Background & Aims: This study investigated whether obesity, diabetes, and other metabolic factors are independently associated with hepatocellular carcinoma (HCC), stratified by hepatitis B virus (HBV) and hepatitis C virus (HCV) serostatus, and explored the possible joint influence of obesity/diabetes and HBV/HCV infections on the risk of HCC. Methods: A total of 23,820 residents in Taiwan were recruited and followed up for 14 years. All analyses were stratified by hepatitis B surface antigen (HBsAg) and antibody to HCV (anti-HCV) at enrollment, and 218 subjects positive for both seromarkers were excluded. Incident HCC cases were identified via linkage to the national cancer registry. Multivariate-adjusted relative risk (RRa) and 95% confidence interval (95% CI) were estimated using Cox proportional hazards models. Results: Extreme obesity (body mass index ≥30 kg/m2) was independently associated with a 4-fold risk of HCC (RRa, 4.13; 95% CI, 1.38–12.4) among anti-HCV–seropositive subjects and a 2-fold risk (RRa, 2.36; 95% CI, 0.91–6.17) in persons without HBV and HCV infections, after controlling for other metabolic components, but not in HBsAg-seropositive subjects (RRa, 1.36; 95% CI, 0.64–2.89). Diabetes was associated with HCC in all 3 groups, with the highest risk in those with HCV infection (RRa, 3.52; 95% CI, 1.29–9.24) and lowest in HBV carriers (RRa, 2.27; 95% CI, 1.10–4.66). We found more than 100-fold increased risk in HBV or HCV carriers with both obesity and diabetes, indicating synergistic effects of metabolic factors and hepatitis. Conclusions: The finding that both obesity and diabetes are predictors of HCC risk, possibly differently depending on HBV and HCV infection status, may shed some light in preventing HCC. Background & Aims: This study investigated whether obesity, diabetes, and other metabolic factors are independently associated with hepatocellular carcinoma (HCC), stratified by hepatitis B virus (HBV) and hepatitis C virus (HCV) serostatus, and explored the possible joint influence of obesity/diabetes and HBV/HCV infections on the risk of HCC. Methods: A total of 23,820 residents in Taiwan were recruited and followed up for 14 years. All analyses were stratified by hepatitis B surface antigen (HBsAg) and antibody to HCV (anti-HCV) at enrollment, and 218 subjects positive for both seromarkers were excluded. Incident HCC cases were identified via linkage to the national cancer registry. Multivariate-adjusted relative risk (RRa) and 95% confidence interval (95% CI) were estimated using Cox proportional hazards models. Results: Extreme obesity (body mass index ≥30 kg/m2) was independently associated with a 4-fold risk of HCC (RRa, 4.13; 95% CI, 1.38–12.4) among anti-HCV–seropositive subjects and a 2-fold risk (RRa, 2.36; 95% CI, 0.91–6.17) in persons without HBV and HCV infections, after controlling for other metabolic components, but not in HBsAg-seropositive subjects (RRa, 1.36; 95% CI, 0.64–2.89). Diabetes was associated with HCC in all 3 groups, with the highest risk in those with HCV infection (RRa, 3.52; 95% CI, 1.29–9.24) and lowest in HBV carriers (RRa, 2.27; 95% CI, 1.10–4.66). We found more than 100-fold increased risk in HBV or HCV carriers with both obesity and diabetes, indicating synergistic effects of metabolic factors and hepatitis. Conclusions: The finding that both obesity and diabetes are predictors of HCC risk, possibly differently depending on HBV and HCV infection status, may shed some light in preventing HCC. See CME quiz on page 293. The incidence of liver cancer is increasing in several developed countries and would continue to increase for some decades.1Bosch F.X. Ribes J. Diaz M. et al.Primary liver cancer: worldwide incidence and trends.Gastroenterology. 2004; 127: S5-S16Abstract Full Text Full Text PDF PubMed Scopus (2139) Google Scholar The estimated attributable risk for the combined effects of hepatitis B and C viral infections accounts for more than 80% of liver cancer cases worldwide.1Bosch F.X. Ribes J. Diaz M. et al.Primary liver cancer: worldwide incidence and trends.Gastroenterology. 2004; 127: S5-S16Abstract Full Text Full Text PDF PubMed Scopus (2139) Google Scholar Obesity and diabetes have been found to be associated with an increased risk of hepatocellular carcinoma (HCC) in several epidemiologic studies.2Moller H. Mellemgaard A. Lindvig K. et al.Obesity and cancer risk: a Danish record-linkage study.Eur J Cancer. 1994; 30A: 344-350Abstract Full Text PDF PubMed Scopus (398) Google Scholar, 3Wolk A. Gridley G. Svensson M. et al.A prospective study of obesity and cancer risk (Sweden).Cancer Causes Control. 2001; 12: 13-21Crossref PubMed Scopus (506) Google Scholar, 4Rapp K. Schroeder J. Klenk J. et al.Obesity and incidence of cancer: a large cohort study of over 145,000 adults in Austria.Br J Cancer. 2005; 93: 1062-1067Crossref PubMed Scopus (270) Google Scholar, 5Calle E.E. Rodriguez C. Walker-Thurmond K. et al.Overweight, obesity and mortality from cancer in a prospective studied cohort of US adults.N Engl J Med. 2003; 348: 1625-1638Crossref PubMed Scopus (5870) Google Scholar, 6Samanic C. Gridley G. Chow W.H. et al.Obesity and cancer risk among white and black United States veterans.Cancer Causes Control. 2004; 15: 35-43Crossref PubMed Scopus (263) Google Scholar, 7Nair S. Mason A. Eason J. et al.Is obesity an independent risk factor for hepatocellular carcinoma in cirrhosis?.Hepatology. 2002; 36: 150-155Crossref PubMed Scopus (286) Google Scholar Five large population studies, 3 in Europe2Moller H. Mellemgaard A. Lindvig K. et al.Obesity and cancer risk: a Danish record-linkage study.Eur J Cancer. 1994; 30A: 344-350Abstract Full Text PDF PubMed Scopus (398) Google Scholar, 3Wolk A. Gridley G. Svensson M. et al.A prospective study of obesity and cancer risk (Sweden).Cancer Causes Control. 2001; 12: 13-21Crossref PubMed Scopus (506) Google Scholar, 4Rapp K. Schroeder J. Klenk J. et al.Obesity and incidence of cancer: a large cohort study of over 145,000 adults in Austria.Br J Cancer. 2005; 93: 1062-1067Crossref PubMed Scopus (270) Google Scholar and 2 in the United States,5Calle E.E. Rodriguez C. Walker-Thurmond K. et al.Overweight, obesity and mortality from cancer in a prospective studied cohort of US adults.N Engl J Med. 2003; 348: 1625-1638Crossref PubMed Scopus (5870) Google Scholar, 6Samanic C. Gridley G. Chow W.H. et al.Obesity and cancer risk among white and black United States veterans.Cancer Causes Control. 2004; 15: 35-43Crossref PubMed Scopus (263) Google Scholar observed that obesity is associated with an increase in HCC incidence2Moller H. Mellemgaard A. Lindvig K. et al.Obesity and cancer risk: a Danish record-linkage study.Eur J Cancer. 1994; 30A: 344-350Abstract Full Text PDF PubMed Scopus (398) Google Scholar, 3Wolk A. Gridley G. Svensson M. et al.A prospective study of obesity and cancer risk (Sweden).Cancer Causes Control. 2001; 12: 13-21Crossref PubMed Scopus (506) Google Scholar, 4Rapp K. Schroeder J. Klenk J. et al.Obesity and incidence of cancer: a large cohort study of over 145,000 adults in Austria.Br J Cancer. 2005; 93: 1062-1067Crossref PubMed Scopus (270) Google Scholar, 6Samanic C. Gridley G. Chow W.H. et al.Obesity and cancer risk among white and black United States veterans.Cancer Causes Control. 2004; 15: 35-43Crossref PubMed Scopus (263) Google Scholar and mortality,5Calle E.E. Rodriguez C. Walker-Thurmond K. et al.Overweight, obesity and mortality from cancer in a prospective studied cohort of US adults.N Engl J Med. 2003; 348: 1625-1638Crossref PubMed Scopus (5870) Google Scholar but a study conducted in an Asian country failed to obtain similar results.8Lai M.S. Hsieh M.S. Chiu Y.H. et al.Type 2 diabetes and hepatocellular carcinoma: a cohort study in high prevalence area of hepatitis virus infection.Hepatology. 2006; 43: 1295-1302Crossref PubMed Scopus (162) Google Scholar A number of case-control and cohort studies have linked diabetes to about a 2-fold increased risk of HCC,8Lai M.S. Hsieh M.S. Chiu Y.H. et al.Type 2 diabetes and hepatocellular carcinoma: a cohort study in high prevalence area of hepatitis virus infection.Hepatology. 2006; 43: 1295-1302Crossref PubMed Scopus (162) Google Scholar, 9Adami H.O. Chow W.H. Nyrn O. et al.Excess risk of primary liver cancer in patients with diabetes mellitus.J Natl Cancer Inst. 1996; 88: 1472-1477Crossref PubMed Scopus (333) Google Scholar, 10Wideroff L. Gridley G. Mellemkjaer L. et al.Cancer incidence in a population-based cohort of patients hospitalized with diabetes mellitus in Denmark.J Natl Cancer Inst. 1997; 89: 1360-1365Crossref PubMed Scopus (568) Google Scholar, 11La Vecchia C. Negri E. Decarli A. et al.Diabetes mellitus and the risk of primary liver cancer.Int J Cancer. 1997; 73: 204-207Crossref PubMed Scopus (139) Google Scholar, 12Lagiou P. Kuper H. Stuver S.O. et al.Role of diabetes mellitus in the etiology of hepatocellular carcinoma.J Natl Cancer Inst. 2000; 92: 1096-1099Crossref PubMed Scopus (147) Google Scholar, 13El-Serag H.B. Richardson P.A. Everhart J.E. The role of diabetes in hepatocellular carcinoma: a case-control study among United States veterans.Am J Gastroenterol. 2001; 96: 2462-2467Crossref PubMed Google Scholar, 14El-Serag H.B. Tran T. Everhart J.E. Diabetes increases the risk of chronic liver disease and hepatocellular carcinoma.Gasteroenterology. 2004; 126: 460-468Abstract Full Text Full Text PDF PubMed Scopus (993) Google Scholar, 15Davila J.A. Morgan R.O. Shaib Y. et al.Diabetes increases the risk of hepatocellular carcinoma in the United States: a population based case control study.Gut. 2005; 54: 533-539Crossref PubMed Scopus (551) Google Scholar, 16Hassan M.M. Hwang L.Y. Hatten C.J. et al.Risk factors for hepatocellular carcinoma: synergism of alcohol with viral hepatitis and diabetes mellitus.Hepatology. 2002; 36: 1206-1213Crossref PubMed Scopus (633) Google Scholar, 17Jee S.H. Ohrr H. Sull J.W. et al.Fasting serum glucose level and cancer risk in Korea men and women.JAMA. 2005; 293: 194-202Crossref PubMed Scopus (792) Google Scholar, 18Coughlin S.S. Calle E.E. Teras L.R. et al.Diabetes mellitus as a predictor of cancer mortality in a large cohort of US adults.Am J Epidemiol. 2004; 159: 1160-1167Crossref PubMed Scopus (659) Google Scholar, 19Yuan J.M. Govindarajan S. Arakawa K. et al.Synergism of alcohol, diabetes, and viral hepatitis on risk of hepatocellular carcinoma in blacks and whites in the U.S..Cancer. 2004; 101: 1009-1017Crossref PubMed Scopus (245) Google Scholar, 20Inoue M. Iwasaki M. Otani T. et al.Diabetes mellitus and the risk of cancer: results from a large-scale population-based cohort study in Japan.Arch Intern Med. 2006; 166: 1871-1877Crossref PubMed Scopus (447) Google Scholar, 21El-Serag H.B. Hampel H. Javadi F. The association between diabetes and hepatocellular carcinoma: a systemic review of epidemiologic evidence.Clin Gastroenterol Hepatol. 2006; 4: 369-430Abstract Full Text Full Text PDF PubMed Scopus (654) Google Scholar including one large cohort study in Taiwan.8Lai M.S. Hsieh M.S. Chiu Y.H. et al.Type 2 diabetes and hepatocellular carcinoma: a cohort study in high prevalence area of hepatitis virus infection.Hepatology. 2006; 43: 1295-1302Crossref PubMed Scopus (162) Google Scholar Among these studies, 3 studies8Lai M.S. Hsieh M.S. Chiu Y.H. et al.Type 2 diabetes and hepatocellular carcinoma: a cohort study in high prevalence area of hepatitis virus infection.Hepatology. 2006; 43: 1295-1302Crossref PubMed Scopus (162) Google Scholar, 15Davila J.A. Morgan R.O. Shaib Y. et al.Diabetes increases the risk of hepatocellular carcinoma in the United States: a population based case control study.Gut. 2005; 54: 533-539Crossref PubMed Scopus (551) Google Scholar, 19Yuan J.M. Govindarajan S. Arakawa K. et al.Synergism of alcohol, diabetes, and viral hepatitis on risk of hepatocellular carcinoma in blacks and whites in the U.S..Cancer. 2004; 101: 1009-1017Crossref PubMed Scopus (245) Google Scholar examined whether diabetes and HCC association varied between different hepatitis infection status, but none investigated the possibility of a modifying effect of hepatitis on obesity/HCC association. With the increasing prevalence of obesity and diabetes, it is important to elucidate the complex relationship between these 2 factors as well as other metabolic factors, such as lipid profile and body composition, and HCC risk. In addition, because hepatitis (including B and C) is a dominant risk factor for HCC, it is very important to determine whether these metabolic factors are associated with HCC differently depending on chronic hepatitis B virus (HBV) and hepatitis C virus (HCV) infection status. We conducted the current analysis using data from a community-based cohort22Chen C.J. Yang H.I. Su J. et al.Risk of hepatocellular carcinoma across a biological gradient of serum hepatitis B virus DNA level.JAMA. 2006; 295: 65-73Crossref PubMed Scopus (2506) Google Scholar, 23Yang H.I. Lu S.N. Liaw Y.F. et al.Hepatitis B e antigen and the risk of hepatocellular carcinoma.N Engl J Med. 2002; 347: 168-174Crossref PubMed Scopus (1057) Google Scholar, 24Iloeje U. Yang H.I. Su J. et al.Predicting cirrhosis risk based on the level of circulating hepatitis B viral load.Gastroenterology. 2006; 130: 678-686Abstract Full Text Full Text PDF PubMed Scopus (1307) Google Scholar that has been followed up for more than 10 years in Taiwan and in which every cohort member had hepatitis B surface antigen (HBsAg) and antibody to HCV (anti-HCV) serostatus tested at enrollment. We aimed to investigate whether obesity and diabetes as well as other metabolic factors are independently associated with HCC, stratified by HBV and HCV serostatus. Furthermore, we sought to explore the possible joint influence of obesity/diabetes and HBV/HCV infections on the risk of HCC to better understand the means through which the elevation in HCC risk is associated with them. To our knowledge, this is the first long-term follow-up study that investigates the relationship between obesity/diabetes as well as other metabolic factors and risk of HCC completely stratifying by HBsAg and anti-HCV serostatus, therefore minimizing the effect of HBV and HCV infections. A detailed description of the recruitment and blood sample collection and assay procedures was reported previously.22Chen C.J. Yang H.I. Su J. et al.Risk of hepatocellular carcinoma across a biological gradient of serum hepatitis B virus DNA level.JAMA. 2006; 295: 65-73Crossref PubMed Scopus (2506) Google Scholar, 23Yang H.I. Lu S.N. Liaw Y.F. et al.Hepatitis B e antigen and the risk of hepatocellular carcinoma.N Engl J Med. 2002; 347: 168-174Crossref PubMed Scopus (1057) Google Scholar, 24Iloeje U. Yang H.I. Su J. et al.Predicting cirrhosis risk based on the level of circulating hepatitis B viral load.Gastroenterology. 2006; 130: 678-686Abstract Full Text Full Text PDF PubMed Scopus (1307) Google Scholar Briefly, 47,079 male and 42,214 female residents (total, 89,293) of 7 townships that covered the northern, central, southern, and remote islet of Taiwan were invited to participate in a cancer screening program between 1991 and 1992. Among them, a total of 11,973 men (25%) and 11,847 women (28%) agreed to participate with written informed consent. Information on demographic and lifestyle factors was collected via a structured personal interview by well-trained public health nurses. Overnight fasting blood samples were collected at baseline, and tests on serum HBsAg, hepatitis B e antigen, HBV DNA levels, and anti-HCV status were performed using commercial kits and detailed laboratory procedures as described previously.22Chen C.J. Yang H.I. Su J. et al.Risk of hepatocellular carcinoma across a biological gradient of serum hepatitis B virus DNA level.JAMA. 2006; 295: 65-73Crossref PubMed Scopus (2506) Google Scholar, 23Yang H.I. Lu S.N. Liaw Y.F. et al.Hepatitis B e antigen and the risk of hepatocellular carcinoma.N Engl J Med. 2002; 347: 168-174Crossref PubMed Scopus (1057) Google Scholar, 24Iloeje U. Yang H.I. Su J. et al.Predicting cirrhosis risk based on the level of circulating hepatitis B viral load.Gastroenterology. 2006; 130: 678-686Abstract Full Text Full Text PDF PubMed Scopus (1307) Google Scholar All participants gave informed consent to participate in this study, and the data collection procedures were reviewed and approved by the Institutional Review Board of the College of Public Health, National Taiwan University. The metabolic factors available for this analysis included obesity, history of diabetes and hypertension, and serum levels of total cholesterol and triglycerides (TG). The presence of obesity was grouped by measurement of body mass index (BMI; kg/m2) and waist circumference (cm). Central obesity was defined as a waist circumference >90 cm for men and >80 cm for women, according to criteria used in the Asian population.25Tan C.E. Ma S. Wai D. et al.Can we apply the National Cholesterol Education Program Adult Treatment Panel definition of the metabolic syndrome to Asians?.Diabetes Care. 2004; 27: 1182-1186Crossref PubMed Scopus (694) Google Scholar BMI was categorized into normal (<23 kg/m2), overweight (23 to <25 kg/m2), obese (25 to <30 kg/m2), and extreme obese (≥30 kg/m2), based on the guidelines for the adult Asian population.26WHO Expert ConsultationAppropriate body-mass index for Asian populations and its implications for policy and intervention strategies.Lancet. 2004; 363: 1077Google Scholar A personal history of diabetes mellitus and hypertension was obtained from the interview. Serum levels of TG ≥150 mg/dL and cholesterol ≥240 mg/dL were considered to be indicative of hyperlipidemia. No assays on high-density lipoprotein and low-density lipoprotein cholesterol were performed at study entry. For participants who tested positive for HBsAg or anti-HCV, baseline screening for liver cancer via abdominal ultrasonography was offered. Unique national identification numbers were used to link with the computerized national cancer registry profiles in Taiwan to identify newly diagnosed HCC cases between January 1, 1985, and December 31, 2004. Subjects who had liver cancer identified through ultrasound screening, through linkage to the national cancer registry, or from personal interview before or at enrollment were excluded from the analysis. The nationwide cancer registry system was implemented in 1978 in Taiwan, which has updated, accurate, and complete information in Taiwan. Because not all liver cancers reported to the national cancer registry had histologic confirmation, chart reviews were performed on 164 liver cancer cases in HBsAg-positive carriers to further validate the primary HCC diagnosis. As a result, all HCC cases in this analysis met at least one of the following criteria: histopathologic confirmation, detection by at least 2 imaging tools (ultrasonography, angiography, or computed tomography), one imaging diagnosis, plus serum α-fetoprotein level ≥400 ng/mL. Person-years for each participant were calculated from the date of questionnaire interview to the date of HCC diagnosis, death, or December 31, 2004, whichever came first. Participants who were alive and not diagnosed with HCC were censored on December 31, 2004. Potential confounding factors that may be associated with both metabolic factors and HCC examined in this analysis included age (modeled as a continuous variable), sex, cigarette smoking status (never/current/past), habitual alcohol consumption (no/yes), habitual betel nuts chewing (no/yes), and education levels that were grouped into 3 categories. All analyses were stratified by serostatus of HBsAg and anti-HCV, and persons who were both HBsAg and anti-HCV positive (106 men and 112 women) were excluded from the current analysis due to small numbers of HCC cases (13 cases). The relative risk (RR) and 95% confidence interval (CI) were estimated by Cox proportional hazards regression models. Interaction terms between HBsAg/anti-HCV status and obesity (BMI <30/≥30 kg/m2), central obesity (no/yes), and history of diabetes (no/yes) were used in the model with adjustment for age, sex, cigarette smoking, habitual alcohol consumption, and education levels attained to investigate whether the existence of HBV or HCV infections modifies the association between obesity/diabetes and HCC risk in a multiplicative scale. Another way to examine the joint effect of obesity, diabetes, and seropositivity of HBsAg or anti-HCV was estimated by etiologic fraction (EF), which indicated the percentage of HCC cases with one metabolic factor and one viral risk factor (eg, diabetes and HBsAg positive) that was due to the synergism ([RR11 – RR01 – RR10 + RR00]/RR11).27Walker A.M. Proportion of disease attributable to the combined effect of two factors.Int J Epidemiol. 1981; 10: 81-85Crossref PubMed Scopus (63) Google Scholar The range of departure from additivity was estimated from the 95% CI of an EF based on the methods described by Walker27Walker A.M. Proportion of disease attributable to the combined effect of two factors.Int J Epidemiol. 1981; 10: 81-85Crossref PubMed Scopus (63) Google Scholar and the inclusion of zero indicating no statistically significant association. Because participants with both chronic HBV and HCV infections were excluded, the EF was based on RRs of seropositivity of HBsAg or anti-HCV alone, diabetes alone, and both factors in comparison with participants who were seronegative for both HBsAg and anti-HCV and diabetes-free. A number of metabolic factors, including central obesity, hypertriglyceridemia, hypercholesterolemia, and history of diabetes and hypertension, were examined among those without chronic HBV and HCV infections to determine whether the number was associated with increasing HCC risk in a dose-response fashion. Trends across levels of categorical variables were assessed by testing the statistical significance of a single trend variable coded as the category of exposure. To further elucidate the relationship between serum TG levels and HCC, as well as other HCC risk factors, an analysis using a general linear regression model with serum TG levels as a dependent variable was conducted. Serum HBV DNA level, HCC diagnosis during follow-up, history of liver cirrhosis, serum alanine aminotransferase level, and the metabolic factors including BMI, central obesity, and history of diabetes mellitus and hypertension were all put in this model to determine if the inverse association of TG level and HCC remained after adjusting for the effects of other metabolic factors. All analyses were performed using Stata statistical software28Stata version 8.0. College Station, TX: Stata Corp.Google Scholar (version 8.0; Stata Corp, College Station, TX). All analyses were repeated without 25 participants who reported a history of liver cirrhosis at enrollment and the results remained the same, indicating that prevalent cirrhosis may not be responsible for the association observed. Similarly, HCC cases that appeared within 6–12 months of follow-up may be present but undiagnosed at recruitment and their BMI and profile of metabolic factors may be different due to HCC. We repeated analyses excluding 30 HCC cases diagnosed within 12 months of follow-up, and the conclusion remained the same. The concern that these undiagnosed HCC cases may be leaner and with different metabolic profiles and may influence the risk estimation should be minimal because the majority (22) of them belonged to the HBsAg-positive/anti-HCV–negative group, which was found to have no association with BMI, and none belonged to the anti-HCV–positive/HBsAg-negative group, in which higher BMI increased HCC risk. Based on statistical power consideration, we chose to keep all study subjects in the analyses. Among the 23,567 participants in the current analysis, the majority (18,541; 78.7%) was seronegative for both HBsAg and anti-HCV, 3931 (16.7%) were seropositive for HBsAg only, and 1095 (4.6%) were seropositive for anti-HCV at enrollment. The frequency distribution or mean of selected HCC risk factors in 3 groups is presented in Table 1. Participants seropositive for anti-HCV were older, were higher in female proportion (57.5%), were heavier, and had the highest proportion of diabetes mellitus (2.7%), hypertension (5.3%), elevated alanine aminotransferase level (16.5%), and history of liver cirrhosis (0.6%). HBsAg-seropositive participants were youngest, were higher in male proportion (59.3%), and had higher proportions of current smokers (29.1%), habitual alcohol drinkers (12.0%), and betel nuts chewers (8.3%). They had the lowest proportion of hypercholesterolemia (7.7%) and hypertriglyceridemia (21.3%) and tended to be leaner than the other 2 groups. Participants who were seronegative for both markers had the highest mean total cholesterol and TG levels as well as the highest proportions with hypercholesterolemia (9.9%) and hypertriglyceridemia (30.3%). A total of 291 newly diagnosed HCC cases were identified during the follow-up period. The proportion developing HCC was similar in 2 groups with chronic HBV or HCV infection (4.8% and 4.7%, respectively) but only 0.3% in the group with neither infection, showing an incidence of HCC per 105 person-years of 387.3, 381.9, and 22.7, respectively.Table 1Frequency Distribution of Selected Risk Factors for HCC Stratified by Chronic HBV and HCV Infection StatusHBsAg negative and anti-HCV negative, n (%)HBsAg positive and anti-HCV negative, n (%)HBsAg negative and anti-HCV positive, n (%)(n = 18,541)(n = 3931)(n = 1095)Age at recruitment (y) <352237 (12.1)604 (15.4)69 (6.3) 35–39.93060 (16.5)693 (17.6)117 (10.7) 40–44.92676 (14.4)618 (15.7)103 (9.4) 45–49.92176 (11.7)473 (12.0)140 (12.8) 50–54.92774 (15.0)600 (15.3)196 (17.9) ≥555618 (30.3)943 (24.0)470 (42.9) Mean ± SD47.4 ± 10.045.9 ± 9.850.8 ± 9.3Sex Male9060 (48.9)2331 (59.3)465 (42.5) Female9481 (51.1)1600 (40.7)630 (57.5)Educational level No formal education3967 (21.4)693 (17.6)349 (31.9) Primary school7712 (41.6)1565 (39.8)515 (47.0) Junior high school or more6852 (37.0)1673 (42.6)231 (21.1) Unknown1000Cigarette smoking at enrollment Never13,270 (71.8)2647 (67.5)793 (72.7) Current4557 (24.7)1140 (29.1)262 (24.0) Past646 (3.5)134 (3.4)36 (3.3) Unknown68104Habitual alcohol consumption No16,544 (89.4)3455 (88.1)1007 (92.1) Yes1953 (10.6)469 (12.0)86 (7.9) Unknown4472Habitual betel nuts chewing No17,423 (94.2)3598 (91.7)1023 (93.7) Yes1079 (5.8)327 (8.3)69 (6.3) Unknown3963Serum TG level (mg/dL) <15012,876 (69.7)3076 (78.7)840 (77.4) ≥1505591 (30.3)832 (21.3)245 (22.6) Mean ± SD138.9 ± 120.4117.6 ± 89.2120.6 ± 103.2 Unknown742310Serum total cholesterol level (mg/dL) <24016,645 (90.1)3610 (92.3)983 (90.7) ≥2401826 (9.9)300 (7.7)101 (9.3) Mean ± SD184.9 ± 44.7178.4 ± 43.3178.4 ± 45.1 Unknown702111BMI (kg/m2) <237429 (40.1)1327 (41.4)425 (38.8) 23–24.94385 (23.7)934 (23.8)238 (21.7) 25–29.95795 (31.3)1180 (30.0)365 (33.3) ≥30932 (5.0)190 (4.8)67 (6.1) Mean ± SD24.0 ± 3.424.0 ± 3.424.2 ± 3.5Central obesityaWaist circumference >90 cm in men and >80 cm in women. No13,366 (72.3)2938 (74.9)728 (66.6) Yes5122 (27.7)985 (25.1)365 (33.4) Unknown5382History of diabetes No18,204 (98.2)3862 (98.2)1060 (96.8) Yes284 (1.5)62 (1.6)30 (2.7) Unknown5375History of hypertension No17,750 (95.7)3793 (96.5)1032 (92.5) Yes734 (4.0)131 (3.3)58 (5.3) Unknown5775Serum aspartate aminotransferase level at enrollment (IU/L) <4518,269 (98.5)3742 (95.2)907 (82.8) ≥45272 (1.5)189 (4.8)188 (17.2)Serum alanine aminotransferase level at enrollment (IU/L) <4518,154 (97.9)3705 (94.3)914 (83.5) ≥45387 (2.1)226 (5.8)181 (16.5)History of liver cirrhosis No18,474 (99.6)3915 (99.6)1084 (99.0) Yes10 (0.1)9 (0.2)6 (0.6) Unknown5775Newly diagnosed HCC during follow-up No18,488 (99.7)3744 (95.2)1044 (95.3) Yes53 (0.3)187 (4.8)51 (4.7)a Waist circumference >90 cm in men and >80 cm in women. Open table in a new tab Among participants without chronic HBV and HCV infections (Table 2), older age, male gender, cigarette smoking, habitual alcohol consumption, and serum aspartate aminotransferase or alanine aminotransferase levels ≥45 U/L were associated with an increased risk of HCC. Extreme obesity and history of diabetes were associated with an increased risk of HCC (RR, 2.32; 95% CI, 0.93–5.81 for BMI ≥30 vs <23 kg/m2; RR, 2.88; 95% CI, 0.89–9.29 for diabetes). For HBsAg-seropositive participants (Table 2), older age, male gender, habitual alcohol consumption, and serum aspartate aminotransferase or alanine aminotransferase levels ≥45 U/L were significantly associated with an increased risk of HCC. Diabetes was associated with a relative HCC risk of 2.17 (95% CI, 1.07–4.43); neither BMI nor central obesity were significantly related to HCC risk. Hypertriglyceridemia was associated with a 32% reduced risk of HCC (RR, 0.6; 95% CI, 0.47–1.00) in HBsAg-seropositive subjects.Table 2Age- and Sex-Adjusted RRs of Incident HCC in Relation to Selected Risk Factors Stratified by Chronic HBV and HCV Infection StatusHBsAg negative and anti-HCV negative (n = 18,541)HBsAg positive and anti-HCV negative (n = 3931)HBsAg negative and anti-HCV positive (n = 1095)HCC (n = 53)Person-years (233,567)RR (95% CI)HCC (n = 187)Person-years (48,284)RR (95% CI)HCC (n = 51)Person-years (13,354)RR (95% CI)Age at recruitment (y) <35328,9301.00678001.000900 35–39.9239,4680.48 (0.08–2.88)1088011.49 (0.54–4.10)11485 40–44.9234,2340.56 (0.09–3.35)1478802.34 (0.90–6.09)213071.00 45–49.9327,7561.03 (0.21–5.12)3957299.07 (3.84–21.4)317772.14 (0.43–10.6) 50–54.91335,1353.46 (0.99–12.2)3372695.98 (2.51–14.3)1123505.83 (1.63–20.9) 55–59.9935,4092.34 (0.63–8.65)4364128.98 (3.82–21.1)1429545.97 (1.71–20.8) ≥602132,6365.53 (1.64–18.6)42439412.2 (5.17–28.6)2025829.51 (2.82–32.0)Sex Female15121,0821.003120,1881.0023784