Helicobacter pylori-independent effect of hyperglycemia on gastric mucosal atrophy

TO THE EDITOR: The relationship between Helicobacter pylori (H. pylori) infection and diabetes mellitus is still controversial (1, 2), although diabetes-related immunosuppression may influence the infection rate and clinical course. We have previously demonstrated that the H. pylori infection rate is not influenced by fasting blood glucose (FBG) level (3). However, diabetes-related immunosuppression may augment gastric colonization by H. pylori and aggravate the resulting gastritis and mucosal atrophy even if it is not related to the infection rate. Therefore, we performed the present study to investigate the correlation between the FBG level and gastric mucosal atrophy, with or without H. pylori infection. A total of 2500 people who visited Shimane Institute of Health Science, Shimane, Japan, for their annual medical checkup from September, 1998 to August, 1999 were prospectively enrolled in the present study. FBG level and H. pylori infection status were tested in all of the subjects. H. pylori infection was determined by measurement of serum H. pylori IgG antibody with an ELISA kit (IMMUNIS anti-PYLORI EIA, Institute of Immunology, Tokyo, Japan). Subjects who showed a FBG level exceeding 126 mg/dL (hyperglycemic group) or less than 85 mg/dL (normoglycemic group) were selected for measurement of serum pepsinogens I and II, and the I/II ratio, with an ELISA kit (AUTO ACE PG pepsinogen I and II, AZWELL, Osaka, Japan). Statistical analysis was performed using analysis of covariance to adjust for confounding values using the SPSS statistical package (version 6.1 for Macintosh, SPSS, Chicago, IL). Differences at p 0.05 were considered to be significant. There were 102 hyperglycemic subjects and 126 normoglycemic subjects, among whom H. pylori infection was detected in 59 (57.8%) and 63 (50.0%), respectively. The difference in the H. pylori infection rate between the groups was not significant. As shown in Table 1, hyperglycemic H. pylori-positive subjects tended to be older than normoglycemic H. pylori-negative subjects. Therefore, we corrected for age and sex as confounding factors in the subsequent statistical analysis of the data. Concentrations of serum pepsinogen I were significantly lower in hyperglycemic than in normoglycemic subjects, and pepsinogen II concentrations were markedly elevated in the H. pylori-positive subjects. Accordingly, the pepsinogen I/II ratio, a marker of the integrity of the gastric fundic mucosa, was significantly lower not only in H. pylori-positive but also hyperglycemic subjects (Table 1 and Fig. 1). Because the pepsinogen I/II ratio was lower in hyperglycemic subjects without H. pylori infection, these results demonstrated that hyperglycemia, like H. pylori infection, was an independent risk factor for the development of gastric mucosal atrophy. Marrollo et al. recently reported that diabetes mellitus aggravated H. pylori-induced gastritis (4). In the present study, however, we clarified that hyperglycemia may not only accelerate H. Table 1. Serum Pepsinogen I and II in Four Groups of Subjects