© The Author 2006. Published by Oxford University Press.
ARTICLE |
Processed Meat Consumption and Stomach Cancer Risk: A Meta-Analysis
Affiliation of authors: Division of Nutritional Epidemiology, The National Institute of Environmental Medicine, Karolinska Institutet, Stockholm, Sweden
Correspondence to: Susanna C. Larsson, MSc, Division of Nutritional Epidemiology, The National Institute of Environmental Medicine, Karolinska Institutet, Box 210, SE-171 77, Stockholm, Sweden (e-mail: susanna.larsson{at}ki.se).
 |
ABSTRACT |
|---|
 TopNotes Abstract IntroductionSubjects and methodsResultsDiscussionReferences |
|---|
Background: The relationship between processed meat consumption and the risk of stomach cancer is controversial. We conducted a meta-analysis to summarize available evidence from cohort and case–control studies on this issue. Methods: We searched Medline for studies of processed meat consumption and stomach cancer published from January 1966 through March 2006. Random-effects models were used to pool the relative risks from individual studies. All statistical tests were two-sided. Results: Six prospective cohort studies (involving 2209 stomach cancer patients) and nine case–control studies (2495 case patients) were eligible for inclusion in the dose–response meta-analysis of processed meat consumption. The estimated summary relative risks of stomach cancer for an increase in processed meat consumption of 30 g/day, approximately half of an average serving, were 1.15 (95% confidence interval [CI] = 1.04 to 1.27) for the cohort studies and 1.38 (95% CI = 1.19 to 1.60) for the case–control studies. There was no statistically significant heterogeneity among the cohort studies (P = .42) or among the case–control studies (P = .19). In three cohort and four case–control studies that examined the association between bacon consumption and stomach cancer, the summary relative risk was 1.37 (95% CI = 1.17 to 1.61) for the highest versus lowest intake categories of bacon, without heterogeneity among these studies (P = .66). Conclusion: Increased consumption of processed meat is associated with an increased risk of stomach cancer. However, the possibility that the association may be confounded or modified by other factors cannot be ruled out.
|
INTRODUCTION |
|---|
|
TopNotesAbstractIntroductionSubjects and methodsResultsDiscussionReferences |
|---|
Despite a steady global decline in stomach cancer incidence and mortality over the last 50 years, this malignancy remains the second leading cause of cancer death worldwide (1). In fact, stomach cancer accounts for nearly 10% of all cancer deaths and claims approximately 700 000 lives annually (1). Because the prognosis is poor, identification of risk factors amenable for modification could have a marked impact on reducing stomach cancer morbidity and mortality. Infection with Helicobacter pylori is strongly implicated in stomach cancer etiology (2,3). Nevertheless, only a small proportion of people infected with the bacterium develop stomach cancer, suggesting that such infection is not sufficient itself to cause this malignancy.
The decline in incidence rates of stomach cancer has been proposed to be attributable, at least in part, to improvements in diet and food storage and, in particular, the introduction of refrigeration. In addition to increased year-round availability of fresh fruits and vegetables, refrigeration has reduced the need for salt and other methods of food preservation. Processed meat includes foods preserved by salting, smoking, or adding nitrates or nitrites. Diets high in salt could damage the gastric mucosa, leading to gastritis, increased DNA synthesis, and excessive cell replication (4–6). Processed meat often contains, besides high amounts of salt, carcinogenic N-nitroso compounds (7–9). Although an association between processed meat consumption and stomach cancer risk is biologically plausible, epidemiologic studies on this relationship have yielded inconsistent results.
A systematic and quantitative assessment of published findings on the association between processed meat consumption and risk of stomach cancer is not available. Therefore, we used meta-analysis as a systematic approach to summarize evidence from cohort and case–control studies on this issue.
|
SUBJECTS AND METHODS |
|---|
|
TopNotesAbstractIntroductionSubjects and methodsResultsDiscussionReferences |
|---|
Search Strategy
To identify epidemiologic studies of processed meat consumption and risk of stomach cancer, we conducted a literature search in Medline for articles published in any language from January 1966 through March 2006. We used the following medical subject heading terms and/or text words: "meat" or "foods" combined with "stomach cancer," "stomach neoplasm," "gastric cancer," or "gastric neoplasm." We also searched the reference lists of pertinent publications manually to identify additional studies.
Study Selection
To be included in our meta-analysis, studies had to 1) be a cohort or case–control study with stomach cancer incidence or mortality as the outcome and 2) provide relative risk estimates with confidence intervals (or information to compute them) of stomach cancer associated with processed meat consumption. Processed meat was taken to include any of the following meat items: bacon, sausage, hot dogs, salami, and ham. We considered "cured meat," "preserved meat," "salted meat," and "smoked meat" as equivalent to "processed meat."
We identified 11 prospective cohort studies (10–20) and 22 case–control studies (21–42) with data that were potentially eligible for inclusion in the meta-analysis. One cohort study (10) and three case–control studies (27,32,34) were excluded for the following reasons: no relative risks (10) or confidence intervals (27) were provided and could not be calculated, the relative risk was for 1 standard deviation difference in processed meat consumption (32), or the exposure was nonspecific (mixed with eggs) (34). The remaining 10 cohort studies (11–20) and 19 case–control studies (21–26,28–31,33,35–42) were included in the meta-analysis.
Data Extraction
We extracted from each publication the following data: the first author's last name, the year of publication, the country in which the study was performed, study design, the sample size, the type of control subjects (in case–control studies), duration of follow-up (in cohort studies), variables controlled for by matching or in the multivariable model, and the relative risks with corresponding confidence intervals for each category of processed meat consumption and/or for processed meat consumption as a continuous variable. From each study, we extracted the relative risk estimates that reflected the greatest degree of control for potential confounders.
Statistical Analysis
Relative risk (RR) was used as the measure of the association of processed meat consumption with stomach cancer risk. Because the absolute risk of stomach cancer is low, the odds ratios in case–control studies approximate the relative risks (43); we therefore report all results as relative risks for simplicity. Relative risks from individual studies and corresponding standard errors (derived from the confidence intervals) were transformed to their natural logarithms to stabilize the variance and to normalize the distributions. We quantified the relationship between processed meat consumption and stomach cancer risk with the method of DerSimonian and Laird (44) by using the assumptions of a random effects model, which incorporates both within- and between-study variability. When separate relative risk estimates were provided for men and women (15,18,40), we pooled the relative risks (weighted by the inverse of their variance) to obtain a single relative risk from each study. For one case–control study (29) that used two control groups, the relative risks based on comparison with population control subjects were used in the overall meta-analysis.
For the dose-response meta-analysis, we used the method proposed by Greenland and associates (45,46) to compute study-specific slopes (linear trends) from the correlated natural log of the relative risks across categories of processed meat consumption. This method requires that the distribution of case patients and control subjects (or person–time) and the relative risk with its variance estimate for at least three quantitative exposure categories be known. For studies that did not provide the number of case patients and control subjects in each exposure category (15,25,29,34,36,39–41), we estimated the slopes using variance-weighted least-squares regression models. We rescaled processed meat consumption to grams per day using 50 g as the approximate average portion (serving) size (47). For each study, the median or mean level of consumption for each category of processed meat was assigned to each corresponding relative risk. When the median or mean consumption per category was not presented in the article, we assigned the midpoint of the upper and lower boundaries in each category as the average consumption. If the upper boundary of the highest category was not provided, we assumed that it had the same amplitude as the preceding category. When the lowest category was open-ended, the lowest boundary was assumed to be zero. For three studies that reported relative risks for processed meat consumption as a continuous variable (19–21), we used these relative risks in the dose–response analysis. We used an increase in processed meat consumption of 30 g/day, which is approximately half of an average serving. This cut point was chosen a priori; the same cut point had been used in a previous meta-analysis of processed meat consumption and risk of colorectal cancer (47). We checked for nonlinearity of the dose–response relationship between processed meat consumption and stomach cancer by estimating polynomial models. However, the best-fitting model was a linear model.
Statistical heterogeneity among studies was evaluated using the Q and I2 statistics (48). For the Q statistic, a P<.1 was considered to be representative of statistically significant heterogeneity. I2 is the proportion of total variation contributed by between-study variation (48). We conducted subgroup analyses by study design (cohort versus case–control), the type of control subjects (hospital-based versus population-based) in case–control studies, the location of the study (North America, Europe, or Asia), and publication year (before 2000 versus after). Separate analyses were performed for cohort and case–control studies when there was statistically significant heterogeneity in the summary results between these study designs. Publication bias was assessed with Egger's regression asymmetry test (49); P<.1 was considered to be representative of statistically significant publication bias. Statistical analyses were performed with Stata, version 9.0 (StataCorp, College Station, TX). All statistical tests were two-sided.
|
RESULTS |
|---|
|
TopNotesAbstractIntroductionSubjects and methodsResultsDiscussionReferences |
|---|
Study Characteristics
Of the 10 prospective cohort studies, four were carried out in the United States (including one in Hawaii and one among men of Japanese ancestry), four in Europe, and two in Japan (Table 1). The study population consisted of men and women in seven cohort studies, of only men in two studies, and of only women in one study. Sample sizes ranged from 3123 (17) to 970 045 (15), and the number of stomach cancer cases varied from 68 (14) to 1349 (15). In six studies, the outcome was stomach cancer incidence (11,13,14,17,19,20), and in four studies, the outcome was stomach cancer mortality (12,15,16,18). The duration of follow-up was longer than 10 years in all but two studies (17,19).
|
Of the 19 case–control studies, three were conducted in the United States, seven in Europe, two in Japan, two in Uruguay, and one each in Canada, Taiwan, China, Puerto Rico, and Mexico (Table 2). The number of case patients enrolled in these studies ranged from 109 (28) to 741 (26), and the number of control subjects varied from 123 (28) to 36 490 (39). Control subjects were drawn from the general population (10 studies), hospitals (eight studies), or both (one study).
|
Processed Meat
Highest versus lowest intake category. Seven cohort studies (11,13–16,19,20) and 14 case–control studies (21,23–25,29,33,35–42) examined the relationship between processed meat consumption and stomach cancer risk. For the meta-analysis comparing the highest with the lowest consumption categories, two case–control studies were excluded because processed meat consumption was analyzed as a continuous variable (21) or the study (37) was superseded by a later publication (42) based on the same population but with a larger sample size [the smaller study (37) was included in the dose–response meta-analysis because processed meat consumption was not quantified in the larger study (42)].
There was statistically significant heterogeneity between the cohort and case–control studies in the summary relative risks (Q = 2.85; P = .09). Combined, the seven cohort studies included 2277 stomach cancer patients. In these studies, the summary relative risk of stomach cancer was 1.24 (95% confidence interval [CI] = 0.98 to 1.56; Fig. 1) for individuals in the highest relative to the lowest category of processed meat consumption. The corresponding summary relative risk for the 12 case–control studies (involving 3030 case patients) was 1.63 (95% CI = 1.31 to 2.01; Fig. 2). Statistically significant heterogeneity was observed among the cohort studies (Q = 12.99; P = .04; I2 = 53.8%) and among the case–control studies (Q = 19.11; P = .06; I2 = 42.4%). Within each study design, the association between processed meat consumption and stomach cancer risk did not differ statistically significantly by the location of the study or publication year (P>.1 for all). No publication bias was observed (P = .68 for cohort studies and P = .93 for case–control studies).
|
|
Dose–response meta-analysis. The dose–response meta-analysis of processed meat consumption included six cohort studies (N = 2209 cases) (11,13,15,16,19,20) and nine case–control studies (N = 2495 cases) (21,25,29,33,35,37,39–41). One cohort study (14) and five case–control studies (23,24,36,38,42) were excluded because processed meat consumption could not be quantified (14,24,38,42) or because the range of exposure was very narrow (23,36), leading to exaggerated regression coefficients.
The estimated summary relative risks of stomach cancer for an increase in processed meat consumption of 30 g/day were 1.15 (95% CI = 1.04 to 1.27) for cohort studies and 1.38 (95% CI = 1.19 to 1.60) for case–control studies. There was no statistically significant heterogeneity among the cohort studies or among the case–control studies (Table 3), but the estimated summary relative risk was statistically significantly higher for case–control than for cohort studies (Q = 3.97; P = .05). The association between processed meat consumption and stomach cancer risk was similar for population-based case–control and hospital based case–control studies (Q = 0.17; P = .68).
|
Individual processed meat items. Bacon consumption was examined in three cohort studies (12,17,20) and four case–control studies (25,30,31,40). All studies reported a positive association, and in two cohort studies (17,20) and one case–control study (31) the relationship was statistically significant (Fig. 3). There was no statistically significant heterogeneity between the summary relative risks for cohort and case–control studies (Q = 0.00; P = .96). When all seven studies were analyzed together, the summary relative risk of stomach cancer was 1.37 (95% CI = 1.17 to 1.61) for individuals in the highest category of bacon consumption compared with those in the lowest category. There was no statistically significant heterogeneity among the seven studies (Q = 4.13; P = .66; I2 = 0), and no publication bias was found (P = .51).
|
Sausage consumption was evaluated in three cohort studies (17,18,20) and six case–control studies (22,26,28,30,31,41). Three case–control studies reported a statistically significant positive association between sausage consumption and stomach cancer risk (26,28,31), and two case–control (22,41) and two cohort studies (18,20) found non–statistically significant increased risks (Fig. 4). There was no statistically significant heterogeneity between the summary relative risks for cohort and case–control studies (Q = 0.58; P = .45). Analysis of all nine studies showed a summary relative risk of 1.39 (95% CI = 1.12 to 1.73) for high versus low sausage consumption; however, there was statistically significant heterogeneity among studies (Q = 19.52; P = .01; I2 = 59.0%). Publication bias was observed for all studies (P =.03) but not by study type (P = .41 for cohort studies and P = .11 for case–control studies).
|
Ham consumption was investigated in two cohort studies (17,20) and three case–control studies (22,25,31). Of the two cohort studies, one observed a non–statistically significant 48% (RR = 1.48; 95% CI = 0.99 to 2.22) increased risk of stomach cancer for high versus low ham consumption (20), whereas the other cohort study (17) did not support such an association (RR = 0.77; 95% CI = 0.56 to 1.07). In the three case–control studies, there was a statistically significant 64% increased risk of stomach cancer for individuals in the highest relative to the lowest category of ham consumption (RR = 1.64, 95% CI = 1.31 to 2.06; Fig. 5). The Egger's test of publication bias for case–control studies was of borderline statistical significance (P = .10).
|
|
DISCUSSION |
|---|
|
TopNotesAbstractIntroductionSubjects and methodsResultsDiscussionReferences |
|---|
Findings of this meta-analysis support a positive relationship between processed meat consumption and risk of stomach cancer. Overall, an increase in processed meat consumption of 30 g/day was associated with statistically significant 15% and 38% increased risks of stomach cancer in cohort studies and case–control studies, respectively. Among the individual processed meat items, findings were most consistent for bacon consumption. Summary results indicate that there was a statistically significant 37% higher risk of stomach cancer among those in the highest relative to the lowest category of bacon consumption.
Our study has several limitations. First, as a meta-analysis of observational studies, it is prone to bias (e.g., recall and selection bias) inherent in the original studies. Prospective cohort studies are less susceptible to bias than case–control studies because, in the prospective design, information on exposures is collected before the diagnosis of the disease. The association between processed meat consumption and stomach cancer risk was statistically significantly stronger in the case–control studies than in the cohort studies. It is possible that the relationships reported from case–control studies may have been overstated due to recall or interviewer bias and possible prediagnostic early symptoms in cancer patients may have led to changes in dietary habits. Moreover, selection bias is of concern in the studies with low participation rates among control subjects (21,29,40) because those who participate are likely to be more health conscious and thus might consume less meat than nonrespondents.
A second limitation is that individual studies may have failed to control for potential confounders, which may introduce bias in an unpredictable direction. For example, infection with H. pylori is an established risk factor for gastric noncardia cancer (3). No study controlled for H. pylori infection status, although one study (19) examined whether the association between processed meat consumption and stomach cancer risk was modified by H. pylori infection (see below). The two prospective studies that showed a statistically significant increased risk of stomach cancer associated with a high consumption of processed meat (19,20) both adjusted for major potential confounders, such as age, education, smoking, body size, and intakes of total energy, alcohol, fruits, and vegetables. In these studies, the age-adjusted relative risks for the highest compared with the lowest category of processed meat consumption [RR = 1.58, 95% CI = 1.09 to 2.29 (19) and RR = 1.69, 95% CI = 1.17 to 2.45 (20), respectively] were similar to the multivariable relative risks [RR = 1.62, 95% CI = 1.08 to 2.41 (19) and RR = 1.66, 95% CI = 1.13 to 2.45 (20), respectively], indicating a lack of confounding from the considered covariates. Of the case–control studies, only approximately half controlled for age, sex, and education (or socioeconomic status) (22,26,30,35,38,40–42) or for smoking (29,35–41), and only four adjusted for fruit and vegetable consumption (22,23,25,40). Hence, the possibility of confounding from unaccounted nondietary and dietary risk factors for stomach cancer cannot be excluded.
A third limitation is that our results were likely to be affected by imprecise measurement of processed meat consumption and of potential confounders in the original studies. Categorization of main exposures and confounders that are measured with nondifferential error (independent of disease status) may induce differential misclassification and may bias the expected relative risk toward or away from the null value (50–52). Thus, misclassification of processed meat consumption and of potential confounders in the studies included in this meta-analysis might have led to an overestimate or an underestimate of the summary relative risks.
A fourth limitation is that the consumption levels in the lowest and highest categories and the range of consumption varied across studies. These differences may have contributed to the heterogeneity among studies in the analysis of the highest versus the lowest intake categories. To account for the different ranges of processed meat consumption, we estimated for each study (that provided the required data) a relative risk for an increase in intake of 30 g/day.
Finally, as with any meta-analysis, possible publication bias is of concern. Only studies of sausage consumption showed statistically significant publication bias. Thus, the summary results may be an overestimate of the relative risk of stomach cancer associated with sausage consumption.
A positive relationship between processed meat consumption and risk of stomach cancer is biologically plausible. In addition to having high amounts of salt, processed meat often contains nitrite or nitrate and N-nitroso compounds (7–9). N-Nitroso compounds are potent carcinogens that can induce tumors in various animal species at several sites (7,53). N-Nitroso compounds are also formed endogenously in the stomach through nitrosation by nitrites of amines or amides. Sugimura and Fujimura (54) showed that gastric carcinomas were induced in rats in response to oral administration of an N-nitroso compound, N-methyl-N'-nitro-N-nitrosoguanidine. Furthermore, several epidemiologic studies have reported a statistically significant positive association between N-nitrosodimethylamine or nitrosamine intake with stomach cancer risk (20,55–58). Although salt is not intrinsically carcinogenic, it may act as an irritant to the gastric mucosa, leading to gastritis, increased DNA synthesis, and cell proliferation (4–6). Excess salt also makes the mucosa cells more susceptible to carcinogens from foods. In experimental studies with rats, salt enhances the carcinogenic effects of gastric carcinogens, such as the N-nitroso compound N-methyl-N'-nitro-N-nitrosoguanidine (59–61).
It is possible that the relationship between processed meat consumption and stomach cancer risk is restricted to certain subgroups of the population, such as those with specific genetic polymorphisms coding for enzymes involved in the metabolism of N-nitroso compounds or those infected with H. pylori. In a nested case–control study within the European Prospective Investigation Into Cancer and Nutrition (19), the positive association between processed meat consumption and stomach cancer risk appeared to be restricted to H. pylori antibody–positive subjects. An association with processed meat consumption may also be modified by intake of vitamins C and E, fruits, and vegetables because these dietary factors can reduce the formation of N-nitroso compounds in the stomach (62). In the Swedish Mammography Cohort (20), the positive association of processed meat consumption with stomach cancer risk was somewhat stronger among women with a low consumption of fruits and vegetables.
Four studies investigated whether the association between processed meat (19,35,39) or sausage (26) consumption and stomach cancer risk differed by histological subtype (intestinal versus diffuse); none showed any appreciable difference. In the only study that examined risk by anatomical subsite, high consumption of processed meat was associated with an increased risk of gastric noncardia cancer (highest versus lowest intake category, RR = 1.92; 95% CI = 1.11 to 3.33) but not of gastric cardia cancer (highest versus lowest intake category, RR = 1.14; 95% CI = 0.52 to 2.49) (19).
In summary, findings of this meta-analysis support the hypothesis that high consumption of processed meat may be associated with an increased risk of stomach cancer. Future studies should control for more potential confounders and examine whether the association between processed meat consumption and stomach cancer risk is modified by other dietary factors, H. pylori infection status, or genetic polymorphisms. In addition, whether the association differs according to histologic subtype or anatomic subsite warrants further study.
|
NOTES |
|---|
|
TopNotesAbstractIntroductionSubjects and methodsResultsDiscussionReferences |
|---|
This study was supported by research grants from the Swedish Cancer Society (A. Wolk) and the Swedish Research Council/Longitudinal Studies (A. Wolk). The study sponsors had no role in the design, collection, analysis, or interpretation of the data, or in the writing or decision to submit the manuscript.
|
REFERENCES |
|---|
|
TopNotesAbstractIntroductionSubjects and methodsResultsDiscussionReferences |
|---|
(1) Parkin DM, Bray F, Ferlay J, Pisani P. Global cancer statistics, 2002. CA Cancer J Clin 2005;55:74–108.
(2) International Agency for Research on Cancer. IARC monographs on the evaluation of carcinogenic risks to humans: Vol 61. Schistosomes, liver flukes and Helicobacter pylori. Lyon (France): IARC Press; 1994.
(3) 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.
(4) Ames BN, Gold LS. Too many rodent carcinogens: mitogenesis increases mutagenesis. Science 1990;249:970–1.
(5) Charnley G, Tannenbaum SR. Flow cytometric analysis of the effect of sodium chloride on gastric cancer risk in the rat. Cancer Res 1985;45:5608–16.
(6) Fox JG, Dangler CA, Taylor NS, King A, Koh TJ, Wang TC. High-salt diet induces gastric epithelial hyperplasia and parietal cell loss, and enhances Helicobacter pylori colonization in C57BL/6 mice. Cancer Res 1999;59:4823–8.
(7) Tricker AR, Preussmann R. Carcinogenic N-nitrosamines in the diet: occurrence, formation, mechanisms and carcinogenic potential. Mutat Res 1991;259:277–89.[CrossRef][Web of Science][Medline]
(8) Hotchkiss JH. Preformed N-nitroso compounds in foods and beverages. Cancer Surv 1989;8:295–321.[Web of Science][Medline]
(9) Scanlan RA. Formation and occurrence of nitrosamines in food. Cancer Res 1983;43 5 Suppl:2435S–2440S.[Medline]
(10) Chyou PH, Nomura AM, Hankin JH, Stemmermann GN. A case-cohort study of diet and stomach cancer. Cancer Res 1990;50:7501–4.
(11) Nomura A, Grove JS, Stemmermann GN, Severson RK. A prospective study of stomach cancer and its relation to diet, cigarettes, and alcohol consumption. Cancer Res 1990;50:627–31.
(12) Kneller RW, McLaughlin JK, Bjelke E, Schuman LM, Blot WJ, Wacholder S, et al. A cohort study of stomach cancer in a high-risk American population. Cancer 1991;68:672–8.[CrossRef][Web of Science][Medline]
(13) Galanis DJ, Kolonel LN, Lee J, Nomura A. Intakes of selected foods and beverages and the incidence of gastric cancer among the Japanese residents of Hawaii: a prospective study. Int J Epidemiol 1998;27:173–80.
(14) Knekt P, Järvinen R, Dich J, Hakulinen T. Risk of colorectal and other gastro-intestinal cancers after exposure to nitrate, nitrite and N-nitroso compounds: a follow-up study. Int J Cancer 1999;80:852–6.[CrossRef][Web of Science][Medline]
(15) McCullough ML, Robertson AS, Jacobs EJ, Chao A, Calle EE, Thun MJ. A prospective study of diet and stomach cancer mortality in United States men and women. Cancer Epidemiol Biomarkers Prev 2001;10:1201–5.
(16) Ngoan LT, Mizoue T, Fujino Y, Tokui N, Yoshimura T. Dietary factors and stomach cancer mortality. Br J Cancer 2002;87:37–42.[CrossRef][Web of Science][Medline]
(17) 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]
(18) Tokui N, Yoshimura T, Fujino Y, Mizoue T, Hoshiyama Y, Yatsuya H, et al. Dietary habits and stomach cancer risk in the JACC Study. J Epidemiol 2005;15 Suppl 2:S98–108.[Medline]
(19) González CA, Jakszyn P, Pera G, Agudo A, Bingham S, Palli D, 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.
(20) 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;119:915–9.[CrossRef][Medline]
(21) Risch HA, Jain M, Choi NW, Fodor JG, Pfeiffer CJ, Howe GR, et al. Dietary factors and the incidence of cancer of the stomach. Am J Epidemiol 1985;122:947–59.
(22) La Vecchia C, Negri E, Decarli A, D'Avanzo B, Franceschi S. A case-control study of diet and gastric cancer in northern Italy. Int J Cancer 1987;40:484–9.[Web of Science][Medline]
(23) Lee HH, Wu HY, Chuang YC, Chang AS, Chao HH, Chen KY, et al. Epidemiologic characteristics and multiple risk factors of stomach cancer in Taiwan. Anticancer Res 1990;10:875–81.[Web of Science][Medline]
(24) Boeing H, Frentzel-Beyme R, Berger M, Berndt V, Gores W, Korner M, et al. Case-control study on stomach cancer in Germany. Int J Cancer 1991;47:858–64.[Web of Science][Medline]
(25) González CA, Sanz JM, Marcos G, Pita S, Brullet E, Saigi E, et al. Dietary factors and stomach cancer in Spain: a multi-centre case-control study. Int J Cancer 1991;49:513–9.[Web of Science][Medline]
(26) Boeing H, Jedrychowski W, Wahrendorf J, Popiela T, Tobiasz-Adamczyk B, Kulig A. Dietary risk factors in intestinal and diffuse types of stomach cancer: a multicenter case-control study in Poland. Cancer Causes Control 1991;2:227–33.[CrossRef][Web of Science][Medline]
(27) Tuyns AJ, Kaaks R, Haelterman M, Riboli E. Diet and gastric cancer: a case-control study in Belgium. Int J Cancer 1992;51:1–6.[Web of Science][Medline]
(28) Sanchez-Diez A, Hernandez-Mejia R, Cueto-Espinar A. Study of the relation between diet and gastric cancer in a rural area of the Province of Leon, Spain. Eur J Epidemiol 1992;8:233–7.[Web of Science][Medline]
(29) Hoshiyama Y, Sasaba T. A case-control study of stomach cancer and its relation to diet, cigarettes, and alcohol consumption in Saitama Prefecture, Japan. Cancer Causes Control 1992;3:441–8.[CrossRef][Web of Science][Medline]
(30) Hansson LE, Nyrén O, Bergström R, Wolk A, Lindgren A, Baron J, et al. Diet and risk of gastric cancer: a population-based case-control study in Sweden. Int J Cancer 1993;55:181–9.[Web of Science][Medline]
(31) Nazario CM, Szklo M, Diamond E, Roman-Franco A, Climent C, Suarez E, et al. Salt and gastric cancer: a case-control study in Puerto Rico. Int J Epidemiol 1993;22:790–7.
(32) Harrison LE, Zhang ZF, Karpeh MS, Sun M, Kurtz RC. The role of dietary factors in the intestinal and diffuse histologic subtypes of gastric adenocarcinoma: a case-control study in the U.S. Cancer 1997;80:1021–8.[CrossRef][Web of Science][Medline]
(33) Ward MH, Sinha R, Heineman EF, Rothman N, Markin R, Weisenburger DD, et al. Risk of adenocarcinoma of the stomach and esophagus with meat cooking method and doneness preference. Int J Cancer 1997;71:14–9.[CrossRef][Web of Science][Medline]
(34) Ji BT, Chow WH, Yang G, McLaughlin JK, Zheng W, Shu XO, et al. Dietary habits and stomach cancer in Shanghai, China. Int J Cancer 1998;76:659–64.[CrossRef][Web of Science][Medline]
(35) Ward MH, Lopez-Carrillo L. Dietary factors and the risk of gastric cancer in Mexico City. Am J Epidemiol 1999;149:925–32.
(36) Takezaki T, Gao CM, Wu JZ, Ding JH, Liu YT, Zhang Y, et al. Dietary protective and risk factors for esophageal and stomach cancers in a low-epidemic area for stomach cancer in Jiangsu Province, China: comparison with those in a high-epidemic area. Jpn J Cancer Res 2001;92:1157–65.[CrossRef][Web of Science]
(37) De Stefani E, Ronco A, Brennan P, Boffetta P. Meat consumption and risk of stomach cancer in Uruguay: a case-control study. Nutr Cancer 2001;40:103–7.[CrossRef][Web of Science][Medline]
(38) Chen H, Ward MH, Graubard BI, Heineman EF, Markin RM, Potischman NA, et al. Dietary patterns and adenocarcinoma of the esophagus and distal stomach. Am J Clin Nutr 2002;75:137–44.
(39) Ito LS, Inoue M, Tajima K, Yamamura Y, Kodera Y, Hirose K, et al. Dietary factors and the risk of gastric cancer among Japanese women: a comparison between the differentiated and non-differentiated subtypes. Ann Epidemiol 2003;13:24–31.[CrossRef][Web of Science][Medline]
(40) Nomura AM, Hankin JH, Kolonel LN, Wilkens LR, Goodman MT, Stemmermann GN. 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]
(41) Lissowska J, Gail MH, Pee D, Groves FD, Sobin LH, Nasierowska-Guttmejer A, et al. Diet and stomach cancer risk in Warsaw, Poland. Nutr Cancer 2004;48:149–59.[CrossRef][Web of Science][Medline]
(42) De Stefani E, Correa P, Boffetta P, Deneo-Pellegrini H, Ronco AL, Mendilaharsu M. Dietary patterns and risk of gastric cancer: a case-control study in Uruguay. Gastric Cancer 2004;7:211–20.[CrossRef][Medline]
(43) Greenland S. Quantitative methods in the review of epidemiologic literature. Epidemiol Rev 1987;9:1–30.[Medline]
(44) DerSimonian R, Laird N. Meta-analysis in clinical trials. Control Clin Trials 1986;7:177–88.[CrossRef][Web of Science][Medline]
(45) Greenland S, Longnecker MP. Methods for trend estimation from summarized dose-response data, with applications to meta-analysis. Am J Epidemiol 1992;135:1301–9.
(46) Orsini N, Bellocco R, Greenland S. Generalized least squares for trend estimation of summarized dose-response data. Stata J 2006;6:40–57.
(47) Norat T, Lukanova A, Ferrari P, Riboli E. Meat consumption and colorectal cancer risk: dose-response meta-analysis of epidemiological studies. Int J Cancer 2002;98:241–56.[CrossRef][Web of Science][Medline]
(48) Higgins JP, Thompson SG. Quantifying heterogeneity in a meta-analysis. Stat Med 2002;21:1539–58.[CrossRef][Web of Science][Medline]
(49) Egger M, Davey Smith G, Schneider M, Minder C. Bias in meta-analysis detected by a simple, graphical test. BMJ 1997;315:629–34.
(50) Wacholder S, Dosemeci M, Lubin JH. Blind assignment of exposure does not always prevent differential misclassification. Am J Epidemiol 1991;134:433–7.
(51) Flegal KM, Keyl PM, Nieto FJ. Differential misclassification arising from nondifferential errors in exposure measurement. Am J Epidemiol 1991;134:1233–44.
(52) Wacholder S. When measurement errors correlate with truth: surprising effects of nondifferential misclassification. Epidemiology 1995;6:157–61.[Web of Science][Medline]
(53) International Agency for Research on Cancer. IARC monographs on the evaluation of the carcinogenic risk of chemicals to humans: Vol 17. Some N-nitroso compounds. Lyon (France): IARC Press; 1978. p. 1–349.
(54) Sugimura T, Fujimura S. Tumour production in glandular stomach of rat by N-methyl-N'-nitro-N-nitrosoguanidine. Nature 1967;216:943–4.[CrossRef][Medline]
(55) González CA, Riboli E, Badosa J, Batiste E, Cardona T, Pita S, et al. Nutritional factors and gastric cancer in Spain. Am J Epidemiol 1994;139:466–73.
(56) Pobel D, Riboli E, Cornee J, Hemon B, Guyader M. Nitrosamine, nitrate and nitrite in relation to gastric cancer: a case-control study in Marseille, France. Eur J Epidemiol 1995;11:67–73.[CrossRef][Web of Science][Medline]
(57) La Vecchia C, D'Avanzo B, Airoldi L, Braga C, Decarli A. Nitrosamine intake and gastric cancer risk. Eur J Cancer Prev 1995;4:469–74.[Web of Science][Medline]
(58) De Stefani E, Boffetta P, Mendilaharsu M, Carzoglio J, Deneo-Pellegrini H. Dietary nitrosamines, heterocyclic amines, and risk of gastric cancer: a case-control study in Uruguay. Nutr Cancer 1998;30:158–62.[Web of Science][Medline]
(59) Takahashi M, Nishikawa A, Furukawa F, Enami T, Hasegawa T, Hayashi Y. Dose-dependent promoting effects of sodium chloride (NaCl) on rat glandular stomach carcinogenesis initiated with N-methyl-N'-nitro-N-nitrosoguanidine. Carcinogenesis 1994;15:1429–32.
(60) Tatematsu M, Takahashi M, Fukushima S, Hananouchi M, Shirai T. 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]
(61) Takahashi M, Kokubo T, Furukawa F, Kurokawa Y, Tatematsu M, Hayashi Y. Effect of high salt diet on rat gastric carcinogenesis induced by N-methyl-N'-nitro-N-nitrosoguanidine. Gann 1983;74:28–34.[Web of Science][Medline]
(62) Bartsch H, Ohshima H, Pignatelli B. Inhibitors of endogenous nitrosation: mechanisms and implications in human cancer prevention. Mutat Res 1988;202:307–24.[Web of Science][Medline]
Manuscript received December 23, 2005; revised May 11, 2006; accepted June 14, 2006.
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  P. T. Campbell, M. Sloan, and N. Kreiger Dietary Patterns and Risk of Incident Gastric Adenocarcinoma Am. J. Epidemiol., February 1, 2008; 167(3): 295 - 304. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||