Skip Navigation


American Journal of Epidemiology Advance Access originally published online on November 28, 2007
American Journal of Epidemiology 2008 167(3):295-304; doi:10.1093/aje/kwm294
This Article
Right arrow Abstract Freely available
Right arrow FREE Full Text (PDF) Freely available
Right arrow All Versions of this Article:
167/3/295    most recent
kwm294v1
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Add to My Personal Archive
Right arrow Download to citation manager
Right arrow Search for citing articles in:
ISI Web of Science (3)
Right arrowRequest Permissions
Right arrow Disclaimer
Google Scholar
Right arrow Articles by Campbell, P. T.
Right arrow Articles by Kreiger, N.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Campbell, P. T.
Right arrow Articles by Kreiger, N.
Social Bookmarking
Add to CiteULike   Add to Connotea   Add to Del.icio.us  
What's this?

American Journal of Epidemiology © The Author 2007. Published by the Johns Hopkins Bloomberg School of Public Health. All rights reserved. For permissions, please e-mail: journals.permissions@oxfordjournals.org.

ORIGINAL CONTRIBUTIONS

Dietary Patterns and Risk of Incident Gastric Adenocarcinoma

Peter T. Campbell1,2, Margaret Sloan1 and Nancy Kreiger1,3,4

1 Division of Preventive Oncology, Cancer Care Ontario, Toronto, Ontario, Canada
2 Cancer Prevention Program, Fred Hutchinson Cancer Research Center, Seattle, WA
3 Department of Public Health Sciences, Faculty of Medicine, University of Toronto, Toronto, Ontario, Canada
4 Department of Nutritional Sciences, Faculty of Medicine, University of Toronto, Toronto, Ontario, Canada

Correspondence to Dr. Nancy Kreiger, Division of Preventive Oncology, Cancer Care Ontario, 620 University Avenue, Toronto, Ontario, Canada M5G 2L7 (e-mail: nancy.kreiger{at}cancercare.on.ca).

Received for publication April 9, 2007. Accepted for publication September 17, 2007.


   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 References
 
Few studies have assessed the relation between dietary patterns and gastric adenocarcinoma risk, yet this approach has advantages over single-nutrient analysis, including the ability to reflect eating patterns in populations and ease of implementing dietary recommendations. The authors evaluated associations between dietary patterns, a food index score, and incident gastric cancer risk in a Canadian study of 1,169 cases and 2,332 controls (1995–1997). Dietary patterns were assessed via factor analysis applied to a food frequency questionnaire. A food index score was derived based on risk factors for gastric cancer. A multivariable-adjusted prudent dietary pattern was associated with decreased risk of gastric cancer in women (odds ratio (OR) = 0.58, 95% confidence interval (CI): 0.37, 0.92); a Western dietary pattern was associated with increased risk in women (OR = 1.86, 95% CI: 1.20, 2.89) and men (OR = 1.44, 95% CI: 1.03, 2.02). The food index score was associated with decreased risk among women (OR = 0.40, 95% CI: 0.27, 0.59) and men (OR = 0.63, 95% CI: 0.45, 0.88). Dietary patterns especially characterized by Western features (soft drinks, processed meats, refined grains, and sugars) were associated with increased risk of gastric adenocarcinoma, whereas dietary patterns characterized by increased consumption of fruits, vegetables, and fish were associated with lessened risk.

diet; nutrition surveys; risk factors; stomach neoplasms

Abbreviations: CI, confidence interval; FFQ, food frequency questionnaire; OR, odds ratio


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 References
 
Dietary factors are implicated in carcinogenesis of the stomach, although the contributions of individual foods and nutrients remain unclear. Previous observational work has been inconsistent but has shown that fruits and vegetables generally protect against gastric cancer risk, with stronger support for this association coming from case-control studies than from cohort studies, whereas salt, processed meats and foods, and sweets have usually been linked with increased risk of the disease (1–9). The cancer-preventive effect of fruits and vegetables has been often attributed to their antioxidant content (2). Randomized controlled trials of antioxidant supplements, however, have shown mixed results. A randomized controlled trial from China demonstrated that supplementation with selenium, β-carotene, and {alpha}-tocopherol reduced the risk of gastric cancer (10). This finding was in agreement with work carried out in Colombia, where an ascorbic acid and β-carotene intervention reduced disease progression among patients with gastric precursor lesions (11). Conversely, randomized controlled trials from Finland and the United States suggested no effect of {alpha}-tocopherol, β-carotene, vitamins C and E, and multivitamin supplementation on gastric cancer incidence (12) or mortality (13).

It is difficult to identify specific influences of individual food items or nutrients on disease risk in observational studies. These complications are probably due to high correlations among foods, food groups, and nutrients and are potentially aggravated by the presence of nutrient-nutrient interactions (14). Dietary pattern analysis circumvents some of these complications by reducing consumption of multiple food items to a few distinct dietary patterns (15–17). Dietary patterns have recently been associated with overall mortality, several cancers, type 2 diabetes, and cardiovascular disease (16, 18–25). The majority of case-control and cohort studies of risk factors for gastric cancer, including studies of dietary patterns, have been conducted among populations outside of North America, in populations that have considerably different indigenous foods, food preparation methods, and gastric cancer rates. In data from a nationwide case-control study conducted in Canada, we investigated associations between dietary patterns, a food index score, and gastric cancer risk.


   MATERIALS AND METHODS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
RESULTS
 DISCUSSION
 References
 
The National Enhanced Cancer Surveillance System was established in the mid-1990s by Health Canada and eight provincial cancer agencies, as described previously (26). Cases of gastric adenocarcinoma were identified and histologically verified from pathologists' records at each cancer registry. Patients were eligible for participation if they were diagnosed with gastric adenocarcinoma (International Classification of Diseases, Ninth Revision (27), codes 151.0–151.9) between the beginning of 1995 and the end of 1997 and were aged 20–74 years at diagnosis. Nonadenocarcinoma gastric cancer cases (e.g., lymphoma and carcinoid tumors) were excluded. Physician consent was required to contact cases. In subsite analyses, cardia cancers were defined as "cardia," "cardia-esophageal junction," and "cardia orifice"; distal cancers were those classified as "distal," "fundus," "body," "antrum," and "pyloris." All subsite classifications were based on pathology/surgical notes, which were available to us for Ontario cases only.

Population controls aged 20–74 years were identified from provincial health plan rolls (Prince Edward Island, Nova Scotia, Manitoba, Saskatchewan, and British Columbia), property assessment files (Ontario), or random digit dialing (Newfoundland and Alberta). The health care plan rolls and property assessment files cover ≥95 percent of the target control population (28). Cases and controls were mailed self-administered questionnaires. Nonrespondents were recontacted by postcard, letter, and telephone call. Returned questionnaires were reviewed for comprehensibility and completeness, and subjects were telephoned for clarification of answers when needed.

A total of 2,813 cases of gastric adenocarcinoma were identified in the eight provinces (960 females and 1,853 males). Questionnaires were mailed to 1,872 of the 2,813 cases. The remaining cases were not included because they were not eligible for the study by age or diagnosis (n = 603), they were not contacted (n = 63), their physicians refused contact or we were unable to locate them (n = 273), or for other reasons (n = 2). A total of 1,181 questionnaires were returned by gastric adenocarcinoma cases (a 63.1 percent response rate). Slightly more than 95 percent of cases from the province of Ontario were contacted within 6 months of diagnosis; this period was slightly longer in the remaining provinces.

In the parent study, 5,030 questionnaires were completed and returned by controls (65 percent and 71 percent response rates from male and female controls, respectively). For this analysis, controls were matched to cases on age (±5 years) and sex. Twelve participants were excluded because they left ≥50 percent of the food frequency items blank on the questionnaire. The final analysis was based on 373 female cases, 1,131 female controls, 796 male cases, and 1,201 male controls. The ratios of cases to controls differed by sex, since gastric cancer was one of 18 neoplasms investigated in the parent case-control study and control selection was conducted such that each of the neoplasms would have at least one frequency-matched control by 5-year age group. Because gastric cancer is much more common among men than among women, the ratio of controls to cases was higher for women than for men for this tumor site.

Diet measures
The self-administered National Enhanced Cancer Surveillance System questionnaire included a semiquantitative food frequency questionnaire (FFQ). The FFQ was based on the short Block (29) and Willett (30) questionnaires, which have been extensively validated in other, similar populations. The FFQ in this study was not validated or calibrated against standard tools or biomarkers, but it probably had good face validity, since results obtained in previous studies using this tool were consistent with the literature for a number of diet-cancer relations. Subjects were asked to report their consumption frequency of 69 different foods/food groups 2 years prior to participation in the study (controls) or 2 years prior to cancer diagnosis (cases). The frequency scale included "never or less than once per month," "one to three times per month," "once per week," "two to four times per week," "five to six times per week," "once per day," "two to three times per day," "four to five times per day," and "six or more times per day."

The 69 foods and beverages on the FFQ were sorted under general headings: "beverages made with water" (e.g., coffee, tea, tap water); "other beverages" (e.g., whole milk, tomato juice, beer, wine); "fruit" (e.g., apples, oranges); "vegetables" (e.g., broccoli, carrots); "breads and cereals" (e.g., cooked cereals, white bread, dark or whole-grain bread); "meat, poultry, fish, eggs, and cheese" (e.g., chicken, hamburger, eggs); "sweets" (e.g., cake, doughnuts); and "miscellaneous" (e.g., nuts, margarine). Total daily energy intake was estimated by summing each food and beverage item according to its energy content as defined in the 1997 Canadian Nutrient File (31).

Factor analysis was utilized as an empirical method to identify dietary patterns based on the consumption frequencies of all 69 FFQ items. Factor analyses were conducted separately among men and women. The method used an orthogonal variance maximizing (varimax) rotation procedure to create uncorrelated factors. We gave subjects a factor score for each pattern by summing their reported intake of each food item weighted by the factor loading of that item. Factor score coefficients are approximately standard normal values and, because of the varimax rotation procedure, are independent of each other. The preliminary series of factor analysis yielded eight factors with eigenvalues greater than 1. These eight factors explained approximately 23 percent of the total variance in the 69 dietary items for each sex. Based on the scree plot, which shows the variance explained by each factor, and interpretability of the factor scores and patterns, we decided that only two factors logically explained the dietary patterns in this sample. The next factor analysis restricted the output to two factors only; these factors explained approximately 8 percent of the total variance in the diet. Factors in this final analysis had eigenvalues between 3 and 4.5.

Because factor analysis is driven by the data on hand (32), a food index score was defined a priori, based on putative dietary risk factors for gastric cancer that were identified from expert summary reports and recent cohort and case-control studies (1–9). Five general food groups were identified from the literature for calculating the index: 1) fruits (apples/pears, oranges, bananas, cantaloupe, and other fruit (fresh or canned)); 2) vegetables (tomatoes, carrots, broccoli, cabbage/cauliflower/brussels sprouts, spinach or greens, yellow squash, other vegetables, sweet potatoes, and lentils); 3) fiber/whole grains (bran or granola cereals/shredded wheat and dark or whole-grain bread); 4) salted/preserved foods (hot dogs, lunch meats, smoked meats, bacon, sausage, and smoked fish); and 5) sweets/snacks (cake, cookies, doughnuts/pastry, pies, ice cream, chocolate, and potato chips). Scoring for the index was as follows: subjects received zero, one, or two points for each food item based on consumption frequency. For each of the fruit, vegetable, and fiber/whole-grain food groups, subjects were given one point if they reported consuming a food item at a frequency of "two to four times per week" or "five to six times per week"; two points were given if a food item was consumed at a frequency of "once per day" or greater. For the salted/preserved foods and sweets/snacks categories, subjects were similarly allotted points for avoiding food items belonging to those groups (i.e., two points were awarded for reporting a food item frequency as "one to three per month" or less; one point was awarded for reporting a food item frequency as "once per week" or "two to four times week"). The 29 individual food item scores were summed, creating a scale of possible scores from 0 to 58. The food index score was divided into four categories of approximately equal size. To better characterize the individual contribution of each of the five food groups comprising the food index score, results for each of the five food groups are also presented.

Strenuous physical activity was defined in detail previously (33). Briefly, participants reported their typical frequency of several common modes of physical activity (e.g., cycling, jogging, and swimming; average metabolic equivalent value, 7.96) 2 years prior to participation in the study. Modes of activity were converted to common units (metabolic equivalents), and subjects were stratified according to the quartile distribution among controls. Because we recently reported on the influence of strenuous physical activity on gastric cancer risk in this study population (33), we were interested in identifying the combined influences, if any, of diet and physical activity on gastric cancer risk.

Statistical procedures
Approximately 30 lifestyle, occupational, demographic, and reproductive (females only) variables were tested for a potentially confounding influence on the main dietary exposures; no variable confounded a main dietary association by 8 percent or more. The final series of covariates for multivariable models was selected to account for modest confounding and study design (e.g., province of accrual). The final covariates included age, strenuous physical activity 2 years prior to diagnosis/participation, highest nonpregnancy body mass index, income, education, province of residence, pack-years of cigarette smoking, and total energy intake. Physical activity and a Western dietary pattern were assessed for statistical interaction; a cross-product term for lifetime strenuous physical activity and a Western dietary pattern was included in multivariable models, as above.

Odds ratios and 95 percent confidence intervals were estimated from unconditional logistic regression models with SAS software (SAS, version 9.1; SAS Institute, Inc., Cary, North Carolina).


   RESULTS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
DISCUSSION
 References
 
The first factor, termed the "prudent pattern" in keeping with earlier studies (16, 18–22, 24), positively loaded on several vegetables, fruits, and fish (table 1). The second factor, termed the "Western pattern," strongly correlated with soft drinks, French fries, white bread, hamburger, eggs, bacon, doughnuts, and hot dogs; the Western pattern was modestly negatively correlated with broccoli, spinach, and tofu (table 1).


View this table:
[in this window]
[in a new window]

  		TABLE 1. Dietary pattern loadings from factor analysis (prudent and Western) of self-reported dietary consumption 2 years prior in a Canadian case-control study, by sex, 1995–1997

 
Table 2 shows descriptive characteristics of the cases and controls, stratified by sex. Cases reported less physical activity 2 years prior to diagnosis than did controls, as discussed elsewhere (33). Cases also reported a higher body mass index, less education (more so among men), a higher income (among women), and more pack-years of cigarette smoking.


View this table:
[in this window]
[in a new window]

  		TABLE 2. Distribution of selected covariates among gastric cancer cases and controls in a Canadian case-control study, by sex, 1995–1997

 
The prudent pattern was not associated with gastric cancer risk in men; however, an approximate one third reduction in disease risk was observed among women in the highest quartile of the prudent pattern relative to the lowest quartile (odds ratio (OR) = 0.58, 95% confidence interval (CI): 0.37, 0.92) (table 3). The Western pattern was positively associated with gastric cancer risk in both men (OR = 1.44, 95 percent CI: 1.03, 2.02) and women (OR = 1.86, 95 percent CI: 1.20, 2.89). Individual scores for vegetables, fiber/whole-grain foods, low intake of salted/preserved foods, and low intake of sweets/snacks were negatively associated with gastric cancer risk in women. Among men, increased intake of fiber/whole grains, low intake of salted/preserved foods, and low intake of snacks/sweets were associated with decreased risk of gastric cancer. The composite food index score, based on all 28 food items representative of the five major food groups identified from the literature, was significantly inversely associated with gastric cancer risk in both women (fourth quartile relative to first: OR = 0.40, 95 percent CI: 0.27, 0.59) and men (OR = 0.63, 95 percent CI: 0.45, 0.88).


View this table:
[in this window]
[in a new window]

  		TABLE 3. Association between dietary pattern and risk of gastric cancer in a Canadian case-control study, by sex, 1995–1997

 
Table 4 shows the combined influence of recent strenuous physical activity and a Western dietary pattern. Among women and men in both the top 50th percentile of physical activity and the bottom 50th percentile of the Western dietary pattern, the risk of gastric cancer was lower than that among respondents with the inverse profile of both risk factors (females: OR = 0.39, 95 percent CI: 0.25, 0.61; males: OR = 0.45, 95 percent CI: 0.32, 0.64). The interaction terms for physical activity and a Western dietary pattern did not reach statistical significance in models for males or females (both p's > 0.11). In subsite analyses (table 5) restricted to a small portion of the overall case sample, dietary variables were relatively more consistently associated with distal gastric cancer than with cardia gastric cancer among women. Among men, dietary variables were relatively more consistently linked with risk of cancer at the gastric cardia subsite. Four-digit International Classification of Diseases, Ninth Revision, codes and surgical/pathology notes were available only from Ontario.


View this table:
[in this window]
[in a new window]

  		TABLE 4. Combined influence of dietary pattern and physical activity on risk of gastric cancer in a Canadian case-control study, by sex, 1995–1997

 

View this table:
[in this window]
[in a new window]

  		TABLE 5. Association between dietary pattern and risk of gastric cancer in a subsample of case-control study participants, by tumor location and sex, Ontario, Canada, 1995–1997

 

   DISCUSSION
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
References
 
Data from a national population-based case-control study from Canada identified two dietary patterns, labeled Western and prudent, that explained approximately 8 percent of the variance in responses to a 69-item FFQ. The Western pattern was positively associated with gastric cancer risk among women and men. The prudent pattern was linked with an inverse risk of gastric cancer in women only. Similarly, risk estimates from individual food group scores for high intakes of vegetables and fiber/whole grains and low intakes of salted/preserved foods and sweets/snacks indicated reduced risk of disease among women. Among men, increased intake of fiber/whole grains and low intakes of salted/preserved foods and sweets/snacks were associated with decreased risk of gastric cancer. The food index score was associated with decreased risk of gastric cancer in women and men.

The Western pattern in the current study, which correlated positively with soft drinks, French fries, white bread, hamburger, hot dogs, and doughnuts, is comparable to the "Western" (34, 35), "starchy" (36), "meat" (37), and "high meat" (38) dietary patterns described in previous studies of gastric cancer risk from Japan (34, 37), Sweden (35), Uruguay (36), and the United States (38). Similar to the case in most previous work (35–38), the Western dietary pattern in this study was associated with increased risk of gastric cancer. Collectively, these results suggest that diets characterized by high intake of preserved meats and refined grains and sugars increase the risk of gastric cancer in several populations. Additionally, low intake of the Western pattern in combination with increased physical activity was strongly associated with decreased risk of gastric cancer. We are aware of no other study that combined diet and physical activity in a similar manner; however, we recently reported results from a more complete analysis of physical activity and risk of incident gastric cancer from these data (33).

Low intake of salted/preserved food was associated with decreased gastric cancer risk in both sexes. In a recent cohort study of female participants, processed meat consumption in the highest tertile was associated with a 66 percent (95 percent CI: 1.13, 2.45) increased risk of gastric cancer (7). Similarly, a large cohort study carried out among half a million men and women in Europe indicated a 62 percent (95 percent CI: 1.08, 2.41) increased risk of gastric cancer for the highest quartile of processed meat intake versus the lowest (9). Summary relative risks for each 30-g/day increase in processed meat consumption were 1.15 (95 percent CI: 1.04, 1.27) from six cohort studies and 1.38 (95 percent CI: 1.19, 1.60) from nine case-control studies (39). Mechanistic support for a link between salted/preserved foods and gastric cancer comes from animal studies in which high salt consumption induced gastritis and mucosal damage (40), potentially initiating gastric cancer, whereas red and processed meats may be further linked to cancer risk because of their heme (haem) and nitrite content, which correlates with formation of N-nitroso compounds (41, 42). N-nitroso compounds, in turn, may cause chronic gastric inflammation, a precursor to cancer (43).

The prudent pattern in the current study is comparable to the "healthy" (34–36, 38) and "vegetable and fruit" (37) dietary patterns described previously. We report an inverse association between the prudent pattern and disease risk among women and no association among men. The differential prudent pattern findings for men and women may be explained by differences in intake of vegetables and salted/preserved foods. Control comparisons between sexes in the fourth quartile of the prudent pattern suggested that males consumed significantly fewer vegetables (mean vegetable intake score: males, 5.0; females, 5.8; t test: p < 0.001) and more salted/preserved foods (mean avoidance of salted/preserved foods intake score (where a low score indicates higher consumption): males, 10.9; females, 11.3; t test: p < 0.0001) than did females. This may indicate that men in the highest quartile of the prudent pattern consumed too much salt/preserved food and too few vegetables to experience any influence on gastric cancer risk. Alternatively, or perhaps concurrently, men in general were heavier and smoked more cigarettes than women, which may have overshadowed any beneficial effect of relatively high fruit and vegetable intake. Similar results were reported from a Japanese cohort study (34), where a healthy dietary pattern was associated with reduced risk of gastric cancer among women but not among men. Slightly contradictory evidence comes from a study in Uruguay, where a healthy dietary pattern was associated with decreased gastric cancer risk among men but not among women (36). The discordant results could indicate different population dietary deficiencies in Uruguay than were observed among the Japanese and Canadian samples. This possibility has been offered to explain the discordant results between randomized controlled trials of antioxidants and gastric cancer in different populations (10–13). Studies from Sweden (35) and Japan (males) (37) have indicated weak, nonsignificant inverse associations between a healthy dietary pattern and gastric cancer risk.

Mechanistic studies have suggested that the cancer-protective effects of fruits and vegetables are probably related to their antioxidant content, including vitamin C and β-carotene, which are capable of scavenging highly oxidative molecules and preventing DNA and lipid membrane damage (2). Meta-analyses of the relation between fruit and vegetable intake (not a prudent dietary pattern per se) and gastric cancer risk have indicated summary relative risks of 0.81 (95 percent CI: 0.75, 0.87) and 0.74 (95 percent CI: 0.69, 0.81) for each 100-g increase in fruit intake and vegetable intake, respectively (1). When the summary estimates were restricted to results from cohort studies, however, the reduced risks were attenuated (relative risks were 0.89 for both fruits and vegetables) and no longer statistically significant. Likewise, a recent large cohort study from Europe indicated no association between fruits and vegetables and gastric cancer risk (8). It is difficult to explain the discrepant results between study designs, but the discrepancy may involve recall bias in case-control studies or changing dietary patterns over time (e.g., measurement error) in cohort studies.

Factor analysis of dietary patterns has a certain degree of arbitrariness, as discussed previously (44). For example, there are no standard criteria for selecting the number of food item variables, the number of factors, the value of factor loadings, the method of rotation, or the naming of the factors. The fact that our study also included an a priori diet assessment method, the food index score, added to the robustness of these analyses.

Given the natural history of gastric cancer, it is possible that dietary recall by cases could be affected by their own perception of what dietary factors cause the disease, leading to recall bias. This is less likely to have been a problem in the current study, however, because this was primarily a study of environmental exposures, and knowledge of diet-stomach cancer associations among the general population of Canada was probably low at the time of data collection. Other potential limitations of this study include the lack of information on gastroesophageal reflux disease and use of nonsteroidal antiinflammatory drugs, as well as the 63–71 percent participation rate, which may have contributed to selection bias.

A further limitation in this study was the lack of available data on Helicobacter pylori infection among study participants. Few studies have examined the potential modifying effects of H. pylori infection status on diet-gastric cancer risk (9). Since H. pylori infection is differentially associated with risk of cancer in the gastric cardia (no association) as opposed to the distal stomach (increased risk) (45), examining diet-disease associations according to infection status is warranted. A recent European cohort study suggested that associations between meat intake and gastric cancer risk were largely restricted to participants infected with H. pylori (9). Generally, the Canadian population has a low incidence of H. pylori infection (46) and low rates of stomach cancer (47); it is therefore plausible that we were able to identify risk factors in this population that may be overshadowed by H. pylori infection in other populations.

In future work, investigators must also consider gastric cardia cancers versus noncardia/distal cancers with more statistical power than was afforded by the current study. Our subsite analyses suggested stronger links between the dietary variables and distal cancer among women, whereas among men, dietary variables were more consistently associated with cardia gastric cancer. However, the small sample sizes (n = 22–110) and relatively crude median splitting of the dietary exposures in these analyses precluded firm interpretation of the data.

Our findings provide support that diets characterized by Western features (e.g., processed meats and refined sugars) are associated with elevated risk of gastric cancer in both females and males. The food index score in our study was associated with lower gastric cancer risk among women and men, a finding with potential clinical utility. Dietary pattern analysis, although open to some subjectivity, may prove useful in developing preventive dietary strategies aimed at reducing chronic disease risk, including risk of gastric cancer.


   ACKNOWLEDGMENTS
 
These data were collected within the context of the National Enhanced Cancer Surveillance System, sponsored by the Laboratory Centre for Disease Control, Health Canada.

Conflict of interest: none declared.


   References
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 References
 

  1. Riboli E, Norat T. Epidemiologic evidence of the protective effect of fruit and vegetables on cancer risk. Am J Clin Nutr (2003) 78(suppl):559S–69S.[Abstract/Free Full Text]
  2. World Cancer Research Fund/American Institute for Cancer Research. Food, nutrition and the prevention of cancer: a global perspective. (1997) Washington, DC: American Institute for Cancer Research.
  3. Nomura AM, Hankin JH, Kolonel LN, et al. Case-control study of diet and other risk factors for gastric cancer in Hawaii (United States). Cancer Causes Control (2003) 14:547–58.[CrossRef][Web of Science][Medline]
  4. Key TJ, Schatzkin A, Willett WC, et al. Diet, nutrition and the prevention of cancer. Public Health Nutr (2004) 7:187–200.[CrossRef][Web of Science][Medline]
  5. Ngoan LT, Mizoue T, Fujino Y, et al. Dietary factors and stomach cancer mortality. Br J Cancer (2002) 87:37–42.[CrossRef][Web of Science][Medline]
  6. van den Brandt PA, Botterweck AA, Goldbohm RA. Salt intake, cured meat consumption, refrigerator use and stomach cancer incidence: a prospective cohort study (Netherlands). Cancer Causes Control (2003) 14:427–38.[CrossRef][Web of Science][Medline]
  7. Larsson SC, Bergkvist L, Wolk A. Processed meat consumption, dietary nitrosamines and stomach cancer risk in a cohort of Swedish women. Int J Cancer (2006) 119:915–19.[CrossRef][Medline]
  8. Gonzalez CA, Pera G, Agudo A, et al. Fruit and vegetable intake and the risk of stomach and oesophagus adenocarcinoma in the European Prospective Investigation into Cancer and Nutrition (EPIC-EURGAST). Int J Cancer (2006) 118:2559–66.[CrossRef][Web of Science][Medline]
  9. Gonzalez CA, Jakszyn P, Pera G, et al. Meat intake and risk of stomach and esophageal adenocarcinoma within the European Prospective Investigation into Cancer and Nutrition (EPIC). J Natl Cancer Inst (2006) 98:345–54.[Abstract/Free Full Text]
  10. Blot WJ, Li JY, Taylor PR, et al. Nutrition intervention trials in Linxian, China: supplementation with specific vitamin/mineral combinations, cancer incidence, and disease-specific mortality in the general population. J Natl Cancer Inst (1993) 85:1483–92.[Abstract/Free Full Text]
  11. Correa P, Fontham ET, Bravo JC, et al. Chemoprevention of gastric dysplasia: randomized trial of antioxidant supplements and anti-Helicobacter pylori therapy. J Natl Cancer Inst (2000) 92:1881–8.[Abstract/Free Full Text]
  12. Varis K, Taylor PR, Sipponen P, et al. Gastric cancer and premalignant lesions in atrophic gastritis: a controlled trial on the effect of supplementation with alpha-tocopherol and beta-carotene. The Helsinki Gastritis Study Group. Scand J Gastroenterol (1998) 33:294–300.[CrossRef][Web of Science][Medline]
  13. Jacobs EJ, Connell CJ, McCullough ML, et al. Vitamin C, vitamin E, and multivitamin supplement use and stomach cancer mortality in the Cancer Prevention Study II cohort. Cancer Epidemiol Biomarkers Prev (2002) 11:35–41.[Abstract/Free Full Text]
  14. Willett W. Nutritional epidemiology. (1998) 2nd ed. New York NY: Oxford University Press.
  15. Hu FB. Dietary pattern analysis: a new direction in nutritional epidemiology. Curr Opin Lipidol (2002) 13:3–9.[CrossRef][Web of Science][Medline]
  16. Hu FB, Rimm EB, Stampfer MJ, et al. Prospective study of major dietary patterns and risk of coronary heart disease in men. Am J Clin Nutr (2000) 72:912–21.[Abstract/Free Full Text]
  17. Hu FB, Rimm E, Smith-Warner SA, et al. Reproducibility and validity of dietary patterns assessed with a food-frequency questionnaire. Am J Clin Nutr (1999) 69:243–9.[Abstract/Free Full Text]
  18. Fung TT, Hu FB, Holmes MD, et al. Dietary patterns and the risk of postmenopausal breast cancer. Int J Cancer (2005) 116:116–21.[CrossRef][Web of Science][Medline]
  19. Fung TT, Rimm EB, Spiegelman D, et al. Association between dietary patterns and plasma biomarkers of obesity and cardiovascular disease risk. Am J Clin Nutr (2001) 73:61–7.[Abstract/Free Full Text]
  20. Fung TT, Schulze M, Manson JE, et al. Dietary patterns, meat intake, and the risk of type 2 diabetes in women. Arch Intern Med (2004) 164:2235–40.[Abstract/Free Full Text]
  21. Fung TT, Willett WC, Stampfer MJ, et al. Dietary patterns and the risk of coronary heart disease in women. Arch Intern Med (2001) 161:1857–62.[Abstract/Free Full Text]
  22. Osler M, Heitmann BL, Gerdes LU, et al. Dietary patterns and mortality in Danish men and women: a prospective observational study. Br J Nutr (2001) 85:219–25.[Web of Science][Medline]
  23. Handa K, Kreiger N. Dietary patterns and the risk of renal cell carcinoma. Public Health Nutr (2002) 5:757–67.[CrossRef][Web of Science][Medline]
  24. Wu K, Hu FB, Fuchs C, et al. Dietary patterns and risk of colon cancer and adenoma in a cohort of men (United States). Cancer Causes Control (2004) 15:853–62.[CrossRef][Web of Science][Medline]
  25. Nkondjock A, Krewski D, Johnson KC, et al. Dietary patterns and risk of pancreatic cancer. Int J Cancer (2005) 114:817–23.[CrossRef][Web of Science][Medline]
  26. Frise S, Kreiger N, Gallinger S, et al. Menstrual and reproductive risk factors and risk for gastric adenocarcinoma in women: findings from the Canadian National Enhanced Cancer Surveillance System. Ann Epidemiol (2006) 16:908–16.[CrossRef][Web of Science][Medline]
  27. World Health Organization. International classification of diseases. The international statistical classification of diseases, injuries, and causes of death. In: Ninth Revision (1977) Geneva, Switzerland: Word Health Organization.
  28. Holowaty E. The Ontario Cancer Registry: a registry with almost completely automated data collection. In: Automated data collection in a cancer registry. (IARC Technical Report no. 32)—Black RJ, Simonato L, Storm HH, eds. (1998) Lyon, France: IARC Press. 39–44.
  29. Block G, Hartman AM, Naughton D. A reduced dietary questionnaire: development and validation. Epidemiology (1990) 1:58–64.[Medline]
  30. Willett WC. Diet, nutrition, and avoidable cancer. Environ Health Perspect (1995) 103(suppl 8):165–70.
  31. Health Canada. Nutrient value of some common foods (1997) Ottawa, Ontario, Canada: Health Canada. (http://www.hc-sc.gc.ca/fn-an/nutrition/fiche-nutri-data/nutrient_value-valeurs_nutritives_e.html).
  32. Trichopoulos D, Lagiou P. Dietary patterns and mortality. Br J Nutr (2001) 85:133–4.[Web of Science][Medline]
  33. Campbell PT, Sloan M, Kreiger N. Physical activity and stomach cancer risk: the influence of intensity and timing during the lifetime. Eur J Cancer (2007) 43:593–600.[CrossRef][Web of Science][Medline]
  34. Kim MK, Sasaki S, Sasazuki S, et al. Prospective study of three major dietary patterns and risk of gastric cancer in Japan. Int J Cancer (2004) 110:435–42.[CrossRef][Web of Science][Medline]
  35. Bahmanyar S, Ye W. Dietary patterns and risk of squamous-cell carcinoma and adenocarcinoma of the esophagus and adenocarcinoma of the gastric cardia: a population-based case-control study in Sweden. Nutr Cancer (2006) 54:171–8.[CrossRef][Web of Science][Medline]
  36. De Stefani E, Correa P, Boffetta P, et al. Dietary patterns and risk of gastric cancer: a case-control study in Uruguay. Gastric Cancer (2004) 7:211–20.[CrossRef][Medline]
  37. Masaki M, Sugimori H, Nakamura K, et al. Dietary patterns and stomach cancer among middle-aged male workers in Tokyo. Asian Pac J Cancer Prev (2003) 4:61–6.[Medline]
  38. Chen H, Ward MH, Graubard BI, et al. Dietary patterns and adenocarcinoma of the esophagus and distal stomach. Am J Clin Nutr (2002) 75:137–44.[Abstract/Free Full Text]
  39. Larsson SC, Orsini N, Wolk A. Processed meat consumption and stomach cancer risk: a meta-analysis. J Natl Cancer Inst (2006) 98:1078–87.[Abstract/Free Full Text]
  40. Tatematsu M, Takahashi M, Fukushima S, et al. Effects in rats of sodium chloride on experimental gastric cancers induced by N-methyl-N-nitro-N-nitrosoguanidine or 4-nitroquinoline-1-oxide. J Natl Cancer Inst (1975) 55:101–6.[Web of Science][Medline]
  41. Bingham SA, Hughes R, Cross AJ. Effect of white versus red meat on endogenous N-nitrosation in the human colon and further evidence of a dose response. J Nutr (2002) 132(suppl):3522S–5S.[Abstract/Free Full Text]
  42. Cross AJ, Pollock JR, Bingham SA. Haem, not protein or inorganic iron, is responsible for endogenous intestinal N-nitrosation arising from red meat. Cancer Res (2003) 63:2358–60.[Abstract/Free Full Text]
  43. Bartsch H, Spiegelhalder B. Environmental exposure to N-nitroso compounds (NNOC) and precursors: an overview. Eur J Cancer Prev (1996) 5(suppl 1):11–17.[Medline]
  44. Terry P, Hu FB, Hansen H, et al. Prospective study of major dietary patterns and colorectal cancer risk in women. Am J Epidemiol (2001) 154:1143–9.[Abstract/Free Full Text]
  45. Helicobacter and Cancer Collaborative Group. Gastric cancer and Helicobacter pylori: a combined analysis of 12 case control studies nested within prospective cohorts. Gut (2001) 49:347–53.[Abstract/Free Full Text]
  46. Sullivan T, Ashbury FD, Fallone CA, et al. Helicobacter pylori and the prevention of gastric cancer. Can J Gastroenterol (2004) 18:295–302.[Web of Science][Medline]
  47. Canadian Cancer Society/National Cancer Institute of Canada. Canadian cancer statistics 2006. (2006) Toronto, Ontario, Canada: Canadian Cancer Society.

Add to CiteULike CiteULike   Add to Connotea Connotea   Add to Del.icio.us Del.icio.us    What's this?


This article has been cited by other articles:


Home page
 Cancer Epidemiol. Biomarkers Prev.Home page
P. Bertuccio, V. Edefonti, F. Bravi, M. Ferraroni, C. Pelucchi, E. Negri, A. Decarli, and C. La Vecchia
Nutrient Dietary Patterns and Gastric Cancer Risk in Italy
Cancer Epidemiol. Biomarkers Prev., November 1, 2009; 18(11): 2882 - 2886.
[Abstract] [Full Text] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow FREE Full Text (PDF) Freely available
Right arrow All Versions of this Article:
167/3/295    most recent
kwm294v1
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Add to My Personal Archive
Right arrow Download to citation manager
Right arrow Search for citing articles in:
ISI Web of Science (3)
Right arrowRequest Permissions
Right arrow Disclaimer
Google Scholar
Right arrow Articles by Campbell, P. T.
Right arrow Articles by Kreiger, N.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Campbell, P. T.
Right arrow Articles by Kreiger, N.
Social Bookmarking
Add to CiteULike   Add to Connotea   Add to Del.icio.us  
What's this?