Background: The role of Helicobacter pylori infection in the development of oesophageal malignancies was investigated through a multivariate conditional logistic regression analysis in a nested case-control study.
Methods: Blood samples and a questionnaire on smoking and alcohol habits were collected from a cohort of 32,906 city residents during a health-screening programme between 1974 and 1992. Forty-four cases of oesophageal cancer and 149 matched controls were selected. The mean interval between screening and cancer diagnosis was 11.9 years. H. pylori seropositivity was determined by an enzyme-linked immunosorbant assay measuring IgG. Occupation was included in the statistical analysis as an indicator of socio-economic status.
Results: Helicobacter pylori seropositivity was present in 10 of the cases (22.7%) and 67 of the controls (45.0%). In a multivariate model, with adjustment for occupation, tobacco and alcohol consumption, the odds ratio for developing an oesophageal malignancy when infected with H. pylori was 0.29 (95% confidence interval (CI): 0.12-0.67). Current smokers had an odds ratio of 17.3 (95% CI: 3.0-99.4) and the odds ratio for ex-smokers was 5.9 (95% CI: 1.15-29.9). High alcohol consumption was no longer significantly associated with oesophageal neoplasms after tobacco smoking was included into the model, odds ratio 1.22 (95% CI: 0.46-3.2). The protective effect of H. pylori was more pronounced for oesophageal adenocarcinoma (seven cases, odds ratio 0.16, 95% CI: 0.00-1.06) than for squamous-cell carcinoma (29 cases, odds ratio 0.41, 95% CI: 0.14-1.2).
Conclusions: Helicobacter pylori infection is associated with a decreased risk of developing an oesophageal malignancy. Current smokers and ex-smokers have instead a definite increased risk of oesophageal neoplasms.