Green Tea Consumption and Chronic Atrophic Gastritis : A Cross-Sectional Study in a Green Tea Production Village

Chronic atrophic gastritis (CAG) is well known as a precancerous lesion of the stomach, and Helicobacter pylori (H. pylori) infection increases the risk of CAG. While recent studies have reported that green tea consumption decreases the risk of gastric cancer, there has been no study analyzing the relationship between green tea consumption and the both risks H. pylori infection and CAG. We conducted a cross-sectional study on 636 subjects living in a farming village in Japan to examine the relationship among green tea consumption, H. pylori infection, and CAG. Smoking, alcohol drinking, consumption of four beverages, including green tea, and of five foods were investigated as lifestyle factors that may affect H. pylori infection and CAG. The measurement of H. pylori-IgG antibodies was used to define H. pylori infection, and serum pepsinogens were used to define of CAG. The unconditional logistic regression model was used for analyzing each odds ratio (OR). H. pylori infection was positively associated with the risk of CAG (OR = 3.73; 95% confidence interval [CI], 2.59-5.36). High green tea consumption (more than 10 cups per day) was negatively associated with the risk of CAG, even after adjustment for H. pylori infection and lifestyle factors associated with green tea consumption (OR = 0.63; 95% CI, 0.43-0.93). These results support the hypothesis that high green tea consumption prevents CAG. J Epidemiol, 2000 ; 10 : 310316


INTRODUCTION
Helicobacter pylori (H.pylori) infection is well known as an important risk factor for chronic atrophic gastritis (CAG) 1,2), and CAG is known as an intermediate stage in the pathological course leading to gastric cancer 3,4).Several epidemiological studies have indicated that not only H. pylori infection but also differences in dietary habits can influence the occurrence of .This means that the identification and control of risk factors for CAG help to prevent the occurrence of CAG and consequently lead to a decrease in the occurrence of gastric cancer.
On the other hand, recent laboratory studies have shown that the green tea extract inhibits carcinogenesis in the skin, stomach, and duodenum of rats and mice 8-10).In epidemiological studies as well, high consumption of green tea has been report-ed to decrease the risk of gastric cancer [11][12][13].In considering the early prevention of gastric cancer, it is very important to investigate the effect of green tea on the stage in which normal mucosa progress to CAG, and then gastric cancer.Although previous studies have analyzed the relationship between green tea consumption and H. pylori infection 14. 15), these studies did not report the relationship between green tea consumption and CAG.In Japan, where both the H. pylori infection rate and the prevalence of CAG are higher than other developed countries [16][17][18] , the elucidation of the relationship between green tea consumption and both H. pylori infection and CAG is very important.We carried out a cross-sectional study on the residents of a Japanese farming village famous for green tea production and which has a low mortality of gastric cancer.The purpose of this study is to investigate the relationship among green tea consumption, H. pylori infection, and CAG.

Subjects
The study area was a farming village in Fukuoka Prefecture, Kyushu, Japan.The chief industry is agriculture and the proportion of residents engaged in agriculture is about eight times higher than that of Japan as a whole (49.5% vs. 6.5% according to a 1995 census).
In 1995, in this village, which is famous for its green tea production, tea plantations occupied 44.2% of the entire commercially cultivated area.The standardized mortality ratio (SMR;1986-1996) of gastric cancer in this village was 70.2, which was significantly low compared to Japan as a whole (P < 0.05).
The total population of the village in 1997 was 4,083.Annual health screenings are conducted for residents aged 30 and over.
We asked all of the 1,696 residents expected to participate in this health screening in 1997 to cooperate with our study and distributed questionnaires on lifestyles, including the consumption of green tea.Of these residents, 1,207 consented to our request, answered questionnaires and gave blood samples.
Although the proportion of female subjects was higher than that of the population aged 30 and over in the village, there was no significant difference in mean age between the population and the male and female subjects combined (59.14 years vs. 60.44 years).Fourteen of the subjects were excluded because of insufficient data on their lifestyles, 45 subjects were excluded because they had a history of gastric resection, and 512 elderly subjects aged 65 or older were excluded to reduce the influence of aging on atrophy 19,20).A final total of 636 residents aged 30-64 years were enrolled as subjects in this study.

Serological Methods
H. pylori infection was define by serum IgG antibodies for H. pylori using an enzyme immunoassay (GAP-IgG, Biomerica, USA).The sensitivity and specificity of this assay have been shown to be 94% and 76%, respectively, in comparison with cultivation of biopsy specimens 21).The level of serum pepsinogen I (PGI) and pepsinogen II (PGII) in the same samples were measured for identification of CAG using a radioimmunoassay (PG I / PG II RIABEAD, Dainabot Co. Ltd., Tokyo).CAG positive was define by both serum PG I < 70ng/ml and PG 11 PG II ratio < 3.0.The sensitivity and specificity of this criterion have been shown to be 86% and 82%, respectively, in comparison with endoscopic atrophy 22).Recently, several epidemiological studies regarding the risk of CAG adopted this criterion, and the reliability has been recognized 5,7).

Statistical Methods
Odds ratios (ORs) for H. pylori infection according to lifestyle variables were calculated after adjustments for gender and age.Then, assuming that H pylori infection itself was an additional risk factor, the OR for CAG according to each variable was also calculated.We conducted multivariate analyses including lifestyles variables, which were significantly associated with green tea consumption based on univariate analysis, in order to examine independent relation of green tea consumption to H. pylori infection and CAG.All variables were assigned dummy variables except for age (actual value).All analyses were conducted by unconditional logistic models.The SAS statistical software package (version 6.11) was used in all statistical analyses.Notes: H.pylori: Helicobacter pylori CAG: Chronic atronic gastritis #: The test for trend was calculated by a logistic regression model.

RESULTS
Table 1 shows the distribution of the H. pylori antibody-positive rate and the CAG prevalence by gender and age group.The H. pylori positive rates in males and in females were 48.6% and 48.8%, respectively.The prevalence of CAG in males and in females were 33.5% and 34.5%, respectively.There were no significant differences between males and females for either the H. pylori positive rate or the CAG prevalence.All of these rates increased significantly with age (P < 0.01).
Table 2 shows the age distribution of CAG prevalence classified according to H. pylori antibody status.In each age group, the prevalence of CAG in H. pylori antibody-positive subjects was higher than that in H. pylori antibody-negative subjects.The prevalence of CAG increased significantly with age in both H. pylori antibody-positive subjects and H. pylori antibody-negative subjects (P < 0.01).
The left column in Table 3 shows the OR for H. pylori infection according to each lifestyle factor, while the right column shows the OR for CAG according to H. pylori infection and each lifestyle factor, after adjustment for gender and age.No significant association was found between any of lifestyle factors and H. pylori infection.
The OR for CAG in H. pylori antibody-positive subjects compared with H pylori antibody-negative subjects was 3.73 (95%CI: 2.59-5.36).The OR for CAG among subjects who consumed more than 10 cups of green tea per day compared with subjects who consumed less than 9 cups of green tea per day was 0.59 (95%CI: 0.42-0.86).There was no significant association between the other lifestyle factors and CAG.
The statistically significant relationship between green tea consumption and other beverages or foods was observed for coffee, soybean products, raw vegetables and salted foods.Therefore, we conducted multiple logistic analyses including these lifestyle factors, in order to examine the independent relation of green tea to H pylori infection and CAG.Even after adjustment for these factors, no significant relationship was found between green tea consumption and H. pylori infection (OR = 0.78; 95% CI: 0.56-1.10,P = 0.16).
The upper section of Table 4 shows the relationship between green tea consumption and CAG, after adjustment for gender, age, lifestyle factors significantly associated with green tea consumption, and H. pylori infection.Even after adjustment for these factors, the OR for CAG among the subjects who consumed more than 10 cups per day was decreased significantly (OR = 0.63, 95% CI: 0.43-0.93).
These subjects were further classified and analyzed according to presence of H. pylori antibodies (Table 4, lower section).Among H. pylori antibody-negative subjects, the OR for CAG decreased in the subjects with high green tea consumption, although this was not significant (OR = 0.72, 95% CI: 0.39-1.32).On the other hand, among the H. pylori antibody-positive subjects, OR for CAG decreased significantly in the subjects with high green tea consumption (OR = 0.57, 95% CI: 0.35-0.93).

DISCUSSION
The occurrence of CAG increases with age 17 23), and infection rate of H. pylori, which is an important risk factor of CAG, also increases with age 16.24).It is controversial whether an increased CAG rate is due to increased H. pylori infection with age or aging itself.For example, a few studies suggest that aging without H. pylori infection does not contribute to CAG risk 26.26), while another study indicates that even with no  Notes: H.pylori: Helicobacter pylori, CAG: Chronic atrophic gastritis a)Asked as tofu and natto (fermented soybean) b )Asked as green-yellow vegetables such as carrot , pumpkin, spinach, and tomatc c )Asked as pickled vegetables and smoked/dried fish Table 4. Adjusted odds ratio(OR) and 95% confidence interval (95% CI) for chronic atrophic gastritis in relation to green tea consumption (n = 636).
Notes: Hpylori: Helicobacter pylori CAG: Chronic atrophic gastritis a) Adjusted for gender , age, consumption of coffee, soybean products, raw vegetables, salted food, and H. pylori status.b) Adjusted for gender , age, consumption of coffee, soybean products, raw vegetables, and salted food.*P< 0 .05H. pylori infection, increased age contributes to the risk of CAG 19).
In the present study, the prevalence of CAG significantly increased with age in both H. pylori antibody-positive subjects and H. pylori antibody-negative subjects, although the prevalence of the antibody-positive subjects were higher than that of the antibody-negative subjects in each age group.These results support the possibility that not only H. pylori infection but also aging itself contributes to the risk of CAG.
In the present study, there were no significant relationship between lifestyle factors and H. pylori infection.As for CAG and lifestyles, significant inverse relationship was only observed between green tea consumption and the risk of CAG.This significant relationship remained after adjustments for gender, age, other lifestyle factors associated with green tea consumption, and H. pylori infection.These results suggest the possibility that high green tea consumption has an inhibitory effect on CAG, independent of other factors.
In our view, these results can be explained as follows.Excessive production of active oxygen promotes inflammation 38).Several hospital-based studies have indicated that active oxygen promotes superficial gastritis (SG), which is a precursor lesion to CAG, and/or CAG 39,40).Resent laboratory studies have shown that the polyphenols contained in green tea has a scavenging effect on active oxygen 41.42) and that this effect is stronger than other antioxidant substances such as vitamins C and E43).In addition, tumor necrosis factor-a (TNF-a), which is a kind of inflammatory cytokine plays an important role in the occurrence of gastritis 44, 45 ).Green tea polyphenols have been reported to inhibit this release of TNF-a 10.46).Furthermore, it has been shown that pheophytin a and b from the non-polyphenolic fraction of green tea act to suppress the activation of human polymorphonuclear neutrophils as major inflammatory cells 47).Thus, we consider that green tea may inhibit the progression from normal mucosa to SG and CAG by the anti-inflammatory effect of these green tea components.
Recently, eradication therapy has been adopted for patients infected with H. pylori in clinical settings in Japan.However, this therapy is mostly adopted in cases of peptic ulcer 48).The wide use of eradication therapy for the treatment or prevention of CAG may be difficult because of side-effects and cost-benefit considerations 49.50).In particular, in rural areas like that surveyed in the present study, inadequate medical facilities make such therapy difficult.Therefore, health education regarding dietary habits is considered to be one promising way of community-based CAG prevention.The results of the present study suggest that green tea consumption, which has no sideeffect, play a role in the prevention of CAG in Japan.
Since the present study is the first to report on the significant relation of green tea consumption to the risk of CAG, additional studies are needed to clarify this relation particularly in the prospective study framework.area.We thank Ms. M. Yamaguchi and Ms. M. Shinohara (public health nurses in this area) for their valuable help.We also thank Mr. Y.Yoshinaga for his statistical advice.

Table 1 .
Gender and age distribution according to H. pylori antibody positive rates and chronic atrophic gastritis prevalence.

Table 2 .
Age distribution of CAG prevalence according to H. pylori antibody status .
Notes: H.pylori: Helicobacterpylori CAG: Chronic atrophic gastritis #: The test for trend was caluculated by a logistic model.

Table 3 .
Gender and age adjusted odds ratio (OR) and 95% confidence interval (95% CI) for H.Pylori infection and CAG according to each factor (n=636).