Ageing Clocks: Challenges When Translating Into Clinical Practice

Introduction. Biological age is an indicator of the functional integrity of the body. It is a more accurate predictor of health status and mortality compared to chronological age, and also has the potential to evaluate the effectiveness of geroprotective interventions. Biological age is calculated using biomarkers of aging, which include various biochemical, genetic, phenotypic and functional characteristics of the body. To measure biological age, statistical analysis methods and neural networks with deep machine learning are used. The resulting formulas and calculation methods are called biological age calculators or “ageing clocks”. Purpose. Our study was aimed to examine existing biological age calculators and describe the potential difficulties of translating them into clinical practice.

Materials and methods. A literature review was performed on the PubMed platform spanning the last 5 years. The search was performed using the keywords “aging clocks, ageing clock, age clock, biological age” for the period from 2018 to 2023. Reviews, systematic reviews, and meta-analyses were included in the search query. A total of 136 articles matching the search query were found.

Research results. The results derived from the literature analysised, demonstrated that the most accurate currently existing “ageing clock” is the second generation epigenetic clock. They estimate both biological age and all-cause mortality (DNAm PhenoAge, DNAm GrimAge), as well as take into account the potential aging phenotype. The data required for these epigenetic biological age calculators includes not only DNA methylation information, but also additional information: some clinical indicators (PhenoAge, GrimAge), as well as a smoking index (GrimAge).

However, the lack of availability of DNA methylation assessment may cause difficulties when translating the epigenetic “aging clock” into clinical practice. From this perspective, the assessment of biological age using spectrometry and chromatography on proteome or metabolome data, as well as the analysis of fecal microbiota associated with human biological age, might be more viable for routine practice.

At the same time, the existing “ageing clocks” do not allow an objective assessment of biological age, since they either consider a limited amount of data (such as Galkin’s deep microbiotic clock), or while it is known that different types of cells have different rates of aging, analyze data from a very large undifferentiated array of biomaterial (like Howarth's multi-tissue epigenetic clock or Manuela Rist's metabolomic clock based on blood and urine data). Finally, it is not entirely clear which exact mechanisms of aging underlie the data obtained on biological age; therefore, the points for geroprotective interventions remain unclear.

In our opinion, the limited predictive precision and challenging implementation process are the primary hindrances to translating the existing "aging clock" into clinical practice. Further studies are required to create a variety of highly specialized calculators that estimate the biological age of various body tissues and do not require complex invasive methods of collecting biomaterials, and a machine model trained to analyze a large array of data to create a more accurate integrative indicator in the form of the biological age of an individual.

Conclusion. The existing “aging clock” assesses the functional integrity of the body and predicts outcomes. However, the low accuracy and low availability of complex information processing methods currently limit the translation of this tool into clinical practice and require improvement of existing models of biological age calculators.