DIABETES/METABOLISM RESEARCH AND REVIEWS RESEARCH ARTICLE

Diabetes Metab Res Rev 2001; 17: 5154.

DOI: 10.1002 /1520-7560(2000)9999 : 9999<: : AID-DMRR156>3.0.CO;2-H

Bovine b-casein antibodies in breast- and bottle-

fed infants: their relevance in Type 1 diabetes

Laura Monetini1 Abstract

Maria G. Cavallo1,4
Lavinia Stefanini1 Background Bovine b-casein is a cow's milk protein that targets both
humoral and cellular immune responses in patients with Type 1 diabetes and,
Federica Ferrazzoli2
to a lesser degree, also in normal subjects. In this study we aimed to
Carla Bizzarri2 determine whether the avoidance of cow's milk consumption early in life
Giovanni Marietti2 could prevent the development of antibody response to bovine b-casein
Vincenzo Curro2 despite the mother being exposed on a daily basis to cow's milk consumption.
Marco Cervoni3
Paolo Pozzilli1,4,5* Materials and Methods We measured the antibody response to bovine b-
the IMDIAB group{ casein using an ELISA method in 28 healthy infants under 4 months of age, of
whom 16 were exclusively breast-fed and 12 were bottle-fed with cow's milk.
1
University of Rome `Tor Vergata', In addition, b-casein antibodies were measured in 37 prepubertal children
Rome, Italy with Type 1 diabetes and in 31 healthy children who were exposed to cow's
2
University of Rome `Cattolica', milk or dairy products to see whether differences in antibody titers exist in
Rome, Italy this young age group. Antibodies binding to b-casein were also evaluated by
3
University of Rome `La Sapienza',
immunoblotting analysis.
Rome, Italy
4 Results Elevated levels of b-casein antibodies were found in bottle-fed
University of Rome `Campus
Biomedico', Rome, Italy infants compared to breast-fed infants ( p<0.0001). Antibody levels to bovine
5 b-casein were also signicantly higher in children with Type 1 diabetes
Department of Diabetes &
compared to age-matched controls ( p=0.03). By western blot analysis we
Metabolism, St Bartholomew's
Hospital, London, UK conrmed specic binding to bovine b-casein in bottle-fed infants, in children
with Type 1 diabetes and in controls exposed to cow's milk, but not in infants
*Correspondence to: Prof. P Pozzilli, who were exclusively breast-fed.
University Campus Biomedico, Via
Longoni 83, 00155 Rome, Italy. Conclusions The results of this study indicate that breastfeeding within the
E-mail: p.pozzilli @ caspur.it rst 4 months of life prevents the generation of antibody response to bovine
b-casein despite the mothers' consumption of cow's milk during the
{IMDIAB group: Paolo Pozzilli, Natalia
Visalli, Maria G. Cavallo, Alberto Signore, breastfeeding period. These ndings may have relevance for disease
Marco G. Baroni, Raffaella Buzzetti, Giusy prevention. Copyright # 2000 John Wiley & Sons, Ltd.
Coppolino, Chiara A. Mesturino, Rossana
Fiori, Lucio Lucentini, Maria C. Matteoli,
Antonino Crino , Stefania Spera, Carlo Keywords Type 1 diabetes; cow's milk; b-casein antibodies
Teodonio, Francesco Paci, Rita Amoretti,
Luigi Pisano, Concetta Suraci, Giuseppe
Multari, Nicoletta Sulli, Marco Cervoni,
Giancarlo De Mattia, Maria R. Cassone
Faldetta, Bruno Boscherini, Maria L. Introduction
Manca Bitti, Giovanni Marietti, Federica
Ferrazzoli, Carla Bizzarri, Dario Pitocco,
Giovanni Ghirlanda. It has been proposed that exposure to cow's milk early in life is an
environmental risk factor pertaining to Type 1 diabetes [1,2]. Although this
hypothesis has yet to be conrmed, a dietary intervention trial has been
implemented to study whether complete avoidance of cow's milk and the
Received: 13 June 2000 introduction of a hydrolysed formula milk during early infancy can prevent
Revised: 12 September 2000
Type 1 diabetes in individuals at genetic risk of developing the disease [3].
Accepted: 14 September 2000
Exposure to oral antigens such as proteins from cow's milk induces a
Published online: 22 November 2000
physiological immune response to these proteins and the antibody titers have

Copyright # 2000 John Wiley & Sons, Ltd.

52
been related to the precocity of exposure and the duration
of cow's milk consumption [4,5].
Signicantly high levels of antibodies to cow's milk
proteins have been reported in children with Type 1
diabetes, and the occurrence of these antibodies has been
retrospectively associated with a short breastfeeding
period and early exposure to cow's milk [6,7]. In addition,
by using a multivariate analysis, antibody levels to cow's
milk proteins including bovine serum albumin (BSA) and
b-lactoglobulin were found as an independent risk factor
for the disease when introducing islet cell antibodies
(ICA) or human leukocyte antigen (HLA) status.
Above all, the introduction of cow's milk at a very early
age appears to be crucial in triggering the immune
response to cow's milk. Therefore infants not exposed to
cow's milk or dairy products during the rst 45 months
of life should be protected from developing an abnormal
immune response to cow's milk proteins early in life.
In this study we measured IgG antibodies to bovine b-
casein (one of the cow's milk proteins inducing an
immune response in Type 1 diabetes) [8,9] in infants
below 4 months of age who were exclusively breast-fed
and in bottle-fed infants of the same age as control. In
addition, the antibody response to b-casein was studied in
prepubertal children with Type 1 diabetes and healthy
children who were both exposed to cow's milk or dairy
products.
Materials and methods
Subjects
Blood samples were collected from 15 exclusively breast-
fed and 12 bottle-fed healthy infants below 4 months of
age. Infants were recruited from the Department of
Paediatrics at the University of Rome `Cattolica' under-
going routine monitoring after delivery.
No family history of diabetes or other autoimmune
diseases was present in the normal infants or in the
healthy children tested as control.
Mothers of breast-fed infants were exposed to cow's
milk or dairy products on a daily basis during the period
of breastfeeding.
Venous blood was also collected, at the time of
diagnosis, from 37 consecutive children with Type 1
diabetes attending the Paediatrics Clinics of different
centres co-ordinated by the IMDIAB Study Group (mean
age 7.5 yearst2.1 SD), and from 31 randomly selected
age-matched healthy children (mean age 7.3 yearst
1.9 SD) recruited from the Department of Paediatrics
at the University of Rome `Cattolica'.
Information regarding cow's milk consumption, aller-
gies or intolerance were collected from infants, their
mothers, healthy and Type 1 diabetes affected children by
means of a questionnaire.
This study was approved by the Ethical Committee.
Written consent was obtained from parents of infants and
children included in the study.
Copyright # 2000 John Wiley & Sons, Ltd.
L. Monetini et al.
ELISA
Sera were obtained after centrifugation and stored at
x20uC until required.
Antibodies to bovine b-casein were measured according
to a previously described method [9]. Briey, at-bottom
96-well plates (MaxiSorp, NUNC, Denmark) were coated
with bovine b-casein (Sigma Chemical, St Louis, USA)
and the sera added to the wells in duplicate [1/200
dilution in bovine serum albumin (BSA) 1% in phosphate-
buffered saline (PBS)] for 1 h at room temperature. After
washing with PBS-Tween 0.05%, peroxidase-conjugated
anti human-IgG antibody (Sigma) was added to the plate
for 1 h at room temperature.
The enzyme reaction was detected by adding 3,3k,5,5k
tetramethylbenzidine (TMB) tablets (Sigma) in phos-
phate-citrate buffer 0.05 M. Finally, the absorbance (OD)
was read at 450 nm after stopping the reaction with
H2SO4 2 M.
Antibodies to b-casein were expressed in units obtained
by normalising against a pool of ten sera from patients
with Type 1 diabetes as a reference preparation [9].
Western blot
Puried cow's milk proteins (Sigma) and de-fat bovine
milk (Marvel, UK) were separated using a 12% acryla-
mide SDS gel under reducing condition. The proteins
were successively transferred onto a nitrocellulose
membrane (Amersham Life Science, Amersham, UK).
Non-specic binding was blocked using BSA 1% in PBS-
Tween 0.1% for 3 h on a rotating shaker at room
temperature.
Sera were incubated with the membrane at 1/500
dilution in BSA 1% in PBS-Tween 0.1% overnight at 4uC
on a rotating shaker. The membranes were successively
washed four times with PBS-Tween 0.1%. Peroxidase-
conjugated anti human-IgG antibody (Sigma) was added
to the membranes at 1/2000 dilution in 1% BSA in PBS-
Tween 0.1% for 1 h at room temperature.
Blots were washed ve times with PBS-Tween 0.1%
before the addition of ECL detection reagents (Amersham
Life Science) then exposed to autoradiographic lm
(Amersham Life Science) and developed.
Statistical analysis
The Mann-Whitney U test was used in the analysis of
statistical signicance between groups.
Results
ELISA
Levels of b-casein antibodies were signicantly higher in
infants exposed to cow's milk compared to those infants
who were exclusively breast-fed ( p<0.0001) (Figure 1).
Diabetes Metab Res Rev 2001; 17: 5154.
b-Casein Antibodies 53

Figure 1. Scattergram of single responses to bovine b-casein

in breast- and bottle-fed infants. Antibody levels were statisti-
cally higher in infants exposed to cow's milk compared
to breast-fed infants (1.19 Ut0.32 SD vs 0.24 Ut0.13 SD,
p<0.0001 Mann Whitney U test)
Table 1. Antibodies to bovine b-casein in children with Type 1
diabetes and in healthy control children
Figure 2. Binding to cow's milk proteins in serum from an
infant exposed to cow's milk. A strong antibody reactivity to
puried b-casein (lane 1) and b-casein present in cow's milk
(lane 2) was detected. Antibody binding to a-casein and
b-lactoglobulin from cow's milk (lane 2) was also observed

Subjects Number Age (yearstSD) Mean (unitstSD)

Patients 37 7.5t2.1 0.797t0.317*

Controls 31 7.3t1.9 0.638t0.334
*p=0.03 vs healthy control children.

Antibody titers in the breast-fed group were very low and

similar to the background binding activity.
One of the breast-fed infants showed high levels of b-
casein (indicated by an arrow in gure 1). This infant was
reported to have received fruit preparation containing
casein during weaning. In this case antibodies to bovine
b-casein might be due to the presence of casein in food
supplementation.
Children with Type 1 diabetes had signicantly higher
levels of antibodies to b-casein compared to the controls
( p=0.03) (Table 1).
Western blot
Western blot analysis was performed in a group of
subjects who were exposed (n=3) or not exposed (n=3)
to cow's milk. Specic binding to bovine b-casein was
detected in bottle-fed infants (Figure 2) and also in both
diabetic and control children exposed to cow's milk or
dietary products (data not shown). Binding to bovine b-
casein was not observed in infants who were exclusively
breast-fed (Figure 3).
Discussion
This is the rst study to examine bovine b-casein
antibodies in infants exposed to different feeding
patterns. We found that avoidance of cow's milk through
exclusive breastfeeding prevents the development of
antibody reactivity to bovine b-casein during the rst 4
Figure 3. Absence of antibody reactivity to puried b-casein
(lane 1) and cow's milk proteins (lane 2) in an exclusively
breast-fed infant
months of life despite the mother being exposed to cow's
milk during the breastfeeding period.
High levels of antibodies to bovine b-casein were
detected in bottle-fed infants under the age of 4 months,
whereas exclusively breast-fed infants of the same age did
not show an antibody response to this protein. This
observation is in concordance with the ndings of other
studies in which antibody response to other cow's milk
proteins including b-lactoglobulin and BSA was reported
in infants receiving cow's milk or hydrolysed formula milk
[4,5].
We have also demonstrated, by western blot analysis,
that antibody reactivity to bovine b-casein as well as to a-
casein and b-lactoglobulin is detectable only in bottle-fed
infants and not in infants who are exclusively breast-fed.
The importance of b-casein antibodies in Type 1 diabetes
is conrmed by the increased levels of such antibodies

Copyright # 2000 John Wiley & Sons, Ltd. Diabetes Metab Res Rev 2001; 17: 5154.
54
found in affected children compared to age-matched
controls. It is worth noting that the mean age of our
diabetic population was 7 years, well below the age
reported in another study [10] suggesting that b-casein
antibodies may develop early in life. However, not only
early exposure but also the duration of exposure to cow's
milk intake may be additional risk factors for Type 1
diabetes development, as recently suggested in another
publication [11].
Raised humoral [9,10] and cellular immune responses
[8,12] to bovine b-casein were previously reported in a
large population of Type 1 diabetic patients compared to
normal subjects. The highest level of these antibodies in
the diabetic group can be a consequence of the defective
development of oral tolerance to cow's milk proteins [1].
An impaired gut immune function has been proposed to
explain the association between cow's milk and Type 1
diabetes [13]. Therefore according to this hypothesis the
exposure to foreign proteins such as cow's milk, which
might be potentially cross-reactive with antigens
expressed by beta-cells, should be avoided. In this respect
bovine b-casein is of particular interest as it possesses a
sequence homology with GLUT-2, the glucose transporter
in beta-cells which is recognized as an autoantigen in
Type 1 diabetes [14].
In conclusion, our results indicate that there is no sign
of humoral immune response to bovine b-casein in breast-
fed infants despite mother's diet suggesting that immu-
nologically signicant amounts of b-casein are not
passively transferred to the infant via breast milk.
For the current debate on cow's milk as an environ-
mental trigger in Type 1 diabetes, our ndings reinforce
the concept that exposure to cow's milk should be avoided
during the rst months of life, at least in individuals at
genetic risk for the disease, as recommended in a recent
publication [15]. The interesting question still unan-
swered remains whether a delayed introduction to cow's
milk might be associated with lower antibody levels at an
older age.
Acknowledgements
This work was supported by grants from the Universities of
Rome `Tor Vergata', `Campus Biomedico' and the Centro
Copyright # 2000 John Wiley & Sons, Ltd.
L. Monetini et al.
Internazionale Studi Diabete (CISD) Rome, Italy. Laura Mon-
etini is supported by a fellowship from the Fondazione `Istituto
Pastuer-Cenci Bolognetti'.
References
1. Akerblom HK, Knip M. Putative environmental factors in Type 1
diabetes. Diabetes Metab Rev 1998; 14: 3167.
2. Kolb H, Pozzilli P. Cow's milk and immune mediated diabetes:
the gut immune system deserves attention. Immunol Today
1999; 20: 108110.
3. Knip M, Akerblom HK. IDDM prevention trials in progress a
critical assessment. J Pediatr Endocrinol Metab 1998; 11:
371377.
4. Keller KM, Wolff AB, Lippold R, Wirth S, Lentze MJ. The
diagnostic signicance of IgG cow's milk protein antibodies re-
evaluated. Eur J Pediatr 1996; 155: 331337.
5. Vaarala O, Saukkonen T, Savilahti E, Klemola T, Akerblom K.
Development of immune response to cow's milk proteins in
infants receiving cow's milk or hydrolyzed formula. J Allergy Clin
Immunol 1995; 96: 917923.
6. Savilahti E, Saukkonen TT, Virtala ET, Tuomilehto J, Akerblom
HK, and the Childhood Diabetes in Finland Study Group.
Increased levels of cow's milk and b-lactoglobulin antibodies in
young children with newly-diagnosed IDDM. Diabetes Care
1993; 16: 984989.
7. Sukkonen T, Virtanen SM, Karppinen M, et al. Signicance of
cow's milk protein antibodies as risk factor for childhood IDDM:
interactions with dietary cow's milk intake and HLA-DQB1
genotype. Diabetologia 1998; 4: 7278.
8. Cavallo MG, Fava D, Monetini L, Barone F, Pozzilli P. Cell
mediated immune response to b-casein in recent onset insulin
dependent diabetes: implications for disease pathogenesis.
Lancet 1996; 348: 926928.
9. Cavallo MG, Monetini L, Walker B, Thorpe R, Pozzilli P. Diabetes
and cow's milk. Lancet 1996; 348: 1655.
10. Sarugeri E, Dozio N, Meschi F, Pastore MR, Bonifacio E. Cellular
and humoral immunity against cow's milk proteins in Type 1
diabetes. J Autoimmun 1999; 3: 365373.
11. Virtanen SM, Hypponen E, Laara E, et al. Cow's milk
consumption, disease-associated autoantibodies and Type 1
diabetes mellitus: a follow-up study of diabetic children. Diabet
Med 1998; 15: 730738.
12. Ellis TM, Ottendorfer E, Jodoin E, et al. Cellular immune
response to b-casein: elevated but not specic for individuals
with Type 1 diabetes mellitus. Diabetologia 1998; 41: 731735.
13. Harrison LC, Honeyman MC. Cow's milk and Type 1 diabetes.
The real debate is about mucosal immune function. Diabetes
1999; 48: 15011507.
14. Inman LR, McAllister CT, Chen L, et al. Autoantibodies to the
GLUT-2 glucose transporter of b cells in insulin dependent
diabetes mellitus of recent onset. Proc Natl Acad Sci U S A 1993;
90: 12811284.
15. American Academy of Pediatrics. Breast-feeding and the use of
human milk. Pediatrics 1997; 100: 10351039.
Diabetes Metab Res Rev 2001; 17: 5154.