A case-control study for estimating the population attributable proportion and association between type 2 diabetes mellitus and hepatocellular carcinoma at a viral hepatitis hyperendemic area

• Main Authors: Wei-Hsu Ko, 柯威旭
• Other Authors: 陳秀熙
• Format: Others
• Language: zh-TW
• Published: 2007
• Online Access: http://ndltd.ncl.edu.tw/handle/87590123741173300652

碩士 === 臺灣大學 === 預防醫學研究所 === 95 === Among causal models, a sufficient cause-component model differs from the other models in that it provides more information on causal mechanisms and defines biologic effects of several risk factors, particularly including strengths of association and their joint interaction effects. The advantage of looking at sufficient causes for the disease of interest is to give a clue to the mechanism of disease in question. We conducted a hospital-based case-control study with the inclusion of DM as a risk factor to elucidate the causal relationship leading to HCC after controlling for other major risk factors of HCC. In this hospital-based case control study, we corroborated the association between DM and HCC in Taiwan, where viral hepatitis B and C are hyperendemic. The estimated population attributable proportion indicated that diabetes accounted for 18.7% of HCC cases given 7% prevalence of diabetes. The independent contribution to the development of HCC cases in the absence of HBV and HCV infection was around 7.5%. In addition, we applied the method from the literature to calculate the adjusted population attributable proportion. We also developed a Bayesian acyclic graphic model to calculate the adjusted PAF and 95% confidence intervals.