Logistic regression models of the association between <i>H. pylori</i> infection, serum pepsinogen levels and vibriocidal seroconversion after immunization with a single oral dose of CVD 103-HgR.

Partially adjusted analysis, in addition to H. pylori serostatus/PG levels, age and sex were added to the analysis. The following variables were entered: H. pylori infection, age (in years as a continuous variable), sex, PG I, PG II, PG I∶PG II ratio, baseline vibriocidal antibody titers (transformed into natural logarithm) S. flexneri 2a IgG and hepatitis A antibodies (as markers for environmental fecal contamination and low socioeconomic status). The final model of children aged <5 years included PG II and PG I∶PG II ratio but they were not significant, and gender (OR 0.42 95% CI 0.19–0.92, for Males vs. females). The final model of children aged 5–9 years included S. flexneri antibodies but it was not significantly associated with vibriocidal seroconversion. In addition to H. pylori/PG II status, the following variables were entered to the analysis: age (in years as a continuous variable), gender, S. flexneri 2a IgG, hepatitis A antibodies and baseline vibriocidal antibody titers. The final model included the variables gender and hepatitis A, but they were not significantly associated with vibriocidal antibody seroconversion,