Given a body of human epidemiological data, one can typically only draw conclusions on correlations rather than causations: biomarkers A and B correlate with age-related condition X, but may or may not be involved in causing that condition. The research community designs animal studies to produce data on causation, and in many cases it is quite reasonable to lean on the results of those studies as support for the same path of causation in humans, but one can't run those same causation-focused studies in humans.
Thus the research community has developed strategies to tease out support for causation from correlational human data. Mendelian randomization brings in data on genetic variants that are known to affect specific biomarkers or outcomes of age-related disease. When segmenting the human data by these variants changes the outcomes, segment by segment, that suggests a causative relationship between biomarker and age-related disease. Linkage disequilibrium score regression is an analogous approach developed to try to break apart the individual contributions of multiple effects to a given outcome by adding data on genetic variants as a point of comparison.
Today's open access paper is an example of the application of these strategies to the challenge of gaining insight into causation between the aging of the gut microbiome and onset and progression of age-related conditions. The composition of the gut microbiome can be measured accurately via sequencing, and researchers now have a great deal of data to correlate specific changes with specific age-related conditions. Proof of causation has been obtained from animal studies in which restoration of a youthful gut microbiome composition improves health and extends life. Now, researchers would like to see more human data along the same lines.
The influence of gut microbes on aging has been reported in several studies, but the mediating pathways of gut microbiota, whether there is a causal relationship between the two, and biomarker screening and validation have not been fully discussed. In this study, Mendelian Randomization (MR) and Linkage Disequilibrium Score Regression (LDSC) are used to systematically investigate the associations between gut microbiota, three aging indicators, and 14 age-related diseases. Additionally, this study integrates machine learning algorithms to explore the potential of MR and LDSC methods for biomarker screening.
Gut microbiota is found to be a potential risk factor for 14 age-related diseases. The causal effects of gut microbiota on chronic kidney disease, cirrhosis, and heart failure are partially mediated by aging indicators. Additionally, gut microbiota identified through MR and LDSC methods exhibit biomarker properties for disease prediction (average area under curve, AUC = 0.731). These methods can serve as auxiliary tools for conventional biomarker screening, effectively enhancing the performance of disease models (average AUC increased from 0.808 to 0.832).
This study provides evidence that supports the association between the gut microbiota and aging and highlights the potential of genetic correlation and causal relationship analysis in biomarker discovery. These findings may help to develop new approaches for healthy aging detection and intervention.
View the full article at FightAging
