Background Patients with rheumatoid arthritis (RA) frequently develop dyspepsia due to peptic ulceration. There has been conflicting data published on the interactive roles of non steroidal anti inflammatory drugs (NSAIDs) and colonisation of the gastric antrum with Helicobacter pylori.

Objectives We wished to assess the prevalence of peptic ulceration in dyspeptic RA patients and matched controls. We intended to calculate the relative risks (RR) of NSAID consumption and Helicobacter infection, both singly and in combination.

Methods We endoscoped 100 dyspeptic patients with RA and 50 age and sex matched controls with dyspepsia but no evidence of RA. We recorded all medication taken over the preceding 6 months and assessed the presence of Helicobacter using a CLO test at endoscopy.

Results Endoscopic evidence of peptic ulceration was found in 39% of RA patients and 12% of controls. (p < 0.01). Multiple ulcers were significantly commoner in the RA patients (15% v 2%). NSAIDs were used by 60% of RA patients and 22% controls (p < 0.01). Helicobacter was found in 41% of RA patients and 36% of controls (NS). The consumption of NSAIDs conferred a RR of 5.2, while the presence of Helicobacter gave an RR of 2.0 in RA patients. The RR for the combination of NSAID consumption and Helicobacter infection was 12.3.

Conclusion RA patients have more major and more multiple pathology at endoscopy than age, sex and symptom matched control subjects. This is mainly due to their increased NSAID consumption but is also contributed to by the presence of Helicobacter infection.