Although the incidence of and mortality due to gastric cancer have declined markedly worldwide over the latter half of this century,1 it is still the second most common fatal cancer in the world.2, 3 Since the incidence of gastric cancer varies markedly among various regions4, 5 and from one generation to the next,1 it is thought that the risk of gastric cancer is determined largely by environmental factors, including dietary factors and infection with Helicobacter pylori.6 Among these dietary factors, the possible role of excess salt intake in the development of gastric cancer was first reported in 1959.7 Since then, the relationship between salt intake and gastric cancer has been investigated in ecological,8 cross-sectional epidemiological9, 10, 11, 12, 13 and in vivo experimental studies.14, 15 Seven prospective studies on this issue have been reported in the literature,16, 17, 18, 19, 20, 21, 22 but unequivocal evidence remains scarce. However, one study of a Japanese male population showed a dose-dependent association between dietary salt intake and the incidence of gastric cancer, although the significant association disappeared after adjustments were made for sampling region.22 Another study of males and females from the Netherlands revealed a weak association between the amount of dietary salt and the incidence of gastric cancer.21 Five other studies found no clear association between high salt foods and the risk of gastric cancer.16, 17, 18, 19, 20 Moreover, no study reported to date has estimated the quantitative effect of dietary salt on the risk of the occurrence of gastric cancer, while also taking into account the confounding effects of other risk factors such as H. pylori infection and atrophic gastritis. Thus, it remains unclear whether or not high salt intake interacts with other risk factors in the pathogenesis of gastric carcinogenesis.
In the present article, we report the findings of a prospective study examining the relationship between the amount of dietary salt intake and the occurrence of gastric cancer in subjects included in a study of a general Japanese sample; other risk factors, including H. pylori infection and atrophic gastritis, were simultaneously taken into account.
Material and methods
The Town of Hisayama is a suburban community adjacent to Fukuoka City, a metropolitan area in the southern region of Japan. The population of the town is approximately 7,500 and has remained stable for over 40 years. According to the 1985 census, the age and occupational distributions of the town population were almost identical to those of Japan as a whole.23 The dietary pattern of the residents was also similar to that of the participants in the National Nutrition Survey, which selected its subjects from 300 areas throughout Japan.24
In 1988, 2,742 Hisayama residents aged 40 years or older (80.1% of the total population in that age group) underwent a screening examination. After excluding 132 individuals with a history of gastrectomy or gastric cancer, 138 who did not undergo a dietary survey and 5 who died during the screening period, a total of 2,467 subjects (1,023 men, mean age: 57.3 years; 1,444 women, mean age: 58.7 years) were enrolled in the study.
This population was followed up for 14 years between December, 1988 and November, 2002 by repeated health checkups that were conducted every 1–2 years.25 Approximately 60–80% of the subjects regularly returned for their checkups, whereas 163 subjects (5.9% of the total number of subjects) moved out of town during the follow-up period. In each case involving those subjects who did not undergo regular checkups or who moved out of town, health status was checked every year by mail or telephone. In addition, a daily monitoring system was established by the study team and by local physicians and members of the Division of Health and Welfare of the town. To identify a new occurrence of gastric cancer in the cohort, we monitored radiographic and endoscopic study records and records of endoscopic biopsy of the stomach carried out at local clinics or general hospitals in and around the town. We also checked all of the annual mass screening records for cases of gastric cancer diagnosed by barium radiographic examination. Furthermore, to detect any cases of concealed gastric cancer, autopsies were performed on 496 (76.5%) of a total of 648 subjects who died during the follow-up period.
The diagnosis of all cases of gastric cancer was confirmed by histological examination of tissue specimens obtained by surgery, including gastrectomy and endoscopic mucosal resection, or autopsy. The tumors were classified as either intestinal-type or diffuse-type tumors, according to the classification of Lauren.26 The location of the tumor within the stomach was determined by combined evaluation of the clinical and histopathological records.
During the follow-up period, only 1 subject was lost to follow-up, and gastric cancer developed in 93 subjects (67 men and 26 women), including 3 (3.2%) concealed cases first diagnosed at autopsy. Among these positive cases, there were 8 subjects (8.6%) who had 2 synchronous gastric cancers, resulting in 101 lesions. Of these lesions, 77 were classified as intestinal-type tumors and the remaining 24 were diffuse-type tumors. As regards the anatomical subsite of the cancers, 23 lesions were found on the proximal one-third of the stomach, including the cardia, and the remaining 78 were on the distal two-thirds of the organ.
The dietary survey was conducted using a semiquantitative food frequency method.27 A self-administered questionnaire concerning food intake over the last year, which consisted of 70 food items, was completed by each participant prior to the initiation of the study, and was checked by experienced dieticians and nutritionists, who presented the subjects with food models of actual size in the survey. The average food intake per day was estimated based on detailed descriptions of the frequency of eating and the quantity of each food. Nutritional intake was calculated using the 4th revision of the Standard Tables of Food Composition in Japan.28 The nutritional elements were adjusted for energy intake, using the method of Willet and Stamper29 in order to assess the independent contributions.
The subjects were divided into 4 groups according to the amount of daily salt intake: namely, <10.0, 10.0–12.9, 13.0–15.9 and ≥16.0 g/day.
Risk factor measurement
To assess the independent effect of dietary salt intake on the occurrence of gastric cancer, baseline risk factors, in addition to age and sex, were used as confounding factors for the analysis. Serum IgG antibodies to H. pylori were measured by means of a quantitative enzyme immunoassay, using a commercial kit (HM-CAP; Enteric Products, Westbury, NY). The assay values were interpreted as positive and negative based on the manufacturer's instructions. Serum pepsinogen was measured by an immunoradiometric assay (PG I/II RIA BEAD, Dainabot Co., Tokyo, Japan). Then, atrophic gastritis was determined by the serum pepsinogen level, according to the criteria proposed by Miki et al.30; that is the presence of atrophic gastritis was determined by both serum pepsinogen I levels of ≤ 70 ng/ml and pepsinogen I/II ratios of ≤ 3. Information about smoking habit, alcohol intake, medical history of peptic ulcer disease and family history of cancer was obtained by means of a questionnaire administered to each subject. Diabetes was revealed by medical history, glucose levels (fasting glucose level ≥ 7.8 mmol/l or postprandial glucose level ≥ 11.1 mmol/l) or a 75-g oral glucose tolerance test (the 1998 WHO criteria), which was administered to most of the subjects aged 40–79 years,23 with plasma glucose measured by the glucose-oxidase method. The serum cholesterol level was determined enzymatically. Body mass index (kg/m2) was used as an indicator of obesity. Those subjects engaging in sports more than 3 times a week, during their leisure time, were classified as belonging to the physically active group.
Mean values of possible risk factors were adjusted for age and sex by means of the covariance method and were compared among different categories of dietary salt intake, using the linear regression model. The age- and sex-adjusted frequencies of risk factors were estimated by the direct method and compared by the Mantel-Haenszel χ2 test, using 10-year age groupings. The incidence of first-ever gastric cancer was estimated with the person–year method and compared by the Cox proportional hazards model.31 All study subjects were used as a standard population for age- and sex-adjustment. The age- and sex- or multivariate-adjusted hazard ratios (HR) and 95% confidence intervals (CI) were also estimated with the use of the Cox proportional hazards model. In the multivariate analysis, we used the stepwise method, with p < 0.1 required for entry into the model and for remaining there. In the analysis carried out according to histological type and location of cancer, each of the double cancers was stratified into the respective category.
The study protocol was approved by the Human Ethics Review Committee of Kyushu University Graduate School of Medical Sciences.
The age- and sex-adjusted mean values or frequencies of potential risk factors according to the amount of dietary salt intake are presented in Table I. The mean values of age, body mass index and total cholesterol levels, as well as the frequencies of male sex, positive H. pylori infection, atrophic gastritis, medical history of peptic ulcer, family history of cancer, diabetes, active physical activity and alcohol intake did not change in association with salt intake levels. The frequency of current smoking increased with elevating salt intake. Among the relevant dietary factors, the mean values of carbohydrate and vitamin C intake showed a positive correlation with salt intake, whereas the intake of vitamin B2 and fiber significantly decreased with elevating salt intake.
Table I. Age- and Sex-Adjusted Mean Values or Frequencies of Potential Risk Factors According to The Amount of Dietary Salt Intake, The Hisayama Study, 1988
The age- and sex-adjusted incidence rates of gastric cancer, according to the amount of dietary salt intake, are shown in the Figure 1. The incidence of gastric cancer was significantly higher among those subjects whose dietary intake of salt was ≥ 10 g/day, as compared to the subjects whose dietary intake of salt amounted to < 10 g/day.
The age- and sex-adjusted HRs among subjects with a dietary intake of salt of ≥ 10 g/day were 2-fold higher than those of subjects with a dietary intake of salt of < 10 g/day (Table II). This association did not substantially change, even after adjustments were made for other confounding factors, such as H. pylori infection, atrophic gastritis, medical history of peptic ulcer, family history of cancer, body mass index, diabetes, total cholesterol, physical activity, alcohol intake, smoking habit and dietary factors (intake of total energy, protein, carbohydrate, vitamin B1, vitamin B2, vitamin C and dietary fiber).
Table II. Age- and Sex- or Multivariate-Adjusted HRs and 95% CIs for The Development of Gastric Cancer by Dietary Salt Intake Levels, The Hisayama Study, 1988–2002
Adjusted for age, sex, Helicobacter pylori infection, atrophic gastritis, medical history of peptic ulcer, family history of cancer, body mass index, diabetes mellitus, total cholesterol, physical activity, alcohol intake, smoking habits and dietary factors (intake of total energy, total protein, carbohydrate, vitamin B1, vitamin B2, vitamin C and dietary fiber), using the stepwise method of Cox's proportional hazards model.
We then analyzed the risk of gastric cancer with respect to high salt intake (≥ 10 g/day) relative to low salt intake (< 10 g/day), according to the status of atrophic gastritis and H. pylori infection at baseline (Table III). The age- and sex- adjusted HR of gastric cancer significantly increased with high salt intake in the subjects who had atrophic gastritis or H. pylori infection, whereas this effect was not significant in those who did not have atrophic gastritis or H. pylori infection. We further estimated the age- and sex-adjusted HRs of gastric cancer for high salt intake in 4 strata, based on the presence or absence of atrophic gastritis and H. pylori infection. As shown in Table IV, a significant association between high salt intake and gastric cancer was observed in the stratum with atrophic gastritis and H. pylori infection (HR, 2.87; 95% CI, 1.14–7.24), but was not present in the strata of atrophic gastritis alone or H. pylori infection alone. We could not estimate the HR of gastric cancer in the stratum without atrophic gastritis and H. pylori infection, since no one case developed cancer among subjects with low salt intake in this stratum.
Table III. Age- and Sex-Adjusted HRs and 95% CIs of Gastric Cancer by Dietary Salt Intake According to Atrophic Gastritis, Helicobacter Pylori Infection, Cancer Location and Cancer Histology, The Hisayama Study, 1988–2002
As regards the location of gastric cancer, the risk of cancer located in the distal two-thirds of the stomach was significantly higher in subjects with high salt intake than in those with low salt intake, whereas such an association was not observed in cases of cancer located in the proximal one-third of the stomach (Table III). When cancer was classified according to histological type, the magnitude of the association with high salt intake was significantly high for the intestinal-type tumors, but not for the diffuse-type tumors.
In a prospective study of a defined Japanese population, we revealed a positive association between the consumption of dietary salt and the incidence of gastric cancer, whereby the risk of cancer significantly increased at salt intake levels of 10 g/day and more. This association remained substantially unchanged, even after adjustments were made for other confounding factors, such as age, sex, H. pylori infection, atrophic gastritis, medical history of peptic ulcer, family history of cancer, body mass index, diabetes, total cholesterol, physical activity, alcohol intake, smoking habit and other dietary factors. In the stratified analysis, a significant salt–cancer association was observed only in subjects who had both H. pylori infection and atrophic gastritis. To our knowledge, this is the first prospective cohort study of the association between the amount of salt intake and the incidence of gastric cancer, while taking into account the interaction between dietary salt intake and the status of atrophic gastritis, as well as that of H. pylori infection.
Several mechanisms of gastric carcinogenesis induced by a high salt diet have been discussed in the literature, although there has been no conclusive evidence reported to date. High dietary salt intake is considered to alter the viscosity of the protective mucous barrier and facilitate exposure to carcinogenic agents such as nitrates.14 Moreover, intragastric high salt concentrations are known to cause mucosal damage and inflammation.32 Persistent inflammatory changes in the stomach may promote temporary cell proliferation and increase the rate of endogenous mutations,33, 34 eventually resulting in the development of cancer. An alternative explanation for the effect of high salt intake on the gastric carcinogenesis is that a highly salty diet may facilitate the colonization of H. pylori,35 which is recognized as a definite risk factor for gastric cancer. However, in our cohort, the prevalence of H. pylori infection did not vary according to the amount of dietary salt intake at baseline (Table I), suggesting that high salt intake per se does not promote H. pylori infection.
Association with gastric atrophy and H. pylori infection
In our cohort, the effect of high salt intake on gastric carcinogenesis was strong in subjects who had both atrophic gastritis and H. pylori infection. The synergistic promoting effects of a high salt diet and H. pylori infection on gastric carcinogenesis were observed in an experimental study with gerbils.36 It is shown that H. pylori infection exacerbates stomach mucosal damage because of a salty diet in mice,37 and the titers of anti-H. pylori antibodies are relatively higher in infected animals with a high salt diet than in those with a normal diet,36 suggesting that the high salt diet affects humoral responses after H. pylori infection, and in some way contributes to a carcinogenesis-enhancing effect. On the other hand, atrophic gastritis is thought to be a precancerous lesion of gastric cancer.38 Thus, the synergistic promoting effects of a high salt diet and H. pylori infection may appear strongly under the conditions of gastric atrophy.
It is generally accepted that gastric adenocarcinoma, particularly the intestinal type, arises through a multistep process originating with chronic gastritis, and progresses through stages of atrophy, intestinal metaplasia and dysplasia, which eventually results in carcinoma.39 Based on the results of previous reports, as well as on the present findings, it is more likely that a high intake of salt is involved primarily in the latter stages of multi-step gastric carcinogenesis and thereby promotes gastric carcinogenicity because atrophic gastritis is one of the morphological hallmarks of these stages.40
Histological type of gastric cancer
In our subjects, the promotional effect of dietary salt intake on the development of intestinal-type cancer was stronger than that on the development of diffuse-type cancer. It is generally recognized that intestinal-type cancer arises via multistep gastric carcinogenesis,39 and a high salt diet is considered to be involved in multistep carcinogenesis, inducing precursor lesions of gastric cancer such as intestinal metaplasia in humans.41 On the other hand, diffuse-type gastric cancer is considered to be less associated with environmental factors, and is postulated to be more genetically predetermined.42 The results of earlier reports thus confirm our finding that high salt intake was clearly associated only with intestinal-type gastric cancer.
Anatomical subsite of gastric cancer
As regards the anatomical subsites of gastric cancer, high dietary salt intake was clearly associated in the present sample with cancer of the distal two-thirds of the stomach, but was not associated with cancer of the proximal one-third of the stomach. Risk factors for gastric cancer are thought to differ according to anatomical subsite.43 It is well known that the above-mentioned multistep type of carcinogenesis is associated with distal gastric cancer, but not with cardial cancer.44 Thus, in our subjects, the promotional effect of a high salt diet is thought to be greater in cases of cancer of the distal two-thirds of the stomach than in cases occurring in the proximal one-third of the stomach, including those in the cardia.
The potential limitations of this study should also be noted. First, for the screening of gastric cancer, we did not perform endoscopy on all subjects. It is undeniable that there were presymptomatic gastric cancer patients at baseline. However, when we analyzed the data after excluding those subjects who developed gastric cancer in the first 2 years of the follow-up period, the results were similar to those of the present study (data not shown). Moreover, the misclassification of any gastric cancer patients would be expected to reduce the effect of high salt on gastric carcinogenesis. Thus, the significant association observed between high salt intake and gastric cancer may have been conservative in our study, and the true association may be even greater than that reflected in our findings. Second, the information regarding nutrient intake, including dietary salt intake, was derived from a semiquantitative food frequency questionnaire that requested information from subjects regarding their average intake over the previous year. The limitations of this method are well known,45 and random measurement errors are more likely to have contributed to a bias towards a finding of no effect. Thus, the estimates of effect that we have reported here are most likely conservative. Third, our results might be biased, because 138 participants did not undergo a baseline nutritional examination and were excluded from our cohort. The mean age of these people was higher than that of the subjects enrolled in the final study. However, among our subjects, the amount of dietary salt intake did not vary by age, suggesting that the amount of dietary salt intake did not differ greatly between those excluded from the study and the participants in the study. Furthermore, the incidence of gastric cancer among nonparticipants was similar to that among the participants (data not shown). Thus, it appears that the exclusion of these nonparticipants did not bias the salt–cancer association observed in this study.
In our study, we observed a significantly increased risk of gastric cancer among subjects with moderate to high levels of dietary salt intake, especially among those who had both atrophic gastritis and H. pylori infection. Although the mechanism by which salt per se is involved in gastric carcinogenesis remains unclear, the restriction of salt intake is thought to be beneficial for the prevention of gastric cancer, especially in countries such as Japan, where dietary salt intake and the risk of gastric cancer are both considerable.