Interpretation
We show that children with more severe, early-onset AD before age 1 year
have an increased risk for developing sensitization to aeroallergens and
allergic rhinitis compared to children with onset of AD after age 1
year. This is in agreement with some previous
studies23–26 as most studies found higher frequencies
of allergy outcomes in early-onset vs. late-onset
AD27,28, whereas one study found no
difference29 and two other studies only found a higher
risk if children with early-onset also had persistent
AD30,31. Further, the latter study found an increased
risk of sensitization to aeroallergens at 18 years among children with
late-onset AD31.
The varying findings in these previous studies can be due to several
factors. Importantly, the severity of AD in early life seems to be the
most important factor according to our study but was not assessed in
previous studies. Also, the diagnosis of AD and allergic rhinitis was
mostly based on questionnaires, and the evaluation of allergy outcomes
was done at different time-points across the studies ranging from 4 to
18 years. Only one other study diagnosed sensitization and allergic
rhinitis at several time points31, but with low
complete follow-up rates of 52% at 7 years, 40% at 12 years and 39%
at 18 years, in contrast to a follow-up rate of 71% at 7 years and 86%
at 12 years in our study. None of the previous studies assessed AD
severity, which showed significantly different effects in early-onset
vs. late-onset in relation to both aeroallergen sensitization and
allergic rhinitis.
The possible mechanisms driving the association between AD and
respiratory allergy have previously been explained by an underlying
Th2-skewed immune system10,32, which leads to an
increased risk of both AD, asthma and allergic rhinitis. An inherent
genetic susceptibility may also influence the risk of developing all
three diseases. Thus, a GWAS study identified shared AD, asthma, and
allergic rhinitis immune-related genetic variants, suggesting that these
disorders may co-exist because they share genetic risk loci that result
in dysregulation of immune-related genes33. However,
this does not explain the mechanism behind age at onset and severity of
AD and development of aeroallergen sensitization and allergic rhinitis.
More recently, the hypothesis explaining the association between AD and
allergy has shifted to include a primary defect in the epidermal barrier
in AD34. The epidermis provides an essential barrier
to the external environment, preventing both water loss from the body
and intrusion of infectious agents and allergens. When allergens are
captured and processed by antigen-presenting Langerhans cells of the
epidermis, they migrate to the draining lymph nodes and if not
down-regulated, they can interact with naïve T cells to promote Th2
immunity leading to allergies35. This epicutaneous
sensitization has also been proposed to cause migration of Th2 memory
cells from the skin to the bronchial lymphoid tissue, where subsequent
inhalation of the sensitizing allergen causes a cellular and humoral
response in the airways, resulting in symptoms of allergic rhinitis and
asthma10. Finally, it has been proposed that chronic
AD lesions may express thymic stromal lymphopoietin and other
proinflammatory mediators, which increase the risk of developing
allergic inflammation and sensitization in the
lungs36.
Our study also showed a trend of higher risk of developing allergic
rhinitis with increasing SCORAD among children with a FLGmutation compared to children without a mutation. Indeed, FLGmutation is known to lead to an impaired skin barrier, which is a
plausible mechanistic explanation of the observed associations. To our
knowledge no previous studies have investigated the role of FLGmutations in the relationship between AD age at onset, severity, and
development of respiratory allergy. Our results should be interpreted
with caution due to low numbers, and it would be interesting to
investigate this association in a larger-scale study.
To ensure that early-onset AD was not present with concomitant allergic
sensitization, and that reverse causation is not driving our results, we
investigated early allergic sensitization at age 6 months. However, only
2 children were sensitized to an aeroallergen at this point - one from
the early-onset AD group and one who never developed AD, so that is not
likely to alter our results.
Our study suggests programming in early life and that appropriate early
interventions may hamper the progression of AD to aeroallergen
sensitization and allergic rhinitis, especially in children with
early-onset, more severe disease and FLG mutation. Following such
thoughts, early strengthening of the epidermal barrier in high-risk
infants using emollients before the onset of AD has been hypothesized to
prevent development of AD and subsequent respiratory allergy, but this
could not be demonstrated in two recent randomized controlled
trials37,38.
To prevent development of aeroallergen sensitization and allergic
rhinitis in children with early-onset, severe AD, an alternative
possible effective intervention may reside in modulating the immune
response through manipulation of the route of earliest encounter with
the allergen, along the same lines as has been shown with food
allergens. Thus, early environmental exposure through the skin may lead
to sensitization whereas early oral exposure may led to
tolerance6–8 by stimulating regulatory T-cells
induction and function in the gut39. A murine study
has supported this hypothesis40, and two human studies
have tested this hypothesis in children using a sublingual immunotherapy
mixture of house dust mite, cat and grass41 or house
dust mite alone42,43. The studies failed to show any
effect on development of allergic sensitization, which was also less
influenced by early-onset and severity of AD in our study compared to
allergic rhinitis, i.e., clinically relevant sensitization. Further,
they had some weaknesses including high drop-out rates, lack of
compliance, the children being too old at the beginning of the study,
difficulty taking the sublingual immunotherapy correctly, and difficulty
identifying children at high risk for developing AD. To establish
whether early sublingual exposure to aeroallergens can prevent
development of sensitization and allergic rhinitis in children with
early-onset, severe AD, randomized controlled trials taking all these
factors into account are needed.