This article was downloaded by: [New York University] On: 17 May 2015, At: 14:30 Publisher: Routledge Informa Ltd Registered in England and Wales Registered Number: 1072954 Registered office: Mortimer House, 37-41 Mortimer Street, London W1T 3JH, UK
Nutrition and Cancer Publication details, including instructions for authors and subscription information: http://www.tandfonline.com/loi/hnuc20
Dairy Consumption and Gastric Cancer Risk: A MetaAnalysis of Epidemiological Studies a
b
c
Yanjun Guo , Zhilei Shan , Hongyu Ren & Weihong Chen
a
a
MOE Key Lab of Environment and Health, School of Public Health, Tongji Medical College, Huazhong University of Science & Technology, Wuhan, P. R. China b
MOE Key Lab of Environment and Health, School of Public Health, Tongji Medical College, and Department of Nutrition and Food Hygiene, Hubei Key Laboratory of Food Nutrition and Safety, Huazhong University of Science & Technology, Wuhan, P. R. China c
Click for updates
Department of Gastroenterology, Union Hospital, Tongji Medical College, Huazhong University of Science & Technology, Wuhan, P.R. China Published online: 29 Apr 2015.
To cite this article: Yanjun Guo, Zhilei Shan, Hongyu Ren & Weihong Chen (2015) Dairy Consumption and Gastric Cancer Risk: A Meta-Analysis of Epidemiological Studies, Nutrition and Cancer, 67:4, 555-568, DOI: 10.1080/01635581.2015.1019634 To link to this article: http://dx.doi.org/10.1080/01635581.2015.1019634
PLEASE SCROLL DOWN FOR ARTICLE Taylor & Francis makes every effort to ensure the accuracy of all the information (the “Content”) contained in the publications on our platform. However, Taylor & Francis, our agents, and our licensors make no representations or warranties whatsoever as to the accuracy, completeness, or suitability for any purpose of the Content. Any opinions and views expressed in this publication are the opinions and views of the authors, and are not the views of or endorsed by Taylor & Francis. The accuracy of the Content should not be relied upon and should be independently verified with primary sources of information. Taylor and Francis shall not be liable for any losses, actions, claims, proceedings, demands, costs, expenses, damages, and other liabilities whatsoever or howsoever caused arising directly or indirectly in connection with, in relation to or arising out of the use of the Content. This article may be used for research, teaching, and private study purposes. Any substantial or systematic reproduction, redistribution, reselling, loan, sub-licensing, systematic supply, or distribution in any form to anyone is expressly forbidden. Terms & Conditions of access and use can be found at http:// www.tandfonline.com/page/terms-and-conditions
Nutrition and Cancer, 67(4), 555–568 Copyright Ó 2015, Taylor & Francis Group, LLC ISSN: 0163-5581 print / 1532-7914 online DOI: 10.1080/01635581.2015.1019634
Dairy Consumption and Gastric Cancer Risk: A Meta-Analysis of Epidemiological Studies Yanjun Guo MOE Key Lab of Environment and Health, School of Public Health, Tongji Medical College, Huazhong University of Science & Technology, Wuhan, P. R. China
Zhilei Shan MOE Key Lab of Environment and Health, School of Public Health, Tongji Medical College, and Department of Nutrition and Food Hygiene, Hubei Key Laboratory of Food Nutrition and Safety, Huazhong University of Science & Technology, Wuhan, P. R. China
Downloaded by [New York University] at 14:30 17 May 2015
Hongyu Ren Department of Gastroenterology, Union Hospital, Tongji Medical College, Huazhong University of Science & Technology, Wuhan, P.R. China
Weihong Chen MOE Key Lab of Environment and Health, School of Public Health, Tongji Medical College, Huazhong University of Science & Technology, Wuhan, P. R. China
Studies investigating the association of dairy consumption with gastric cancer risk have reported inconsistent findings. We conducted this systematic review and meta-analysis to review and summarize the epidemiologic evidence on the relation of total dairy and milk consumption with risk of gastric cancer. We summarized the available literature on this topic using metaanalysis of relative risks (RR) associated with total dairy and milk intake. The total of 17 case-control and 6 cohort studies (3256 cases) were eligible for inclusion. When comparing the highest with the lowest category of total dairy intake, the results of cohort studies indicated that increased consumption of total dairy food was associated with a reduced risk of gastric cancer (RR D 0.76; 95% CI: 0.64–0.91), whereas case-control studies provided no association. In subgroup analysis, significantly inverse associations between total diary food consumption and gastric cancer risk were observed in Europe subgroup (RR D 0.73; 95% CI: 0.54–0.99), U.S. subgroup (RR D 0.78; 95% CI: 0.63–0.98) but not in Asia subgroup. However, milk consumption was not associated with gastric cancer risk no matter in main or
Submitted 28 March 2014; accepted in final form 10 February 2015. Address correspondence to Weihong Chen, MOE Key Lab of Environment and Health, School of Public Health, Tongji Medical College, Huazhong University of Science & Technology, Wuhan 430030, Hubei, China. Phone: C86 27 83691677. Fax: C86 27 83692333. E-mail: [email protected] Color versions of one or more of the figures in the article can be found online at www.tandfonline.com/hnuc.
subgroup analysis. The results of cohort studies, but not casecontrol studies, suggested that total dairy might be related to the reduction of gastric cancer risk. Milk consumption was not associated with gastric cancer risk.
INTRODUCTION Although the incidence of gastric cancer has decreased markedly in the past few decades, it is still the fifth most common malignancy worldwide with an expected incidence of 952,000 cases in 2012 (1). Meanwhile, approximately 8% of total cases and 10% of annual cancer deaths worldwide are attributed to gastric cancer, which translates into a high fatality to case ratio of 70%, significantly higher than other prevalent cancers (2). Dairy products are rich in minerals (calcium, potassium, and magnesium) and vitamins (riboflavin and vitamin B-12) that can exert beneficial effects of cancer. In addition milk from ruminants contains conjugated linoleic acid (3), which was shown to have anticancer effects and to decrease growth of tumors in rodent models (4). On the other hand, high calcium and estrogen in dairy products may contribute to the etiology of cancer (5,6). Recently, systematic reviews have summarized the association, showing high intake of dairy products may be associated with a reduced risk of breast cancer (7), bladder cancer (8), and colorectal cancer (9), but an 555
556
Y. GUO ET AL.
Downloaded by [New York University] at 14:30 17 May 2015
increased risk of prostate cancer (10) and ovarian cancer (11). Numbers of epidemiologic studies (3,12–34) have examined the association between dairy products consumption and risk of gastric cancer. However, the relationship between dairy products consumption and gastric cancer risk remains controversial. In this study, we conducted a meta-analysis of published epidemiological studies with the following objectives: 1) to review and summarize the epidemiologic evidence on the relation of dairy consumption with risk of gastric cancer; and 2) to examine the dairy consumption and gastric cancer association according to study characteristics.
SEARCH STRATEGY We conducted a literature search of PubMed (Medline) and Embase up to September 2013 for studies examining the association between dairy intake and risk of gastric cancer. PubMed search terms were (“Gastric Cancer” [MeSH] or “Stomach Cancer” [MeSH]) and (“Dairy” [MeSH] or “Milk” [MeSH]). Similar search terms were used for Embase. In addition, we scrutinized references from relevant original papers and review articles to identify further pertinent studies. For inclusion in this analysis, the identified articles should be published in English and had to provide the odds ratio (OR) or relative risk (RR) and its corresponding 95% confidence interval (CI), or had sufficient data to permit their calculations.
DATA EXTRACTION The following data were extracted: the name of the first author, the year of publication, the country, study design, years of data collection, sample size, range of dairy and milk consumption, risk estimates with corresponding 95% CI comparing the highest with the lowest category, adjusted covariates, and exposure assessment. Quality assessment was performed according to the Newcastle-Ottawa Quality Assessment Scale (35), which is a validated scale for nonrandomized studies in meta-analyses. Data extraction was carried out independently by two authors (Yanjun Guo and Zhilei Shan) using a unanimous extraction form. To resolve discrepancies, group consensus and consulted with a third reviewer (Hongyu Ren) were used. We used adjusted RRs in cohort studies and ORs in case-control studies for meta-analysis. Adjusted RRs/ORs were extracted directly from the original reports.
STATISTICAL ANALYSIS Our main analyses were focused on the associations between consumption of total dairy food and milk and risk of gastric cancer. Any results stratified by sex were treated as two separated reports for analysis. Total dairy food was defined as skim/low-fat milk, whole/high-fat milk, yogurt, cottage, cheese, butter, and other dairy products. Milk was defined as
skim/low-fat milk and whole/high-fat milk. Because certain exposures, such as cheese, yogurt, or butter, were seldom assessed in individual reports, the analyses on above exposures were not performed. The measure of effect of interest is the RR and the corresponding 95% CI. Because the risk of gastric cancer is low, OR from case-control studies was used to accurately estimate RR. We report all results as RR for simplicity. Pooled RR with 95% CI based on both fixed effects and random-effects models were calculated, depending on the heterogeneity of the analysis. The Cochran Q test and the I2 statistic were used to estimate the heterogeneity between studies (36). Heterogeneity was confirmed with a significance level of P < 0.10. I2 statistic describes the percentage of total variation in point estimates that can be attributed to heterogeneity (37). Forest plot and funnel plot were used to observe the overall effect and assess the publication bias, respectively. A sensitivity analysis was used to evaluate the influence of each study by omitting one study at one time. All statistical analyses were performed with Stata version 12 (Stata Corp, College Station, TX) and all tests were two sided with a significant level of 0.05.
RESULTS Study Characteristics Fig. 1 summarizes the process of identifying and selecting studies. We identified 33 articles on dairy intake and the incidence of gastric cancer published between 1974 and 2013 (3,12–34,38–46). Upon closer examination, 10 articles did not provide sufficient information to estimate a summary odds ratio and its 95% CIs (32,38–46). The remaining 23 articles are presented in Table 1, including a total of 3256 cases. Four articles (3,13,22,30) were conducted to explore the associations of both total dairy and milk intake with gastric cancer, 9 (12,14,15,17–20,28,29) for total dairy intake and 10 (16,21,23–27,31,33,34) for milk intake. We included 6 cohort studies (3,28–31,33), 12 hospital-based case-control studies (12,13,15–17,20,22,24–27,34), and 5 population-based casecontrol studies (14,18,19,21,23) in the analyses. Of the selected studies, 12 studies were conducted before 2000 (3,18–27,33), 11 after 2000 (12–17,28–31,34); 11 in Asia (12,14–16,20,21,25,26,28,31,33), 5 in Europe (13,19,22,27,30), 5 in the United States (3,23,24,29,34), and 2 in South America (17, 18). Information on dairy or milk consumption was obtained by interview (12–27,33,34) or self-administered questionnaire (3,28–31).
Association Between Total Dairy Intake and Risk of Gastric Cancer Main Analysis Nine case-control studies and 4 cohort studies reported their results comparing the highest with the lowest categories of
Downloaded by [New York University] at 14:30 17 May 2015
DAIRY CONSUMPTION AND GASTRIC CANCER RISK
557
FIG. 1. Flow chart of study selection for meta-analysis.
total dairy food consumption in relation to gastric cancer risk. As shown in Fig. 2, the results of included studies were conflicting. No significant association was observed between dairy consumption and gastric cancer risk based on case-control studies: the pooled RR was 1.30 (95% CI: 0.91–1.85) in a random-effects analysis with a statistically significant heterogeneity across studies (I2 D 87.3%). However, there was an inverse association based on cohort studies: the summary RR was 0.76 (95% CI: 0.64–0.91) in a fixed-effects analysis with no heterogeneity (I2 D 0%). The overall RR was 1.09 (95% CI: 0.83– 1.43) for case-control and cohort studies combined with a statistically significant heterogeneity across studies (I2 D 85.5%).
consumption, significantly inverse associations with gastric cancer risk were observed in cohort subgroup (RR D 0.76; 95% CI: 0.64–0.91), Europe subgroup (RR D 0.73; 95% CI: 0.54–0.99), U.S. subgroup (RR D 0.78; 95% CI: 0.63–0.98), and self-administered questionnaire (RR D 0.76; 95% CI: 0.64–0.91) but not in other subgroups. We performed sensitivity analyses by sequentially excluding one study in each turn to examine the influence of a single study in each turn on the overall estimate. The results suggested the overall risk estimates were not substantially modified by any single study (Supplementary Fig. S1). The statistically significant heterogeneity never changed whichever study was excluded.
Subgroup and Sensitivity Analyses In Table 2, we pooled the RR estimates by study design (cohort, hospital-based case-control and population-based case-control studies), study location (United States, Europe and Asia), and exposure assessment (interview and self-administered questionnaire) using adjusted data, which reflected the greatest degree of control for confounders. For total diary food
Association Between Milk Intake and Risk of Gastric Cancer Main Analysis Ten case-control and 4 cohort studies on milk consumption and gastric cancer risk were identified. As shown in Fig. 3, the
558 Japan
Cohort
Cohort
Cohort
Cohort
Pham 2010
Park 2009
Pols 2007
Ngoan 2002
1995 to 2003
1988 to 2003
Study period
Japan
1989 to 1999
Great 1937 to Britain 2005
US
Country
Authors and Study publication year design Milk and dairy products
Items
Highest quintile vs. lowest quintile
Highest quartile vs. lowest quartile
Range of consumption
3/24374
Quantity assessment
Qualitative score (0–9 points)
(continued on next page)
Self-administered 6.5 questionnaire
Self-administered 7.5 Age, smoking, questionnaire history of gastric ulcer, stomach cancer screening, BMI, educational level, and total energy intake. Race, education, Self-administered 7.5 questionnaire marital status, BMI, family history of cancer, vigorous physical activity, alcohol consumption, intakes of red meat, total energy, smoking and antacid use. 8 7-d household Age, sex, and inventory energy, fruit, method and calcium intakes.
Variables of adjustment
All liquid milks, Dairy: 471 g/d vs. 89 g/d; infant Milk: 1.2 formulas, cups/d vs. cream, cheese, ice creams, and once/day vs. < 2 times/ week
Highest quartile vs. lowest quartile 150/7990 Milk, ice cream, > 5 times/week butter vs.17 times/week Age, gender, total calories, chili vs. 5 times/week vs. once/day vs. < 3 times/ month
Milk and cheese Highest vs. lowest tertile
Milk
Age and sex
Questionnaire interview
Age, sex, marital Questionnaire status, ethnic interview group and schooling year Age, sex, Questionnaire smoking interview history, education and county of residence Age, sex Questionnaire interview
>2 meals/week vs. < once/ week
Quantity assessment Questionnaire interview
Variables of adjustment
>5 times/week vs. < once/ week
Whole milk, 2% Daily vs. none milk and skim milk
Milk
Items
Range of consumption
GC D gastric cancer; BMI D body mass index; SES D social-economic status; GCA D gastric cardia adenocarcinoma; GNCA D gastric noncardia adenocarcinoma.
Vecchia 1987
Kono 1988
Mettlin 1990
Lee 1990
Country
Populationbased casecontrol study HospitalTaiwan based case control study HospitalUS based case control study
Authors and Study publication year design
TABLE 1 Characteristics of 23 studies included in this meta-analysis (Continued)
Downloaded by [New York University] at 14:30 17 May 2015
7
7
7
7
Qualitative score (0–9 points)
563
Study design Cohort studies Hospital–based casecontrol studies Population-based casecontrol studies Publish year After 2000 Before 2000 Study location Asia Europe United States Exposure assessment Questionnaire interview Self-administered questionnaire
Subgroup 0.76 (0.64–0.91) 1.16 (0.76–1.79) 0.88 (0.55–1.41)
1.08 (0.76–1.54) 1.08 (0.71–1.67) 1.16 (0.78–1.75) 0.73 (0.54–0.99) 0.78 (0.62–0.99) 1.30 (0.91–1.85) 0.76 (0.64–0.91)
3
8 5 5 4 2 9 3
RR (95% CI)
4 6
No. of studies
70.96 2.99
63.65 1.09 2.81
88.82 13.36
9.97
2.99 60.44
Heterogeneity
Total dairy food
87.30% 0.00%
90.60% 0.00% 28.90%
88.70% 70.10%
79.90%
0.00% 90.10%
I2 (%)
11 3
6 4 4
5 9
3
4 7
No. of studies
1.27 (0.93–1.75) 1.27 (0.87–1.86)
0.96 (0.79–1.16) 1.57 (0.86–2.88) 1.28 (0.98–1.67)
1.37 (0.75–2.50) 1.25 (0.92–1.71)
1.04 (0.72–1.52)
1.23 (0.95–1.59) 1.36 (0.83–2.23)
RR (95% CI)
Milk
39.46 4.26
4.87 9.39 13.2
14.07 29.72
1.82
4.31 36.95
Heterogeneity
TABLE 2 Relative risks (RR) of gastric cancer in relation to total dairy food and milk consumption according to study characteristics
Downloaded by [New York University] at 14:30 17 May 2015
74.70% 29.60%
0.00% 68.00% 77.30%
64.50% 73.10%
0.00%
7.10% 83.80%
I2 (%)
Downloaded by [New York University] at 14:30 17 May 2015
564
Y. GUO ET AL.
FIG. 2. Forest plots showing estimated relative risks (highest vs. lowest category) of gastric cancer associated with total dairy consumption.
results for association between milk intake and risk of gastric cancer in various studies were inconsistent. Compared with the lowest categories of milk intake, the summary RRs for the highest categories of milk intake were 1.29 (95% CI: 0.89– 1.88) in a random-effects analysis based on case-control studies and 1.23 (95% CI: 0.95–1.59) in a fixed-effects analysis based on cohort studies, respectively. There was significant heterogeneity between case-control studies (I2 D 76.9%), but low heterogeneity observed (I2 D 7.1%) between cohort studies. The overall RR was 1.28 (95% CI: 0.97–1.67) for casecontrol and cohort studies combined with a statistically significant heterogeneity across studies (I2 D 68.0%). Subgroup and Sensitivity Analyses As shown in Table 2, in subgroup analyses of milk intake and risk of gastric cancer, there was no association found in all strata of the study characteristics. The results of sensitivity analyses suggested the overall risk estimates did not substantially modified by any single study (Supplementary Fig. S2). The statistically significant heterogeneity never changed whichever study was excluded. Publication Bias There was no evidence of publication bias with regard to consumption of total dairy or milk in relation to gastric cancer
risk no matter based on case-control or cohort studies, as suggested by Begg rank correlation test and Egger linear regression test (all P > 0.05). DISCUSSION In our meta-analysis, we observed discrepancies between results of case-control and cohort studies concerning the relationship between total dairy consumption and gastric cancer risk. The result of cohort studies indicated that increased consumption of total dairy food might be associated with a reduced risk of gastric cancer, whereas case-control studies provided no association. Milk consumption was not associated with increased gastric cancer risk no matter in case-control studies or in cohort studies. Subgroup analyses based on limited numbers of studies suggested that the inverse associations between total dairy consumption and gastric cancer existed in Europe subgroup and U.S. subgroup but not in Asia subgroup. Some mechanisms exist to explain the protective effect of dairy on gastric cancer. First, phospholipids in dairy food may increase gastric cellular restitution (47), and both probiotics (48) and yogurt (49) may interfere with Helicobacter pylori colonization, which is considered as a key promoter of gastric antral and body cancer (50,51). In addition, dairy could have its protective effect may attribute to some ingredients that have anticarcinogenic properties, such as vitamin D, calcium,
Downloaded by [New York University] at 14:30 17 May 2015
DAIRY CONSUMPTION AND GASTRIC CANCER RISK
565
FIG. 3. Forest plots showing estimated relative risks (highest vs. lowest category) of gastric cancer associated with milk consumption.
and b-carotene (52–54). Vitamin D, which is tightly related to calcium regulation in the body, has been inversely associated with digestive system cancers in another study (54). The metaanalysis based on the data of randomized control trials suggested a significantly decreased risk for stomach cancer in individuals supplemented with doses of b-carotene below 20 mg per day in comparison with the controls. However, no overall association was indicated in this study, and there was even a significantly increased risk of gastric cancer in individuals with high doses of b-carotene supplementary (55). In addition, no association was observed between milk intake and risk of gastric cancer in this study. Associations between milk and gastric cancer risk may have been underestimated in case-control studies for that gastric cancer cases may drink more milk to control symptoms of the disease such as dyspepsia. Furthermore, in the subgroup analyses, the inverse associations between total dairy consumption and gastric cancer were found in Europe subgroup and U.S. subgroup but not in Asia subgroup. The different observations may be explained, at least in part, by the variations of dairy consumption across the world. Higher dairy consumption is found in Europe and North America when compared with Asia (56,57). In addition, a possible role of ethnic differences in genetic backgrounds should also be taken into account. For example, lactose intolerance is at its highest frequency in some parts of East Asia,
whereas northern Europeans generally have the lowest frequency of this dietary problem (58). There are several possible explanations why case-control studies and cohort studies provide conflicting results pertaining to the association between total dairy intake and gastric cancer. In case-control studies, dietary intake is assessed after the cancer diagnosis. Consequently, these studies may be subject to recall bias and inaccurate measurements of dietary intake attributed to changes in the dietary practices among cases after the cancers were diagnosed. Even though the cases were inquired about their diet-before-disease, their recall may well be influenced by their current diet. Another difference between case-control and cohort studies may be attributed to the time interval between the period covered by the questionnaire and the cancer diagnosis. The time interval of dietary intake is usually 1 or 2 years before diagnose in case-control studies, whereas in cohort studies it could be as long as 10– 20 years (recent diet at baseline). If the diet over the past 10–20 years before the diagnosis of gastric cancer may be most relevant and if any change in dietary patterns have occurred, an association with dairy food consumption may be hidden in case-control studies. It is also likely that long-term dietary intake may be most relevant to chronic diseases like cancer with a long duration of development.
Downloaded by [New York University] at 14:30 17 May 2015
566
Y. GUO ET AL.
However, contrary to our results, high intake of dairy products was found to be associated with increased risk of some specific cancers (11,59). Increased plasma calcium due to intake of dairy or milk could suppress the production of plasma 1,25-dihydroxyvitamin D3, which might be one possible mechanism underlying the association (6). Moreover, high dairy or milk consumption could increase blood levels of insulin-like growth factor-I (60), which has been associated with cancer risk (61,62). Several potential limitations of our study are worth considering. First, this meta-analysis is based on observational studies, and the inherent limitations of such studies may affect our findings. The possibility of bias and uncontrolled or residual confounding cannot be excluded. Although most studies included in the meta-analysis controlled possible confounding factors, inherent problems with the variation of confounders in the studies included should be addressed. Most articles included in this meta-analysis had adjusted for major potential confounders, including age, gender, alcohol consumption, body mass index, and history of family gastric cancer, but only 4 studies had adjusted for total energy intake and one study included the confounder of H. pylori. The inconsistent adjustments among different studies may lead to bias of the pooled results. However, there was no difference between the study adjusting for H. pylori and the pooled results of other studies. Because H. pylori infection is such an important risk factor for gastric cancer, further studies are needed to explore the interactions between dairy intake and H. pylori infection on gastric cancer. Second, dairy and milk products are a collection of several products; the ambiguous or different definition in each questionnaire of study may result in inaccurate estimates. Inevitably, dietary assessments suffer from measurement errors. Within single studies, nondifferential misclassification may occur in classifying categories of dairy product consumption. Across studies, differential misclassification may be introduced as dietary assessments were based on different questionnaires and different nutrient databases. Third, most of the included studies defined patients with all types of gastric cancers as gastric cancer cases and did not examine the association by gastric cancer subtype. It is possible that some associations that may exist only for certain tumor subtypes. In addition, we could not rule out the influences of diet change on the risk estimates among participants with sub-clinical gastric cancer. Few studies assessed dairy and milk consumption more than once during the followup periods, even in cohort studies. Finally, potential publication bias might influence the findings, yet we found no evidence of publication bias involved in our findings. In conclusion, the results of this meta-analysis provided some support for the hypothesis that total dairy food may have a potential protective effect for gastric cancer. However, this evidence is largely limited to cohort studies, whereas case-
control studies have produced conflicting results. Further research is warranted to confirm these findings and elucidate the likely biological mechanisms. ACKNOWLEDGMENTS We gratefully acknowledge the directions of Prof. Liegang Liu and Dr Wei Bao. FUNDING This work was supported by National Basic Research Program (Grant 2011CB503804). SUPPLEMENTAL DATA Supplemental data for this article can be accessed on the publisher’s website. REFERENCES 1. World Health Organization: GLOBOCAN 2012: estimated incidence, mortality and prevalence worldwide in 2012. Available from http://globo can.iarc.fr/Pages/fact_sheets_cancer.aspx. 2. Guggenheim DE and Shah MA: Gastric cancer epidemiology and risk factors. J Surg Oncol 107, 230–236, 2013. 3. Kneller RW, McLaughlin JK, Bjelke E, Schuman LM, Blot WJ, et al.: A cohort study of stomach cancer in a high-risk American population. Cancer 68, 672–678, 1991. 4. Lampe JW: Dairy products and cancer. J Am Coll Nutr 30(5, Suppl 1), 464S–470S, 2011. 5. Qin LQ, Wang PY, Kaneko T, Hoshi K, and Sato A: Estrogen: one of the risk factors in milk for prostate cancer. Med Hypotheses 62, 133–142, 2004. 6. Chan JM and Giovannucci EL: Dairy products, calcium, and vitamin D and risk of prostate cancer. Epidemiol Rev 23, 87–92, 2001. 7. Dong JY, Zhang L, He K, and Qin LQ: Dairy consumption and risk of breast cancer: a meta-analysis of prospective cohort studies. Breast Cancer Res Treat 127, 23–31, 2011. 8. Mao QQ, Dai Y, Lin YW, Qin J, Xie LP, et al.: Milk consumption and bladder cancer risk: a meta-analysis of published epidemiological studies. Nutr Cancer 63, 1263–1271, 2011. 9. Aune D, Lau R, Chan DS, Vieira R, Greenwood DC, et al.: Dairy products and colorectal cancer risk: a systematic review and meta-analysis of cohort studies. Ann Oncol 23, 37–45, 2012. 10. Gao X, LaValley MP, and Tucker KL: Prospective studies of dairy product and calcium intakes and prostate cancer risk: a meta-analysis. J Natl Cancer Inst 97, 1768–1777, 2005. 11. Genkinger JM, Hunter DJ, Spiegelman D, Anderson KE, Arslan A, et al.: Dairy products and ovarian cancer: a pooled analysis of 12 cohort studies. Cancer Epidemiol Biomarkers Prev 15, 364–372, 2006. 12. Gao Y, Hu N, Han XY, Ding T, Giffen C, et al.: Risk factors for esophageal and gastric cancers in Shanxi Province, China: a case-control study. Cancer Epidemiol 35, e91–e99, 2011. 13. Lazarevic K, Nagorni A, Rancic N, Milutinovic S, Stosic L, et al.: Dietary factors and gastric cancer risk: hospital-based case control study. J BUON 15, 89–93, 2010. 14. Pourfarzi F, Whelan A, Kaldor J, and Malekzadeh R: The role of diet and other environmental factors in the causation of gastric cancer in Iran—a population based study. Int J Cancer 125, 1953–1960, 2009. 15. Fei SJ and Xiao SD: Diet and gastric cancer: a case-control study in Shanghai urban districts. Chin J Dig Dis 7, 83–88, 2006.
Downloaded by [New York University] at 14:30 17 May 2015
DAIRY CONSUMPTION AND GASTRIC CANCER RISK 16. Kim HJ, Chang WK, Kim MK, Lee SS, and Choi BY: Dietary factors and gastric cancer in Korea: a case-control study. Int J Cancer 97, 531–535, 2002. 17. Munoz N, Plummer M, Vivas J, Moreno V, De Sanjose S, et al.: A casecontrol study of gastric cancer in Venezuela. Int J Cancer 93, 417–423, 2001. 18. Ward MH and Lopez-Carrillo L: Dietary factors and the risk of gastric cancer in Mexico City. Am J Epidemiol 149, 925–932, 1999. 19. Hansson LE, Nyren O, Bergstrom R, Wolk A, Lindgren A, et al.: Diet and risk of gastric cancer. A population-based case-control study in Sweden. Int J Cancer 55, 181–189, 1993. 20. Hoshiyama Y and 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 3, 441–448, 1992. 21. Yu GP and Hsieh CC: Risk factors for stomach cancer: a populationbased case-control study in Shanghai. Cancer Causes Control2, 169–174, 1991. 22. Boeing H, Frentzel-Beyme R, Berger M, Berndt V, Gores W, et al.: Casecontrol study on stomach cancer in Germany. Int J Cancer 47, 858–864, 1991. 23. Wu-Williams AH, Yu MC, and Mack TM: Life-style, workplace, and stomach cancer by subsite in young men of Los Angeles County. Cancer Res 50, 2569–2576, 1990. 24. Mettlin CJ, Schoenfeld ER, and Natarajan N: Patterns of milk consumption and risk of cancer. Nutr Cancer 13(1–2), 89–99, 1990. 25. Lee HH, Wu HY, Chuang YC, Chang AS, Chao HH, et al.: Epidemiologic characteristics and multiple risk factors of stomach cancer in Taiwan. Anticancer Res 10, 875–881. 26. Kono S, Ikeda M, Tokudome S, and Kuratsune M: A case-control study of gastric cancer and diet in northern Kyushu, Japan. Jpn J Cancer Res 79, 1067–1074, 1988. 27. La Vecchia C, Negri E, Decarli A, D’Avanzo B, and Franceschi S: A case-control study of diet and gastric cancer in northern Italy. Int J Cancer 40, 484–489, 1987. 28. Pham TM, Fujino Y, Kikuchi S, Tamakoshi A, Matsuda S, et al.: Dietary patterns and risk of stomach cancer mortality: the Japan collaborative cohort study. Ann Epidemiol 20, 356–363, 2010. 29. Park Y, Leitzmann MF, Subar AF, Hollenbeck A, and Schatzkin A: Dairy food, calcium, and risk of cancer in the NIH-AARP Diet and Health Study. Arch Intern Med 169, 391–401, 2009. 30. van der Pols JC, Bain C, Gunnell D, Smith GD, Frobisher C, et al.: Childhood dairy intake and adult cancer risk: 65-y follow-up of the Boyd Orr cohort. Am J Clin Nutr 86, 1722–1729, 2007. 31. Ngoan LT, Mizoue T, Fujino Y, Tokui N, and Yoshimura T: Dietary factors and stomach cancer mortality. Br J Cancer 87, 37–42, 2002. 32. Ursin G, Bjelke E, Heuch I, and Vollset SE: Milk consumption and cancer incidence: a Norwegian prospective study. Br J Cancer 61, 454–459, 1990. 33. Nomura A, Grove JS, Stemmermann GN, and Severson RK: A prospective study of stomach cancer and its relation to diet, cigarettes, and alcohol consumption. Cancer Res 50, 627–631, 1990. 34. Chen H, Ward MH, Graubard BI, Heineman EF, Markin RM, et al.: Dietary patterns and adenocarcinoma of the esophagus and distal stomach. Am J Clin Nutr 75, 137–144, 2002. 35. Wells GA, Shea B, O’Connell D, Peterson J, Welch V, et al.: The Newcastle-Ottawa Scale (NOS) for assessing the quality of nonrandomized studies in meta-analyses. Accessed from www.ohri.ca/programs/clinica l_epidemiology/oxford.asp, 2011. 36. Higgins JP, Thompson SG, Deeks JJ, and Altman DG: Measuring inconsistency in meta-analyses. BMJ 327, 557–560, 2003. 37. Higgins JP: Commentary: Heterogeneity in meta-analysis should be expected and appropriately quantified. Int J Epidemiol 37, 1158–1160, 2008.
567
38. Xibin S, Moller H, Evans HS, Dixing D, Wenjie D, et al.: Residential environment, diet and risk of stomach cancer: a case-control study in Linzhou, China. Asian Pac J Cancer Prev 3, 167–172, 2002. 39. Zhuo XG and Watanabe S: Factor analysis of digestive cancer mortality and food consumption in 65 Chinese counties. J Epidemiol 9, 275–284, 1999. 40. Tuyns AJ, Kaaks R, Haelterman M, and Riboli E: Diet and gastric cancer. A case-control study in Belgium. Int J Cancer 51, 1–6, 1992. 41. Gonzalez CA, Sanz JM, Marcos G, Pita S, Brullet E, et al.: Dietary factors and stomach cancer in Spain: a multi-centre case-control study. Int J Cancer 49, 513–519, 1990. 42. Demirer T, Icli F, Uzunalimoglu O, and Kucuk O: Diet and stomach cancer incidence. A case-control study in Turkey. Cancer 65, 2344–2348, 1990. 43. Buiatti E, Palli D, Decarli A, Amadori D, Avellini C, et al.: A case-control study of gastric cancer and diet in Italy. Int J Cancer 44, 611–616, 1989. 44. Hu JF, Zhang SF, Jia EM, Wang QQ, Liu SD, et al.: Diet and cancer of the stomach: a case-control study in China. Int J Cancer 41, 331–335, 1988. 45. Correa P, Fontham E, Pickle LW, Chen V, Lin YP, et al.: Dietary determinants of gastric cancer in south Louisiana inhabitants. J Natl Cancer Inst 75, 645–654, 1985. 46. Nagai M, Hashimoto T, Yanagawa H, Yokoyama H, and Minowa M: Relationship of diet to the incidence of esophageal and stomach cancer in Japan. Nutr Cancer 3,257–268, 1982. 47. Dial EJ and Lichtenberger LM: A role for milk phospholipids in protection against gastric acid. Studies in adult and suckling rats. Gastroenterology 87, 379–385, 1984. 48. Kabir AM, Aiba Y, Takagi A, Kamiya S, Miwa T, et al.: Prevention of Helicobacter pylori infection by lactobacilli in a gnotobiotic murine model. Gut 41, 49–55, 1997. 49. Sheu BS, Cheng HC, Kao AW, Wang ST, Yang YJ, et al.: Pretreatment with Lactobacillus-and Bifidobacterium-containing yogurt can improve the efficacy of quadruple therapy in eradicating residual Helicobacter pylori infection after failed triple therapy. Am J Clin Nutr 83, 864–869, 2006. 50. Lee SA, Kang D, Shim KN, Choe JW, Hong WS, et al.: Effect of diet and Helicobacter pylori infection to the risk of early gastric cancer. J Epidemiol 13, 162–168, 2003. 51. Singh K and Ghoshal UC: Causal role of Helicobacter pylori infection in gastric cancer: an Asian enigma. World J Gastroenterol 12, 1346–1351, 2006. 52. Peterlik M, Grant WB, and Cross HS: Calcium, vitamin D and cancer. Anticancer Res 29, 3687–3698, 2009. 53. Garland CF, Gorham ED, Mohr SB, and Garland FC: Vitamin D for cancer prevention: global perspective. Ann Epidemiol 19, 468–483, 2009. 54. Giovannucci E, Liu Y, Rimm EB, Hollis BW, Fuchs CS, et al.: Prospective study of predictors of vitamin D status and cancer incidence and mortality in men. J Natl Cancer Inst 98, 451–459, 2006. 55. Druesne-Pecollo N, Latino-Martel P, Norat T, Barrandon E, Bertrais S, et al.: Beta-carotene supplementation and cancer risk: a systematic review and metaanalysis of randomized controlled trials. Int J Cancer 127, 172– 184, 2010. 56. Ma Q, Xin J, Zhao Z, Guo Q, Yu S, et al.: Value of 18F-FDG PET/CT in the diagnosis of primary gastric cancer via stomach distension. European Journal of Radiology 82, e302–e306, 2013. 57. Wang Y and Li S: Worldwide trends in dairy production and consumption and calcium intake: is promoting consumption of dairy products a sustainable solution for inadequate calcium intake? Food Nutr Bull 29, 172–185, 2008. 58. Wang XQ, Yan H, Terry PD, Wang JS, Cheng L, et al.: Interaction between dietary factors and helicobacter pylori infection in noncardia gastric cancer: A population-based case-control study in China. J Am Coll Nutr 31, 375–384, 2012.
568
Y. GUO ET AL.
Downloaded by [New York University] at 14:30 17 May 2015
59. Larsson SC, Bergkvist L, and Wolk A: Milk and lactose intakes and ovarian cancer risk in the Swedish Mammography Cohort. Am J Clin Nutr 80, 1353–1357, 2004. 60. Giovannucci E, Pollak M, Liu Y, Platz EA, Majeed N, et al.: Nutritional predictors of insulin-like growth factor I and their relationships to cancer in men. Cancer Epidemiol Biomarkers Prev 12, 84–89, 2003.
61. Renehan AG, Zwahlen M, Minder C, O’Dwyer ST, Shalet SM, et al.: Insulin-like growth factor (IGF)-I, IGF binding protein-3, and cancer risk: systematic review and meta-regression analysis. Lancet 363(9418), 1346– 1353, 2004. 62. Lukanova A, Lundin E, Toniolo P, Micheli A, Akhmedkhanov A, et al.: Circulating levels of insulin-like growth factor-I and risk of ovarian cancer. Int J Cancer 101, 549–554, 2002.
Nutrition and Cancer Publication details, including instructions for authors and subscription information: http://www.tandfonline.com/loi/hnuc20
Dairy Consumption and Gastric Cancer Risk: A MetaAnalysis of Epidemiological Studies a
b
c
Yanjun Guo , Zhilei Shan , Hongyu Ren & Weihong Chen
a
a
MOE Key Lab of Environment and Health, School of Public Health, Tongji Medical College, Huazhong University of Science & Technology, Wuhan, P. R. China b
MOE Key Lab of Environment and Health, School of Public Health, Tongji Medical College, and Department of Nutrition and Food Hygiene, Hubei Key Laboratory of Food Nutrition and Safety, Huazhong University of Science & Technology, Wuhan, P. R. China c
Click for updates
Department of Gastroenterology, Union Hospital, Tongji Medical College, Huazhong University of Science & Technology, Wuhan, P.R. China Published online: 29 Apr 2015.
To cite this article: Yanjun Guo, Zhilei Shan, Hongyu Ren & Weihong Chen (2015) Dairy Consumption and Gastric Cancer Risk: A Meta-Analysis of Epidemiological Studies, Nutrition and Cancer, 67:4, 555-568, DOI: 10.1080/01635581.2015.1019634 To link to this article: http://dx.doi.org/10.1080/01635581.2015.1019634
PLEASE SCROLL DOWN FOR ARTICLE Taylor & Francis makes every effort to ensure the accuracy of all the information (the “Content”) contained in the publications on our platform. However, Taylor & Francis, our agents, and our licensors make no representations or warranties whatsoever as to the accuracy, completeness, or suitability for any purpose of the Content. Any opinions and views expressed in this publication are the opinions and views of the authors, and are not the views of or endorsed by Taylor & Francis. The accuracy of the Content should not be relied upon and should be independently verified with primary sources of information. Taylor and Francis shall not be liable for any losses, actions, claims, proceedings, demands, costs, expenses, damages, and other liabilities whatsoever or howsoever caused arising directly or indirectly in connection with, in relation to or arising out of the use of the Content. This article may be used for research, teaching, and private study purposes. Any substantial or systematic reproduction, redistribution, reselling, loan, sub-licensing, systematic supply, or distribution in any form to anyone is expressly forbidden. Terms & Conditions of access and use can be found at http:// www.tandfonline.com/page/terms-and-conditions
Nutrition and Cancer, 67(4), 555–568 Copyright Ó 2015, Taylor & Francis Group, LLC ISSN: 0163-5581 print / 1532-7914 online DOI: 10.1080/01635581.2015.1019634
Dairy Consumption and Gastric Cancer Risk: A Meta-Analysis of Epidemiological Studies Yanjun Guo MOE Key Lab of Environment and Health, School of Public Health, Tongji Medical College, Huazhong University of Science & Technology, Wuhan, P. R. China
Zhilei Shan MOE Key Lab of Environment and Health, School of Public Health, Tongji Medical College, and Department of Nutrition and Food Hygiene, Hubei Key Laboratory of Food Nutrition and Safety, Huazhong University of Science & Technology, Wuhan, P. R. China
Downloaded by [New York University] at 14:30 17 May 2015
Hongyu Ren Department of Gastroenterology, Union Hospital, Tongji Medical College, Huazhong University of Science & Technology, Wuhan, P.R. China
Weihong Chen MOE Key Lab of Environment and Health, School of Public Health, Tongji Medical College, Huazhong University of Science & Technology, Wuhan, P. R. China
Studies investigating the association of dairy consumption with gastric cancer risk have reported inconsistent findings. We conducted this systematic review and meta-analysis to review and summarize the epidemiologic evidence on the relation of total dairy and milk consumption with risk of gastric cancer. We summarized the available literature on this topic using metaanalysis of relative risks (RR) associated with total dairy and milk intake. The total of 17 case-control and 6 cohort studies (3256 cases) were eligible for inclusion. When comparing the highest with the lowest category of total dairy intake, the results of cohort studies indicated that increased consumption of total dairy food was associated with a reduced risk of gastric cancer (RR D 0.76; 95% CI: 0.64–0.91), whereas case-control studies provided no association. In subgroup analysis, significantly inverse associations between total diary food consumption and gastric cancer risk were observed in Europe subgroup (RR D 0.73; 95% CI: 0.54–0.99), U.S. subgroup (RR D 0.78; 95% CI: 0.63–0.98) but not in Asia subgroup. However, milk consumption was not associated with gastric cancer risk no matter in main or
Submitted 28 March 2014; accepted in final form 10 February 2015. Address correspondence to Weihong Chen, MOE Key Lab of Environment and Health, School of Public Health, Tongji Medical College, Huazhong University of Science & Technology, Wuhan 430030, Hubei, China. Phone: C86 27 83691677. Fax: C86 27 83692333. E-mail: [email protected] Color versions of one or more of the figures in the article can be found online at www.tandfonline.com/hnuc.
subgroup analysis. The results of cohort studies, but not casecontrol studies, suggested that total dairy might be related to the reduction of gastric cancer risk. Milk consumption was not associated with gastric cancer risk.
INTRODUCTION Although the incidence of gastric cancer has decreased markedly in the past few decades, it is still the fifth most common malignancy worldwide with an expected incidence of 952,000 cases in 2012 (1). Meanwhile, approximately 8% of total cases and 10% of annual cancer deaths worldwide are attributed to gastric cancer, which translates into a high fatality to case ratio of 70%, significantly higher than other prevalent cancers (2). Dairy products are rich in minerals (calcium, potassium, and magnesium) and vitamins (riboflavin and vitamin B-12) that can exert beneficial effects of cancer. In addition milk from ruminants contains conjugated linoleic acid (3), which was shown to have anticancer effects and to decrease growth of tumors in rodent models (4). On the other hand, high calcium and estrogen in dairy products may contribute to the etiology of cancer (5,6). Recently, systematic reviews have summarized the association, showing high intake of dairy products may be associated with a reduced risk of breast cancer (7), bladder cancer (8), and colorectal cancer (9), but an 555
556
Y. GUO ET AL.
Downloaded by [New York University] at 14:30 17 May 2015
increased risk of prostate cancer (10) and ovarian cancer (11). Numbers of epidemiologic studies (3,12–34) have examined the association between dairy products consumption and risk of gastric cancer. However, the relationship between dairy products consumption and gastric cancer risk remains controversial. In this study, we conducted a meta-analysis of published epidemiological studies with the following objectives: 1) to review and summarize the epidemiologic evidence on the relation of dairy consumption with risk of gastric cancer; and 2) to examine the dairy consumption and gastric cancer association according to study characteristics.
SEARCH STRATEGY We conducted a literature search of PubMed (Medline) and Embase up to September 2013 for studies examining the association between dairy intake and risk of gastric cancer. PubMed search terms were (“Gastric Cancer” [MeSH] or “Stomach Cancer” [MeSH]) and (“Dairy” [MeSH] or “Milk” [MeSH]). Similar search terms were used for Embase. In addition, we scrutinized references from relevant original papers and review articles to identify further pertinent studies. For inclusion in this analysis, the identified articles should be published in English and had to provide the odds ratio (OR) or relative risk (RR) and its corresponding 95% confidence interval (CI), or had sufficient data to permit their calculations.
DATA EXTRACTION The following data were extracted: the name of the first author, the year of publication, the country, study design, years of data collection, sample size, range of dairy and milk consumption, risk estimates with corresponding 95% CI comparing the highest with the lowest category, adjusted covariates, and exposure assessment. Quality assessment was performed according to the Newcastle-Ottawa Quality Assessment Scale (35), which is a validated scale for nonrandomized studies in meta-analyses. Data extraction was carried out independently by two authors (Yanjun Guo and Zhilei Shan) using a unanimous extraction form. To resolve discrepancies, group consensus and consulted with a third reviewer (Hongyu Ren) were used. We used adjusted RRs in cohort studies and ORs in case-control studies for meta-analysis. Adjusted RRs/ORs were extracted directly from the original reports.
STATISTICAL ANALYSIS Our main analyses were focused on the associations between consumption of total dairy food and milk and risk of gastric cancer. Any results stratified by sex were treated as two separated reports for analysis. Total dairy food was defined as skim/low-fat milk, whole/high-fat milk, yogurt, cottage, cheese, butter, and other dairy products. Milk was defined as
skim/low-fat milk and whole/high-fat milk. Because certain exposures, such as cheese, yogurt, or butter, were seldom assessed in individual reports, the analyses on above exposures were not performed. The measure of effect of interest is the RR and the corresponding 95% CI. Because the risk of gastric cancer is low, OR from case-control studies was used to accurately estimate RR. We report all results as RR for simplicity. Pooled RR with 95% CI based on both fixed effects and random-effects models were calculated, depending on the heterogeneity of the analysis. The Cochran Q test and the I2 statistic were used to estimate the heterogeneity between studies (36). Heterogeneity was confirmed with a significance level of P < 0.10. I2 statistic describes the percentage of total variation in point estimates that can be attributed to heterogeneity (37). Forest plot and funnel plot were used to observe the overall effect and assess the publication bias, respectively. A sensitivity analysis was used to evaluate the influence of each study by omitting one study at one time. All statistical analyses were performed with Stata version 12 (Stata Corp, College Station, TX) and all tests were two sided with a significant level of 0.05.
RESULTS Study Characteristics Fig. 1 summarizes the process of identifying and selecting studies. We identified 33 articles on dairy intake and the incidence of gastric cancer published between 1974 and 2013 (3,12–34,38–46). Upon closer examination, 10 articles did not provide sufficient information to estimate a summary odds ratio and its 95% CIs (32,38–46). The remaining 23 articles are presented in Table 1, including a total of 3256 cases. Four articles (3,13,22,30) were conducted to explore the associations of both total dairy and milk intake with gastric cancer, 9 (12,14,15,17–20,28,29) for total dairy intake and 10 (16,21,23–27,31,33,34) for milk intake. We included 6 cohort studies (3,28–31,33), 12 hospital-based case-control studies (12,13,15–17,20,22,24–27,34), and 5 population-based casecontrol studies (14,18,19,21,23) in the analyses. Of the selected studies, 12 studies were conducted before 2000 (3,18–27,33), 11 after 2000 (12–17,28–31,34); 11 in Asia (12,14–16,20,21,25,26,28,31,33), 5 in Europe (13,19,22,27,30), 5 in the United States (3,23,24,29,34), and 2 in South America (17, 18). Information on dairy or milk consumption was obtained by interview (12–27,33,34) or self-administered questionnaire (3,28–31).
Association Between Total Dairy Intake and Risk of Gastric Cancer Main Analysis Nine case-control studies and 4 cohort studies reported their results comparing the highest with the lowest categories of
Downloaded by [New York University] at 14:30 17 May 2015
DAIRY CONSUMPTION AND GASTRIC CANCER RISK
557
FIG. 1. Flow chart of study selection for meta-analysis.
total dairy food consumption in relation to gastric cancer risk. As shown in Fig. 2, the results of included studies were conflicting. No significant association was observed between dairy consumption and gastric cancer risk based on case-control studies: the pooled RR was 1.30 (95% CI: 0.91–1.85) in a random-effects analysis with a statistically significant heterogeneity across studies (I2 D 87.3%). However, there was an inverse association based on cohort studies: the summary RR was 0.76 (95% CI: 0.64–0.91) in a fixed-effects analysis with no heterogeneity (I2 D 0%). The overall RR was 1.09 (95% CI: 0.83– 1.43) for case-control and cohort studies combined with a statistically significant heterogeneity across studies (I2 D 85.5%).
consumption, significantly inverse associations with gastric cancer risk were observed in cohort subgroup (RR D 0.76; 95% CI: 0.64–0.91), Europe subgroup (RR D 0.73; 95% CI: 0.54–0.99), U.S. subgroup (RR D 0.78; 95% CI: 0.63–0.98), and self-administered questionnaire (RR D 0.76; 95% CI: 0.64–0.91) but not in other subgroups. We performed sensitivity analyses by sequentially excluding one study in each turn to examine the influence of a single study in each turn on the overall estimate. The results suggested the overall risk estimates were not substantially modified by any single study (Supplementary Fig. S1). The statistically significant heterogeneity never changed whichever study was excluded.
Subgroup and Sensitivity Analyses In Table 2, we pooled the RR estimates by study design (cohort, hospital-based case-control and population-based case-control studies), study location (United States, Europe and Asia), and exposure assessment (interview and self-administered questionnaire) using adjusted data, which reflected the greatest degree of control for confounders. For total diary food
Association Between Milk Intake and Risk of Gastric Cancer Main Analysis Ten case-control and 4 cohort studies on milk consumption and gastric cancer risk were identified. As shown in Fig. 3, the
558 Japan
Cohort
Cohort
Cohort
Cohort
Pham 2010
Park 2009
Pols 2007
Ngoan 2002
1995 to 2003
1988 to 2003
Study period
Japan
1989 to 1999
Great 1937 to Britain 2005
US
Country
Authors and Study publication year design Milk and dairy products
Items
Highest quintile vs. lowest quintile
Highest quartile vs. lowest quartile
Range of consumption
3/24374
Quantity assessment
Qualitative score (0–9 points)
(continued on next page)
Self-administered 6.5 questionnaire
Self-administered 7.5 Age, smoking, questionnaire history of gastric ulcer, stomach cancer screening, BMI, educational level, and total energy intake. Race, education, Self-administered 7.5 questionnaire marital status, BMI, family history of cancer, vigorous physical activity, alcohol consumption, intakes of red meat, total energy, smoking and antacid use. 8 7-d household Age, sex, and inventory energy, fruit, method and calcium intakes.
Variables of adjustment
All liquid milks, Dairy: 471 g/d vs. 89 g/d; infant Milk: 1.2 formulas, cups/d vs. cream, cheese, ice creams, and once/day vs. < 2 times/ week
Highest quartile vs. lowest quartile 150/7990 Milk, ice cream, > 5 times/week butter vs.17 times/week Age, gender, total calories, chili vs. 5 times/week vs. once/day vs. < 3 times/ month
Milk and cheese Highest vs. lowest tertile
Milk
Age and sex
Questionnaire interview
Age, sex, marital Questionnaire status, ethnic interview group and schooling year Age, sex, Questionnaire smoking interview history, education and county of residence Age, sex Questionnaire interview
>2 meals/week vs. < once/ week
Quantity assessment Questionnaire interview
Variables of adjustment
>5 times/week vs. < once/ week
Whole milk, 2% Daily vs. none milk and skim milk
Milk
Items
Range of consumption
GC D gastric cancer; BMI D body mass index; SES D social-economic status; GCA D gastric cardia adenocarcinoma; GNCA D gastric noncardia adenocarcinoma.
Vecchia 1987
Kono 1988
Mettlin 1990
Lee 1990
Country
Populationbased casecontrol study HospitalTaiwan based case control study HospitalUS based case control study
Authors and Study publication year design
TABLE 1 Characteristics of 23 studies included in this meta-analysis (Continued)
Downloaded by [New York University] at 14:30 17 May 2015
7
7
7
7
Qualitative score (0–9 points)
563
Study design Cohort studies Hospital–based casecontrol studies Population-based casecontrol studies Publish year After 2000 Before 2000 Study location Asia Europe United States Exposure assessment Questionnaire interview Self-administered questionnaire
Subgroup 0.76 (0.64–0.91) 1.16 (0.76–1.79) 0.88 (0.55–1.41)
1.08 (0.76–1.54) 1.08 (0.71–1.67) 1.16 (0.78–1.75) 0.73 (0.54–0.99) 0.78 (0.62–0.99) 1.30 (0.91–1.85) 0.76 (0.64–0.91)
3
8 5 5 4 2 9 3
RR (95% CI)
4 6
No. of studies
70.96 2.99
63.65 1.09 2.81
88.82 13.36
9.97
2.99 60.44
Heterogeneity
Total dairy food
87.30% 0.00%
90.60% 0.00% 28.90%
88.70% 70.10%
79.90%
0.00% 90.10%
I2 (%)
11 3
6 4 4
5 9
3
4 7
No. of studies
1.27 (0.93–1.75) 1.27 (0.87–1.86)
0.96 (0.79–1.16) 1.57 (0.86–2.88) 1.28 (0.98–1.67)
1.37 (0.75–2.50) 1.25 (0.92–1.71)
1.04 (0.72–1.52)
1.23 (0.95–1.59) 1.36 (0.83–2.23)
RR (95% CI)
Milk
39.46 4.26
4.87 9.39 13.2
14.07 29.72
1.82
4.31 36.95
Heterogeneity
TABLE 2 Relative risks (RR) of gastric cancer in relation to total dairy food and milk consumption according to study characteristics
Downloaded by [New York University] at 14:30 17 May 2015
74.70% 29.60%
0.00% 68.00% 77.30%
64.50% 73.10%
0.00%
7.10% 83.80%
I2 (%)
Downloaded by [New York University] at 14:30 17 May 2015
564
Y. GUO ET AL.
FIG. 2. Forest plots showing estimated relative risks (highest vs. lowest category) of gastric cancer associated with total dairy consumption.
results for association between milk intake and risk of gastric cancer in various studies were inconsistent. Compared with the lowest categories of milk intake, the summary RRs for the highest categories of milk intake were 1.29 (95% CI: 0.89– 1.88) in a random-effects analysis based on case-control studies and 1.23 (95% CI: 0.95–1.59) in a fixed-effects analysis based on cohort studies, respectively. There was significant heterogeneity between case-control studies (I2 D 76.9%), but low heterogeneity observed (I2 D 7.1%) between cohort studies. The overall RR was 1.28 (95% CI: 0.97–1.67) for casecontrol and cohort studies combined with a statistically significant heterogeneity across studies (I2 D 68.0%). Subgroup and Sensitivity Analyses As shown in Table 2, in subgroup analyses of milk intake and risk of gastric cancer, there was no association found in all strata of the study characteristics. The results of sensitivity analyses suggested the overall risk estimates did not substantially modified by any single study (Supplementary Fig. S2). The statistically significant heterogeneity never changed whichever study was excluded. Publication Bias There was no evidence of publication bias with regard to consumption of total dairy or milk in relation to gastric cancer
risk no matter based on case-control or cohort studies, as suggested by Begg rank correlation test and Egger linear regression test (all P > 0.05). DISCUSSION In our meta-analysis, we observed discrepancies between results of case-control and cohort studies concerning the relationship between total dairy consumption and gastric cancer risk. The result of cohort studies indicated that increased consumption of total dairy food might be associated with a reduced risk of gastric cancer, whereas case-control studies provided no association. Milk consumption was not associated with increased gastric cancer risk no matter in case-control studies or in cohort studies. Subgroup analyses based on limited numbers of studies suggested that the inverse associations between total dairy consumption and gastric cancer existed in Europe subgroup and U.S. subgroup but not in Asia subgroup. Some mechanisms exist to explain the protective effect of dairy on gastric cancer. First, phospholipids in dairy food may increase gastric cellular restitution (47), and both probiotics (48) and yogurt (49) may interfere with Helicobacter pylori colonization, which is considered as a key promoter of gastric antral and body cancer (50,51). In addition, dairy could have its protective effect may attribute to some ingredients that have anticarcinogenic properties, such as vitamin D, calcium,
Downloaded by [New York University] at 14:30 17 May 2015
DAIRY CONSUMPTION AND GASTRIC CANCER RISK
565
FIG. 3. Forest plots showing estimated relative risks (highest vs. lowest category) of gastric cancer associated with milk consumption.
and b-carotene (52–54). Vitamin D, which is tightly related to calcium regulation in the body, has been inversely associated with digestive system cancers in another study (54). The metaanalysis based on the data of randomized control trials suggested a significantly decreased risk for stomach cancer in individuals supplemented with doses of b-carotene below 20 mg per day in comparison with the controls. However, no overall association was indicated in this study, and there was even a significantly increased risk of gastric cancer in individuals with high doses of b-carotene supplementary (55). In addition, no association was observed between milk intake and risk of gastric cancer in this study. Associations between milk and gastric cancer risk may have been underestimated in case-control studies for that gastric cancer cases may drink more milk to control symptoms of the disease such as dyspepsia. Furthermore, in the subgroup analyses, the inverse associations between total dairy consumption and gastric cancer were found in Europe subgroup and U.S. subgroup but not in Asia subgroup. The different observations may be explained, at least in part, by the variations of dairy consumption across the world. Higher dairy consumption is found in Europe and North America when compared with Asia (56,57). In addition, a possible role of ethnic differences in genetic backgrounds should also be taken into account. For example, lactose intolerance is at its highest frequency in some parts of East Asia,
whereas northern Europeans generally have the lowest frequency of this dietary problem (58). There are several possible explanations why case-control studies and cohort studies provide conflicting results pertaining to the association between total dairy intake and gastric cancer. In case-control studies, dietary intake is assessed after the cancer diagnosis. Consequently, these studies may be subject to recall bias and inaccurate measurements of dietary intake attributed to changes in the dietary practices among cases after the cancers were diagnosed. Even though the cases were inquired about their diet-before-disease, their recall may well be influenced by their current diet. Another difference between case-control and cohort studies may be attributed to the time interval between the period covered by the questionnaire and the cancer diagnosis. The time interval of dietary intake is usually 1 or 2 years before diagnose in case-control studies, whereas in cohort studies it could be as long as 10– 20 years (recent diet at baseline). If the diet over the past 10–20 years before the diagnosis of gastric cancer may be most relevant and if any change in dietary patterns have occurred, an association with dairy food consumption may be hidden in case-control studies. It is also likely that long-term dietary intake may be most relevant to chronic diseases like cancer with a long duration of development.
Downloaded by [New York University] at 14:30 17 May 2015
566
Y. GUO ET AL.
However, contrary to our results, high intake of dairy products was found to be associated with increased risk of some specific cancers (11,59). Increased plasma calcium due to intake of dairy or milk could suppress the production of plasma 1,25-dihydroxyvitamin D3, which might be one possible mechanism underlying the association (6). Moreover, high dairy or milk consumption could increase blood levels of insulin-like growth factor-I (60), which has been associated with cancer risk (61,62). Several potential limitations of our study are worth considering. First, this meta-analysis is based on observational studies, and the inherent limitations of such studies may affect our findings. The possibility of bias and uncontrolled or residual confounding cannot be excluded. Although most studies included in the meta-analysis controlled possible confounding factors, inherent problems with the variation of confounders in the studies included should be addressed. Most articles included in this meta-analysis had adjusted for major potential confounders, including age, gender, alcohol consumption, body mass index, and history of family gastric cancer, but only 4 studies had adjusted for total energy intake and one study included the confounder of H. pylori. The inconsistent adjustments among different studies may lead to bias of the pooled results. However, there was no difference between the study adjusting for H. pylori and the pooled results of other studies. Because H. pylori infection is such an important risk factor for gastric cancer, further studies are needed to explore the interactions between dairy intake and H. pylori infection on gastric cancer. Second, dairy and milk products are a collection of several products; the ambiguous or different definition in each questionnaire of study may result in inaccurate estimates. Inevitably, dietary assessments suffer from measurement errors. Within single studies, nondifferential misclassification may occur in classifying categories of dairy product consumption. Across studies, differential misclassification may be introduced as dietary assessments were based on different questionnaires and different nutrient databases. Third, most of the included studies defined patients with all types of gastric cancers as gastric cancer cases and did not examine the association by gastric cancer subtype. It is possible that some associations that may exist only for certain tumor subtypes. In addition, we could not rule out the influences of diet change on the risk estimates among participants with sub-clinical gastric cancer. Few studies assessed dairy and milk consumption more than once during the followup periods, even in cohort studies. Finally, potential publication bias might influence the findings, yet we found no evidence of publication bias involved in our findings. In conclusion, the results of this meta-analysis provided some support for the hypothesis that total dairy food may have a potential protective effect for gastric cancer. However, this evidence is largely limited to cohort studies, whereas case-
control studies have produced conflicting results. Further research is warranted to confirm these findings and elucidate the likely biological mechanisms. ACKNOWLEDGMENTS We gratefully acknowledge the directions of Prof. Liegang Liu and Dr Wei Bao. FUNDING This work was supported by National Basic Research Program (Grant 2011CB503804). SUPPLEMENTAL DATA Supplemental data for this article can be accessed on the publisher’s website. REFERENCES 1. World Health Organization: GLOBOCAN 2012: estimated incidence, mortality and prevalence worldwide in 2012. Available from http://globo can.iarc.fr/Pages/fact_sheets_cancer.aspx. 2. Guggenheim DE and Shah MA: Gastric cancer epidemiology and risk factors. J Surg Oncol 107, 230–236, 2013. 3. Kneller RW, McLaughlin JK, Bjelke E, Schuman LM, Blot WJ, et al.: A cohort study of stomach cancer in a high-risk American population. Cancer 68, 672–678, 1991. 4. Lampe JW: Dairy products and cancer. J Am Coll Nutr 30(5, Suppl 1), 464S–470S, 2011. 5. Qin LQ, Wang PY, Kaneko T, Hoshi K, and Sato A: Estrogen: one of the risk factors in milk for prostate cancer. Med Hypotheses 62, 133–142, 2004. 6. Chan JM and Giovannucci EL: Dairy products, calcium, and vitamin D and risk of prostate cancer. Epidemiol Rev 23, 87–92, 2001. 7. Dong JY, Zhang L, He K, and Qin LQ: Dairy consumption and risk of breast cancer: a meta-analysis of prospective cohort studies. Breast Cancer Res Treat 127, 23–31, 2011. 8. Mao QQ, Dai Y, Lin YW, Qin J, Xie LP, et al.: Milk consumption and bladder cancer risk: a meta-analysis of published epidemiological studies. Nutr Cancer 63, 1263–1271, 2011. 9. Aune D, Lau R, Chan DS, Vieira R, Greenwood DC, et al.: Dairy products and colorectal cancer risk: a systematic review and meta-analysis of cohort studies. Ann Oncol 23, 37–45, 2012. 10. Gao X, LaValley MP, and Tucker KL: Prospective studies of dairy product and calcium intakes and prostate cancer risk: a meta-analysis. J Natl Cancer Inst 97, 1768–1777, 2005. 11. Genkinger JM, Hunter DJ, Spiegelman D, Anderson KE, Arslan A, et al.: Dairy products and ovarian cancer: a pooled analysis of 12 cohort studies. Cancer Epidemiol Biomarkers Prev 15, 364–372, 2006. 12. Gao Y, Hu N, Han XY, Ding T, Giffen C, et al.: Risk factors for esophageal and gastric cancers in Shanxi Province, China: a case-control study. Cancer Epidemiol 35, e91–e99, 2011. 13. Lazarevic K, Nagorni A, Rancic N, Milutinovic S, Stosic L, et al.: Dietary factors and gastric cancer risk: hospital-based case control study. J BUON 15, 89–93, 2010. 14. Pourfarzi F, Whelan A, Kaldor J, and Malekzadeh R: The role of diet and other environmental factors in the causation of gastric cancer in Iran—a population based study. Int J Cancer 125, 1953–1960, 2009. 15. Fei SJ and Xiao SD: Diet and gastric cancer: a case-control study in Shanghai urban districts. Chin J Dig Dis 7, 83–88, 2006.
Downloaded by [New York University] at 14:30 17 May 2015
DAIRY CONSUMPTION AND GASTRIC CANCER RISK 16. Kim HJ, Chang WK, Kim MK, Lee SS, and Choi BY: Dietary factors and gastric cancer in Korea: a case-control study. Int J Cancer 97, 531–535, 2002. 17. Munoz N, Plummer M, Vivas J, Moreno V, De Sanjose S, et al.: A casecontrol study of gastric cancer in Venezuela. Int J Cancer 93, 417–423, 2001. 18. Ward MH and Lopez-Carrillo L: Dietary factors and the risk of gastric cancer in Mexico City. Am J Epidemiol 149, 925–932, 1999. 19. Hansson LE, Nyren O, Bergstrom R, Wolk A, Lindgren A, et al.: Diet and risk of gastric cancer. A population-based case-control study in Sweden. Int J Cancer 55, 181–189, 1993. 20. Hoshiyama Y and 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 3, 441–448, 1992. 21. Yu GP and Hsieh CC: Risk factors for stomach cancer: a populationbased case-control study in Shanghai. Cancer Causes Control2, 169–174, 1991. 22. Boeing H, Frentzel-Beyme R, Berger M, Berndt V, Gores W, et al.: Casecontrol study on stomach cancer in Germany. Int J Cancer 47, 858–864, 1991. 23. Wu-Williams AH, Yu MC, and Mack TM: Life-style, workplace, and stomach cancer by subsite in young men of Los Angeles County. Cancer Res 50, 2569–2576, 1990. 24. Mettlin CJ, Schoenfeld ER, and Natarajan N: Patterns of milk consumption and risk of cancer. Nutr Cancer 13(1–2), 89–99, 1990. 25. Lee HH, Wu HY, Chuang YC, Chang AS, Chao HH, et al.: Epidemiologic characteristics and multiple risk factors of stomach cancer in Taiwan. Anticancer Res 10, 875–881. 26. Kono S, Ikeda M, Tokudome S, and Kuratsune M: A case-control study of gastric cancer and diet in northern Kyushu, Japan. Jpn J Cancer Res 79, 1067–1074, 1988. 27. La Vecchia C, Negri E, Decarli A, D’Avanzo B, and Franceschi S: A case-control study of diet and gastric cancer in northern Italy. Int J Cancer 40, 484–489, 1987. 28. Pham TM, Fujino Y, Kikuchi S, Tamakoshi A, Matsuda S, et al.: Dietary patterns and risk of stomach cancer mortality: the Japan collaborative cohort study. Ann Epidemiol 20, 356–363, 2010. 29. Park Y, Leitzmann MF, Subar AF, Hollenbeck A, and Schatzkin A: Dairy food, calcium, and risk of cancer in the NIH-AARP Diet and Health Study. Arch Intern Med 169, 391–401, 2009. 30. van der Pols JC, Bain C, Gunnell D, Smith GD, Frobisher C, et al.: Childhood dairy intake and adult cancer risk: 65-y follow-up of the Boyd Orr cohort. Am J Clin Nutr 86, 1722–1729, 2007. 31. Ngoan LT, Mizoue T, Fujino Y, Tokui N, and Yoshimura T: Dietary factors and stomach cancer mortality. Br J Cancer 87, 37–42, 2002. 32. Ursin G, Bjelke E, Heuch I, and Vollset SE: Milk consumption and cancer incidence: a Norwegian prospective study. Br J Cancer 61, 454–459, 1990. 33. Nomura A, Grove JS, Stemmermann GN, and Severson RK: A prospective study of stomach cancer and its relation to diet, cigarettes, and alcohol consumption. Cancer Res 50, 627–631, 1990. 34. Chen H, Ward MH, Graubard BI, Heineman EF, Markin RM, et al.: Dietary patterns and adenocarcinoma of the esophagus and distal stomach. Am J Clin Nutr 75, 137–144, 2002. 35. Wells GA, Shea B, O’Connell D, Peterson J, Welch V, et al.: The Newcastle-Ottawa Scale (NOS) for assessing the quality of nonrandomized studies in meta-analyses. Accessed from www.ohri.ca/programs/clinica l_epidemiology/oxford.asp, 2011. 36. Higgins JP, Thompson SG, Deeks JJ, and Altman DG: Measuring inconsistency in meta-analyses. BMJ 327, 557–560, 2003. 37. Higgins JP: Commentary: Heterogeneity in meta-analysis should be expected and appropriately quantified. Int J Epidemiol 37, 1158–1160, 2008.
567
38. Xibin S, Moller H, Evans HS, Dixing D, Wenjie D, et al.: Residential environment, diet and risk of stomach cancer: a case-control study in Linzhou, China. Asian Pac J Cancer Prev 3, 167–172, 2002. 39. Zhuo XG and Watanabe S: Factor analysis of digestive cancer mortality and food consumption in 65 Chinese counties. J Epidemiol 9, 275–284, 1999. 40. Tuyns AJ, Kaaks R, Haelterman M, and Riboli E: Diet and gastric cancer. A case-control study in Belgium. Int J Cancer 51, 1–6, 1992. 41. Gonzalez CA, Sanz JM, Marcos G, Pita S, Brullet E, et al.: Dietary factors and stomach cancer in Spain: a multi-centre case-control study. Int J Cancer 49, 513–519, 1990. 42. Demirer T, Icli F, Uzunalimoglu O, and Kucuk O: Diet and stomach cancer incidence. A case-control study in Turkey. Cancer 65, 2344–2348, 1990. 43. Buiatti E, Palli D, Decarli A, Amadori D, Avellini C, et al.: A case-control study of gastric cancer and diet in Italy. Int J Cancer 44, 611–616, 1989. 44. Hu JF, Zhang SF, Jia EM, Wang QQ, Liu SD, et al.: Diet and cancer of the stomach: a case-control study in China. Int J Cancer 41, 331–335, 1988. 45. Correa P, Fontham E, Pickle LW, Chen V, Lin YP, et al.: Dietary determinants of gastric cancer in south Louisiana inhabitants. J Natl Cancer Inst 75, 645–654, 1985. 46. Nagai M, Hashimoto T, Yanagawa H, Yokoyama H, and Minowa M: Relationship of diet to the incidence of esophageal and stomach cancer in Japan. Nutr Cancer 3,257–268, 1982. 47. Dial EJ and Lichtenberger LM: A role for milk phospholipids in protection against gastric acid. Studies in adult and suckling rats. Gastroenterology 87, 379–385, 1984. 48. Kabir AM, Aiba Y, Takagi A, Kamiya S, Miwa T, et al.: Prevention of Helicobacter pylori infection by lactobacilli in a gnotobiotic murine model. Gut 41, 49–55, 1997. 49. Sheu BS, Cheng HC, Kao AW, Wang ST, Yang YJ, et al.: Pretreatment with Lactobacillus-and Bifidobacterium-containing yogurt can improve the efficacy of quadruple therapy in eradicating residual Helicobacter pylori infection after failed triple therapy. Am J Clin Nutr 83, 864–869, 2006. 50. Lee SA, Kang D, Shim KN, Choe JW, Hong WS, et al.: Effect of diet and Helicobacter pylori infection to the risk of early gastric cancer. J Epidemiol 13, 162–168, 2003. 51. Singh K and Ghoshal UC: Causal role of Helicobacter pylori infection in gastric cancer: an Asian enigma. World J Gastroenterol 12, 1346–1351, 2006. 52. Peterlik M, Grant WB, and Cross HS: Calcium, vitamin D and cancer. Anticancer Res 29, 3687–3698, 2009. 53. Garland CF, Gorham ED, Mohr SB, and Garland FC: Vitamin D for cancer prevention: global perspective. Ann Epidemiol 19, 468–483, 2009. 54. Giovannucci E, Liu Y, Rimm EB, Hollis BW, Fuchs CS, et al.: Prospective study of predictors of vitamin D status and cancer incidence and mortality in men. J Natl Cancer Inst 98, 451–459, 2006. 55. Druesne-Pecollo N, Latino-Martel P, Norat T, Barrandon E, Bertrais S, et al.: Beta-carotene supplementation and cancer risk: a systematic review and metaanalysis of randomized controlled trials. Int J Cancer 127, 172– 184, 2010. 56. Ma Q, Xin J, Zhao Z, Guo Q, Yu S, et al.: Value of 18F-FDG PET/CT in the diagnosis of primary gastric cancer via stomach distension. European Journal of Radiology 82, e302–e306, 2013. 57. Wang Y and Li S: Worldwide trends in dairy production and consumption and calcium intake: is promoting consumption of dairy products a sustainable solution for inadequate calcium intake? Food Nutr Bull 29, 172–185, 2008. 58. Wang XQ, Yan H, Terry PD, Wang JS, Cheng L, et al.: Interaction between dietary factors and helicobacter pylori infection in noncardia gastric cancer: A population-based case-control study in China. J Am Coll Nutr 31, 375–384, 2012.
568
Y. GUO ET AL.
Downloaded by [New York University] at 14:30 17 May 2015
59. Larsson SC, Bergkvist L, and Wolk A: Milk and lactose intakes and ovarian cancer risk in the Swedish Mammography Cohort. Am J Clin Nutr 80, 1353–1357, 2004. 60. Giovannucci E, Pollak M, Liu Y, Platz EA, Majeed N, et al.: Nutritional predictors of insulin-like growth factor I and their relationships to cancer in men. Cancer Epidemiol Biomarkers Prev 12, 84–89, 2003.
61. Renehan AG, Zwahlen M, Minder C, O’Dwyer ST, Shalet SM, et al.: Insulin-like growth factor (IGF)-I, IGF binding protein-3, and cancer risk: systematic review and meta-regression analysis. Lancet 363(9418), 1346– 1353, 2004. 62. Lukanova A, Lundin E, Toniolo P, Micheli A, Akhmedkhanov A, et al.: Circulating levels of insulin-like growth factor-I and risk of ovarian cancer. Int J Cancer 101, 549–554, 2002.
