Acute Associations between PM2.5 and Allergic and Nonallergic Asthma Exacerbations in Children and Adults

Abstract

Introduction: Asthma results from airway inflammation and has allergic and nonallergic phenotypes. It is unknown whether susceptibility to the acute effects of air pollution differs by asthma phenotype. Methods: Associations of ambient PM2.5 concentrations with acute asthma visits (including unscheduled office, urgent care, emergency department, and hospitalization) were investigated using electronic medical records from Renown Health in Reno, the largest healthcare network in Northern Nevada. International Classification of Disease (ICD) codes were used to identify patients with at least one asthma visit (N=23,331), and to classify asthma patients by phenotype based on the presence or absence of an allergic comorbidity (atopic dermatitis, allergic rhinitis, or allergy) in their medical records (allergic N=7,744, nonallergic N=15,587). Daily mean PM2.5 and temperature were obtained from a centralized monitor. Time-stratified case-crossover analyses were conducted matching on month and day of week for the time period December 2011 - June 2017. Associations of 3-day moving average (lag 0, lag 1, lag2) PM2.5 concentrations were estimated separately by asthma phenotype and adjusted for temperature and holidays. Analyses were further stratified by age (children/adult) and gender. Results: There were 49,832 acute asthma visits over the study period (21,887 among allergic patients and 27,945 among nonallergic patients). In preliminary analysis, the odds ratio (OR) for all asthma visits combined per 10 µg/m3 increase in 3-day moving average PM2.5 was 1.064 (95%CI: 1.043, 1.085). The ORs for allergic and nonallergic asthma were 1.070 (95%CI: 1.038, 1.102) and 1.060 (95%CI: 1.032, 1.088) respectively. ORs for allergic and nonallergic asthma were also similar when estimated within strata of age and gender. Conclusion: Asthma visits increased 6% per 10 µg/m3 increase in 3-day moving average PM2.5. Preliminary analysis showed little evidence of heterogeneity by asthma phenotype.