Introduction
Eosinophilic esophagitis (EoE) is an atopy-related chronic inflammatory disease whose activity may be influenced by environmental aeroallergen exposure. Prior low quality evidence has suggested a seasonal pattern in EoE diagnosis [1], but whether food bolus impactions specifically display seasonal variability remains controversial. We aimed to investigate the seasonal distribution of food bolus impaction events in a registry-based EoE cohort and to explore potential associations with regional pollen exposure.
Aims & Methods
We analysed prospective data from patients recruited at Tomelloso General Hospital (Spain) registered in EOE CONNECT [2,3] (snapshot November 2025). Food bolus impaction events with a known month and year were extracted and assigned to meteorological seasons (winter: Dec–Feb; spring: Mar–May; summer: Jun–Aug; autumn: Sep–Nov), and classified as pre- or post-diagnosis. Seasonal variability was assessed by chi-square goodness-of-fit against a uniform distribution and Edwards' test for unimodal cyclic trends. Differences between pre- and post-diagnosis seasonal distributions were evaluated by chi-square test of independence. Aeroallergen correlation analyses used daily pollen counts from two regional monitoring stations: AEROCAM (Toledo, 2019–2023) [4] and the Comunidad de Madrid open-data network (Aranjuez, 2010–2025) [5], with same-month Spearman rank correlations and Poisson regression between monthly pollen concentration and monthly impaction count per taxon.
Results
Among 448 registered patients, 170 (37.9%) had ≥1 dated food bolus impaction event, yielding 217 analysed events (146 pre-diagnosis, 71 post-diagnosis). Seasonal distribution is shown in the table. Edwards' test identified a statistically significant unimodal annual cycle (χ²=11.82, df=2, p=0.003), with a peak in July and relative amplitude of 0.117. The pattern was consistent across pre- and post-diagnosis events (p=0.361), arguing against a clinical-scheduling artefact. In the aeroallergen analysis, same-month spring–summer pollen was positively correlated with monthly impaction counts (Poaceae ρ=+0.18, p=0.020; Olea ρ=+0.18, p=0.017; Quercus ρ=+0.15, p=0.037), whereas Cupressaceae showed a significant inverse association (Poisson IRR=0.84, p=0.014), consistent with its winter peak.
| Season | Events (n) | % | Expected (uniform, n) | p-value |
|---|
| Winter | 47 | 21.7 | 54.2 |
|
| Spring | 49 | 22.6 | 54.2 |
|
| Summer | 68 | 31.3 | 54.2 |
|
| Autumn | 53 | 24.4 | 54.2 |
|
| Total | 217 | 100.0 | 217 |
|
| Chi-square goodness-of-fit |
|
| df=3 | 0.172 |
| Edwards' test (cyclic trend) |
|
| df=2 | 0.003 |
| Pre- vs. post-diagnosis |
|
| df=3 | 0.361 |
Conclusion
Food bolus impactions in this large, single-center cohort displayed a significant unimodal seasonal pattern peaking in July, independent of the diagnostic phase. The concordance with spring–summer aeroallergen exposure supports a biological rather than artifactual basis for EoE seasonal variability, suggesting that pollen-driven mucosal inflammation may modulate impaction risk. Seasonality in EoE appears to be a reality, not a myth.
References
1. Lucendo AJ, et al. Allergy. 2015;70:1640–1650.
2. Lucendo AJ, Santander C, et al. Therap Adv Gastroenterol. 2022;15:17562848221074204.
3. Navarro P, Laserna-Mendieta EJ, et al. United European Gastroenterol J. 2022;10:507–517.
4. Pérez-Badia R, et al. Ann Agric Environ Med. 2011;18:99–111.
5. Comunidad de Madrid. Histórico de mediciones de polen [dataset]. datos.comunidad.madrid. Licence CC-BY 4.0; 2010–2025.