How Uncertain Should We Be Regarding the Validity of Epigenetic Clocks as a Measure of Biological Age?
How well does an epigenetic clock measure biological age? The research community cares about whether an epigenetic clock can be trusted in the measure of biological age because being able to use clocks to assess potential rejuvenation therapies would greatly speed up research and development. This is an important question, but because of the way in which epigenetic clocks are constructed, using machine learning to fit algorithms to epigenetic data from populations at various ages, it is not obviously the case that researchers can quantify the risk that a specific clock is not a good measure of biological age. Here find a discussion of this issue and some thoughts on the way forward.
The primary assumption of aging clocks is that the deviation ∆ of predicted age from the chronological age C represents an accelerated or decelerated aging, that is, an increase or decrease in the biological age B. Since biological age cannot be measured directly, the epigenetic age estimated by the clocks is therefore considered a proxy measure of the biological age. However, before aging clocks could be integrated into clinical practice, these models should provide an estimate of uncertainty for their own predictions.
Uncertainty manifests itself in three ways: (i) Model choice uncertainty, part of a broader category known as epistemic uncertainty, represents how well a proposed model reflects the actual underlying process. (ii) Out-of-distribution (OOD) uncertainty, another type of epistemic uncertainty, emerges when the testing data are not represented in the training data distribution, leading to a high risk of model prediction failure (iii) Aleatoric uncertainty originates from data variations that cannot be reduced to zero by the model.
From the clinical perspective, epistemic uncertainty must be estimated to make reliable conclusions about whether to trust a model or not. Specifically, epistemic uncertainty resulting from the dataset shift should be scrutinized, considering the prevalence of batch effects in biological data. Dataset shift describes the case of OOD sampling where the testing population is under-represented within the training distribution. However, most popular DNA methylation aging clocks fail to meet this criterion because they are typically built using algorithms from the penalized multivariate linear regression (MLR) family. Such algorithms do not yield information on any of the uncertainties, except for the error between chronological and predicted ages in the training data.
In this work, we question the applicability of existing aging clock methodology for measuring rejuvenation by specifically examining prediction uncertainty. We present an analytical framework to consider rejuvenation predictions from the uncertainty perspective. Our analysis reveals that the DNA methylation profiles across reprogramming are poorly represented in the aging data used to train clock models, thus introducing high epistemic uncertainty in age estimations. Moreover, predictions of different published clocks are inconsistent, with some even suggesting zero or negative rejuvenation. While not questioning the possibility of age reversal, we show that the high clock uncertainty challenges the reliability of rejuvenation effects observed during in vitro reprogramming before pluripotency and throughout embryogenesis. Conversely, our method reveals a significant age increase after in vivo reprogramming. We recommend including uncertainty estimation in future aging clock models to avoid the risk of misinterpreting the results of biological age prediction.
I see the Longecity forum server is down.
As I see it, the fundamental, and insurmountable, problem with epigenetic testing to access age is that biological age is not the exact same thing as remaining life/healthspan. We do have access to a tool which can judge remaining life/healthspan. It may sound on a certain level superficial, but the human brain evolved over millions of years to make that assessment based upon visual cues at both conscious and unconscious levels. How old you look to people that do not know you is currently the most accurate biomarker of aging. I think a digital model should be built around this existing framework.
@johnD how old you look to other people is very subjective determination -- therefore very inaccurate and very uncertain. Many women put on make up to look younger than their chronological age and/or had plastic surgery performed on them-- to look younger.
Few men put on make up to look younger, but some men are clean shaven-- to look younger, some had plastic surgery done on them, and some are unshaven and/or grow beard -- to look older and more distinguished ( as for example Aubrey deGrey).
Epigenetic clock could measure VITALITY (how healthy and alive you are)
It is useful to measure how fast you are aging compared to normal aging ("normal" is difficult to establish ) Are you aging faster than normal or slower [accelerated or decelerated aging] for the purpose of anti-aging treatment -- doctors should be able to measure slower pace of aging or faster rejuvenation to determine effectiveness of anti-aging treatment {and it must be standardized and objective measure} International Anti-aging Conference must establish the standard of measure of velocity of aging {how fast or slow you are aging or de-aging [rejuvenating]}
I always found the fascination with epigenetic clocks as a preferred, primary evaluation tool strange. It seems to me that in pursuing a reduction in death, dysfunction, and decline (in assumed deceasing importance) would be easier to track and more relevant to measuring the success of an anti-aging intervention. As we look at time/ track/ functionality to failure (or of stability and/or improvement) of any scale of generic human biology from the cellular machinery to groups of major organs, it would seem reasonable to accumulate that data and presume that increased 'stabilization' (over decline) that various parts exhibited, would naturally provide a path forward to larger scale survivability and even preferred inter-realtionship success and overall system thriving. Surely, an AI literature review of all the various means of localized failure (and possibly unusual stability) could shed (more) light on 'pro-active' intervention, monitoring, and follow-up strategy. My 2c.