Validity of results
This population-based cohort study was based on a random sample of
children identified from the roster of Statistics Finland. Additionally,
the relatively high rates of participation (80.3% at baseline and
63.2% at follow-ups) and the similar participant characteristics at
baseline and follow-ups,18 suggest that any major
selection bias was unlikely. We assessed exposure to green spaces at
individual level, independently from the assessment of allergic
rhinitis. We were able to use both the prenatal and life-time
residential addresses in the exposure assessment, thus minimizing the
likelihood of misclassification of exposure. Furthermore, we evaluated
the effects of cumulative exposure to green spaces separately for spring
and summer seasons. Season-specific exposures to allergenic pollen may
affect the risk of allergic rhinitis differently from each
other,15 which can explain the heterogeneity in
associations between green spaces and allergic rhinitis at least partly.
The outcome assessment was based on parental and self-administered
questionnaire information on the presence and age of onset of allergic
rhinitis up to 27 years and thus, this study captured the effect of
prenatal and early-life exposure to green spaces across life course. Any
misclassification of the outcome was independent from exposure
assessment and could introduce non-differential error, potentially
leading to underestimation of the studied effects. We were able to
adjust for several potential confounders, including individual
characteristics, socioeconomic status, and environmental exposures. This
is also one of the few studies assessing the mediation and modification
effects of air pollution on the association between green spaces and
allergic rhinitis.34-36 Finally, sensitivity analyses
were conducted to assess the impact of the selected distance buffers
(100 m, 500 m, and 1000 m) and quartiles for exposure to NDVI, and their
consistent results demonstrate the robustness of the study findings.
The current study had some limitations. NDVI is an objective measure
used to assess green vegetation cover and density, but it does not
provide qualitative information on vegetation characteristics (e.g.,
plant species), feature, nor on quality or accessibility to green
spaces. We did not measure the pollen counts/concentrations or their
quality, nor specific weather characteristics that could affect pollen
and air pollutant concentrations and NDVI exposures. These may limit the
assessment of potential mechanistic pathways. However, considering that
the season may affect the type and the conditions of
greenness/vegetation as well as the levels of
pollens,37 the season-specific effect estimates
indirectly provide information on the role of the quality of vegetation
or on changes in the abundance of vegetation on the association between
exposure to green spaces and allergic rhinitis. Additionally, similarly
with other studies on neighborhood health effects, our study may not
capture an individual’s true geographic context (Uncertain Geographic
Context Problem), since some of mothers’ and children’s activities may
take place outside of their residential environment.38