Multivariate analysis of risk factors for development of duodenal ulcer in Helicobacter pylori-infected patients

Abstract

Background: Although Helicobacter pylori is a significant etiologic factor of peptic ulcer disease, it remains unknown why ulcers develop only in the minority of infected individuals. Aim: The aim of this cross-sectional study was to evaluate the association between the presence of duodenal ulcer in H. pylori-infected patients and different risk factors. Methods: A total of 122 H. pylori-infected patients were enrolled; 79 had duodenal ulcer and 43 gastritis. Univariate analysis was conducted using either Fisher's exact test or exact Cochrane-Armitage trend test. In multivariate analysis the logistic model was used. Results: Univariate analysis indicated six factors (male sex, smoking, antral H. pylori density, cagA presence in antrum, and vacA s1a presence in antrum and corpus). Four factors (sex, smoking-alcohol index, H. pylori density index, and cagA index) were found to be significant in multivariate analysis. The best model predicting duodenal ulcer included male sex, smoking, presence of H. pylori on histopathology in antrum and cagA presence in corpus. Conclusion: Although several risk factors were significantly associated with duodenal ulcer, we failed in the identification of either a single risk factor or a set of factors that can unequivocally differentiate patients with ulcer from those with gastritis.