Abstract
Background
Coffee has inverse relationships with both type 2 diabetes and hepatic fibrosis in patients with nonalcoholic fatty liver disease (NAFLD).
Aim
Relationships were explored between coffee intake and insulin resistance (IR) with respect to NAFLD histologic severity.
Methods
We analyzed data from 782 adults (≥18 years) in the Nonalcoholic Steatohepatitis Clinical Research Network (NASH CRN) from 2004–2008. IR was assessed using the HOMA-IR. We modeled associations between coffee intake and NAFLD histologic severity using multiple logistic regression and interactions between coffee and IR on NAFLD histology were explored.
Results
Among 782 participants, 38% (n=295) were men, 12% (n=97) were Latino, mean age (±standard deviation) was 48±12 years. Median BMI was 33.5 kg/m2 [interquartile range, 29.7–38.3] and median HOMA-IR was 4.3 [2.7–7.2]. Diabetes was present in 24% (n=189). NASH was present in 79% (n=616), and 25% (n=199) had advanced fibrosis. The frequency of coffee intake (cups/day, cpd) was as follows: 0 cpd, n=230 (29%); <1 cpd, n=219 (28%); 1 -<2 cpd, n=116 (15%); ≥2 cpd, n=217 (28%). The effect of coffee on fibrosis varied with degree of IR (interaction p=0.001). Coffee consumers with less IR, defined as HOMA-IR<4.3, had a lower odds of advanced fibrosis (OR=0.64; 95% CI, [0.46–0.88], p=0.001). There was no protective effect of coffee on advanced fibrosis among individuals with higher HOMA-IR (OR=1.06, 95% CI [0.87–1.28], p=0.6).
Conclusions
Coffee intake is inversely associated with advanced fibrosis among NAFLD patients with lower HOMA-IR. Our findings warrant further investigation given the worldwide ubiquity of coffee intake.
Keywords: Nonalcoholic Steatohepatitis, Histology, HOMA, Diet, Lifestyle
Introduction
Nonalcoholic fatty liver disease (NAFLD) is the leading cause of chronic liver disease affecting up to 30% of the U.S. adult population.(1–4) NAFLD encompasses a histopathological spectrum ranging from bland steatosis to nonalcoholic steatohepatitis (NASH). NASH is characterized by hepatic inflammation, ballooning degeneration of hepatocytes, and hepatic steatosis -- with or without fibrosis. NASH is associated with decreased long-term survival, higher propensity for progression to advanced fibrosis, and cirrhosis, and higher risk of hepatocellular carcinoma (HCC).(5, 6)
Although the pathophysiology of NAFLD remains incompletely defined, a number of clinical factors are intimately associated with this disease -- obesity, insulin resistance, diabetes mellitus, hypertension, and hyperlipidemia. Lifestyle habits, such as fructose consumption, physical activity level, alcohol, and tobacco use, have also been studied in NAFLD and shown to impact disease severity.(7–12) Further insights into relationships between NAFLD and other modifiable lifestyle factors may contribute to the development of intervention strategies for NAFLD.
Coffee consumption is a popular lifestyle choice worldwide; at least 50% of U.S. adults consume coffee on a daily basis.(13) Beneficial effects of coffee have been observed in multiple chronic medical conditions including type 2 diabetes, Parkinson’s disease, prostate cancer, hepatitis C (HCV), HCC, and more recently, NAFLD.(14–33) Moreover, a large population-based study found that increasing coffee intake is associated with a modest decrease in all-cause mortality.(31) Therefore, we hypothesized that coffee would be associated with decreased NAFLD histologic severity. We further hypothesized that the beneficial effects of coffee on NAFLD histologic severity would be modified by severity of insulin resistance.
Patients and Methods
Research Design and Study Population
This is a cross-sectional study using prospectively collected data from participants enrolled in the U.S. multicenter collaborative research consortium known as the Nonalcoholic Steatohepatitis Clinic Research Network (NASH CRN) from 2004 to 2008.(34–36) The NASH CRN is a National Institute of Diabetes and Digestive and Kidney Diseases (NIDDK) funded consortium to study the pathogenesis of NAFLD and the determinants of disease progression and severity. We included in our analyses NASH CRN participants who were enrolled in either the NASH CRN Database or the PIVENS Trial (Pioglitazone Versus Vitamin E Versus Placebo for the Treatment of Non-Diabetic Patients with NASH; ClinicalTrials.gov Identifier: NCT00063622).(35, 36) Only baseline data at the time of enrollment into the NASH CRN and PIVENS trial were utilized so that the results were not impacted by their involvement in the NASH CRN related studies. All laboratory, dietary, and hepatic histological data used for our analyses were concurrent, being obtained within 6-months of enrollment into the NASH CRN. Only those participants who were 18 years of age or older and had data available pertaining to both NAFLD liver histology and dietary data documenting coffee consumption were used for these analyses.
Informed consent was obtained in writing from all participants in the NASH CRN and the study protocol conformed to the ethical guidelines of the 1975 Declaration of Helsinki as reflected in a priori approval by all the institutional review boards at all NASH CRN participating sites.
NAFLD Diagnosis
As part of the entry criteria, all patients enrolled in the NASH CRN must meet the following criteria: (1) documented absence of other etiologies of chronic liver disease (viral, autoimmune, drug-induced, cholestatic, genetic), (2) no clinically significant history of alcohol consumption, which was defined as: > 20 or > 10 grams/day for men and women, respectively, in the NASH CRN Database; or >30 or > 20 grams/day for men and women, respectively, in the PIVENS Trial, and (3) histologically confirmed NAFLD according to the NASH-CRN criteria (see below).
Coffee Consumption
Dietary information was obtained using a validated dietary questionnaire (Block Food Frequency Questionnaire, version 1998), based upon self-reported typical eating habits over the past year.(37) Coffee consumption was reported as average number of cups per day (cpd). Data collection with respect to coffee consumption in the Block Food Frequency Questionnaire was capped at a maximum of 5 cpd. No distinction was made between caffeinated and decaffeinated coffee in the Block Food Frequency Questionnaire.
Outcomes
The primary outcome of interest for the current investigation was advanced fibrosis based on liver histology, categorized as: none to moderate (≤ Stage 2) or advanced (> Stage 2). A secondary outcome of interest was the presence versus absence of NASH histology.
Liver biopsy slides for all study participants were stained with hematoxylin and eosin and Masson’s trichrome and were reviewed and scored centrally by the members of the NASH CRN Pathology Committee who were blinded to clinical data. The Nonalcoholic Fatty Liver Disease Activity Score (NAS) and stage of fibrosis were based on the histological scoring system for NAFLD developed by the NASH CRN.(38) The NAS includes a systematic assessment of the following histologic features: steatosis, hepatocyte ballooning, and lobular inflammation. The final diagnosis for each biopsy was determined by consensus of the NASH CRN Pathology Committee as follows: (1) definite steatohepatitis; (2) definitely not steatohepatitis; and (3) borderline steatohepatitis. Fibrosis on liver biopsy was staged from 0 to 4 as follows: 0 = none; 1a = mild zone 3 (central) perisinusoidal fibrosis; 1b = moderate zone 3 perisinusoidal fibrosis; 1c = periportal and portal fibrosis (zone 1 only); 2 = both zone 3 perisinusoidal and periportal or portal fibrosis; 3 = bridging fibrosis; and 4 = cirrhosis. Liver biopsy length (mm) was also recorded.
Covariates
Information regarding the following variables was also obtained for all individuals included in our analyses:
Demographics
age, gender, self-reported race and ethnicity (categorized as Non-Latino, Latino, Caucasian, Black, Asian, American Indian, Native Hawaiian/Pacific Islander, and Multiracial), education level (categorized as < high school, Some College, and Advanced Degree), and annual income (categorized as < $30,000, $30,000–$49,999, ≥ $50,000).
Anthropometrics
Height (meters, m), weight (kilograms, kg) and waist circumference (centimeters, cm) were measured at the time of enrollment into the NASH CRN. Body mass index (BMI) kg/m2 was calculated for each participant.
Laboratory Data
Laboratory data were collected at the time of enrollment and included measurements of aspartate and alanine aminotransferases (AST and ALT, respectively), gamma-glutamyl transferase (GGT), serum albumin, total bilirubin, alkaline phosphatase, platelet count, lipid profile (including total cholesterol, low-density lipoprotein [LDL], high-density lipoprotein [HDL] and triglycerides), and fasting insulin and fasting glucose levels. The homeostasis model assessment of insulin resistance (HOMA-IR) was calculated from fasting glucose and insulin levels, using the following equation: (fasting insulin [μU/mL])*(fasting glucose [mg/dL])/405. The HOMA-IR was not calculated for individuals taking insulin or for individuals in whom fasting insulin data was not available.
Clinical Data
Information regarding comorbidities, including diabetes mellitus and hyperlipidemia were also collected.
Smoking, Alcohol Intake and Physical Activity
Participants’ smoking history was documented and the number of pack-years smoked was recorded. Alcohol use was also documented and recorded as average number of drinks per month. Physical activity levels were calculated based upon self-reported leisure-time physical activity using a questionnaire derived from the National Health and Nutrition Examination Survey.(39–41) Using a standard reference for metabolic equivalent (MET) intensities for specific activities and the reported duration of each activity, a score for each individual’s total physical activity, expressed as metabolic hours per week (MET hours/week), was generated.(40)
Statistical Analyses
Continuous data are reported as mean ± standard deviation (sd) or median [Interquartile range, IQR] values, as appropriate. Student’s t-test, ANOVA, Kruskal Wallis and Mann Whitney U testing were applied for comparisons, as appropriate. Chi2 testing was used to compare categorical variables.
Associations between clinical characteristics and advanced fibrosis (> stage 2) were investigated using univariate and multivariate logistic regression models. Data are reported as odds ratios (OR) with 95% Confidence Intervals (95% CI). Stepwise logistic regression analysis was used to identify significant predictors of advanced fibrosis. The p-value for addition or elimination from the models was p < 0.05, and the models were forced to include terms for age, gender and ethnicity. Interactions between coffee intake and insulin resistance (HOMA-IR) with respect to fibrosis stage were also explored. For these analyses, the participant population was divided into two groups based on the population’s median value for HOMA-IR, (< 4.3 versus ≥ 4.3), and multivariate logistic regression models were conducted for each of the two HOMA-IR groups, adjusting for variables selected from the stepwise logistic regression model. Similar logistic regression analyses were also performed for the secondary outcome of presence versus absence of NASH histology on liver biopsy.
For all analyses, STATA 11.0 (College Station, TX) was utilized. Nominal two-sided p values were used. We considered differences statistically significant when p values were < 0.05. Statistical interactions were considered statistically significant when p values were < 0.01.
Results
Patient Characteristics
Characteristics of the Overall Population
Seven hundred and eighty two adults had biopsy-proven NAFLD with complete coffee consumption history.(Table 1) The mean age of the cohort was 48 (±12) years and 38% (n=295) were men. Eighty-four percent (n=633) of participants self-identified as Caucasian, and 12% (n=97) were Latino. The majority of participants were either overweight or obese with the median BMI of the overall patient population being 33.5 kg/m2 [IQR 29.7–38.3] and median waist circumference of 108 cm [IQR 99–117]. Diabetes was present in 24% (n=189) of participants and hyperlipidemia in 56% (n=435).
Table 1.
Characteristics of the Overall Study Population (n=782)
| Age, years (mean ± sd) | 48±12 |
| Male, n (%) | 295 (38%) |
| Latino, n (%) | 97 (12%) |
| Race, n (%) | |
| Caucasian | 633 (84%) |
| Black | 20 (2.7%) |
| Asian | 36 (4.8%) |
| American Indian | 27 (3.6%) |
| Native Hawaiian/Pacific Islander | 5 (0.7%) |
| Multiracial | 32 (4.3%) |
| BMI, kg/m2 (median, IQR) | 33.5 [29.7–38.3] |
| Waist Circumference, cm (median, IQR) | 108 [99–117] |
| Biopsy Length, mm (median, IQR) | 17 [12–25] |
| NASH Histology (Definite or Probable), n (%) | 616 (79%) |
| Cirrhosis, n (%) | 59 (8%) |
| Fibrosis, n (%) | |
| ≤ Stage 2 | 583 (75%) |
| > Stage 2 | 199 (25%) |
| Steatosis, n (%) | |
| < 5% | 34 (4%) |
| 5–33% | 295 (38%) |
| 34–66% | 271 (35%) |
| >66% | 182 (23%) |
| Lobular Inflammation, n (%) | |
| 0 | 3 (0.4%) |
| < 2 under 20x | 399 (51%) |
| 2–4 under 20x | 290 (37%) |
| >4 under 20x | 90 (12%) |
| Ballooning, n (%) | |
| None | 259 (33%) |
| Few | 206 (26%) |
| Many | 317 (41%) |
| NAS, n (%) | |
| < 5 | 400 (52%) |
| ≥ 5 | 382 (48%) |
| Cups of Coffee/Day, cpd (mean ± sd) | 1.0±1.2 |
| AST, U/L (median, IQR) | 44 [32–64] |
| ALT, U/L (median, IQR) | 62 [41–92] |
| Alkaline Phosphatase, U/L (median, IQR) | 81 [66–103] |
| Total Bilirubin, mg/dL (median, IQR) | 0.7 [0.5–0.9] |
| Gamma Glutamyl Transferase, U/L (median, IQR) | 47 [29–82] |
| Glucose, mg/dL (median, IQR) | 96 [86–110] |
| Insulin, μU/mL (median, IQR) | 17.9 [11.8–27.5] |
| HOMA-IR (median, IQR) | 4.3 [2.7–7.2] |
| Diabetes, n (%) | 189 (24%) |
| Hyperlipidemia, n (%) | 435 (56%) |
| Alcohol, drinks/month (mean ± sd) | 5.3±11.6 |
| Smoking, pack-years (mean ± sd) | 7.7±16.1 |
| Physical Activity, METS/week (median, IQR) | 115 [85–151] |
| Education Level, n (%) | |
| < High School | 239 (31%) |
| Some College | 293 (37%) |
| Advanced Degree | 250 (32%) |
| Annual Income, n (%) | |
| < $30,000 | 158 (21%) |
| $30,000–$49,999 | 162 (21%) |
| ≥ $50,000 | 449 (58%) |
The median AST and ALT levels were 44 U/L [IQR, 32–64] and 62 U/L [IQR, 41–92], respectively. The median alkaline phosphatase and total bilirubin levels were 81 U/L [IQR, 66–103] and 0.7 mg/dL [IQR, 0.5–0.9], respectively. The HOMA-IR was calculated for individuals who were not taking exogenous insulin and who had the necessary data available for calculation of the HOMA-IR (n=769). The median HOMA-IR for the population was 4.3 [2.7–7.2].
The mean number of cups of coffee consumed per day (cpd) was 1.0±1.2, mean number of pack-years smoked was 7.7±16.1, and mean number of alcoholic drinks per month was 5.3±11.6. NASH histology (definite or probable) was present in 79% of participants (n=616). Advanced fibrosis (> stage 2) was present in 25% (n=199). Fifty-nine individuals (8%) had cirrhosis.
Participant Characteristics by Coffee Consumption
To determine the impact of coffee on various metabolic parameters, participants were further characterized by their daily coffee consumption (Table 2). Twenty-nine percent of participants (n=230) drank 0 cpd; 28% (n=219) drank < 1 cpd; 15% (n=116) drank 1- <2 cpd; and 28% (n=217) drank ≥2 cpd. Between these 4 groups, there was a statistically significant difference in age (p=0.0001), with individuals drinking ≥2 cpd being older than those who drank fewer cpd. Fewer individuals who self-identified as Latino drank ≥2 cpd (p=0.003). Additionally, AST levels were inversely related to daily coffee intake (p=0.02). Number of pack-years smoked and number of alcoholic beverages consumed per month also increased with increasing coffee consumption (p=0.0001 and p=0.002, respectively).
Table 2.
Comparison of Population Characteristics by Coffee Intake*
| 0 cpd (n=230) | <1 cpd (n=219) | 1- < 2 cpd (n=116) | ≥ 2 cpd (n=217) | p | |
|---|---|---|---|---|---|
| Age, years | 45 ± 12 | 45 ± 13 | 51 ± 12 | 53 ± 9 | 0.0001 |
| Male, (%) | 87 (38%) | 91 (42%) | 40 (34%) | 77 (35%) | 0.5 |
| Latino, n (%) | 25 (11%) | 35 (16%) | 22 (19%) | 15 (7%) | 0.003 |
| Race, n (%) | 0.03 | ||||
| Caucasian | 119 (89%) | 155 (76%) | 91 (83%) | 188 (87%) | |
| Black | 4 (2%) | 8 (4%) | 2 (2%) | 6 (3%) | |
| Asian | 8 (4%) | 16 (8%) | 4 (4%) | 8 (4%) | |
| American Indian | 7 (3%) | 12 (6%) | 5 (5%) | 3 (1%) | |
| Native Hawaiian/Pacific Islander | 0 (0%) | 4 (2%) | 1 (1%) | 0 (0%) | |
| Multiracial | 6 (3%) | 10 (5%) | 6 (6%) | 10 (5%) | |
| BMI, kg/m2 | 34.2 [29.5–38.7] | 33.7 [29.2–38.7] | 33.0 [29.7–38.0] | 33.3 [29.8–37.8] | 0.8 |
| Waist Circumference, cm | 108 [99–119] | 108 [98–116] | 106 [98–117] | 109 [99–118] | 0.7 |
| Biopsy Length, mm | 17 [11–25] | 19 [12–25] | 17 [13–23] | 16 [11–24] | 0.4 |
| NASH Histology (Definite or Probable), n (%) | 178 (77%) | 172 (79%) | 96 (83%) | 170 (78%) | 0.7 |
| Cirrhosis, n (%) | 19 (8%) | 16 (7%) | 6 (5%) | 18 (8%) | 0.7 |
| Fibrosis, n (%) | 0.2 | ||||
| ≤ Stage 2 | 173 (76%) | 170 (78%) | 79 (69%) | 156 (72%) | |
| > Stage 2 | 55 (24%) | 48 (22%) | 36 (31%) | 60 (28%) | |
| AST, U/L | 43 [32–62] | 45 [31–65] | 54 [35–73] | 40 [31–58] | 0.02 |
| ALT, U/L | 60 [42–88] | 62 [39–95] | 67 [44–104] | 60 [40–91] | 0.5 |
| Alkaline Phosphatase, U/L | 84 [67–106] | 79 [67–100] | 80 [67–97] | 60 [40–91] | 0.8 |
| Total Bilirubin, mg/dL | 0.6 [0.5–0.9] | 0.7 [0.5–0.9] | 0.7 [0.5–0.9] | 0.6 [0.5–0.9] | 0.3 |
| Gamma Glutamyl Transferase, U/L | 46 [29–75] | 49 [29–83] | 50 [33–91] | 47 [29–77] | 0.4 |
| HOMA-IR | 4.3 [2.6–7.7] | 4.3 [2.5–6.8] | 4.8 [3.2–7.1] | 4.2 [2.7–7.0] | 0.5 |
| Diabetes, n (%) | 53 (23%) | 50 (23%) | 24 (21%) | 62 (29%) | 0.3 |
| Hyperlipidemia, n (%) | 127 (55%) | 118 (54%) | 59 (51%) | 131 (60%) | 0.3 |
| Alcohol, drinks/month | 3.7 ± 9.1 | 4.9 ± 11.5 | 6.3 ± 13.3 | 7.0 ± 12.9 | 0.002 |
| Smoking, pack-years | 4.6 ± 14.0 | 5.1 ± 11.2 | 10.3 ± 20.7 | 12.2 ± 18.4 | 0.0001 |
| Physical Activity, METS/week | 113 [88–146] | 121 [91–161] | 108 [80–143] | 113 [78–149] | 0.3 |
| Education Level, n(%) | 0.8 | ||||
| < High School | 65 (28%) | 70 (32%) | 34 (29%) | 70 (32%) | |
| Some College | 92 (40%) | 73 (33%) | 47 (41%) | 81 (37%) | |
| Advanced Degree | 73 (32%) | 76 (35%) | 35 (30%) | 66 (30%) | |
| Annual Income, n (%) | 0.4 | ||||
| < $30,000 | 47 (21%) | 51 (24%) | 19 (17%) | 41 (19%) | |
| $30,000–$49,999 | 44 (19%) | 43 (20%) | 33 (29%) | 42 (20%) | |
| ≥ $50,000 | 135 (60%) | 123 (57%) | 62 (54%) | 129 (61%) |
Unless otherwise noted, data presented are mean ± sd; or median [IQR]
cpd, cups per day
Participant Characteristics by Advanced Fibrosis
Comparisons of study participants categorized by presence or absence of advanced fibrosis (> stage 2) are demonstrated in Table 3. Compared to those without advanced fibrosis, individuals with advanced fibrosis, had significantly higher AST, alkaline phosphatase and GGT levels (p<0.0001, for all). Diabetes was more prevalent, and HOMA-IR was significantly higher, among individuals with advanced fibrosis (p<0.0001 for both). The number of alcoholic beverages consumed per month was significantly lower among individuals with advanced fibrosis, while the number of pack-years smoked was significantly greater among individuals with advanced fibrosis (p<0.0001 and p=0.0004, respectively).
Table 3.
Comparison of Population by Advanced Fibrosis (≤ Stage 2 vs. > Stage 2)
| ≤ Stage 2 | > Stage 2 | p | |
|---|---|---|---|
| Age, years (mean ± sd) | 46.5±12.2 | 54.0±10.4 | <0.0001 |
| Male, n (%) | 230 (40%) | 62 (31%) | 0.03 |
| Latino, n (%) | 82 (14%) | 13 (7%) | 0.004 |
| Race, n (%) | 0.04 | ||
| Caucasian | 464 (84%) | 165 (85%) | |
| Black | 14 (3%) | 6 (3%) | |
| Asian | 30 (5%) | 6 (3%) | |
| American Indian | 24 (4%) | 2 (1%) | |
| Native Hawaiian/Pacific Islander | 4 (1%) | 1 (1%) | |
| Multiracial | 18 (3%) | 14 (7%) | |
| BMI, kg/m2 (median, IQR) | 33.2 [29.6–37.8] | 34.6 [29.9–39.2] | 0.05 |
| Waist Circumference, cm (median, IQR) | 108 [99–116] | 110 [99–122] | 0.010 |
| Biopsy Length, mm (median, IQR) | 17 [11–24] | 19 [14–26] | 0.004 |
| Coffee (cpd) | 0.9 ± 1.2 | 1.1 ± 1.2 | 0.1 |
| AST, IU/L | 41 [30–61] | 53 [39–74] | <0.0001 |
| ALT, IU/L | 63 [41–92] | 60 [43–90] | 0.5 |
| Alkaline Phosphatase, IU/L | 79 [64–97] | 90 [70–116] | <0.0001 |
| Total Bilirubin, mg/dL | 0.7 [0.5–0.9] | 0.7 [0.5–1.0] | 0.4 |
| GGT (IU/L) | 42 [28–72] | 66 [45–107] | <0.0001 |
| Glucose (mg/dL) | 95 [86–108] | 99 [86–119] | 0.01 |
| Insulin (μIU/mL) | 17 [11–25] | 22 [15–36] | <0.0001 |
| HOMA-IR | 3.9 [2.5–6.3] | 5.9 [37–9.9] | <0.0001 |
| HOMA, dichotomized, n (%) | |||
| <4.3 | 291 (51%) | 57 (29%) | <0.0001 |
| ≥4.3 | 277 (49%) | 138(71%) | |
| Diabetes, n (%) | 111 (19%) | 77 (39%) | <0.0001 |
| Hyperlipidemia, n (%) | 319 (55%) | 113 (57%) | 0.7 |
| Alcohol, drinks/month | 5.8±11.8 | 3.5±10.5 | <0.0001 |
| Smoking, ≥ 10 pack-years | 6.4±15.0 | 11.5±19.5 | 0.0004 |
| Physical Activity, METS/week | 118 [89–153] | 108 [79–140] | 0.008 |
| Education Level, n(%) | 0.02 | ||
| < High School | 164 (28%) | 73 (37%) | |
| Some College | 214 (37%) | 76 (38%) | |
| Advanced Degree | 200 (35%) | 50 (25%) | |
| Annual Income, n(%) | 0.01 | ||
| < $30,000 | 106 (19%) | 52 (27%) | |
| $30,000–$49,999 | 113 (20%) | 47 (24%) | |
| ≥ $50,000 | 349 (61%) | 97 (49%) |
Unless otherwise noted, data presented are mean ± sd; or median [IQR]
Univariate Analyses for Advanced Fibrosis
To identify features associated with advanced fibrosis, univariate analyses were performed (Table 4). Age, BMI, waist circumference, liver biopsy length, AST, alkaline phosphatase, glucose, insulin, HOMA-IR, diabetes, and smoking were significantly, positively associated with advanced fibrosis. In contrast, male gender, Latino ethnicity, alcohol consumption, physical activity level, education level, and annual income were significantly inversely associated with advanced fibrosis.
Table 4.
Univariate Analyses for Risk of Advanced Fibrosis (≤ Stage 2 vs. > Stage 2)
| Odds Ratio | 95% CI | p | |
|---|---|---|---|
| Age, years | 1.06 | [1.04–1.07] | <0.0001 |
| Male | 0.68 | [0.49–0.97] | 0.03 |
| Latino | 0.42 | [0.23–0.78] | 0.006 |
| Race | 1.03 | [0.91–1.18] | 0.6 |
| BMI, kg/m2 | 1.02 | [1.00–1.05] | 0.08 |
| Waist, cm | 1.02 | [1.004–1.03] | 0.006 |
| Biopsy length, mm | 1.03 | [1.01–1.04] | <0.0001 |
| Coffee, cpd | 1.10 | [0.96–1.24] | 0.2 |
| AST, IU/L | 1.009 | [1.004–1.012] | <0.0001 |
| ALT, IU/L | 0.99 | [0.99–1.00] | 0.4 |
| Alk Phos, IU/L | 1.01 | [1.01–1.02]] | <0.0001 |
| Total Bili, mg/dL | 1.09 | [0.73–1.61] | 0.7 |
| GGT (IU/L) | 1.004 | [1.002–1.006] | <0.0001 |
| Glucose, (mg/dL) | 1.01 | [1.00–1.01] | 0.004 |
| Insulin, (μIU/mL) | 1.02 | [1.01–1.03] | <0.0001 |
| HOMA-IR | 1.06 | [1.03–1.09] | <0.0001 |
| Diabetes | 2.66 | [1.87–3.78] | <0.0001 |
| Hyperlipidemia | 1.07 | [0.77–1.47] | 0.7 |
| Alcohol, drinks/month | 0.98 | [0.96–0.99] | 0.02 |
| Smoking, ≥10 pack-years | 2.01 | [1.40–2.89] | <0.0001 |
| Physical Activity, METS/week | 0.995 | [0.992–0.998] | 0.007 |
| Education Level | 0.61 | [0.57–0.92] | 0.006 |
| Annual Income | 0.74 | [0.61–0.91] | 0.003 |
Interaction between Coffee and HOMA-IR in Increasing the Odds of Advanced Fibrosis
The interaction between coffee intake and insulin resistance (HOMA-IR) with respect to odds of advanced fibrosis was investigated. We found a statistically significant interaction between HOMA-IR and coffee intake with respect to increasing the odds of advanced fibrosis (p=0.001). Due to this significant interaction, we examined the effect of coffee intake on the odds of advanced fibrosis in multivariate modeling after stratifying by the degree of insulin resistance, as measured by HOMA-IR (see below). Specifically, we dichotomized the participant population into two groups based in insulin resistance using the population’s median HOMA-IR value (< 4.3 versus ≥4.3) as the cutoff.
Multivariate Analyses for Advanced Fibrosis
The results from multivariate logistic regression modeling for advanced fibrosis, which included an interaction term between coffee intake and HOMA-IR, are demonstrated in Table 5. In the multivariate model, the following factors were significantly associated with increased odds of advanced fibrosis: age (OR 1.05, 95% CI [1.03–1.07], p<0.0001), AST (OR 1.01, 95% CI [1.004–1.011], p<0.0001), GGT (OR 1.003, 95% CI [1.003–1.005], p=0.02), diabetes (OR 1.62, 95% CI [1.05–2.09], p=0.03), smoking (OR 1.60, 95% CI [1.03–2.47], p=0.04), and biopsy length (OR 1.03, 95% CI [1.01–1.05], p=0.002). Conversely, alcoholic beverage consumption (OR 0.98, 95% CI [0.96–0.99], p=0.04) and coffee intake (OR 0.68, 95% CI [0.52–0.89], p=0.005) were significantly associated with decreased odds for advanced fibrosis.
Table 5.
Multivariate Analyses for Risk for Advanced Fibrosis (≤ Stage 2 vs. > Stage 2)
| Covariate | Odds Ratio [95% CI] | p |
|---|---|---|
| Age (years) | 1.05 [1.03–1.07] | <0.0001 |
| Male | 1.04 [0.69–1.58] | 0.8 |
| Latino | 0.66 [0.33–1.31] | 0.2 |
| Waist Circumference (cm) | 1.01 [0.99–1.03] | 0.06 |
| AST (U/L) | 1.01 [1.004–1.01] | <0.0001 |
| GGT (U/L) | 1.003 [1.003–1.005] | 0.02 |
| Diabetes | 1.62 [1.05–2.49] | 0.03 |
| Smoking (≥ 10 pack-years) | 1.60 [1.03–2.47] | 0.04 |
| Alcohol (drinks/month) | 0.98 [0.96–0.99] | 0.04 |
| Biopsy Length (mm) | 1.03 [1.01–1.05] | 0.002 |
| Coffee (cpd) | 0.68 [0.52–0.89] | 0.005 |
| HOMA-IR | 1.004 [0.97–1.04] | 0.8 |
| Coffee*HOMA-IR# | 1.07 [1.03–1.10] | <0.0001 |
Interaction term between Coffee intake (cpd) and HOMA-IR
There was a significant interaction between coffee intake and HOMA-IR with respect to the odds of advanced fibrosis (p<0.0001). Among individuals with less insulin resistance (HOMA-IR < 4.3), increasing coffee intake was associated with decreased odds of advanced fibrosis, OR 0.64, 95% CI: 0.46–0.88, p=0.001. However, among individuals with greater insulin resistance (HOMA-IR ≥4.3) there was no significant association between coffee intake and odds of advanced fibrosis, OR 1.06, 95% CI: 0.87–1.28, p=0.6, (Figure 1).
Figure 1.
Differential effect of coffee consumption on risk for advanced fibrosis by degree of insulin resistance (HOMA-IR < 4.3 versus HOMA-IR ≥ 4.3).
Multivariate Analyses for NASH Histology
Our analyses of the secondary outcome of NASH histology found no significant associations between coffee consumption and odds of NASH histology (data not shown).
Discussion
In this study we demonstrate that coffee intake has a differential effect on the odds of advanced fibrosis in NAFLD and is dependent on underlying insulin resistance. Among individuals with less insulin resistance as indicated by a lower HOMA-IR (<4.3), increasing coffee intake is associated with decreased odds of advanced fibrosis. However, among individuals with more severe insulin resistance as indicated by a higher HOMA-IR (≥ 4.3), there was no significant effect of coffee intake on the odds of advanced fibrosis. The mechanisms underlying the differential effect of coffee intake on hepatic fibrosis are not immediately evident from our data. However, type 2 diabetes, a state of extreme insulin resistance, has been cross-sectionally associated with advanced fibrosis in patients with NAFLD.(42) It is plausible that the detrimental effects of greater insulin resistance on hepatic fibrosis overshadow the beneficial effects of coffee on hepatic fibrosis. Specifically, many patients with greater insulin resistance are actually diabetic, and the preventive effects of coffee may be related to its effects on improving insulin sensitivity early on in the spectrum of progression of loss of function of pancreatic beta-cell. One could further speculate that once a patient develops overt diabetes there may be limited or no benefits with coffee intake.
Aside from fibrosis, the daily coffee consumption did not affect any of the other histological features associated with NASH. Interestingly, our data also demonstrate that the length of the liver biopsy was significantly and positively associated with the odds of advanced fibrosis. Biopsy length maintained its significant association with advanced fibrosis in multivariate analysis. These results are consistent with the well-known sampling variability and sampling errors that may be associated with liver biopsies, and also underscore the need for the development of accurate and valid noninvasive markers for NASH and fibrosis in NASH.(43)
There are only a few publications investigating the associations between coffee and/or caffeine consumption, and NAFLD; two of these studies incorporated hepatic histologic data.(28–30, 32, 33). In a study of patients with biopsy-proven NASH, Molloy et al. demonstrated that individuals with stage 0–1 fibrosis consumed significantly greater amounts of coffee caffeine per day, compared to individuals with stage 2–4 fibrosis (p=0.035).(28) These authors concluded that coffee may confer a decreased risk for advanced fibrosis and speculated that moderate coffee consumption may be a readily implemented adjunct to the management of patients with NASH. Anty et al. also conducted a study investigating the influence of coffee and caffeine consumption on hepatic fibrosis among bariatric surgery patients who underwent intra-operative hepatic wedge biopsy.(29) These investigators found that the consumption of regular coffee, but not espresso, was independently associated with less fibrosis (< stage 2). The reasons for the differential effects of regular coffee and espresso on hepatic fibrosis found in their bariatric surgery population remain unclear, but these investigators speculated that the differences might be attributable to chemical alterations induced during the preparation of these two forms of caffeinated beverage.
It is not fully known what substances in coffee may be beneficial in the amelioration of severity of hepatic fibrosis, but it is known that coffee contains hundreds of components in addition to caffeine. These components include: phenolic compounds; chlorogenic acids; melanidins; maillard reaction products; lignans; ferulic acid; diterpenes; and vitamins and minerals. It is worth noting that chlorogenic acids have been previously demonstrated to have the following effects: anti-oxidant properties; increase insulin sensitivity; decrease inflammation; and decrease gluconeogenesis.(44) However, the specific mechanisms by which coffee may exert any beneficial effects in NAFLD and other disease processes remain to be unraveled.
The strengths of our study include its large sample size with a substantial number of patients with biopsy-proven NASH, including a broad range of hepatic fibrosis severity. The data for this study were prospectively collected through the NASH CRN and included extensive clinical, laboratory and histologic information on study participants. The availability of these data allowed us to adjust for clinically relevant potential confounders in our multivariate analyses. The issue of confounding is particularly important in the context of dietary coffee intake as heavy coffee consumption may be strongly associated with other factors such as lack of exercise and/or cigarette smoking and, perhaps, with an overall less healthy lifestyle.(14) Cigarette smoking has been demonstrated to be associated with increased risk for advanced fibrosis in NASH.(7) Conversely, moderate alcohol consumption and vigorous physical activity levels have been demonstrated to be associated with lower risk for advanced fibrosis in NASH.(8, 39) In our analyses, we found that tobacco and alcohol use in our patient population increased with increasing coffee intake. As such, we felt it was imperative to include these potential confounders in our multivariate models. Although we included a measure of physical activity in our univariate models, this variable did not contribute to the generation of the final multivariate models.
Our study also had some limitations including: a paucity of information regarding caffeinated versus decaffeinated coffee; truncation of coffee intake at a maximum of 5 cups per day in the Block Food Frequency Questionnaire; the self-reported nature of coffee drinking habits based on dietary recall; and lack of information on coffee cup size, type of coffee, and strength of the brew. As has been described previously in studies investigating associations between dietary intake and disease processes, some measurement error is inevitable.(45). In large-scale epidemiological studies, such as the NASH CRN, self-administered food frequency questionnaires, rather than exhaustive food diaries, are often used. Although food diaries may have the advantage of reducing reliance on memory of dietary intake, validation studies comparing food frequency questionnaires to food diaries have shown good agreement between the two.(46, 47) We also note that, despite the absence of data pertaining to: exact serving size; type of coffee; and strength of the brew in our analyses; we were able to demonstrate an effect of coffee on advanced fibrosis in NAFLD. However, the potential for misclassification in the amount of coffee consumed does limit the ability to draw valid conclusions from our data on the specific amounts of coffee needed to exert a beneficial effect.
In conclusion, we demonstrate that increasing coffee intake is associated with decreased odds of advanced fibrosis among patients with less insulin resistance. No beneficial effect of coffee intake was seen among patients with greater insulin resistance. The potential for beneficial effects of coffee intake among patients with chronic liver diseases is becoming more evident in the literature. However, longitudinal studies are needed to further investigate the impact of coffee consumption on hepatic fibrosis, and whether coffee may be a useful adjunct therapy in NAFLD.
Acknowledgments
Financial Support:
The Nonalcoholic Steatohepatitis Clinical Research Network (NASH CRN) is supported by the National Institute of Diabetes and Digestive and Kidney Diseases (NIDDK) (grants U01DK061718, U01DK061728, U01DK061731, U01DK061732, U01DK061734, U01DK061737, U01DK061738, U01DK061730, U01DK061713), and the National Institute of Child Health and Human Development (NICHD).
Several clinical centers use support from the National Center for Advancing Translational Sciences (NCATS) in conduct of NASH CRN Studies (grants UL1TR000439, UL1TR000077, UL1TR000436, UL1TR000150, UL1TR000424, UL1TR000006, UL1TR000448, UL1TR000040, UL1TR000100, UL1TR000004, UL1TR000423, UL1TR000058, UL1TR000067, UL1TR000454).
References
- 1.Williams CD, Stengel J, Asike MI, Torres DM, Shaw J, Contreras M, Landt CL, et al. Prevalence of nonalcoholic fatty liver disease and nonalcoholic steatohepatitis among a largely middle-aged population utilizing ultrasound and liver biopsy: a prospective study. Gastroenterology. 2011;140:124–131. doi: 10.1053/j.gastro.2010.09.038. [DOI] [PubMed] [Google Scholar]
- 2.Browning JD, Szczepaniak LS, Dobbins R, Nuremberg P, Horton JD, Cohen JC, Grundy SM, et al. Prevalence of hepatic steatosis in an urban population in the United States: impact of ethnicity. Hepatology. 2004;40:1387–1395. doi: 10.1002/hep.20466. [DOI] [PubMed] [Google Scholar]
- 3.Ekstedt M, Franzen LE, Mathiesen UL, Thorelius L, Holmqvist M, Bodemar G, Kechagias S. Long-term follow-up of patients with NAFLD and elevated liver enzymes. Hepatology. 2006;44:865–873. doi: 10.1002/hep.21327. [DOI] [PubMed] [Google Scholar]
- 4.Rafiq N, Bai C, Fang Y, Srishord M, McCullough A, Gramlich T, Younossi ZM. Long-term follow-up of patients with nonalcoholic fatty liver. Clin Gastroenterol Hepatol. 2009;7:234–238. doi: 10.1016/j.cgh.2008.11.005. [DOI] [PubMed] [Google Scholar]
- 5.De Minicis S, Day C, Svegliati-Baroni G. From NAFLD to NASH and HCC: Pathogenetic Mechanisms and Therapeutic Insights. Curr Pharm Des. 2013 [PubMed] [Google Scholar]
- 6.White DL, Kanwal F, El-Serag HB. Association between nonalcoholic fatty liver disease and risk for hepatocellular cancer, based on systematic review. Clin Gastroenterol Hepatol. 2012;10:1342–1359. e1342. doi: 10.1016/j.cgh.2012.10.001. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Zein CO, Unalp A, Colvin R, Liu YC, McCullough AJ. Smoking and severity of hepatic fibrosis in nonalcoholic fatty liver disease. J Hepatol. 2011;54:753–759. doi: 10.1016/j.jhep.2010.07.040. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Dunn W, Sanyal AJ, Brunt EM, Unalp-Arida A, Donohue M, McCullough AJ, Schwimmer JB. Modest alcohol consumption is associated with decreased prevalence of steatohepatitis in patients with non-alcoholic fatty liver disease (NAFLD) J Hepatol. 2012;57:384–391. doi: 10.1016/j.jhep.2012.03.024. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Rodriguez B, Torres DM, Harrison SA. Physical activity: an essential component of lifestyle modification in NAFLD. Nat Rev Gastroenterol Hepatol. 2012;9:726–731. doi: 10.1038/nrgastro.2012.200. [DOI] [PubMed] [Google Scholar]
- 10.Ryan MC, Itsiopoulos C, Thodis T, Ward G, Trost N, Hofferberth S, O’Dea K, et al. The Mediterranean Diet Improves Hepatic Steatosis and Insulin Sensitivity in Individuals with Nonalcoholic Fatty Liver Disease. J Hepatol. 2013 doi: 10.1016/j.jhep.2013.02.012. [DOI] [PubMed] [Google Scholar]
- 11.Vos MB, Lavine JE. Dietary fructose in nonalcoholic fatty liver disease. Hepatology. 2013 doi: 10.1002/hep.26299. [DOI] [PubMed] [Google Scholar]
- 12.Abdelmalek MF, Suzuki A, Guy C, Unalp-Arida A, Colvin R, Johnson RJ, Diehl AM. Increased fructose consumption is associated with fibrosis severity in patients with nonalcoholic fatty liver disease. Hepatology. 2010;51:1961–1971. doi: 10.1002/hep.23535. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Storey ML, Forshee RA, Anderson PA. Beverage consumption in the US population. J Am Diet Assoc. 2006;106:1992–2000. doi: 10.1016/j.jada.2006.09.009. [DOI] [PubMed] [Google Scholar]
- 14.Greenberg JA, Axen KV, Schnoll R, Boozer CN. Coffee, tea and diabetes: the role of weight loss and caffeine. Int J Obes (Lond) 2005;29:1121–1129. doi: 10.1038/sj.ijo.0802999. [DOI] [PubMed] [Google Scholar]
- 15.van Dam RM, Feskens EJ. Coffee consumption and risk of type 2 diabetes mellitus. Lancet. 2002;360:1477–1478. doi: 10.1016/S0140-6736(02)11436-X. [DOI] [PubMed] [Google Scholar]
- 16.Salazar-Martinez E, Willett WC, Ascherio A, Manson JE, Leitzmann MF, Stampfer MJ, Hu FB. Coffee consumption and risk for type 2 diabetes mellitus. Ann Intern Med. 2004;140:1–8. doi: 10.7326/0003-4819-140-1-200401060-00005. [DOI] [PubMed] [Google Scholar]
- 17.van Dam RM, Dekker JM, Nijpels G, Stehouwer CD, Bouter LM, Heine RJ. Coffee consumption and incidence of impaired fasting glucose, impaired glucose tolerance, and type 2 diabetes: the Hoorn Study. Diabetologia. 2004;47:2152–2159. doi: 10.1007/s00125-004-1573-6. [DOI] [PubMed] [Google Scholar]
- 18.Paynter NP, Yeh HC, Voutilainen S, Schmidt MI, Heiss G, Folsom AR, Brancati FL, et al. Coffee and sweetened beverage consumption and the risk of type 2 diabetes mellitus: the atherosclerosis risk in communities study. Am J Epidemiol. 2006;164:1075–1084. doi: 10.1093/aje/kwj323. [DOI] [PubMed] [Google Scholar]
- 19.Pereira MA, Parker ED, Folsom AR. Coffee consumption and risk of type 2 diabetes mellitus: an 11-year prospective study of 28 812 postmenopausal women. Arch Intern Med. 2006;166:1311–1316. doi: 10.1001/archinte.166.12.1311. [DOI] [PubMed] [Google Scholar]
- 20.Lopez-Garcia E, van Dam RM, Li TY, Rodriguez-Artalejo F, Hu FB. The relationship of coffee consumption with mortality. Ann Intern Med. 2008;148:904–914. doi: 10.7326/0003-4819-148-12-200806170-00003. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Ruhl CE, Everhart JE. Coffee and tea consumption are associated with a lower incidence of chronic liver disease in the United States. Gastroenterology. 2005;129:1928–1936. doi: 10.1053/j.gastro.2005.08.056. [DOI] [PubMed] [Google Scholar]
- 22.Ruhl CE, Everhart JE. Coffee and caffeine consumption reduce the risk of elevated serum alanine aminotransferase activity in the United States. Gastroenterology. 2005;128:24–32. doi: 10.1053/j.gastro.2004.09.075. [DOI] [PubMed] [Google Scholar]
- 23.Larsson SC, Wolk A. Coffee consumption and risk of liver cancer: a meta-analysis. Gastroenterology. 2007;132:1740–1745. doi: 10.1053/j.gastro.2007.03.044. [DOI] [PubMed] [Google Scholar]
- 24.Bravi F, Bosetti C, Tavani A, Bagnardi V, Gallus S, Negri E, Franceschi S, et al. Coffee drinking and hepatocellular carcinoma risk: a meta-analysis. Hepatology. 2007;46:430–435. doi: 10.1002/hep.21708. [DOI] [PubMed] [Google Scholar]
- 25.Gallus S, Bertuzzi M, Tavani A, Bosetti C, Negri E, La Vecchia C, Lagiou P, et al. Does coffee protect against hepatocellular carcinoma? Br J Cancer. 2002;87:956–959. doi: 10.1038/sj.bjc.6600582. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Wilson KM, Kasperzyk JL, Rider JR, Kenfield S, van Dam RM, Stampfer MJ, Giovannucci E, et al. Coffee consumption and prostate cancer risk and progression in the Health Professionals Follow-up Study. J Natl Cancer Inst. 2011;103:876–884. doi: 10.1093/jnci/djr151. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Liu R, Guo X, Park Y, Huang X, Sinha R, Freedman ND, Hollenbeck AR, et al. Caffeine intake, smoking, and risk of Parkinson disease in men and women. Am J Epidemiol. 2012;175:1200–1207. doi: 10.1093/aje/kwr451. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Molloy JW, Calcagno CJ, Williams CD, Jones FJ, Torres DM, Harrison SA. Association of coffee and caffeine consumption with fatty liver disease, nonalcoholic steatohepatitis, and degree of hepatic fibrosis. Hepatology. 2012;55:429–436. doi: 10.1002/hep.24731. [DOI] [PubMed] [Google Scholar]
- 29.Anty R, Marjoux S, Iannelli A, Patouraux S, Schneck AS, Bonnafous S, Gire C, et al. Regular coffee but not espresso drinking is protective against fibrosis in a cohort mainly composed of morbidly obese European women with NAFLD undergoing bariatric surgery. J Hepatol. 2012;57:1090–1096. doi: 10.1016/j.jhep.2012.07.014. [DOI] [PubMed] [Google Scholar]
- 30.Birerdinc A, Stepanova M, Pawloski L, Younossi ZM. Caffeine is protective in patients with non-alcoholic fatty liver disease. Aliment Pharmacol Ther. 2012;35:76–82. doi: 10.1111/j.1365-2036.2011.04916.x. [DOI] [PubMed] [Google Scholar]
- 31.Freedman ND, Park Y, Abnet CC, Hollenbeck AR, Sinha R. Association of coffee drinking with total and cause-specific mortality. N Engl J Med. 2012;366:1891–1904. doi: 10.1056/NEJMoa1112010. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32.Gutierrez-Grobe Y, Chavez-Tapia N, Sanchez-Valle V, Gavilanes-Espinar JG, Ponciano-Rodriguez G, Uribe M, Mendez-Sanchez N. High coffee intake is associated with lower grade nonalcoholic fatty liver disease: the role of peripheral antioxidant activity. Ann Hepatol. 2012;11:350–355. [PubMed] [Google Scholar]
- 33.Catalano D, Martines GF, Tonzuso A, Pirri C, Trovato FM, Trovato GM. Protective role of coffee in non-alcoholic fatty liver disease (NAFLD) Dig Dis Sci. 2010;55:3200–3206. doi: 10.1007/s10620-010-1143-3. [DOI] [PubMed] [Google Scholar]
- 34.Neuschwander-Tetri BA, Clark JM, Bass NM, Van Natta ML, Unalp-Arida A, Tonascia J, Zein CO, et al. Clinical, laboratory and histological associations in adults with nonalcoholic fatty liver disease. Hepatology. 2010;52:913–924. doi: 10.1002/hep.23784. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35.Nonalcoholic steatohepatitis clinical research network. Hepatology. 2003;37:244. doi: 10.1002/hep.510370203. [DOI] [PubMed] [Google Scholar]
- 36.Chalasani NP, Sanyal AJ, Kowdley KV, Robuck PR, Hoofnagle J, Kleiner DE, Unalp A, et al. Pioglitazone versus vitamin E versus placebo for the treatment of non-diabetic patients with non-alcoholic steatohepatitis: PIVENS trial design. Contemp Clin Trials. 2009;30:88–96. doi: 10.1016/j.cct.2008.09.003. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 37.Block G, Subar AF. Estimates of nutrient intake from a food frequency questionnaire: the 1987 National Health Interview Survey. J Am Diet Assoc. 1992;92:969–977. [PubMed] [Google Scholar]
- 38.Kleiner DE, Brunt EM, Van Natta M, Behling C, Contos MJ, Cummings OW, Ferrell LD, et al. Design and validation of a histological scoring system for nonalcoholic fatty liver disease. Hepatology. 2005;41:1313–1321. doi: 10.1002/hep.20701. [DOI] [PubMed] [Google Scholar]
- 39.Kistler KD, Brunt EM, Clark JM, Diehl AM, Sallis JF, Schwimmer JB. Physical activity recommendations, exercise intensity, and histological severity of nonalcoholic fatty liver disease. Am J Gastroenterol. 2011;106:460–468. doi: 10.1038/ajg.2010.488. quiz 469. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 40.Ainsworth BE, Haskell WL, Whitt MC, Irwin ML, Swartz AM, Strath SJ, O’Brien WL, et al. Compendium of physical activities: an update of activity codes and MET intensities. Med Sci Sports Exerc. 2000;32:S498–504. doi: 10.1097/00005768-200009001-00009. [DOI] [PubMed] [Google Scholar]
- 41.Centers for Disease Control and Prevention. [Accessed on January 7, 2013];National Health and Nutrition Examination Survey. Available at: http://www.cdc.gov/nchs/nhanes.htm.
- 42.Loomba R, Abraham M, Unalp A, Wilson L, Lavine J, Doo E, Bass NM. Association between diabetes, family history of diabetes, and risk of nonalcoholic steatohepatitis and fibrosis. Hepatology. 2012;56:943–951. doi: 10.1002/hep.25772. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 43.Rockey DC, Caldwell SH, Goodman ZD, Nelson RC, Smith AD. Liver biopsy. Hepatology. 2009;49:1017–1044. doi: 10.1002/hep.22742. [DOI] [PubMed] [Google Scholar]
- 44.Tunnicliffe JM, Shearer J. Coffee, glucose homeostasis, and insulin resistance: physiological mechanisms and mediators. Appl Physiol Nutr Metab. 2008;33:1290–1300. doi: 10.1139/H08-123. [DOI] [PubMed] [Google Scholar]
- 45.Willett WC. Nutritional Epidemiology. New York: Oxford University Press; 1998. [Google Scholar]
- 46.Salvini S, Hunter DJ, Sampson L, Stampfer MJ, Colditz GA, Rosner B, Willett WC. Food-based validation of a dietary questionnaire: the effects of week-to-week variation in food consumption. Int J Epidemiol. 1989;18:858–867. doi: 10.1093/ije/18.4.858. [DOI] [PubMed] [Google Scholar]
- 47.Feskanich D, Rimm EB, Giovannucci EL, Colditz GA, Stampfer MJ, Litin LB, Willett WC. Reproducibility and validity of food intake measurements from a semiquantitative food frequency questionnaire. J Am Diet Assoc. 1993;93:790–796. doi: 10.1016/0002-8223(93)91754-e. [DOI] [PubMed] [Google Scholar]