Firewall link : The use and misuse of ‘biological aging’ in health research | Nature Medicine
From Satchin Panda on X
“ Biological Age Vs. Chronological age is often discussed in academic and public discourse. This article in @NatureMedicine is a good read to understand the use and misuse of various “Aging clocks””
I like the idea of the “Propensity score” which is what these bioclocks seem to be trying to measure… propensity to disfunction / disease.
Was listening to the interview with Morgan Levine today while I was working out. In part 2 she discusses that far more than is discussed, these groups that identify success with a given therapeutic approach (as measured by aging clock) there is likely much more “shopping” around of the various clocks (testing with many different ones - and only citing the one that shows benefit). So, be skeptical of groups claiming success with a given bioclock.
It is not really surprising that people quote the best results that they can get.
I think, however, that there are biomarkers that are really worth concentrating on one baseline CRP is one of those.
The problem with this biomarker is that you cannot just measure CRP and find the baseline CRP. It needs to be measured a number of times to exclude the effect of temporary infection.
Similarly with things like DNA methylation that is affected by external temporary factors which do not reflect the underlying status of the organism.
A few thoughts on CRP & Aging Clocks:
Multiple tests are only necessary to establish a baseline when CRP is elevated, to rule out the possibility of a transient infection. However, a single annual test showing CRP consistently below 0.5 mg/L provides substantial information, including confirmation that there are no subclinical infections, such as chronic infections that could elevate CRP. These might include common viruses, bacteria linked to UTIs, dental conditions, rhinoviruses, or even H. pylori.
Although I haven’t developed a full explanation, evidence related to ASCVD, and perhaps certain cancers and pulmonary functions, suggests that inflammation—as reflected in metrics like CRP—acts as a control variable in the development of many diseases. Extrapolating from multiple studies, it appears that only 10-15% of significant ASCVD events occur in individuals with CRP levels <0.5 mg/L. This holds even in studies where only one CRP measurement was taken. Given the proportion of deaths attributable to ASCVD, if this estimate is accurate, it indicates that CRP could be one of the most critical longevity metrics to manage.
This reasoning aligns with my belief that since most people die from specific diseases, it’s more productive to focus on preventing or managing the diseases for which our genetic heritage and environment make us most susceptible. This is more useful than chasing an aging clock index, particularly because many of these indices are products marketed by private companies, and none have demonstrated significant predictive power in experiments where death is the dependent variable.
This makes sense.
I am about to publish a podcast with Wayne Jackson PhD about combating chronic inflammation for better health and athletic performance in older athletes. He leans on these 4 blood markers to get at chronic inflammation status in athletes, and he emphasizes diet to address digestion related root causes as well as calorie balance (too few calories is the more common problem in athletes) and micronutrient deficiencies.
Well, I disagree with Ikram. An interesting point is how heavily he leans on Meyer/Schumacher 2024 here. I haven’t independently combed through that, but one might read Mitteldorf’s comments (below) for context on the issue.
https://joshmitteldorf.scienceblog.com/2024/08/14/stochastic-methylation-clocks/
I get bloodwork done regularly and often plug the results into various aging calculators, some of which have disappeared. The results varied between tests, sometimes as much as ten years. I started to believe that the results were BS because I certainly didn’t feel 10 to 20 years younger, as the tests suggested.
What difference does it make if an aging clock suggests that you are 20 years younger than you are when you know for a fact that you are not physically 20 years younger based on your current athletic abilities?
I’m not very well versed on how these clocks work but few times I looked at them (at least a couple of them) they measure say the age of single organs and kind of come with an average/estimate age. for example, a 50-year-old male his heart markers showing that of a 20-year-old, lungs the level of 55, and kidney that of a 35 and so on, and their overall age ends up showing 37 (13 years younger than actual). Well, all good but I would NOT want to be that 50-year-old lol even though it showed that I’m 37, because to me your age is that of the worst performing major organ. In this case 55 and not 37.
https://onlinelibrary.wiley.com/doi/full/10.1111/acel.14377
“ Given that the bulk of molecular aging is tissue-specific and aging itself is a remarkably complex, multifarious process, it is unsurprising that most surveyed scientists agree that aging cannot be quantified via a single metric. We share this sentiment and argue that, just like it would not be reasonable to assume that an individual with an ideal grip strength, VO2 max, or any other aging biomarker is biologically young, we should be careful not to conflate an aging clock with whole-body biological aging. ”
Mike Lustgarten goes deep on DunedinPace biological clock, and his results using it:
And another good discussion on biological aging clocks, this time from Vadim N. Gladyshev, Ph.D., and it includes hypothetical scenarios with rapamycin use.
https://gladyshevlab.bwh.harvard.edu
My use of “aging clocks” has been interesting. I don’t do this to evaluate disease risk at all. I have regular blood work and other tests done for that.
Trends over time are more my interest. At our last DNAm test (Nov 23) several markers improved significantly. Interestingly at about the same time (Jan 24) as that DNAm test, my blood work was the best it’s been in 15 years.
While I struggled to improve my blood tests over the past 15 years and did make some progress, it was the “radical” use of 3 or 4 protocols that finally moved both needles in a more healthy direction. What motivated me more? my Dr and the tests she can prescribe? or the DNAm tests that show different aspects of aging than the blood tests do?
Did I miss out on something by doing DNAm testing? or did I use the tech that was available to measure as much as I could to improve all my markers?
I don’t see DNAm testing as a substitute for “conventional” testing but as an interesting adjunct that further motivates me to be better.
For me this discourse about the use of DNAm tests is more intellectual, and may or may not be helpful to the longevity community. If gamification of longevity as we see with the rapidly growing number of participants in the Rejuvenation Olympics, helps the movement gain greater traction in the public mind, that should be a good thing.
Also controversy and debate bring eyeballs 
It introduces the concept of “if you don’t measure it, you can’t manage it” to more people who might never think of that aspect of aging.
To me this is a same debate as the health-span vs life-span camps.
Aren’t many of these measures derivative of others, either empirically or statistically? This highlights another weakness in portions of the aging clock movement. They need to produce a single metric that demonstrates both convergent and discriminant validity for the DV. The “clock’s” model is unscientific if it consists of a laundry list of potentially relevant metrics from which any one can be casewise isolated as an IV.
To take this in yet another direction, the concept of an aging clock is appealing to many, including myself. However, despite my efforts, the work on this to date has been less than impressive, and in some ways, it seems lacking in scientific rigor and logic. Given the high caliber of this group (the best, in my opinion), I’m sure some of you recall Wittgenstein’s caution against searching for a single criterion that unites all uses of a term (or construct). He argued persuasively that many terms do not share one common feature but instead are linked by numerous overlapping threads, forming what he called a “family resemblance.” A metaphor often used to illustrate this idea is that of a long, strong rope, through which no single strand runs end-to-end.
The notion of a true aging clock presupposes or implies some sort of master orchestration of all structures and functions—something that “directs” or otherwise leads to the gradual decline in their integrity.
I still appreciate the elegance of this model, but it has yet to be substantiated, even weakly, and support for the current contenders, such as methylation, seems to be waning.
There is an alternative to the “clock” model. Based on current evidence, it seems equally plausible that structures and functions decline both independently and interactively, driven by a variety of unrelated causes, some of which are random. For example, declining kidney function caused by the acquisition of a rogue virus while traveling might trigger a cascade of aging events, eventually leading to death. The inheritance of a BRCA gene can result in numerous aging-related events. Background radiation levels differ between geographic locations. One person’s chronic stress may lead to hardened and blocked arteries, while another individual with the same lipid profile, who has mastered stress management, might not experience this outcome. We can all think of such examples. Is there a master clock behind these events? Not necessarily. Our genes, our environment—including epigenetic factors—what we learn, and what we believe, all play a role, and none of these are orchestrated by or answer to a master clock. Aging may simply represent a gradual decline in structural and functional integrity, driven by a multiplicity of causes, some of which are unpredictable or unknowable.
Viewing aging as a complex, non-clock-driven process does not dismiss the utility of so-called “aging clocks,” which often consist of lists of scientifically validated health metrics. However, these lists are not clocks in any meaningful sense. What they do offer is a collection of factors to monitor and manage. Through empirical research, these lists can be refined, made more precise, and perhaps even simplified, with a focus on atomic (as opposed to derived) metrics. This type of research could eventually lead to a more parsimonious index, benefiting the management of age-related challenges. I believe this is a goal we all share.
Of course, I’m open to alternative perspectives or further development of this idea.
It makes sense that aging researchers would like to have aging clocks to have a way to quickly assess the impact on aging for short term interventions. But there are lots of assumptions in that idea. Science doesn’t know how a single cell works yet but it is trying to predict how all cells change over a long time (aging). It might work but I’m not putting that information into my head until I believe it is good information.
I am confused enough by the complexity of what IS known. Throwing in a heap of possibly bad data won’t help me to sort out what to do.
In the meantime I’ll keep investing most of my energy in doing the things that are very likely to help.
Don’t major in the minors is the best advice I’ve received so far.
I agree on the majoring in the minors challenge. Even though it is exactly the right advice, I am sometimes challenged to stick with it. I recall Attia mentioning how little he thought we knew about aging once you take down the mirrors and clear the smoke. What we do know quite a bit about is how to reduce ASCVD risk and, to a lesser extent, a few other things. But even ASCVD is changing. It looks like static measurements may capture a decent amount of variance but, as we know, it leaves notable outliers unexplained. We might be on the edge of developing a “lipid clearing rate metric” which will more reflect a true cause. I like the idea because it might explain why some with terrible metrics never suffer events and vice versa.
Usage of the phrase “biological age” has picked up considerably since the advent of aging clocks and it has become commonplace to describe an aging clock’s output as biological age. In contrast to this labeling, biological age is also often depicted as a more abstract concept that helps explain how individuals are aging internally, externally, and functionally. Given that the bulk of molecular aging is tissue-specific and aging itself is a remarkably complex, multifarious process, it is unsurprising that most surveyed scientists agree that aging cannot be quantified via a single metric. We share this sentiment and argue that, just like it would not be reasonable to assume that an individual with an ideal grip strength, VO2 max, or any other aging biomarker is biologically young, we should be careful not to conflate an aging clock with whole-body biological aging. To address this, we recommend that researchers describe the output of an aging clock based on the type of input data used or the name of the clock itself. Epigenetic aging clocks produce epigenetic age, transcriptomic aging clocks produce transcriptomic age, and so forth. If a clock has a unique name, such as our recently developed epigenetic aging clock CheekAge, the name of the clock can double as the output. As a compromise solution, aging biomarkers can be described as indicators of biological age. We feel that these recommendations will help scientists and the public differentiate between aging biomarkers and the much more elusive concept of biological age.
