Predictors of duodenal eosinophil counts among subjects undergoing diagnostic endoscopy.

Introduction: Duodenal eosinophilia has been implicated in the pathophysiology of functional dyspepsia. In a retrospective observational study, we previously reported that duodenal eosinophilia (as defined by a mucosal count of greater than 15 eosinophils per 5 high power fields), was associated with symptomatic erosive gastroesophageal reflux disease (GERD), concomitant co-morbidities and Chinese ethnicity but not functional dyspepsia among 289 multiracial subjects undergoing diagnostic endoscopy in 2019 before the COVID-19 pandemic. We tested the reproducibility of those findings on a larger sample that included the original cohort and another 221 subjects who underwent endoscopy in 2022 after the easing of pandemic restrictions.

Materials and methods: Archived duodenal histology slides were assessed by a pathologist blind to demographic and clinical data gleamed retrospectively from clinical chart review. Logistic regression analysis was used to explore associations between duodenal eosinophilia and the variables age, gender, ethnicity, year of sampling (2019 vs 2022), concomitant co-morbidities, functional dyspepsia, symptomatic erosive GERD (Los Angeles Grades A to D), endoscopic oesophagitis, gallstone disease, Helicobacter pylori infection, irritable bowel syndrome and NSAID consumption. Three different thresholds for defining duodenal eosinophilia (>15, >22 and >30 eosinophils per 5 high power fields) were tested.

Results: Year of sampling (2019, pre-pandemic) strongly predicted duodenal eosinophilia across all thresholds (OR 11.76, 13.11 and 21.41 respectively; p = 0.000). The presence of concomitant co-morbidities was a modest predictor across all thresholds whereas Chinese ethnicity only predicted at the lowest threshold. Absolute duodenal eosinophil counts predicted symptomatic erosive GERD (OR 1.03; p = 0.015) but not functional dyspepsia (OR 1.00; p = 0.896) after adjusting for age, gender, ethnicity, concomitant comorbidities and year of endoscopy. None of the subjects reached the threshold for the diagnosis of eosinophilic duodenitis.

Conclusion: The cumulative impact of environmental exposures on duodenal eosinophil counts may be much greater than of putative factors linked to functional dyspepsia. A signal linking duodenal eosinophil counts and symptomatic erosive GERD was detected.