A gut aging clock using microbiome multi-view profiles is associated with health and frail risk

ABSTRACT

Age-related changes in the microbiome have been reported in previous studies; however, direct evidence for their association with frailty is lacking. Here, we introduce biological age based on gut microbiota (gAge), an integrated prediction model that integrates gut microbiota data from different perspectives with potential background factors for aging assessment. Simulation results show that, compared with a single model, the ensemble model can not only significantly improve the prediction accuracy, but also make full use of the data in unpaired samples. From this, we identified markers associated with age development and grouped markers into accelerated aging and mitigated aging according to their effect on the prediction. Importantly, the application of gAge to an elderly cohort with different frailty levels confirmed that gAge and its predictive residuals are closely related to the individual’s health status and frailty stage, and age-related markers overlap significantly with disease and frailty characteristics. Furthermore, we applied the gAge prediction model to another independent cohort of the elderly population for aging assessment and found that gAge could effectively represent the aging population. Overall, our study explains the association between the gut microbiota and frailty, providing potential targets for the development of gut microbiota-based targeted intervention strategies for aging.

KEYWORDS:

• Gut microbiome
• metagenomics
• machine learning
• aging
• frailty

Acknowledgments

We acknowledge the active involvement of all the participants of the cohort. We are also thankful for the support of the cohort recruitment and processing teams.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Authors’ contributions

Hongchao Wang and Yutao Chen: Conceptualization, Software, Validation. Yutao Chen, Hongchao Wang and Ling Feng: Writing original manuscript, Methodology, Formal analysis. Wenwei Lu and Shourong Lu:Collecting population cohorts and Analysis. Wenwei Lu and Jinlin Zhu: Investigation, Methodology, and Formal analysis. Jianxin Zhao, Hao Zhang Wei Chen, and Wenwei Lu: Supervision, writing – review & editing.

Availability of data and materials

This study incorporates data from previously published studies. The vast majority of sequence data comes from curatedMetagenomicData3 repository, the taxonomic and pathway profiles were downloaded for the analysis in this study. The Eldermet cohort was used as the validation cohort data in this study, the raw sequence data was available at the European Nucleotide Archive(ENA) via accession numbers PRJEB37017, and the metadata was available as part of the original publication. *1. Adjusting for age improves identification of gut microbiome alterations in multiple diseases. All relevant codes (or scripts) used for this analysis are available at https://github.com/hcwang-jn/gut-age.

Ethics approval and consent to participate

All participants in this study provided their written informed consent. The cohort was established at Wuxi People’s Hospital Affiliated to Nanjing Medical University, and was approved by the hospital ethics committee (approval number: KS2019039).

Additional information

Funding

This work was supported by the National Key Research and Development Program of China (2022YFF1100403).