Molecules and Cells in Allergy – Original Paper

Int Arch Allergy Immunol 2017;173:147–152 Received: January 12, 2017

Accepted after revision: May 23, 2017
DOI: 10.1159/000477725
Published online: August 9, 2017

Association between Caesarean

Delivery and Isolated Doses of Formula
Feeding in Cow Milk Allergy
Francisco Gil a Ana Amezqueta b Diana Martinez b Elena Aznal b
Veronica Etayo c Teodoro Durá b Félix Sánchez-Valverde b
a
Department of Pediatrics, Estella Hospital, Estella, b Department of Pediactrics, Navarra Hospital Complex,
Pamplona, and c Health Center of Tafalla, Tafalla, Spain

Keywords (OR 0.29; 95% CI 0.09–0.92) to be a protective factor. Conclu-

Cow milk allergy · Caesarean delivery · Breastfeeding · sions: Perinatal factors play a key role in the development of
Formula feeding CMA IgE+, with an influence of breastfeeding duration, FFH
and caesarean delivery as risk factors and prematurity as a
protective factor. While family history had no important role,
Abstract environmental factors were more decisive.
Background: Cow milk allergy (CMA) is the most common © 2017 S. Karger AG, Basel
food allergy in breastfed infants. The aim of this study is to
verify whether certain perinatal factors may influence the
development of CMA immunoglobulin E (IgE)+. Methods: A Introduction
retrospective, observational study of case and control groups
was carried out. Information was collected of patients with Cow milk allergy (CMA) is the most common food al-
CMA IgE+ from our department during the years 1990–2013. lergy in breastfeeding (BF) infants and young children
Patients of the same age and sex were recruited for the con- and it can affect 2% of children under 5 years of age and
trol group. Information on the following variables was col- 0.9% of the general population [1–3]. CMA presents a
lected: sex, age, pregnancy tolerance, duration of pregnan- wide variety of clinical symptoms as immune responses
cy, type of delivery, isolated doses of formula feeding in hos- to cow’s milk proteins (CMP) can be mediated by immu-
pital (FFH), duration of breastfeeding, and family history of noglobulin E (IgE) and not mediated by IgE, fundamen-
allergy (defined as ≥1 first-degree family member with aller- tally type III (immune complex) and type IV reactions
gic disease). Statistical analysis was performed using multi- (mediated by T cells or a delayed type) [4, 5]. The diagno-
variate logistic regression techniques. Results: A total of 211 sis is based on clinical and allergy studies, with provoca-
cases were included in this study. Multivariate analysis tion testing being the gold standard for diagnosis [6]. The
showed an influence of duration of breastfeeding, FFH to be treatment of choice is initially a diet free from CMP [4, 6].
a risk factor (OR 4.94; 95% CI 2.68–9.08), especially in caesar- CMA is a disease that generates high health care costs
ean delivery (OR 11.82; 95% CI 2.64–47.50), and prematurity [7–10]. In primary prevention strategies recommended
131.211.208.19 - 8/13/2017 3:12:13 PM

© 2017 S. Karger AG, Basel Correspondence to: Dr. Francisco Gil

Departament of Pediatrics, Estella Hospital
University Library Utrecht

22 Santa Soria Street

E-Mail karger@karger.com
ES–31200 Estella (Spain)
www.karger.com/iaa
Downloaded by:

E-Mail fj.gil.saenz @ cfnavarra.es
by clinical guidelines, children at risk are defined as those Table 1. Descriptive analysis of age and gender in the CMA and
whose parents or siblings have a history of allergic disease control groups
[6]. These strategies have been based mainly on dietary
CMA IgE+ Control p
restrictions for mothers during pregnancy and lactation,
suggesting that they avoid or delay introducing certain Age, years 14.41 ± 5.42 14.59 ± 4.09 0.713a
foods that are considered allergenic [6]. However, such Male gender 128/211 (60.7) 128/211 (60.7) 1.000b
recommendations have been found to not be effective
[11–14]. BF has been postulated as a protective factor Values are presented as means ± SD or numbers/total (%) un-
less otherwise stated. CMA, cow milk allergy; IgE, immunoglobu-
against the development of CMA, although evidence lev- lin E. a Student t test for independent samples. b Pearson χ2 test.
els are somewhat inconsistent [14]. There is also not
much evidence regarding the use of special formulas [15,
16], probiotics [17], or prebiotics [18] by patients at risk.
The control sample was found in the same healthy reference pop-
Other factors that have been studied are caesarean deliv- ulation according to a criterion of no difference in the distribution
ery [19–23] and perinatal exposure to small amounts of of basic biological variables (age and gender) from the study cases.
CMP [24–27], though the results are conflicting. Other The control sample was recruited using a direct questionnaire at
factors such as duration of BF [14, 28, 29], prematurity, schools and health centers. Control recruitment was performed after
threatened premature delivery, preeclampsia [30], hyper- planning stratification by gender and age, with different numbers of
participants in every group according to the case distribution. They
tension, gestational diabetes [31], and antibiotics [21] were collected unselectively in order of arrival to complete every stra-
during labor have been much less studied. The purpose tum previously established and then the recruitment was stopped, so
of this study is to analyze the impact different perinatal a database of the excluded controls was not generated. The inclusion
factors may have on the development of IgE-mediated criteria were: individuals from the same reference population who
CMA. regularly tolerated the consumption of dairy products, excluding pa-
tients with a previous or current history of CMA.
The following variables were collected for both the case and the
control groups: date of birth, pregnancy tolerance (poor pregnan-
Materials and Methods cy tolerance was defined as presence of any of the following condi-
tions in the mother during pregnancy: hypertension, preeclamp-
A retrospective, observational case-control study was conduct- sia/eclampsia, and threatened preterm birth or gestational diabe-
ed in Navarra, Spain, after receiving approval from the Clinical tes; the absence of these conditions was the reference category),
Research Ethics Committee. The cases were recruited in the Pedi- duration of pregnancy (pregnancy was defined as premature when
atric Gastroenterology and Nutrition Unit of the Hospital of Na- the baby was born at a gestational age of less than 37 weeks and as
varra, the tertiary referral hospital of the Navarra Health Service, term when they baby was born at a gestational age of more than 37
which offers universal assistance to the people of Navarra (Au- weeks, with term delivery being the reference category), type of
tonomous Community of Spain, estimated population: 640,154) delivery (categorized as vaginal birth or caesarean delivery, with
[32]. All patients were recruited in the outpatient pediatric gastro- vaginal birth being the reference category), formula feeding in hos-
enterology consult at the time of diagnosis in the first year of life, pital (FFH) (categorized as patients who received some isolated
and any refusal to be included in this observational study was reg- doses of formula during maternity and then continued with BF
istered. Clinical data of all of the patients diagnosed with CMA exclusively; the reference category was comprised of patients who
were collected from June 1990 to June 2013 via consultation of received no formula supplement and were discharged with exclu-
medical records and a recruitment questionnaire. The inclusion sive BF or those who were discharged with formula feeding or
criteria for the cases were: previously healthy patients with a clini- mixed BF), duration of BF (categorized as less than 1 month, from
cal profile suggestive of CMA (anaphylactic shock, urticaria, an- 1 to 4 months, from 4 to 6 months, and more than 4 months, with
gioedema, perioral syndrome, atopic dermatitis, vomiting, diar- BF duration longer than 6 months as the reference category), fam-
rhea, rejection of feeding, spasmodic cough, allergic rhinitis, otitis, ily history of allergy (FH) (defined as one or more first-degree rel-
and other compatible symptoms) who also presented specific IgE atives with a history of allergic disease; no first-degree relatives
antibodies to CMP (IgE specific to casein, β-lactoglobulin and with a history of allergic disease was the reference category). Ad-
α-lactalbumin, IgE cutoff >0.35 kU/L) in the initial evaluation with ditionally, for the cases, age and clinical manifestations when di-
confirmation via a positive provocation test (except for those cases agnosed were collected.
in which a provocation test was considered contraindicated). For statistical analysis, the Pearson χ2 test was used for dichot-
Provocation tests were performed on all of the patients, except omous variables for comparison of proportions. To calculate OR
those with a history of anaphylaxis or severe allergic reaction after and 95% CI, logistic regression was used. A multivariate logistic
exposure to CMP or who presented very high levels of specific IgE regression model was employed for different predictor variables
in the initial determination (>2.5 kU/L). We excluded all patients (i.e., pregnancy tolerance, duration of pregnancy, type of delivery,
with a history of chronic digestive diseases and patients who had FFH, duration of BF, and FH, tested a priori for confusion and in-
already been diagnosed with CMA IgE+ who had undergone pre- teraction), performing calibration and adjustment with ROC
vious treatments. curves and the Hosmer-Lemeshow test.
131.211.208.19 - 8/13/2017 3:12:13 PM

148 Int Arch Allergy Immunol 2017;173:147–152 Gil/Amezqueta/Martinez/Aznal/Etayo/

DOI: 10.1159/000477725 Durá/Sánchez-Valverde
University Library Utrecht
Downloaded by:
 Table 2. Perinatal variables for the CMA and control groups

CMA Control ORa (CI 95%) pb

PPT 24 (11.5) 37 (18) 0.59 (0.34 – 1.03) 0.062

Prematurity 9 (4.3) 27 (13.2) 0.29 (0.13 – 0.64) 0.001
Caesarean 94 (45) 33 (15.9) 4.33 (2.73 – 6.87) <0.001
FFH 134 (73.6) 35 (17.9) 12.76 (7.79 – 20.88) <0.001
FHc 100 (42.1) 89 (47.3) 1.23 (0.84 – 1.81) 0.282
BF duration
<1 month 41 (19.4) 25 (11.8) 1.79 (1.04 – 3.07) 0.034
1 – 4 months 69 (32.7) 66 (31.3) 1.06 (0.70 – 1.60) 0.754
4 – 6 months 65 (30.8) 39 (18.5) 1.96 (1.24 – 3.09) 0.004
>6 months 30 (14.2) 57 (27.0) 0.44 (0.27 – 0.73) 0.001

Values are presented as numbers (%) unless otherwise stated. CMA, cow milk allergy; PPT, poor pregnancy
tolerance; FFH, formula feeding in hospital; FH, family history of allergy; BF, breastfeeding. a The control group
was used as reference. b Pearson χ2 test. c One or more family members with an allergic disease.

Results Table 3. Multivariate analysis model for the variable outcome

CMA
A total of 211 cases were included. The gender distri-
OR (95% CI) p
bution was 128 (60.7%) men and 83 (39.3%) women. The
age at diagnosis was 5.07 ± 2.67 months and the current FFHa 4.94 (2.68 – 9.08) <0.001
mean age (June 2015) was 14.41 ± 5.42 years. The 211 Caesarean 0.91 (0.36 – 2.31) 0.857
healthy controls showed no significant differences in gen- FFH × caesarean 11.62 (2.84 – 47.50) 0.001
der or age compared to the cases (Table 1). Prematurity 0.29 (0.09 – 0.92) 0.037
BF duration 0.005
Table 2 shows the univariate analysis of the different
<1 month 3.87 (1.58 – 9.50) 0.003
perinatal variables that were studied in the case and con- 1 – 4 months 1.81 (0.84 – 3.90) 0.127
trol groups. Generally, a higher proportion of poor preg- 4 – 6 months 3.54 (1.58 – 7.89) 0.002
nancy tolerance, increased prematurity, and a higher pro- >6 months Ref.
portion of BF duration >6 months were observed in the FH 0.86 (0.49 – 1.48) 0.594
control group compared to the CMA group. Conversely, Constant 0.17 <0.001
a higher proportion of caesarean deliveries, FFH, BF du- FFH, formula feeding in hospital; BF, breastfeeding; FH, fam-
ration <1 month, and BF for 4–6 months in the CMA ily history of allergy; CMA, cow milk allergy. a Isolated formula
group was evident. feeding during maternity with subsequent exclusive BF.
When performing the multivariate analysis on pa-
tients with CMA and the control group of independent
predictors of CMA, the model shown in Table 3 was ob-
tained. A strong association of CMA with FFH as a risk Discussion
factor was observed, and it also interacted with caesarean
delivery, in the presence of which the OR value was sig- Comparison of the different perinatal variables ana-
nificantly multiplied. Other factors such as prematurity lyzed between the CMA IgE+ and control groups suggests
were negatively associated. With respect to BF duration a number of interesting considerations. The most signifi-
(used BF longer than 6 months as the reference category cant is the strong association of this disease with isolated
[33]), we observed that BF duration <1 month as well as administration of FFH in women who subsequently con-
BF for 4–6 months were associated with a higher risk of tinued to exclusively breastfeed their children. Among
CMA, while there were no differences in BF from 1 to 4 the patients with CMA IgE+, 73.6% reported having con-
months. The model showed goodness of fit, with an area sumed isolated doses of formula during maternity, com-
under the curve of 0.852 and a nonsignificant Hosmer- pared to only 17.9% of the healthy control group. Similar
Lemeshow test result (p = 0.816). data have been reported in case and control studies [34],
131.211.208.19 - 8/13/2017 3:12:13 PM

Perinatal Risk Factors and CMA Int Arch Allergy Immunol 2017;173:147–152 149
DOI: 10.1159/000477725
University Library Utrecht
Downloaded by:
a series of case studies [25], and cohort studies [24, 35– In our study the duration of BF was shown to have a
36]. However, De Jong et al. [26] refuted this association significant influence on the occurrence of CMA. The du-
in a double-blind randomized trial of 1,533 infants who ration of BF was longer in the control group. We found a
were given either CMP or a placebo for 3 days and were lower risk of CMA in infants who were BF for longer than
monitored for the occurrence of atopic diseases for 2 [26] 6 months, which we used as the reference category, but in
and 5 years [27]. The authors concluded that early expo- fact over 90% of patients with CMA are diagnosed before
sure to CMP was not associated with atopic diseases. De- the age of 6 months, so we stratified the BF duration vari-
spite this, the infants randomized into the CMP group able into 4 categories. We found a higher risk of CMA in
from that study were BF exclusively for a very short pe- infants with a very brief BF duration (less than 1 month)
riod (33.4% were exclusively BF for less than 2 weeks, 48% and in those with a BF duration of 4–6 months, but no
were exclusively BF for less than 6 weeks, and more than significant differences were identified in infants with a BF
70% were exclusively BF for less than 3 months), in con- of 1–4 months. Liao et al. [40] reported a lower risk of
trast to our study in which the mean period of exclusive CMP sensitization in infants with prolonged BF, although
BF in the CMA group was more than 4 months. The fact their cutoff was 4 months. The results of a meta-analysis
that the rate of exclusive BF was so low in such a high per- [41] addressing BF as a protective factor against allergic
centage of cases may have contributed to inducing toler- diseases also suggested that prolonged BF is a protective
ance in these patients, as occurred in patients who con- factor for the development of CMA IgE+. Regardless of
sumed breastmilk mixed with formula or those who only the effects of BF, the special bimodal distribution of risk
consumed formula. Not considering the duration of the in the duration of BF points to an important effect of time
BF period may have led to underestimation of the asso- of introduction of CMP in these patients. Clinical experi-
ciation between FFH and CMA, which in our study was ence has shown us that most mothers who discontinue BF
very strong, with an OR in the multivariate analysis of do so during the first month (due to BF problems) or at
4.94 for vaginal delivery and 11.62 for caesarean delivery, 4–6 months (coinciding with the end of their maternity
as has been noted by various authors. Katz et al. [37] pub- leave). These medical and socioeconomic reasons could
lished a study of a cohort of more than 13,000 children condition the time of introduction of CMP and, at the
and found that children who consumed formula from an same time, the duration of BF. Different factors such as
earlier age were less likely to develop IgE-mediated CMA. the benefits of BF, the time of introduction of CMP, and
Possibly the key problem is isolated consumption of for- others like the maturity of the digestive and immune sys-
mula and a subsequent extended period of exclusive BF. tems could explain the effects we found to be associated
The other risk factor in our study associated with CMA with BF duration in our study.
was caesarean delivery. There is some controversy re- Prematurity also appeared to be a protective factor
garding this in the literature, and mixed results have been against CMA in the multivariate analysis. The relevant
published [19–23]. The univariate analysis of our study fact is that only 4.3% of patients with CMA were prema-
showed that a caesarean delivery is a very significant risk ture, and this figure is much lower than the proportion in
factor. However, when performing the multivariate anal- the general population (13.2% in the control group). This
ysis this association was only observed for interactions fact has already been pointed out by Zachariassen et al.
with the variable FFH, such that the OR increased sig- [29] who, when studying the incidence of allergic disease
nificantly in the group of patients born by caesarean who in a cohort of extremely premature infants, did not find a
consumed FFH, while in the group of patients born by single case of food allergy. The explanation for this is pos-
caesarean who did not consume FFH it was not shown to sibly immunological: the premature patient’s immune
be a risk factor. In the studies mentioned above, FFH was system is most likely not developed enough to generate
not considered as a variable, and perhaps this could ex- IgE-mediated reactions to allergens through the digestive
plain the contradictory results found in previous studies. tract, a fact that was already observed in a study published
It therefore seems that both early and isolated introduc- in 1975 by Rieger and Rothberg [42] in which it was found
tion of CMP with a subsequent period of prolonged BF that up until 35–36 weeks newborns exposed to CMP are
and the presumable changes in the intestinal microflora not able to generate specific IgE antibodies.
produced by a caesarean delivery [38] (in addition to the It is worth mentioning that FH was not a significant
concomitant use of antibiotherapy [39]) act together, risk factor for IgE-mediated CMA, even in the univariate
with the latter being a predisposing factor while the first analysis. Although a family history of atopy is a univer-
would be the trigger. sally accepted risk factor [6, 14, 41] in the development of
131.211.208.19 - 8/13/2017 3:12:13 PM

150 Int Arch Allergy Immunol 2017;173:147–152 Gil/Amezqueta/Martinez/Aznal/Etayo/

DOI: 10.1159/000477725 Durá/Sánchez-Valverde
University Library Utrecht
Downloaded by:
allergic diseases, in view of the current results we could the correlation was correct. Selection bias in the controls
hypothesize that the development of CMA in our context is possible, because parents with allergies could have been
has a greater association with environmental factors. more motivated to participate in this study, but the prev-
Consumption of FFH has been (and remains) a very com- alence of FH among cases and controls was similar to the
mon practice in our context, especially in caesarean de- allergy prevalence described in developed countries [43],
liveries, where the mother is separated from the child for so despite this we think that ours results can be applied to
a few hours, and it is in these situations in which patients the general population. Another limitation of this study
usually receive the first doses of formula. This, combined was the frequency of disease in our context, which meant
with subsequent more or less prolonged exclusive BF, is that an extended recruitment period was needed to reach
largely responsible for many of these children developing a sufficient sample size. Finally, some important condi-
IgE-mediated CMA. It seems that the development of tions, like the socioeconomic or educational level of the
CMA is more related to this “iatrogenic care” than to a mothers, have not been studied and could be related to
high genetic or familial risk. Thus, in our opinion isolated other variables like BF duration. A lack of these variables
doses of formula should be avoided in newborns whose could make the current results subject to residual con-
mothers wish to give exclusive BF, especially in children founding.
born by caesarean. If isolated doses are necessary for
medical reasons, extensively hydrolyzed formulas could
be a good option to avoid sensitization in these children. Conclusions
The limitations of this study come from its own meth-
odological design. This was a retrospective study of case Sporadic FFH consumption during the first days of life
and control groups, and there may have been bias with is a key risk factor in the development of CMA, especial-
regard to memory, i.e., the parents of the cases may have ly in those patients born by caesarean delivery, such that
remembered certain details about the risk factors better its consumption should be avoided. Prolonged BF has
than those of the controls. In some cases, such as for the been shown to be a protective factor against the develop-
variable type of delivery, this bias is irrelevant, but it may ment of CMA and should therefore be encouraged.
have affected this study in terms of other variables. How-
ever, as the birth of a child is a vital event, parents tend to
remember all of the details surrounding the event well. Disclosure Statement
Furthermore, the responses given in the questionnaire The authors confirm that they have no conflict of interest re-
were verified via the available medical record register and garding the content of this article.

References
1 Rona RJ, Keil T, Summers C, et al: The preva- 6 Martorell-Aragonésa L, Echeverría-Zudaireb 11 Kramer MS, Kakuma R: Maternal dietary an-
lence of food allergy: a meta-analysis. J Aller- E, Alonso-Lebreroc J, et al: Position docu- tigen avoidance during pregnancy or lacta-
gy Clin Immunol 2007;120:638–646. ment: IgE-mediated cow’s milk allergy. Aller- tion, or both, for preventing or treating atop-
2 Sicherer SH: Epidemiology of food allergy. J gol Immunopathol (Madr) 2015;43:507–526. ic disease in the child. Cochrane Database Syst
Allergy Clin Immunol 2011;127:594–602. 7 Sladkevicius E, Nagy E, Lack G, Guest JF: Re- Rev 2012;3:CD000133.
3 Liu AH, Jaramillo R, Sicherer SH, et al: Na- source implications and budget impact of 12 Lovegrove JA, Hampton SM, Morgan JB: The
tional prevalence and risk factors for food al- managing cow milk allergy in the UK. J Med immunological and long-term atopic out-
lergy and relationship to asthma: results from Econ 2010;13:119–128. come of infants born to women following a
the National Health and Nutrition Examina- 8 Sladkevicius E, Guest JF: Budget impact of milk-free diet during late pregnancy and lac-
tion Survey 2005–2006. J Allergy Clin Immu- managing cow milk allergy in the Nether- tation: a pilot study. Br J Nutr 1994; 71: 223–
nol 2010;126:798–806. lands. J Med Econ 2010;13:273–283. 238.
4 Plaza Martín AM: Alergia a proteínas de leche 9 Sladkevicius E, Guest JF: Modelling the health 13 Fiocchi A, Burks W, Bahna SL, et al: Clinical
de vaca. Protoc Diagn Ter Pediatr 2013;1:51– economic impact of managing cow milk al- use of probiotics in pediatric allergy (CUP-
61. lergy in South Africa. J Med Econ 2010; 13: PA): a world allergy organization position pa-
5 Host A, Halken S: Cow’s milk allergy: where 257–272. per. World Allergy Org J 2012;5:148–167.
have we come from and where are we going? 10 Guest JF, Nagy E: Modelling the resource im- 14 Von Berg A: Dietary interventions for prima-
Endocr Metab Immune Disord Drug Targets plications and budget impact of managing ry allergy prevention – what is the evidence?
2014;14:2–8. cow milk allergy in Australia. Curr Med Res World Rev Nutr Diet 2013;108:71–78.
Opin 2009;25:339–349.
131.211.208.19 - 8/13/2017 3:12:13 PM

Perinatal Risk Factors and CMA Int Arch Allergy Immunol 2017;173:147–152 151
DOI: 10.1159/000477725
University Library Utrecht
Downloaded by:
15 Osborn DA, Sinn J: Formulas containing hy- 25 Cantani A, Micera M: Neonatal cow milk sen- 34 Stintzing G, Zetterström R: Cow’s milk aller-
drolysed protein for prevention of allergy and sitization in 143 case-reports: role of early ex- gy, incidence and pathogenetic role of early
food intolerance in infants. Cochrane Data- posure to cow’s milk formula. Eur Rev Med exposure to cow’s milk formula. Acta Paediatr
base Syst Rev 2006, p CD003664. Pharmacol Sci 2005;9:227–230. Scand 1979;68:383–387.
16 Osborn DA, Sinn J: Soy formula for preven- 26 De Jong MH, Scharp-van der Linden VT, Aal- 35 Høst A: Importance of the first meal on the
tion of allergy and food intolerance in in- berse RC, Oosting J, Tijssen JG, de Groot CJ: development of cow’s milk allergy and intol-
fants. Cochrane Database Syst Rev 2006; 4: Randomised controlled trial of brief neonatal erance. Allergy Proc 1991;12:227–232.
CD003741. exposure to cows’ milk on the development of 36 Saarinen KM, Savilahti E: Infant feeding pat-
17 Osborn DA, Sinn JK: Probiotics in infants for atopy. Arch Dis Child 1998;79:126–130. terns affect the subsequent immunological
prevention of allergic disease and food hyper- 27 De Jong MH, Scharp-Van Der Linden VT, features in cow’s milk allergy. Clin Exp Al-
sensitivity. Cochrane Database Syst Rev 2007, Aalberse R, Heymans HS, Brunekreef B: The lergy 2000;30:400–406.
p CD006475. effect of brief neonatal exposure to cows’ milk 37 Katz Y, Rajuan N, Goldberg MR, et al: Early
18 Osborn DA, Sinn JK: Prebiotics in infants for on atopic symptoms up to age 5. Arch Dis exposure to cow’s milk protein is protective
prevention of allergy. Cochrane Database Syst Child 2002;86:365–369. against IgE-mediated cow’s milk protein al-
Rev 2013;3:CD006474. 28 Lodge CJ, Tan DJ, Lau M, et al: Breastfeeding lergy. J Allergy Clin Immunol 2010; 126: 77–
19 Eggesbø M, Botten G, Stigum H, Samuelsen and asthma and allergies: a systematic review 82.
SO, Brunekreef B, Magnus P: Cesarean deliv- and meta-analysis. Acta Paediatr 2015; 104: 38 Salminen S, Gibson GR, McCartney AL, Iso-
ery and cow milk allergy/intolerance. Allergy 38–53. lauri E: Influence of mode of delivery on gut
2005;60:1172–1173. 29 Zachariassen G, Faerk J, Esberg BH, et al: Al- microbiota composition in seven year old
20 Metsälä J, Lundqvist A, Kaila M, Gissler M, lergic diseases among very preterm infants ac- children. Gut 2004;53:1388–1389.
Klaukka T, Virtanen SM: Maternal and peri- cording to nutrition after hospital discharge. 39 Azad MB, Konya T, Persaud RR, et al: Impact
natal characteristics and the risk of cow’s milk Pediatr Allergy Immunol 2011;22:515–520. of maternal intrapartum antibiotics, method
allergy in infants up to 2 years of age: a case- 30 Byberg KK, Ogland B, Eide GE, Oymar K: of birth and breastfeeding on gut microbiota
control study nested in the Finnish popula- Birth after preeclamptic pregnancies: associa- during the first year of life: a prospective co-
tion. Am J Epidemiol 2010;171:1310–1316. tion with allergic sensitization and allergic hort study. BJOG 2016;123:983–993.
21 Toro Monjaraz EM, Ramírez Mayans JA, Cer- rhinoconjunctivitis in late childhood: a his- 40 Liao SL, Lai SH, Yeh KW, et al: Exclusive
vantes Bustamante R, et al: Perinatal factors torically matched cohort study. BMC Pediatr breastfeeding is associated with reduced
associated with the development of cow’s 2014;14:101. cow’s milk sensitization in early childhood.
milk protein allergy. Rev Gastroenterol Mex 31 Kumar R, Ouyang F, Story RE, et al: Gesta- Pediatr Allergy Immunol 2014;25:456–461.
2015;80:27–31. tional diabetes, atopic dermatitis, and aller- 41 Van Odijk J, Kull I, Borres MP, et al: Breast-
22 Sánchez-Valverde F, Gil F, Martinez D, et al: gen sensitization in early childhood. J Allergy feeding and allergic disease: a multidisci-
The impact of caesarean delivery and type of Clin Immunol 2009;124:1031–1038. plinary review of the literature (1966–-2001)
feeding on cow’s milk allergy in infants and 32 Instituto de Estadística de Navarra: Población on the mode of early feeding in infancy and its
subsequent development of allergic march in y demografía. http://www.navarra.es/Apps- impact on later atopic manifestations. Allergy
childhood. Allergy 2009;64:884–889. Ext/GN.InstitutoEstadistica.Web/informa- 2003;58:833–843.
23 Bager P, Wohlfahrt J, Westergaard T: Caesar- cionestadistica.aspx?R=1&E=1 (accessed Sep- 42 Rieger CH, Rothberg RM: Development of
ean delivery and risk of atopy and allergic dis- tember 16, 2015) the capacity to produce specific antibody to
ease: meta-analyses. Clin Exp Allergy 2008; 33 World Health Organization: Global strategy an ingested food antigen in the premature in-
38:634–642. for infant and young child feeding. 2003. fant. J Pediatr 1975;87:515–518.
24 Høst A, Husby S, Osterballe O: A prospective http://www.who.int/maternal_child_adoles- 43 Pawankar R, Canonica GW, Holgate ST,
study of cow’s milk allergy in exclusively cent/documents/9241562218/en/ (accessed Lockey RF: World Health Organization
breast-fed infants: incidence, pathogenetic February 28, 2017). White Book on Allergy: 2011–2012 Executive
role of early inadvertent exposure to cow’s Summary. Milwaukee, WAO, 2011.
milk formula, and characterization of bovine
milk protein in human milk. Acta Paediatr
Scand 1988;77:663–670.

131.211.208.19 - 8/13/2017 3:12:13 PM

152 Int Arch Allergy Immunol 2017;173:147–152 Gil/Amezqueta/Martinez/Aznal/Etayo/

DOI: 10.1159/000477725 Durá/Sánchez-Valverde
University Library Utrecht
Downloaded by: