Establishment of the intestinal microbiota and its role
for atopic dermatitis in early childhood
Citation for published version (APA):
Penders, J., Gerhold, K., Stobberingh, E. E., Thijs, C., Zimmermann, K., Lau, S., & Hamelmann, E.
(2013). Establishment of the intestinal microbiota and its role for atopic dermatitis in early childhood.
Journal of Allergy and Clinical Immunology, 132(3), 601-607. https://doi.org/10.1016/j.jaci.2013.05.043
Document status and date:
Published: 01/01/2013
DOI:
10.1016/j.jaci.2013.05.043
Document Version:
Publisher's PDF, also known as Version of record
Document license:
Taverne
Please check the document version of this publication:
• A submitted manuscript is the version of the article upon submission and before peer-review. There can
be important differences between the submitted version and the official published version of record.
People interested in the research are advised to contact the author for the final version of the publication,
or visit the DOI to the publisher's website.
• The final author version and the galley proof are versions of the publication after peer review.
• The final published version features the final layout of the paper including the volume, issue and page
numbers.
Link to publication
General rights
Copyright and moral rights for the publications made accessible in the public portal are retained by the authors and/or other copyright
owners and it is a condition of accessing publications that users recognise and abide by the legal requirements associated with these
rights.
• Users may download and print one copy of any publication from the public portal for the purpose of private study or research.
• You may not further distribute the material or use it for any profit-making activity or commercial gain
• You may freely distribute the URL identifying the publication in the public portal.
If the publication is distributed under the terms of Article 25fa of the Dutch Copyright Act, indicated by the “Taverne” license above,
please follow below link for the End User Agreement:
www.umlib.nl/taverne-license
Take down policy
If you believe that this document breaches copyright please contact us at:
repository@maastrichtuniversity.nl
providing details and we will investigate your claim.
Download date: 09 Aug. 2022
Atopic dermatitis and skin disease
Establishment of the intestinal microbiota and its role for
atopic dermatitis in early childhood
John Penders, PhD,
a,b
Kerstin Gerhold, MD, PhD,
c
Ellen E. Stobberingh, PhD,
a
Carel Thijs, MD, PhD,
b
Kurt Zimmermann, PhD,
d
Susanne Lau, MD, PhD,
c
and Eckard Hamelmann, MD, PhD
e
Maastricht, The Netherlands, and
Berlin, Herborn, and Bochum, Germany
Background: Perturbations in the intestinal microbiota may
disrupt mechanisms involved in the development of
immunologic tolerance. The present study aimed to examine
the establishment of the infant microbiota and its association
to the development of atopic dermatitis (AD).
Methods: Within a randomized, placebo-controlled trial on the
prevention of AD by oral supplementation of a bacterial lysate
between week 5 and the end of month 7, feces was collected at
the ages of 5 weeks (n 5 571), 13 weeks (n 5 332), and 31 weeks
(n 5 499) and subjected to quantitative PCRs to detect
bifidobacteria, bacteroides, lactobacilli, Escherichia coli,
Clostridium difficile, and Clostridium cluster I.
Results: Birth mode, breast-feeding but also birth order had a
strong effect on the microbiota composition. With increasing
number of older siblings the colonization rates at age 5 weeks of
lactobacilli (P < .001) and bacteroides (P 5 .02) increased,
whereas rates of clostridia decreased (P <.001). Colonization with
clostridia, at the age of 5 and 13 weeks was also associated with an
increased risk of developing AD in the subsequent 6 months of life
(odds ratio
adjusted
5 2.35; 95% CI, 1.36-3.94 and 2.51; 1.30-4.86,
respectively). Mediation analyses demonstrated that there was a
statistically significant indirect effect via Clostridium cluster I
colonization for both birth mode and birth order in association
to AD.
Conclusion: The results of this study are supportive for a role of
the microbiota in the development of AD. Moreover, the
‘‘beneficial’’ influence of older siblings on the microbiota
composition suggests that this microbiota may be one of the
biological mechanisms underlying the sibling effect. (J Allergy
Clin Immunol 2013;132:601-7.)
Key words: Microbiota, atopic dermatitis, birth mode, siblings,
mediation analysis
The intestinal microbiota is a key source of immune develop-
ment and regulation early in life. Deprivation of microbial
exposure is thought to predispose to immune dysregulation and
the development of atopic diseases.
1
Animal studies have found
that oral tolerance is difficult to achieve in germ-free animals
2
and that administration of lipopolysaccharides (constituents of
the outer membrane of gram-negative bacteria) together with
food antigens increases the tolerizing effect of feeding.
3
In
addition, a complex intestinal microbiota, rather than coloniza-
tion with a single microorganism, seems to be required to support
oral tolerance development.
4
Numerous epidemiologic studies showed indeed that the
microbiota of infants with allergies differs from the microbiota
of infants without allergies.
5
Although most of these studies were
case–control studies, some, but not all, of the longitudinal studies
found that these differences in the composition and diversity
of the microbiota actually preceded the development of
allergic manifestations.
5-7
Thus, the immune modulation by
gastrointestinal (GI) microbiota is still one of the key candidates
that may explain the increase of allergies (and other immune
disorders) in terms of the hygiene hypothesis.
The fetal intestine is sterile and bathed in swallowed amniotic
fluid. After delivery, the colonization of the intestines by a variety
of microorganisms begins.
8
Intestinal colonization involves a
succession of bacterial populations waxing and waning as the
diet changes and the host develops.
9
Factors that influence the intestinal microbiota composition
can be divided into host factors (such as pH, bile acids, pancreatic
enzymes, mucus composition, and transit time), nonhost factors
(such as diet, medication, and environmental factors), and bac-
terial factors (such as adhesion capacity, enzymes, and metabolic
capacities).
10
Especially changes in nonhost factors due to
Western lifestyle (antibiotic use, diet, smaller family sizes,
increased hygiene) may result in perturbations in the GI micro-
biota composition and thus may interfere with the mechanisms
involved in the development of immunologic tolerance.
11
In the present study, we investigated the influence of nonhost
factors on the establishment of the intestinal microbiota in
From the
a
Department of Medical Microbiology, School for Nutrition, Toxicology and
Metabolism, Maastricht University Medical Centre, Maastricht;
b
the Department of
Epidemiology, School for Public Health and Primary Care, Maastricht University,
Maastricht;
c
the Department of Pediatric Pneumology and Immunology, Charit
e–
Universit
€
atsmedizin Berlin, Berlin;
d
SymbioPharm, Herborn; and
e
the University
Children’s Hospital, Ruhr-Universit
€
at Bochum, Bochum.
Supported by the German Research Foundation DFG (DFG HA 2162/4-1 to E.H.) and
SymbioPharm Herborn.
Disclosure of potential conflict of interest: K. Gerhold has received grants from the
German Research Foundation (DFG); is employed by the biologics register RABBIT
which is supported by joint, unconditional grants from AbbVie, Bristol-Myers Squibb,
MSD Sharp & Dohme, Pfizer, Roche, and UCB; has received payment for lectures, in-
cluding service on speakers bureaus from BMS; has received payment for a case report
and review article manuscript preparation for ‘‘Arthritis und Rheuma,’’ schattaur Ver-
lag. S. Lau has received grants and consulting fees or honoraria from Symbiopharm, is
a board member for Merck DMC, has consultant arrangements and grants pending
with Allergopharma, has received payment for lectures, including service on speakers’
bureaus from Nestl
e. E. Hamelmann has received grants from DFG. The rest of the au-
thors declare that they have no relevant conflicts of interest.
Received for publication December 9, 2012; revised May 29, 2013; accepted for publi-
cation May 31, 2013.
Available online July 27, 2013.
Corresponding author: John Penders, PhD, Departments of Epidemiology and
Medical Microbiology, Maastricht University, PO Box 616, 6200 MD Maastricht,
The Netherlands. E-mail: j.penders@maa strichtuniversity.nl.
0091-6749/$36.00
Ó 2013 American Academy of Allergy, Asthma & Immunology
http://dx.doi.org/10.1016/j.jaci.2013.05.043
601
Abbreviations used
AD: Atopic dermatitis
C-section: Cesarean section
GI: Gastrointestinal
OR: Odds ratio
CFU: Colony-forming units
infancy, within a randomized, placebo-controlled trial of primary
prevention of atopic dermatitis (AD) by oral supplementation
of a bacterial lysate in very early infancy. Furthermore, we
prospectively examined the composition of the infant intestinal
microbiota in association to the subsequent development of AD
and sensitization to common food allergens.
METHODS
Study population
The present study was conducted within the context of a randomized,
placebo-controlled trial (registration no. ISRCTN60475069) on the primary
prevention of AD by an orally applied bacterial lysate that contained
heat-killed Escherichia coli Symbio DSM 17252 and Enterococcus faecalis
Symbio DSM 16440 (Pro-Symbioflor). The study was approved by the Charit
e
Ethics Committee in 2002, and all parents gave informed consent. The design
of this trial has been described in detail elsewhere.
12
Briefly, 606 healthy newborns (at term and birth weight
>_
2500 g) with a
single or double heredity for atopy (AD, allergic rhinitis, and/or asthma) were
included in the study. Exclusion criteria were antibiotic treatment or other
medication directly after birth, lymphocytopenia or thrombocytopenia, inten-
sive care after birth, or parents lacking knowledge of the German language.
After an initial screening phase (age birth to 4 weeks), enrolled infants were
randomly assigned at 4 to 5 weeks of age. From week 5 until the end of week
31 postpartum, infants were orally supplemented with the bacterial lysate or
placebo daily.
Parents were asked to sample the infant’s feces at the age of 5 weeks (start
of intervention; n 5 571), at 13 weeks (in a random subgroup only; n 5 332),
and at 31 weeks (end of intervention period; n 5 499). Participants were
provided with standard stool tubes with spoons attached to the lid (Sarstedt,
Hilden, Germany) and were instructed to collect the fecal sample before the
next visit during which times samples were handed to the researchers.
During the intervention period and thereafter until the age of 3 years,
children were clinically examined at a regular basis by a pediatrician for signs
of AD.
DNA purification from feces
At the laboratory 1 spatula of feces (approximately 200 mg) was diluted in
2 mL of Crowser-Medium (5 g of Lab Lemco [meat extract 3.0 g/L and Pepton
5 g/L] 1 50 mL of Gycerol and 450 H
2
O; ;pH 7.3) and stored at 2808C until
further analysis.
For DNA isolation, 0.2 mL of the diluted feces was added to a 2-mL vial
that contained approximately 300 mg of glass beads (diameter, 0.1 mm) and
1.4 mL of ASL buffer from the QIAamp DNA stool minikit (Qiagen, Hilden,
Germany), and the samples were disrupted in a mechanical bead beater at 5000
rpm for 3 minutes. Subsequently, the bacterial DNA was isolated from the
samples with the QIAamp DNA stool mini kit, according to the instructions
provided by the manufacturer. The DNA was eluted in a final volume of
200 mL. DNA yields (ng/mL) were measured with an Eppendorf Photometer.
Microbial analysis of fecal samples
DNA from the fecal samples was subjected to quantitative real-time PCR
assays for the quantification of bifidobacteria, E coli, Clostridium difficile,
Clostridium cluster I (Clostridium sensu stricto), Bacteroides fragilis group,
and lactobacilli targeting 16S rDNA gene sequences (see Table E1
for primer and probe sequences in this article’s Online Repository at
www.jacionline.org) as described previously.
13
Counts of the bacterial groups and species were calculated for each stool
sample from the threshold cycle values by using constructed standard curves
and were expressed as the log
10
colony-forming units (CFU) per milliliter of
diluted feces. The prevalence of colonization was expressed as the percentage
of infants colonized with a specific bacterial group or species.
Diagnosis of AD
Infants were clinically examined by a pediatrician during the intervention
phase at the ages of 13, 21, and 31 weeks (end of the intervention phase). In the
follow-up phase, participants were seen for additional visits at 1, 2, and 3 years
of age. AD was clinically assessed.
Sensitization to food allergens
Sensitization to common food allergens (soy, peanut, cow’s milk, hen’s
egg, wheat, and cod fish) was tested by panel ImmunoCAP fx5 on blood
samples taken at 31 weeks, 1 year, and 2 years of age. Children who tested
positive to any of the food allergens (>0.35 IU/mL) at any time point were
labeled sensitized. Children were regarded nonsensitized when they were
tested at least at age 2 years and were found negative at this time point and
were negative at the other time points at which they were tested (31 weeks
and/or 1 year).
Statistical analysis
Effects of birth characteristics, environmental
factors, and intervention on gut microbiota.
The following
potential determinants were examined in association to the GI microbiota at
the age of 5 and 13 weeks: sex (male/female), birth mode (spontaneous
vaginal, assisted vaginal [forceps/vacuum extraction], cesarean section
[C-section]), number of siblings (0, 1, 2 or more), atopy mother, or atopy
father. For GI microbiota at the age of 31 weeks, this list was complemented
with duration of breast-feeding (0-3 months, 3-6 months, or >6 months) and
day care attendance (group size
>_
3 children) during the first 6 months of life.
The Mann–Whitney rank sum test was used for the associations between
these determinants and the counts of the bacteria under study (including the
noncolonized infants with counts defined as zeros). The same method was
used to examine the influence of the intervention on the bacterial counts
in those infants who completed follow-up until the end of the treatment
(age 31 weeks).
GI microbiota composition in association with AD
and sensitization.
Logistic regression analyses were used to test
for associations between colonization with gut bacteria (colonized or
noncolonized) under study and the development of AD or sensitization to
food allergens respectively.
The following covariates were taken into account in the logistic regression
models: sex, birth weight, maternal and paternal atopy, (duration of) breast-
feeding, number of siblings, mode of delivery, and treatment group (placebo vs
active group).
Logistic regression analyses were also used for associations between the
concentrations (counts) of the gut bacteria and AD. Here, we additionally
adjusted for the DNA concentrations of the samples to normalize the data. To
test for trend bacterial counts were categorized (noncolonized infants were
used as a reference category, and the remaining colonized infants were
accommodated in 3 equal groups). These analyses were all limited to the
completers group for the specific end points.
Survival analysis by Cox regression was used to examine the effects of the
GI microbial composition on AD-free survival time.
Follow-up time for subjects who developed AD was calculated as the
number of days between birth and the date of the visit at which AD was first
diagnosed. The follow-up time of children who had not developed AD (yet)
until they were lost to follow-up during the follow-up was the age in days at the
moment of the last performed study visits.
To check the potential effect-modifying role of the treatment groups
(placebo vs active group), we initially incorporated interaction terms between
J ALLERGY CLIN IMMUNOL
SEPTEMBER 2013
602 PENDERS ET AL
the variable ‘‘treatment group’’ and the variables for the different bacteria
under study in all statistical models. Because none of these interaction terms
appeared statistically significant, they were removed from the models, and
associations are reported for the entire study population without stratification
for treatment group.
Mediation analyses. To investigate whether Clostridium cluster I
mediated the associations between birth mode, respectively, birth order
and AD, we used the ab product-coefficient method.
14
This entails estimat-
ing the product of 2 coefficients: that of the association between birth mode/
siblings and Clostridium cluster I (the a path) and that of the association
between Clostridium cluster I and AD (the b path). Standardized coefficients
and standard errors were obtained from these analyses. To test for statistical
significance of the ab-product coefficient, the Sobel test was used.
RESULTS
At the start of the intervention there were neither differences
between the intervention groups for baseline characteristics nor
differences between baseline characteristics of the entire study
population (n 5 303 in both the active and placebo groups) and
the study population that was included for the present study
(those children of whom fecal samples were collected at baseline;
n 5 285 and n 5 286 in the active and placebo group,
respectively; Table I).
Effects of birth characteristics, environmental
factors, and intervention on gut microbiota
A strong association between birth by C-section and the
GI microbiota composition was found: infants delivered by
C-section were less often colonized by bifidobacteria, bacteroi-
des, and E coli, but more frequently colonized by both
Clostridium cluster I and C difficile. If colonized, infants
delivered by C-section had also lower counts (CFU/mL diluted
feces) of bifidobacteria and bacteroides and a higher count of
clostridia than infants delivered spontaneously (Table II). At the
age of 13 weeks and even at the age of 31 weeks the effects of
C-section were still prominent, with a reduced prevalence of
TABLE I. Baseline characteristics of both treatment groups in the entire study population and the present study population
Baseline characteristics
Total study population Present study population*
Active (n 5 303) Placebo (n 5 303) Active (n 5 285) Placebo (n 5 286)
Age of newborns (wk), median (25%-75% quartile) 5.1 (4.6-5.7) 5.1 (4.6-5.7) 5.1 (4.6-5.7) 5.1 (4.6-5.7)
Proportion males, no. (%) 161 (53.1) 152 (50.2) 155 (54.4) 140 (49.0)
Weight at birth (g), median (25%-75% quartile) 3480 (3140-3780) 3480 (3230-3800) 3480 (3140-3790) 3495 (3230-3800)
Gestational age (wk), median (25%-75% quartile) 40 (39-40) 40 (38-40) 40 (39-40) 40 (38-40)
Cesarean section, no. (%) 75 (24.8) 76 (25.1) 70 (24.6) 74 (26.0)
Breast-fed
Never, no. (%) 9 (3.0) 6 (2.0) 7 (2.5) 5 (1.7)
<3 mo, no. (%) 59 (19.5) 41 (13.5) 57 (20.0) 36 (12.6)
3-6 mo, no. (%) 38 (12.5) 38 (12.5) 34 (11.9) 36 (12.6)
6-9 mo, no. (%) 71 (23.4) 81 (26.7) 67 (23.5) 78 (27.3)
9-12 mo, no. (%) 63 (20.8) 60 (19.8) 60 (21.1) 58 (20.3)
>12 mo, no. (%) 63 (20.8) 77 (25.4) 60 (21.1) 73 (25.5)
No. of siblings, median (25%-75% quartile) 1 (1-1) 1 (1-1) 1 (1-1) 1 (1-1)
Frequency of siblings
1 sibling, no. 122 110 110 99
2 siblings, no. 22 28 21 22
3 siblings, no. 5 3 5 3
4 siblings, no. — — — —
5 siblings, no. — 1 — 1
Smoking mother
Before pregnancy, no. (%) 81 (26.7) 73 (24.1) 77 (27.0) 67 (23.4)
During pregnancy, no (%) 75 (24.8) 66 (21.8) 71 (24.9) 62 (21.7)
After pregnancy, no. (%) 71 (23.4) 74 (24.4) 69 (24.2) 70 (24.5)
Family history of atopy
Both parents, no. (%) 148 (49.2) 157 (52.5) 137 (48.1) 148 (51.7)
One of both parents, no. (%) 154 (50.5) 145 (47.2) 147 (51.6) 137 (47.9)
Mother, no. (%) 79 (25.7) 85 (27.4) 76 (26.7) 79 (27.6)
Father, no. (%) 75 (24.8) 60 (19.8) 71 (24.9) 58 (20.3)
Single mother 2 (0.6) 1 (0.3)à 1 (0.4)à 1 (0.3)à
Underlying parental disease§
Mother
Atopic eczema 117 (38.7) 109 (36.1) 110 (38.6) 105 (36.7)
Allergic rhinitis 171 (56.6) 194 (64.2) 163 (57.2) 181 (63.3)
Allergic asthma 89 (29.5) 99 (32.8) 83 (29.1) 92 (32.2)
Father
Atopic eczema 55 (18.2) 65 (21.5) 52 (18.4) 64 (22.5)
Allergic rhinitis 204 (67.5) 189 (62.6) 189 (66.5) 179 (62.8)
Allergic asthma 70 (23.2) 67 (22.2) 63 (22.1) 64 (22.5)
*All children with fecal samples collected at age 5 weeks.
Family history of atopy was unknown for 1 father but known for the second father.
àFamily history of atopy was unknown for the father.
§Underlying diseases in parents ranged between 1 and 3.
J ALLERGY CLIN IMMUNOL
VOLUME 132, NUMBER 3
PENDERS ET AL 603
colonization by bacteroides and a slightly higher prevalence
of colonization by Clostridium cluster I. Furthermore, at 31 weeks
of age approximately one-half of the children delivered spontane-
ously were colonized by lactobacilli, whereas this number was
significantly lower (37.5%) in children delivered by C-section.
Next to the mode of delivery, the number of older siblings
showed a strong association with the establishment of the GI
microbiota. With increasing number of siblings the colonization
rate of clostridia decreased (P
for trend
< .001) and lactobacilli
(P
for trend
< .001) and bacteroides (P
for trend
5 .02) at the age of
TABLE II. Median counts and prevalence of colonization with selected gut bacteria in feces of infants at age 5 (n 5 571), 13, and 31
weeks (n 5 499)
No.
Bifidobacteria,
counts* (%)
Clostridium cluster I,
counts* (%)
C difficile,
counts* (%)
Lactobacilli,
counts* (%)
B fragilis group,
counts* (%)
E coli,
counts* (%)
Age 5 wk
Birth weight
<3000 g 78 8.46 (80.8) 5.58 (50.0) 7.47 (25.6) 6.15 (19.2) 9.66 (41.0) 8.72 (55.1)
3000-4000 gà 421 8.68 (90.7) 5.65 (42.5) 6.98 (24.0) 6.07 (19.7) 9.40 (57.0) 8.46 (61.5)
>_
4000 g 72 8.82 (90.3) 5.44 (49.3) 5.06 (19.4) 6.28 (26.4) 9.44 (70.8) 8.50 (75.0)
Delivery
Spontaneousà 391 8.79 (90.5) 5.54 (37.2) 6.55 (19.4) 6.03 (22.0) 9.48 (65.0) 8.45 (67.3)
Assisted vaginal 34 8.38 (97.1) 6.28 (38.2) 6.88 (41.2)§ 6.89 (8.8) 9.80 (79.4) 8.69 (64.7)
C-section 144 8.33 (84.0)§ 5.62 (65.3)§ 7.48 (31.3)§ 6.28 (19.4) 7.01 (29.2)§ 8.63 (48.6)§
No. of siblings
0à 310 8.57 (85.8) 5.72 (52.3) 6.84 (26.1) 6.15 (15.8) 9.42 (53.9) 8.60 (59.0)
1 209 8.76 (95.2)§ 5.44 (37.0)§ 7.01 (22.0) 6.03 (24.9) 9.44 (56.0) 8.17 (65.6)
>_
2 52 8.80 (86.5) 5.41 (26.9)§ 5.18 (15.4) 7.11 (30.8)§ 9.34 (75.0) 8.48 (69.2)
P
for trend
.001 .046 .001 .02
Age 13 wk
Birth weight
<3000 g 44 8.91 (86.4) 6.19 (65.9) 7.18 (36.4) 7.11 (25.0) 10.21 (50.0) 9.00 (68.2)
3000-4000 gà 252 9.03 (89.7) 5.92 (51.0) 6.93 (20.9) 6.93 (26.5) 9.85 (54.5) 8.99 (76.6)
>_
4000 g 35 9.20 (94.3) 5.52 (34.3) 6.64 (20.0) 7.42 (37.1) 9.80 (62.9) 9.05 (82.9)
P
for trend
.001
Delivery
Spontaneousà 232 9.02 (90.1) 5.73 (46.8) 6.88 (19.3) 7.00 (26.2) 9.85 (60.1) 9.00 (80. 2)
Assisted vaginal 17 9.42 (88.2) 6.11 (52.9) 7.04 (29.4) 7.06 (35.3) 10.39 (64.7) 9.04 (64.7)
C-section 83 8.91 (89.2) 6.19 (63.9)§ 7.17 (32.5) 6.88 (28.9) 10.05 (37.3)§ 8.98 (67.5)
No. of siblings
0à 198 8.99 (88.4) 6.03 (58.8) 7.04 (23.6) 7.04 (21.6) 9.87 (52.3) 9.03 (73.7)
1 102 9.03 (92.2) 5.75 (44.1)§ 6.89 (26.5) 6.91 (36.3) 9.89 (54.9) 8.97 (81.4)
>_
2 32 9.32 (90.6) 5.93 (28.1)§ 7.99 (9.4) 7.00 (34.4) 9.75 (68.8) 9.02 (75.0)
P
for trend
.001
Age 31 wk
Birth weight
<3000 g 66 8.91 (90.9) 5.33 (78.8) 7.31 (53.0) 6.31 (43.9) 10.38 (56.1) 8.84 (90.9)
3000-4000 g 371 9.04 (94.6) 5.51 (74.1) 6.99 (39.1) 6.31 (46.1) 10.07 (68.7) 8.74 (88.9)
>_
4000 g 62 9.22 (92.1) 5.36 (65.1) 6.03 (39.7) 6.35 (52.4) 10.04 (77.8) 8.77 (92.1)
Delivery
Spontaneousà 346 9.11 (93.6) 5.41 (72.5) 6.83 (39.9) 6.32 (49.1) 10.18 (74.0) 8.82 (91.0)
Assisted vaginal 32 9.11 (96.9) 5.16 (62.5) 7.32 (34.4) 6.27 (56.2) 10.29 (62.5) 9.09 (78.1)
C-section 119 8.87 (94.2) 5.60 (80.0) 7.23 (45.8) 6.22 (37.5) 9.80 (54.2)§ 8.71 (88.3)
Breast-feeding
0-3 moà 55 8.75 (92.7) 5.42 (83.6) 7.03 (78.2) 6.00 (40.0) 10.31 (83.6) 8.59 (100.0)
3-6 mo 39 8.91 (92.3) 5.60 (82.1) 6.97 (64.1) 6.09 (35.9) 10.22 (84.6) 8.53 (97.4)
>_
6 mo 405 9.11 (94.1)§ 5.42 (71.6) 6.75 (33.6)§ 6.38 (48.6) 10.04 (64.4)§ 8.87 (87.4)
No. of siblings
0à 276 8.99 (94.9) 5.46 (74.4) 6.98 (40.4) 6.24 (40.4) 10.01 (61.7) 8.71 (87.7)
1 178 9.05 (93.3) 5.60 (72.5) 6.96 (43.3) 6.44 (51.7) 10.19 (73.0)§ 8.90 (91.6)
>_
2 45 9.42 (88.9) 5.07 (73.3) 7.33 (35.6) 6.30 (64.4)§ 10.18 (88.9)§ 8.92 (93.3)
P
for trend
.001 .001
*Counts expressed as median (log
10
CFU/mL feces). Counts were calculated from positive samples only.
P < .05, as determined with the Mann–Whitney rank-sum test, calculated from all samples (the statistical significance refers to an overall difference incorporating counts,
including noncolonized infants with counts being zero).
àReference category.
§P < .01, as determined with the Mann–Whitney rank-sum test, calculated from all samples (the statistical significance refers to an overall difference incorporating counts,
including noncolonized infants with counts being zero).
J ALLERGY CLIN IMMUNOL
SEPTEMBER 2013
604 PENDERS ET AL
