18. Epidemiology and Biostatistics Division

Formed in October 1994, the Epidemiology and Biostatistics Division has planned and conducted independent and collaborative studies on cancer etiology and prevention, with special focus on dietary factors. In this respect, several types of epidemiological projects are currently in progress, such as descriptive, ecological, case-control and cohort studies. The Japan Clinical Oncology Group (JCOG) Statistical Center was moved to this Division. In addition, collaborative studies in the field of clinical and basic sciences are conducted through providing statistical design, analysis and data processing support.

Ecological Study for Clues to Geographic Differences in Stomach Cancer in Japan (Ecocancer Study)

To determine the cause of the 3-fold variation in stomach cancer mortality in five areas in Japan, 1,004 randomly selected residents (631 men and 373 women) were intensively surveyed through questionnaires and biological specimen collection. The serum level of selenium was not a predictor of geographic differences in stomach cancer.(131) The risk factors of Helicobactor pylori infection were investigated with this data set and salty food intake was suggested.(129)

Population-based Prospective Study (Multi-purpose Koseisho Cohort I)

The five areas of the above ecological study were expanded to a total of a 60,000-population cohort study in 1990, and 50,000 questionnaires and 24,000 blood samples were collected to give baseline information. To obtain more precise information on dietary habits, a semi-quantitative food frequency questionnaire was developed in 1994 and will be used for the second survey in February 1995. The dietary record survey to test the validity of the questionnaire has been completed for a subsample of 200 cohort members.

Epidemiological Studies in Brazil (São Paulo Project)

Since 1987, epidemiological studies have been conducted on Japanese immigrants and their descendants in Brazil. After determining the mortality and cancer incidence among Japanese Brazilians, a cross-sectional survey of randomly selected Japanese Brazilians was conducted to collect data on prevaling lifestyle characteristics and certain biological markers. Serum levels of macro- and micronutrients, body mass index, and blood pressure were significantly different from those of Japanese in Japan.(130,131) A case-control study of lung cancer showed that the use of black tobacco was associated with an increase in risk(127) and that Msp I type polymorphism of cytochrome P-450 IA1 was not related to susceptibility.(126)

Randomized Controlled Field Trial with Antioxidant Micronutrients to Prevent Stomach Cancer in High Risk Japanese Populations (HIRAKA Project)

A chemoprevention trial is planned to test the efficacy of nutrient supplementation for preventing stomach cancer among subjects with gastric atrophy living in an area with a high stomach cancer mortality and low serum beta-carotene level. A pilot study is being conducted among the recipients of health check-ups and the feasibility of a randomized controlled field trial will be tested.

Other Epidemiological Studies

An international collaborative study in 17 populations showed that the adduct O6-methylguanine in peripheral blood leukocyte DNA was present in 16% of individuals among populations with a high stomach cancer rate, while it was 1% among low risk populations.(137) A multi-center case-control study revealed a quantitative association between alcohol consumption and risk of esophageal cancer.(135)

JCOG Statistical Center

The center is responsible for (1) assisting with design of study protocols, (2) registering and randomizing patients in each protocol, (3) managing and editing all study data, (4) generating annual reports of protocol study findings, and (5) researching statistical aspects of clinical trials. The center's activities are closely coordinated with those of the JCOG, the largest clinical cooperative cancer research group in Japan. The center cooperated in 40 trials including 15 phase III trials and registered 2,117 patients in 1993-94, while generating the system of the statistical center.

Collaborative Studies through Statistical Support

Prognostic factors were statistically analyzed for cancers of the uterine cervix(7) and ovary.(138) Observer disagreement in histological classification of ovarian tumors was revealed among three certified pathologists and indicated the necessity of a review system in multi-institutional clinical studies.(139)


List of papers from this division
Table of Contents