Objectives: The incidence of eosinophilic esophagitis is unknown in our area. The aim of our study was to determine the incidence of eosinophilic esophagitis and its possible association with the most frequent absolute annual pollen counts.
Methods: A descriptive retrospective multicenter observational study was designed to calculate the incidence of eosinophilic esophagitis in children aged under 15 years in the southwest region of Madrid, Spain in 2002-2013 (data were provided by the Statistics Institute of Madrid). We collected data on age, sex, clinical presentation, and date of endoscopic diagnosis. Relative risk (RR) was estimated (Stata v.11) using negative binomial regression models to assess the association between incidence and pollen counts (provided by Subiza Clinic).
Results: The study population comprised 254 patients (192 male [75.6%], aged 0.5-14.99 years). The clinical presentation was esophageal impaction in 23.6%, dysphagia in 22%, gastroesophageal reflux–like symptoms in 44.9%, and other findings in 9.4%. The annual incidence from 2002 to 2013 per 100 000 children aged <15 years per year was, respectively, 0.81, 1.5, 0.37, 3.17, 3.07, 4.36, 6.87, 7.19, 8.38, 9.05, 9.14, and 9.68. The incidence of eosinophilic esophagitis increased by an average of 19% annually (RR, 1.19; 95%CI, 1.14-1.25; P<.001). In the overall analysis, the relationship between incidence and absolute annual and monthly counts during the pollen seasons
of the respective pollen types was analyzed only for Platanus species, which had an RR >1 (1.17 and 1.06, respectively) (P<.05).
Conclusion: The incidence of diagnosis of pediatric eosinophilic esophagitis increased by an average of 19% annually. No significant association was found between incidence and pollen counts, except for a weak association with Platanus species.

Key words: Pollen, Eosinophilic esophagitis, Epidemiology