Page 275 - Contributed Paper Session (CPS) - Volume 2
P. 275

CPS1855 Nobutane H.
            groups as shown in the Figure 1. However, the rates for 75-79 and those for
            50-54 have a different trend after 2000-14 as seen in the Figure 2.
               Such trend in breast cancer death rates for Japanese women aged 50 and
            over might be because age factors and socio-economic factors (Lagerlund, et
            al, 2005) intricately affect Japanese women. For example, we consider the age
            factor associated with that most invasive breast cancers are found in women
            55  years  and  older.  On  the  other  hand,  the  socio-economic  factors  are
            associated, for example, with westernized food culture, low fertility, increase
            of  working  women  and  increasing  of  woman’s  first  pregnancy  over  40.  In
            addition, also incorporation of mammography to public health check is also
            considered as socio-economic factors. So our aim in this study is to propose a
            statistical model for decomposing data given by age and period for breast
            cancer deaths into age and socio-economic factors. And the second aim is to
            apply the proposed model to actual data for Japanese females. From the result
            of fitting the proposed mode, it is found that there exist environmental effects
            on  the  numbers  of  breast  cancer  deaths  associated  our  westernized  food
            culture or low fertility or increase of the number of working women or increase
            of  the  number  of  pregnancies  over  40  years  old  or  incorporation  of
            mammography.
                The proposed model can be applied to data given by the same format, that
            is, age-by-period data. Because of that, we can compare trends in environment
            effects estimated from data for people’s traits or features. In the discussion
            part of this article, we focused on breast cancer and food preferences, and
            show the result of analysing causal relations between the food preferences
            and breast cancer based on estimated environment effects for breast cancer
            and food preferences.
                       Table 1. Numbers of deaths in 10,000 for Japanese women
             Age-                                   Period
             group      1985-1989  1990-1994  1995-1999  2000-2004  2005-2009  2010-2014
               50-54       18.8      21.1       26.6       29.2      28.6       27.2
               55-59       21.5      23.5       27.3       31.5      35.6       35.8
               60-64       20.9      22.3       26.0       34.0      34.0       36.8
               65-69       19.0      21.7       24.8       31.3      31.3       36.1
               70-74       18.9      20.9       24.0       30.1      30.1       34.0
               75-79       18.6      20.1       24.9       30.6      30.6       34.8
















                                                               264 | I S I   W S C   2 0 1 9
   270   271   272   273   274   275   276   277   278   279   280