A metabolomics approach to the validation of predictive metabolites and phenotypic expression in non-alcoholic fatty liver disease

Aims: Nonalcoholic fatty liver disease (NAFLD) is becoming more common and severe. Individuals with NAFLD have an altered composition of gut-microbial metabolites. We used metabolomics profiling to identify microbial metabolites that could indicate gut-liver metabolic severity. Noninvasive biomarkers are required for NAFLD, especially for patients at high risk of disease progression. Main methods: We compared the stool metabolomes, untargeted metabolomics, and clinical data of 80 patients.Patients with nonalcoholic fatty liver (NAFL: n = 16), nonalcoholic steatohepatitis (NASH: n = 26), and cirrhosis (n = 19) and healthy control individuals (HC: n = 19) were enrolled. The identified metabolites in NAFLD were evaluated by multivariate statistical analysis and metabolic pathotypic expression. Gas chromatography-mass spectrometry (GC-MS) and liquid chromatography coupled to time-of-flight-mass spectrometry (LC-TOF-MS) were used to analyze metabolites.Key findings: Untargeted metabolomics was used to identify and quantify 103 metabolites. Principal component analysis (PCA) was used to assess the metabolic discrimination of NAFL, NASH, and cirrhosis. Short-chain fatty acids (SCFA) levels were significantly lower in NAFLD patients, including those of acetate (p = 0.03), butyrate (p = 0.0008), and propionate. The stool cholic acid (p = 0.001) level was significantly increased in NAFLD patients. Palmitoylcarnitine and L-carnitine levels were significantly increased in NASH and cirrhosis patients. The phenotypic expression of these metabolites was linked to beta-oxidation.Significance: We demonstrated a distinct metabolome profile in NAFLD patients with NAFL, NASH, and cirrhosis. We also discovered that the expression of certain metabolites and metabolic pathways was linked to NAFLD.