Characteristics of differences in Helicobacter pylori serology and 13C-urea breath-testing in an asymptomatic sample of blood donors.

Numerous tests, both invasive and non-invasive, are available for the diagnosis of Helicobacter pylori infection. The aim of this study was to evaluate the extent and determinants of differences in serology and 13C-urea breath test (13C-UBT) in diagnosing H. pylori. Four-hundred-and-seventy-four asymptomatic blood donors aged 40 to 68 years (mean age 55.8 years) and of German nationality were recruited between October 1996 and November 1997. H. pylori infection was measured with 13C-UBT and with a commercial IgG-ELISA test (Medac; Hamburg, Germany). A standardized questionnaire was applied to identify factors which could explain discrepant results of 13C-UBT and serology. Prevalence of infection was similar according to both tests (33.1% for 13C-UBT and 30.8% for serology). From the 146 subjects with a positive serology, 121 (82.7%) had a positive 13C-UBT. Previous treatment of H. pylori infection was the strongest predictor of a negative 13C-UBT given positive IgG serology (odds ratio (OR)= 12.87, 95% CI=4.10-40.36). Coffee consumption of more than 3 cups/day was also associated with an increased odds of a negative 13C-UBT given positive serology (OR = 3.12, 95% CI=1.16-8.43). No significant determinants of positive 13C-UBT given negative serology could be identified. These findings suggest a delayed fall in serum antibodies following eradication of the infection to be a major source of differences in H. pylori diagnosis using non invasive 13C-UBT and serology.