We thank Drs Rajendra and Robertson for their supplementary commentary, analysis, work and interest in this intriguing field of research. Although their data had somewhat similar results to our case–control study, the critical differences in study design warrant explication.

The associations between Helicobacter pylori, Barrett’s oesophagus and gastro-oesophageal reflux disease (GORD) have been contradictory: some studies have shown no association, some a positive association and some, like the case series by Drs Rajendra and Robertson, suggested an inverse association.1^–3 However, the case series design used by almost all prior studies (which, unfortunately, could not all be cited due to space limitations) are very susceptible to bias, which may have created the conflicting results.

Given this knowledge, we used a community-based case–control design to attempt a less biased estimate of the association between H pylori …