Evidence Based Review Article

The Journal of Informed Pharmacotherapy 2001;5:223-225.

Green Tea and the Risk of Gastric Cancer in Japan; Is there Any Relationship or Are We Simply Reading Tea Leaves?

Download a PDF version of this article

Reviewer: Suzanne C. Malfair Taylor, BSc(Pharm), PharmD, BCPS
Reviewer's email address:
Reviewer's profession/specialty:

Original Citation

Yoshitaka Tsubono, Yoshikazu Nishino, Shoko Komatsu, Chung-Cheng Hsieh, Seiki Kanemura, Ichiro Tsuji, Haruo Nakatsuka, Akira Fukao, Hiroshi Satoh, Shigeru Hisamichi.  Green Tea and the Risk of Gastric Cancer in Japan.   N Engl J Med 2001;344:632-6.  PubMed Cit

Overall Study Question

The polyphenols in green tea are believed to have anticarcinogenic effects. The objective of this study was to examine whether consumption of green tea decreases the relative risk of gastric cancer.

This was a prospective cohort study involving 26,311 residents aged 40 years or older who were

of Miyagi Prefecture (an area in northern Japan where the incidence of gastric cancer is high) during the period January 1984 to

December 1992.   A multivariate analysis

that adjusted for sex, age, history of peptic ulcer, smoking status, alcohol consumption, several dietary elements, and type of health insurance was conducted to determine the relative risk of gastric cancer associated with 3 levels of green tea consumption: 1-2 cups/day, 3-4 cups/day, and >=5 cups/day. There was no actual intervention.  This was a population study where exposure data was gathered via self-administered questionnaires.  The questionnaires were delivered to and collected from residences by government health-promotion committee members. 

Follow-up data on patient status and cancer incidence was gathered from population registries and the Miyagi Prefectural Cancer Registry.

Are the Results of the Study Valid?

1. Was assignment of patients randomized?


The self-administered questionnaires were delivered to 33,453 residents aged 40 years or greater, residing in any of 3 municipalities in Miyagi Prefecture. 

As there is no mention of any selection criteria (other than the aforementioned age), for these residents, it seems that the questionnaire was offered to all residents meeting the above description. This however, is not directly identified

in the paper.

2. Were all patients who entered the trial properly accounted for and attributed at its conclusion?

Unable to determine.  Although all subjects entering the study were clearly accounted for (33,453 offered questionnaires; 94% returned; 31,345 usable; 30,804 did not have pre-existing cancer; 26,311 answered the green tea question on the questionnaire), the results are reported in person-years.  Since subjects were followed until date of gastric cancer diagnosis, date of emigration from the study district, date of death, or end date of the study period, there is insufficient data provided to verify the number of person-years.  To their credit, the authors do reveal that an evenly distributed 16% of subjects fell into the 'emigration from study district' category and were thus considered lost to follow-up.

3. Were patients, their clinicians, and study personnel 'blind' to treatment?

There were no treatment groups in this study. The questionnaire provided to subjects included questions on various health and diet habits so they were not aware that green tea specifically was being examined.

4. Were the groups similar at the start of the trial?

Since this was a population study, there were no study groups, per se. Once data on green tea consumption was available, the subjects who reportedly drank <1 cup/day were deemed the control group.  Although p values are not reported, it is evident from the percentages reported in the demographics table that there were obvious differences in several habits between men and women, as well as within each sex.  Some elements that differed included: pickled vegetable consumption (risk), fruit consumption (protectant), and smoking (risk).

5. Aside from the experimental intervention, were the groups treated equally?

Data was collected in the same manner for all subjects.

6. Overall, are the results of the study valid?


What were the Results?

1. How large was the treatment effect? 

The investigators found no inverse association between green tea consumption and gastric cancer risk. Relative risk was calculated for men, women, and the combination of men and women.  In each of these groups, three different ratios were calculated for each of the different levels of green tea consumption: 1) sex and age adjusted; 2) multivariate: sex, age, health insurance, peptic ulcer history, smoking and diet adjusted; and finally, 3) the same multivariate analysis excluding subjects with gastric cancer diagnosed during the first 3 years of follow-up.

All but one of the 27 relative risk ratios calculated had 95% confidence intervals that contained unity.  The age-adjusted relative risk for men who consumed at least 5 cups/day was 1.6 (95% confidence interval 1.1-2.2).  Only the age-adjusted trend towards increased gastric cancer in men with increasing levels of green tea consumption reached statistical significance (p=0.007).  This would imply that green tea actually increased the relative risk of gastric cancer for this stratification.  When the multivariate analysis was conducted, the authors reported that the trend was still significant (p=0.03), however all reported relative risk ratios were associated with 95% confidence intervals that included unity.  The authors concluded that this study found no association (inverse or otherwise) between green tea consumption and the risk of gastric cancer in Japan.

The authors identified some possible confounding factors that could limit this study.  For example, there was no attempt to record other dietary habits or history of Helicobacter pylori infection.  Other potential confounders that were not discussed include drugs, vitamins, herbal remedies, and other complementary therapies. 

Finally, while this study spanned a 8-year period, the dietary and health habit information was apparently collected in the first year only.  In contrast, the gastric cancer data was collected during the 1984 to 1992 period.  Accordingly, it is possible that the dietary health and habits of the participants changed changed over the course of the follow-up period, although this would not have been recorded by the investigators. 

Will the Results Help Me in Caring for My Patients?

1. Can the results be applied to my patient care?

Despite the potential for unmeasured confounders (including changes in dietary and/or health habits over the 9-year follow-up period), this study provides data that is often non-existent for other dietary/complementary therapies. This prospective cohort study carries more weight than case-control studies and anecdotal reports.  In the absence of an intervention study, this is about as good as it gets.

The accompanying editorial (1) describes how gastric cancer characteristics are different (in terms of location and some risk factors), between Japan and Western nations. This should be taken into consideration and mentioned during patient counseling sessions.

2. Were all clinically important outcomes considered?

The gastric cancers identified were not characterized in terms of location in the stomach (i.e. proximal vs. distal). Had this been done, the results would have greater external validity.  As discussed earlier, controlling for some of the potential confounding variables would also have strengthened the study results.

3. Are the likely treatment benefits worth the potential harms and costs?

This study did not find any benefit of lowered relative risk of gastric cancer (indeed, there was a trend towards a worsening) with increasing green tea consumption.  Based on the theory of the health benefits of green tea (including antioxidant activity and potential antiangiogensis activity(2)), and the limitations of this study, green tea consumption as a beverage generally need not be discouraged.


Many patients, especially those with cancer, have questions about dietary and complementary therapies.  Often, there is only theory or anecdotal reports from which to try to formulate answers. Green tea is an antioxidant with free radical scavenging ability greater than vitamin E.(3)  Generally antioxidants are considered desirable.  There are however, some possible interactions between antioxidants and radiation therapy as well as between antioxidants and some chemotherapy agents.(3,4)

Unless it is decaffeinated, green tea contains about 60mg caffeine/cup.  Adverse effects can include stomach upset, constipation, as well as the usual effects of caffeine.  There are also potential drug and disease interactions, which are generally related to the caffeine content as well.(2)

Taking all of this into consideration, it seems that moderation would be the logical answer. Patients may be counseled on both the potential advantages and disadvantages so that they are more able to make their own informed decision about how much green tea they wish to consume. For cancer patients undergoing radiation therapy or chemotherapy with an agent whose mechanism relies on oxidation, it would be prudent to suggest abstention from or minimization of concurrent green tea consumption.


1. Sano T, Sasako M. Green tea and gastric cancer (editorial). N Engl J Med 2001;344:675-6.

2. Green tea. Natural medicines comprehensive database. Therapeutic Research 1995-2001. http://www.naturaldatabase.com

3. Boik J. Cancer and natural medicine: a textbook of basic science and clinical research. Oregon Medical Press, Minnesota, 1996.

4. Labriola D, Livingston R. Possible interactions between dietary antioxidants and chemotherapy. Oncology 1999;13:1003-8.

Copyright © 2001 by the Journal of Informed Pharmacotherapy. All rights reserved.