Advertisement

Coffee and Caffeine Are Associated With Decreased Risk of Advanced Hepatic Fibrosis Among Patients With Hepatitis C

  • Natalia Khalaf
    Affiliations
    Section of Gastroenterology and Hepatology, Michael E. DeBakey VA Medical Center and Baylor College of Medicine, Houston, Texas

    Section of Health Services Research, Michael E. DeBakey VA Medical Center and Baylor College of Medicine, Houston, Texas
    Search for articles by this author
  • Donna White
    Affiliations
    Section of Gastroenterology and Hepatology, Michael E. DeBakey VA Medical Center and Baylor College of Medicine, Houston, Texas

    Section of Health Services Research, Michael E. DeBakey VA Medical Center and Baylor College of Medicine, Houston, Texas
    Search for articles by this author
  • Fasiha Kanwal
    Affiliations
    Section of Gastroenterology and Hepatology, Michael E. DeBakey VA Medical Center and Baylor College of Medicine, Houston, Texas

    Section of Health Services Research, Michael E. DeBakey VA Medical Center and Baylor College of Medicine, Houston, Texas
    Search for articles by this author
  • David Ramsey
    Affiliations
    Section of Gastroenterology and Hepatology, Michael E. DeBakey VA Medical Center and Baylor College of Medicine, Houston, Texas

    Section of Health Services Research, Michael E. DeBakey VA Medical Center and Baylor College of Medicine, Houston, Texas
    Search for articles by this author
  • Sahil Mittal
    Affiliations
    Section of Gastroenterology and Hepatology, Michael E. DeBakey VA Medical Center and Baylor College of Medicine, Houston, Texas

    Section of Health Services Research, Michael E. DeBakey VA Medical Center and Baylor College of Medicine, Houston, Texas
    Search for articles by this author
  • Shahriar Tavakoli-Tabasi
    Affiliations
    Section of Infectious Diseases, Michael E. DeBakey VA Medical Center and Baylor College of Medicine, Houston, Texas
    Search for articles by this author
  • Jill Kuzniarek
    Affiliations
    Section of Health Services Research, Michael E. DeBakey VA Medical Center and Baylor College of Medicine, Houston, Texas
    Search for articles by this author
  • Hashem B. El-Serag
    Correspondence
    Reprint requests Address requests for reprints to: Hashem B. El-Serag, MD, MPH, Michael E. DeBakey VA Medical Center, 2002 Holcombe Boulevard (MS152), Houston, Texas 77030. fax: (713) 748-7359.
    Affiliations
    Section of Gastroenterology and Hepatology, Michael E. DeBakey VA Medical Center and Baylor College of Medicine, Houston, Texas

    Section of Health Services Research, Michael E. DeBakey VA Medical Center and Baylor College of Medicine, Houston, Texas
    Search for articles by this author
Published:March 14, 2015DOI:https://doi.org/10.1016/j.cgh.2015.01.030

      Background & Aims

      Coffee or caffeine has been proposed to protect against hepatic fibrosis, but few data are available on their effects in patients with chronic hepatitis C virus (HCV) infection.

      Methods

      We conducted a cross-sectional study of veterans with chronic HCV infection to evaluate the association between daily intake of caffeinated and decaffeinated coffee, tea, and soda, and level of hepatic fibrosis, based on the FibroSURE test (BioPredictive, Paris, France) (F0–F3, mild [controls] vs F3/F4–F4, advanced). Models were adjusted for multiple potential confounders including age, alcohol abuse, and obesity.

      Results

      Among 910 patients with chronic HCV infection, 98% were male and 38% had advanced hepatic fibrosis. Daily intake of caffeinated coffee was higher among controls than patients with advanced fibrosis (1.37 vs 1.05 cups/d; P = .038). In contrast, daily intake of caffeinated tea (0.61 vs 0.56 cups/d; P = .651) or soda (1.14 vs 0.95 cans/d; P = .106) did not differ between the groups. A higher percentage of controls (66.0%) than patients with advanced fibrosis (57.9%) consumed 100 mg or more of caffeine daily from all sources (P = .014); controls also received a larger proportion of their caffeine from coffee (50.2% vs 43.0%; P = .035). Hepatoprotective effects of an average daily intake of 100 mg or more of caffeine (adjusted odds ratio, 0.71; 95% confidence interval, 0.53–0.95; P = .020) and 1 cup or more of caffeinated tea by non–coffee drinkers (adjusted odds ratio, 0.56; 95% confidence interval, 0.34–0.94; P = .028) persisted after adjustment for confounders, including insulin resistance.

      Conclusions

      A modest daily caffeine intake (as little as 100 mg) may protect against advanced hepatic fibrosis in men with chronic HCV infection. Additional research is needed to confirm these findings in women and in people with other chronic liver diseases.

      Keywords

      Abbreviations used in this paper:

      BMI (body mass index), CI (confidence interval), HALT-C (Hepatitis C Antiviral Long-Term Treatment Against Cirrhosis), HCC (hepatocellular carcinoma), HCV (hepatitis C virus), HOMA-IR (homeostasis model assessment–estimated insulin resistance), IDU (injection drug use), MEDVAMC (Michael E. DeBakey Veterans Affairs Medical Center), MELD (Model for End-Stage Liver Disease), OR (odds ratio)
      Podcast interview: www.gastro.org/cghpodcast. Also available on iTunes.
      Coffee is one of the most popular beverages worldwide, and there is growing evidence that it may provide helpful health benefits in a variety of conditions. Recently, a large, prospective study of more than 50,000 participants reported an inverse, dose-dependent association between coffee drinking and overall mortality.
      • Freedman N.D.
      • Park Y.
      • Abnet C.C.
      • et al.
      Association of coffee drinking with total and cause-specific mortality.
      Other studies have suggested that coffee drinkers are at decreased risk of developing heart disease
      • Andersen L.F.
      • Jacobs Jr., D.R.
      • Carlsen M.H.
      • et al.
      Consumption of coffee is associated with reduced risk of death attributed to inflammatory and cardiovascular diseases in the Iowa Women's Health Study.
      and metabolic syndrome
      • Matsuura H.
      • Mure K.
      • Nishio N.
      • et al.
      Relationship between coffee consumption and prevalence of metabolic syndrome among Japanese civil servants.
      compared with non–coffee drinkers.
      For the past 2 decades, studies have supported that coffee may be protective against the development of liver injury, as manifested by abnormal liver function tests and/or cirrhosis from a variety of hepatic parenchymal disorders.
      • Klatsky A.L.
      • Armstrong M.A.
      Alcohol, smoking, coffee, and cirrhosis.
      • Casiglia E.
      • Spolaore P.
      • Ginocchio G.
      • et al.
      Unexpected effects of coffee consumption on liver enzymes.
      • Tanaka K.
      • Tokunaga S.
      • Kono S.
      • et al.
      Coffee consumption and decreased serum γ-glutamyltransferase and aminotransferase activities among male alcohol drinkers.
      • Nakanishi N.
      • Nakamura K.
      • Suzuki K.
      • et al.
      Effects of coffee consumption against the development of liver dysfunction a 4-year follow-up study of middle-aged Japanese male office workers.
      • Honjo S.
      • Kono S.
      • Coleman M.P.
      • et al.
      Coffee consumption and serum aminotransferases in middle-aged Japanese men.
      • Tverdal A.
      • Skurtveit S.
      Coffee intake and mortality from liver cirrhosis.
      • Ruhl C.E.
      • Everhart J.E.
      Coffee and caffeine consumption reduce the risk of elevated serum alanine aminotransferase activity in the United States.
      • Modi A.A.
      • Feld J.J.
      • Park Y.
      • et al.
      Increased caffeine consumption is associated with reduced hepatic fibrosis.
      Several studies including meta-analyses of case–control and cohort studies reported an inverse association between coffee consumption and hepatocellular carcinoma (HCC) in both the presence and absence of chronic hepatitis C virus (HCV) infection.
      • Ohfuji S.
      • Fukushima W.
      • Tanaka T.
      • et al.
      Coffee consumption and reduced risk of hepatocellular carcinoma among patients with chronic type C liver disease: a case-control study.
      • Bravi F.
      • Bosetti C.
      • Tavani A.
      • et al.
      Coffee drinking and hepatocellular carcinoma risk: a meta-analysis.
      • Larsson S.C.
      • Wolk A.
      Coffee consumption and risk of liver cancer: a meta-analysis.
      • Bravi F.
      • Bosetti C.
      • Tavani A.
      • et al.
      Coffee reduces risk for hepatocellular carcinoma: an updated meta-analysis.
      Coffee has several ingredients that could account for its reported health benefits. Previous studies have suggested that caffeine may account for the hepatoprotective effects of coffee in HCV-related chronic liver disease.
      • Modi A.A.
      • Feld J.J.
      • Park Y.
      • et al.
      Increased caffeine consumption is associated with reduced hepatic fibrosis.
      • Freedman N.D.
      • Everhart J.E.
      • Lindsay K.L.
      • et al.
      Coffee intake is associated with lower rates of liver disease progression in chronic hepatitis C.
      • Costentin C.E.
      • Roudot-Thoraval F.
      • Zafrani E.S.
      • et al.
      Association of caffeine intake and histological features of chronic hepatitis C.
      Other studies have shown that insulin resistance worsens hepatic inflammation in HCV patients,
      • Negro F.
      Insulin resistance and HCV: will new knowledge modify clinical management?.
      and that the protective effect of coffee, at least in part, may be a consequence of coffee intake–associated reduction in insulin resistance and type 2 diabetes mellitus.
      • Van Dam R.M.
      • Hu F.B.
      Coffee consumption and risk of type 2 diabetes a systematic review.
      Few studies have examined the effect of coffee, as well as caffeine intake from noncoffee beverages, on the severity of liver fibrosis in patients with untreated chronic HCV infection after adjustment for insulin resistance and diabetes status.
      • Freedman N.D.
      • Everhart J.E.
      • Lindsay K.L.
      • et al.
      Coffee intake is associated with lower rates of liver disease progression in chronic hepatitis C.
      • Costentin C.E.
      • Roudot-Thoraval F.
      • Zafrani E.S.
      • et al.
      Association of caffeine intake and histological features of chronic hepatitis C.
      By using data from a large cross-sectional study among US veterans with chronic HCV infection, we investigated the association between caffeinated and decaffeinated coffee, tea, and soda on the development of HCV-related advanced liver fibrosis. We also examined the extent to which insulin resistance mediated the association between caffeine intake and the severity of liver fibrosis.

      Methods

       Study Population and Design

      We performed a cross-sectional study at the Michael E. DeBakey Veterans Affairs Medical Center (MEDVAMC) in Houston, Texas. Study participants were chronically HCV-infected veterans, prospectively recruited from a dedicated HCV clinic at the MEDVAMC between January 5, 2009, and November 30, 2013.
      Details about study design have been described previously.
      • White D.L.
      • Tavakoli-Tabasi S.
      • Kuzniarek J.
      • et al.
      Higher serum testosterone is associated with increased risk of advanced hepatitis C-related liver disease in males.
      • Kanwal F.
      • White D.L.
      • Tavakoli-Tabasi S.
      • et al.
      Many patients with interleukin 28B genotypes associated with response to therapy are ineligible for treatment because of comorbidities.
      Veterans ages 18 to 70 years with confirmed HCV viremia and who were not receiving antiviral therapy at the time of recruitment were eligible for inclusion. Exclusion criteria included the following: (1) hepatitis B virus (serum hepatitis B surface antigen positivity) or human immunodeficiency virus co-infection; (2) self-reported or medical record–reported history of liver transplant, decompensated liver disease, or HCC; and (3) a diagnosis of dementia or psychosis that would preclude providing consent. The study was approved by the Institutional Review Boards of the MEDVAMC and Baylor College of Medicine.

       Data Collection and Study Measures

      All study participants completed a detailed computerized questionnaire administered by a research assistant, had fasting venipuncture for performance of clinical laboratory tests, and had anthropometric measurements taken. The questionnaire surveyed lifetime history of alcohol, tobacco, injection drug use (IDU), marijuana use, other recreational drug use, and the presence of comorbid conditions, including diabetes mellitus. Alcohol use was classified as never, current, former, and cumulative years of use, with chronic alcohol abuse defined as 10 or more years of drinking an average of 3 or more drinks per day for men and 2 or more drinks per day for women. Participants who smoked fewer than 100 cigarettes over their entire life were classified as nonsmokers. Participants who self-reported IDU or marijuana use in the study questionnaire were classified as users. All participants with missing information on marijuana use responded to a question about recreational drug use. For those participants who self-reported “yes” to recreational drug use, their VA electronic medical record was searched for the following: International Classification of Diseases, 9th revision, codes for cannabis or marijuana abuse; or documented former or current marijuana use in the progress notes by using the search terms “marijuana” or “cannabis.” We assessed participant levels of physical activity using the validated International Physical Activity Questionnaire.
      • Craig C.L.
      • Marshall A.L.
      • Sjöström M.
      • et al.
      International physical activity questionnaire: 12-country reliability and validity.
      We classified participants as diabetic if they self-reported ever having received a physician’s diagnosis of type 2 diabetes mellitus, were excluded from fasting testing as a result of medical record documentation of a diagnosis of diabetes mellitus type 2, had a fasting blood sugar level higher than 126 mg/dL, or a nonfasting blood sugar level higher than 200 mg/dL based on serologic testing performed at enrollment. In nondiabetic patients, we calculated a homeostasis model assessment–estimated insulin resistance (HOMA-IR) score. A HOMA-IR score of 3 or higher was used as a surrogate marker of insulin resistance in nondiabetic participants. The definition of metabolic syndrome was adopted from the National Cholesterol Education Panel-Third Adult Treatment Panel as the presence of 3 or more of the following: (1) waist circumference greater than 40 inches in men or greater than 35 inches in women; (2) increased serum triglyceride level of 150 mg/dL or higher; (3) reduced high-density lipoprotein–cholesterol level less than 40 mg/dL in men or less than 50 mg/dL in women; (4) use of antihypertensive medications; and (5) increased fasting serum glucose level of 110 mg/dL or higher and/or use of insulin or hypoglycemic medication.
      National Cholesterol Education Program (NCEP) Expert Panel on Detection, Evaluation, and Treatment of High Blood Cholesterol in Adults (Adult Treatment Panel III)
      Third Report of the National Cholesterol Education Program (NCEP) Expert Panel on Detection, Evaluation, and Treatment of High Blood Cholesterol in Adults (Adult Treatment Panel III) final report.
      Information regarding prior antiviral treatments was collected. All participants also had their serologically determined Model for End-Stage Liver Disease (MELD) and body mass index (BMI) scores calculated.
      Information on intake of caffeinated and decaffeinated coffee, tea, and carbonated soda, including use of creamers and sweeteners, also was collected. We calculated the average daily intake of coffee, tea, and soda using standard measures of consumption (eg, 8 oz cups for coffee and tea and 12 oz cans for sodas) in the year preceding study recruitment. Cumulative lifetime exposure to coffee also was gathered by asking about daily intake in each decade of life starting in the 20s. Total caffeine intake was calculated from the sum of reported daily intake from all sources, and further stratified as caffeine from coffee and noncoffee sources. Average caffeine content was estimated as follows: 137 mg for each 8-oz cup of coffee,
      • Michels K.B.
      • Willett W.C.
      • Fuchs C.S.
      • et al.
      Coffee, tea, and caffeine consumption and incidence of colon and rectal cancer.
      30 mg for each 8-oz cup of tea,
      • Iso H.
      • Date C.
      • Wakai K.
      • et al.
      The relationship between green tea and total caffeine intake and risk for self-reported type 2 diabetes among Japanese adults.
      46 mg for each 12-oz can of caffeinated soda,
      • Michels K.B.
      • Willett W.C.
      • Fuchs C.S.
      • et al.
      Coffee, tea, and caffeine consumption and incidence of colon and rectal cancer.
      and 0 mg for each cup or can of a decaffeinated beverage.
      We determined the degree of hepatic fibrosis and inflammation using the FibroSURE test (BioPredictive, Paris, France). The FibroSURE test estimates the level of hepatic fibrosis and inflammation by using a proprietary algorithm incorporating serum levels of α2-macroglobulin, apolipoprotein A1, haptoglobin, total bilirubin, γ-glutamyl-transpeptidase, and alanine aminotransferase; this test has been validated in HCV patients.
      • Halfon P.
      • Bourliere M.
      • Deydier R.
      • et al.
      Independent prospective multicenter validation of biochemical markers (fibrotest-actitest) for the prediction of liver fibrosis and activity in patients with chronic hepatitis C: the fibropaca study.
      • Ngo Y.
      • Munteanu M.
      • Messous D.
      • et al.
      A prospective analysis of the prognostic value of biomarkers (FibroTest) in patients with chronic hepatitis C.
      • Morali G.
      • Maor Y.
      • Klar R.
      • et al.
      Fibrotest-Actitest: the biochemical marker of liver fibrosis–the Israeli experience.
      • Imbert-Bismut F.
      • Ratziu V.
      • Pieroni L.
      • et al.
      Biochemical markers of liver fibrosis in patients with hepatitis C virus infection: a prospective study.
      • Grigorescu M.
      • Rusu M.
      • Neculoiu D.
      • et al.
      The FibroTest value in discriminating between insignificant and significant fibrosis in chronic hepatitis C patients. The Romanian experience.
      • Poynard T.
      • Imbert-Bismut F.
      • Ratziu V.
      • et al.
      Biochemical markers of liver fibrosis in patients infected by hepatitis C virus: longitudinal validation in a randomized trial.
      • Poynard T.
      • Munteanu M.
      • Ngo Y.
      • et al.
      ActiTest accuracy for the assessment of histological activity grades in patients with chronic hepatitis C, an overview using Obuchowski measure.
      FibroSURE test scores are categorized into METAVIR biopsy-based equivalent degrees of hepatic fibrosis with ranges from F0 (no fibrosis) to F4 (cirrhosis), and hepatic inflammation with ranges from A0 (no inflammatory activity) to A3 (severe inflammatory activity). We classified all study participants by FibroSURE results as either advanced fibrosis cases (F3/F4–F3) or mild/absent fibrosis controls (F0–F3), and as advanced inflammatory activity cases (A2/A3–A3) or mild/no inflammatory activity controls (A0–A2).

       Statistical Analyses

      The main exposure variables were intake of caffeinated and decaffeinated coffee (cups per day, any use, ≥1, 2, or 3 cups daily), tea and sodas (cups or cans per day, any use, ≥1 cup or can daily), and estimated amount of caffeine (≥100 mg or 200 mg/d) overall and from each beverage. The main outcome variable was advanced hepatic fibrosis (F3/F4–F4) as determined by FibroSURE test scoring. We compared exposure variables between cases and controls using t tests and analysis of variance for means and the chi-square and Fisher exact test for categoric variables.
      We constructed unconditional logistic-regression models to examine the individual associations between coffee, tea, soda, and caffeine intake and risk of advanced hepatic fibrosis. The first multivariable model adjusted for the following potential confounders: age, chronic alcohol use, and BMI category. Because it is possible that patients with more severe liver disease would avoid drinking coffee, in the second model we further adjusted for MELD scores. In the third model, we also examined the degree of insulin resistance status categorized as diagnosed diabetic, insulin resistant (HOMA-IR ≥3 in nondiabetic participants), and non–insulin resistant (HOMA-IR <3 in nondiabetic participants) as a possible mediating variable. In the fourth model, we adjusted for the presence of the metabolic syndrome. We calculated odds ratios (ORs) and associated 95% confidence intervals (95% CIs) for each model.
      All tests were 2-sided, with statistical significance determined at an α value of .05. Statistical analyses were conducted using IBM SPSS version 21 (SPSS Inc, Chicago, IL).

      Results

       Study Population

      A total of 910 veterans with chronic HCV infection met our inclusion criteria. Based on FibroSURE scoring, 342 of those patients (37.6%) had advanced fibrosis (F3/F4–F4), and 568 (62.4%) had mild fibrosis (controls: F0–F3) (Table 1). Most participants were male (97.6%), African American (54.2%), and chronic alcohol users (54.2%), with no significant differences between cases and controls. Compared with mild fibrosis controls, advanced fibrosis cases were significantly older (average, 1.8 y), had higher BMIs, and were more likely to have type 2 diabetes mellitus and metabolic syndrome. Among nondiabetic participants, advanced fibrosis cases were more likely than controls to be insulin-resistant based on a baseline fasting HOMA-IR score of 3 or higher. Cases also were more likely to have higher MELD scores (>9) and to have received prior HCV antiviral therapy than controls; however, all patients had detectable HCV RNA levels at the time of the study and none were actively receiving treatment. Marijuana use, IDU, level of physical activity, and smoking history were not significantly different between the 2 groups (Table 1).
      Table 1Demographic Characteristics of Advanced Fibrosis HCV Cases Compared With Mild Fibrosis HCV Controls
      Controls (F0–F3)

      N = 568 (%)
      Cases (F3/F4–F4)

      N = 342 (%)
      P value
      Mean age, y (SD)55.89 (5.60)57.74 (4.83)<.001
      Sex
       Male551 (97.0)337 (98.5).145
       Female17 (3.0)5 (1.5)
      Race
       White233 (41.0)148 (43.3).769
       Black313 (55.1)180 (52.6)
       Other22 (3.9)14 (4.1)
      HCV genotype
       1a327 (57.6)199 (58.2).434
       1b119 (21.0)81 (23.7)
       1 NOS25 (4.4)12 (3.5)
       254 (9.5)22 (6.4)
       328 (4.9)19 (5.6)
       Other5 (0.9)1 (0.3)
       Missing10 (1.7)8 (2.3)
      MELD score
       Mean (SD)7.87 (2.42)8.74 (2.62)<.001
       <9495 (87.2)258 (75.4)<.001
       10–1965 (11.4)80 (23.4)
       ≥207 (1.2)4 (1.2)
       Missing1 (0.2)0 (0.0)
      Prior HCV treatment
       Yes50 (8.8)47 (13.7).020
       No515 (90.7)294 (86.0)
       Missing3 (0.5)1 (0.3)
      Years of alcohol abuse, mean (SD)14.49 (14.56)14.29 (14.45).838
      Chronic alcohol abuse
      Ten or more consecutive years of drinking an average of 3 or more drinks per day for men and 2 or more drinks per day for women.
       Yes308 (54.2)185 (54.1).924
       No258 (45.4)157 (45.9)
       Missing2 (0.4)0 (0.0)
      Marijuana use
      Self-reported use or documented ever use of marijuana in the electronic medical record (International Classification of Diseases, 9th revision, code for cannabis or marijuana abuse; or documented former or current marijuana use in the electronic medical record when searched for the following key words: MJ, marijuana, and cannabis).
       Yes234 (41.2)128 (37.4).351
       No312 (54.9)195 (57.0)
       Missing22 (3.9)19 (5.6)
      IDU use
       Yes326 (57.4)205 (59.9).450
       No242 (42.6)137 (40.1)
      BMI categories
       <25181 (31.9)75 (21.9).005
       Overweight (25–29)211 (37.1)145 (42.4)
       Obese (≥30)170 (29.9)119 (34.8)
       Missing6 (1.1)3 (0.9)
      DM/HOMA-IR status
       Diabetic
      Self-reported physician’s diagnosis, medical record documentation, or presence of a fasting blood sugar level of greater than 126 mg/dL, or a nonfasting blood sugar level of greater than 200 mg/dL at enrollment.
      118 (20.8)99 (28.9)<.001
       HOMA-IR ≥3 in nondiabetic participants149 (26.2)129 (37.8)
       HOMA-IR <3 in nondiabetic participants270 (47.5)102 (29.8)
       Missing31 (5.5)12 (3.5)
      Metabolic syndrome
      Metabolic syndrome was defined as the presence of 3 or more of the following: (1) waist circumference greater than 40 inches in men or greater than 35 inches in women; (2) increased serum triglyceride level of 150 mg/dL or higher; (3) reduced high-density lipoprotein cholesterol level less than 40 mg/dL in men or less than 50 mg/dL in women; (4) use of antihypertensive medications; and (5) increased fasting serum glucose level of 110 mg/dL or greater, and/or use of insulin or hypoglycemic medication.
       Yes134 (23.6)142 (41.5)<.001
       No426 (75.0)198 (57.9)
       Missing8 (1.4)2 (0.6)
      Smoking
       Nonsmoker180 (31.7)121 (35.4).404
       <30 pack-years190 (33.4)115 (33.6)
       ≥30 pack-years196 (34.5)105 (30.7)
       Missing2 (0.4)1 (0.3)
      Physical activity
      Based on participant response to the International Physical Activity Questionnaire.
       Low activity171 (30.1)117 (34.2).374
       Moderate activity175 (30.8)94 (27.5)
       High activity222 (39.1)131 (38.3)
      DM, diabetes mellitus; NOS, not otherwise specified.
      a Ten or more consecutive years of drinking an average of 3 or more drinks per day for men and 2 or more drinks per day for women.
      b Self-reported use or documented ever use of marijuana in the electronic medical record (International Classification of Diseases, 9th revision, code for cannabis or marijuana abuse; or documented former or current marijuana use in the electronic medical record when searched for the following key words: MJ, marijuana, and cannabis).
      c Self-reported physician’s diagnosis, medical record documentation, or presence of a fasting blood sugar level of greater than 126 mg/dL, or a nonfasting blood sugar level of greater than 200 mg/dL at enrollment.
      d Metabolic syndrome was defined as the presence of 3 or more of the following: (1) waist circumference greater than 40 inches in men or greater than 35 inches in women; (2) increased serum triglyceride level of 150 mg/dL or higher; (3) reduced high-density lipoprotein cholesterol level less than 40 mg/dL in men or less than 50 mg/dL in women; (4) use of antihypertensive medications; and (5) increased fasting serum glucose level of 110 mg/dL or greater, and/or use of insulin or hypoglycemic medication.
      e Based on participant response to the International Physical Activity Questionnaire.

       Relationship Between Coffee, Tea, and Soda Consumption and Hepatic Fibrosis in the Year Preceding Study Enrollment

       Coffee

      Most participants (54.6%) reported drinking some caffeinated coffee within the year preceding study enrollment, with 47.2% drinking on average 1 or more cups per day (Table 2). In contrast, few participants (<8%) reported decaffeinated coffee use, with less than 5% reporting an average intake of 1 or more cups per day. Mild fibrosis controls had a higher average daily intake of caffeinated coffee compared with advanced fibrosis cases (mean, 1.37 vs 1.05 cups/d; P = .038), and were more likely to drink an average of 1 or more cups of caffeinated coffee daily (49.8% vs 43.0%; P = .045). However, although the proportions of patients consuming 2 or more or 3 or more cups of caffeinated coffee a day were higher in mild fibrosis controls than advanced fibrosis cases, these differences were not statistically different. Decaffeinated coffee intake was not different between the groups (Table 2).
      Table 2Comparison of Beverage Intake Between Advanced Fibrosis HCV Cases and Mild Fibrosis HCV Controls in One Year Preceding Study Enrollment
      Controls (F0–F3)Cases (F3/F4–F4)P value
      Coffee intake, n (%)568342
       Caffeinated coffee
      Cups per day, mean (SD)1.37 (2.39)1.05 (1.92).038
      Any use
       Yes320 (56.3)177 (51.8).179
       No248 (43.7)165 (48.2)
      ≥1 cup daily
       Yes283 (49.8)147 (43.0).045
       No285 (50.2)195 (57.0)
      ≥2 cups daily
       Yes163 (28.7)85 (24.9).208
       No405 (71.3)257 (75.1)
      ≥3 cups daily
       Yes91 (16.0)42 (12.3).123
       No477 (84.0)300 (87.7)
       Decaffeinated coffee
      Cups per day, mean (SD)0.08 (0.37)0.07 (0.39).923
      Any use
       Yes42 (7.4)26 (7.6).908
       No526 (92.6)316 (92.4)
      ≥1 cup daily
       Yes26 (4.6)14 (4.1).730
       No542 (95.4)328 (95.9)
      Tea intake
       Caffeinated tea
      Cups per day, mean (SD)0.61 (1.25)0.56 (1.36).651
      Any use
       Yes401 (70.6)235 (68.7).592
       No167 (29.4)106 (31.0)
       Missing0 (0.0)1 (0.3)
      ≥1 cup daily
       Yes135 (23.8)72 (21.1).356
       No433 (76.2)269 (78.6)
       Missing0 (0.0)1 (0.3)
       Decaffeinated tea
      Any use.423
       Yes22 (3.9)17 (5.0)
       No546 (96.1)324 (94.7)
       Missing0 (0.0)1 (0.3)
      ≥1 cup per day
       Yes562 (98.9)336 (98.2).755
       No6 (1.1)5 (1.5)
       Missing0 (0.0)1 (0.3)
      Soda intake
       Caffeinated soda
      Cans per day, mean (SD)1.14 (1.82)0.95 (1.60).106
      Any use
       Yes445 (78.3)254 (74.3).158
       No123 (21.7)88 (25.7)
      ≥1 can per day
       Yes230 (40.5)111 (32.5).015
       No338 (59.5)231 (67.5)
       Decaffeinated soda
      Cans per day, mean (SD)0.17 (0.62)0.18 (0.65).753
      Any use
       Yes138 (24.3)88 (25.7)
       No430 (75.7)254 (74.3)
      ≥1 can per day.963
       Yes52 (9.2)31 (9.1)
       No516 (90.8)311 (90.9)
      Intake in non–coffee drinkers, n (%)248165
       Caffeinated tea in non–coffee drinkers
      Any use
       Yes169 (68.1)97 (58.8).062
       No79 (31.9)67 (40.6)
       Missing0 (0.0)1 (0.6)
      ≥1 cup daily
       Yes65 (26.2)28 (17.0).030
       No183 (73.8)136 (82.4)
       Missing0 (0.0)1 (0.6)
       Decaffeinated tea in non–coffee drinkers
      Any use
       Yes14 (5.6)11 (6.7).658
       No234 (94.4)153 (92.7)
       Missing0 (0.0)1 (0.6)
      ≥1 cup daily
       Yes4 (1.6)4 (2.4).553
       No244 (98.4)160 (97.0)
       Missing0 (0.0)1 (0.6)
       Caffeinated soda in non–coffee drinkers
      Any use
       Yes183 (73.8)114 (69.1).298
       No65 (26.2)51 (30.9)
      ≥1 can per day
       Yes94 (37.9)56 (33.9).412
       No154 (62.1)109 (66.1)
       Decaffeinated soda in non-coffee drinkers
      Any use
       Yes65 (26.2)46 (27.9).708
       No183 (73.8)119 (72.1)
      ≥1 can per day
       Yes26 (10.5)17 (10.3).953
       No222 (89.5)148 (89.7)
      Cream/sugar intake, n (%)568342
       Creamer or whitener use
      Any use214 (37.7)123 (35.9).595
      No use
      No use includes noncoffee/tea/soda drinkers.
      350 (61.6)217 (63.5)
      Missing4 (0.7)2 (0.6)
       Artificial sweetener use
      Any use125 (22.0)92 (26.9).173
      No use
      No use includes noncoffee/tea/soda drinkers.
      430 (75.7)246 (71.9)
      Missing13 (2.3)4 (1.2)
      a No use includes noncoffee/tea/soda drinkers.

       Tea

      Approximately 70% of participants reported drinking caffeinated tea (hot or iced) within the prior year, with 22.7% consuming on average 1 or more cups per day (Table 2). Overall, caffeinated and decaffeinated tea intake was not notably different between advanced fibrosis cases and mild fibrosis controls. However, among the subset of non–coffee drinkers (n = 413), caffeinated tea use was more common among controls (26.2% vs 17.0%; P = .030) (Table 2).

       Soda

      An average caffeinated soda intake of 1 or more cans daily was reported more frequently in mild fibrosis controls vs advanced fibrosis cases (40.5% vs 32.5%; P = .015); however, among the subset of non–coffee drinkers (n = 413), intake of 1 or more cans daily of caffeinated soda use was not significant (37.9% vs 33.9%; P = .412). On multivariate analysis, more than 1 can of caffeinated soda daily remained significantly associated with a decreased risk of hepatic fibrosis, with some attenuation (adjusted OR, 0.74; 95% CI, 0.55–1.00; P = .047) after adjustment for age, BMI, and alcohol use, with similar findings when also adjusted for MELD score and the presence of metabolic syndrome. When adjusted for self-reported diabetes and HOMA-IR scores, there was further attenuation of the soda–fibrosis association (adjusted OR, 0.75; 95% CI, 0.55–1.02; P = .063) (Table 3). There was no notable difference in decaffeinated soda intake between cases and controls (Table 2).
      Table 3Association of Caffeinated and Decaffeinated Beverage Intake With Advanced Fibrosis HCV Results of Logistic Regression
      Fibrosis (F3/F4–F4 vs F0–F3)
      N
      Discrepancy in the number between unadjusted and adjusted models reflects the loss of individuals with missing required data on an adjustment factor.
      OR95% CIP value
      ≥100 mg of caffeine daily +from all sources
       Unadjusted9100.7080.537–0.933.014
       Adjusted
      Adjustment 1 included age, alcohol use (never, former, or current), and BMI category (<25, 25–29, or ≥30).
      8990.7120.534–0.949.020
       Adjusted
      Adjustment 2 included MELD score range in addition to the variables in adjustment 1 models.
      8980.7280.544–0.975.033
       Adjusted
      Adjustment 3 included self-reported diabetes status and HOMA-IR less than 3 and HOMA-IR of 3 or greater categories in addition to the variables in adjustment 1 models.
      8580.7470.555–1.005.054
       Adjusted
      Adjustment 4 included the presence of the metabolic syndrome in addition to the variables in adjustment 1 models.
      8900.6940.519–0.930.014
      ≥100 mg of caffeine daily from coffee only
       Unadjusted9100.7490.571–0.981.036
       Adjusted
      Adjustment 1 included age, alcohol use (never, former, or current), and BMI category (<25, 25–29, or ≥30).
      8990.7670.580–1.014.062
       Adjusted
      Adjustment 2 included MELD score range in addition to the variables in adjustment 1 models.
      8980.7880.594–1.047.100
       Adjusted
      Adjustment 3 included self-reported diabetes status and HOMA-IR less than 3 and HOMA-IR of 3 or greater categories in addition to the variables in adjustment 1 models.
      8580.8310.623–1.108.207
       Adjusted
      Adjustment 4 included the presence of the metabolic syndrome in addition to the variables in adjustment 1 models.
      8900.7470.562–0.993.044
      ≥200 mg of caffeine daily from all sources
       Unadjusted9100.7540.571–0.995.046
       Adjusted
      Adjustment 1 included age, alcohol use (never, former, or current), and BMI category (<25, 25–29, or ≥30).
      8990.7730.580–1.030.079
       Adjusted
      Adjustment 2 included MELD score range in addition to the variables in adjustment 1 models.
      8980.8170.610–1.094.175
       Adjusted
      Adjustment 3 included self-reported diabetes status and HOMA-IR less than 3 and HOMA-IR of 3 or greater categories in addition to the variables in adjustment 1 models.
      8580.8150.606–1.096.175
       Adjusted
      Adjustment 4 included the presence of the metabolic syndrome in addition to the variables in adjustment 1 models.
      8900.7460.557–1.001.051
      ≥1 cup of caffeinated coffee daily
       Unadjusted9100.7590.580–0.994.045
       Adjusted
      Adjustment 1 included age, alcohol use (never, former, or current), and BMI category (<25, 25–29, or ≥30).
      8990.7800.590–1.031.081
       Adjusted
      Adjustment 2 included MELD score range in addition to the variables in adjustment 1 models.
      8980.8000.602–1.062.123
       Adjusted
      Adjustment 3 included self-reported diabetes status and HOMA-IR less than 3 and HOMA-IR of 3 or greater categories in addition to the variables in adjustment 1 models.
      8580.8450.634–1.127.252
       Adjusted
      Adjustment 4 included the presence of the metabolic syndrome in addition to the variables in adjustment 1 models.
      8900.7590.571–1.008.057
      ≥1 cup of caffeinated tea daily in non–coffee drinkers
       Unadjusted4120.5800.353–0.951.031
       Adjusted
      Adjustment 1 included age, alcohol use (never, former, or current), and BMI category (<25, 25–29, or ≥30).
      4030.5610.335–0.939.028
       Adjusted
      Adjustment 2 included MELD score range in addition to the variables in adjustment 1 models.
      4030.5410.320–0.915.022
       Adjusted
      Adjustment 3 included self-reported diabetes status and HOMA-IR less than 3 and HOMA-IR of 3 or greater categories in addition to the variables in adjustment 1 models.
      3830.4830.279–0.835.009
       Adjusted
      Adjustment 4 included the presence of the metabolic syndrome in addition to the variables in adjustment 1 models.
      3950.5440.317–0.933.027
      ≥1 can of caffeinated soda daily
       Unadjusted9100.7060.533–0.936.015
       Adjusted
      Adjustment 1 included age, alcohol use (never, former, or current), and BMI category (<25, 25–29, or ≥30).
      8990.7390.550–0.991.043
       Adjusted
      Adjustment 2 included MELD score range in addition to the variables in adjustment 1 models.
      8980.7370.548–0.992.044
       Adjusted
      Adjustment 3 included self-reported diabetes status and HOMA-IR less than 3 and HOMA-IR of 3 or greater categories in addition to the variables in adjustment 1 models.
      8580.7500.554–1.016.063
       Adjusted
      Adjustment 4 included the presence of the metabolic syndrome in addition to the variables in adjustment 1 models.
      8900.7370.546–0.995.047
      a Discrepancy in the number between unadjusted and adjusted models reflects the loss of individuals with missing required data on an adjustment factor.
      b Adjustment 1 included age, alcohol use (never, former, or current), and BMI category (<25, 25–29, or ≥30).
      c Adjustment 2 included MELD score range in addition to the variables in adjustment 1 models.
      d Adjustment 3 included self-reported diabetes status and HOMA-IR less than 3 and HOMA-IR of 3 or greater categories in addition to the variables in adjustment 1 models.
      e Adjustment 4 included the presence of the metabolic syndrome in addition to the variables in adjustment 1 models.

       Overall caffeine intake

      Although quite variable, daily caffeine consumption from all beverages (coffee, tea, and soda) was, on average, higher in mild fibrosis controls than in advanced fibrosis cases (273.8 vs 218.2 mg; P = .013) (Table 4). The majority of caffeine consumption in both groups came from coffee (67.7% of total, with an average of 1.2 cups consumed per day), followed by soda (19.5%, with an average of 0.8 cans consumed per day), and then tea (12.8%, with an average of 0.6 cups consumed per day; data not shown); however, controls had a higher percentage of their daily caffeine consumption derived from coffee sources. When further stratified by dose, a daily average intake of 100 mg or more and 200 mg or more of caffeine from all sources was associated with a significantly reduced risk of advanced fibrosis (P = .014 and P = .045, respectively). After excluding coffee drinkers, an intake of 100 mg or more and 200 mg or more of caffeine daily from noncoffee sources (tea and soda) was slightly higher in mild fibrosis controls than in advanced fibrosis cases, but the difference was not significant (Table 4).
      Table 4Comparison of Average Caffeine Intake Between Advanced Fibrosis HCV Cases and Mild Fibrosis HCV Controls in Year Preceding Study Enrollment
      Controls (F0–F3)Cases (F3/F4–F4)P value
      Caffeine intake, n (%)568342
       Total caffeine from any beverage per day, mean (SD), mg273.8 (346.7)218.2 (291.1).013
       Total caffeine from coffee per day, mean (SD), mg187.2 (327.7)143.8 (263.0).028
       ≥100 mg of caffeine daily from all sources
      Yes375 (66.0)198 (57.9).014
      No193 (34.0)144 (42.1)
       ≥100 mg of caffeine daily from coffee only
      Yes285 (50.2)147 (43.0).035
      No283 (49.8)195 (57.0)
       ≥100 mg of caffeine daily from noncoffee sources only
      Yes164 (28.9)92 (26.9).522
      No404 (71.1)250 (73.1)
       ≥200 mg of caffeine daily from all sources
      Yes239 (42.1)121 (35.4).045
      No329 (57.9)221 (64.6)
       ≥200 mg of caffeine daily from coffee only
      Yes164 (28.9)85 (24.9).188
      No404 (71.1)257 (75.1)
      Intake in non–coffee drinkers, n (%)248165
       ≥100 mg of caffeine daily from noncoffee sources only in non–coffee drinkers
      Yes70 (28.2)40 (24.2).370
      No178 (71.8)125 (75.8)
       ≥200 mg of caffeine daily from noncoffee sources only in non–coffee drinkers
      Yes25 (10.1)11 (6.7).228
      No223 (89.9)154 (93.3)

       Creamers/whiteners

      Use of creamers/whiteners and added sweeteners was reported in 37.0% and 23.8% of participants, respectively, with no significant differences between cases and controls in the proportions of participants who reported use of these additives (Table 2).

       Lifetime Exposure to Coffee in Advanced Fibrosis Hepatitis C Virus Cases and Mild Fibrosis Hepatitis C Virus Controls

      The average daily intake of coffee in prior life decades was higher in mild fibrosis controls than in advanced fibrosis cases, but none of these differences reached statistical significance. We also did not find a difference between cumulative years or lifetime consumption of coffee drinking between cases and controls (Table 5).
      Table 5Comparison of Average Daily Cups of Caffeinated Coffee Across Decades of Life in Advanced Fibrosis HCV Cases and Mild Fibrosis HCV Controls
      Controls (F0–F3)

      N = 568 (%)
      Cases (F3–F4/F4)

      N = 342 (%)
      P value
      Respondents

      n/N (%)
      Mean (SD)Respondents

      n/N (%)
      Mean (SD)
      Cups per day in previous decades of life, y
      One-way analysis of variance was used for each time period.
       20–29567/907 (62.5)1.52 (2.95)340/907 (37.5)1.45 (3.36).733
       30–39566/905 (62.5)1.82 (3.08)339/905 (37.5)1.66 (2.87).444
       40–49559/901 (62.0)1.93 (2.90)342/901 (38.0)1.72 (3.15).313
       50–59496/822 (60.3)1.69 (2.42)326/822 (39.7)1.48 (2.50).227
       60–69157/287 (54.7)1.30 (1.74)130/287 (45.3)1.27 (2.50).927
      Cumulative use, y24.4 (16.5)24.5 (17.8).945
      Lifetime average daily consumption1.99 (2.70)1.76 (2.76).221
      Tertiles of lifetime coffee consumption
       First, 0–4380 cups182 (32.0)122 (35.7).502
       Second, 4381–23,481 cups191 (33.6)112 (32.7)
       Third, >23,481 cups195 (34.3)108 (31.6)
      a One-way analysis of variance was used for each time period.

       Relationship Between Coffee, Tea, Soda, and Caffeine Ingestion With Homeostasis Model Assessment–Estimated Insulin Resistance in Nondiabetic Subjects

      Among nondiabetic subjects (n = 650), those with a HOMA-IR of less than 3 were significantly more likely to consume more than 100 mg of caffeine daily compared with those with insulin resistance/HOMA-IR of 3 or greater (68.8% vs 59.4%; P = .012). This association between caffeine intake and decreased risk of insulin resistance remained significant after adjustment for age, BMI, and alcohol use (adjusted OR, 0.64; 95% CI, 0.45–0.94; P = .022); MELD score (adjusted OR, 0.66; 95% CI, 0.45–0.95; P = .024); and the presence of the metabolic syndrome (adjusted OR, 0.63; 95% CI, 0.43–0.92; P = .017). However, there were no significant differences in the daily intake of caffeinated coffee, tea, or soda between the 2 groups (data not shown).

       Association Between Coffee, Tea, Soda and Caffeine Intake and Risk of Advanced Hepatic Fibrosis After Controlling for Confounders

      In the first multivariable model adjusting for age, alcohol use, and BMI, an average daily intake of 100 mg or more of caffeine from all sources remained associated with a significantly decreased risk of advanced fibrosis (adjusted OR, 0.71; 95% CI, 0.53–0.95; P = .020). With additional adjustment for MELD scores in the second multivariable model, the inverse associations between a daily intake of 100 mg or more of caffeine and hepatic fibrosis remained statistically significant. Further adjustment for our insulin resistance variable (ie, diagnosed diabetes, HOMA-IR ≥ 3 in nondiabetic participants, HOMA-IR < 3 in nondiabetic participants) in the third model minimally attenuated the associations between coffee/caffeine intake and advanced fibrosis in those who consumed an average of 100 mg or more of caffeine, although still closely approached significance (P = .054). However, the association between caffeine intake and decreased risk of advanced hepatic fibrosis remained significant after adjusting for the metabolic syndrome in the multivariate regression model (adjusted OR, 0.69; 95% CI, 0.52–0.93; P = .014) (Table 3).
      In the subset of non–coffee drinkers, caffeinated tea intake of 1 or more cups per day became more significant after adjustment for age, BMI, and alcohol use (adjusted OR, 0.56; 95% CI, 0.36–0.94; P = .028) and remained significant with the additional adjustments of MELD score (adjusted OR, 0.54; 95% CI, 0.32–0.92; P = .022), insulin resistance status (adjusted OR, 0.48; 95% CI, 0.28–0.84; P = .009), or the presence of the metabolic syndrome (adjusted OR, 0.54; 95% CI, 0.32–0.93; P = .027) (Table 3).

       Relationship Between Coffee, Tea, and Soda Consumption and Hepatic Inflammation

      Based on FibroSURE scoring, 252 participants (27.7%) were classified as advanced inflammatory activity cases (A2/A3–A3), and 658 (72.3%) were classified as mild inflammatory activity controls (A0–A2). There was no significant association between caffeine intake from coffee, tea, or soda in the year preceding study enrollment and severity of hepatic inflammation (Supplementary Table 1, Supplementary Table 2, Supplementary Table 3) There were also no significant differences between the 2 groups in terms of average daily coffee intake in prior life decades, cumulative years of coffee drinking, or lifetime coffee consumption (data not shown).

      Discussion

      We found that self-reported coffee drinking and caffeine consumption from beverages were associated with a lower risk of advanced hepatic fibrosis in patients with chronic HCV but had no association with degree of hepatic inflammation. An average daily intake of an estimated 100 mg of caffeine from coffee, tea, or soda was associated with an approximately one-third reduction in odds of advanced fibrosis, although higher intake did not seem to confer any additional benefit. Interestingly, tea intake in those who did not consume coffee also was found to be associated with a decreased risk of advanced fibrosis. The inverse association between 100 mg or more of caffeine intake daily and advanced hepatic fibrosis remained significant after adjustment for potential confounders (eg, age, alcohol use, BMI, MELD score, and metabolic syndrome), and was attenuated when adjusted for HOMA-IR status. The inverse association between caffeine consumption and degree of fibrosis was found using caffeine intake data from the year preceding study recruitment, however, lifetime caffeine intake did not seem to influence this relationship.
      Our finding that caffeine intake is associated with a decreased risk of advanced HCV-related fibrosis is in line with several other studies that have suggested a beneficial effect of caffeine on hepatic function in various liver diseases, including nonalcoholic steatohepatitis, alcoholic liver disease, and primary sclerosing cholangitis.
      • Klatsky A.L.
      • Armstrong M.A.
      Alcohol, smoking, coffee, and cirrhosis.
      • Tanaka K.
      • Tokunaga S.
      • Kono S.
      • et al.
      Coffee consumption and decreased serum γ-glutamyltransferase and aminotransferase activities among male alcohol drinkers.
      • Ruhl C.E.
      • Everhart J.E.
      Coffee and caffeine consumption reduce the risk of elevated serum alanine aminotransferase activity in the United States.
      • Birerdinc A.
      • Stepanova M.
      • Pawloski L.
      • et al.
      Caffeine is protective in patients with non-alcoholic fatty liver disease.
      • Chen S.
      • Teoh N.C.
      • Chitturi S.
      • et al.
      coffee and non-alcoholic fatty liver disease: brewing evidence for hepatoprotection?.
      • Andersen I.M.
      • Tengesdal G.
      • Lie B.A.
      • et al.
      Effects of coffee consumption, smoking, and hormones on risk for primary sclerosing cholangitis.
      • Lammert C.
      • Juran B.D.
      • Schlicht E.
      • et al.
      Reduced coffee consumption among individuals with primary sclerosing cholangitis but not primary biliary cirrhosis.
      • Corrao G.
      • Lepore A.R.
      • Torchio P.
      • et al.
      The effect of drinking coffee and smoking cigarettes on the risk of cirrhosis associated with alcohol consumption. A case-control study.
      • Klatsky A.L.
      • Morton C.
      • Udaltsova
      • et al.
      Coffee, cirrhosis, and transaminase enzymes.
      It also is consistent with an analysis of the Hepatitis C Antiviral Long-Term Treatment Against Cirrhosis (HALT-C) trial, which found that daily coffee consumption was associated with slower disease progression in patients with chronic HCV-related bridging fibrosis and cirrhosis who had failed to respond to treatment with pegylated interferon and ribavirin. However, in contrast to our findings, which suggest no substantial additional benefit of caffeine intake of 100 mg/day or more, they found a dose-dependent response with coffee drinkers who reported 3 or more cups daily experiencing the greatest (∼53%) reduced risk of disease progression (ie, to decompensation or HCC) compared with non–coffee drinkers.
      • Freedman N.D.
      • Everhart J.E.
      • Lindsay K.L.
      • et al.
      Coffee intake is associated with lower rates of liver disease progression in chronic hepatitis C.
      We hypothesize that the primary explanation for this discrepancy is the difference in the spectrum of HCV-related liver disease among studies because the HALT-C trial was restricted to cirrhotic patients, whereas our study population included the full spectrum of HCV-related liver disease. However, other differences in study design and among populations also may have contributed to this discrepancy. Our results suggesting that a modest dose of caffeine intake (∼100 mg/d) was associated with a significantly reduced risk of advanced fibrosis in veterans with chronic HCV are similar to an earlier comparable pilot study in 91 HCV-positive veterans.
      • White D.L.
      • Richardson P.A.
      • Al-Saadi M.
      • et al.
      Dietary history and physical activity and risk of advanced liver disease in veterans with chronic hepatitis C infection.
      The consistency in these results over time further supports the internal validity of our findings.
      In contrast to our results, Costentin et al
      • Costentin C.E.
      • Roudot-Thoraval F.
      • Zafrani E.S.
      • et al.
      Association of caffeine intake and histological features of chronic hepatitis C.
      found no significant protective association in their study of 238 French patients with treatment-naive chronic HCV. However, they did find that daily caffeine consumption of more than 408 mg (equivalent to ≥3 cups of coffee) was an independent predictor of lower histologic activity grade of hepatitis. Several factors may explain this discrepancy, including differences in populations such as higher average caffeine intake but lower rates of obesity and alcohol abuse overall in the French study population, differences in methods of estimating caffeine intake, and known substantial variability in caffeine found in coffee and tea based on methods of production and preparation.
      Our findings suggest that a total combined caffeine intake or dose of 100 mg or more daily is associated with a significant decrease in risk of hepatic fibrosis, but without further risk reduction with higher coffee or caffeine doses. However, the optimal hepatoprotective dose of caffeine in HCV-infected patients overall is unclear. Although some studies suggested the strongest benefit from coffee consumption was at 3 or more cups daily, or approximately 400 mg of caffeine,
      • Freedman N.D.
      • Everhart J.E.
      • Lindsay K.L.
      • et al.
      Coffee intake is associated with lower rates of liver disease progression in chronic hepatitis C.
      • Costentin C.E.
      • Roudot-Thoraval F.
      • Zafrani E.S.
      • et al.
      Association of caffeine intake and histological features of chronic hepatitis C.
      other studies reported a significant reduction in the risk in HCV-related hepatic fibrosis associated with an intake of 1 or 2 cups of cups of coffee daily.
      • Modi A.A.
      • Feld J.J.
      • Park Y.
      • et al.
      Increased caffeine consumption is associated with reduced hepatic fibrosis.
      • Machado S.R.
      • Parise E.R.
      • de Carvalho L.
      Coffee has hepatoprotective benefits in Brazilian patients with chronic hepatitis C even in lower daily consumption than in American and European populations.
      Our results suggest that as little as 100 mg or more of caffeine daily may be beneficial in a general HCV-infected population with a high prevalence of other risk factors for advanced liver disease. However, our power to evaluate the benefit of higher intake likely was limited by the small number of our study participants who reported drinking higher quantities of caffeinated beverages. Nonetheless, if validated in other HCV-infected populations in the United States, our results suggest that a relatively low (and therefore potentially more tolerable dose) of caffeine, particularly from caffeinated coffee and possibly from tea, may convey a substantial reduction in fibrosis progression.
      Several mechanisms have been suggested by which caffeine may act to reduce the risk of advanced fibrosis.
      • Kumari R.
      • Kim W.R.
      Coffee: a panacea or snake oil for the liver?.
      Caffeine functions as a nonselective adenosine-receptor antagonist and a phosphodiesterase inhibitor with a broad range of attendant effects including inhibition of growth factors that contribute to hepatic fibrosis.
      • Peng Z.
      • Fernandez P.
      • Wilder T.
      • et al.
      Ecto-5′-nucleotidase (CD73)-mediated extracellular adenosine production plays a critical role in hepatic fibrosis.
      • Arauz J.
      • Moreno M.G.
      • Cortes-Reynosa P.
      • et al.
      Coffee attenuates fibrosis by decreasing the expression of TGF-β and CTGF in a murine model of liver damage.
      • Gressner O.A.
      • Lahme B.
      • Rehbein K.
      • et al.
      Pharmacological application of caffeine inhibits TGF-beta-stimulated connective tissue growth factor expression in hepatocytes via PPARgamma and SMAD2/3-dependent pathways.
      Caffeine is metabolized in the liver primarily by CYP1A2, with reduced expression of CYP1A2 correlated with fibrosis progression in a study of HCV patients.
      • Nakai K.
      • Tanaka H.
      • Hanada K.
      • et al.
      Decreased expression of cytochromes P450 1A2, 2E1, and 3A4 and drug transporters Na+-taurocholate-cotransporting polypeptide, organic cation transporter 1, and organic anion-transporting peptide-C correlates with the progression of liver fibrosis in chronic hepatitis C patients.
      Caffeine also has been shown in animal and in vitro studies to act as an antioxidant.
      • Devasagayam T.P.
      • Kamat J.P.
      • Mohan H.
      • et al.
      Caffeine as an antioxidant: inhibition of lipid peroxidation induced by reactive oxygen species.
      Other studies have suggested that the primary hepatoprotective benefit of caffeine is associated with caffeinated coffee intake.
      • Klatsky A.L.
      • Armstrong M.A.
      Alcohol, smoking, coffee, and cirrhosis.
      • Tanaka K.
      • Tokunaga S.
      • Kono S.
      • et al.
      Coffee consumption and decreased serum γ-glutamyltransferase and aminotransferase activities among male alcohol drinkers.
      • Modi A.A.
      • Feld J.J.
      • Park Y.
      • et al.
      Increased caffeine consumption is associated with reduced hepatic fibrosis.
      • Freedman N.D.
      • Everhart J.E.
      • Lindsay K.L.
      • et al.
      Coffee intake is associated with lower rates of liver disease progression in chronic hepatitis C.
      • Gallus S.
      • Tavani A.
      • Negri E.
      • et al.
      Does coffee protect against liver cirrhosis?.
      However, this could be the effect of generally higher daily cumulative caffeine intake in coffee drinkers compared with consumers of other caffeinated beverages overall, or with pharmacodynamic differences in effects given the substantial differences in caffeine per unit intake of caffeinated coffee in comparison with tea or soda. Finally, it also has been postulated that other constituents of coffee, such as diterpenes, polyphenols, kahweol, and cafestrerol, which have anti-oxidant activity, also may be responsible for the particular hepatoprotective effect of coffee.
      • Scalbert A.
      • Williamson G.
      Dietary intake and bioavailability of polyphenols.
      • Cavin C.
      • Holzhaeuser D.
      • Scharf G.
      • et al.
      Cafestol and kahweol, two coffee specific diterpenes with anticarcinogenic activity.
      • Huber W.W.
      • Scharf G.
      • Rossmanith W.
      • et al.
      The coffee components kahweol and cafestol induce gamma-glutamylcysteine synthetase, the rate limiting enzyme of chemoprotective glutathione synthesis, in several organs of the rat.
      • Lee C.
      Antioxidant ability of caffeine and its metabolites based on the study of oxygen radical absorbing capacity and inhibition of LDL peroxidation.
      However, studies, including ours, have not shown that decaffeinated coffee exerts the same effect on liver disease as caffeinated coffee.
      • Modi A.A.
      • Feld J.J.
      • Park Y.
      • et al.
      Increased caffeine consumption is associated with reduced hepatic fibrosis.
      • Gallus S.
      • Tavani A.
      • Negri E.
      • et al.
      Does coffee protect against liver cirrhosis?.
      Inflammation is another known risk factor for progression to fibrosis in chronic HCV,
      • Leandro G.
      • Mangia A.
      • Hui J.
      • et al.
      Relationship between steatosis, inflammation, and fibrosis in chronic hepatitis C: a meta-analysis of individual patient data.
      and the role of coffee as an anti-inflammatory agent in nonliver diseases has been suggested.
      • Andersen L.F.
      • Jacobs Jr., D.R.
      • Carlsen M.H.
      • et al.
      Consumption of coffee is associated with reduced risk of death attributed to inflammatory and cardiovascular diseases in the Iowa Women's Health Study.
      In our study, we found no significant association between caffeine intake and the severity of hepatic inflammation, in line with findings by Freedman et al
      • Freedman N.D.
      • Everhart J.E.
      • Lindsay K.L.
      • et al.
      Coffee intake is associated with lower rates of liver disease progression in chronic hepatitis C.
      that coffee intake was not associated with lower baseline hepatic inflammation. Prior work has shown that insulin resistance worsens hepatic inflammation in HCV patients,
      • Negro F.
      Insulin resistance and HCV: will new knowledge modify clinical management?.
      and a meta-analysis suggested that part of the hepatoprotective effect of coffee in part may be related to an associated reduction in insulin resistance and type 2 diabetes mellitus.
      • Van Dam R.M.
      • Hu F.B.
      Coffee consumption and risk of type 2 diabetes a systematic review.
      In our study, caffeine intake was associated inversely with insulin resistance via the surrogate marker of the HOMA-IR score in nondiabetic participants. Therefore, one potential mechanism linking caffeine intake or the associated lifestyle with decreased risk of advanced fibrosis is the amelioration of insulin resistance. Further studies are needed to better elucidate the underlying pathophysiological mechanisms by which caffeine or its metabolites and other constituents of coffee account for the positive health benefits seen particularly with caffeinated coffee consumption in liver disease patients.
      We found that an average of 100 mg or more of caffeine daily from sodas and teas does not have the same protective effect as 100 mg or more of caffeine daily from combined sources (coffee, tea, soda) or from coffee alone, suggesting that caffeine alone may not entirely explain the effect of coffee on liver disease. Although several prior studies have not shown a lower risk of liver disease progression in those who consume caffeine from noncoffee sources such as tea
      • Klatsky A.L.
      • Armstrong M.A.
      Alcohol, smoking, coffee, and cirrhosis.
      • Modi A.A.
      • Feld J.J.
      • Park Y.
      • et al.
      Increased caffeine consumption is associated with reduced hepatic fibrosis.
      • Freedman N.D.
      • Everhart J.E.
      • Lindsay K.L.
      • et al.
      Coffee intake is associated with lower rates of liver disease progression in chronic hepatitis C.
      • Gallus S.
      • Tavani A.
      • Negri E.
      • et al.
      Does coffee protect against liver cirrhosis?.
      or soda,
      • Gallus S.
      • Tavani A.
      • Negri E.
      • et al.
      Does coffee protect against liver cirrhosis?.
      this could be related to differing caffeine content, other ingredients in addition to caffeine that may be partly responsible for the proposed hepatoprotective effects of coffee, or to substantial differences in study design or populations. Among the 413 noncoffee drinkers in our current study, intake of at least 1 cup of caffeinated tea daily was more common in the mild fibrosis control group, and after controlling for potential confounders, this significance was even more pronounced. This finding provides further evidence for the protective role of caffeine from any source against the progression of liver disease.
      In regards to soda, an understudied source of caffeine, we found consumption of more than 1 can of caffeinated soda daily to be associated with a decreased risk of hepatic fibrosis, supporting that caffeine likely plays a major role in coffee’s hepatoprotective effects. On multivariate nonstratified analysis, more than 1 can of caffeinated soda daily remained significantly associated with a decreased risk of hepatic fibrosis, with some attenuation after controlling for possible confounders (age, alcohol, BMI, MELD score, and the presence of the metabolic syndrome). When adjusted for self-reported diabetes and HOMA-IR scores, there was a nonsignificant trend toward a decreased risk of fibrosis (adjusted OR, 0.75; 95% CI, 0.55–1.02; P = .063). However, in analyzing soda intake in the subgroup of non–coffee drinkers (n = 413), there was no notable difference in the risk of hepatic fibrosis (P = .41). The findings may indicate that caffeine from soda has hepatoprotective effects, but we cannot exclude the possibility that increased soda consumption reflects an increased intake of caffeinated beverages overall, including coffee. As with coffee, decaffeinated tea and soda were not found to be associated with a decreased risk of liver disease.
      Our study had several strengths, including being a large study that examined the association between caffeinated coffee intake and HCV-related liver disease, and adds to the limited research in US populations. It also adds to the sparse literature on the effects of caffeinated tea and soda intake. In addition, we obtained and analyzed information on the consumption of decaffeinated coffee, tea, and soda to better elucidate the mechanisms underlying the potential hepatoprotective effects of caffeine and coffee on liver disease. Finally, our novel use of a validated noninvasive marker for liver fibrosis, the validated FibroSURE test,
      • Halfon P.
      • Bourliere M.
      • Deydier R.
      • et al.
      Independent prospective multicenter validation of biochemical markers (fibrotest-actitest) for the prediction of liver fibrosis and activity in patients with chronic hepatitis C: the fibropaca study.
      • Ngo Y.
      • Munteanu M.
      • Messous D.
      • et al.
      A prospective analysis of the prognostic value of biomarkers (FibroTest) in patients with chronic hepatitis C.
      • Morali G.
      • Maor Y.
      • Klar R.
      • et al.
      Fibrotest-Actitest: the biochemical marker of liver fibrosis–the Israeli experience.
      • Imbert-Bismut F.
      • Ratziu V.
      • Pieroni L.
      • et al.
      Biochemical markers of liver fibrosis in patients with hepatitis C virus infection: a prospective study.
      • Grigorescu M.
      • Rusu M.
      • Neculoiu D.
      • et al.
      The FibroTest value in discriminating between insignificant and significant fibrosis in chronic hepatitis C patients. The Romanian experience.
      • Poynard T.
      • Imbert-Bismut F.
      • Ratziu V.
      • et al.
      Biochemical markers of liver fibrosis in patients infected by hepatitis C virus: longitudinal validation in a randomized trial.
      • Poynard T.
      • Munteanu M.
      • Ngo Y.
      • et al.
      ActiTest accuracy for the assessment of histological activity grades in patients with chronic hepatitis C, an overview using Obuchowski measure.
      allowed evaluation of a large and diverse HCV patient group, many of whom would have been precluded from biopsy based on highly prevalent comorbid conditions in this population.
      Our study also had some limitations. Given the observational retrospective nature of the study, a causal association cannot be made. We studied a select patient population of HCV-monoinfected, largely male veterans, and therefore the findings may not be applicable to women with HCV or co-infected patients. Beverage consumption was self-reported, and caffeine intake was estimated based on the average reported intake, and therefore may not accurately reflect the actual intake given the known variability in caffeine levels in coffee and tea of the same type owing to differences in production and preparation. However, any misclassification should be nondifferential (ie, similar among cases and controls) and therefore any possible bias should be toward the null or finding of no association. A potential for reverse causality may be present if patients with more advanced liver disease are averse to drinking caffeine-containing beverages including coffee; however, we did not enroll patients with decompensated liver disease. We also found that the decreased risk of advanced fibrosis with an average daily intake of 100 mg or more of caffeine or 1 or more cups of tea in non–coffee drinkers remained significant or nearly significant after adjusting for MELD scores and insulin resistance. In total, 40% of our participants were classified as marijuana users, and we found no significant association with the degree of hepatic fibrosis or inflammation and marijuana use. However, we did not have information on the duration or frequency of marijuana use, a limitation of our study.
      • Hézode C.
      • Roudot-Thoraval F.
      • Nguyen S.
      • et al.
      Daily cannabis smoking as a risk factor for progression of fibrosis in chronic hepatitis C.
      Prior studies also have suggested women who consume alcohol may have a less pronounced response to the hepatoprotective effects of caffeine as measured by γ-glutamyltransferase levels,
      • Danielsson J.
      • Kangastupa P.
      • Laatikainen T.
      • et al.
      Dose- and gender-dependent interactions between coffee consumption and serum GGT activity in alcohol consumers.
      and in animal models, females may recover more slowly from hepatic injury than males.
      • Wang Y.
      • Ye F.
      • Ke Q.
      • et al.
      Gender-dependent histone deacetylases injury may contribute to differences in liver recovery rates of male and female mice.
      Our study participants were largely male (n = 888), with only 22 females represented. As such, our study lacked the statistical power to find a notable gender-specific difference in the hepatoprotective effects of caffeine or coffee.
      In conclusion, we found that an average daily intake of 100 mg or more of caffeine was associated with a lower risk of hepatic fibrosis in a general clinical population with chronic HCV infection. We further showed that in non–coffee drinkers, daily caffeinated tea intake also may protect against progressive liver disease in this population. Our study further suggests that caffeinated coffee overall and caffeinated tea in non–coffee drinkers likely provide the most benefit in liver disease compared with other caffeinated beverages or decaffeinated coffee. Future prospective studies are warranted to determine the optimal dose and preparation of caffeinated beverage intake that could be safely and tolerably recommended for prevention of progressive liver disease in HCV patients in routine clinical practice.

      Supplementary Material

      Supplementary Table 1Demographic Characteristics of Advanced Inflammatory Activity HCV Cases Compared With Mild Inflammatory Activity HCV Controls
      CharacteristicsInflammatory activity
      Controls (A0–A2)

      N = 658 (%)
      Cases (A2/A3–A3)

      N = 252 (%)
      P value
      Mean age, y (SD)53.64 (5.42)56.43 (5.36).604
      Sex
       Male641 (97.4)247 (98.0).598
       Female17 (2.6)5 (2.0)
      Race
       White253 (38.4)128 (50.8)<.001
       Black384 (58.4)109 (43.2)
       Other21 (3.2)15 (6.0)
      HCV genotype
       1a388 (59.0)138 (54.8).237
       1b146 (22.2)54 (21.4)
       1 NOS26 (3.9)11 (4.4)
       252 (7.9)24 (9.5)
       327 (4.1)20 (7.9)
       Other5 (0.8)1 (0.4)
       Missing14 (2.1)4 (1.6)
      MELD score
       Mean (SD)8.23 (2.73)8.11 (1.93).474
       <9540 (82.0)213 (84.5).334
       10–19107 (16.3)38 (15.1)
       ≥2010 (1.5)1 (0.4)
       Missing1 (0.2)0 (0.0)
      Prior HCV treatment
       Yes68 (10.3)29 (11.5).633
       No586 (89.1)223 (88.5)
       Missing4 (0.6)0 (0.0)
      Years of alcohol abuse, mean (SD)14.31 (14.60)14.68 (14.29).731
      Chronic alcohol abuse
      Ten or more consecutive years of drinking an average of 3 or more drinks per day for men and 2 or more drinks per day for women.
       Yes352 (53.5)141 (56.0).482
       No305 (46.3)110 (43.6)
       Missing1 (0.2)1 (0.4)
      Marijuana use
      Self-reported use or documented ever use of marijuana in the electronic medical record (International Classification of Diseases, 9th revision, code for cannabis or marijuana abuse; or documented former or current marijuana use in the electronic medical record when searched for the following key words: marijuana and cannabis).
       Yes260 (39.5)102 (40.5).756
       No369 (56.1)138 (54.8)
       Missing29 (4.4)12 (4.7)
      IVD use
       Yes386 (58.7)145 (57.5).758
       No272 (41.3)107 (42.5)
      BMI categories
       <25194 (29.5)62 (24.6).011
       Overweight (25–29)267 (40.6)89 (35.3)
       Obese (≥30)190 (28.9)99 (39.3)
       Missing7 (1.0)2 (0.8)
      DM/HOMA-IR status
       Diabetic
      Self-reported physician’s diagnosis, medical record documentation, or the presence of a fasting blood sugar level of greater than 126 mg/dL, or a nonfasting blood sugar level of greater than 200 mg/dL at enrollment.
      148 (22.5)69 (27.4)<.001
       HOMA-IR ≥3 in nondiabetic participants298 (45.3)74 (29.4)
       HOMA-IR <3 in nondiabetic participants183 (27.8)95 (37.7)
       Missing29 (4.4)14 (5.5)
      Metabolic syndrome
      Metabolic syndrome was defined as the presence of 3 or more of the following: (1) waist circumference greater than 40 inches in men or greater than 35 inches in women; (2) increased serum triglyceride level of 150 mg/dL or greater; (3) reduced high-density lipoprotein cholesterol level less than 40 mg/dL in men or less than 50 mg/dL in women; (4) increased blood pressure of 130/85 mm Hg or greater and/or use of antihypertensive medications; and (5) increased fasting serum glucose level of 110 mg/dL or greater and/or use of insulin or hypoglycemic medication.
       Yes178 (27.1)98 (38.9)<.001
       No473 (71.9)151 (59.9)
       Missing7 (1.0)3 (1.2)
      Smoking
       Nonsmoker221 (33.6)80 (31.8).489
       <30 pack-years213 (32.4)92 (36.5)
       ≥30 pack-years222 (33.7)79 (31.3)
       Missing2 (0.3)1 (0.4)
      Physical activity
      Based on participant response to the International Physical Activity Questionnaire.
       Low activity212 (32.2)76 (30.1).545
       Moderate activity198 (30.1)71 (28.2)
       High activity248 (37.7)105 (41.7)
      DM, diabetes mellitus; IVD, intravenous drug use; NOS, not otherwise specified.
      a Ten or more consecutive years of drinking an average of 3 or more drinks per day for men and 2 or more drinks per day for women.
      b Self-reported use or documented ever use of marijuana in the electronic medical record (International Classification of Diseases, 9th revision, code for cannabis or marijuana abuse; or documented former or current marijuana use in the electronic medical record when searched for the following key words: marijuana and cannabis).
      c Self-reported physician’s diagnosis, medical record documentation, or the presence of a fasting blood sugar level of greater than 126 mg/dL, or a nonfasting blood sugar level of greater than 200 mg/dL at enrollment.
      d Metabolic syndrome was defined as the presence of 3 or more of the following: (1) waist circumference greater than 40 inches in men or greater than 35 inches in women; (2) increased serum triglyceride level of 150 mg/dL or greater; (3) reduced high-density lipoprotein cholesterol level less than 40 mg/dL in men or less than 50 mg/dL in women; (4) increased blood pressure of 130/85 mm Hg or greater and/or use of antihypertensive medications; and (5) increased fasting serum glucose level of 110 mg/dL or greater and/or use of insulin or hypoglycemic medication.
      e Based on participant response to the International Physical Activity Questionnaire.
      Supplementary Table 2Comparison of Beverage Intake Between Advanced Inflammatory Activity HCV Cases and Mild Inflammatory Activity HCV Controls in Year Preceding Study Enrollment
      Inflammatory activity
      Controls (A0–A2)Cases (A2/A3–A3)P value
      Coffee intake, n (%)658252
       Caffeinated coffee
      Cups per day, mean (SD)1.27 (2.36)1.19 (1.85).609
      Any use
       Yes347 (52.7)150 (59.5).066
       No311 (47.3)102 (40.5)
      ≥1 cup daily
       Yes301 (45.7)129 (51.2).141
       No357 (54.3)123 (48.8)
      ≥2 cups daily
       Yes182 (27.7)66 (26.4).656
       No476 (72.3)186 (73.8)
      ≥3 cups daily
       Yes100 (15.2)33 (13.1).422
       No558 (84.8)219 (86.9)
       Decaffeinated coffee
      Cups per day, mean (SD)0.08 (0.41)0.07 (0.33).629
      Any use
       Yes52 (7.9)16 (6.3).425
       No606 (92.1)236 (93.7)
      ≥1 cup daily
       Yes29 (4.4)11 (4.4).978
       No629 (95.6)241 (95.6)
      Tea intake
       Caffeinated tea
      Cups per day, mean (SD)0.62 (1.37)0.52 (1.05).256
      Any use
       Yes460 (69.9)176 (69.8).959
       No197 (30.0)76 (30.2)
       Missing1 (0.1)0 (0.0)
      ≥1 cup daily
       Yes151 (22.9)56 (22.2).807
       No506 (77.0)196 (77.8)
       Missing1 (0.1)0 (0.0)
       Decaffeinated tea
      Cups per day, mean (SD)0.03 (0.28)0.03 (0.22).819
      Any use
       Yes27 (4.1)12 (4.8).664
       No630 (95.8)240 (95.2)
       Missing1 (0.1)0 (0.0)
      ≥1 cup per day
       Yes7 (1.0)4 (1.6).509
       No650 (98.9)248 (98.4)
       Missing1 (0.1)0 (0.0)
      Soda intake
       Caffeinated soda
      Cans per day, mean (SD)1.08 (1.69)1.07 (1.88).939
      Any use
       Yes504 (76.6)195 (77.4).802
       No154 (23.4)57 (22.6)
      ≥1 can per day
       Yes254 (38.6)87 (34.5).255
       No404 (61.4)165 (65.5)
       Decaffeinated soda
      Cans per day, mean (SD)0.19 (0.64)0.14 (0.60).349
      Any use
       Yes172 (26.1)54 (21.4).141
       No486 (73.9)198 (78.6)
      ≥1 can per day
       Yes68 (10.3)15 (6.0).040
       No590 (89.7)237 (94.0)
      Intake in non–coffee drinkers, n (%)311102
       Caffeinated tea in non–coffee drinkers
      Any use
       Yes207 (66.8)59 (57.8).102
       No103 (33.1)43 (42.2)
       Missing1 (0.1)0 (0.0)
      ≥1 cup daily
       Yes73 (23.4)20 (19.6).409
       No237 (76.5)82 (80.4)
       Missing1 (0.1)0 (.0)
       Decaffeinated tea in non–coffee drinkers
      Any use
       Yes21 (6.8)4 (3.9).295
       No289 (93.1)98 (96.1)
       Missing1 (0.1)0 (0.0)
      ≥1 cup daily
       Yes6 (1.9)2 (2.0).987
       No304 (98.0)100 (98.0)
       Missing1 (0.1)0 (0.0)
       Caffeinated soda in non–coffee drinkers
      Any use
       Yes222 (71.4)75 (73.5).675
       No89 (28.6)27 (26.5)
      ≥1 can per day
       Yes115 (37.0)35 (34.3).627
       No196 (63.0)67 (65.7)
       Decaffeinated soda in non–coffee drinkers
      Any use
       Yes87 (28.0)24 (23.5).380
       No224 (72.0)78 (76.5)
      ≥1 can per day
       Yes36 (11.6)7 (6.9).176
       No275 (88.4)95 (93.1)
      Cream/sugar intake, n (%)658252
       Creamer or whitener use
      Any use235 (35.7)102 (40.5).195
      No use
      No use includes non–coffee/tea/soda drinkers.
      418 (63.5)149 (59.1)
      Missing5 (0.8)1 (0.4)
       Artificial sweetener use
      Any use166 (25.2)59 (23.4).493
      No use
      No use includes non–coffee/tea/soda drinkers.
      477 (72.5)191 (75.8)
      Missing15 (2.3)2 (0.8)
      a No use includes non–coffee/tea/soda drinkers.
      Supplementary Table 3Comparison of Average Caffeine Intake Between Advanced Inflammatory Activity HCV Cases and Mild Inflammatory Activity HCV Controls in Year Preceding Study Enrollment
      Inflammatory activity
      Controls (A0–A2)Cases (A2/A3–A3)P value
      Caffeine intake, n (%)658252
       Total caffeine from any beverage per day, mean (SD), mg257.6 (341.6)240.6 (289.1).483
       Total caffeine from coffee per day, mean (SD), mg174.1 (323.5)162.5 (253.2).609
       ≥100 mg of caffeine daily from all sources
      Yes411 (62.5)160 (63.5).774
      No247 (37.5)92 (36.5)
       ≥100 mg of caffeine daily from coffee only
      Yes302 (45.9)130 (51.6).124
      No356 (54.1)122 (48.4)
       ≥100 mg of caffeine daily from noncoffee sources only
      Yes194 (29.5)62 (24.6).143
      No464 (70.5)190 (75.4)
       ≥200 mg of caffeine daily from all sources
      Yes263 (40.0)97 (38.5).683
      No395 (60.0)155 (61.5)
       ≥200 mg of caffeine daily from coffee only
      Yes183 (27.8)66 (26.2).624
      No475 (72.2)186 (73.8)
      Intake in non–coffee drinkers, n (%)311102
       ≥100 mg of caffeine daily from noncoffee sources only in non–coffee drinkers
      Yes87 (28.0)23 (22.5).282
      No224 (72.0)79 (77.5)
       ≥200 mg of caffeine daily from noncoffee sources only in non–coffee drinkers
      Yes29 (9.3)7 (6.9).444
      No282 (90.7)95 (93.1)

      References

        • Freedman N.D.
        • Park Y.
        • Abnet C.C.
        • et al.
        Association of coffee drinking with total and cause-specific mortality.
        N Engl J Med. 2012; 366: 1891-1904
        • Andersen L.F.
        • Jacobs Jr., D.R.
        • Carlsen M.H.
        • et al.
        Consumption of coffee is associated with reduced risk of death attributed to inflammatory and cardiovascular diseases in the Iowa Women's Health Study.
        Am J Clin Nutr. 2006; 83: 1039-1046
        • Matsuura H.
        • Mure K.
        • Nishio N.
        • et al.
        Relationship between coffee consumption and prevalence of metabolic syndrome among Japanese civil servants.
        J Epidemiol. 2012; 22: 160-166
        • Klatsky A.L.
        • Armstrong M.A.
        Alcohol, smoking, coffee, and cirrhosis.
        Am J Epidemiol. 1992; 136: 1248-1257
        • Casiglia E.
        • Spolaore P.
        • Ginocchio G.
        • et al.
        Unexpected effects of coffee consumption on liver enzymes.
        Eur J Epidemiol. 1993; 9: 293-297
        • Tanaka K.
        • Tokunaga S.
        • Kono S.
        • et al.
        Coffee consumption and decreased serum γ-glutamyltransferase and aminotransferase activities among male alcohol drinkers.
        Int J Epidemiol. 1998; 27: 438-443
        • Nakanishi N.
        • Nakamura K.
        • Suzuki K.
        • et al.
        Effects of coffee consumption against the development of liver dysfunction a 4-year follow-up study of middle-aged Japanese male office workers.
        Ind Health. 2000; 38: 99-102
        • Honjo S.
        • Kono S.
        • Coleman M.P.
        • et al.
        Coffee consumption and serum aminotransferases in middle-aged Japanese men.
        J Clin Epidemiol. 2001; 54: 823-829
        • Tverdal A.
        • Skurtveit S.
        Coffee intake and mortality from liver cirrhosis.
        Ann Epidemiol. 2003; 13: 419-423
        • Ruhl C.E.
        • Everhart J.E.
        Coffee and caffeine consumption reduce the risk of elevated serum alanine aminotransferase activity in the United States.
        Gastroenterology. 2005; 128: 24-32
        • Modi A.A.
        • Feld J.J.
        • Park Y.
        • et al.
        Increased caffeine consumption is associated with reduced hepatic fibrosis.
        Hepatology. 2010; 51: 201-209
        • Ohfuji S.
        • Fukushima W.
        • Tanaka T.
        • et al.
        Coffee consumption and reduced risk of hepatocellular carcinoma among patients with chronic type C liver disease: a case-control study.
        Hepatol Res. 2006; 36: 201-208
        • Bravi F.
        • Bosetti C.
        • Tavani A.
        • et al.
        Coffee drinking and hepatocellular carcinoma risk: a meta-analysis.
        Hepatology. 2007; 46: 430-435
        • Larsson S.C.
        • Wolk A.
        Coffee consumption and risk of liver cancer: a meta-analysis.
        Gastroenterology. 2007; 132: 1740-1745
        • Bravi F.
        • Bosetti C.
        • Tavani A.
        • et al.
        Coffee reduces risk for hepatocellular carcinoma: an updated meta-analysis.
        Clin Gastroenterol Hepatol. 2013; 11: 1413-1421
        • Freedman N.D.
        • Everhart J.E.
        • Lindsay K.L.
        • et al.
        Coffee intake is associated with lower rates of liver disease progression in chronic hepatitis C.
        Hepatology. 2009; 50: 1360-1369
        • Costentin C.E.
        • Roudot-Thoraval F.
        • Zafrani E.S.
        • et al.
        Association of caffeine intake and histological features of chronic hepatitis C.
        J Hepatol. 2011; 54: 1123-1129
        • Negro F.
        Insulin resistance and HCV: will new knowledge modify clinical management?.
        J Hepatol. 2006; 45: 514-519
        • Van Dam R.M.
        • Hu F.B.
        Coffee consumption and risk of type 2 diabetes a systematic review.
        JAMA. 2005; 294: 97-104
        • White D.L.
        • Tavakoli-Tabasi S.
        • Kuzniarek J.
        • et al.
        Higher serum testosterone is associated with increased risk of advanced hepatitis C-related liver disease in males.
        Hepatology. 2012; 55: 759-768
        • Kanwal F.
        • White D.L.
        • Tavakoli-Tabasi S.
        • et al.
        Many patients with interleukin 28B genotypes associated with response to therapy are ineligible for treatment because of comorbidities.
        Clin Gastroenterol Hepatol. 2014; 12: 327-333
        • Craig C.L.
        • Marshall A.L.
        • Sjöström M.
        • et al.
        International physical activity questionnaire: 12-country reliability and validity.
        Med Sci Sports Exerc. 2003; 35: 1381-1395
        • National Cholesterol Education Program (NCEP) Expert Panel on Detection, Evaluation, and Treatment of High Blood Cholesterol in Adults (Adult Treatment Panel III)
        Third Report of the National Cholesterol Education Program (NCEP) Expert Panel on Detection, Evaluation, and Treatment of High Blood Cholesterol in Adults (Adult Treatment Panel III) final report.
        Circulation. 2002; 106: 3143-3421
        • Michels K.B.
        • Willett W.C.
        • Fuchs C.S.
        • et al.
        Coffee, tea, and caffeine consumption and incidence of colon and rectal cancer.
        J Natl Cancer Inst. 2005; 97: 282-292
        • Iso H.
        • Date C.
        • Wakai K.
        • et al.
        The relationship between green tea and total caffeine intake and risk for self-reported type 2 diabetes among Japanese adults.
        Ann Intern Med. 2006; 144: 554-562
        • Halfon P.
        • Bourliere M.
        • Deydier R.
        • et al.
        Independent prospective multicenter validation of biochemical markers (fibrotest-actitest) for the prediction of liver fibrosis and activity in patients with chronic hepatitis C: the fibropaca study.
        Am J Gastroenterol. 2006; 101: 547-555
        • Ngo Y.
        • Munteanu M.
        • Messous D.
        • et al.
        A prospective analysis of the prognostic value of biomarkers (FibroTest) in patients with chronic hepatitis C.
        Clin Chem. 2006; 52: 1887-1896
        • Morali G.
        • Maor Y.
        • Klar R.
        • et al.
        Fibrotest-Actitest: the biochemical marker of liver fibrosis–the Israeli experience.
        Isr Med Assoc J. 2007; 9: 588-591
        • Imbert-Bismut F.
        • Ratziu V.
        • Pieroni L.
        • et al.
        Biochemical markers of liver fibrosis in patients with hepatitis C virus infection: a prospective study.
        Lancet. 2001; 357: 1069-1075
        • Grigorescu M.
        • Rusu M.
        • Neculoiu D.
        • et al.
        The FibroTest value in discriminating between insignificant and significant fibrosis in chronic hepatitis C patients. The Romanian experience.
        J Gastrointest Liver Dis. 2007; 16: 31-37
        • Poynard T.
        • Imbert-Bismut F.
        • Ratziu V.
        • et al.
        Biochemical markers of liver fibrosis in patients infected by hepatitis C virus: longitudinal validation in a randomized trial.
        J Viral Hepat. 2002; 9: 128-133
        • Poynard T.
        • Munteanu M.
        • Ngo Y.
        • et al.
        ActiTest accuracy for the assessment of histological activity grades in patients with chronic hepatitis C, an overview using Obuchowski measure.
        Gastroenterol Clin Biol. 2010; 34: 388-396
        • Birerdinc A.
        • Stepanova M.
        • Pawloski L.
        • et al.
        Caffeine is protective in patients with non-alcoholic fatty liver disease.
        Aliment Pharmacol Ther. 2012; 35: 76-82
        • Chen S.
        • Teoh N.C.
        • Chitturi S.
        • et al.
        coffee and non-alcoholic fatty liver disease: brewing evidence for hepatoprotection?.
        J Gastroenterol Hepatol. 2014; 29: 435-441
        • Andersen I.M.
        • Tengesdal G.
        • Lie B.A.
        • et al.
        Effects of coffee consumption, smoking, and hormones on risk for primary sclerosing cholangitis.
        Clin Gastroenterol Hepatol. 2014; 12: 1019-1028
        • Lammert C.
        • Juran B.D.
        • Schlicht E.
        • et al.
        Reduced coffee consumption among individuals with primary sclerosing cholangitis but not primary biliary cirrhosis.
        Clin Gastroenterol Hepatol. 2014; 12: 1562-1568
        • Corrao G.
        • Lepore A.R.
        • Torchio P.
        • et al.
        The effect of drinking coffee and smoking cigarettes on the risk of cirrhosis associated with alcohol consumption. A case-control study.
        Eur J Epidemiol. 1994; 10: 657-664
        • Klatsky A.L.
        • Morton C.
        • Udaltsova
        • et al.
        Coffee, cirrhosis, and transaminase enzymes.
        Arch Intern Med. 2006; 166: 1190-1195
        • White D.L.
        • Richardson P.A.
        • Al-Saadi M.
        • et al.
        Dietary history and physical activity and risk of advanced liver disease in veterans with chronic hepatitis C infection.
        Dig Dis Sci. 2011; 56: 1835-1847
        • Machado S.R.
        • Parise E.R.
        • de Carvalho L.
        Coffee has hepatoprotective benefits in Brazilian patients with chronic hepatitis C even in lower daily consumption than in American and European populations.
        Braz J Infect Dis. 2014; 18: 170-176
        • Kumari R.
        • Kim W.R.
        Coffee: a panacea or snake oil for the liver?.
        Clin Gasteroenterol Hepatol. 2014; 12: 1569-1571
        • Peng Z.
        • Fernandez P.
        • Wilder T.
        • et al.
        Ecto-5′-nucleotidase (CD73)-mediated extracellular adenosine production plays a critical role in hepatic fibrosis.
        FASEB J. 2008; 22: 2263-2272
        • Arauz J.
        • Moreno M.G.
        • Cortes-Reynosa P.
        • et al.
        Coffee attenuates fibrosis by decreasing the expression of TGF-β and CTGF in a murine model of liver damage.
        J Appl Toxicol. 2013; 33: 970-979
        • Gressner O.A.
        • Lahme B.
        • Rehbein K.
        • et al.
        Pharmacological application of caffeine inhibits TGF-beta-stimulated connective tissue growth factor expression in hepatocytes via PPARgamma and SMAD2/3-dependent pathways.
        J Hepatol. 2008; 49: 758-767
        • Nakai K.
        • Tanaka H.
        • Hanada K.
        • et al.
        Decreased expression of cytochromes P450 1A2, 2E1, and 3A4 and drug transporters Na+-taurocholate-cotransporting polypeptide, organic cation transporter 1, and organic anion-transporting peptide-C correlates with the progression of liver fibrosis in chronic hepatitis C patients.
        Drug Metab Dispos. 2008; 36: 1786-1793
        • Devasagayam T.P.
        • Kamat J.P.
        • Mohan H.
        • et al.
        Caffeine as an antioxidant: inhibition of lipid peroxidation induced by reactive oxygen species.
        Biochim Biophys Acta. 1996; 1282: 63-70
        • Gallus S.
        • Tavani A.
        • Negri E.
        • et al.
        Does coffee protect against liver cirrhosis?.
        Ann Epidemiol. 2002; 12: 202-205
        • Scalbert A.
        • Williamson G.
        Dietary intake and bioavailability of polyphenols.
        J Nutr. 2000; 130: 2073S-2085S
        • Cavin C.
        • Holzhaeuser D.
        • Scharf G.
        • et al.
        Cafestol and kahweol, two coffee specific diterpenes with anticarcinogenic activity.
        Food Chem Toxicol. 2002; 40: 1155-1163
        • Huber W.W.
        • Scharf G.
        • Rossmanith W.
        • et al.
        The coffee components kahweol and cafestol induce gamma-glutamylcysteine synthetase, the rate limiting enzyme of chemoprotective glutathione synthesis, in several organs of the rat.
        Arch Toxicol. 2002; 75: 685-694
        • Lee C.
        Antioxidant ability of caffeine and its metabolites based on the study of oxygen radical absorbing capacity and inhibition of LDL peroxidation.
        Clin Chim Acta. 2000; 295: 141-154
        • Leandro G.
        • Mangia A.
        • Hui J.
        • et al.
        Relationship between steatosis, inflammation, and fibrosis in chronic hepatitis C: a meta-analysis of individual patient data.
        Gastroenterology. 2006; 130: 1636-1642
        • Hézode C.
        • Roudot-Thoraval F.
        • Nguyen S.
        • et al.
        Daily cannabis smoking as a risk factor for progression of fibrosis in chronic hepatitis C.
        Hepatology. 2005; 42: 63-71
        • Danielsson J.
        • Kangastupa P.
        • Laatikainen T.
        • et al.
        Dose- and gender-dependent interactions between coffee consumption and serum GGT activity in alcohol consumers.
        Alcohol Alcohol. 2013; 48: 303-307
        • Wang Y.
        • Ye F.
        • Ke Q.
        • et al.
        Gender-dependent histone deacetylases injury may contribute to differences in liver recovery rates of male and female mice.
        Transplant Proc. 2013; 45: 463-473