Associations between Total and Speciated Pollen Counts and Several Morbidity Measures in the Contiguous United States from 2008 to 2015
Sunday, March 4, 2018: 3:00 PM
South Hall A1 (Convention Center)
Jeremy J. Hess, MD MPH, , , , , , , , ,
Rationale: Pollen allergies cause substantial morbidity, yet knowledge of specifics regarding the timing and magnitude of risks associated with pollen exposure across the US is limited.

Methods: We combined a previously-assembled daily time series database of speciated pollen counts from 36 US stations from 2003 to 2016 with claims data from Truven Health Analytics MarketScan Research databases from 2008-2015. We identified the Metropolitan Statistical Area (MSA) for the stations and extracted daily MSA-specific health information. We divided exposure metrics at 50th, 75th, 90th, and 95th percentiles and used case-crossover methods to examine odds of outpatient visits for asthma and allergic rhinitis, emergency department (ED) visits for asthma, and allergy medications refills for high versus low exposure quintiles.

Results: Combined data covered the US with some gaps in the Midwest and Southwest. We found increased odds of outpatient visits for allergic rhinitis and asthma, ED visits for asthma, and medication purchases in association with high exposures to total pollen and several highly allergenic genera. Odds ratios ranged from 1.05 (95% CI 1.00-1.11) to 1.08 (1.04-1.33) for total pollen count 1-3 day moving averages for the various endpoints. Relationships were generally consistent across US ecozones.

Conclusions: Counts of total and allergenic pollens are moderately and consistently associated with several morbidity measures across the US. These results are consistent with more localized studies and extend the catalogue of quantitative estimates of pollen disease burden in the US. Results have implications for clinical management, public health surveillance, disease burden estimates, and health impact forecasting.