Hepatitis B virus infection and decreased risk of nonalcoholic fatty liver disease: A cohort study

• Published in: Hepatology (Baltimore, Md.) Vol. 65; no. 3; pp. 828 - 835
• Main Authors: Joo, Eun‐Jeong, Chang, Yoosoo, Yeom, Joon‐Sup, Ryu, Seungho
• Format: Journal Article
• Language: English
• Published: United States Wiley Subscription Services, Inc 01.03.2017
• Subjects: Adult; Age Distribution; Alcohol use; Cohort Studies; Comorbidity; Confidence Intervals; Female; Health risk assessment; Hepatitis; Hepatitis B; Hepatitis B Surface Antigens - analysis; Hepatitis B virus; Hepatitis B, Chronic - epidemiology; Hepatitis B, Chronic - pathology; Humans; Infections; Liver diseases; Male; Middle Aged; Non-alcoholic Fatty Liver Disease - epidemiology; Non-alcoholic Fatty Liver Disease - pathology; Prevalence; Prognosis; Proportional Hazards Models; Republic of Korea; Retrospective Studies; Risk Assessment; Sensitivity and Specificity; Severity of Illness Index; Sex Distribution
• ISSN: 0270-9139; 1527-3350
• DOI: 10.1002/hep.28917

The presence of an association between chronic hepatitis B virus (HBV) infection and fatty liver is controversial. We examined the association between HBV infection and the development of nonalcoholic fatty liver disease (NAFLD). We conducted a cohort study of 83,339 participants without NAFLD at baseline who underwent serologic testing for hepatitis B surface antigen (HBsAg) between 2002 and 2006 and were followed annually or biennially until December 2014. NAFLD was defined as the presence of ultrasonographic fatty liver in the absence of excessive alcohol use or other identifiable causes. We used a parametric Cox model to estimate adjusted hazard ratios with 95% confidence intervals of incident NAFLD. During 484,736.1 person‐years of follow‐up, 20,200 incident NAFLD cases were identified. In models adjusted for age, sex, year of visit, smoking status, alcohol intake, regular exercise, education level, and body mass index, the adjusted hazard ratio (95% confidence interval) for incident NAFLD comparing HBsAg‐positive to HBsAg‐negative participants was 0.83 (0.73‐0.94). After introducing HBV infection and confounders (including homeostasis model assessment of insulin resistance and metabolic factors) as time‐dependent exposures, the association between HBV infection and decreased risk of incident NAFLD was attenuated but persisted. These associations were consistently observed across clinically relevant, prespecified subgroups. Conclusion: In this large cohort of apparently healthy Korean adults, HBsAg seropositivity was associated with lower risk of developing NAFLD, indicating a possible effect of HBV infection on the pathogenesis of NAFLD development. (Hepatology 2017;65:828‐835).