Interactions between birth cohort and urbanization on gastric cancer mortality in Taiwan

Birth-cohort analysis is of particular importance for gastric cancer since migrant studies have indicated that early life experiences play a dominant role in the occurrence of the disease. However, none of the birth-cohort analyses conducted in various countries have provided information on variation in the birth-cohort phenomenon. To examine the interaction of birth cohort and urbanization, mortality data for gastric cancer between 1971 and 1990 in differentially urbanized areas in Taiwan were analysed. Traditional birth-cohort analysis and age-period-cohort (APC) analysis were employed in the study. The identification problem inherent in the APC analysis was circumvented by the 'individual record' method. Grading of the degree of urbanization of each township and district in Taiwan was based on an urbanization index derived from several demographic and socioeconomic variables. The birth-cohort effects diffused from urban to rural areas after fitting the APC model. In the older generations, born before 1910, the relative risks of gastric cancer were higher in urban areas than in rural ones, but in recent generations, i.e. those born after 1916-1922, the reverse was true. The age curves of gastric cancer mortality emerged as almost straight lines when plotted on a double logarithmic scale. Sex ratios increased with age up to age 60 and then remained constant. The spread of the birth-cohort effects suggested that dietary factors, e.g. dietary habits and food processing practices, in the early life of the population play an important role in the occurrence of gastric cancer.(ABSTRACT TRUNCATED AT 250 WORDS)