Helicobacter pylori infection and chronic atrophic gastritis: associations according to severity of disease. | Read by QxMD (original) (raw)
BACKGROUND: Helicobacter pylori infection is an established risk factor for chronic atrophic gastritis. However, estimates of the strength of this association have varied widely, possibly due to clearance of the infection in severe stages of chronic atrophic gastritis, which may lead to underestimation of the association. We assessed the association of H. pylori infection with chronic atrophic gastritis according to severity of disease.
METHODS: We measured serum pepsinogen I and II (as surrogates for chronic atrophic gastritis) and antibodies against H. pylori by ELISA in 9444 men and women aged 50-74 years in a population-based study in Saarland, a state of Germany. The association between H. pylori and chronic atrophic gastritis (defined as pepsinogen I <70 ng/mL and pepsinogen I/II-ratio <3) was analyzed after stratification of chronic atrophic gastritis cases by quintiles of pepsinogen I as proxy marker for severity of chronic atrophic gastritis.
RESULTS: When all cases were included, the odds ratio for the association with Chronic atrophic gastritis for H. pylori infection alone was 2.9 (95% confidence interval 2.3-3.6); it was 4.1 (3.2-5.2), for H. pylori infection that was positive for the presence of Ig G antibodies specific to the cytotoxin-associated gene A (CagA) protein-a well-established virulence factor of H. pylori. These ORs ranged from 11 (5.2-22) and 16 (7.7-34) for the quintile of cases with highest pepsinogen I (least severe cases) to 1.0 (0.7-1.6) and 0.9 (0.5-1.5) for the quintile of cases with lowest pepsinogen I (most severe cases). Five of 7 cases with CagA-seropositivity but negative H. pylori serostatus (a pattern indicative of past infection) were in the group of most severe cases.
CONCLUSIONS: Our results support the hypothesis of major underestimation of the association of H. pylori and chronic atrophic gastritis, due to clearance of the infection in advanced stages of the disease. These results suggest that the association is much stronger than estimated by most epidemiologic studies to date.