Practically Oriented Study on the Deviation Pattern Similarity in Analysis of Dietary Habits and Life

The deviation pattern similarity proposed by Marui et al. requires a subsequent quantitative developmental approach if its practical value for dietary life analysis and widespread adoption are to be encouraged. In the present study we sought to investigate a means to garner useful information from the deviation similarity pattern and at the same time applied it to actual food consumption data from Gifu Prefecture. The following method and results were obtained. 1. For the method, first the deviation pattern similarity matrix was converted to a Euclidean distance matrix. Next, using the distance matrix as input data, an analysis was run with the multidimensional scaling method (Torgerson method) and space coordinates were obtained. These were interpreted in terms of space relatively easy to grasp subjectively. Then, using multiple regression analysis, the correlation between the space coordinates and food consumption was studied. 2. Application to the food consumption data from Gifu Prefecture yielded three patterns (called A, B and C groups) for the 15 districts surveyed. These three corresponded to the government sectors of village, town and city. They also showed an A group-B group-C group chain-like construction. 3. Multiple regression analysis showed that this linking of the district groups bore a close relationship with the “ongoing evolution in dietary life. ” Findings suggested that this “ongoing evolution in dietary life” is unfolding along with urbanization.