Relationship Between Metabolic Syndrome, Alanine Aminotransferase Levels, and Liver Disease Severity in a Multiethnic North American Cohort With Chronic Hepatitis B

Hepatitis B virus (HBV) infection is associated with a substantial global health burden: roughly 250 million persons are chronically infected worldwide (1). Recent data suggest that the number of foreign-born individuals with chronic HBV living in the U.S. may be greater than previously reported, and the total number of Americans with chronic HBV may be as high as 2.2 million (2). As such, HBV represents a significant public health burden in North America. Chronic HBV has a variable course, ranging from inactive infection with minimal viral replication and/or mild or no liver damage to aggressive disease with severe hepatic inflammation, leading to progressive fibrosis and risk of cirrhosis, liver failure, and hepatocellular carcinoma (HCC) (3). In addition to viral factors, host factors have been identified that influence disease course and long-term prognosis (4).

Metabolic syndrome (MS) is defined as a constellation of metabolic abnormalities including central obesity, hypertriglyceridemia, low HDL, hypertension, and impaired fasting glucose (5). MS has emerged as an important and potentially modifiable risk factor for chronic liver disease progression. In addition to a direct association with nonalcoholic fatty liver disease, MS may influence the outcomes of other chronic liver diseases (4,6). Among patients with chronic hepatitis C virus infection, those with evidence of MS have an accelerated disease course and an increased risk for HCC (7). Although components of MS have been associated with overall patient mortality and HCC risk in HBV (4,8), the effects of MS on other outcomes in chronic HBV infection are not as well documented.

Population-based studies from Asia have reported variable findings; some show an increased prevalence of MS among individuals with HBV infection (9), whereas others find no or a negative association (10). With increasing rates of obesity and its associated metabolic complications, including type 2 diabetes, in North America, the prevalence of MS is also likely to increase among persons with chronic HBV infection (11). Large observational studies have shown that chronic HBV infection is also associated with an increased risk for type 2 diabetes. Although the association of chronic HBV infection and type 2 diabetes was initially observed in Asian but not in U.S. cohorts, we recently reported that the prevalence of type 2 diabetes was higher in HBV-infected individuals than in the general North American population and was associated with longer residence in North America, especially among non-Asians (12). These results suggest that environmental factors, likely primarily dietary, influence the risk for type 2 diabetes and other metabolic risk factors among HBV-infected immigrants. However, limited data exist on whether the presence of MS alters the course of HBV infection (13), especially among North Americans with chronic HBV infection. Moreover, the impact of MS on laboratory tests of liver disease activity and noninvasive markers of liver fibrosis within the context of HBV has not been well defined. Indeed, assessment of the phase of chronic HBV infection, HBV disease monitoring, and decision to initiate HBV therapy rely on serial serum alanine aminotransferase (ALT) levels, HBV viral parameters, and liver disease severity (14). In addition, achieving sustained suppression of HBV viral replication with anti-HBV therapy is associated with normalization of serum ALT (defined as ≤19 units/L in females and ≤30 units/L in males) and improvement in liver histology (14).

In this study we examined the prevalence of MS and its association with liver disease parameters in a large cohort of ethnically diverse North American patients with chronic HBV infection who were followed for a median of 3.7 years.

Subjects enrolled in the Hepatitis B Research Network (HBRN) Adult Cohort Study from January 2011 to February 2014 were considered for inclusion in this analysis. The HBRN Adult Cohort Study is a prospective study of hepatitis B surface antigen–positive adult patients sponsored by the National Institute of Diabetes and Digestive and Kidney Diseases, and it comprises 21 adult liver centers in the U.S. and Canada, as previously described (15). In brief, the HBRN Adult Cohort Study enrolled hepatitis B surface antigen–positive persons 18 years of age or older who did not have a history of hepatic decompensation, HCC, solid organ or bone marrow transplantation, or HIV coinfection and who were not receiving hepatitis B antiviral therapy. If antiviral therapy was initiated any time after enrollment, subjects continued in the study. Participants were evaluated at enrollment, week 12, week 24, and every 24 weeks thereafter. For this study we excluded participants who had acute hepatitis B infection, who were pregnant, or in whom the presence/absence of MS, as defined below, could not be assessed. Up to 240 weeks (5 years) of follow-up were included in this analysis.

All protocols were approved by the HBRN Steering Committee and the institutional review boards (research ethics board in the case of the Toronto site) of the participating sites. All participants provided written informed consent.

MS was defined as the presence of at least three of five factors at baseline (5): 1) abnormal glucose metabolism (fasting glucose ≥100 mg/dL, use of antidiabetes agents, and/or a history of diabetes); 2) high-risk waist circumference (≥80 cm for Asian women and ≥88 cm for women of all other races, ≥90 cm for Asian men and ≥102 cm for men of all other races); 3) low HDL cholesterol level (<50 mg/dL for women and <40 mg/dL for men); 4) elevated triglyceride level (fasting value ≥150 mg/dL or taking lipid-lowering agents at baseline); and 5) elevated blood pressure (systolic pressure ≥130 mmHg and/or diastolic pressure ≥85 mmHg, taking antihypertensive medications at baseline, and/or a history of hypertension). Alcohol consumption was graded as none or minimal (<1 drink/month), moderate (more than none or minimal but ≤4 drinks/day or 14 drinks/week in men, ≤3 drinks/day or 7 drinks/week in women), or heavy (more than moderate or binge drinking, defined as ≥5 drinks on at least 1 day in the previous month) (16). The upper limit of normal (ULN) for ALT was 30 units/L for males and 20 units/L for females. Because transient elastography was only approved for use in the United States in 2013, this noninvasive assessment of liver fibrosis was not available when the HBRN Adult Cohort Study was initiated. Therefore, the aspartate aminotransferase (AST)–to–platelet ratio index (APRI) score was used to estimate liver fibrosis. The APRI was calculated using the equation (17)

APRI score was chosen over the Fibrosis-4 score for evaluating fibrosis to avoid collinearity with age that was used in the models assessing the impact of MS on fibrosis. Cirrhosis was defined as APRI >2.0.

Biopsies performed for clinical indication within 1 year of entry into the HBRN Adult Cohort Study were reviewed by a central pathology committee comprising pathologists from participating clinical centers. Hematoxylin and eosin and Masson trichrome stains were available for all cases. Biopsies were scored for inflammation and fibrosis using the method described by Ishak et al. (18). Steatosis was graded based on the proportion of steatotic hepatocytes assessed at low magnification: grade 0 indicates no steatosis; grade 1, <5% steatosis; grade 2, from 5% to 33% steatosis; grade 3, >33% to 67% steatosis; and grade 4, >67% steatosis. Perisinusoidal fibrosis was assessed as 0 for none, 1 for mild (visible only with trichrome staining), and 2 for moderate (evident with hematoxylin and eosin staining). The presence of concurrent steatohepatitis was based on findings of steatosis, ballooning injury with or without Mallory-Denk bodies, and perisinusoidal fibrosis in an appropriate architectural pattern (19). Possible steatohepatitis was distinguished from definite steatohepatitis based on the quality of the characteristic features.

Descriptive statistics included median and range and mean ± SD, as appropriate. Continuous variables were compared between those with and those without MS using the nonparametric Kruskal-Wallis test, and categorical variables were compared using the χ² test or Fisher exact test, as appropriate. Linear regression models were used to estimate the adjusted (for age, sex, race, alcohol consumption, HBV DNA levels, and hepatitis B e antigen [HBeAg] status) association between MS and both ALT level and the APRI score at baseline. Linear mixed models were used for a similar analysis that included ALT values and APRI scores both at baseline and during follow-up. All the ALT values and APRI scores were analyzed as individual and separate dependent variables in these models. Note, however, that to meet normality assumptions, the log₂ scale for both ALT and the APRI score was used in the models, and therefore results are presented as ratios rather than differences. All HBV DNA levels included in the models were assessed serially during follow-up. All the other predictor variables were assessed at baseline.

Of the 1,547 subjects enrolled in the HBRN Adult Cohort Study during the study period who met the study inclusion criteria, the presence or absence of MS could be ascertained in 777 (50.2%). The inability to ascertain MS was typically due to a lack of fasting glucose or lipid levels at enrollment. The participants for whom MS could (n = 777) or could not (n = 770) be ascertained did not differ significantly with respect to age (median 43.5 vs. 42.2 years; P = 0.82), sex (47.2% vs. 46.9% female; P = 0.89), race (72.6% vs. 69.7% Asian; P = 0.21), or APRI score (3.5% vs. 3.4% with APRI >2; P = 0.90).

Of 777 participants included in the analysis, 171 (22%) had MS at baseline. Table 1 summarizes the characteristics of participants with and those without MS at baseline. The participants with MS were older, more likely to be born in the U.S. or Canada or to have been immigrants longer in these countries, and more likely to have a known family history of diabetes. With respect to HBV viral parameters, those with MS had significantly lower prevalence of HBeAg (14.2% vs. 30.3%; P = 0.0002) and lower HBV DNA levels (median log₁₀ value 3.23 vs. 3.96; P < 0.0001) than those without MS, presumably as a result of their older age and longer estimated duration of HBV infection. No significant differences were found between groups with respect to baseline ALT or AST levels and APRI scores. However, a significantly larger proportion of participants with MS at baseline had an indeterminant HBV phenotype (20) (those who do not meet the criteria for immune-active HBeAg positive or negative, immune-tolerant, or inactive carrier states, 41.5% vs. 29.5%). This difference was primarily related to a larger proportion of participants with MS at baseline who were HBeAg negative, with low levels of HBV DNA (≤10⁴ IU/mL) but elevated ALT values (38% vs. 23.8% in those without MS at baseline). Of note, nearly one-third of participants with this type of indeterminant phenotype had MS—a proportion significantly larger than that with HBeAg-negative immune-active and inactive carrier HBV phenotypes combined (31.1% vs. 23.1%; P = 0.046).

With respect to histology, liver biopsy within a year of study entry was available for 73 participants (50 without MS and 23 with MS). Median time elapsed from liver biopsy to study entry was similar among those participants without and those with MS (−123 vs. −117 days; P = 0.5). Participants with MS had higher grades of steatosis and higher scores for perisinusoidal fibrosis, and were more likely to have definite steatohepatitis, than those without MS (30% vs. 2%), consistent with the presence of metabolic liver disease. However, scores for inflammation and Ishak fibrosis were similar in the two groups.

Of the 777 participants included in the data analysis, a total of 734 participants had available baseline ALT levels (of whom 160 had MS) and 687 had available baseline APRI scores (of whom 148 had MS) (Supplementary Fig. 1). In multivariable regression analysis, the presence of MS at baseline was not associated with ALT value or APRI score at baseline. HBV DNA level was positively associated with baseline ALT (1 log₁₀ higher HBV DNA was associated with 14% higher ALT) and APRI score (1 log₁₀ higher HBV DNA was associated with 10% higher APRI score). Women had, on average, lower ALT and APRI scores than men (Table 2). In addition, older age was associated with higher baseline APRI scores (8% higher for every decade greater age).

During a median follow-up of 3.7 years, 634 participants (501 without MS and 133 with MS) did not undergo HBV therapy. Overall, relatively small proportions of these participants (18% without MS and 12% with MS) met the American Association for the Study of Liver Diseases (AASLD) laboratory criteria for treatment initiation based on ALT level, HBV DNA, and HBeAg status at baseline. The proportion of untreated participants who met treatment criteria decreased over time to 4% by the 5th year of follow-up. However, a larger percentage of participants with MS received HBV therapy compared with those without MS (23% vs. 16%; P = 0.03). Figure 1 shows ALT levels in the participants in whom HBV therapy was not initiated during the entire follow-up period. ALT levels tended to be higher at most follow-up time points in participants with MS at baseline compared with those without MS. However, this difference was predominantly seen among women (Supplementary Fig. 2), and the absolute value of the differences between the two groups was small. In multivariable models that included all participants and controlled for years of follow-up, receipt and duration of HBV therapy, and other covariates (age, sex, race, alcohol intake, HBV DNA, and HBeAg status), MS was independently associated with higher ALT levels over time, with a magnitude of association similar to that seen at baseline (Table 3). Participants with MS at baseline had, on average, 12% higher ALT levels during follow-up (when combining all time points) compared with those without MS (P = 0.02). As expected, women had on average 27% lower ALT levels during follow-up than men (P < 0.0001), and higher HBV DNA levels were also associated with higher ALT levels over time (21% higher for every log₁₀ HBV DNA increase; P < 0.0001).

Given the influence of sex on ALT levels, we further evaluated the primary effects of both MS and sex, and the interactions between them, in the same multivariable model that included all participants and controlled for years of follow-up, receipt and duration of HBV therapy, and other covariates (age, sex, race, alcohol intake, HBV DNA, and HBeAg status). This showed that MS at baseline was significantly associated with higher ALT levels in women (ratio 1.26; 95% CI 1.10–1.45; P = 0.0011) but not in men (ratio 1.03; 95% CI 0.92–1.16; P = 0.6). In the multivariable model for ALT shown in Table 3, HBeAg positivity seems to be associated with lower ALT values compared with those observed in HBeAg-negative participants. While it should be noted that the linear relationship between HBV DNA and ALT levels is preserved for both the HBeAg-positive and HBeAg-negative groups, the data suggest that the magnitude of this relationship is attenuated in the HBeAg-positive group.

APRI scores during follow-up were associated with older age (8% higher for every decade increase in age; P < 0.0001) and female sex (19% lower; P < 0.0001). However, no statistically significant association was found between MS at baseline and APRI scores over time (P = 0.06).

We recognized that MS status can change during follow-up, so we repeated the multivariable modeling for the longitudinal analysis using only data from participants whose baseline MS status remained constant over time. For this modeling, a total of 624 subjects with available data and unchanged MS status during follow-up were included for the ALT outcome analysis and 587 participants were included for the APRI outcome analysis (Supplementary Fig. 1). When evaluating participants whose MS status remained constant during the follow-up period, the presence of MS was associated with even higher ALT levels—an average of 19% higher (P = 0.003)—whereas the estimates for APRI scores did not change (Supplementary Table 1). Sex did not have a statistically significant differential influence on the relationship between MS and ALT (P = 0.08) or APRI (P = 0.7) in these models.

This large, prospective cohort study of individuals with chronic HBV infection in North America followed for up to 5 years offers a unique opportunity to assess the relationship between MS and serum ALT and surrogate markers of liver disease severity at baseline and longitudinally. We found that the overall prevalence of MS was 22% in this largely Asian (∼70%) cohort. This observed prevalence is similar to those in reports of other Asian populations, in which MS prevalence ranged from 18% to 23% (21). While baseline ALT levels were not increased in the overall cohort in the presence of MS, nearly one-third of patients with high ALT but low HBV DNA levels (≤10⁴ IU/mL) had MS, a proportion that was significantly higher than that in other HBeAg-negative groups with chronic HBV infection. Another important finding was an association between MS at baseline and higher ALT levels over time, even after controlling for patient and viral factors including age, sex, HBV DNA levels, and HBeAg status. This relationship was stronger among those with persistent MS status. While the difference in the absolute values of ALT was numerically small in those with MS and those without MS over time (e.g., at year 1, median ALT 34.5 vs. 30.0 units/L; at year 2, median ALT 32.5 vs. 29 units/L; at year 3, median ALT 28.0 vs. 26.5 units/L), it is recognized that small changes may lead to patients with chronic HBV infection falling within versus outside the cutoff values recommended by treatment guidelines for the initiation of antiviral therapy (14,22,23).

AASLD guidelines recommend the use of ALT levels ≥2 × ULN (the ULN for ALT in healthy adults is defined as 30 units/L for males and 19 units/L for females) as one of the criteria for HBV therapy (14). The risks and benefits of treatment versus observation in persons who are in the “gray zone,” with ALT 1–2 × ULN and low HBV DNA level, have not been well defined. Thus it is recommended that the severity of liver disease (defined by biopsy or noninvasive testing for fibrosis) be considered in treatment decisions for patients in the gray zone (14). Our results highlight the contribution of MS to modest ALT changes and call attention to the need for clinicians to consider metabolic profiles when interpreting changes in ALT. In addition, identification and optimal management (including lifestyle modification and weight loss) of common metabolic abnormalities such as diabetes and hyperlipidemia are critical for those with coexisting HBV disease. In our study, a larger proportion of participants with MS (23% vs. 16% with no MS) received HBV therapy. Although the reasons for treatment initiation per se were not assessed, it is possible that a higher percentage of participants with MS met the ALT criterion for initiating therapy despite HBV DNA levels that may have been below the guideline cutoff or within the gray zone. Alternatively, participants with MS may have had more persistent ALT elevations that prompted a higher rate of treatment initiation when HBV DNA criteria were met.

A significant association was not observed between the presence of MS and a surrogate measure of liver disease severity (APRI score) at baseline and upon follow-up. Several potential explanations are available for this lack of association. First, our follow-up duration may be insufficient to detect modest changes in fibrosis. The HBRN Adult Cohort Study excluded patients receiving antiviral therapy at enrollment, thereby selecting for patients with less active or advanced liver disease. Indeed, most participants had milder liver disease (∼90% had an APRI score <1.0) at baseline. Second, while APRI has been shown to accurately exclude advanced fibrosis, it is not as accurate in differentiating stages of fibrosis or in detecting small changes in liver disease severity over time (24,25). Use of transient elastography (13) or liver biopsy would have been a more ideal means of capturing modest changes in disease severity, but FibroScan was not approved in the U.S. until 2013. Among the subset with available liver biopsies, however, the participants with MS were more likely to have evidence of concomitant fatty liver disease, including higher grades of steatosis and higher rates of perisinusoidal fibrosis and steatohepatitis, compared with those without MS, supporting our hypothesis that concomitant MS contributes to higher ALT and greater disease severity in chronic HBV infection.

A clear association exists between chronic hepatitis C infection with diabetes and insulin resistance (26), and this in turn likely provides the linkage with MS. By contrast, the relationship between HBV infection and MS remains inconclusive; while some studies show a positive association, most show no or inverse associations (10,27–30). Obesity, diabetes, MS, and HBV infection have all been associated with an increased risk for cirrhosis and HCC (31,32). In addition, certain components of MS—namely, diabetes and/or insulin resistance—are also associated with overall mortality in chronic HBV infection (4). These findings suggest that MS and its pathogenesis are important to HBV disease, but further longitudinal follow-up of these patients is required to understand better the mechanisms by which MS can influence the natural history of HBV.

This study has several limitations. We used the standard definition of MS that requires the presence of at least three of five clinical and laboratory components, but ∼50% of our participants had to be excluded from the analysis because data were missing for two or three of these components. However, no differences were found in age, sex, race, and liver fibrosis (APRI) between participants included in this analysis and those excluded because of missing data related to ascertainment of MS. Contrary to our expectations, the presence of MS at baseline was not associated with baseline serum ALT values or APRI scores. This may reflect the selection criterion of not receiving treatment at study entry, thereby potentially excluding participants with more advanced fibrosis or active chronic HBV infection, for whom antiviral therapy is recommended. Also, we anticipated that both MS and HBV would influence ALT levels at baseline, given that participants were not receiving HBV treatment at enrollment. However, longitudinal ALT data that showed an association between MS and higher ALT over time included participants who received HBV therapy, and longitudinal analyses controlled for receipt of HBV treatment. Thus, by adjusting for treatment (which was not possible at baseline), MS is likely to make a greater contribution to the longitudinal ALT values than to those at baseline. Another limitation of this analysis is that only a limited number of participants had undergone a liver biopsy as a standard of care, requiring reliance on APRI score as a surrogate measure of liver fibrosis. It is reassuring, however, that in the subset of participants who had a liver biopsy, we confirmed that features of steatosis, steatohepatitis, and perisinusoidal fibrosis were more prevalent in participants with MS. Last, the median duration of follow-up of 3.7 years was too short to fully appreciate the impact of MS on the progression of HBV-related liver disease. Nevertheless, this is to our knowledge the largest multiethnic North American cohort with detailed demographic, clinical, and virologic parameters to be prospectively monitored for objective evidence of liver disease progression over time.

In conclusion, approximately one in four North American patients with chronic HBV infection has MS. In evaluating the natural history of the disease in patients with untreated chronic HBV infection, we found MS to be independently associated with higher ALT levels over time. While absolute increases in ALT values were modest, they may be of clinical consequence in persons with ALT values at or near the threshold for consideration of antiviral therapy. Understanding the impact of MS on clinically important end points such as cirrhosis requires a longer duration of follow-up, and our multiethnic population, representative of North American patients with chronic HBV infection, will be useful in this endeavor.

Acknowledgments. In addition to the authors, the HBRN acknowledges the contributions of the following: Harvard Consortium: Jianghe Niu, PhD, Asad Javaid, MBBS, Bilal Nasir, MBBS, Ammu Susheela, MBBS, Imad Nasser, MD (Beth Israel Deaconess Medical Center, Boston, MA); and Nifasha Rusibamayila and Cara Foley (Massachusetts General Hospital, Boston, MA). Minnesota Alliance for Research in Chronic Hepatitis B: Alisha C. Stahler and Linda Stadheim, RN (Mayo Clinic Rochester, Rochester, MN); John Lake, MD, and Philip Lacher (University of Minnesota, Minneapolis, MN). Midwest Hepatitis B Consortium: Kathryn Rushing, RN, Rosemary A. Nagy, RDN, LD, MBA, and Jacki Cerkoski, (Saint Louis University School of Medicine, St. Louis, MO); Debra DeMarco Shaw, RN, BSN, Lisa Kessels, RN, and Michael K. Klebert, PhD, RN, ANP-BC (Washington University School of Medicine in St. Louis, St. Louis, MO). University of Toronto Consortium: Seham Noureldin, PhD, Danie La, RN, Lucie Liu, MSc, CCRP, Diana Kaznowski, RN, Jiayun Chen, Doinita Vladutu, and Orlando Cerocchi (Toronto General Hospital, Toronto, Ontario). HBV CRN North Texas Consortium: Debra Rowan, LVN (Division of Digestive and Liver Diseases, University of Texas Southwestern Medical Center at Dallas, Dallas, TX), Sheila Bass (University of Texas Southwestern, Dallas, TX), and Barbara Lilly, BS (Baylor University Medical Center, Dallas, TX). Los Angeles Hepatitis B Consortium: Samuel French, MD, Velma Peacock, RN (David Geffen School of Medicine at UCLA, Los Angeles, CA). San Francisco Hepatitis B Research Group Consortium: Ashley Shobe, MS, Rayshawnda Davis, Romuald Kuras, Claudia Ayala, MS, Ivy Lau, BS (University of California, San Francisco, San Francisco, CA); Veronika Podolskaya, BS, NCPT, Anna von Bakonyi, LVN, CCRC, and Nata DeVole, RN (California Pacific Medical Center Research Institute). Michigan Hawaii Consortium: Barbara McKenna, MD, Karen Choi, MD, Kelly Oberhelman, PAC, Sravanthi Kaza, BPharm, Abigail Bowen, BS (University of Michigan, Ann Arbor, MI); Sumodh Kalathil, MD, Leslie Huddleston, NP, Richmond Wong (The Queen’s Medical Center, University of Hawaiʽi, Honolulu, HI). Chapel Hill, NC, Consortium: A. Sidney Barritt, MD, Tiffany Marsh, BA, Vikki Metheny, ANP, Danielle Cardona, PA-C (University of North Carolina at Chapel Hill, Chapel Hill, NC). Virginia Commonwealth University Medical Center: Velimir A. Luketic, MD, Paula G. Smith, RN, BSN, Charlotte Hofmann, RN (Virginia Commonwealth University Health System, Richmond, VA). PNW/Alaska Clinical Center Consortium: Alycia Wolfstone, RN, MN (University of Washington Medical Center, Seattle, WA), Jody Mooney, Lupita Cardona-Gonzalez (Virginia Mason Medical Center, Seattle WA). Liver Diseases Branch, National Institute of Diabetes and Digestive and Kidney Diseases (NIDDK), National Institutes of Health (NIH): Nancy Fryzek, RN, BSN, Elenita Rivera, BSN, Nevitt Morris, Vanessa Haynes-Williams, Amy Huang, RN. Liver Disease Research Branch, NIDDK, NIH: Averell H. Sherker, MD, Rebecca J. Torrance, RN, MS, Sherry R. Hall, MS. Immunology Center: Mary E. Valiga, RN, Keith Torrey, BS, Danielle Levine, BS, James Keith, BS, Michael Betts, PhD (University of Pennsylvania, Philadelphia, PA), Luis J. Montaner, DVM, DPhil (Wistar Institute, Philadelphia, PA). Data Coordinating Center: Frani Averbach, MPH, RDN, Tamara Haller, Regina Hardison, MS, Stephanie Kelley, MS, Sharon Lawlor, MBA, Hsing-Hua (Sylvia) Lin, MS, Stephen Liu, MS, Andrew Pelesko, BS, Donna Stoliker, Melissa Weiner, MPH, Ella Zadorozny, MS, Qian Zhao, PhD (Graduate School of Public Health, University of Pittsburgh, Pittsburgh, PA).

The authors also acknowledge the HBRN pathology committee members for their contribution to histologic evaluation of the liver biopsies. The HBRN pathology committee members are Atul K. Bhan, MBBS, MD (Massachusetts General Hospital, Boston, MA); Michael S. Torbenson, MD (Mayo Clinic Rochester, Rochester, MN); Barbara McKenna, MD (University of Michigan, Ann Arbor, MI); Elizabeth Brunt, MD (Washington University School of Medicine in St. Louis, St. Louis, MO); Sandra Fischer, MD (Toronto General Hospital, Toronto, Ontario, Canada); Samuel French, MD (David Geffen School of Medicine at UCLA, Los Angeles, CA); Rageshree Ramachandran, MD, PhD (University of California, San Francisco, San Francisco, CA); Cynthia D. Guy, MD (Duke University Medical Center, Durham, NC); Mathew M. Yeh, MD, PhD (University of Washington Medical Center, Seattle, WA); Lan Peng, MD (University of Texas Southwestern Medical Center at Dallas, Dallas, TX); Michael Idowu, MD, MPH (Virginia Commonwealth University Health System, Richmond, VA); Robert Anders, MD, PhD (Johns Hopkins University, Baltimore, MD); Michael A. Nalesnik, MD (University of Pittsburgh Medical Center, Pittsburgh, PA); and David Kleiner, MD, PhD (NIH, Bethesda, MD).

Funding. The HBRN was funded by a U01 grant from the National Institute of Diabetes and Digestive and Kidney Diseases (DK082843 to Lewis R. Roberts, DK082863 to Anna Suk-Fong Lok, DK082864 to Steven H. Belle, DK082866 to Kyong-Mi Chang, DK082867 to Michael W. Fried, DK082871 to Adrian M. Di Bisceglie, DK082872 to William M. Lee, DK082874 to Harry L. A. Janssen, DK082919 to Daryl T.-Y. Lau, DK082923 to Richard K. Sterling, DK082927 to Steven-Huy B. Han, DK082943 to Robert C. Carithers, DK082944 to Norah A. Terrault); an interagency agreement with NIDDK (A-DK-3002-001 to Lilia M. Ganova-Raeva); and support from the intramural program, NIDDK, NIH (to Marc G. Ghany). Additional funding to support this study was provided by the NIH/NIDDK Center of Molecular Studies in Digestive and Liver Diseases (P30DK50306 to Kyong-Mi Chang, the Immunology Center) and a National Institutes of Health Public Health Service Research grant (M01-RR00040 to Kyong-Mi Chang, the Immunology Center); the National Center for Advancing Translational Sciences, National Institutes of Health (UL1TR000058 to Richard K. Sterling); and Clinical and Translation Science Awards (grant nos. UL1TR000004 to Norah A. Terrault, UL1TR001111 to Michael W. Fried, and UL1RR024986 and U54TR001959 to Anna Suk-Fong Lok). Additional support was provided by Gilead Sciences, Inc., and Roche Molecular Systems through a Cooperative Research and Development Agreement through NIDDK.

Duality of Interest. M.K. is a recipient of a research grant (to her institution) from Gilead Sciences, Inc.; Intercept Pharmaceuticals; and AbbVie, Inc.; and she has served on scientific advisory panels for Gilead Sciences, Inc., and Intercept Pharmaceuticals. J.J.F. is a recipient of a research grant (to his institution) or consults for AbbVie, Inc.; ContraVir Pharmaceuticals; Gilead Sciences, Inc.; Janssen Pharmaceutica; Merck; and MedImmune. R.T.C. is a recipient of research grants (to his institution) from Gilead Sciences, Inc.; Janssen Pharmaceutica; and Bristol-Myers Squibb. N.A.T. is a recipient of research grants (to her institution) from Gilead Sciences, Inc., and Bristol-Myers Squibb. M.L.-M. serves on the speaker bureau and scientific advisory board for AbbVie, Inc., and the speaker bureau for Gilead Sciences, Inc.; Merck; and Simply Speaking, Inc. A.S. is president of Sanyal Biotechnology and has stock options in Genfit, Akarna, Tiziana Life Sciences, Indalo, and Durect; has served as a consultant to AbbVie, AstraZeneca, Nitto Denko, Ardelyx, Conatus, Nimbus, Amarin, Salix, Tobira, Takeda, FibroGen, Jannsen, Gilead, Boehringer Ingelheim, Eli Lilly, Zafgen, Novartis, Pfizer, Immuron, Exalenz, and Genfit; has been an unpaid consultant to Intercept, Echosens, Immuron, Galectin, Fractyl, Synlogic, Novo Nordisk, Afimmune, ChemomAb, Nordic Bioscience, and Bristol-Myers Squibb; his institution has received grant support from Gilead, Salix, Tobira, Bristol-Myers Squibb, Shire, Intercept, Merck, AstraZeneca, Malinckrodt, Cumberland, and Novartis; and he receives royalties from Elsevier and UptoDate. A.S.L. is a recipient of research grants (to her institution) from Gilead Sciences, Inc., and Bristol-Myers Squibb. No other potential conflicts of interest relevant to this article were reported.

Author Contributions. All authors designed the study concept, interpreted and analyzed the data and results, and critically reviewed and revised the manuscript. M.K., M.C.S., M.L., J.J.F., N.A.T., and A.S.L. drafted the manuscript. M.L. cleaned up the data. M.K. is the guarantor of this work and, as such, had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.