31.   According to previous studies, gastric metaplasia is an almost constant finding in patients with duodenal ulceration.

32.   An important problem in the detection of gastric metaplasia is its patchy distribution.

33.   In addition, underestimation of gastric metaplasia may occur if biopsy specimens have been collected from the distal bulb.

34.   Other factors may also have contributed to the large variability in prevalence rate of gastric metaplasia as mentioned in published works.

35.   In this study, the extent of gastric metaplasia was estimated independently by two investigators to limit interobserver variability.

36.   In agreement with previous studies, we found that gastric metaplasia was significantly more prevalent in patients with duodenal ulcer disease than in patients with functional dyspepsia.

37.   In contrast with some of these studies, we also found a comparatively high prevalence of gastric metaplasia in the second group of patients.

38.   In accordance with Wyatt et al we found a significant association between the extent of gastric metaplasia and microscopic signs of duodenitis.

39.   We could not show a significant difference between the extent of gastric metaplasia at sites around ulcer craters or scars and macroscopically normal appearing mucosa.

40.   It has been suggested that gastric metaplasia develops as a result of high intragastric acidity or rapid gastric emptying, or both.