Outdoor pollen-related changes in lung function and markers of airway inflammation: A systematic review and meta-analysis

Background: Experimental challenge studies have shown that pollen can have early and delayed effects on the lungs and airways. Here, we qualitatively and quantitatively synthesize the evidence of outdoor pollen exposure on various lung function and airway inflammation markers in community-based studies.

Methods: Four online databases were searched: Medline, Web of Science, CINAHL and Google Scholar. The search strategy included terms relating to both exposure and outcomes. Inclusion criteria were human-based studies published in English that were representative of the community. Additionally, we only considered cross-sectional or short-term longitudinal studies which investigated pollen exposure by levels or season. Study quality assessment was performed using the Newcastle-Ottawa scale. Meta-analysis was conducted using random-effects models.

Results: We included 27 of 6551 studies identified from the search. Qualitative synthesis indicated associations between pollen exposure and predominantly type-2 inflammation in both the upper and lower airways, but little evidence for lung function changes. People with ever asthma and/or seasonal allergic rhinitis (SAR) were at higher risk of such airway inflammation. Meta-analysis confirmed a positive relationship between pollen season, eosinophilia and eosinophil cationic protein (ECP) in people with ever SAR but the results between studies were highly variable. Heterogeneity was reduced after further subgrouping by age, and the forest plots indicated that eosinophilic airway inflammation to outdoor pollen exposure increased with age.

Conclusion: Among people with ever asthma and ever SAR, exposure to increased ambient pollen triggers type-2 upper and lower airway inflammation rather than a non-specific or innate inflammation. These findings can lead to the formulation of specific pollen immunotherapy for susceptible individuals. Future research should be directed towards investigating lagged associations and effect modifications using larger and more generalized populations.