Different risk factors influence peptic ulcer disease development in a Brazilian population

Figure 1 Incidence of chronic gastritis, duodenal ulcer and gastric ulcer disease and distribution of the investigated independent variables. A: Incidence of gastric diseases in patients who underwent endoscopy (n = 1466); B: Distribution of the histopathological parameters of atrophy (n = 1020); C: Intestinal metaplasia (n = 1094); D: Positive histological diagnosis of Helicobacter pylori (H. pylori) (n = 1045); E: H. pylori vacA m1 genotypes (n = 668); F: H. pylori vacA m2 genotypes (n = 668). CG: Chronic gastritis; DU: Duodenal ulcer; GU: Gastric ulcer; vacA: Vacuolating cytotoxin A. Values outside circles are negative tests.

Figure 2 Logistic regression curves of age and sex as independent factors to explain gastric ulcer disease incidence. The model has a significant fit, with contributions of both coefficients (χ² = 30.5, P < 0.0001). Additional test for the age/sex interaction had a nonsignificant result (β₃ = -0.011, SE = 0.0161, P = 0.4844, OR = 0.989, CI- = 0.958, CI+ = 1.021). OR: Odds ratios; CI: Confidence intervals; GU: Gastric ulcer.

Figure 3 Logistic regression curves of age as an independent factor to explain atrophy and intestinal metaplasia incidences. All patients who underwent endoscopy and who had at least one of the three diseases were included in the analysis. Both models had highly significant fits (atrophy OR = 1.0297, P < 0.0001; intestinal metaplasia OR = 3.6077, P < 0.0001). OR: Odds ratios.

Figure 4 Logistic regression curves of age and duodenal ulcer disease as independent factors to explain Helicobacter pylori incidence. The model had a highly significant fit (age OR = 0.9827, P < 0.0001; DU OR = 3.6077, P < 0.0001). A posterior test for the age/DU interaction had a nonsignificant result (β₃ = 0.0085, SE = 0.0168, P = 0.6142, OR = 1.0085, CI- = 0.9758, CI+ = 1.0423). DU: Duodenal ulcer; OR: Odds ratios; CI: Confidence intervals; H. pylori: Helicobacter pylori.