DISCUSSION
This study aimed to elucidate and characterize the development of
allergic rhinitis (AR) in young children below 2 y.o. who had atopic
dermatitis and/or food allergy. Those children are regarded as a
high-risk population for atopic march leading to respiratory allergies,
including asthma and AR, in later life. Since there are no definitive
diagnostic criteria for AR in young children, we tentatively defined 8
subgroups based on 3 major clinical features of HDM-related AR, which is
the most prevalent type of AR in Japanese children below the age of 6
years.12 We found that the prevalence of AR-like
phenotypes, which meet 2 or more of the 3 clinical criteria (i.e.,
persistent nasalsymptoms, nasal eosinophilia and HDM sensitization),
increased significantly during the 2-year follow-up, from an average age
of 13 months to 27 months, and was associated with the cumulative
prevalence of physician-diagnosed asthma.
Several papers have reported on the prevalence of AR in infants and
young children. The PARIS birth cohort study found that 9.1% of infants
at age 18 months had AR-like symptoms (runny nose, blocked nose,
sneezing; apart from a cold).10 Positive associations
of AR-like symptoms with inhalant allergen sensitization, HDM
sensitization and a blood eosinophil count ≥400/mL were described, with
adjusted odds ratios of 2.2, 2.9 and 1.5, respectively. Osawa et
al.21 reported that general health examinations at age
18 months found that 8 (2%) of 408 infants had both inhalant allergen
sensitization and nasal eosinophils, and 6 (1.5%) infants also had
rhinorrhea. Zeiger et al.22 investigated nasal
basophilic cells and eosinophils in a cohort of high-risk infants with
atopic parents who participated in an intervention study of maternal
allergenic food avoidance for allergy prevention. They found that the
prevalence of nasal eosinophils increased from 0.7% at age 4 months to
38.2% at 48 months, and the prevalence was higher in infants with an
allergic disease such as FA. In the MAS birth cohort, the prevalence of
AR (based on the ISAAC questionnaire) was investigated from age 3 years
to 13 years.23 The prevalence of symptom-based AR
increased continuously from 6% at age 3 to 24% at age 13 in children
with no parental history of allergy and from 13% to 44% in those with
parental allergy. Sensitization to aeroallergens was a significant risk
for AR, with an odds ratio of 18.9. Collectively, AR starts in infancy
and becomes more common as children grow older, especially in those with
a genetic predisposition to allergy and/or aeroallergen sensitization.
Our present results from a high-risk cohort of infants with AD and/or FA
were in accord with those earlier findings. An added strength of our
study is that we prospectively investigated AR development in regard to
not only the nasal symptoms but also nasal eosinophilia and
sensitization during a critical 2-year period of infancy (up to 37.3
months of age).
AR is a Th2-type inflammatory disease of the upper
airways.24,25 It is diagnosed on the basis of the
pathogenesis, i.e, IgE antibody to an aeroallergen, IgE-mediated mast
cell activation leading to mucosal symptoms such as rhinorrhea, nose
itching, sneezing, blocked nose, and eosinophilic inflammation of the
nasal mucosa leading to nasal hyper-reactivity. Positive findings for
the ”triad” of AR criteria, i.e., 1) “allergic” nasal symptoms, 2)
aeroallergen sensitization and 3) nasal eosinophilia, can establish the
diagnosis. However, in the early stage of AR development, the triad may
not be present due to the non-specific nature of symptoms and variable
nasal mucosal pathology. Moreover, in preschool children, viral upper
respiratory infections are common and symptom-recognition relies on a
mother’s observation, not the perception or complaints of the child
him/herself. To circumvent those
problems, we employed a classification system of AR-like phenotypes that
allows for fluctuation of AR manifestations in children. We found that
our classifications corresponded well with previous observations in
several birth cohort studies and, importantly, correlated positively
with the prevalence of asthma.
Nasal eosinophilia has been variously defined. In adults, more than
10%17 or 20%26 eosinophils in
inflammatory cells were regarded to be eosinophilia. In 11- to
15-year-old children, more than 10 eosinophils per high-power field was
defined as nasal eosinophilia, with a diagnostic specificity of 96% and
a sensitivity of 62% for AR.27 We defined nasal
eosinophilia in young children as when more than a few eosinophils were
present in the whole field, which may be a lower cut-off level for nasal
eosinophilia compared with those in adults and older children. Because
the number of eosinophils in nasal smears was reported to be much lower
in young children with rhinorrhea and no signs of acute
infection,28 we employed our above definition. Yet, we
found that a higher cut-off of ≥10% for the eosinophil/neutrophil ratio
yielded similar results, although the prevalence of eosinophilia with
this definition was lower than that with the former definition. Overall,
we believe that our definition of positive nasal eosinophilia may be
optimal due to its higher sensitivity.
Multimorbidity is a characteristic feature of allergies in
children.25 The MeDALL (Mechanisms of the development
of ALLergy) study, involving 12 cohort studies, reported that
coexistence of asthma, rhinitis and eczema in the same child was more
prevalent than would be expected by chance.29 An
important aspect of multimorbidity is development of asthma, because it
places the largest burden on children. We previously reported that 78%
of children younger than 10 years old with asthma had AR, and onset of
AR preceded asthma onset in one-third of them.9 In a
German MAS birth cohort, AR at 2 years and 5 years of age predicted
later development of wheeze.6 Our present study, in
which we focused on “multimorbid” infants with AD and/or FA, confirmed
those earlier findings.
This study has several limitations. First, the number of subjects was
relatively small compared with large birth cohort studies. However,
prospective investigation of the development of AR relied on not only
the nasal symptoms but also HDM sensitization and nasal eosinophilia,
which we believe enhanced the reliability of our findings. Second,
definitive diagnosis of asthma at an early age is difficult, and the
asthma outcome findings of this study may not be accurate. For that
reason, we continue to follow the subjects for development of school-age
asthma and will confirm our results by using more objective findings,
such as lung function. Third, our cohort was high-risk infants with
AD/FA, and we may not be able to extrapolate the findings to natural
history in the general population. However, because of the focused
population, we were able to describe the clinical course of early AR in
our relatively small number of subjects.
In conclusion, we described the early phase of AR development in a
high-risk cohort of children and proposed practical diagnostic criteria
for AR in young children. Our findings show that AR increased markedly
during 1 to 3 years of age, and that increase was positively associated
with the prevalence of asthma. Additional studies will be needed to
clarify the natural history of AR in children in the context of atopic
march.