Does Helicobacter pylori Protect against Eosinophilic Esophagitis in Children?

BackgroundHelicobacter pylori infection and eosinophilic esophagitis (EoE) in children seem to have a reversed association with socioeconomic status (hygienic condition) and allergy conditions. While Hp infection (Hp) is highly associated with poor hygiene and/or poor socioeconomic status, but not with allergic conditions (asthma, rhinitis, etc.), EoE has the opposite epidemiological relationship (high association with allergy but low with low hygienic conditions). AimTo investigate the association between Hp infection and EoE in children. MethodsA retrospective chart review of all children who undergo the first upper endoscopy procedure in the gastroenterology clinic, between 2007 and 2012, was performed. Demographic, endoscopic and histological data were collected. The data was divided into 4 diagnostic groups: Hp infection, EoE, reflux esophagitis, and children who had normal histology. The relationship between Hp positive children and the other groups was performed. ResultsA total of 966 charts were available for review. Esophagitis, idiopathic gastritis, EoE, and Hp infection were detected in 268 (28%), 480 (49%), 62 (6%), and 31 (3%) children, respectively. The mean age of the EoE group was significantly lower compared to all reference groups (p<.002), but no significant different was detected among the reference groups (gastritis, GERD, and Hp infection; p=1.00). Simple logistic regression analysis using Hp infection as a predictor for EoE did not find a significant relationship between these two variables (p-value=.471, OR=0.478, 95% CI 0.06-3.56). However, multivariable logistic regression analysis between EoE and the reference groups indicated a significant negative relationship between Hp infection and EoE (p-value=.023, adjusted OR=0.096, 95%CI 0.013-0.72). Neither gastritis nor GER showed significant relationship with EoE (p-values are 1.000 and .992, respectively). ConclusionA reversed association between Hp and EoE was found in a cohort of West Virginia children. The possible explanations for these findings are discussed.