Original Articles
Copyright ©The Author(s) 1998. Published by Baishideng Publishing Group Inc. All rights reserved.
World J Gastroenterol. Dec 15, 1998; 4(6): 516-518
Published online Dec 15, 1998. doi: 10.3748/wjg.v4.i6.516
Diet and gastric cancer: a case-control study in Fujian Province, China
Wei-Min Ye, Ying-Nan Yi, Ren-Xia Luo, Tian-Shu Zhou, Ru-Tao Lin, Gui-Dong Chen
Wei-Min Ye, Ying-Nan Yi, Ru-Tao Lin, Department of Epidemiology, Fujian Medical University, Fuzhou 350004, Fujian Province, China
Ren-Xia Luo, PLA Fuzhou College of Medicine, Fuzhou 350001, Fujian Province, China
Tian-Shu Zhou, Gui-Dong Chen, Hygiene and Anti-epidemic Station of Fujian Province, Fuzhou 350001, Fujian Province, Chia
Wei-Min Ye, male, born on 1986-12-06 in Youxi, Fujian, graduated and earned a master degree from Fujian Medical University in 1991, now lecturer of epidemiology, majoring cancer epidemiology, having 10 papers published.
Author contributions: All authors contributed equally to the work.
Correspondence to: Dr. Wei-Min Ye, Department of Epidemiology, Fujian Medical University, 88 Jiaotong Road. Fuzhou 350004, Fujian Province,China
Telephone: +86-591-3357231 Fax: +86-591-3351345
Received: June 6, 1998
Revised: September 10, 1998
Accepted: November 24, 1998
Published online: December 15, 1998


AIM: To explore the relationship between consumption of fish sauce, other dietary factors, living habits and the risk of gastric cancer.

METHODS: From May 1994 to July 1995, a population-based 1:2 case-control study was in Carried out in high-risk areas of gastric cancer, Changle and Fuqing cities, Fujian Province. Totally 272 cases and 544 age, gender-matched controls were included. Risk state analyses were made by ASRS package.

RESULTS: Risk state single-factor analysis indicated that gastric cancer risk rose with high intake of fish sauce (OR = 2.57), salted vegetables (OR = 1.41), salted/fried fish and small shrimps (OR = 1.57), low consumption of fresh vegetables (OR = 1.95), fresh citrus fruits (OR = 1.41), other fresh fruits (OR = 1.31), green tea (OR = 1.72), exposure to moldy foods (OR = 2.32), irregular dinners (OR = 5.47) and familial history of malignancy (OR = 3.27). No significant relationship was observed between smoking, drinking, salt intake, use of refrigerator and gastric cancer risk. The results of risk state conditional Logistic regression showed that fish sauce, salted dried fish and small shrimps, irregular dinners, familial history of malignancy were included in the best risk set. The summary ARS for the four factors was 75.49%.

CONCLUSION: High intake of fish sauce, salted foods, moldy foods, irregular dinners and familial history of malignancy were possible risk factors for gastric cancer, whereas fresh vegetables and fruits. And green tea might have protective effects for gastric cancer.

Key Words: stomach neoplasms/etiology, living habits, food habits, risk facto


Changle and Fuqing cities are located in the southeastern part of Fujian Province, China with a high incidence of gastric cancer. However, the causes of gastric cancer still remain unclear. Previous studies indicate that environmental factors may play an important role in the carcinogenesis of gastric cancer, among which, dietary risk factors for gastric cancer were most extensively investigated. Our hypothesis is that the high incidence of gastric cancer may be attributed, to some extent, to some unique dietary habits. Recently a statistically significant relationship between fish sauce consumption, a condiment commonly used by local residents, and the mortality rates from gastric cancer was observed by our ecological study[1]. The mutagenicity of fish sauce was also reported by experimental studies[2]. N-nitrosamines can also be detected in fish sauce[3]. In order to explore further the relationship between consumption of fish sauce and gastric cancer, a population-based 1:2 matched case-control study was carried out from May 1994 to July 1995.

Selection of cases and controls

This study was conducted in Changle and Fuqing cities with populations around 600 thousands and 1 million, respectively. All cases histologically confirmed or diagnosed by operation from January 1993 to July 1995 were collected from cancer registry and a quick-reporting system from hospitals. Each case was matched by two randomly selected controls who resided in the same village as index case, with same gender, nationality and age (± 3 years). Those who have ever been diagnosed having gastric diseases within the past 3 years were not eligible as controls. Study subjects must have resided in the two cities for more than 20 years, and can answer questions clearly.


Face to face interviews were made by specially trained interviewers with a structured questionnaire. The items of questionnaire included demographic and socio-economic factors, occupational and medical histories, family occurrence of cancer, use of alcohol and tobacco, and dietary habits. As for dietary habits, we emphasized on the exposure 20 years before. Diet was assessed with a food frequency questionnaire. The subjects were interviewed about the usual frequency of consumption of certain foodstuffs, supplemented with questions about the actual amount consumed per time unit. Then total amount per year was estimated accordingly. The subjects within a matched pair were interviewed by one interviewer.

Statistical analysis

Data werre handled by Epi-info. The statistical analyses, including univariate analysis and conditional logistic analysis were done using ASRS software[4]. In univariate analysis, the best cut-points for exposure levels were searched by automatic or forced adjustment, combination, on the criteria of CPDS ans AIC. Those factors, which showed association with the risk of gastric cancer in univariate analysis, were further investigated by conditional logistic analysis of risk states to establish a main-effect model.


Totally 272 pairs were investigated, among which 157 were from Changle City, 115 from Fuqing City; 233 were male and 39 female. The age range for cases was 30-78 years, averageing 58.67 years. No obvious difference in marital status, occupation and education level was observed between cases and controls.

The results of univariate analysis are shown in Table 1. As for dietary habits, consumption of fish sauce, salted vegetables, salted fermented sea products and moldy foods may increase the risk of gastric cancer. However, fresh vegetables, fruits and green tea may have protective effects against gastric cancer. Additionally, irregular dinner and family occurrence of cancer are also risk factors for gastric cancer. No association between the use of alcohol and tobacco, the amount of salt intake, the use of refrigerator and the risk of gastric cancer was observed. To select a possible best subset of risk factors for gastric cancer, conditional logistic analyses of risk states were also made. The results showed that the best subset of risk factors included fish sauce, irregular dinner, salted fermented sea products and family occurrence of cancer (Table 2). The summary attributable risk for these four risk factors is 75.49%, indicating that these four factors may play an important role in the carcinogenesis of gastric cancer.

Table 1 Results of univariate analysis on the relationship between fish sauce, other factors and the risk of gastric cancer.
Index of smokinga> 101182411
≤ 101543031.040.95-1.13
Hard distilled spirit ≤ 20 kg/y2655341
> 20 kg/y7101.410.63-3.14
Soft distilled spirit> 25 kg/y13291
≤ 25 kg/y2595151.120.86-1.47
Wine ≤ 30 kg/y2454941
> 30 kg/y27501.090.89-1.33
Beer> 50 bottle/y421061
≤50 bottle/y2304381.330.93-1.88
Green tea> 0.75 kg/y471441
≤ 0.75 kg/y2254001.721.26-2.36e
Fish sauce< 0.4 kg/m1984751
≥ 0.4 kg/m74692.571.89-3.50e
Saltb ≤ 0.25 kg/m1593471
> 0.25 kg/m1131971.250.96-1.63
Moldy foodsNo1884561
Irregular dinners< 3 times/w1144341
≥ 3 times/w1581105.474.22-7.09e
Use of refrigeratorYes35781
Salted vegetables< 2 kg/y1573581
≥ 2 kg/y1151861.411.09-1.83d
Salted fermented sae foods< 1.5 kg/y1443471
≥ 1.5 kg/y1281971.571.21-2.02e
Citrus fruitsc> 2.5 kg/y551431
≤ 2.5 kg/y2174011.411.03-1.92d
Other fruits> 2.5 kg/y1663661
≤ 2.5 kg/y1061781.311.01-1.71d
Fresh meat, fish, egg, poultry> 25 kg/y1052381
≤ 25 kg/y1673061.240.95-1.61
Fresh vegetables> 25 kg/y2124751
≤ 25 kg/y60691.951.41-2.70e
Family occurrence of cancerNo1654541
Table 2 The results of conditional logistic analysis of risk states.
FactorsRegression coefficientStandardized regression coefficientAdjusted attributable riska
Fish sauce1.083.4917.81%c
Irregular dinners1.858.2648.93%c
Salted sea foods0.542.1119.69%b
Familial history of malignancy1.195.4527.41%c

Fish sauce is one kind of condiments consumed daily by local residents. It is produced by long-term fermentation from several kinds of sea frsh. Due to the proteins with amino in the fishes, salted fermented fish products may contain a large amount of important precursors of N-nitroso compounds-amines[5]. These precursors may react with nitrite in gastric juice to form N-nitroso compounds internally. Deng et al[2] reported that abstracts of fish sauce from Changle have carcinogenicity and mutagenicity after nitration. The amount of N-nitro compounds increased greatly after nitration, and genotoxins can be detected[3]. Our ecological study indicated that there was a statistically significant relationship between frsh sauce consumption and mortality rates from gastric cancer among 14 counties in Fujian Province[1]. The results of this study further supported the point that fish sauce consumption may be an important cause for the high incidence of gastric cancer.

Several case-control studies have indicated that long-term use of refrigerators may decrease the risk of gastric cancer[6-8]. However, in our study, refrigerators were not commonly used by local residents. Even among those users, the history of refrigerator use is very shout. Therefore, it is impossible to evaluate the role of refrigerator use in the etiology of gastric cancer. However, the deficiency of refrigerator indicated that the consumption of salted foods was very common. Especially in the study area which is located on seaboard, the consumption of salted or fermented sea foodstuffs is very common. In our study, consumption of salted foodstuffs, especially of salted fermented sea products was found to increase the risk of gastric cancer. This is in accordance with the results of Buiatti et al[7]. Additionally, our study indicated that irregular dinner may be one of etiological factors for gastric cancer. This finding supported the results of our previous ecoloical study[1]. Irregular dinner may cause injuries of gastric mucosa and promote the effects of carcinogens. A prospective study showed that those with familial history of malignancy, especially gastric cancer, havea higher risk of gastric cancer[9]. This is in agreement with our study results. In our study, fresh vegetables showed a protective effect against gastric cancer. This is also true for citrus fruits and other fruits. Their protective effects may be attributed to the vitamin C, which may interrupt the internal formation of N-nitroso compounds. Some studies reported that only consumption of raw vegetables had a protective effect[6,7]. However, in our study area, residents do not have such a habit. There were conflicting views on the relationship between the use of alcohol and tobacco and the risk of gastric cancer[6,7,9,10]. In our study, no association was observed between the use of alcohol, tobacco and the gastric cancer risk. No interaction between these two factors was found either.


Project supported by “8.5” national major project, No. 95-914-01-10.

1.  Ye WM, Ti YN, Lin RT, Zhou TS, Cai L, Chen CG et al. A study on relationship between fish sauce consumption and mortality from gastric cancer in Fujian piovince. J Fujian Med Coll. 1994;28:402-405.  [PubMed]  [DOI]
2.  Deng DJ, Zhang RF, Chen Y, Chen CS, Jin S, Zhu SX. Mutagenicity and carci-nogenicity of fish sauce from a county with the high risk for gastric cancer in China. Carcinogenesis, Teratogenesis and Mutagenests. 1991;3:13-17.  [PubMed]  [DOI]
3.  Chen CS, Pignatelli B, Malaveille C, Bouvier G, Shuker D, Hautefeuille A et al. Levels of direct-acting mutagens, total N-nitrosocompounds in nitrosated fermented fish products, consumed in a high-rish area for gastric cancer in southern China. Mutat Res. 1992;265:211-221.  [PubMed]  [DOI]
4.  Liu YY. Risk state analysis and it's application in biomedicine (I). Beijing: Science Press. 1990;83-123.  [PubMed]  [DOI]
5.  Zhang RF, Deng DJ, Chen Y, Chen CS, Fan ZF. Analysis on precursors of N-nitrosamins in fish sauce from a high-risk area of gastric cancer. Chin J Cancer. 1993;12:395-398.  [PubMed]  [DOI]
6.  Boeing H, Frentzel-Beyme R, Berger M, Berndt V, Göres W, Körner M, Lohmeier R, Menarcher A, Männl HF, Meinhardt M. Case-control study on stomach cancer in Germany. Int J Cancer. 1991;47:858-864.  [PubMed]  [DOI]
7.  Buiatti E, Palli D, Decarli A, Amadori D, Avellini C, Bianchi S, Biserni R, Cipriani F, Cocco P, Giacosa A. A case-control study of gastric cancer and diet in Italy. Int J Cancer. 1989;44:611-616.  [PubMed]  [DOI]
8.  Graham S, Haughey B, Marshall J, Brasure J, Zielezny M, Freudenheim J, West D, Nolan J, Wilkinson G. Diet in the epidemiology of gastric cancer. Nutr Cancer. 1990;13:19-34.  [PubMed]  [DOI]
9.  Kato I, Tominaga S, Matsumoto K. A prospective study of stomach cancer among a rural Japanese population: a 6-year survey. Jpn J Cancer Res. 1992;83:568-575.  [PubMed]  [DOI]
10.  Hu JF, Zhang SF, Jia EM, Wang QQ, Liu SD, Liu YY, Wu YP, Cheng YT. Diet and cancer of the stomach: a case-control study in China. Int J Cancer. 1988;41:331-335.  [PubMed]  [DOI]