摘要：We report here a historical cohort study on cancer mortality among Minamata disease (MD) patients (n=1, 351) in Kagoshima and Kumamoto Prefectures of Japan. Taking into account their living area, sex, age and fish eating habits, the residents (n=5, 667; 40 years of age or over at 1966) living in coastal areas of Kagoshima, who consumed fish daily, were selected as a reference group from the six-prefecture cohort study conducted by Hirayama et al. The observation periods of the MD patients and of the reference group were from 1973 to 1984 and from 1970 to 1981, respectively. Survival analysis using the Poisson regression model was applied for comparison of mortality between the MD patients and the reference group. No excess of relative risk (RR) adjusted for attained age, sex and follow-up period was observed for mortality from all causes, all cancers, and non-cancers combined. Analysis of site-specific cancers showed a statistically significant decrease in mortality from stomach cancer among MD patients (RR, 0.49; 95 % confidence interval, 0.26-0.94). In addition, a statistically significant eight-fold excess risk, based on 5 observed deaths, was noted for mortality from leukemia (RR, 8.35; 95 % confidence interval, 1.61-43.3). It is, however, unlikely for these observed risks to be derived from methylmercury exposure only. Further studies are needed to understand the mechanisms involved in the observed risks among MD patients. J Epidemiol , 1996 ; 6 : 134-138.
