Primary Care Corner with Geoffrey Modest MD: Microbiome Changes and Severity of NAFLD

By Dr. Geoffrey Modest

There is not a great understanding as to which patients with non-alcoholic fatty liver disease (NAFLD), the most common liver disorder in the US/Western countries, are among the 20-30% who progress to nonalcoholic steatohepatitis (NASH), or those who progress even further to fibrosis and cirrhosis. Certain genetic polymorphisms have been identified which predispose people to more aggressive disease: there are some data that the I148M variant of the PNPLA3 gene confers a 3.5-fold increased risk of NASH and a 3.2-fold increased risk of fibrosis. And there may be roles for epigentics, gender/hormone status, and nutrition as well. This article looked at the potential role of the gut microbiome (see HEPATOLOGY 2016;63:764), since as noted in several blogs (see below), there may also be a role of microbiome changes in some related conditions (e.g., obesity, metabolic syndrome, diabetes).


  • 30 patients with no/only periportal F0/F1 fibrosis (10 with NASH) and 27 patients with significant F≥2 fibrosis (25 with NASH) had 16S ribosomal RNA gene sequencing of stool samples. All patients had liver biopsy diagnosis, did not drink alcohol >210 g/week for men or >140 g/week for women, did not have chronic hepatitis B or C, and did not have evidence of other chronic liver diseases on biopsy.
  • Mean age 57, 34 males/23 females, BMI 31, 40% with diabetes, 50% elevated triglycerides, 75% depressed HDL, 81% with metabolic syndrome, mean AST of 40 and ALT of 64.


  • Bacteroides was significantly more abundant in those with NASH and F≥2 fibrosis; Prevotella was decreased (these two genera are competitors with an inverse abundance relationship). Also, Ruminococcus  was increased.
  • On multivariate analysis (adjusting for BMI, BP, triglycerides, HDL, metabolic syndrome), Bacteroides was independently associated with NASH, and Ruminococcus with F≥2 fibrosis.
  • ​There was a dose-response curve: both the rate of NASH and F≥2 fibrosis tracked with increasing abundance of Bacteroides and Ruminococcus, respectively
  • ​In the subgroup with metabolic syndrome (overall a group with higher risk of more severe NAFLD lesions), those with more abundant Bacteroides and Ruminococcus had more severe NAFLD lesions. This was also found in those with diabetes and several of the components of metabolic syndrome
  • In assessing the presumed functional effects of these bacterial changes, those with NASH had a microbiome enriched in ability to metabolize carbohydrates, lipids, and amino acids.
  • This study adds to the potential role of microbiome changes in causing clinical diseases. The study itself cannot imply causality, just an association: for example, did the microbiome cause the changes in the microbiome, or vs vice versa? Were there other nonetiologic factors that caused both the microbiome changes in these patients and the NAFLD/NASH changes in the liver,such as eating too much pâté (this was a French study, after all)? But, arguments in favor of the microbiome changes being causal:
    • Animal studies show that manipulating the gut microbiome can lead to changes in the liver lesions of NAFLD, including steatosis, NASH, fibrosis and liver cancer
    • ​Bacteroides is associated with increased fecal content of deoxycholic acid, D-pinitol, choline, farrrinose, and stachyose (the last 2 containing glucose and fructose), and lower amounts of short-chain fatty acids (SFCAs) and amino acids. And deoxycholic acid induces apoptosis in the rat liver and is increased in NASH patients, fructose is associated with increased liver inflammation and fibrosis in NAFLD patients, and decreased SCFAs (esp. butyrate, propionate and acetate — see Curr Opin Clin Nutr Metab Care 2014, 17:139) may be detrimental to NAFLD
    • Bacteroides is favored overall in the gut when people eat western diets (high in fat, animal proteins, and sugar), vs Prevotella being favored in those eating foods rich in fiber, starch and plant polysaccharides. In fact changing the diet to an animal-based one leads to a rapid shift in the gut microbiome to Bacteroides, and promotes insulin resistance.

So, what are the implications of this study:

  • This study and others on the microbiome (for example, see blogs:​ ) reinforce that there seems to be a dynamic interrelationship between the gut microbiome and human disease, and that the changes in the microbiome lead to changes in bacterial metabolism and byproducts, which may be the mediators/abettors of disease processes.
  • There are several key players affecting the composition of the gut microbiome, including diet [see above microbiome blogs for articles on red meat and heart disease, medications (including one on metformin suggesting that much of its hypoglycemic role is as a promoter of healthy changes in the microbiome) as well as on antibiotics, and exercise]
  • And, I think that the associations with NAFLD above (even if not clearly causal) will expand my approach to patients with NAFLD to include reinforcing a vegetable-based diet, as well as my usual approach of weight loss, lipid control, and exercise.

(Visited 1 times, 1 visits today)