[Skip to Navigation]
Sign In
Comment & Response
August 8, 2019

Whole Grains and Risk of Hepatocellular Carcinoma—Missing the Forest for the Trees?—In Reply

Author Affiliations
  • 1Channing Division of Network Medicine, Department of Medicine, Brigham and Women’s Hospital, Harvard Medical School, Boston, Massachusetts
  • 2School of Public Health, Department of Nutrition, Anhui Medical University, Hefei, Anhui, PR China
  • 3Department of Nutrition and Department of Epidemiology, Harvard T. H. Chan School of Public Health, Boston, Massachusetts
JAMA Oncol. 2019;5(10):1509-1510. doi:10.1001/jamaoncol.2019.2988

In Reply We thank Dr Braillon for the comments on our article.1 The dietary etiology of liver cancer remains poorly understood. We had an a priori hypothesis that higher intake of whole grains and dietary fiber reduces risk of hepatocellular carcinoma (HCC). This hypothesis was partially based on 2 different lines of evidence: (1) higher intake of whole grains and fiber, especially cereal fiber, may reduce risk of metabolic diseases, improve gut integrity, and alter gut microbiota composition,1 and (2) these health conditions may play a role in hepatocarcinogenesis.2 A randomized trial design is considered the gold standard to establish causality but is likely infeasible in the study of diet and cancer because decades of follow-up are needed to study outcomes with a low incidence like HCC, and long-term adherence to assigned diets is usually poor.2,3 Hence, prospective cohorts2,3 with repeated measurements of diet based on validated food frequency questionnaires1 can provide useful information on the dietary etiology of HCC. Although confounding cannot be totally ruled out, the similar results between age-adjusted and multivariable models1 likely argue against substantial confounding by other behaviors and social determinants related to diet. Because this is a hypothesis-driven study, multiple testing is less of a concern. Nonetheless, given multiple exposures and limited case numbers, results should be interpreted with caution. Our findings are generally consistent with the only other European study on this topic.4 Additionally, the observed inverse associations of cereal but not fruit or vegetable fiber with HCC risk were also reported for other disease outcomes (eg, total mortality, cardiovascular disease, type 2 diabetes).1 Taken together, the findings that we reported are biologically plausible.

Add or change institution