-by Raghav Sehgal, Scientific Venture Consultant and Jyothi Devakumar, Chief Scientific Officer
A key question in biology is to understand why and how we age. But asking that question starts with defining aging and scientists have been debating, what constitutes aging, for decades. A particular definition which I personally find the most comprehensive is one given by Luigi Ferrucci, Scientific Director of NIA and it goes as follows:
“Aging is the ratio between
damage accumulation and compensatory mechanisms.
If you think about it, damage accumulation occurs constantly
in our biology and physiology.
We have DNA damage and protein misfolding,
we have organelle deterioration, and so on.
But over millions of years we have
developed mechanisms that allow us to repair this damage…”
But here is the thing: a phenomenon that cannot be quantified cannot be researched upon. This is where aging biomarkers come in to help us quantify the rate of aging in our bodies. In a way any biomarker that predicts aging is effectively trying to capture the ratio between damage accumulation and compensatory mechanism. Take for example, any Omic Aging Biomarkers also called Omic clocks (proteomic, transcriptomic, metabolomic, epigenomic), all of them have the concept of accelerated or decelerated aging, which is essentially quantifying the question how much higher or lower is the ratio between damage accumulation and compensatory mechanisms in your body for your given age!
Omic Aging Biomarkers or clocks are based on the simple principle that “your body keeps a count” in other words, a wide variety of environmental and biological factors cause dysregulation in your normal bodily functions which are imprinted at the very molecular levels of the body. Thus by looking at these molecular levels such as genes, proteins, metabolites and methylations, we can predict how much dysregulation has happened with respect to the compensatory mechanisms of the body, in turn giving us an estimate of the aging rate of our body. That being said, of all the molecular clocks DNA methylation-derived epigenetic clocks are currently better in estimating aging than transcriptomic and proteomic data, or telomere length based clocks.
It is thus worthwhile to go through a brief history of epigenetic aging clocks and also talk about how they can be made better.
The journey of building epigenetic aging clocks was started almost a decade ago by Bocklandt et al when they showed CpG methylation predicts age The Hannum and Horvath multi-Tissue clock was the first to predict chronological age and showed that having a higher predicted age than your chronological age is associated with mortality and comorbidities. The next generation of clocks redefined aging. Till then all clocks had been predicting chronological age, Levine et al defined a new age called Biological age based on bodily biomarkers and their relationship to mortality. This novel Biological age was then predicted using methylations to build the first version of biological age or mortality prediction clocks. These clocks were far superior at predicting aging and a lot better associated with multiple comorbidities. Since then multiple clocks have been built to predict a wide variety of traits in a wide variety of systems ranging from blood based mitotic rate clocks to brain based, brain aging clocks. Each of these clocks are great at predicting very specific measures or traits of aging.
In the last 10 years epigenetic clocks have come a long way from being simple predictors of age to being more specific that predict aging in specific data modalities (eg. single cell vs organ vs whole body) or different types of aging (eg. Frailty vs mortality vs cognitive decline). The reason for this shift is that over the years we have understood that aging is not a monolithic homogeneous process but a mixture of multiple heterogeneous processes. Each of these processes deteriorate at different rates across humans as well as within humans. Thus to measure aging on each of these “dimensions” of aging we need clocks that specifically measure aging on these dimensions. This has given rise to the present trend of bespoke clocks that are very specific in measuring a specific dimension of aging using specific data modalities. Off recent this has given the rise to clocks such as:
The next few months or years will see more and more bespoke epigenetic clocks that target specific aging mechanisms or data modalities.
Even though Epigenetic Aging clocks have been shown to capture a wide range of environmental and biological stressors there is a lot more to be done before they can be used en-masse in clinical settings ranging from the discovery of therapies to enhance healthspan to being used by Physicians in order to determine medical course of action. A few reasons for this are as follows:
At present over 50 different epigenetic aging clocks exist predicting for different aging measures, ranging from mortality to cognitive function to physical function and more. This poses a huge challenge in terms of their usage. Most often researchers from other fields that wish to use these clocks do not have a clear idea as to which clock is best for their research. More importantly, the variety of traits that these clocks predict make it almost impossible to use in a clinic since a physician will have to go through 50 different measures just to figure out what interventions a patient might need. Thus there is a clear need for a singular clock which is able to give a broad overview of biological aging and not of specific traits.
Another critical issue with epigenetic aging clocks is reliability. Many clocks that are available in the present day have huge variations up to 9 years for a single individual taking an epigenetic test on the same day. These variations are partly due to swift changing of our epigenetic landscape due to different bodily activities. Such large variations make it ineffective to use these clocks for any clinical purpose. Thus there is a clear need to reduce noise and variability in epigenetic aging clocks so that they can be used with high reliability to prescribe and build interventions for increasing healthspan. That being said, epigenetic clocks are already better than their clinical chemistry biomarker counterparts at predicting disease outcomes, that doesn’t mean we shouldn’t try to improve them further to become more accurate and reliable.
Even though we need unified clocks that can be used for a broad overview of biological aging, at the other end there is also a need to explain why that biological aging might be occurring. One might argue that clocks that predict specific traits might be able to do so but to actually have explainability we do not need clocks that predict specific aging traits pertaining to particular bodily systems but rather biological aging pertaining to the different biological systems. In turn, a combination of aging in all of these systems should be used to give a whole body biological age. This way, physicians and researchers can actually understand which systems in the body are actually driving the overall biological age which in turn can be used to prescribe or build system specific interventions for patients to increase their healthspan.
Aging is not homogenous throughout our body but rather very heterogeneous with certain systems aging quicker than others in turn predisposing us to a certain group of age-related diseases. For example some of us will die of cardiac disorders while others of age related neurological degeneration resulting in diseases such as Parkinson's. This heterogeneity in aging is not captured by epigenetic aging clocks. In some ways this links back to explainability of clocks. As we build clocks that explain our overall bodily aging as a function of different systems in our body, we will realize that there are multiple aging trajectories that exist and epigenetic aging clocks should be able to predict our aging rate on each of these different aging trajectories.
Epigenetic clocks over the last decade have transformed our thinking of aging by allowing us to give quantifiable metrics to understand the rate of deterioration in our bodies. That being said, a lot more work needs to be done before these tools can be used en masse in research and clinics to build and prescribe novel healthspan prolonging interventions. Rather in the rise of “cell rejuvenation” therapies (and startups built around the therapies) the question is that can Epigenetic aging clocks prove their metal as superior biomarkers for conducting research for such therapies!
 - Jylhävä J, Pedersen NL, Hägg S. Biological Age Predictors. EBioMedicine. 2017 Jul;21:29-36. doi: 10.1016/j.ebiom.2017.03.046. Epub 2017 Apr 1. PMID: 28396265; PMCID: PMC5514388.
 - Bocklandt S, Lin W, Sehl ME, Sánchez FJ, Sinsheimer JS, Horvath S, Vilain E. Epigenetic predictor of age. PLoS One. 2011;6(6):e14821. doi: 10.1371/journal.pone.0014821. Epub 2011 Jun 22. PMID: 21731603; PMCID: PMC3120753.
 - Hannum G, Guinney J, Zhao L, Zhang L, Hughes G, Sadda S, Klotzle B, Bibikova M, Fan JB, Gao Y, Deconde R, Chen M, Rajapakse I, Friend S, Ideker T, Zhang K. Genome-wide methylation profiles reveal quantitative views of human aging rates. Mol Cell. 2013 Jan 24;49(2):359-367. doi: 10.1016/j.molcel.2012.10.016. Epub 2012 Nov 21. PMID: 23177740; PMCID: PMC3780611.
 - Horvath S. DNA methylation age of human tissues and cell types. Genome Biol. 2013;14(10):R115. doi: 10.1186/gb-2013-14-10-r115. Erratum in: Genome Biol. 2015;16:96. PMID: 24138928; PMCID: PMC4015143.
 - Levine ME, Lu AT, Quach A, et al. An epigenetic biomarker of aging for lifespan and healthspan. Aging (Albany NY). 2018;10(4):573-591. doi:10.18632/aging.101414
 - Trapp, A., Kerepesi, C. & Gladyshev, V.N. Profiling epigenetic age in single cells. Nat Aging 1, 1189–1201 (2021). https://doi.org/10.1038/s43587-021-00134-3
 - DNAmFitAge: Biological Age Indicator Incorporating Physical Fitness
Kristen M. McGreevy, Zsolt Radak, Ferenc Torma, Matyas Jokai, Ake T. Lu, Daniel W. Belsky, Alexandra Binder, Luigi Ferrucci, Riccardo E. Marioni, Simon Cox, Michael Kobor, David L. Corcoran, Steve Horvath
medRxiv 2022.03.21.22272043; doi: https://doi.org/10.1101/2022.03.21.22272043
 - Aging the Brain: Multi-Region Methylation Principal Component Based Clock in the Context of Alzheimer’s Disease
Kyra L. Thrush, David A. Bennett, Christopher Gaiteri, Steve Horvath, Christopher H. van Dyck, Albert T. Higgins-Chen, Morgan E. Levine
bioRxiv 2022.02.28.481849; doi: https://doi.org/10.1101/2022.02.28.481849
 - Liu Z, Leung D, Thrush K, Zhao W, Ratliff S, Tanaka T, Schmitz LL, Smith JA, Ferrucci L, Levine ME. Underlying features of epigenetic aging clocks in vivo and in vitro. Aging Cell. 2020 Oct;19(10):e13229. doi: 10.1111/acel.13229. Epub 2020 Sep 15. PMID: 32930491; PMCID: PMC7576259.