Aeroallergen modeling and the effects of aeroallergen concentrations on asthmatic symptoms and ER visits

Background. Public reporting of daily ambient pollen and mold spore concentrations are made from a limited network of aeroallergen monitor stations. Use of statistical models may improve estimates of allergens in areas where monitors do not exist. These estimates can enhance ambient allergen exposure assessments and further the understanding of the association between ambient allergens and allergic respiratory illness.;Methods. The first study in this dissertation utilizes daily pollen measurements from 14 aeroallergen monitoring stations located in a region consisting of New England and the Mid-Atlantic states to develop methods to spatially estimate pollen levels in this area. Inverse distance weighting, kriging, mixed effects models and generalized estimating equations were considered. The second study analyzes the association between daily ambient pollen and mold spore exposures and the likelihood of chest tightness, night symptoms, persistent cough, shortness of breath, wheeze and rescue medication use among a cohort of 463 asthmatic children. The third analysis is an ecologic study examining the association between ambient pollen and mold spore exposures and the likelihood of emergency room (ER) visits for allergic respiratory symptoms in New York State.;Results. Estimates from spatial interpolation and longitudinal regression models were found to be in good agreement with observed pollen measurements, except for days of absent or very high counts. The second study found associations between grass and tree pollen exposures and asthmatic symptoms among children classified as more severe asthmatics. In contrast, grass pollen and mold spore exposures demonstrated associations with a limited number of symptoms among less severe asthmatics. Study three found tree, grass and overall pollen exposures were associated with increased risk of ER visits. No association was observed with weed exposures. Magnitude of the association between allergen exposure and risk of ER visits were also found to vary by race and age.;Conclusions. Statistical models can improve the accuracy of ambient pollen concentrations in areas where monitors do not exist. Individual level tree, grass and weed pollens as well as mold spore exposures revealed associations with daily asthmatic symptoms that differed by asthma severity classification. Exposures to tree, grass and overall pollen were associated with increased risk of ER visits.