Food Science & Technology 17 (2006) 423–437


Fig. 1. Organization of the native milk fat globule membrane (MFGM) compared with the interfacial organization of an homogenized fat
droplet, and proposed general organization of lipid particles in homogenized milk: ( ) native milk fat globule, ( ) casein micelle, ( ) fragment
of casein micelle, ( ) whey protein, ( ) fragments of MFGM (structure and location of the latter in the skim phase remain to be characterized).
Adapted from Michalski et al. (2001, 2002b).

Microfluidization is another homogenization technique
by which the fluid is forced under high pressure
in the reaction chamber and is divided into two jets
colliding at 1808 at high speed. At a given pressure, a
microfluidizer produces significantly narrower fat globule
size distributions compared with a regular homogenizer and
the mean diameter is also smaller (Hardham, Imison, &

French, 2000; Pouliot et al., 1991). This improves emulsion
long-term stability, therefore, microfluidization is advantageous for long shelf-life such as UHT products (Hardham
et al., 2000; Pouliot et al., 1991).
High pressure homogenization (HPH) is based on the
regular homogenization technique but is operated at higher
pressure (O50–100 MPa). It is used to disperse non-miscible

M.-C. Michalski, C. Januel / Trends in Food Science & Technology 17 (2006) 423–437


Table 2. Major milk changes induced by the processing chain (adapted from Korhonen and Korpela, 1994; Morr & Richter, 1988)
Unit process

Related reaction


Machine milking

Lipid oxidation
Dissolution of casein micelles
Fat crystallization
Fat globule disruption
Dispersion of casein micelles
Activation of some enzymes
Destruction of microorganisms
Whey protein denaturation
Lactone formation
Enzyme inactivation
Destruction of water-soluble vitamins
Destruction of water-soluble vitamins
Maillard reaction
Lactose isomerization
Reactivation of enzymes
Growth of psychrotrophic bacteria
Destruction of water-soluble vitamins
Age gelation
Destruction of water-soluble vitamins

Peroxides, oxidized taste
Free fatty acids, rancid taste
Aggregation of casein and calcium phosphate
Alteration of the milk fat globule membrane
Free fatty acids, rancid taste, oxidative off-flavor
Peptides, free amino acids
Smaller fat globules with a new interface
Formation of fat-protein complexes
Oxidized taste, rancid taste
Increased microbiological quality and shelf-life
Formation of casein–whey protein complexes
Enhanced flavor and taste
Increased quality and shelf-life
!10% Vitamin B; !25% Vitamin C
!20% Vitamin B; !30% Vitamin C
Lactose–protein complexes, partial loss of lysine
Formation of lactulose
Organoleptic defects (proteolysis, lipolysis)
Bitter taste due to proteolysis
!30% of Vitamins B and C
Formation of protein-mineral complexes
!50% Vitamin B; O90% Vitamin C

Cooling, agitation, cold


Heat treatment


Storage of packed:
Pasteurized milk
UHT milk

phases, stabilize emulsions and/or prepare products with
appropriate rheological properties (Floury, Desrumaux, &
Lardieres, 2000). The fat droplet size decreases when the
HPH pressure increases (50–200 MPa) and at given temperature and pressure, the HPH fat droplets are significantly
smaller than regularly homogenized ones (Hayes & Kelly,
2003a). However, HPH (40–60 MPa) results in off-flavors in
milk, probably due to oxidation. Homogenization at 20 MPa
does not cause such off-flavor (Humbert, Driou, Guerin, &
Alais, 1980). One advantage of HPH is to reduce bacterial
microflora (Hayes, Fox, & Kelly, 2005; Thiebaud, Dumay,
Picart, Guiraud, & Cheftel, 2003). The size of casein micelles
decreases at pressures greater than 200 MPa. HPH inactivates
plasmin (Hayes & Kelly, 2003b) and reduces alcaline
phosphatase and lactoperoxidase activities (Hayes et al.,
2005). The major drawback of high-pressure treatment is the
high cost of the equipment required. HPH is suggested to be
possible novel milk processing technique-combining advantages of homogenization and pasteurization in a single
process (Hayes et al., 2005).
Consequences of homogenization on milk
The composition of the MFGM is altered by heat
treatments and homogenization (Mulder & Walstra, 1974).
The effects of these treatments on milk organization are
often controversial. Fig. 1 schemes the organization of the
native MFGM and the solely homogenized fat droplet.
Moreover, the consequences of homogenization on the
organization of milk depend on the sequence and type of
homogenization and heat treatments, as also pointed out
using HPH with skim milk (Sandra & Dalgleish, 2005).

Effect of homogenization and heat treatments on
proteins, phospholipids and vitamins
The main effect of homogenization on soluble milk
components is the disruption of casein micelles while
adsorbing at the interface, in micellar form or as fragments.
Moreover, at least one of the agglutination factors is
inactivated (Walstra, 1980) and some MFGM components
are displaced to the skim milk phase during homogenization
(Keenan, Moon, & Dylewski, 1983). Bovine xanthine
oxidase (BXO) and butyrophilin are covalently bound to
fatty acids, contributing to the lipophilicity of these proteins
and the preservation of their affinity with the interface
despite the physical stress caused by homogenization. The
phospholipid content of the new membrane appears slightly
lower than that of the MFGM (McPherson, Dash, &
Kitchen, 1984a). The material loss appears not to be
selective. HPH decreases the casein micelle size in skim
milk, e.g. from w209 nm at 41 MPa down to w190 nm at
186 MPa (Sandra & Dalgleish, 2005); though, it is not
known whether micelle size decreases in HPH homogenized
whole milk.
During milk heating, chemical and enzymatic reaction
rates increase, as well as bacteria disruption (Renner, 1988).
Whey proteins are more heat sensitive than caseins. The
extent of denaturation depends on heat treatment type
(neglectable during pasteurization; 20% of whey proteins
denatured at 80 8C for 1 min vs 60% with UHT process) and
protein type, b-Lg being more heat sensitive than a-La
(Kinsella & Whitehead, 1989). Moreover, in skim milk,
heat-induced binding of denatured proteins to casein
micelles provokes pH-dependent increased micelle size
and inter-micelle interactions (Anema & Li, 2003; Jeurnink

M.-C. Michalski, C. Januel / Trends in Food Science & Technology 17 (2006) 423–437

& De Kruif, 1993). Caseins are much more heat resistant,
particularly b-casein. Several studies have shown that milk
heating induces complex formation between k-casein and aLa via a b-Lg–a-La complex (Elfagm & Wheelock, 1978).
Other studies have shown, using in vitro models (Haque,
Kristjanssan, & Kinsella, 1987; Jang & Swaisgood, 1990)
and skimmed milk (Dalgleish, 1990), that heat-induced
b-Lg–k-casein interactions are mainly due to hydrophobic
interactions and disulfide bond formation. During the
formation of such k-casein–whey protein complexes, both
soluble and micelle-linked aggregates are formed (Anema
& Li, 2003; Guyomarc’h, Law, & Dalgleish, 2003;
Vasbinder, Alting, & de Kruif, 2003). Primary b-Lg–a-La
aggregates seem to be implied in micellar aggreates as well
as k and as2-caseins. Heat-dissociated micellar k-casein is
implied in the formation of soluble aggregates and a
significant part of k-casein is not complexed after heat
Heat also favors glycation (also called non-enzymatic
glycosylation or Maillard reaction): the initial reaction
consists in nucleophilic condensation of a sugar freealdehyde group and a protein amine group (terminal or
often from lysine). Heated milk is sensitive to glycation
due to its lactose content (Morgan, Leonil, Molle, &
Bouhallab, 1999). Among vitamins, only vitamins B1, B6,
B12 and C and folic acid are heat sensitive, but watersoluble vitamins are sensitive to storage time (Renner,
1988; Table 2).
Effect of homogenization and heat treatments
on fat droplet size and z-potential
Shear stress and inertial forces induced by pumping are
maximal during homogenization (Corredig & Dalgleish,
1996). This induces fat globule disruption since the former
are greater than the Laplace pressure of the native milk fat
globules. In the following, we will thus refer to fat droplets,
different from the native milk fat globules. The native milk
fat globule size distribution spans from !1 to w20 mm,
with an average volumic size initially around 3–5 mm. Upon
homogenization, the latter is reduced to around 1 mm,
resulting in a 4-to 10-fold increase of the interface between
fat droplets and the aqueous medium (Keenan et al., 1983).
A relationship is given between the volume–surface average
diameter of the fat droplets (d32) and the homogenizing
pressure P: log10 d32ZaKb log10 P, with parameters such
as aZK2 to K1.8 and bZ0.6–0.71 (Michalski, Michel, &
Briard, 2002; Walstra, 1995). Particles between 100 and
400 nm even appear during regular homogenization
(Michalski, Michel, & Geneste, 2002), while the smallest
native milk fat globules (w100–200 nm; Walstra (1969))
should not be affected due to their higher Laplace pressure.
Heat treatment, when not associated with homogenization,
does not induce changes in the milk fat globule size. When
both processes are operated, regardless of order, fat droplets
tend to be smaller than in milk solely homogenized,
resulting in a greater droplet surface area (Lee, 1997;


Sharma & Dalgleish, 1994). In commercial pasteurized or
UHT full-fat milk, the d32 is in the range 0.20–0.25 mm,
calculated from the measured specific surface area (S) of
27–33 m2 gK1 (Lopez, 2005); the corresponding volumic
average diameter (d43) is in the range 0.37–0.49 vs 4.07 mm
for raw whole milk. Fave´, Coste, and Armand (2004) report
d43Z0.46–1.02 mm in half-skimmed commercial UHT milk
vs 4.36 mm in whole pasteurized organic milk. In
microfluidized milk, d32 values of 0.30 and 0.24 mm are
reported at 50 and 200 MPa, respectively (Olson, White, &
Richter, 2004).
The electrokinetic potential (z-potential) of a particle is
defined as the potential at the shear layer located farther out
the Stern layer. For native milk fat globules, average znative
is K13.5 mV (Michalski, Michel, Sainmont, & Briard,
2001). The absolute value of z increases with homogenizing
pressure for milk fat droplets, up to a plateau value of
z around K20 mV for strongly homogenized ones
(PO30 MPa). This value corresponds to the z-potential of
casein micelles adsorbed at the droplet interface (Fig. 1).
Indeed, z of homogenized droplets is linked to the surface
fraction F that is not covered by the native MFGM
anymore: zZ znative ½1C ðKLnð1KFÞ=10:82Þ1=2 (Michalski
et al., 2001). Heating at 80 8C (15 min) results in only small
changes in the micellar z (Anema & Klostermeyer, 1997).
The z-potential of whey protein–k-casein complexes is
reported to be K17 mV (Jean, Renan, Famelart, &
Guyomarc’h, 2006).
According to these structural measurements, homogenized milk is found to be composed of three types of particles
(Michalski et al., 2002b): (i) regular homogenized milk fat
droplets (disrupted globules from the main population,
whose surface fraction covered by caseins can be calculated
from their increase in specific surface area, the rest of
the surface being still covered by MFGM); (ii) small
(!500 nm) lipid–protein complexes having a new
membrane, presumably mainly composed of caseins; and
(iii) tiny native milk fat globules around 100 nm (that were
originally present in milk as a separate population and
should not be affected by homogenization due to their
small size).
Effect of homogenization and heat treatments
on the fat droplet interface
The rupture of fat globules occurring during homogenization creates a new interface that cannot be entirely
covered by the MFGM and can be measured by the
increased S of fat droplets. Therefore, other surface active
components adsorb and form a new membrane (Darling &
Butcher, 1978). Casein micelles are the major protein
fraction adsorbed, even if part of the native MFGM remains
associated to the fat droplets (Jackson & Brunner, 1960). In
a proportion increasing with P (Fox, Holsinger, Caha, &
Pallansch, 1960; Henstra & Schmidt, 1970), casein micelles
would spread onto the fat surface when colliding during

M.-C. Michalski, C. Januel / Trends in Food Science & Technology 17 (2006) 423–437


A fourfold increase of total proteins occurs in the
membrane when milk is either solely homogenized or
homogenized and heated (regardless of the sequence of
both treatments) compared to the total membrane proteins in
untreated or solely heated milk (Lee, 1997). After heating at
80–85 8C (10 min), total protein increases in the fat droplet
membrane (Dalgleish & Banks, 1991; Houlihan, Goddard,
Kitchen, & Masters, 1992). Heating increases the ability of
whey proteins to interact with MFGM proteins and/or caseins
adsorbed onto the membrane during homogenization. This
would not always be compensated by the desorption of
MFGM proteins that was highlighted by Houlihan et al.
(1992). If milk is homogenized (50 8C, 17 MPa) then
pasteurized (HTST), caseins represent 99% of adsorbed
proteins, among which b-casein represents 40.8%, as-casein
35.7% and k-casein 23.5% (Zahar & Smith, 1996). This does
not reflect untreated milk (Table 3): a preferential adsorption
of k- and b-caseins occurs during homogenization. Significant amounts of para-k-casein (N-terminal fragment of
k-casein) were detected in homogenized droplet membranes
(McPherson, Dash, & Kitchen, 1984b). This could be partly
explained by the action of heat stable proteinases on the
b-Lg–k-casein complexes associated with the membrane
(Garcia-Risco, Ramos, & Lopez-Fandino, 2002). However,
the relative amount of para-k-casein compared to b-Lg is
higher than observed in pasteurized milk (McPherson et al.,
1984b). Para-k-casein could also be formed by direct
k-casein hydrolysis at the fat droplet surface during
homogenization. It could preferentially adsorb onto lipid
droplet surface due to its higher hydrophobicity compared
with k-casein. However, several teams have found the b-Lg–
k-casein complexes formed during heat treatments in the
membranes of homogenized milk (Houlihan et al., 1992).
Denatured b-Lg is linked to the casein micelles adsorbed on
the fat droplets via k-casein (Dalgleish & Sharma, 1993).
Heat not only causes whey protein binding to adsorbed
micelles, but also reorganizations among casein micelles
themselves (Dalgleish & Sharma, 1993).
Caseins are the major protein fraction adsorbed.
Regarding whey proteins, b-Lg is the main one associated
with the lipid droplets but a small quantity of a-La was also
detected (Lee & Sherbon, 2002). Using infant formula
pasteurized then homogenized, the pasteurization step
favors b-Lg–k-casein interactions. The caseins and whey
proteins interact with the fat droplet membrane after
homogenization of these formulae (Guo, Hendricks, &
Table 3. Distribution of casein species in untreated milk (Hui,
1993, Chapter 5) and at the interface of homogenized and
pasteurized milk fat droplets (Zahar & Smith, 1996)


milk (%)

Homogenized and pasteurized
milk fat droplets (%)




Kindstedt, 1998). At average homogenization applied to
pasteurized milk, whey proteins make up about 5% of the
adsorbed protein and about 20% of the surface area covered;
for higher pressures, the proportion becomes increasingly
smaller (Sharma & Dalgleish, 1994). In commercial UHT
milk, about 25% of the droplet surface would still be coated
with MFGM (Lopez, 2005). The casein layer around fat
droplets appears thinner when milk is microfluidized rather
than regularly homogenized, suggesting micelle fragmentation (Dalgleish, Tosh, & West, 1996).
Observed differences depending on the sequence of
homogenization and heating steps
The differences depending on the sequence of homogenization and heating steps are rather controversial due to the
various treatments applied and to the sample preparation
procedure. Van Boekel and Walstra (1989) did not detect
whey proteins in the membrane of homogenized fat droplets
before the heating step. On the other hand, these proteins
associate to lipid droplets from 70 8C when denaturation
begins. Several reactions can take place: (i) interaction with
other denatured whey proteins, (ii) interactions with
k-casein on the surface of casein micelles in the skim
milk phase, (iii) interactions with k-casein located at the
exterior of the casein micelles adsorbed on the fat globule
surface, (iv) interactions with residual native MFGM
material, and (v) direct interaction with the fat droplet
surface. The availability of k-casein is crucial in these
different phenomena. In homogenized milk, the interface
between the adsorbed micelle fragments and the fat surface
is formed of non-k-caseins. Therefore, k-casein is exposed
on the outside, which favors interactions between whey
proteins and proteins at the interface (Dalgleish & Banks,
1991). When milk is homogenized then heated, less caseins
and whey proteins would be adsorbed onto the lipid droplet
surface than when milk is heated then homogenized (Lee,
1997). This is conflicting with the study of Sharma and
Dalgleish (1994) highlighting more whey protein–lipid
droplet membrane interactions when milk is homogenized
then heated, suggesting that the newly formed fat droplet
membrane offers more available binding sites for whey
proteins after homogenization than under its native MFGM
Heat treatment at 80 8C (3–18 min) induces incorporation of whey proteins, particularly b-Lg, in the MFGM
(Lee & Sherbon, 2002). This implies increased membrane
protein concentration. The glycoproteins PAS-6 and PAS-7
(so-called lactadherin) would disappear and the linkage of
b-Lg to the MFGM could be due to disulfide bonds with
membrane proteins. Homogenization (two stages, 50 8C, 17
and 3.5 MPa) causes casein adsorption onto the MFGM, but
no whey protein adsorption if not associated to heat
treatment (Lee & Sherbon, 2002). The total protein amount
(caseins and whey proteins) at the MFGM is not
significantly different whether homogenization is performed
before or after heat treatment. In another study (Lee, 1997),

M.-C. Michalski, C. Januel / Trends in Food Science & Technology 17 (2006) 423–437

whey protein amount in the lipid droplet membrane
increases when milk is heated then homogenized. This
suggests that the k-casein–whey protein complexes formed
during heat treatment adsorb onto the newly formed
membrane during homogenization. Conversely, if heating
is performed before homogenization, then the proteins are
already denatured and are less prone to take part to
interactions at the fat droplet surface (Dalgleish & Sharma,
1993). The membrane of homogenized droplets is thus
thinner and globule aggregation is favored. Two-stage
homogenization limits this phenomenon (van Boekel &
Walstra, 1989). If heat treatment is performed prior to HPH
of skim milk (41–186 MPa), more micellar material and
b-Lg–k-casein complexes are displaced to the skim milk
phase (Sandra & Dalgleish, 2005); this would turn them
potentially available for adsorption at the fat interface if the
same occurred in whole milk but this hypothesis is not
justifiable to date.
Briefly, when heat treatment is performed prior to
homogenization: (i) whey proteins are denatured and
interact with the native proteins of the MFGM and the
micellar caseins, particularly k-casein, and (ii) the casein–
whey protein complexes adsorb onto the lipid droplet
interface. When homogenization is performed prior to heat
treatment: (i) the semi-intact casein micelles or micellar
fragments cover the fat droplet interface, and (ii) the
denatured whey proteins link to the native MFGM proteins
and adsorbed caseins via disulfide bonds. However, the
compositional changes in the fat droplet membrane
depending on the order of homogenization and heat
treatments do not seem to influence cream separation.
Cream separation is identical whether milk is (i) homogenized, (ii) heated then homogenized, or (iii) homogenized
then heated, and always lower than for unhomogenized milk
(Hillbrick, Mcmahon, & Mcmanus, 1999; Lee, 1997). The
homogenization step can thus be performed prior to UHT
treatment, allowing lower asepsis rules and thus lower
industrial costs.

Current data related to homogenization effects
on the health properties of milk
Taste and digestion
A link exists between sensory stimulation and postprandial lipid metabolism (Mattes, 1996): the amount of
post-prandial plasma triacylglycerols after the ingestion of
oil capsules was measured in subjects exposed to various
oral stimuli masticated but not ingested. Higher plasma
triacylglycerols were observed in subjects in contact with
the fattest stimulus (cream cheese). Therefore, the sensory
properties homogenized vs unhomogenized milk could
affect the metabolic response. Due to increased photochemical sensitivity and lipolysis, homogenized milk is
more sensitive to off-flavor formation during storage
(Humbert et al., 1980). Combined homogenization and
heat treatment also increase the viscosity of whole milk


(Lee, 1997). The possible sensory-stimulated effects of
these differences on milk digestion should be investigated.
During digestion of homogenized milk, a simultaneous
coagulation of caseins and lipid droplets occurs in the
stomach. The structure of coagulated matter is much finer
than for untreated milk and the protein transfer to the small
intestine is easier: for subjects suffering intestinal disease,
homogenized milk is more easily digestible than untreated
milk (Sieber, Eyer, & Luginbuhl, 1997). In minipigs, raw
milk and pasteurized milk give a very firm curd and present
slower gastric emptying rate than pasteurized homogenized
milk, UHT milk or cultured milk (Meisel & Hagemeister,
1984). Moreover, the proteolysis of casein is enhanced with
pasteurized homogenized milk and UHT milk (Pfeil, 1984).
UHT milk also results in a greater absorption in this animal
model (Kaufmann, 1984).
Regarding lipid digestion, the gastric step is crucial since
it facilitates subsequent triacylglycerol hydrolysis by the
pancreatic lipase (Fave´ et al., 2004). This is particularly
important in infants and in adult patients suffering
pancreatic insufficiency. In minipigs, Buchheim (1984)
found extended lamellar structures of mono-glycerides in
the gastric coagulum of pasteurized milk (homogenized and
non-homogenized) and of UHT milk, but only rarely in raw
milk and cultured milk, providing direct evidence for
lipolysis that occurs to a considerable extent in the former
milks. After feeding raw milk and (pasteurized and
homogenized) cultured milk, only slight gastric lipolysis
is observed in minipigs (Timmen & Precht, 1984). In
humans, the lipid droplet size is a key physico-chemical
factor governing fatty acid bioavailability: smaller droplets
result in greater lipolysis via their surface excess on a larger
interface area (Armand et al., 1999; Fave´ et al., 2004). The
small sized droplets in homogenized milk would thus favor
milk fat lipolysis. But in premature infants, human milk fat
globules (surrounded by a human native MFGM similar to
Fig. 1) result in a more efficient gastric lipolysis than the
much smaller homogenized lipid droplets of infant formula
(Fave´ et al., 2004). Human milk fat globules are larger in
colostrums (d32Z4.3 mm) than in mature breastmilk (d32Z
3.5 mm), and much larger than infant formula droplets
(d32Z0.3 mm; Michalski, Briard, Michel, Tasson, &
Poulain, 2005). The ultrastructure of milk fat droplets
appears thus to be of utmost importance (Fave´ et al., 2004);
it can be related to the above-mentioned changes in
z-potential that could affect lipase access to the interface.
In rats, small homogenized fat droplets fed as a cream result
in a slower triacylglycerol metabolization than large
phospholipid-coated droplets or unemulsified fat
(Michalski, Briard, Desage, & Ge´loe¨n, 2005; Michalski
et al., 2006). The slower metabolization can be linked to the
delayed gastric emptying due to the gastric clot structure,
even if small droplets are more efficiently lipolyzed. The
discrepancies with minipig studies can be due to the
different animal models and the different fat content of the
gastric clots. Long-term effects of these metabolization


M.-C. Michalski, C. Januel / Trends in Food Science & Technology 17 (2006) 423–437

differences compared with untreated milk fat globules
remain to be elucidated in humans.
Atherosclerosis and coronary heart disease
In atherosclerosis, arteries are partially or totally
obstructed due to the formation of plaques rich in
cholesterol at the internal face of the vascular wall. Along
time, clotting can occur, obstructing arteries and depriving
vital organs of oxygen. This is why it is advised to control
cholesterol level by limiting saturated fat consumption via
butter, milk and fat meat. However, the role of these foods
in promoting cardiovascular diseases is controversial. The
amount of absorbed cholesterol via dairy products consumed daily only represents 15% of the daily recommended
cholesterol intake and the beneficial role of milk fat has
been highlighted in an in-depth review by Berner (1993b).
Studies that used single dietary fat sources to compare
effects of fats on blood lipids should be taken with caution
since the effects of any single fat source will be diluted when
in a mixed diet. The food energy intake as lipids should be
7.5% saturated fatty acids, 15% mono-unsaturated fatty
acids and 7.5% polyunsaturated fatty acids (PUFA). But
within one group, different fatty acids do not act the same
way, and the impact of milk products on plasma lipids and
on the risk of cardiovascular disease is different from
expected considering their lipid content and composition
(Berner, 1993b).
Atherosclerosis develops during the post-prandial lipemia stage and some studies have recently shown that
different structured dairy products result in different lipemia
profiles. Consequently, it is not justified to qualify milk as
anti- or pro-atherogenic regarding solely its lipid composition. Milk, mozzarella–cheese and butter in test meals do
not result in the same timing of triacylglycerol peak in type
II diabetic patients (Clemente et al., 2003). Fermented milk
results in a slower gastric emptying rate than regular milk
(certainly both homogenized although not stated), and in a
greater increase and a quicker decrease of the triacylglycerol content in all lipoprotein fractions (Sanggaard
et al., 2004). Controlled dietary studies in humans have
shown no difference in the effect on plasma cholesterol of
milk and butter with equal fat content and adjusted
regarding lactose and casein content (Tholstrup, Høy,
Normann Andersen, Christensen, & Sandstro¨m, 2005). In
a careful review, Tholstrup (2006) concludes that there is no
strong evidence that dairy products (i.e. including homogenized milk) increase the risk of coronary heart disease in
healthy men of all ages or young and middle-aged healthy
women. Overall, studies would be needed in humans to
investigate the effect of homogenization on the anti- or proatherogenic properties of milk.
Oster (1972) hypothesized that BXO released from the
MFGM due to homogenization would favor atherosclerosis.
The role of BXO in the generation of reactive oxygen
species in the cardiovascular system was also emphasized
(Berry & Hare, 2004). However, the former hypothesis is

criticized on many grounds such as the inactivation BXO at
the acidic gastric pH (Mangino & Brunner, 1976). The
possibly harmful effects of BXO have finally been the
subject of questions to the European Parliament in
September 2000 and April 2001 (written questions
E-2907/00 and E-0864/02). In both cases, the Commission
responded (i) not to have sufficient proofs regarding harmful
enzyme effects and (ii) not to consider new labelling rules.
Today, even if this hypothesis is often discussed, the
possible atherogenic role of BXO enhanced by homogenization appears to be rejected by the scientific community.
Coronary diseases are due to the development of
arteriosclerosis in coronary arteries (arteries that bring the
oxygen necessary for the heart to the myocardium).
Arteriosclerosis is characterized by the deposit of more or
less calcified plaques in the internal wall of coronary
arteries. Gradual narrowing (stenosis) of arteries results,
possibly until their final destruction. Some casein-derived
peptides present anti-thrombotic and anti-hypertensive
features; e.g. the k-casein fragment f103–111 can prevent
blood clotting through inhibition of platelet aggregation
(Silva & Maltaca, 2005). Also, recent studies point out that
(i) milk drinking may be associated with a small but
worthwhile reduction in heart disease and stroke risk
(Elwood, Pickering, Hughes, Fehily, & Ness, 2004; even
though the definition of ‘milk intake’ presents some
difficulties, Tholstrup, 2006), and (ii) milk product intake
is negatively associated with cardiovascular disease risk
factors (Warensjo¨ et al., 2004). Overall, no unfavorable
effect of dairy product could be found in these studies
involving the consumption of heat-treated and homogenized
milk. Since, these treatments involve the reorganization of
casein components (especially k-casein) within the milk
structure, one could suggest that the effect of milk
processing factors should be examined in respect with the
desirable bioactivity of casein-derived peptides.
On the other hand, Moss and Freed (2003) recently
studied the link between coronary disease occurrence and
circulating antibodies against the MFGM proteins. The
latter might by atherogenic by causing the aggregation of
lymphocytes and platelets. However, Spitsberg (2005)
criticizes this suggestion on analytical grounds. Besides,
we should stress that some populations, such as in the
French region of Brittany, use to consume high amounts of
buttermilk that is rich in MFGM fragments, while they are
not associated with the highest coronary mortality within
Northern France (Oberlin, Moquet, & Folliguet, 2004).
Also, hard cheese consumption is negatively correlated with
coronary heart disease (Moss & Freed, 2003; Tholstrup,
2006), although this product is rich in MFGM. Though, we
should highlight that buttermilk and hard cheese are
manufactured from unhomogenized milk. Since, homogenization changes the organization of the MFGM and the
exposure of its proteins, this treatment might trigger the
putative atherogenic effect of these proteins. Studies are
needed to elucidate this point.

M.-C. Michalski, C. Januel / Trends in Food Science & Technology 17 (2006) 423–437

Finally, the A1 variant of b-casein in cow’s milk yields
b-casomorphin 7 (Jinsmaa & Yoshikawa, 1999), a bioactive
peptide with the ability to catalyze the oxidation of LDL that
is implicated in the cardiovascular risk (Allison & Clarke,
in press). The issue of the A1-b–casein variant effect on
coronary heart disease appears to be highly controversial
(Truswell, 2005; Woodford, 2006). Since homogenization
has an effect on milk structure, particularly regarding casein
distribution, we should wonder whether milk processing has
an impact on the possible role of A1-b-casein in coronary
heart disease.
Lactose intolerance and milk allergy
Lactose intolerance results from a lack of lactase
necessary for its proper digestion. Delayed gastric
emptying has been proposed as one possible explanation
for improved lactose tolerance after ingestion of milk with
a meal instead of milk on its own (Vesa, Marteau, &
Korpela, 2000) but no clear conclusion can be drawn
(Korhonen & Korpela, 1994; Vesa et al., 2000). Paajanen,
Tuure, Poussa, and Korpela (2003) studied adults tolerant
to lactose, with a subjective tolerance to unhomogenized
milk but describing subjective intolerance to homogenized
milk. This study revealed no symptom difference between
homogenized and unhomogenized milk consumers. A
following study dealt with adults intolerant to lactose
(Korpela, Paajanen, & Tuure, 2005). No significant
difference is observed in the symptomatic response
between unprocessed organic milk and processed milk.
Even though some subjects subjectively experience a better
tolerance of unhomogenized than homogenized milk, this
is not the case in lactose intolerant subjects in general
(Korpela et al., 2005).
Cow’s milk protein allergy (CMPA) is an abnormal
reaction of the immune system to proteins contained in
cow’s milk. The incidence of allergy in early childhood is
2–3% (Host, 2002). The major allergizing proteins are b-Lg,
a-La and caseins, causing anaphylactic reactions (immune
phenomena induced by a type I hypersensitivity reaction to
the IgE mediation). Decreasing levels of milk-specific IgE
might signify allergy resolution. In animal models,
homogenization seems to favors hypersensitivity. Homogenization and pasteurization enhance the humoral
immune response of rats charged intraperitoneally with
milk (Feng & Collins, 1999). Homogenized milk orally fed
to hypersensitive mice induces an anaphylactic chock
(Poulsen & Hau, 1987), increases milk-specific IgE
production (Nielsen, Poulsen, & Hau, 1989), increases the
mass of intestinal segment and induces mastocyte degranulation (Poulsen, Nielsen, Basse, & Hau, 1990). Moreover,
the allergenicity of homogenized milk in mice increases
with increasing fat content (Poulsen, Hau, & Kollerup,
1987). On the other hand, unhomogenized cow’s milk
induces few or no such symptoms and immune responses.
However, when milk is given intravenously (Poulsen &


Hau, 1987) or subcutaneously (Poulsen et al., 1990), the
same reactions are observed regardless of milk treatment.
In Northern countries, many consumers and also parents
of allergic children state that they tolerate untreated cow’s
milk and pasteurized non-homogenized milk, conversely to
homogenized milk. The explanation could be that during
homogenization, the milk fat presents a dramatically
increased surface onto which allergenic milk proteins
adsorb. In untreated milk, many of the antigenic proteins
are located inside casein micelles. In homogenized milk, the
amount of exposed antigenic proteins increases (Poulsen &
Hau, 1987). Besides, there is also some release of MFGM
proteins (listed in Table 4) in the aqueous phase (the latter
were suggested to be potential allergens, though with no
clear-cut evidence, in a viewpoint by Riccio, 2004).
However, the amount and exposure of allergenic proteins
in untreated milk appear to be sufficient to induce allergic
reactions in some subjects. Moreover, clinical studies reveal
no difference between homogenized and unhomogenized
milk in children allergic to milk (Host & Samuelsson, 1988)
or in adults intolerant to lactose or hypersensitive to milk
(Pelto, Rantakokko, Lilius, Nuutila, & Salminen, 2000). In
one study, homogenized pasteurized milk was found less
suitable than unhomogenized milk in 10% of the children
subjects with milk protein allergy (Hansen, Host, &
Osterballe, 1987). But few studies concern subjects with a
better milk tolerance. No difference was found in the
immunological responses to homogenized and unhomogenized milk in healthy adults with a good tolerance of milk
(Paajanen, Tuure, Vaarala, & Korpela, 2005). Accordingly,
a recent review points out that homogenization does not
change the allergenic potency of cow’s milk (Paschke &
Besler, 2002). However, Paajanen et al. (2005) point out the
possibility that homogenized and unhomogenized milk
could induce different types of primary immunization to
cow’s milk antigens in immunologically intact individuals,
i.e. in infants. Moreover, most milk proteins, even minor
proteins, are potential allergens (Wal, 2004). This can
explain why the effect of homogenization may be difficult to
observe, since individuals can be sensitive to various
epitopes and since some human groups can be more
sensitive than others. The available evidence is not sufficient
to predict reliably the effect of food processing on allergenic
potential of milk proteins (Wal, 2004).
Food processing and interactions between constituents
and additives are strongly suspected to be responsible for at
least part of the increase of allergy incidence (Sanchez &
Fre´mont, 2003). Heating may have no effect or it may
decrease or increase allergenicity. Even in the absence of
heating, interactions between proteins and other components of food can cause conformational changes in
allergens, thereby affecting their thermal stability. The
effect of such interactions on the allergenicity of proteins is
practically unknown today (Sanchez & Fre´mont, 2003). The
most important consequence of heating at common milk
processing temperatures seems to be the increased

Related documents

raw milk
casein and caseinate market
amniotic membrane extract for acute chemical burns
lamour colsurfb 2009
datbiology russelldw

Related keywords