OBJECTIVE

Helicobacter pylori infection and atrophic gastritis are inversely related to erosive esophagitis. Whether these factors affect the pathogenesis of endoscopy-negative reflux disease is not clear. We aimed to elucidate the differences in clinical characteristics between endoscopy-negative erosive disease and erosive esophagitis.

METHODS

253 subjects (89 with endoscopy-negative reflux disease and 164 with erosive esophagitis) were studied. Gastric atrophy was assessed by measurement of serum pepsinogen. Logistic regression was used to calculate the odds ratio (OR) and 95% confidence intervals (CI) of endoscopy-negative reflux disease compared with erosive esophagitis.

RESULTS

Among GERD patients, female gender (OR = 2.27, 95% CI, 1.25-4.10), smoking (OR = 0.45, 95% CI, 0.22-0.91), and the presence of hiatal hernia (OR = 0.30, 95% CI, 0.17-0.56) were significantly associated with endoscopy-negative reflux disease compared with male gender, not smoking, and absence of hiatal hernia, respectively. Body mass index (BMI) was also significantly associated with a decreased OR for endoscopy-negative reflux disease. Although H. pylori infection and gastric atrophy were significantly more common in patients with endoscopy-negative reflux disease, these associations did not persist in a multiple-adjustment model. After adjustment for gender, BMI, smoking, and hiatal hernia, a decrease in serum pepsinogen I/II ratio was significantly associated with an increased OR for endoscopy-negative reflux disease (p for trend = 0.018).

CONCLUSIONS

Female gender, low BMI, not smoking, absence of hiatal hernia, and severity of gastric atrophy were positively associated with endoscopy-negative reflux disease compared with erosive esophagitis among Japanese patients.