Abstract
The cancer deaths among agricultural households in Kanto area were characterized in this series of papers from the two standpoints: the age-dependences and geographical distributions of their cancer risk. Our previous paper had investigated the cancers of high or low risk among agricultural households in the whole of Japan and Kanto area, in terms of odds ratios compared with the non-agricultural households (Nakachi, 1983). The age-de pendences of odds ratios had been analyzed there to estimate the periods in ages as to when the causes of ascending or decending risks for those cancers affected the agricultural households. In this paper, we wish to call further attention to the geographical distribution of odds ratios in the small divisions of Kanto area with respect to the cancers of high or low risk which were disclosed in the previous paper. Focusing on the eating habits of the agricultural and non-agricultural people, we next examine the following two assumptions for cancers of the digestive organs and the breast. (1) The high or low risk for the cancers among agricultural households can be ascribable mainly to their particular eating habits. (2) The geographical variation of standardized mortality ratios (SMR) in districts can be explained by the variation of eating habits. When these assumptions are proved to be valid for some of the cancers, it is also necessary to examine a consistency as to whether the risk (or preventive) factors in dietary life are common or different between the agricultural and general people.
