Learn more about Peter Attia’s digital program
One of the biggest challenges in aging science is that we don’t have any biomarkers that can reliably track the aging process. Such biomarkers – were they to exist – would have obvious uses in predicting lifespan, but perhaps more importantly, they would finally allow us to assess the efficacy of interventions intended to slow the aging process. When you take a drug to lower blood glucose, apoB, or blood pressure, you can titrate the dosage up or down as needed because you have a readout, a biomarker, to monitor the magnitude of the drug’s effect. But when it comes to the most interesting geroprotective molecules – rapamycin, metformin, or NR/NMN, to name a few (we use the term “geroprotective” to denote that such drugs target the basic hallmarks of aging rather than specific diseases) – we have no biomarker to tell us (1) if the drug is having any effect, and (2) if we are taking too much or not enough.
However, in the past decade, a technology has emerged that might offer a clue. Several consumer-facing “biological clocks” claim to give the user insight into their “biological age,” which may be different from their “chronological age.” If biological clocks accurately predict future lifespan and could demonstrate that an intervention postpones mortality, they might prove to be a useful tool in determining the efficacy of geroprotective therapies.
Would you like to read the full article?
This content is an exclusive benefit to premium members.
Premium members get access to articles like this one, monthly “Ask Me Anything” podcast episodes, extensive show notes for every episode of The Drive, and more.
If you are a member who is logged in and are receiving this message, please ensure the cookies are enabled on your browser.
Dr. Attia,
Based on your swimming background and overall fitness level, I was wondering if you ever thought about giving water polo a try?
Excellent article! This is the most thorough breakdown of biological clocks I’ve come across, well done.
I do disagree slightly with the concluding remark, ‘ While epigenetic clocks might someday be a vital tool for the field of aging science, as of now, they offer zero insight into either future lifespan or as a biomarker for geroprotective interventions.’
We currently have very few reliable biomarkers, and I view these clocks similarly to others like A1C, they don’t tell the whole story, but taken in the aggregate they start to paint a picture. I personally use Elysium’s, once a year over the last few years, and have found them to be pretty good at cross verifying my particular health issues. I also do not view the age as a binary (when will I die) rather a probability – ‘I have as much chance of getting a disease as some at age X’. In other words, if my chronological age is 45, and my heart bio-age is 33, broadly speaking I’d say I have as much of a chance of having a heart attack as a 33 yr old.
I look forward to many of the questions and suggestions in this article being teased out in the future! I also would love to see a more liquid-biopsy like approach.
“‘I have as much chance of getting a disease as some at age X’. In other words, if my chronological age is 45, and my heart bio-age is 33, broadly speaking I’d say I have as much of a chance of having a heart attack as a 33 yr old.”
I would really like to see an age of & cause of death study comparing the following 3 groups over a 20+ year period (starting at 60-80) :
– ideally a 30-40 year study would be even better but significantly more difficult to execute
(i) age normal – i.e. those whose biological clock = biomarker clock
(ii) age abnormal poor i.e. those whose biomarker clock is > biological clock
(iii) age abnormal good i.e. those whose biomarker clock is < biological clock (such as yourself)
Obvious Questions:
(1) compare ages of death
(2) compare causes of death (i) between groups) and (ii) wrt biomarkers
I'd appreciate Peter's input on the above plausibility of such a study.
You may want to consider Ben Garcia for a future podcast guest. His mass spectrometric methods give truly novel insights into epigenetics.
https://www.bengarcialab.com/
Thanks for this amazing dive into epigenetic clocks. You mention that longitudinal data sets are “critical for advancing our understanding of the relationship between DNA methylation, cellular aging, and mortality risk, which in turn is necessary for the creation of reliable epigenetic clocks…” Can you comment on the 3rd generation clock, DunedinPace of Aging DNA methylation algorithm, by Belsky and colleagues that is based on longitudinal sampling of the same group of participants at ages 26, 32, 38, and 45. I would love to hear your opinion of whether this approach seems more promising.
Thanks,
Alison
https://pubmed.ncbi.nlm.nih.gov/32367804/
https://pubmed.ncbi.nlm.nih.gov/33796868/
https://elifesciences.org/articles/73420
https://github.com/danbelsky/DunedinPACE
Should have explained at the beginning of the article what ‘methylation’ means. should have clarified first what hyper- or hypo-methylation is. Article suddenly talks about rate of methylation change, which is confusing. When I first started reading, I was thinking “is hypermethylated state good or bad”.
I applaud your dissection of yet another attempt to elevate a biological marker
to a prognostic indication of aging and to imply that it can define one’s future.
Population studies give us an indication of how we age but do nothing to
define those who exceed the expectations of the statistics.
I would love for you to concentrate on those aspects of living that lead to a
“life well lived” …. humility, gratitude, love and sacrifice …
A modern day Christmas Carol ….