Higher intake of whole grains and dietary fiber are associated with lower risk of liver cancer and chronic liver disease mortality

Primary liver cancer is a major global health burden, ranking sixth in incidence and third in cancer-related mortality. In the U.S., incidence and mortality have been rising, and 5-year survival remains poor, underscoring the need for primary prevention. Hepatocellular carcinoma (HCC) and intrahepatic cholangiocarcinoma (ICC) are the predominant histologic types. Established risk factors include chronic HBV/HCV infection, aflatoxin, heavy alcohol consumption, smoking, obesity, NAFLD, and type 2 diabetes. Chronic liver diseases (CLD) such as cirrhosis and fibrosis are important precursors and causes of mortality. Non-viral HCC is common in the U.S., with obesity-, diabetes-, and NAFLD-related etiologies increasingly implicated, many of which are diet-related. Whole grains provide fiber and micronutrients, and both whole grains and dietary fiber have been linked to lower risks of T2DM, cardiovascular disease, some cancers, and possibly NAFLD. However, evidence directly relating whole grain and dietary fiber to liver cancer risk is limited, and no prior study examined their associations with CLD mortality. This study investigates whether higher intake of whole grains and dietary fiber are associated with reduced risks of incident liver cancer and CLD mortality in a large U.S. cohort.

Prior evidence suggests potential protective roles of whole grains and dietary fiber against liver cancer. A systematic review indicated an inverse association between whole grain intake and HCC risk. Two prospective cohorts reported inverse associations between dietary fiber and liver cancer incidence, with EPIC (191 HCC cases) finding significant associations for total fiber and cereal fiber, and NHS/HPFS (141 HCC cases) reporting suggestive inverse associations for cereal fiber and for whole grain intake, but not for fruit or vegetable fiber. Evidence on CLD mortality in relation to these dietary factors had been lacking before this study.

Design: Prospective cohort analysis within the NIH-AARP Diet and Health Study. Participants: 566,398 members aged 50–71 years were initially enrolled (1995–1996). Exclusions: prevalent cancer at baseline (n=50,118), extreme energy intake (women <500 or >3,500 kcal/day; men <800 or >4,000 kcal/day; n=29,983), and participants diagnosed with liver cancer or who died from CLD before questionnaire scanning (n=580). Final analytic cohort: 485,717 (290,484 men; 195,233 women). Follow-up: through December 31, 2011; median 15.5 years. Exposures: Whole grain intake (ounce equivalents/day) and total dietary fiber and fiber from specific sources (fruits, vegetables, beans, grains) assessed at baseline using a validated 124-item FFQ. Whole grains were quantified using USDA MyPyramid Equivalents Database and Pyramid Servings Database; nutrients derived from USDA CSFII 1994–1996. FFQ validation showed energy-adjusted correlations for fiber of 0.72 (men) and 0.66 (women). Outcomes: Incident primary liver cancer (ICD-10 C22) identified via linkage to state cancer registries; HCC defined by morphology codes 8170–8175; ICC by codes 8032, 8033, 8070, 8071, 8140, 8141, 8160, 8161, 8260, 8480, 8481, 8490, 8560. CLD mortality identified via National Death Index Plus (ICD-9 571.0, 571.2–571.6, 571.8, 571.9; ICD-10 K70, K73, K74). Validation studies indicated high sensitivity/specificity for cancer registry linkage and good specificity for CLD classification. Statistical analysis: Cox proportional hazards regression estimated hazard ratios (HRs) and 95% CIs across quintiles of intake and per SD increase. Model 1: stratified by sex and adjusted for age. Model 2: additionally adjusted for education, race/ethnicity, BMI (<25, 25–30, ≥30 kg/m²), alcohol use (non-drinker; 0.1–4.9; 5–9.9; ≥10 g/day), smoking (never, former, current), physical activity (frequency categories), history of diabetes (yes/no), and total energy intake (continuous). Stratified analyses by BMI, diabetes, alcohol, smoking, and physical activity; interactions tested. Sensitivity analyses included adjustment for HEI-2015, exclusion of early cases (first 2 or 5 years), and exclusion of heavy alcohol drinkers.

- Cases: 940 incident liver cancers (635 HCC, 164 ICC) and 993 CLD deaths over a median 15.5 years follow-up among 485,717 participants. - Liver cancer incidence (multivariable Model 2, Q5 vs Q1): • Whole grains HR=0.78 (95% CI: 0.63–0.96), P-trend=0.01. • Total dietary fiber HR=0.69 (0.53–0.90), P-trend<0.001. • Fiber from vegetables HR=0.65 (0.52–0.83), P-trend<0.001. • Fiber from grains HR=0.78 (0.62–0.99), P-trend=0.04. • Fiber from fruits HR=0.94 (0.76–1.17), P-trend=0.82 (null). • Fiber from beans HR=0.89 (0.71–1.10), P-trend=0.02 (Q5 vs Q1 not significant). Associations were stronger for HCC than ICC; inverse associations for total fiber and vegetable fiber observed particularly in men. - CLD mortality (Model 2, Q5 vs Q1): • Whole grains HR=0.44 (0.35–0.55), P-trend<0.001 (56% lower risk). • Total dietary fiber HR=0.37 (0.29–0.48), P-trend<0.001 (63% lower risk). • Fiber from vegetables HR=0.55 (0.44–0.69), P-trend<0.001. • Fiber from beans HR=0.67 (0.54–0.84), P-trend=0.028. • Fiber from grains HR=0.29 (0.23–0.36), P-trend<0.001. • Fiber from fruits HR=0.98 (0.79–1.20), P-trend=0.923 (null). - Stratified analyses: Inverse associations generally consistent across BMI, diabetes, alcohol, smoking, and physical activity strata, with some interactions noted (e.g., for CLD mortality, interactions with BMI, alcohol, smoking for whole grains and total fiber). - Sensitivity analyses: Results robust to additional adjustment for HEI-2015, exclusion of early cases (2 or 5 years), and exclusion of heavy alcohol drinkers. - Baseline: Higher quintiles of whole grain/fiber associated with greater physical activity, lower alcohol and smoking, and slightly higher prevalence of self-reported diabetes; mean age 61.5 years; 59.8% men; 92.8% white.

The findings address the hypothesis that higher whole grain and dietary fiber intake reduce risk of liver cancer and CLD mortality. Significant inverse associations were observed for whole grains and total fiber with both outcomes, and particularly for fiber from vegetables, beans, and grains. No protective association was seen for fruit fiber, suggesting heterogeneity by fiber source, potentially due to fructose content attenuating benefits. Results were consistent across most subgroups and robust in sensitivity analyses, indicating that observed associations are not explained by early disease or heavy alcohol intake and persist after accounting for overall diet quality. Biological plausibility includes effects of whole grains and fiber on glycemic control, insulin sensitivity, liver fat content, and metabolic syndrome, as well as modulation of gut microbiota and the gut–liver axis via short-chain fatty acid production and reduced inflammation. The stronger associations for HCC compared with ICC align with differing etiologies. The study expands prior limited evidence by demonstrating associations with CLD mortality and delineating fiber source-specific effects.

In this large U.S. cohort, higher intake of whole grains and total dietary fiber was associated with significantly lower risks of incident liver cancer and mortality from chronic liver disease. Protective associations were evident for fiber from vegetables, beans, and grains, but not from fruits. These findings support dietary recommendations emphasizing whole grains and fiber-rich foods to reduce liver cancer incidence and CLD mortality. Future research should clarify causal mechanisms, disentangle effects of specific fiber types and accompanying nutrients (e.g., fructose), assess dose–response relationships, and evaluate generalizability across diverse racial/ethnic populations and age groups.

- HBV/HCV infection status and history of hepatitis were not available for the full cohort; however, prior studies suggest limited confounding by these infections with included lifestyle factors and diet. - Dietary intake measured once at baseline via self-administered FFQ; potential measurement error and inability to account for dietary changes over time. FFQ is better for ranking than precise quantification. - Residual confounding possible due to healthier overall lifestyle among high whole grain/fiber consumers despite multivariable adjustment and adjustment for HEI-2015 in sensitivity analyses. - Outcomes included liver cancer incidence and CLD mortality; CLD incidence could not be assessed due to lack of registries and comprehensive records. - Cohort predominantly of European ancestry and aged 50–71 at baseline, potentially limiting generalizability.