New publication: Robust Biomarkers for microbiome-based stratification in lifestyle interventions
Summary
Shaping the gut microbiota can help prevent disorders and improve health, and scientists have been exploring several lifestyle interventions targeting the gut microbiota as a treatment for diverse diseases. However, in some cases, these interventions fail to modify the gut microbiota composition, and as a consequence, the desired therapeutic role is not fulfilled. In this article, first authors Jiarui Chen and Sara Leal Siliceo led by Prof. Gianni Panagiotou demonstrated that we can predict microbiome resistance to change in response to lifestyle interventions by simply using the baseline microbiome composition. We are confident that this knowledge may help in the future to further improve the design of personalized lifestyle approaches.
Figure caption
Lifestyle interventions like diet change and exercise, do not always succeed in their goal of changing the gut microbiota (Non-responder). With the power of Artificial Intelligence, we developed a model to predict if the gut microbiota will be resistant to change in response to lifestyle intervention. Figure created with BioRender (Sara Leal Siliceo -Leibniz-HKI)
Background
A growing body of evidence suggests that the gut microbiota is strongly linked to general human health. Microbiome-directed interventions, such as diet and exercise, are acknowledged as a viable and achievable strategy for preventing disorders and improving human health. However, due to the significant inter-individual diversity of the gut microbiota between subjects, lifestyle recommendations are expected to have distinct and highly variable impacts to the microbiome structure.
Results
Here, through a large-scale meta-analysis including 1448 shotgun metagenomics samples obtained longitudinally from 396 individuals during lifestyle studies, we revealed Bacteroides stercoris, Prevotella copri, and Bacteroides vulgatus as biomarkers of microbiota’s resistance to structural changes, and aromatic and non-aromatic amino acid biosynthesis as important regulator of microbiome dynamics. We established criteria for distinguishing between significant compositional changes from normal microbiota fluctuation and classified individuals based on their level of response. We further developed a machine learning model for predicting “responders” and “non-responders” independently of the type of intervention with an area under the curve of up to 0.86 in external validation cohorts of different ethnicities.
Conclusions
We propose here that microbiome-based stratification is possible for identifying individuals with highly plastic or highly resistant microbial structures. Identifying subjects that will not respond to generalized lifestyle therapeutic interventions targeting the restructuring of gut microbiota is important to ensure that primary end-points of clinical studies are reached.
Original Publication
Chen J, Siliceo SL, Ni Y, Nielsen HB, Xu A, Panagiotou G. Identification of robust and generalizable biomarkers for microbiome-based stratification in lifestyle interventions. Microbiome. 2023 Aug 8; 11(1):178. doi: 10.1186/s40168-023-01604-z. PMID: 37553697.