Five “biological age clocks” were linked to different patterns of gene activity in blood
Five “biological age clocks” were linked to different patterns of gene activity in blood
On July 20, npj Aging published a study of 3 227 participants in the Health and Retirement Study, a long-running US study of people over 50. The researchers measured both DNA methylation, which consists of chemical marks on DNA, and RNA levels, which indicate which genes are active in blood cells.
Epigenetic clocks are formulas that combine DNA marks into an estimate of biological age. They were trained for different purposes. Horvath and Hannum were designed to predict chronological age, PhenoAge and GrimAge to predict health measures and mortality risk, and DunedinPACE to estimate the rate of age-related change across several body systems. The authors examined which changes in blood gene activity were associated with high values for each measure.
On July 24, eight epigenetic clocks produced age estimates for the same blood samples that differed by an average of 17 years. That study documented the disagreement among the estimates. The new paper shows which gene activity programs are associated with each of the five measures.
A higher Horvath score was statistically associated with 49 genes, while DunedinPACE was associated with 3 204. The authors found no gene shared by all five clocks. They obtained the same result when they compared pathways in Reactome, a catalogue of biological processes. At a broader level, the estimates converge: all are associated with metabolism, development, immune function, and cellular regulation. The common term “epigenetic clocks” therefore covers measures that each capture a different combination of processes.
Using these profiles, the authors developed transcriptomic aging scores, or TAGS. These scores are calculated from the RNA levels of genes associated with the corresponding clocks. The researchers used 2 584 participants to determine which genes would enter each formula and how much weight each gene would receive, then tested the result in the remaining 645. TAGS showed a moderate association with their corresponding clocks. TAGS reflect current gene activity in blood, whereas the original clocks measure chemical marks on DNA.
In the group of 645 people, TAGS were more strongly associated with several outcomes than the corresponding DNA methylation clocks. These outcomes included frailty, defined as the accumulation of health deficits, difficulty with daily activities, diabetes, heart and lung diseases, and mortality. When the researchers included TAGS and the clocks in the same analysis, both GrimAge measures and both DunedinPACE measures were independently associated with mortality and frailty.
When testing an intervention, researchers can select a measure according to the process and outcome that the intervention is expected to change, such as the rate of functional decline, immune function, metabolism, or accumulated risk. The RNA profile provides a map of current gene activity in blood that can inform this choice.