ABSTRACT
Background
Diet plays a key role in the pathogenesis of nonalcoholic fatty liver disease. Limited data exist regarding specific nutrients and food groups and liver fat continuously, particularly among different ethnicities.
Objectives
We aimed to determine the relationship between usual dietary intake and accurately measured liver fat content in a multiethnic population.
Methods
Participants from the Multiethnic Cohort were recruited into the cross-sectional Adiposity Phenotype Study including women and men aged 60–77 y and 5 race/ethnic groups (African American, Japanese American, Latino, Native Hawaiian, and white). They filled out a detailed FFQ and underwent abdominal MRI for liver fat quantification and whole-body DXA for total adiposity. Intake of a priori–selected dietary factors (total and macronutrient energy, specific micronutrients, and food groups) was analyzed in relation to liver fat by estimating the mean percentage liver fat for quartiles of each dietary factor in a general linear model that adjusted for age, sex, race/ethnicity, percentage body fat, and daily energy intake (kcal/d).
Results
In total, 1682 participants (mean age: 69.2 y; 51% female) were included. Mean ± SD liver fat percentage was 5.7 ± 4.6. A significant positive association with liver fat was found across quartiles of percentage energy from fat, saturated fat, cholesterol, total red meat, red meat excluding processed red meat, and coffee (Bonferroni-adjusted P-trend < 0.05). A significant inverse association was observed for dietary fiber, vitamin C, and vitamin E (Bonferroni-adjusted P-trend < 0.05).
Conclusions
This study of ethnically diverse older adults shows that certain dietary factors, in particular red meat and saturated fat from red meat, were strongly associated with liver fat, whereas dietary fiber was inversely associated with liver fat, replicating some of the previous studies conducted mostly in whites.
Keywords: liver, liver fat, nonalcoholic fatty liver disease, diet, nutrients
Introduction
Nonalcoholic fatty liver disease (NAFLD; >5% hepatic steatosis in the absence of excessive alcohol consumption or other known etiologies such as viral hepatitis) is the leading cause of liver disease worldwide, with a global prevalence of 25% (1, 2). In obese individuals, this prevalence increases to 70% (3). In adults, 40% of individuals with NAFLD will go on to develop progressive liver fibrosis (2). Other complications of NAFLD include nonalcoholic steatohepatitis (NASH; >5% hepatic steatosis with hepatocellular injury), cirrhosis, and hepatocellular carcinoma. NAFLD is now the leading cause of hepatocellular carcinoma, surpassing viral hepatitis and alcohol use (4).
Current treatment recommendations for NAFLD include lifestyle modifications, namely improving nutrition and increasing physical activity (5). There are currently no FDA-approved drugs available for NAFLD treatment. At present, no guidelines exist in the United States regarding specific dietary therapies for NAFLD, although many cross-sectional, longitudinal, and interventional studies have investigated this question (6). Because NAFLD is a reversible process and with the understanding that nutrition plays a key role in NAFLD pathophysiology, increased research on prevention strategies is warranted. Identifying dietary components that are associated with high and low liver fatness may play a key part in understanding NAFLD development. This in turn may aid in devising early interventions aimed at providing nutritional guidance on preventing NAFLD progression.
Studies have identified a variety of dietary factors associated with liver fatness and NAFLD. Large cross-sectional studies have found higher intakes of dietary fiber, vegetables, and certain micronutrients to be inversely associated with NAFLD, whereas higher intakes of animal protein, red meat, soft drinks, and total fat, and greater resemblance to the Western diet, have been positively associated with NAFLD (7–10). In a meta-analysis of diet and likelihood of NAFLD in observational studies using radiographic imaging or liver biopsy, there was a positive association of liver fatness with red meat and soft drinks, whereas there was an inverse association with nut consumption (11).
One limitation of existing studies is the lack of racial/ethnic heterogeneity in their study populations. Many existing cross-sectional studies focus on diet and NAFLD in a homogeneous population owing to the study country's demographics. This approach does not account for the fact that diet is dependent on different ethnic and racial groups’ dietary practices and customs. In the United States, Hispanics have the highest prevalence of NAFLD, whereas African Americans have the lowest prevalence (12). Lim et al. (13) in a study of a large multiethnic cohort which we examine in this present study found liver fat adjusted for body mass to be highest in Latinos and lowest in African Americans. Large-scale studies that assess nutrient and food intake among multiethnic populations can add key information to existing literature and improve the generalizability of findings.
The Multiethnic Cohort (MEC) study is an ongoing prospective study of diet, lifestyle, and genetic risk factors for cancer and other chronic diseases that started in the 1990s (14). The Adiposity Phenotype Study (APS) evaluated a subset of the MEC by reassessing diet and obtaining abdominal MRI for quantification of liver fat (13). Dietary patterns and certain dietary factors have been associated with Medicare claims–defined NAFLD and NAFLD-related cirrhosis in the MEC (7); however, the association of nutrients and foods with liver fat across the range of direct measurement in the population has not been evaluated in this multiethnic cohort. In this study, a priori selection was used to analyze nutrients and foods which have demonstrated association with liver fat fairly consistently in other studies, mostly in whites (Supplemental Table 1). Here, we evaluate diet in terms of nutrient and food group intake in relation to liver fat with the goal of improving our understanding of diet's role in the prevention of NAFLD.
Methods
The MEC included >215,000 men and women aged 45–75 y at baseline (1993–1996). These individuals were recruited in Hawaii and Los Angeles, California and were largely of Japanese-American, Native-Hawaiian, white, African-American, and Latino ancestry. From 2013 to 2016, the APS re-recruited a subset of MEC members who were 60–77 y of age, stratified into 6 BMI categories [reported BMI (in kg/m2): 18.5–21.9, 22–24.9, 25–26.9, 27–29.9, 30–34.9, and 35–40] within each sex and race/ethnicity to obtain a balanced sample across ethnic groups (13). Invitations were initially mailed, then followed by telephone calls to assess eligibility. Exclusions included reported BMI <18.5 and >40, smoking in the past 2 y, body implants or amputations, potentially confounding medical conditions (viral hepatitis, diabetes receiving insulin, thyroid conditions receiving medications), heavy drinkers given the possibility that alcoholic liver disease may coexist in individuals with metabolic syndrome (alcohol intake > 60 g/d for men and >40 g/d for women), and claustrophobia. For study participants who underwent weight change >9 kg over the past 6 months or treatments that could affect adiposity or their gut microbiome (antibiotics, corticosteroids, weight-loss drugs, androgen/estrogen blockers, chemotherapy, colonoscopy, radiation of the abdomen/pelvis), eligibility was deferred for 6 months and reconsidered at that time.
Of 12,602 persons contacted, 1861 participants (23% participation rate among all eligible persons) underwent a clinic visit, which involved anthropometric measurements, fasting blood sample collection, whole-body DXA for body composition, abdominal MRI for liver fat quantification, and study questionnaires including an FFQ (13).
Dietary assessment
The self-administered survey at the APS clinic visit included a 20-page FFQ with >180 food items, as well as questions on demographics, medical conditions, anthropometric measures, physical activity, and lifestyle (15, 16). The FFQ was specifically developed and validated for the diverse ethnic populations in the MEC (16). In a calibration study in the MEC, individuals within sex and ethnic groups were randomly selected to perform three 24-h food recalls, revealing relatively high correlation between the FFQ and the 24-h recalls (16).
For this study, wea priori selected nutrients, foods, and beverages in the literature found to be associated with NAFLD (Supplemental Table 1). We assessed these foods in relation to percentage liver fat across a continuum to help identify potential quantitative relations across the entire range of liver fat and dietary intake. We searched the PubMed database with a focus on large cross-sectional studies; reviewed meta-analyses, randomized controlled studies, and reviews; and included studies that identified dietary components statistically significantly associated with NAFLD. These studies were conducted mostly in whites (Supplemental Table 1).
Imaging
Abdominal MRI scans were used to obtain liver fat measurements in percentage volume (13). Liver fat content was quantified from a series of axial triple gradient-echo Dixon scans on 3-Tesla MRI scanners (Siemens TIM Trio, General Electric HDx) by averaging the estimates from duplicate scans for each of the 2–3 regions of interest in the lateral right lobe of the liver. Liver fat content was calculated using the method by Guiu et al. (17), without involvement of the spleen signal intensity. MRI measures were calibration-adjusted for minimal differences between the scanners at the 2 study sites based on 15 healthy volunteers who were scanned at both sites.
Statistical analysis
For the current analysis of diet in relation to liver fat, further exclusions were made as follows: DXA scans yielding invalid estimates (previously unreported joint replacements and other implants; n = 21), invalid MRI scans to measure visceral fat (n = 59) or liver fat (n = 24) mostly due to motion artifacts, and missing information on diet (n = 36) (18). After these exclusions, 1682 participants remained for the analysis.
General linear models were used to estimate multivariable-adjusted mean percentage liver fat for quartiles of a priori–selected dietary intake variables. Fewer categories were used for dietary variables with limited numbers of users (e.g., diet soda, green tea, decaffeinated coffee). Multivariable models were adjusted for age (continuous), sex, race/ethnicity (whites as reference), total percentage body fat from DXA (continuous), and total energy intake (kcal/d; continuous). These covariates were selected a priori. Liver fat was log transformed to meet model assumptions. The trend test (P-trend) was calculated using a variable that was assigned the median value for each quantile. A 2-tailed P < 0.05 was considered statistically significant. To account for multiple testing, we used Bonferroni correction (i.e., calculation of the critical value = 0.05/number of hypothesis tests). Analyses were conducted using R version 3.6.2 (R Foundation for Statistical Computing, Vienna, Austria).
Results
Participant characteristics
Table 1 describes the main characteristics of the study participants by quartile of liver fat. Across quartiles, age at clinic visit (mean: 69.2 y among all study participants) and the distribution of men and women (roughly even distribution) were similar. Participants with higher liver fat percentage had higher BMI and total fat based on DXA. In terms of ethnicity, Latinos and Japanese Americans were noted to have higher liver fat percentages, whereas whites and African Americans had lower liver fat percentages.
TABLE 1.
Characteristics of participants according to percentage liver fat quartiles in the Multiethnic Cohort Adiposity Phenotype Study1
| Liver fat quartiles2 | |||||
|---|---|---|---|---|---|
| All (n = 1682) | Q1 (n = 420) | Q2 (n = 421) | Q3 (n = 421) | Q4 (n = 420) | |
| Mean liver fat based on MRI, % | 5.7 ± 4.6 | 2.1 ± 0.3 | 3.2 ± 0.4 | 5.1 ± 0.8 | 12.1 ± 4.6 |
| Age at clinic visit, y | 69.2 ± 2.7 | 69.2 ± 2.9 | 69.4 ± 2.7 | 69.1 ± 2.6 | 68.9 ± 2.6 |
| Sex | |||||
| Male | 48.7 | 43.1 | 55.1 | 48.9 | 47.6 |
| Female | 51.3 | 56.9 | 44.9 | 51.1 | 52.4 |
| Ethnicity | |||||
| White | 22.3 | 32.9 | 23.8 | 17.6 | 15.0 |
| African American | 16.1 | 21.2 | 20.7 | 16.9 | 5.5 |
| Native Hawaiian | 15.7 | 16.9 | 15.0 | 13.5 | 17.4 |
| Japanese American | 25.0 | 20.0 | 22.1 | 22.8 | 35.0 |
| Latino | 21.0 | 9.0 | 18.5 | 29.2 | 27.1 |
| Maximum education attained,3 y | 14.8 ± 2.8 | 15.5 ± 2.5 | 15.0 ± 2.7 | 14.3 ± 2.9 | 14.5 ± 2.8 |
| Smoking | |||||
| Never smoker | 61.3 | 63.3 | 60.6 | 63.7 | 57.6 |
| Former smoker | 38.7 | 36.7 | 39.4 | 36.3 | 42.4 |
| BMI, kg/m2 | 27.8 ± 4.7 | 24.9 ± 4.1 | 27.1 ± 4.4 | 29.0 ± 4.4 | 30.3 ± 4.2 |
| Total fat based on DXA, % | 33.5 ± 7.8 | 31.3 ± 8.0 | 32.7 ± 8.2 | 35.1 ± 7.4 | 35.0 ± 6.8 |
| Alcohol intake, g/d | 6.1 ± 10.1 | 5.9 ± 9.5 | 7.3 ± 10.7 | 5.7 ± 9.5 | 5.4 ± 10.7 |
| Physical activity,3 METs/d | 1.7 ± 0.3 | 1.7 ± 0.3 | 1.7 ± 0.3 | 1.7 ± 0.3 | 1.6 ± 0.3 |
Values are means ± SDs for continuous variables or percentages for categorical variables. MET, metabolic equivalent.
Cutoffs for liver fat quartiles, median (range): Q1: 2.2 (0.95–2.62); Q2: 3.2 (2.63–3.89); Q3: 5.1 (3.90–6.89); and Q4: 10.5 (6.90–30.3).
Minimal missing data include 3% for physical activity and 0.7% for maximal education attained.
Association of energy and nutrient intakes with liver fat
Table 2 shows the adjusted geometric mean percentage liver fat values for quartiles of energy and nutrient intake with the P-trend to indicate the strength of a linear dose–response association across quartiles. After accounting for multiple testing, significant positive associations were found for percentage energy from total fat (Bonferroni-adjusted P-trend = 0.00028), saturated fat (P-trend = 0.00012), and cholesterol (P-trend = 0.00016). Significant inverse associations were observed for intakes of dietary fiber (P-trend = 1.10E-05), vitamin C (P-trend = 0.0013), and vitamin E (P-trend = 0.00088).
TABLE 2.
Associations of energy and nutrient intakes with liver fat in the Multiethnic Cohort Adiposity Phenotype Study1
| Variable | Median (range) | Participants, n | Percentage liver fat, geometric mean (95% CI) |
|---|---|---|---|
| Energy, kcal/d | Q1: 1002 (508–1264) | 420 | 4.24 (4.00, 4.50) |
| Q2: 1485 (1264–1675) | 421 | 4.23 (3.99, 4.49) | |
| Q3: 1904 (1677–2212) | 421 | 4.26 (4.02, 4.52) | |
| Q4: 2730 (2213–8804) | 420 | 4.74 (4.47, 5.02) | |
| P-trend | 0.0052 | ||
| Carbohydrate, g/d | Q1: 117 (45–152) | 420 | 4.51 (4.19, 4.85) |
| Q2: 177 (152–203) | 421 | 4.20 (3.95, 4.46) | |
| Q3: 233 (203–273) | 421 | 4.41 (4.16, 4.67) | |
| Q4: 339 (273–1114) | 420 | 4.34 (4.00, 4.72) | |
| P-trend | 0.78 | ||
| Carbohydrate, % of energy | Q1: 39.3 (17–43) | 420 | 4.66 (4.39, 4.94) |
| Q2: 45.2 (43–48) | 421 | 4.41 (4.17, 4.68) | |
| Q3: 50 (48–53) | 421 | 4.24 (4.00, 4.49) | |
| Q4: 57.1 (53–75) | 420 | 4.16 (3.92, 4.41) | |
| P-trend | 0.0059 | ||
| Dietary fiber, g/d | Q1: 10.7 (3.1–14.3) | 420 | 4.78 (4.47, 5.11) |
| Q2: 17.3 (14.3–20.5) | 421 | 4.69 (4.42, 4.98) | |
| Q3: 24.0 (20.5–28.9) | 421 | 4.26 (4.02, 4.51) | |
| Q4: 36.3 (28.9–120) | 420 | 3.79 (3.53, 4.07) | |
| P-trend | 1.10E-05* | ||
| Sucrose, g/d | Q1: 14 (4–19) | 420 | 4.42 (4.13, 4.72) |
| Q2: 24 (19–29) | 421 | 4.28 (4.03, 4.54) | |
| Q3: 34 (29–41) | 421 | 4.42 (4.17, 4.69) | |
| Q4: 53 (41–183) | 420 | 4.33 (4.04, 4.65) | |
| P-trend | 0.87 | ||
| Fructose, g/d | Q1: 8 (1–11) | 420 | 4.58 (4.30, 4.88) |
| Q2: 14 (11–18) | 421 | 4.41 (4.16, 4.67) | |
| Q3: 21 (18–27) | 421 | 4.46 (4.21, 4.73) | |
| Q4: 34 (27–144) | 420 | 4.03 (3.78, 4.30) | |
| P-trend | 0.012 | ||
| Protein, g/d | Q1: 39 (16–51) | 420 | 4.26 (3.95, 4.59) |
| Q2: 59 (51–68) | 421 | 4.41 (4.15, 4.69) | |
| Q3: 78 (68–90) | 421 | 4.34 (4.09, 4.60) | |
| Q4: 114 (90–383) | 420 | 4.45 (4.09, 4.85) | |
| P-trend | 0.60 | ||
| Protein, % of energy | Q1: 13.2 (8.3–14.3) | 420 | 4.31 (4.07, 4.57) |
| Q2: 15.2 (14.3–15.9) | 421 | 4.24 (4.00, 4.49) | |
| Q3: 16.6 (15.9–17.6) | 421 | 4.45 (4.20, 4.71) | |
| Q4: 18.8 (17.6–34.1) | 420 | 4.46 (4.21, 4.73) | |
| P-trend | 0.27 | ||
| Fat, g/d | Q1: 36 (11–46) | 420 | 4.03 (3.75, 4.33) |
| Q2: 54 (46–63) | 421 | 4.13 (3.89, 4.39) | |
| Q3: 73 (63–88) | 421 | 4.42 (4.18, 4.69) | |
| Q4: 110 (88–398) | 420 | 4.92 (4.53, 5.34) | |
| P-trend | 0.0015 | ||
| Fat, % of energy | Q1: 26.0 (9.3–29.4) | 420 | 4.09 (3.86, 4.34) |
| Q2: 31.9 (29.4–33.9) | 421 | 4.17 (3.93, 4.41) | |
| Q3: 35.9 (33.9–38.1) | 421 | 4.50 (4.25, 4.77) | |
| Q4: 40.9 (38.1–59.6) | 420 | 4.71 (4.44, 4.99) | |
| P-trend | 0.00028* | ||
| Saturated fat, g/d | Q1: 11 (3–13) | 420 | 3.83 (3.57, 4.11) |
| Q2: 16 (13–19) | 421 | 4.33 (4.08, 4.60) | |
| Q3: 22 (19–27) | 421 | 4.39 (4.14, 4.65) | |
| Q4: 35 (27–118) | 420 | 4.97 (4.60, 5.38) | |
| P-trend | 0.00012* | ||
| Monounsaturated fat, g/d | Q1: 14 (3–17) | 420 | 4.04 (3.76, 4.34) |
| Q2: 21 (17–24) | 421 | 4.17 (3.92, 4.43) | |
| Q3: 29 (24–34) | 421 | 4.46 (4.21, 4.73) | |
| Q4: 43 (34–152) | 420 | 4.82 (4.45, 5.23) | |
| P-trend | 0.0042 | ||
| Polyunsaturated fat, g/d | Q1: 8 (2–10) | 421 | 4.50 (4.19, 4.83) |
| Q2: 12 (10–14) | 420 | 4.26 (4.01, 4.52) | |
| Q3: 17 (14–20) | 421 | 4.43 (4.18, 4.69) | |
| Q4: 25 (20–95) | 420 | 4.27 (3.95, 4.62) | |
| P-trend | 0.59 | ||
| n–3 fatty acids, g/d | Q1: 0.8 (0.2–1.1) | 420 | 4.15 (3.87, 4.45) |
| Q2: 1.3 (1.1–1.5) | 421 | 4.30 (4.05, 4.57) | |
| Q3: 1.7 (1.5–2.1) | 421 | 4.37 (4.12, 4.63) | |
| Q4: 2.6 (2.1–11.1) | 420 | 4.66 (4.31, 5.03) | |
| P-trend | 0.069 | ||
| Cholesterol, mg/d | Q1: 104 (7–136) | 420 | 3.87 (3.63, 4.13) |
| Q2: 171 (136–208) | 421 | 4.35 (4.10, 4.62) | |
| Q3: 247 (208–297) | 421 | 4.48 (4.23, 4.75) | |
| Q4: 389 (298–1528) | 420 | 4.81 (4.49, 5.15) | |
| P-trend | 0.00016* | ||
| Alcohol, g/d | Q1: 0 (0–0) | 421 | 4.47 (4.20, 4.74) |
| Q2: 0 (0–1) | 420 | 4.62 (4.35, 4.91) | |
| Q3: 3 (1–8) | 421 | 4.22 (3.99, 4.48) | |
| Q4: 18 (8–59) | 420 | 4.16 (3.91, 4.42) | |
| P-trend | 0.041 | ||
| Zinc, mg/d | Q1: 6 (3–7) | 420 | 4.42 (4.11, 4.75) |
| Q2: 9 (7–10) | 421 | 4.38 (4.12, 4.65) | |
| Q3: 12 (10–13) | 421 | 4.41 (4.16, 4.68) | |
| Q4: 18 (14–98) | 420 | 4.24 (3.92, 4.59) | |
| P-trend | 0.54 | ||
| Vitamin C, mg/d | Q1: 49 (11–74) | 420 | 4.70 (4.41, 5.00) |
| Q2: 95 (74–117) | 421 | 4.41 (4.16, 4.68) | |
| Q3: 147 (117–188) | 421 | 4.38 (4.13, 4.64) | |
| Q4: 252 (188–1188) | 420 | 3.99 (3.75, 4.26) | |
| P-trend | 0.0013* | ||
| Vitamin E, mg/d | Q1: 4.3 (1.4–5.5) | 420 | 4.72 (4.41, 5.05) |
| Q2: 6.7 (5.5–8.0) | 421 | 4.40 (4.14, 4.66) | |
| Q3: 9.5 (8.0–11.2) | 421 | 4.57 (4.31, 4.84) | |
| Q4: 14.3 (11.2–114) | 420 | 3.83 (3.56, 4.12) | |
| P-trend | 0.00088* | ||
| Iron, mg/d | Q1: 7.5 (3.3–9.5) | 420 | 4.64 (4.33, 4.98) |
| Q2: 11.4 (9.5–13.4) | 421 | 4.42 (4.16, 4.69) | |
| Q3: 15.8 (13.4–18.7) | 421 | 4.50 (4.25, 4.77) | |
| Q4: 23.6 (18.7–145) | 420 | 3.92 (3.64, 4.22) | |
| P-trend | 0.0059 |
The P-trend to assess a dose–response relation was obtained in a general linear model of liver fat on each dietary variable with 4 values, i.e., medians for the quartiles, and other covariates (age, sex, race/ethnicity, percentage body fat, and total energy intake for nonenergy diet variables). Q, quartile.
*Dietary factor met the threshold for significance using Bonferroni correction (P-trend < 0.00135).
Association of foods with liver fat
Table 3 depicts the association between adjusted geometric mean percentage liver fat and intake of specific foods or food groups. Significant positive associations after multiple test correction were detected for total red meat (P-trend = 2.38E-09), red meat excluding processed meat (P-trend = 2.71E-06), and coffee (P-trend = 0.00081). Associations for other items such as diet soda, tea, all fruits plus juice, and sugar-sweetened beverages (SSBs) did not reach the significance level accounting for multiple testing.
TABLE 3.
Associations of food items with liver fat in the Multiethnic Cohort Adiposity Phenotype Study1
| Variable | Median (range) | Participants, n | Percentage liver fat, geometric mean (95% CI) |
|---|---|---|---|
| Total red meat, g/d | Q1: 12 (0–20) | 420 | 3.78 (3.55, 4.02) |
| Q2: 27 (20–35) | 421 | 4.22 (3.98, 4.47) | |
| Q3: 44 (35–58) | 421 | 4.43 (4.18, 4.69) | |
| Q4: 84 (59–408) | 420 | 5.13 (4.80, 5.49) | |
| P-trend | 2.38E-09* | ||
| Red meat excluding processed meat, g/d | Q1: 7 (0–12) | 420 | 3.93 (3.70, 4.19) |
| Q2: 16 (12–21) | 421 | 4.24 (4.00, 4.49) | |
| Q3: 28 (21–38) | 421 | 4.39 (4.14, 4.65) | |
| Q4: 53 (38–198) | 420 | 4.97 (4.65, 5.30) | |
| P-trend | 2.71E-06* | ||
| All fruits plus juice, g/d | Q1: 68 (0–117) | 420 | 4.63 (4.36, 4.93) |
| Q2: 167 (117–222) | 421 | 4.43 (4.18, 4.70) | |
| Q3: 287 (222–376) | 421 | 4.38 (4.13, 4.64) | |
| Q4: 535 (377–2507) | 420 | 4.03 (3.78, 4.29) | |
| P-trend | 0.0032 | ||
| Total vegetables, g/d | Q1: 147 (28–204) | 420 | 4.52 (4.24, 4.82) |
| Q2: 258 (204–310) | 421 | 4.34 (4.09, 4.60) | |
| Q3: 373 (310–464) | 421 | 4.46 (4.21, 4.73) | |
| Q4: 626 (464–2936) | 420 | 4.14 (3.88, 4.43) | |
| P-trend | 0.12 | ||
| Nuts excluding coconut, g/d | Q1: 0.2 (0.0–0.4) | 420 | 4.16 (3.91, 4.41) |
| Q2: 0.9 (0.4–1.6) | 421 | 4.45 (4.19, 4.71) | |
| Q3: 2.8 (1.6–4.3) | 421 | 4.35 (4.10, 4.61) | |
| Q4: 7.3 (4.3–80.2) | 420 | 4.51 (4.24, 4.80) | |
| P-trend | 0.20 | ||
| Sugar-sweetened beverages, g/d | Q1: 0 (0–0) | 439 | 4.36 (4.12, 4.62) |
| Q2: 4 (0–7) | 419 | 4.45 (4.19, 4.71) | |
| Q3: 16 (7–32) | 424 | 4.39 (4.14, 4.65) | |
| Q4: 68 (32–467) | 400 | 4.26 (4.01, 4.52) | |
| P-trend | 0.38 | ||
| Diet soda, g/d | Q1: 0 (0–0) | 1145 | 4.24 (4.09, 4.39) |
| Q2: 12 (6–12) | 112 | 4.81 (4.30, 5.38) | |
| Q3: 128 (15–2160) | 425 | 4.60 (4.34, 4.89) | |
| P-trend | 0.034 | ||
| Total tea, g | Q1: 0 (0–0) | 426 | 4.10 (3.87, 4.34) |
| Q2: 84 (7.8–205) | 358 | 4.65 (4.37, 4.96) | |
| Q3: 237 (237–240) | 478 | 4.57 (4.32, 4.82) | |
| Q4: 591 (244–1547) | 420 | 4.18 (3.94, 4.44) | |
| P-trend | 0.54 | ||
| Black tea, g/d | Q1: 0 (0–0) | 520 | 4.14 (3.93, 4.36) |
| Q2: 84 (7.8–205) | 325 | 4.64 (4.34, 4.95) | |
| Q3: 237 (237–237) | 524 | 4.59 (4.35, 4.83) | |
| Q4: 591 (296–947) | 313 | 4.12 (3.85, 4.41) | |
| P-trend | 0.56 | ||
| Green tea, g/d | Q1: 0 (0–0) | 1351 | 4.40 (4.26, 4.55) |
| Q2: 85.3 (7.9–961) | 331 | 4.22 (3.95, 4.51) | |
| P-trend | NA** | ||
| Total coffee, g/d | Q1: 0 (0–0) | 538 | 4.03 (3.82, 4.24) |
| Q2: 7.6 (3.5–16.4) | 266 | 4.30 (3.99, 4.62) | |
| Q3: 30.4 (17.7–76.3) | 474 | 4.50 (4.26, 4.75) | |
| Q4: 230 (76.9–2160) | 404 | 4.74 (4.46, 5.03) | |
| P-trend | 0.00081* | ||
| Regular coffee, g/d | Q1: 0 (0–0) | 655 | 4.05 (3.87, 4.25) |
| Q2: 7.1 (3.5–8.8) | 184 | 4.35 (3.99, 4.75) | |
| Q3: 29.5 (9.5–50.9) | 436 | 4.68 (4.42, 4.95) | |
| Q4: 153 (58.0–2160) | 407 | 4.56 (4.30, 4.85) | |
| P-trend | 0.024 | ||
| Decaffeinated coffee, g/d | Q1: 0 (0–0) | 1363 | 4.36 (4.22, 4.50) |
| Q2: 18.9 (7.6–575) | 319 | 4.37 (4.09, 4.68) | |
| P-trend | NA** |
The P-trend to assess a dose–response relation was obtained in a general linear model of liver fat on each dietary variable with 4 values, i.e., medians for the quartiles, and other covariates (age, sex, race/ethnicity, percentage body fat, and total energy intake for nonenergy diet variables). NA, not applicable; Q, quartile.
*Dietary factor met the threshold for significance using Bonferroni correction (P-trend < 0.00135).
**P-trends were not calculated because there are only 2 categories of intake. There was no significant difference in multivariable-adjusted liver fat means across categories of green tea (P = 0.26) or decaffeinated coffee (P = 0.93) intake.
Discussion
In this cross-sectional study, we assessed the associations of intake of energy, nutrients, and select foods with liver fat content as measured by gold-standard MRI in a large multiethnic adult population. A significant positive association with liver fat was found across quartiles of percentage energy from fat, saturated fat, cholesterol, red meat, and coffee (Bonferroni-adjusted P-trend < 0.05). A significant inverse association was observed for dietary fiber, vitamin C, and vitamin E (Bonferroni-adjusted P-trend < 0.05). This study differs from other studies in that it evaluates diet and liver fat as continuous variables in healthy individuals, rather than focusing on patients diagnosed with NAFLD, a dichotomous variable. Our goal was to identify liver fat–associated dietary components that may prevent or aggravate accumulation of liver fat and NAFLD development in the general population, as opposed to the selected group of patients who come to clinical attention.
Our results were consistent with the findings of Noureddin et al. (7), who investigated the association between diet and NAFLD determined by Medicare claims in the parent MEC cohort. They found positive associations for total red meat, processed red meat, poultry, and cholesterol and an inverse association for fiber. Our study was also consistent with a Dutch study by Rietman et al. (8), where investigators studied healthy adults cross-sectionally for FFQ-based dietary intake and found that individuals with a diagnosis of NAFLD, compared with those without, had higher intakes of animal protein, total fat, soft drinks, and snacks and a lower intake of fiber. However, the association with SSBs was not replicated in our study.
Our findings of a significant positive association of liver fat with total red meat and red meat excluding processed meat agree with those of other cross-sectional studies (10, 19). We had hypothesized that iron also might be associated with increased liver fat given its presence in red meat. Evidence shows iron overload is associated with insulin resistance and steatosis, whereas animal models suggest that iron may accelerate the progression of NAFLD to NASH (20, 21). Our data do not support an association of iron with liver fat. However, we did find that saturated fat (but not monounsaturated or polyunsaturated fats) was statistically significantly associated with increased liver fat. These findings suggest that the strong positive association between higher red meat and increased liver fat seen across many different studies may be driven largely by fat, rather than iron, intake. This is also supported by Rosqvist et al. (22), who found that saturated fats increased liver fat 2-fold more than polyunsaturated fats in an intervention study.
In accordance with previous studies (7, 8), we found a strong inverse association between dietary fiber and liver fat. In addition, previous studies have noted a potentially protective role of antioxidants in NAFLD (23), including vitamins C and E (24). In our study, we also found a significant inverse relation with liver fat across quartiles of dietary intake for vitamin C and vitamin E. Fruit similarly had a trend of inverse association with liver fat, although it did not meet our multiple comparison–corrected significance level, which is consistent with the weak and inconsistent evidence generated in past studies (19, 25, 26).
Somewhat surprisingly, we found that coffee was positively associated with liver fat. In contrast, other observational studies have noted an inverse association between coffee and NAFLD (27, 28). A systematic literature review and meta-analysis by Shen et al. (29) did not find an association between caffeine intake and NAFLD, but there was a significant association between regular coffee intake and decreased hepatic fibrosis in NAFLD patients. Veronese et al. (30), in a population study in South Italy (n = 2819), found a null association between drinking coffee and NAFLD.
We also did not find an association between tea consumption and liver fat. Koch et al. (31) found tea intake to be inversely associated with MRI-assessed liver fat among German adults. In a meta-analysis of randomized clinical trials investigating the association between green tea and tea components and NAFLD, subgroup analyses showed that green tea reduced liver enzymes in patients with NAFLD, although a slight increase in liver enzymes was noted in healthy subjects (32).
Several studies have shown a positive association between SSBs and NAFLD (33–35). Chen et al. (35) in a meta-analysis noted a dose-dependent response, with risk of NAFLD increasing by 14%, 26%, and 53% with low (<1 cup/wk), medium (1–6 cups/wk), and high (≥7 cups/wk) dosages of SSBs, respectively. Basic science and epidemiologic studies support the role of sugar and fructose in the development and progression of NAFLD (36). In our study, SSBs were not associated with percentage liver fat across dietary intake. In our population, intake of SSBs was relatively low (median = 68.3 g/d or one-fifth of a can for the highest quartile), which may explain the lack of association. This low intake was comparable with other studies assessing sugar intake in an elderly population (37, 38).
Our results augment the body of literature available to develop clinical recommendations for dietary modification to prevent NAFLD. Currently, there is limited formal dietary guidance regarding NAFLD. Based on findings to date, recommendations to reduce foods associated with high liver fat (e.g., high-fat foods and red meat in particular) and to increase foods high in dietary fiber warrant follow-up in clinical trials of diet and liver fat.
NAFLD susceptibility varies by race/ethnicity (39), and Lim et al. (13) observed marked ethnic differences in the propensity for NAFLD at a given amount of total adiposity. Thus, we examined the diet–liver fat associations in our multiethnic population considering the diverse diets and lifestyles among ethnic groups. Our study had several strengths in addition to the population-based and multiethnic characteristics of the MEC itself. Firstly, the APS assessed a large, diverse population across a range of BMI. Secondly, the FFQ used is comprehensive, designed for use in this cohort, well-validated, and includes ethnic-specific food groups. Finally, we used MRI, which has higher sensitivity in measuring liver fat content and also allows the assessment of liver fat content as a continuous measurement (39, 40).
There were also limitations to our study that warrant consideration. We understand that studying individual components of diet may not be representative of dietary patterns and the overall quality of diet; however, it affords the opportunity to identify key contributors to diet–liver fat associations and likely mechanisms. The cross-sectional nature of our study design did not allow us to infer temporality or causation. Although our study included a relatively large number of total participants, the sample size was still limited for conducting direct comparisons in analyses stratified by ethnicity.
Given the increasing prevalence of NAFLD and its progression to fibrosis, chronic liver disease, and liver cancer and/or potentially liver transplant, there is a need for prevention before the development of NAFLD, NASH, and fibrosis. In summary, we found strong positive associations between liver fat and intakes of total fat, percentage energy from fat, and saturated fat, with meat intake a likely contributor. We noted a strong inverse association between fiber and liver fat. Whereas prior studies have shown a positive association between intake of simple carbohydrates and NAFLD, this was not seen in our study, possibly because of the lower consumption in our study population. This study contributes to the current understanding by replicating some of the dietary component associations with liver fat in nonwhites and provides an important basis for future prospective and intervention studies.
Supplementary Material
Acknowledgments
The authors’ responsibilities were as follows—TKK and JWL: conceptualized and drafted the manuscript; LLM, LRW, JAS, TE, and UL: carried out the Multiethnic Cohort Adiposity Phenotype study; MAM: performed the biostatistical analysis; MAM, UL, LLM, and CJB: edited the manuscript; and all authors: read and approved the final manuscript.
Notes
The Adiposity Phenotype Study was funded by National Institutes of Health (NIH) National Cancer Institute (NCI) grant P01 CA168530 (to LLM). The Multiethnic Cohort Study was funded by NIH grant U01 CA164973 (to LLM and LRW). The University of Hawaii Cancer Center Shared Resources (Analytical Biochemistry, Biostatistics, and Nutrition Support) were supported in part by NCI grant P30 CA0717890. Recruitment activities at the University of Southern California were supported in part by NIH National Center for Advancing Translational Sciences grant UL1 TR000130 to the Southern California Clinical and Translational Science Institute.
Author disclosures: the authors report no conflicts of interest.
Supplemental Table 1 is available from the “Supplementary data” link in the online posting of the article and from the same link in the online table of contents at https://academic.oup.com/jn.
Abbreviations used: APS, Adiposity Phenotype Study; MEC, Multiethnic Cohort; NAFLD, nonalcoholic fatty liver disease; NASH, nonalcoholic steatohepatitis; SSB, sugar-sweetened beverage.
Contributor Information
Tanyaporn K Kaenkumchorn, Division of Pediatric Gastroenterology, Hepatology, and Nutrition, Department of Pediatrics, Seattle Children's Hospital, Seattle, WA, USA.
Melissa A Merritt, Cancer Epidemiology Program, University of Hawaii Cancer Center, Honolulu, HI, USA.
Unhee Lim, Cancer Epidemiology Program, University of Hawaii Cancer Center, Honolulu, HI, USA.
Loïc Le Marchand, Cancer Epidemiology Program, University of Hawaii Cancer Center, Honolulu, HI, USA.
Carol J Boushey, Cancer Epidemiology Program, University of Hawaii Cancer Center, Honolulu, HI, USA.
John A Shepherd, Cancer Epidemiology Program, University of Hawaii Cancer Center, Honolulu, HI, USA.
Lynne R Wilkens, Cancer Epidemiology Program, University of Hawaii Cancer Center, Honolulu, HI, USA.
Thomas Ernst, Department of Diagnostic Radiology and Nuclear Medicine, School of Medicine, University of Maryland, Baltimore, MD, USA.
Johanna W Lampe, Division of Public Health Sciences, Fred Hutchinson Cancer Research Center, Seattle, WA, USA.
References
- 1.Charlton MR, Burns JM, Pedersen RA, Watt KD, Heimbach JK, Dierkhising RA. Frequency and outcomes of liver transplantation for nonalcoholic steatohepatitis in the United States. Gastroenterology. 2011;141(4):1249–53. [DOI] [PubMed] [Google Scholar]
- 2.Younossi ZM, Koenig AB, Abdelatif D, Fazel Y, Henry L, Wymer M. Global epidemiology of nonalcoholic fatty liver disease—meta-analytic assessment of prevalence, incidence, and outcomes. Hepatology. 2016;64(1):73–84. [DOI] [PubMed] [Google Scholar]
- 3.Grgurevic I, Podrug K, Mikolasevic I, Kukla M, Madir A, Tsochatzis EA. Natural history of nonalcoholic fatty liver disease: implications for clinical practice and an individualized approach. Can J Gastroenterol Hepatol. 2020:9181368. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Makarova-Rusher O, Altekruse SF, McNeel TS, Ulahannan S, Duffy AG, Graubard BI, Greten TF, McGlynn KA. Population attributable fractions of risk factors for hepatocellular carcinoma in the United States. Cancer. 2016;122(11):1757–65. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Berná G, Romero-Gomez M. The role of nutrition in non-alcoholic fatty liver disease: pathophysiology and management. Liver Int. 2020;40(S1):102–8. [DOI] [PubMed] [Google Scholar]
- 6.Fischer K, Pick JA, Moewes D, Nöthlings U. Qualitative aspects of diet affecting visceral and subcutaneous abdominal adipose tissue: a systematic review of observational and controlled intervention studies. Nutr Rev. 2015;73(4):191–215. [DOI] [PubMed] [Google Scholar]
- 7.Noureddin M, Zelber-Sagi S, Wilkens LR, Porcel J, Boushey CJ, Le Marchand L, Rosen HR, Setiawan VW. Diet associations with nonalcoholic fatty liver disease in an ethnically diverse population: the Multiethnic Cohort. Hepatology. 2020;71(6):1940–52. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Rietman A, Sluik D, Feskens EJM, Kok FJ, Mensink M. Associations between dietary factors and markers of NAFLD in a general Dutch adult population. Eur J Clin Nutr. 2018;72(1):117–23. [DOI] [PubMed] [Google Scholar]
- 9.Oddy WH, Herbison CE, Jacoby P, Ambrosini GL, O'Sullivan TA, Ayonrinde OT, Olynyk JK, Black LJ, Beilin JJ, Mori TAet al. The Western dietary pattern is prospectively associated with nonalcoholic fatty liver disease in adolescence. Am J Gastroenterol. 2013;108(5):778–85. [DOI] [PubMed] [Google Scholar]
- 10.Zelber-Sagi S, Ivancovsky-Wajcman D, Fliss Isakov N, Webb M, Orenstein D, Shibolet O, Kariv R. High red and processed meat consumption is associated with non-alcoholic fatty liver disease and insulin resistance. J Hepatol. 2018;68(6):1239–46. [DOI] [PubMed] [Google Scholar]
- 11.He K, Li Y, Guo X, Zhong L, Tang S. Food groups and the likelihood of non-alcoholic fatty liver disease: a systematic review and meta-analysis. Br J Nutr. 2020;124:1–13. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Browning JD, Szczepaniak LS, Dobbins R, Nuremberg P, Horton JD, Cohen JC, Grundy SM, Hobbs HH. Prevalence of hepatic steatosis in an urban population in the United States: impact of ethnicity. Hepatology. 2004;40(6):1387–95. [DOI] [PubMed] [Google Scholar]
- 13.Lim U, Monroe KR, Buchthal S, Fan B, Cheng I, Kristal BS, Lampe JW, Hullar MA, Franke AA, Stram DOet al. Propensity for intra-abdominal and hepatic adiposity varies among ethnic groups. Gastroenterology. 2019;156(4):966–75.e10. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Maskarinec G, Lim U, Jacobs S, Monroe KR, Ernst T, Buchthal SD, Shepherd JA, Wilkens LR, Le Marchard L, Boushey CJ. Diet quality in midadulthood predicts visceral adiposity and liver fatness in older ages: the Multiethnic Cohort study. Obesity (Silver Spring). 2017;25(8):1442–50. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Kolonel LN, Henderson BE, Hankin JH, Nomura AMY, Wilkens LR, Pike MC, Stram DO, Monroe KR, Earle ME, Nagmaine FS. A multiethnic cohort in Hawaii and Los Angeles: baseline characteristics. Am J Epidemiol. 2000;151(4):346–57. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Stram DO, Hankin JH, Wilkens LR, Pike MC, Monroe KR, Park S, Henderson BE, Nomura AM, Earle ME, Nagamine FSet al. Calibration of the dietary questionnaire for a multiethnic cohort in Hawaii and Los Angeles. Am J Epidemiol. 2000;151(4):358–70. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Guiu B, Loffroy R, Petit JM, Aho S, Ben Salem D, Masson D, Hillon P, Cercueil J, Krause D. Mapping of liver fat with triple-echo gradient echo imaging: validation against 3.0-T proton MR spectroscopy. Eur Radiol. 2009;19(7):1786–93. [DOI] [PubMed] [Google Scholar]
- 18.Ahmed A, Wong RJ, Harrison SA. Nonalcoholic fatty liver disease review: diagnosis, treatment, and outcomes. Clin Gastroenterol Hepatol. 2015;13(12):2062–70. [DOI] [PubMed] [Google Scholar]
- 19.Alferink LJ, Kiefte-de Jong JC, Erler NS, Veldt BJ, Schoufour JD, de Knegt RJ, Ikram MA, Metselaar HJ, Janssen HL, Franco OHet al. Association of dietary macronutrient composition and non-alcoholic fatty liver disease in an ageing population: the Rotterdam Study. Gut. 2019;68(6):1088–98. [DOI] [PubMed] [Google Scholar]
- 20.Salaye L, Bychkova I, Sink S, Kovalic AJ, Bharadwaj MS, Lorenzo F, Jain S, Harrison AV, Davis AT, Turnbull Ket al. A low iron diet protects from steatohepatitis in a mouse model. Nutrients. 2019;11(9):2172. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Dongiovanni P, Fracanzani AL, Fargion S, Valenti L. Iron in fatty liver and in the metabolic syndrome: a promising therapeutic target. J Hepatol. 2011;55(4):920–32. [DOI] [PubMed] [Google Scholar]
- 22.Rosqvist F, Iggman D, Kullberg J, Cedernaes J, Johansson H, Larsson A, Johansson L, Ahlström H, Arner P, Dahlman Iet al. Overfeeding polyunsaturated and saturated fat causes distinct effects on liver and visceral fat accumulation in humans. Diabetes. 2014;63(7):2356–68. [DOI] [PubMed] [Google Scholar]
- 23.Musso G, Gambino R, De Michieli F, Cassader M, Rizzetto M, Durazzo M, Fagà E, Silli B, Pagano G. Dietary habits and their relations to insulin resistance and postprandial lipemia in nonalcoholic steatohepatitis. Hepatology. 2003;37(4):909–16. [DOI] [PubMed] [Google Scholar]
- 24.Ivancovsky-Wajcman D, Fliss-Isakov N, Salomone F, Webb M, Shibolet O, Kariv R, Zelber-Sagi S. Dietary vitamin E and C intake is inversely associated with the severity of nonalcoholic fatty liver disease. Dig Liver Dis. 2019;51(12):1698–705. [DOI] [PubMed] [Google Scholar]
- 25.Tajima R, Kimura T, Enomoto A, Saito A, Kobayashi S, Masuda K, Iida K. No association between fruits or vegetables and non-alcoholic fatty liver disease in middle-aged men and women. Nutrition. 2019;61:119–24. [DOI] [PubMed] [Google Scholar]
- 26.Jia Q, Xia Y, Zhang Q, Wu H, Du H, Liu L, Wang C, Shi H, Guo X, Liu Xet al. Dietary patterns are associated with prevalence of fatty liver disease in adults. Eur J Clin Nutr. 2015;69(8):914–21. [DOI] [PubMed] [Google Scholar]
- 27.Birerdinc A, Stepanova M, Pawloski L, Younossi ZM. Caffeine is protective in patients with non-alcoholic fatty liver disease. Aliment Pharmacol Ther. 2012;35(1):76–82. [DOI] [PubMed] [Google Scholar]
- 28.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(11):3200–6. [DOI] [PubMed] [Google Scholar]
- 29.Shen H, Rodriguez AC, Shiani A, Lipka S, Shahzad G, Kumar A, Mustacchia P. Association between caffeine consumption and nonalcoholic fatty liver disease: a systemic review and meta-analysis. Therap Adv Gastoenterol. 2016;9(1):113–20. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Veronese N, Notarnicola M, Cisternino AM, Reddavide R, Inguaggiato R, Guerra V, Rotolo O, Zinzi I, Leandro G, Correale Met al. Coffee intake and liver steatosis: a population study in a Mediterranean area. Nutrients. 2018;10(1):89. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Koch M, Borggrefe J, Barbaresko J, Groth G, Jacobs G, Siegert S, Lieb W, Müller MJ, Bosy-Westphal A, Heller Met al. Dietary patterns associated with magnetic resonance imaging–determined liver fat content in a general population study. Am J Clin Nutr. 2014;99(2):369–77. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32.Mahmoodi M, Hosseini R, Kazemi A, Ofori-Asenso R, Mazidi M, Mazloomi SM. Effects of green tea or green tea catechin on liver enzymes in healthy individuals and people with nonalcoholic fatty liver disease: a systematic review and meta-analysis of randomized clinical trials. Phytother Res. 2020;34(7):1587–98. [DOI] [PubMed] [Google Scholar]
- 33.Abid A, Taha O, Nseir W, Farah R, Grosovski M, Assy N. Soft drink consumption is associated with fatty liver disease independent of metabolic syndrome. J Hepatol. 2009;51(5):918–24. [DOI] [PubMed] [Google Scholar]
- 34.Zelber-Sagi S, Nitzan-Kaluski D, Goldsmith R, Webb M, Blendis L, Halpern Z, Oren R. Long term nutritional intake and the risk for non-alcoholic fatty liver disease (NAFLD): a population based study. J Hepatol. 2007;47(5):711–17. [DOI] [PubMed] [Google Scholar]
- 35.Chen H, Wang J, Li Z, Lam CWK, Xiao Y, Wu Q, Zhang W. Consumption of sugar-sweetened beverages has a dose-dependent effect on the risk of non-alcoholic fatty liver disease: an updated systematic review and dose-response meta-analysis. Int J Environ Res Public Health. 2019;16(12):2192. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 36.Vos MB, Lavine JE. Dietary fructose in nonalcoholic fatty liver disease. Hepatology. 2013;57(6):2525–31. [DOI] [PubMed] [Google Scholar]
- 37.Yun HY, Lampe JW, Tinker LF, Neuhouser ML, Beresford SAA, Niles KR, Mossavar-Rahmani Y, Snetselaar LG, Van Horn L, Prentice RLet al. Serum nitrogen and carbon stable isotope ratios meet biomarker criteria for fish and animal protein intake in a controlled feeding study of a Women's Health Initiative cohort. J Nutr. 2018;148(12):1931–7. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 38.Bao Y, Stolzenberg-Solomon R, Jiao L, Silverman DT, Subar AF, Park Y, Leitzmann MF, Hollenbeck A, Schatzkin A, Michaud DS. Added sugar and sugar-sweetened foods and beverages and the risk of pancreatic cancer in the National Institutes of Health–AARP Diet and Health Study. Am J Clin Nutr. 2008;88(2):431–40. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 39.Anstee QM, Targher G, Day CP. Progression of NAFLD to diabetes mellitus, cardiovascular disease or cirrhosis. Nat Rev Gastroenterol Hepatol. 2013;10(6):330–44. [DOI] [PubMed] [Google Scholar]
- 40.Schwenzer NF, Springer F, Schraml C, Stefan N, Machann J, Schick F. Non-invasive assessment and quantification of liver steatosis by ultrasound, computed tomography and magnetic resonance. J Hepatol. 2009;51(3):433–45. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.