Description of the Evidence
Because of the evidence for an inverse association between gastric cancer and dietary intake of fruits and vegetables, especially those rich in antioxidants, there has been interest in dietary supplementation with antioxidants.
Dietary indices of micronutrient intake have been calculated and indicate possible protective effects of beta carotene, vitamin A, vitamin E, selenium, vitamin C or foods that contain these compounds. A chemoprevention trial in China reported a statistically significant reduction in the gastric cancer mortality rate after supplementation with beta carotene, vitamin E, and selenium. The population studied, however, may have been nutritionally deficient, raising questions of generalizability to other populations such as that of the United States. In addition, the experimental design did not permit assessment of the relative effects of beta carotene, vitamin E, and selenium.
Likewise, there was a randomized placebo-controlled trial of 200 mg of oral allitridium (a component of garlic) every day combined with 100 mcg of oral selenium every other day for 1 month of each year over a 3-year period in Qixia County (Shadong Province, China), an area with low intake of garlic and low selenium content in their garlic compared to other areas of China. Although designed as a double-blinded trial, allitridium causes a distinctive odor of garlic. A total of 5,033 people, who met at least one of the following criteria: (1) medical history of stomach disorder, (2) family history of tumor, (3) history of smoking, or (4) history of alcohol consumption, were randomly assigned. The study, published in a Chinese medical journal, was not well described. After a follow-up for as many as 11 years, there were a total of 23 gastric cancer cases in the allitridium/selenium group and 30 cases in the placebo group (RR after adjustment for a number of baseline characteristics = 0.48; 95% CI, 0.21–1.06). Only 60% of the gastric cancers were diagnosed by histopathology. There was a qualitative difference in outcome by sex: RRmen = 0.36 (95% CI, 0.14–0.92); RRwomen = 1.14 (95% CI, 0.22–5.76). Given the problems with design and reporting of the study, the evidence of benefit (including men) is weak and may not be generalizable to Western countries.