Objective: Although reflux esophagitis is a multifactorial disease, the relative importance of these pathogenetic factors has not been clearly established. In this study, regression analysis was used to model the major determinants of esophagitis in patients with symptomatic gastroesophageal reflux disease (GERD).
Methods: Sixty-six GERD patients and 16 asymptomatic controls were evaluated. All patients underwent upper endoscopy, esophageal manometry, and 24-h pH monitoring. Esophagrams were performed in 38 of the GERD patients and all controls. Stepwise regression was performed using esophagitis severity as the dependent variable. Logistic regression was performed grouping subjects as controls, nonerosive GERD, or erosive esophagitis.
Results: Hiatal hernia size, lower esophageal sphincter pressure, esophageal acid exposure, and number of reflux episodes >5 min significantly correlated with esophagitis severity. Stepwise regression identified hiatal hernia size (p = 0.0001) and lower esophageal sphincter pressure (p = 0.0024) as significant predictors of esophagitis. Logistic regression also identified hiatal hernia size (chi2 = 17.07, p < 0.0001) and lower esophageal sphincter pressure (chi2 = 5.97, p = 0.0146) as significant predictors of erosive esophagitis.
Conclusion: Esophagitis severity is best predicted by hiatal hernia size and lower esophageal sphincter pressure. Of these, hiatal hernia size is the strongest predictor.