Two new studies demonstrate that characterizing key gut microbiota taxa can predict or diagnose type 2 diabetes mellitus (T2DM) and cirrhosis, respectively. These microbiome signatures could lead to non-invasive tools that improve disease diagnosis and the identification of high-risk individuals.
In the first study, Oh et al. compared the faecal microbiomes of 163 individuals, covering the spectrum of nonalcoholic fatty liver disease (NAFLD) and including non-NAFLD controls and patients with NAFLD-cirrhosis, as well as their first degree relatives. Metagenomic data was combined with untargeted metabolomic data, and a machine learning algorithm, along with differential abundance analyses, was used to identify signatures that were effective in detecting cirrhosis.
The signature defined by the investigators detected cirrhosis with an area under the curve (AUC) diagnostic accuracy of 0.91, and it was used to identify patients with cirrhosis from two independent cohorts that included multiple aetiologies from ethnically and geographically diverse areas. Furthermore, when serum aspartate aminotransferase levels were included, the researchers were able to discriminate cirrhosis from earlier stages of fibrosis. “Our finding of a universal microbiome-derived signature lays the foundations for a stool-based diagnostic test for advanced fibrosis or cirrhosis,” write the authors.
In the second study, Reitmeier et al. examined metabolic health and the diurnal rhythmicity of gut microbiota in a German population cohort of 1,976 individuals. Faecal microbiota profiling of the cohort was used to find taxa that show 24 h oscillations in their abundances. The investigators identified a risk signature of 13 microbial taxa that showed disrupted rhythmicity in T2DM and constructed a model to classify T2DM. Using 699 paired faecal samples from 5 years apart, the investigators found their model was able to predict individuals at risk of developing T2DM with an AUC of 0.78 when BMI was included.
“microbiome signatures could lead to non-invasive tools that improve disease diagnosis”
The arrhythmic microbial signature also successfully discriminated T2DM from healthy individuals in an independent German cohort of 1,363 individuals. “These findings clearly highlight the need to consider diurnal changes in the gut microbiome for diagnostic and prognostic investigations,” write the authors.
About this article
Cite this article
Dickson, I. Microbiome signatures for cirrhosis and diabetes.
Nat Rev Gastroenterol Hepatol (2020). https://doi.org/10.1038/s41575-020-0351-3