![Orthogonal projection to latent structures loading plot showing associations between the number of household cats and dogs when the child was 6 months old (Y variable), and a set of 15 X variables
The outcome variables for lung function (forced expiratory volume in 1 s [FEV1]/forced vital capacity [FVC]), bronchial hyperresponsiveness (BHR), blood eosinophil count (B-Eos), percentage of blood eosinophils (Eosproc), total immunoglobulin (IgE), and skin-prick tests (SPTs) were from the age 8–9 years follow-up. SPTs are given as weal diameter. X variable bars pointing in the same direction as the Y variable are positively associated with the Y variable, and bars pointing in the opposite direction are negatively associated. The height of the bars shows the B-coefficients for scaled and centered data, with 95% confidence intervals.](https://www.researchgate.net/publication/329789548/figure/fig3/AS:11431281272047619@1723855317796/Orthogonal-projection-to-latent-structures-loading-plot-showing-associations-between-the_Q320.jpg)
Access to this full-text is provided by PLOS.
Content available from PLOS One
This content is subject to copyright.
RESEARCH ARTICLE
Pet-keeping in early life reduces the risk of
allergy in a dose-dependent fashion
Bill HesselmarID
1
*, Anna Hicke-Roberts
1☯
, Anna-Carin Lundell
2☯
, Ingegerd Adlerberth
3☯
,
Anna Rudin
2☯
, Robert Saalman
1☯
, Go
¨ran Wennergren
1☯
, Agnes E. Wold
3☯
1Department of Paediatrics, Institute of Clinical Sciences, Sahlgrenska Academy, University of Gothenburg,
Gothenburg, Sweden, 2Department of Rheumatology and Inflammation Research, Institute of Medicine,
Sahlgrenska Academy, University of Gothenburg, Gothenburg, Sweden, 3Department of Infectious
Diseases, Institute of Biomedicine, Sahlgrenska Academy, University of Gothenburg, Gothenburg, Sweden
☯These authors contributed equally to this work.
*bill.hesselmar@vgregion.se
Abstract
Objectives
Several studies have indicated that early pet keeping could protect the infant from later
allergy development. Here, we investigate if there is a dose-dependent association between
cat- and dog-keeping during the first year of life and subsequent allergy development.
Methods
Two cohorts were investigated: a cross-sectional questionnaire-based study of 7- to 8-year-
old children (N = 1029) from Mo
¨lndal and Kiruna, and a birth-cohort of children from the
Va
¨stra Go
¨taland county clinically evaluated for asthma and allergy by paediatricians up to
the age of 8–9 years (N = 249). The cross-sectional study asked validated questions on
asthma and allergy that had been used in two previous studies of children from the
same areas. In the birth-cohort study, a diagnosis of asthma and allergy was based on pre-
defined clinical criteria, and laboratory evaluation included blood eosinophils, skin-prick
tests and specific immunoglobulin E analyses. Information on pets during first year of life
was collected retrospectively in the Cross-Sectional Cohort and prospectively in the Birth
Cohort.
Results
A dose-response association was seen, with less allergic manifestations (any of asthma,
allergic rhinoconjunctivitis, or eczema) with increasing number of household cats and dogs
during the first year of life. In the Cross-Sectional Cohort, allergy ever decreased from 49%
in those with no pets to zero in those with five or more pets (P-value for trend 0.038), and
from 32% to zero for allergy last year (P-value for trend 0.006). The same pattern was seen
in Birth Cohort. Sensitization to animals, as well as pollens, also decreased with increasing
number of animals in the household.
PLOS ONE | https://doi.org/10.1371/journal.pone.0208472 December 19, 2018 1 / 13
a1111111111
a1111111111
a1111111111
a1111111111
a1111111111
OPEN ACCESS
Citation: Hesselmar B, Hicke-Roberts A, Lundell A-
C, Adlerberth I, Rudin A, Saalman R, et al. (2018)
Pet-keeping in early life reduces the risk of allergy
in a dose-dependent fashion. PLoS ONE 13(12):
e0208472. https://doi.org/10.1371/journal.
pone.0208472
Editor: Lucienne Chatenoud, Universite
´Paris
Descartes, FRANCE
Received: September 5, 2018
Accepted: November 16, 2018
Published: December 19, 2018
Copyright: ©2018 Hesselmar et al. This is an open
access article distributed under the terms of the
Creative Commons Attribution License, which
permits unrestricted use, distribution, and
reproduction in any medium, provided the original
author and source are credited.
Data Availability Statement: All relevant data are
within the manuscript and its Supporting
Information files.
Funding: The studies were funded by the
Sahlgrenska Academy at the University of
Gothenburg; the Swedish Asthma and Allergy
Association Research Foundation (BH); the
Swedish Research Council; the Vårdal Foundation;
the European Commission (QLK4-2000-00538);
the Torsten and Ragnar So¨derberg Foundation;
Gothenburg Medical Society; the Cancer and
Conclusion
The prevalence of allergic disease in children aged 7–9 years is reduced in a dose-depen-
dent fashion with the number of household pets living with the child during their first year of
life, suggesting a “mini-farm” effect, whereby cats and dogs protect against allergy
development.
Introduction
The clinical consequences of exposure to different allergens in early life have long been a mat-
ter of discussion, especially if infants are exposed to pets such as cats and dogs during their
first year of life. Early pet-keeping was previously considered to be a risk factor for allergy
development, but several studies from the last 20 years have highlighted that this is probably
not the case [1–8], even in individuals with a strong family history of atopy [9]. Today, early
pet-keeping is generally not considered to be a risk factor for allergy in families with otherwise
healthy infants.
Conversely, pet-keeping during early life may instead protect from later allergy [1], espe-
cially exposure to more than one dog or to both a cat and a dog [3,4]. We were the first to
demonstrate, in 1999 [1], that children in families keeping (a) cat(s) or (a) dog(s) during the
child´s first year of life had less asthma at 7–9 years as compared to children with no such ani-
mals, and that this difference remained also after adjusting for selection mechanisms due to
allergy among parents or siblings. The existence of an allergy-protective effect from pet-keep-
ing is also supported by immunological data. In studies analysing the effect of cat exposure on
asthma and allergy development, a high-dose exposure to cat allergens [10], or keeping of cats
[11], were associated with clinical tolerance and cat-specific IgG4, but not IgE.
Immunological tolerance facilitated by keeping of cats and dogs during early life is, how-
ever, still a hypothesis, despite some support for this assumption in the aforementioned stud-
ies. Not all studies report a long-term protective effect [7], and if such an effect exists, it is still
not known how induction of this immunological tolerance is mediated. In principle, we
hypothesized that two different mechanisms–not mutually exclusive–could contribute to a
protective effect of pet-keeping. First, exposure to cat or dog dander, containing massive
amounts of allergens from the respective species, could induce high-dose clinical tolerance to
the allergens, i.e. reduced risk of cat-allergy in the children exposed to cats and dog-allergy in
children with dogs. Second, cohabiting pet animals could provide a “mini-farm” environment,
with microbes or other immunoregulatory factors that provide a broad modifying effect on
immune development in the child, leading to tolerance not only to the pet itself, but also to
food and airborne allergens. In this study we try to address this question, hypothesising that
high-dose allergen exposure should induce tolerance only to that specific type of animal,
whereas a mini-farm induced tolerance is supposed to be protective not only to a specific ani-
mal but also to other environmental allergens.
Most often research focus on identifying risk factors for allergy development. But in mod-
ern society, finding lifestyle factors that could protect from allergy has become equally impor-
tant. The main aim of this study was to investigate if pet-keeping during early life affects later
allergy development and, if so, whether a dose-response association was detectable. Second, if
the protective effect was species-specific, suggesting an allergen-driven tolerance induction, or,
if it is species-unspecific suggesting an allergy-protective “mini-farm” environment. We used
Pet-keeping in infancy reduces allergy
PLOS ONE | https://doi.org/10.1371/journal.pone.0208472 December 19, 2018 2 / 13
Allergy Foundation; Swedish Research Council for
Environmental, Agricultural Sciences and Spatial
Planning; the Ekhaga Foundation; Food and Health
Concept Centre, Gothenburg, West Gothia Region;
The study was also financed by grants from the
Swedish state under the agreement between the
Swedish government and the county councils, the
ALF-agreement (AR, AW, GW, IA). No one from the
funding sources was involved in the design,
execution, or analysis of the study. Anna Rudin
reports that part of her salary for her university full
professor position at The Sahlgrenska Academy at
the University of Gothenburg is covered by grant
from AstraZeneca IMed RIA (Respiratory
Inflammation, Autoimmunity) in compensation for
advice regarding basic research in inflammation at
the company.
Competing interests: The authors have declared
that no competing interests exist.
data from a cross-sectional cohort and a birth cohort for the analyses to minimize influences
from common methodological shortcomings, e.g. selection bias and reverse causation.
Methods
The analysis was based on two study populations. A cross-sectional questionnaire study was
performed in 2007 in 7–8 year old children (Cross-Sectional Cohort, N = 1029). The other
study population was the Birth Cohort, recruited between 1998 and 2007.
Cross-Sectional Cohort
In the cross-sectional cohort, a questionnaire on asthma and allergy was distributed to all 7- to
8-year-old children in Mo¨lndal, a small town which is part of the Gothenburg urban area on
the South West Sweden, and Kiruna, a town in the far north of Sweden. Of 1838 question-
naires distributed, 1029 (56%) were returned. We used the same questions on asthma, eczema,
and allergic rhinoconjunctivitis (ARC) as had been used in two previous studies of children
from the same regions and of similar ages in 1979 and in 1991 [12,13]. Diagnostic criteria and
information gathered on pet exposure are shown in Table 1.
Birth Cohort
The Birth-Cohort was pooled data from two birth-cohorts in the Va¨stra Go¨taland county.
Between 1998 and 2003, 184 children in the ALLERGYFLORA were recruited from Mo¨lndal
in the Gothenburg urban area [14]. The ALLERGYFLORA was designed to analyse the effects
of early life events and early gut colonisation on later allergy development. The second group
was the FARMFLORA. The study is a copy of the ALLERGYFLORA, but the children are liv-
ing in a rural region. Children were recruited between 2005 and 2007, from a farming area in
Skaraborg, northeast of Gothenburg, comprising 28 children living on dairy farms and 37 chil-
dren living in the same rural area but not on farms [15]. The parents were all contacted before
the birth, and children born �38 gestational weeks were included in the study on day 0–3 after
delivery. The parents were interviewed when their children were aged 6 and 12 months; clini-
cal examinations by paediatric allergologists were done at age 18 months and 3 and 8–9 years.
Diagnostic criteria and information gathered on pet exposure are detailed in Table 1.
Lung function tests
Lung function tests were done in the Birth Cohort. Before lung function tests, children were
not permitted: tea, coffee, or cola drinks within 4 hours; short-acting beta-agonists within 8
hours; ipratropium bromide within 24 hours; long-acting beta-agonists, theophylline, or nasal
steroids within 48 hours; or antihistamines within 72 hours to 1 week, depending on the type
of drug. Methacholine challenges were not performed during the pollen season in pollen-aller-
gic children; if the child had a viral infection or common cold within 14 days; if oral steroids
had been given within 14 days; if forced expiratory volume in 1 s (FEV
1
) was <65% predicted;
or if the child had a heart disorder. All lung function tests were done in a sitting position and a
nose clamp was used.
Flow-volume curves and reversibility tests were performed in accordance with American
Thoracic Society and European Respiratory Society guidelines [17] using Spida 5 spirometry
software (Micro Medical Limited, Rochester, UK). A bronchodilator response was considered
positive if FEV
1
increased by >12% from baseline [18].
Airway hyperresponsiveness was determined by direct methacholine challenge [19], using a
tidal volume-triggered dosimetric method (Spira Elektro 2 jet nebulizer; Spira Respiratory
Pet-keeping in infancy reduces allergy
PLOS ONE | https://doi.org/10.1371/journal.pone.0208472 December 19, 2018 3 / 13
Care Centre Ltd, Ha¨meenlinna, Finland). Basic FEV
1
was determined after inhalation of iso-
tonic saline. Methacholine was subsequently inhaled in increasing doses at intervals of at least
1 minute until FEV
1
had decreased by �20%, or a cumulative dose of 6.1875 mg had been
given. At the end of the challenge, all subjects received an inhalation of salbutamol and FEV
1
was measured to ensure recovery (FEV
1
>90% of baseline value). The provocative dose induc-
ing a fall of �20% in FEV
1
(PD
20
) was determined by interpolating the dose-response curve;
airway hyperresponsiveness was defined as PD
20
<0.6 mg. The slope was calculated from the
maximum fall in FEV
1
divided by the cumulative dose.
Eosinophils, specific immunoglobulin E, and Skin-prick tests
Blood tests and Skin-Prick Tests were done in the Birth Cohort. Blood eosinophil cells, specific
immunoglobulin E (IgE), and total IgE were all analyzed at the Sahlgrenska University Hospi-
tal. All analyses were accredited by the Swedish Board for Accreditation and Conformity
Assessment. For specific IgE and total IgE, Phadiatop and ImmunoCAP tests were used
(Thermo Fisher Scientific, Uppsala, Sweden). Skin-prick tests (SPTs) were carried out for com-
mon airborne allergens (cat, dog, horse, rabbit, birch, grass, mugwort, Dermatophagoides pter-
onyssinus,Dermatophagoides farinae, and Cladosporium herbarum) according to the standards
of the Subcommittee on Skin Tests of the European Academy of Allergy and Clinical Immu-
nology [20]. Allergen extracts were all manufactured by ALK (Hørsholm, Denmark). A posi-
tive SPT corresponds to a weal with a diameter exceeding the negative control by �3 mm.
Statistical analysis
Analyses were performed with SPSS statistical software (version 24; IBM Corp., Armonk, NY,
USA); for the multivariate analyses we used SIMCA-P+ software (version 14.1; MKS Umetrics
AB, Umeå, Sweden).
χ
2
tests were used to compare differences between proportions. Trend analyses were based
on linear-by-linear association and exact tests. Backward logistic regression models were used
Table 1. Diagnostic criteria and information on pet exposure in the Cross-Sectional Cohort and Birth Cohort.
Criterion Cross-Sectional Cohort Birth Cohort
Diagnosis of
asthma
“Asthma ever” diagnosed if there was a positive response to: “Has
your child had asthma or asthmatic bronchitis”?
“Current asthma” diagnosed if there was a positive response to
“Has your child had asthma or asthmatic bronchitis in the
previous year”?
Asthma diagnosed at age 8–9 years if the child in the last 12 months had
symptoms of wheeze/heavy breathing together with: FEV
1
reversibility >12%,
or bronchial hyperresponsiveness to methacholine (PD
20
<0.6 mg), or ongoing
controller medication with inhaled corticosteroids or leukotriene antagonist
Diagnosis of
ARC
ARC diagnosed if there was a positive response to: “Has your child
had allergic rhinitis or allergic conjunctivitis”?
“Current ARC” diagnosed if there was a positive response to: “Has
your child had allergic rhinitis or allergic conjunctivitis in the
previous year”?
ARC diagnosed at age 8–9 years if the child in the last 12 months had eye or
nose symptoms suggestive of allergic disease together with a positive skin-prick
test or specific IgE to the relevant allergen
Diagnosis of
eczema
“Eczema ever” diagnosed if there was a positive response to: “Did
your child ever have eczema”?
“Current eczema” diagnosed if there was a positive response to:
“Did your child have eczema in the previous year”?
Eczema diagnosed at age 8–9 years if the child in the last 12 months had a skin
condition fulfilling Williams criteria [16], or an itching dermatitis that had
been chronic or relapsing for �6 months
Diagnosis of
allergy
“Allergy ever” and “allergy last year” included any of asthma, ARC,
or eczema ever or last year, respectively
“Allergy last year” included any of asthma, ARC, or eczema
Cats and dogs in
household
Number of cats and dogs in the household during the first year of
life
Data on number of cats and dogs was obtained from the 6-month telephone
interview
ARC, allergic rhinoconjunctivitis; FEV
1
, forced expiratory volume in 1 s; Ig, immunoglobulin; PD
20
: provocative dose inducing a fall of �20% in FEV
1
.
https://doi.org/10.1371/journal.pone.0208472.t001
Pet-keeping in infancy reduces allergy
PLOS ONE | https://doi.org/10.1371/journal.pone.0208472 December 19, 2018 4 / 13
to control for covariates and possible confounders. A two-sided P-value <0.05 was considered
statistically significant.
Orthogonal projection to latent structures (OPLS), an extension of PLS-regression (Partial
Least Square regression) in order to improve interpretability, was used in the birth-cohort
study to analyse the relationship between the number of pets at 6 months of age, parental his-
tory of allergy, and 12 independent outcomes from the follow-up at 8–9 years. B coefficients
on scaled and centered data were calculated with 95% confidence intervals.
Ethics
Written informed consent was obtained from all parents. The study was approved by the Eth-
ics Committee of the University of Gothenburg, Sweden (R448-97 and O
¨446–00) and the
Human Research Ethics Committee of the Medical Faculty, University of Gothenburg, Sweden
(Dnr. 321–05, 363–05, 105–07 and 674–14).
Results
In both the Cross-Sectional Cohort and the Birth Cohort, the sex ratios were 50:50 or close to
it (Table 2). A parental history of allergy was slightly less common in the Birth Cohort, proba-
bly due to the stricter diagnostic criteria used requiring a doctor’s diagnosis of allergic disease.
In children, the prevalence of allergic disease (allergy last year) was similar in the Cross-Sec-
tional Cohort and the Birth Cohort.
In the Cross-Sectional Cohort, allergy was based on a history of asthma, ARC, or eczema
(allergy ever), or, asthma, ARC, or eczema with symptoms in the last 12 months (allergy last
year). In the Birth Cohort, allergy ever was based on a diagnosis of asthma, ARC, or eczema at
any of the follow-ups (18 months, 3 years, or 8–9 years), and allergy last year was based on
Table 2. Characteristics of the study populations.
Cross-Sectional Cohort (n = 1029) Birth Cohort (n = 249)
Boys, n (%) 483 (47) 125 (50)
History of allergy
a
, n (%)
Mother 498 (48) 110 (44)
Father 399 (39) 90 (36)
Parent with a university degree, n (%)
Mother 321 (31) –
Father 277 (28) –
Number of pets
0 767 181
1 165 40
2 64 –
�2 – 28
3 21 –
4 7 –
�5 2 –
Children with allergy, n (%)
Ever 481 (47) 95 (38)
In the last year 314 (31) 73 (29)
a
In the cross-sectional study: A history of asthma or allergic rhinoconjunctivitis ever. In the birth-cohort study, a
doctor’s diagnosis of asthma, allergic rhinoconjunctivitis, or eczema.
https://doi.org/10.1371/journal.pone.0208472.t002
Pet-keeping in infancy reduces allergy
PLOS ONE | https://doi.org/10.1371/journal.pone.0208472 December 19, 2018 5 / 13
current asthma, ARC, or eczema with symptoms in the 12-month period preceding the follow-
up at age 8–9 years.
The number of household dogs and cats during first year of life was set to range from zero
to �5 in the Cross-Sectional Cohort; in the smaller Birth Cohort, the number of pets at 6
months of age was recorded in a range from zero to �2.
Fig 1 shows the cumulative incidence (allergy ever) and prevalence (allergy last year) of
allergic disease in relation to the number of household cats and dogs during the first year of
life for the Cross-Sectional Cohort. Both allergy ever and allergy last year decreased with
increasing number of cats and dogs (P-value for trend with exact test: 0.006 for allergy ever
and 0.038 for allergy last year).
A similar pattern was seen in the Birth Cohort, with a decreasing frequency of allergic dis-
ease (both current and ever) with increasing number of household cats and dogs (Fig 2;P-
value for trend: 0.007 for allergy ever and 0.008 for allergy last year).
Backward multiple logistic regression analyses, with allergy ever as independent variable,
were used on both the Cross-Sectional and Birth Cohort. In the Cross-Sectional Cohort, inde-
pendent variables were sex, parental history of allergy, number of siblings, and number of pets
during first year of life. In the final step, only pets and parental history of allergy remained, giv-
ing an odds ratio of 0.80 for every additional animal (P= 0.012). In the Birth Cohort, the same
independent variables were included. In the final step, only pets during first year of life and
parental history of allergy remained, giving an odds ratio of 0.65 for each additional animal
(P= 0.058).
To further analyse a possible influence of parental allergic disease on the families’ choice to
have pets, parental sensitization was analysed in relation to number of household pets. In the
first 184 Birth Cohort-children from the Gothenburg-Mo¨lndal area, parents were tested for
sensitization with the Phadiatop test. Blood samples were obtained from 149 mothers and 141
fathers. There was no statistically significant difference in the frequency of positive Phadiatop
tests from parents with no household pets when their child was 6 months old versus parents
with increasing number of animals (Table 3).
In the Birth Cohort, the relationship between the number of household pets at 6 months
old and sensitisation at 8–9 years old was tested in an OPLS analysis (Fig 3), a regression
model suited to test how a large set of X-variables relate to Y-variable(s). The number of pets
was used as the Y variable (the left bar). The figure shows how the other variables (X variables)
are related to the Y variable. X-variable bars in the same direction as the Y variable bar are pos-
itively associated; bars pointing in the opposite direction to the Y-variable bar are negatively
associated. The main finding was that the degree of sensitization in children, expressed as SPT
diameter, decreased with increasing number of pets, and that this association was seen not
only for sensitization to pets but also for sensitization to pollen (birch and grass). No signifi-
cant association was found between number of pets and presence of allergy in mother or
father.
Discussion
In 1999 we published a study showing that early pet-keeping was associated with less allergy
development in children [1]. Since then, several studies have been published supporting our
finding [2–6,8], but others did not [7]. A common interpretation of published data has, to
date, been that early keeping of cats or dogs does not increase the risk of allergy but we do not
definitely know if it has any protective effect, in the same way as farm animals and farm-living
have [21–23]. However, the main findings from this extended study support our previous
results, that pet-keeping during early childhood is associated with less allergy, and that the
Pet-keeping in infancy reduces allergy
PLOS ONE | https://doi.org/10.1371/journal.pone.0208472 December 19, 2018 6 / 13
protective effect from pet-keeping increased with increasing number of animals. Furthermore,
the protective effect influenced not only clinical allergy but also sensitization to both animals
and pollen, suggesting an underlying “mini-farm” mechanism.
In this study we used results from two different study populations in order to address com-
mon questions raised when interpreting results from studies investigating allergy prevention.
The first, and most obvious question is if the findings could be due to a type-1 error, i.e. a false
positive. As we found similar results in both populations studied, we argue that our main find-
ing was not due to a type-1 error. Epidemiological cross-sectional studies have several advan-
tages as they usually include a large population, making them suitable for both univariate and
multivariate analyses, but other biases and diagnostic validity are always issues to be consid-
ered. We used the Cross-Sectional Cohort for the main analyses, and these were then repeated
in the Birth Cohort. As the results from the Cross-Sectional Cohort were reproduced in the
Birth Cohort, we argue that neither recall bias nor reverse causation explain the results. Nei-
ther was diagnostic validity a major problem in the Cross-Sectional Cohort, as the main find-
ing was repeated in the Birth Cohort in which strict diagnostic criteria were used. A similar
argument could also be used when assessing parental history of allergy. In the Cross-Sectional
Cohort, a healthy pet-owner effect might be an issue, but in the Birth Cohort the information
on parental allergy was collected when the child was just a few days old. Furthermore, parental
sensitization data from the Birth Cohort does not indicate any major difference in sensitization
pattern between parents with versus without pets. Selection bias is another issue often dis-
cussed for this type of study, as allergic parents are not supposed to own cats or dogs. One way
to handle this issue is to ask parents about their reasons for not owing pets, as we did in our
Fig 1. Data from the cross-sectional study. Allergy (any of asthma, allergic rhinoconjunctivitis, or eczema) in relation to the
number of household cats and dogs during the child’s first year of life. Allergy last year required current symptoms, i.e. symptoms in
the last 12 months.
https://doi.org/10.1371/journal.pone.0208472.g001
Pet-keeping in infancy reduces allergy
PLOS ONE | https://doi.org/10.1371/journal.pone.0208472 December 19, 2018 7 / 13
1999 study [1]. Another method is to conduct a dose-response analysis, as we have done in this
study. The rationale behind this approach is that selection may occur between families having
versus not having animals, rather than between having one or two animals versus two or three
animals. To summarise, it is our view that selection bias, recall bias, reverse causation, or
imprecise diagnostic validity do not explain our finding of an inverse correlation between the
number of household cats and dogs during a child’s first year of life and allergy prevalence.
The dose-response effect and a similar protective effect for sensitisation to animals and pol-
len, indicate that the protective effect is mediated by the keeping of animals, and is not a spe-
cies-specific effect. It is our suggestion that the allergy-protective effect mediated by pet-
keeping should be considered as a “mini-farm” effect, equating our findings to those found in
the numerous farm studies performed [24]. A “mini-farm” effect could also explain why a
Fig 2. Data from the Birth Cohort. Allergy (any of asthma, allergic rhinoconjunctivitis, or eczema) in relation to the number of
household cats and dogs when the child was 6 months old. Allergy last year required current symptoms, i.e. symptoms in the last 12
months.
https://doi.org/10.1371/journal.pone.0208472.g002
Table 3. Sensitisation in parents, measured with Phadiatop tests, in relation to the number of household cats and
dogs the family had when their child was 6 months old.
Number of cats or dogs when the child was 6 months old Positive Phadiatop test result, n/N (%)
Mother (n = 149) Father (n = 141)
0 65/127 (51) 73/121 (60)
1 9/18 (50) 7/17 (41)
�2 3/4 (75) 2/3 (67)
P-value for trend
a
0.590 0.425
a
Based on exact tests
https://doi.org/10.1371/journal.pone.0208472.t003
Pet-keeping in infancy reduces allergy
PLOS ONE | https://doi.org/10.1371/journal.pone.0208472 December 19, 2018 8 / 13
protective effect is found in some studies, but not all. The effectiveness of an unspecific allergy-
preventive (or immune-stimulating) agent should be seen in the light of other protective fac-
tors. This was demonstrated elegantly by Matricardi et al, who showed that siblings only had
an allergy-protective effect in subjects seronegative for hepatitis A and not in those who were
seropositive, i.e. those who already had a strong allergy-protective effect from hepatitis A or an
environment where hepatitis A is common [25]. A dog or a cat may thus have a protective
effect in children who have few other protective factors, provided that the child has close con-
tact with the animal during their early years. If the child already has several other protective
factors, a dog or cat may not add any extra protection, unless the child is exposed to several
animals, i.e. a “mini-farm”. Although several studies have shown that pet-keeping, mainly
from direct exposure [26], in early life is associated with less asthma or allergy [2–6,8,26],
especially if exposed to more than one animal [3,4], such an effect is not found in all studies.
In a large study with pooled data from several birth cohorts, neither a protective effect, nor an
increased allergy risk from early pet-keeping was found [7]. However, not finding a protective
effect in a study population does not necessarily mean that a protective effect does not exist
during certain circumstances, as previously mentioned. Protection seems more likely if expo-
sure occurs at close quarters, i.e. direct exposure [8], if other strong allergy-protective factors
are missing, or, as in our study, if the child is exposed to more than one animal [3,4]. Thus, it
is plausible that tolerance induction via a mini-farm mechanism require a close contact with
the animal(s), otherwise the child will only be exposed to allergens from the animal, not the
microbes and endotoxins shred by the animal, components that seems to be important in
Fig 3. Orthogonal projection to latent structures loading plot showing associations between the number of household cats and
dogs when the child was 6 months old (Y variable), and a set of 15 X variables. The outcome variables for lung function (forced
expiratory volume in 1 s [FEV
1
]/forced vital capacity [FVC]), bronchial hyperresponsiveness (BHR), blood eosinophil count (B-Eos),
percentage of blood eosinophils (Eosproc), total immunoglobulin (IgE), and skin-prick tests (SPTs) were from the age 8–9 years
follow-up. SPTs are given as weal diameter. X variable bars pointing in the same direction as the Y variable are positively associated
with the Y variable, and bars pointing in the opposite direction are negatively associated. The height of the bars shows the B-
coefficients for scaled and centered data, with 95% confidence intervals.
https://doi.org/10.1371/journal.pone.0208472.g003
Pet-keeping in infancy reduces allergy
PLOS ONE | https://doi.org/10.1371/journal.pone.0208472 December 19, 2018 9 / 13
tolerance induction. And close contact with the pet animal is probably more common in
urban areas, where families use to keep their pet animals inside the house or flat. In rural areas,
dogs and cats are more often kept outdoors. In such cases, allergens from cats and dogs will
still be spread inside the house, causing sensitisation, but not the microbes and microbial prod-
ucts that follow a close contact with the animal. Cleaning habits may also affect the effective-
ness of pet-induced tolerance induction. Allergens are seldom reduced by excessive cleaning
[27], but the mini-farm environment might be less effective.
The mechanisms behind the proposed “mini-farm” effect from dogs and cats can, of course,
only be speculated on, but according to the hygiene hypothesis [28], immune stimulation by
microbial exposure might be one possible mechanism. We have found support in various stud-
ies for allergy protection by early microbial exposure [29,30] or presumed early microbial
exposure [31], and dogs and other pet animals seem to have this capability [8].
With the study design, we have been able to show a negative association between the num-
ber of animals in the child’s home during the first year of life and allergy development, but the
study has limitations. The Cross-Sectional Cohort had a response rate of slightly less than 60%,
which may select a more allergy-prone population, even though we have not found any indica-
tions for such a selection [32]. Furthermore, recall bias and diagnostic validity may be limita-
tions in cross-sectional questionnaire studies, but these limitations are balanced by the
concordant results found in the Birth Cohort. Similarly, the smaller Birth Cohort population is
balanced by the much larger population size in the Cross-Sectional Cohort.
In conclusion, the prevalence of allergic disease in children aged 7–9 years is reduced in a
dose-response pattern with increasing number of cats and dogs in the home during the first
year of life, suggesting a “mini-farm” effect whereby pet-keeping protects against allergy
development.
Supporting information
S1 Table. Cross-Sectional Cohort.
(DOCX)
S2 Table. Birth Cohort.
(DOCX)
S3 Table. Cross-Sectional Cohort logistic regression.
(DOCX)
S4 Table. Birth Cohort logistic regression.
(DOCX)
S5 Table. Birth Cohort OPLS.
(DOCX)
S6 Table. Birth Cohort parental sensitization.
(DOCX)
Acknowledgments
The authors would like to thank school personnel and teachers in Mo¨lndal and Kiruna for
helping in the distribution and collection of questionnaires in the cross-sectional study.
In the birth-cohort study, the authors would like to thank all the children and families who
took part in the study. We also want to thank the staff at the Delivery Ward, Mo¨lndal Hospital,
the study nurses Mainor Åmark, Helen Andersson, and Anders Nordberg, and pediatricians
Pet-keeping in infancy reduces allergy
PLOS ONE | https://doi.org/10.1371/journal.pone.0208472 December 19, 2018 10 / 13
Susanne Johansen, Margareta Ceder, Gunhild Lindhagen, Stefan Stentoft, and Carl-Johan
To¨rnhage. We also thank Ms Caroline Landon for her professional and excellent work in edit-
ing and styling the manuscript.
Author Contributions
Data curation: Bill Hesselmar, Anna Hicke-Roberts, Anna-Carin Lundell, Ingegerd Adler-
berth, Anna Rudin, Robert Saalman, Agnes E. Wold.
Formal analysis: Bill Hesselmar.
Funding acquisition: Bill Hesselmar, Anna-Carin Lundell, Ingegerd Adlerberth, Anna Rudin,
Go¨ran Wennergren, Agnes E. Wold.
Investigation: Bill Hesselmar, Anna Hicke-Roberts, Anna-Carin Lundell, Ingegerd Adler-
berth, Anna Rudin, Robert Saalman, Agnes E. Wold.
Methodology: Bill Hesselmar, Anna Hicke-Roberts, Anna-Carin Lundell, Ingegerd Adler-
berth, Anna Rudin, Robert Saalman, Agnes E. Wold.
Project administration: Bill Hesselmar, Anna-Carin Lundell, Ingegerd Adlerberth, Anna
Rudin, Agnes E. Wold.
Writing – original draft: Bill Hesselmar.
Writing – review & editing: Bill Hesselmar, Anna Hicke-Roberts, Anna-Carin Lundell, Inge-
gerd Adlerberth, Anna Rudin, Robert Saalman, Go¨ran Wennergren, Agnes E. Wold.
References
1. Hesselmar B, Åberg N, Åberg B, Eriksson B, Bjo
¨rkste
´n B. Does early exposure to cat or dog protect
against later allergy development? Clinical and experimental allergy. 1999; 29:611–7. PMID: 10231320
2. Remes ST, Castro-Rodriguez JA, Holberg CJ, Martinez FD, Wright AL. Dog exposure in infancy
decreases the subsequent risk of frequent wheeze but not of atopy. The Journal of allergy and clinical
immunology. 2001; 108:509–15. https://doi.org/10.1067/mai.2001.117797 PMID: 11590373
3. Ownby DR, Johnson CC, Peterson EL. Exposure to dogs and cats in the first year of life and risk of aller-
gic sensitization at 6 to 7 years of age. JAMA. 2002; 288:963–72. PMID: 12190366
4. Mandhane PJ, Sears MR, Poulton R, Greene JM, Lou WY, Taylor DR, et al. Cats and dogs and the risk
of atopy in childhood and adulthood. The Journal of allergy and clinical immunology. 2009; 124:745–50
e4. https://doi.org/10.1016/j.jaci.2009.06.038 PMID: 19703709
5. Almqvist C, Garden F, Kemp AS, Li Q, Crisafulli D, Tovey ER, et al. Effects of early cat or dog ownership
on sensitisation and asthma in a high-risk cohort without disease-related modification of exposure. Pae-
diatr Perinat Epidemiol. 2010; 24:171–8. https://doi.org/10.1111/j.1365-3016.2010.01095.x PMID:
20415774
6. Lodge CJ, Lowe AJ, Gurrin LC, Matheson MC, Balloch A, Axelrad C, et al. Pets at birth do not increase
allergic disease in at-risk children. Clinical and experimental allergy. 2012; 42:1377–85. https://doi.org/
10.1111/j.1365-2222.2012.04032.x PMID: 22925324
7. Lødrup Carlsen KC, Roll S, Carlsen KH, Mowinckel P, Wijga AH, Brunekreef B, et al. Does pet owner-
ship in infancy lead to asthma or allergy at school age? Pooled analysis of individual participant data
from 11 European birth cohorts. PloS one. 2012; 7:e43214. https://doi.org/10.1371/journal.pone.
0043214 PMID: 22952649
8. Nermes M, Niinivirta K, Nylund L, Laitinen K, Matomaki J, Salminen S, et al. Perinatal pet exposure, fae-
cal microbiota, and wheezy bronchitis: is there a connection? ISRN Allergy. 2013; 2013:827934. https://
doi.org/10.1155/2013/827934 PMID: 23724248
9. Millqvist E, Johansson A, Mansson T, Bende M. A prospective study of allergy development in 158 chil-
dren and 128 adults with new extensive exposure to furred animals. Clinical and experimental allergy.
2007; 37:948–53. https://doi.org/10.1111/j.1365-2222.2007.02731.x PMID: 17517109
10. Platts-Mills TA, Vaughan JW, Blumenthal K, Pollart Squillace S, Sporik RB. Serum IgG and IgG4 anti-
bodies to Fel d 1 among children exposed to 20 microg Fel d 1 at home: relevance of a nonallergic
Pet-keeping in infancy reduces allergy
PLOS ONE | https://doi.org/10.1371/journal.pone.0208472 December 19, 2018 11 / 13
modified Th2 response. International archives of allergy and immunology. 2001; 124:126–9. https://doi.
org/10.1159/000053689 PMID: 11306947
11. Hesselmar B, Åberg B, Eriksson B, Bjo
¨rkste
´n B, Åberg N. High-dose exposure to cat is associated with
clinical tolerance—a modified Th2 immune response? Clinical and experimental allergy. 2003;
33:1681–5. PMID: 14656355
12. Åberg N, Hesselmar B, Åberg B, Eriksson B. Increase of asthma, allergic rhinitis and eczema in Swed-
ish schoolchildren between 1979 and 1991. Clinical and experimental allergy. 1995; 25:815–9. PMID:
8564719
13. Åberg N, Engstro
¨m I, Lindberg U. Allergic diseases in Swedish school children. Acta Paediatr Scand.
1989; 78:246–52. PMID: 2929348
14. Hesselmar B, Saalman R, Rudin A, Adlerberth I, Wold A. Early fish introduction is associated with less
eczema, but not sensitization, in infants. Acta Paediatr. 2010; 99:1861–7. https://doi.org/10.1111/j.
1651-2227.2010.01939.x PMID: 20670305
15. Stro
¨mbeck A, Rabe H, Lundell AC, Andersson K, Johansen S, Adlerberth I, et al. High proportions of
FOXP3(+) CD25(high) T cells in neonates are positively associated with allergic sensitization later in
childhood. Clinical and experimental allergy. 2014; 44:940–52. https://doi.org/10.1111/cea.12290
PMID: 24528482
16. Williams HC. Diagnostic criteria for atopic dermatitis. Lancet. 1996; 348:1391–2.
17. Miller MR, Hankinson J, Brusasco V, Burgos F, Casaburi R, Coates A, et al. Standardisation of spirome-
try. The European respiratory journal. 2005; 26:319–38. https://doi.org/10.1183/09031936.05.
00034805 PMID: 16055882
18. Pellegrino R, Viegi G, Brusasco V, Crapo RO, Burgos F, CasaburiR, et al. Interpretativestrategies for
lung function tests. The European respiratory journal. 2005; 26:948–68. https://doi.org/10.1183/
09031936.05.00035205 PMID: 16264058
19. Crapo RO, Casaburi R, Coates AL, Enright PL, Hankinson JL, Irvin CG, et al. Guidelines for methacho-
line and exercise challenge testing-1999. This official statement of the American Thoracic Society was
adopted by the ATS Board of Directors, July 1999. American journal of respiratory and critical care med-
icine. 2000; 161:309–29. https://doi.org/10.1164/ajrccm.161.1.ats11-99 PMID: 10619836
20. Skin tests used in type I allergy testing Position paper. Sub-Committee on Skin Tests of the European
Academy of Allergology and Clinical Immunology. Allergy. 1989; 44 Suppl 10:1–59.
21. Braun-Fahrlander C, Gassner M, Grize L, Neu U, Sennhauser FH, Varonier HS, et al. Prevalence of
hay fever and allergic sensitization in farmer’s children and their peers living in the same rural commu-
nity. SCARPOL team. Swiss Study on Childhood Allergy and Respiratory Symptomswith Respect to
Air Pollution. Clinical and experimental allergy. 1999; 29:28–34. PMID: 10051699
22. Riedler J, Eder W, Oberfeld G, Schreuer M. Austrian children living on a farm have less hay fever,
asthma and allergic sensitization. Clinical and experimental allergy. 2000; 30:194–200. PMID:
10651771
23. Riedler J, Braun-Fahrlander C, Eder W, Schreuer M, Waser M, Maisch S, et al. Exposure to farming in
early life and development of asthma and allergy: a cross-sectional survey. Lancet. 2001; 358:1129–33.
https://doi.org/10.1016/S0140-6736(01)06252-3 PMID: 11597666
24. Campbell BE, Lodge CJ, Lowe AJ, Burgess JA, Matheson MC, Dharmage SC. Exposure to ‘farming’
and objective markers of atopy: a systematic review and meta-analysis. Clinical and experimental
allergy. 2015; 45:744–57. https://doi.org/10.1111/cea.12429 PMID: 25270644
25. Matricardi PM, Rosmini F, Ferrigno L, Nisini R, Rapicetta M, Chionne P, et al. Cross sectional retrospec-
tive study of prevalence of atopy among Italian military students with antibodies against hepatitis A
virus. BMJ. 1997; 314:999–1003. PMID: 9112843
26. Chen CM, Morgenstern V, Bischof W, Herbarth O, Borte M, Behrendt H, et al. Dog ownership and con-
tact during childhood and later allergy development. The European respiratory journal. 2008; 31:963–
73. https://doi.org/10.1183/09031936.00092807 PMID: 18256062
27. Stemeseder T, Schweidler B, Doppler P, Klinglmayr E, Moser S, Lueftenegger L, et al. Exposure to
Indoor Allergens in Different Residential Settings and Its Influence on IgE Sensitization in a Geographi-
cally Confined Austrian Cohort. PloS one. 2017; 12:e0168686. https://doi.org/10.1371/journal.pone.
0168686 PMID: 28045938
28. Strachan DP. Hay fever, hygiene, and household size. Bmj. 1989; 299:1259–60. PMID: 2513902
29. Wang M, Karlsson C, Olsson C, Adlerberth I, Wold AE, Strachan DP, et al. Reduced diversity in the
early fecal microbiota of infants with atopic eczema. The Journal of allergy and clinical immunology.
2008; 121:129–34. https://doi.org/10.1016/j.jaci.2007.09.011 PMID: 18028995
Pet-keeping in infancy reduces allergy
PLOS ONE | https://doi.org/10.1371/journal.pone.0208472 December 19, 2018 12 / 13
30. Hesselmar B, Sjo
¨berg F, Saalman R, Åberg N, Adlerberth I, Wold AE. Pacifier cleaning practices and
risk of allergy development. Pediatrics. 2013; 131:e1829–37. https://doi.org/10.1542/peds.2012-3345
PMID: 23650304
31. Hesselmar B, Hicke-Roberts A, Wennergren G. Allergy in children in hand versus machine dishwash-
ing. Pediatrics. 2015; 135:e590–7. https://doi.org/10.1542/peds.2014-2968 PMID: 25713281
32. Hicke-Roberts A, Aberg N, Wennergren G, Hesselmar B. Allergic rhinoconjunctivitis continued to
increase in Swedish children up to 2007, but asthma and eczema levelled off from 1991. Acta Paediatr.
2017; 106:75–80. https://doi.org/10.1111/apa.13433 PMID: 27102081
Pet-keeping in infancy reduces allergy
PLOS ONE | https://doi.org/10.1371/journal.pone.0208472 December 19, 2018 13 / 13
Available via license: CC BY 4.0
Content may be subject to copyright.
