Prospective study of diabetes mellitus and liver cancer in Japan

Background Previous studies have identified the association between diabetes mellitus and liver cancer. However, the detail of this association is still unclear, in terms of confounding factors, the trend according to the duration of diabetes, and the interaction between diseases associated with the liver cancer and this association. The purpose of the present study was to examine the association between diabetes and liver cancer in view of the trend and the interaction. Methods The baseline survey was conducted during the period 1986-1989 among the general population of Fukuoka Prefecture, Japan (15 417 persons, aged 30-79 years). The respondents were assessed for history of diabetes, age at which they had had diabetes, and other covariates by means of a baseline questionnaire. A total of 7308 persons aged 40-79 years were retrieved for the main analysis and 4902 persons for a subcohort from which the information on history of diseases associated with liver cancer were obtained. The relative risks (RR) and their 95% confidence intervals (CIs) were estimated using the Cox proportional hazards model. Results After adjustment for smoking, alcohol and the diseases associated with liver cancer, the RR for liver cancer was 2.06 (95% CI = 1.01-4.19). Diabetes increased the risk of liver cancer in persons with hepatitis and/or cirrhosis (RR=2.90, 95% CI=113-7.41). However, the RR of diabetes, for liver cancer was 1.35 (95% CI = 0.41-4.43) in persons without hepatitis and cirrhosis. The trend according to the duration of diabetes was not seen. Conclusions A significant association between diabetes and liver cancer was observed. Moreover, this association was modified by hepatitis and cirrhosis. Copyright (C) 2001 John Wiley & Sons, Ltd.