Particulate Matter Under 2.5 Micrometers as a Possible Risk Factor for Eosinophilic Esophagitis
Citation:
Rappazzo, K., C. Cotton, C. Gaber, K. Hoffman, K. Turner, R. Genta, E. Jensen, AND E. Dellon. Particulate Matter Under 2.5 Micrometers as a Possible Risk Factor for Eosinophilic Esophagitis. Digestive Disease Week, NA, NC, May 21 - 23, 2021.
Impact/Purpose:
Investigate the association between PM2.5 and eosinophilic esophagitis.
Description:
Background: Particulate matter smaller than 2.5 micrometers (PM2.5) is known to cause heart and lung diseases including asthma. Poor environmental quality and rural locations have been shown to be associated with eosinophilic esophagitis (EoE), but the reasons why are unknown, as is the relationship between PM2.5 and EoE. We aimed to estimate the effect of PM2.5 concentration on EoE. Methods: We performed a case control study in a large pathology database of esophageal biopsies from across the United States. Cases were defined as ≥15 eosinophils per high-powered field in the absence of other histopathologic causes; controls were all other patients with esophageal biopsies. All patient data was geocoded by home address to the census tract of residence. Exposure to PM2.5 was determined by census tract location using estimates from the EPA’s Fused Air Quality Surface Using Downscaling model, averaged over the last five single-year concentrations prior to case occurrence. To estimate the effect of long-term PM2.5 concentration on the prevalence of EoE, we performed logistic regression adjusting for age, census tract population per square mile, census tract race/ethnicity, and census tract age distribution, using an auto-regressive correlation structure for clustering by census tract. We estimated the case control odds ratio with 95% robust confidence limits (CL). Results: There were 29,956 included EoE cases and 587,381 non-EoE controls. EoE was more common in census tracts with a higher proportion of non-Hispanic White residents and in census tracts with a younger age distribution (Table). EoE was more common in rural populations where PM2.5 concentration was often lower (Figure 1). The case control odds ratio of EoE for each 10 additional micrograms of PM2.5 per meter cubed was 0.84 (95% CL: 0.78 – 0.90) and the matched, adjusted prevalence ratio was 0.63 (95% CL: 0.60 – 0.68). Discussion: In this case control study, there is an inverse association between higher PM2.5 and EoE, making PM2.5 an unlikely contributor to EoE diagnosis. These results are also consistent with the previously described association between EoE and rural areas with low population density, where PM2.5 is generally lower. Further studies should be undertaken to determine different environmental risk factors for EoE.