Evaluating the clinical relevance of the enterotypes in the Estonian microbiome cohort

Abstract

Human gut microbiome is subject to high inter-individual and temporal variability, which complicates building microbiome-based applications, including applications that can be used to improve public health. Categorizing the microbiome profiles into a small number of distinct clusters, such as enterotyping, has been proposed as a solution that can ameliorate these shortcomings. However, the clinical relevance of the enterotypes is poorly characterized despite a few studies marking the potential for using the enterotypes for disease diagnostics and personalized nutrition. To gain a further understanding of the clinical relevance of the enterotypes, we used the Estonian microbiome cohort dataset (n = 2,506) supplemented with diagnoses and drug usage information from electronic health records to assess the possibility of using enterotypes for disease diagnostics, detecting disease subtypes, and evaluating the susceptibility for developing a condition. In addition to the previously established 3-cluster enterotype model, we propose a 5-cluster community type model based on our data, which further separates the samples with extremely high Bacteroides and Prevotella abundances. Collectively, our systematic analysis including 231 phenotypic factors, 62 prevalent diseases, and 33 incident diseases greatly expands the knowledge about the enterotype-specific characteristics; however, the evidence suggesting the practical use of enterotypes in clinical practice remains scarce.

A novel framework for assessing causal effect of microbiome on health: long-term antibiotic usage as an instrument

“…We and others have previously shown that hAB is associated with the microbiome composition additively, meaning that higher hAB is associated with larger changes in the microbiome composition 2,16 . Here, we focused our analysis on the Prevotella / Bacteroides ratio (P/B ratio) that represents a summarised state of microbiome by reflecting the enterotype composition 18 . We confirm the additive effect of total hAB on P/B ratio and show that it also holds for different classes of antibiotics, namely macrolides (ATC code J01FA), penicillins (J01CR) and fluoroquinolones (J01MA), which we will use for the sensitivity analysis ( Fig.…”

“…We and others have previously shown that hAB is associated with the microbiome composition additively, meaning that higher hAB is associated with larger changes in the microbiome composition 2,16 . Here, we focused our analysis on the Prevotella/Bacteroides ratio (P/B ratio) that represents a summarised state of microbiome by reflecting the enterotype composition 18 .…”