Serum Pepsinogen Values and Helicobacter pylori Status among Control Subjects of a Nested Case-Control Study in the JACC study

BACKGROUD: Helicobacter pylori infection and serum pepsinogen values are strongly related with stomach cancer. The aim of this study was to know what were these factors among general population. METHODS: Subjects were randomly selected 633 control subjects in a nested case-control study for risk of stomach cancer. Most of them were from rural areas of Japan. Using frozen sera, pepsinogen I (PG I) and II (PG II) values and H. pylori antibody were measured. Those with PG I less than 50 ng/mL and the ratio of PG I to PG II (PG I/II) was less than 2.0 were defined as severe, those with PG I less than 70 ng/ml and PG I/II less than 3.0 were defined as mild and the other subjects were defined as no serological atrophy. RESULTS: About 70% of the subjects were H. pylori seropositive and the seroprevalence did no depend on age or sex. Percentages of those with severe serological atrophy increased with age from 10% in those aged 40-49 years to 38% in 70 and more, and percentages of those with mild serological atrophy were about 30% independent of age. CONCLUSIONS: The subjects, who were expected to represent populations of rural area of Japan, had high prevalence of both H. pylori infection and serological atrophy of gastric mucosa. These facts should be considered in discussing results of the nested case-control study.

assayed in 1999 by trained staff at a single laboratory who were blinded to the case/control status of the individuals. Serum levels of H. pylori antibody and PG I and PG II were measured using J-HM-CAP (Kyowa Medex Co. Ltd., Tokyo) and RIAbeads Pepsinogen I and II kits (Dainabot Co. Ltd., Tokyo), respectively. H. pylori status was defined according to the manufacturer's instructions. When the titer of H. pylori antibody was less than 2.3, it was defined as negative, and defined as positive when not less than 2.3. Using serum pepsinogen values, the subjects were classified into three levels of inflammation/atrophy of gastric mucosa (expressed as serological atrophy below): severe, mild and no. Those with PG I less than 50 ng/mL and the ratio of PG I to PG II (PG I/II) was less than 2.0 were defined as severe, those with PG I less than 70 ng/mL and PG I/II less than 3.0 were defined as mild and the other subjects were defined as no serological atrophy. The subjects were divided into 6 groups by levels of gastric mucosal serological atrophy and H. pylori status: severe/positive, severe/negative, mild/positive, mild/negative, no/positive and no/negative. Then age (40-49, 50-59, 60-69, 70-79 years) and sex distribution of H. pylori infection and serological atrophy determined by serum pepsinogen values was analyzed. In order to confirm difference between rural and urban area, prevalence of H. pylori infection and serological atrophy was compared between the subjects from a city with a population of a million people and the other subjects.
Relationships between H. pylori infection or severe serological atrophy and factors in the baseline questionnaire were evaluated using logistic regression models with adjustment for age and sex. The evaluated factors were father's and/or mother's history of stomach cancer, history of mass-survey for stomach cancer, pregnancy, intake of salted foods, dried fish, drinking and smoking habits.
Age and sex distribution and H. pylori serology of the subjects are shown in Figure 1. H. pylori prevalence did not depend on age or sex. About 70% of the subjects were H. pylori seropositive. Male subjects in their 40s showed exceptional 53% of seroprevalence. Percentages of those with severe serological atrophy increased with age from about 10% in those aged 40-49 years to about 38% in 70 and more in both male and female subjects ( Figure 2). Percentages of those with mild serological atrophy were about 30% and did not depend on age or sex. In those aged 40-49 years, percentages of those without serological atrophy was about 60%, whereas that was less than 50% in those over 50 years of age. Figure 3 shows age distribution of the 6 groups determined by serological atrophy and H. pylori status. As similar results were obtained between both sexs of subjects, only total results are shown. Percentages of the no/negative decreased with age from 30% to 15%, while those of the severe/positive increased from 7% to 27%. Percentages of the mild/negative and the mild/positive groups did not change over age. The severe/negative Kikuchi S, et al. Helicobacter pylori infection is one of the causes of stomach cancer. Many epidemiologic studies so far have argued relationship between H. pylori and risk of stomach cancer. 1,2 However, several negative results have been obtained even in East Asian countries, where both mortality of stomach cancer and prevalence of H. pylori are high. 3 One explanation for the negative results is very high prevalence of H. pylori in the general population or control subjects. 3 Another explanation is spontaneous eradication of H. pylori and consequent seroreversion between the time of diagnosis and when H. pylori plays a role in gastric carcinogenesis. 4 Serum pepsinogen is a marker of both inflammation and atrophy of gastric mucosa. 5 It is also related to the risk of stomach cancer, and either a low value of pepsinogen I (PG I), 6,7 a high value of pepsinogen II (PG II), or a low value of PG I to PG II ratio indicates an increased risk of stomach cancer. 8 The Japan Collaborative Cohort Study (JACC Study) for Evaluation of Cancer Risk sponsored by the Ministry of Education, Science, Sports and Culture of Japan (Monbusho) is a cohort study initiated between 1988 and 1990 when apparently healthy residents aged 40-79 years were enrolled as a basic cohort population from 45 areas throughout Japan. 9 The subjects answered a questionnaire including lifestyle factors and past histories as baseline information of the study. A peripheral blood sample was collected from some of the subjects. A total of 39,242 subjects provided their sera when they attended general health check programs. The main aim of the JACC Study is to investigate risk factors of cancers. A nested case-control study was conducted in the JACC Study to evaluate the relationships between serum markers and risk of stomach cancer, and serum pepsinogen and H. pylori antibody was measured using sera from 164 death and 161 incidence cases of stomach cancer and 635 control subjects. 10,11 The current study was conducted to know what was the H. pylori and serum pepsinogen status, what factors were related to status of the serum markers among the control subjects, and what was the effect of them on the case-control study.
Subjects were selected from those who provided their sera for the JACC Study in 1988-1990. 9 They were 635 control subjects who were matched for 164 death and 161 incidence cases of stomach cancer in the JACC Study. Matched for age ( 6 year), sex and study area with each case subject, the control subjects were randomly selected from those who were alive without diagnosis or history of stomach cancer at the time of death or diagnosis of the case subject. Two subjects, whose pepsinogen values were not measured because of lack in sera, were excluded from the study and 633 subjects were enrolled.
Serum PG I, PG II values, and H. pylori antibody were measured using the sera from the cases and control subjects. Sera were separated from the blood samples at laboratories in each study area and stored at -80 until analysis. All the samples were No factors were related with H. pylori status ( Table 2). On the other hand, history of mass survey for stomach cancer and increased intake of salted foods compared with when young were positively, and intake of dried fish were negatively related with severe serological atrophy.
increased with age and were frequent among those with severe serological atrophy in those aged 70 years.
Subjects from a large city showed lower seroprevalence of H. pylori (54%) than those from the other areas (72%), but no difference was observed in frequency of serological atrophy (Table 1).

Figure 1.
Helicobacter pylori seroprevalence with respect to age and sex.   difference. In this study, H. pylori antibody was measured using a domestic kit, which has shown better accuracy than imported kits. 13,14 Our previous study used an imported kit "Pilika-Plate G Helicobacter" which is the same kit as "IgG-GAP." However, it seems to be impossible to explain the difference in seroprevalence only by the used kits, because differences in sensitivity and specificity between the two kinds of kits are at most 10% and 6%, respectively. 13,14 Another explanation is that prevalence of H. pylori is different between in rural areas and urban areas. 15 Actually, the subjects from a large city with a population of a million people showed similar seroprevalence to our previous study. As most of the subjects were from rural area, seroprevalence in the JACC Study may be high compared with seroprevalence in 1988-90 in urban areas including our previous study.
This cross sectional serological study has shown prevalence of H. pylori, mild and severe serological atrophy among control subjects of a nested case-control study, who were not diagnosed as stomach cancer until death or diagnosis of their matched case subjects.
Our previous study, collecting sera from workers in the Tokyo Metropolitan Area, Japan in 1989, reported that seroprevalence was 40% in those aged 39-44 years, 49% in 45-54, 56% in 55-64 and 56 % in 65 and older. 12 Even though sera were collected almost concurrently, there were differences in seroprevalence between this study and our previous study whose subjects were from urban area. Several explanations seem to be possible for the cal atrophy. Intake of both salted foods and grilled dried fish is known to elevate risk for stomach cancer. 25,26 It seems to be difficult to explain the inverse association of these factors. Only possible explanation is as follows. A part of those with severe gastric mucosal atrophy suffered some discomfort from intake of grilled dried fish, which did not accelerate atrophy. On the other hand, the seven subjects who answered that intake of salted foods increased than when young really had highly salted frequently, and atrophy of gastric mucosa was accelerated in their stomach. They did not feel any discomfort from intake of highly salted foods even though they have severe gastric mucosal atrophy. However, the latter relationship may be unstable compared with the former one because the number of subjects who increased salt intake was too small.
The subjects of the current study are control subjects of a nested case-control study who were not diagnosed as stomach cancer during the follow-up period. They were born before 1950 and were randomly selected from almost all study areas of the cohort study. Although there might be a bias that the subjects of the current study are more interested in their health conditions than the other residents of the areas because they gave their sera mainly at the time of general health check programs, they are expected to represent the population of their resident areas. It is expected that the population of the study areas, who were born before 1950, have high prevalence of both H. pylori infection and serological atrophy of gastric mucosa. These facts should be considered in discussing results of the nested case-control study using the subjects.