Here’s a crisp summary of Sierra’s Nature Aging correspondence, “Are we getting closer to understanding why we age?”:
Framing the field. Aging theories cluster into two camps—damage accumulation vs programmed aging. The 2013 Hallmarks of Aging and geroscience framework organized the field and helped move it toward clinical testing, aided by epigenetic clocks and newer pace-of-aging biomarkers.
Core claim. While macromolecular damage (mutations, telomere loss, protein/lipid damage) rises with age, its pattern is not linear. Examples such as brain amyloid show little accumulation before ~40, then acceleration later. This suggests aging isn’t just passive damage buildup; rather, there’s an age-dependent failure of protective systems (the hallmarks), producing a system-level loss of “molecular resilience”—cells’ capacity to cope with diverse insults.
Why resilience fails. Thermodynamics alone can’t explain the late-life acceleration or species differences. Evolutionary tuning does: species invest in repair/resilience only to the extent favored by their ecological niche and reproductive timing. Long-lived species (e.g., naked mole rat, whales, elephants) possess more robust defenses because longevity benefits reproduction in their niches. In humans, declining female reproductive fitness around 40–50 implies resilience built to last roughly to 60–70, after which it degrades.
Implications.
Prioritize molecular resilience as a unifying target above individual hallmarks and develop predictive measures of resilience and future health.
Expand beyond standard lab models to comparative species that are “successful agers” for their niche, to learn diverse resilience strategies.
Continue translating geroscience into human trials while refining biomarkers of aging trajectories, not just chronological age.
Bottom line: Aging reflects an evolutionarily tuned decline in molecular resilience across interconnected maintenance systems, leading to accelerated damage later in life. Advancing the field now hinges on measuring, comparing, and therapeutically bolstering resilience.
I’ve never really understood why evolution-driven programmed aging is so controversial. It always seem obvious at a ‘just so story’ level: Tribe A blessed with ‘programmed aging’ genes will have fewer decrepit oldies and will struggle less to survive a period of scarce resources than Tribe B. The trade off being the cultural wisdom stored in the brains of said oldies. Tribe A will also have better genetic fitness by limiting the age of reproductivity.
So a non-linear increase in macro-molecular damage makes sense. Thermodynamics explain some aging, and explain the need for programmed senescence - but not all.
Which leaves two big questions - what is the background pattern of damage accumulation, and when will AI figure out how to circumvent programmed aging?
My high level assumption is that if aging is evolutionarily programmed to some extent through some ‘species selection’ effect, then it’s likely that a variety mechanisms are involved.
In evolutionary biology terms a local peak of genetic fitness would be more robust if that were the case. Or put more simply:
the DNA of a prehistoric individual who say evolved immortal mitochondria through some citrate mechanism would be less likely to dominate if multiple mechanisms had evolved each sufficient to program aging.
and also, presumably, the evolutionary imperative to prune the elderly for the benefit of the next generations arose independently multiple times over the eons.
I hope I’m wrong, and that we’re on the cusp of mitochondrial-therapy longevity escape velocity.
So keep up the good work!
And even if I’m right, I also assume that programmed aging will prove more tractable to both human and Artificial intelligence than the underlying thermodynamic entropy. It shouldn’t be beyond our wits to reverse engineer a giant tortois
I think there are other issues such as nucDNA damage, but what we understand today to be aging is primarily a mitochondrial issue with senescence as the cherry on top.
What evidence is there that mitochondria primarily drives aging. I can imagine that question is susceptible to scientific study and I’d love to see some.
Presumably if we can preserve or improve mitochondrial function universally, even for a short time, we should be able to see broad and systemic aging biomarker preservation or improvement.
Or maybe other species evidence?
sorry, I realize I’m all questions and no answers
In my opinion, building a strong extracellular matrix similar to that of naked mole rats and then inducing telomerase with potent inducers could already be sufficient in extending lifespan indefinitely without requiring constant growth. The former requires replenishing collagen and elastin as well as ultra high molecular weight hyaluronic acid and will build a cancer-proof environment. The latter actually de-ages human cells without inducing cancer.
Sorry I wasn’t clear. The poster cites lots of papers showing how mitochondrial damage contributes aging. And how those impacts can be major and widespread.
But what I was asking about was evidence that mitochondrial damage is the “primary” driver of aging. ie that mitochondrial damage contributes to aging more than any other factors.
Ideally, I’d also love to know of any evidence which quantifies the degree to which mitochondrial damage contributes to aging cf to all other factors. That would give us some idea of the longevity impact of ‘solving mitochondria’ alone.
There are obviously lots of candidates for this: senescence, stem cell exhaustion, genomic instability, altered intercellular communication, epigenetic deregulation, accumulated DNA damage, histone disregulation, glycation, protein aggregation, loss of proteostasis etc.
But I think I understand where you’re coming from: that acetylation issues are the underlying mechanism which drives all of these. Is that correct?
But is there any data we can point to that gives us confidence that fixing acetylation will “fix” all these other ‘causes’ of phenotypic aging? As opposed to understanding the mechanisms by which fixing acetylation will merely help these other mechanisms.
I guess I’m assuming a 100% fix is unlikely, but it would interesting to have some insight into how much of the aging problem would be solved by fixing acetylation alone. I don’t have any insight whatsoever based on what I’ve read. But I’m hoping you or others might be able to point me towards some research data which would help build that insight.
That’s correct. Acetylation drives all of these. Glycation is slightly different in that it is driven by glucose levels, but the repair systems are driven by acetylation.
But is there any data we can point to that gives us confidence that fixing acetylation will “fix” all these other ‘causes’ of phenotypic aging? As opposed to understanding the mechanisms by which fixing acetylation will merely help these other mechanisms.
I don’t think that can be proven at this point. However, i take a simplistic view which is to fix acetylation and then see what remains.
But I’m hoping you or others might be able to point me towards some research data which would help build that insight.
You can look at the links between splicing and diseases of aging as a first step.
It’s because physics comes first and evolution only comes on top of physics. A lot of evolutionary programmed theories tend to deny this and promote the idea that it’s all programmed, which is definitely not true. Yet many people can’t let go of that belief. I think in some cases it’s wishful thinking, because if aging were 100% programmed, it would be much easier to solve than if it’s driven largely by damage accumulation.
Yes 100% programmed aging is a completely different argument. Presumably, the only reason evolution would favour programmed aging is if there were underlying physics/entropy causes of aging too. So 100% seems very unlikely. But I’ve never come across anyone arguing that 100% of aging is programmed.
I’d say the area of debate is partly around how much of the underlying entropy is reversible/repairable through biological mechanisms. And that could be 100% or it could be much less. And in any event, the % of aging damage we need to fix for sizeable longevity improvements is also up for debate.
I’d love to see charts showing decade by decade decline in various hallmarks of aging. I suspect those charts, if they show a cliff edge/tipping points (with say a sharp decline after a certain age), would give us some insight into how much gain can be made from reprogramming evolutionary aging mechanisms.
You would be surprised. I have debated with several prominant figures in the longevity industry that are such strong believers in programmed aging that they appear to dismiss the role of damage in aging entirely and think that all we have to do is fix to the programming and then everything else will get fixed by the body.
Yes, IMO that should be the main debate. There is no question that some entropic damage is repairable. I think there is a lot that cannot ever be fixed by reversing programming. I say this not only because I know of examples of damages that can’t be fixed, but because it makes no sense for evolution to waste energy on fixing everything, when there is no need to do so to pass the genes on.
I agree. I would phrase it a bit differently though. I would say that evolution doesn’t exaclty set the average lifespan per se. More accurately, it sets the minimal resilience against entropy that is necessary to pass the genes on. The average lifespan then results from that.
There comes a time when you need to think in terms of “design”. It’s not enough to try to fix individual weak links in the machinery, like atherosclerosis, but fundamentally re-engineer the whole design. “Design” by evolution is entirely opportunistic, where the whole point is to allow species perpetuation, and if a problem occurs that interferes with that, a “quick and dirty” solution, a patch evolves and puts a bandaid on the problem. Look at cancer. Cancer can result from faulty reading of the genetic code, if you perform a number of readings, eventually purely stochastically you’ll misread something and you get cancer. So the patch here is to have telomeres that count down the number of readings (divisions) and eventually at an arbitrary number (long enough to allow species propagation) the telomere runs out and the cell senesces so it can’t divide anymore and prevents cancer. Yes, senescent cells are all part of “old”, but that doesn’t matter, because the species already propagated. And so on for everything, nature finds a temporary fix that works just long enough for species propagation and then the whole rickety structure falls apart, there is no further fixes because they are not needed and there is no evolutionary pressure.
What we need instead is to rething the whole mechanism. If the problem in cancer is misreading, then instead of senescing the cell, how about we put in a mechanism that makes misreading less likely through a superior mechanism - that’s true genetic engineering, where you are thinking in terms of a re-design, instead of a “quick fix” - as is, our bodies at present are a big bag of ‘quick fixes’ that stuck around long enough, a rube goldberg device that works on inertia - an original design that keeps getting one patch on top of another, there is no evolutionary mechanism for starting from scratch and asking “how can we design this to work with longevity as a goal”. No more patching miles of ancient spaghetti rotting code, and start with a complete re-write.
I have no doubt that such an approach will one day come about. Right now, we are at the stage of trying to improve on nature’s quick fixes by tinkering with them - that’s what pharma does - statins, BP meds, GLP-1RA etc., all look at how to improve on a quick fix patch - and as we know from the engineering joke “there is no solution quite as permanent as a temporary solution”. We need to think higher vantage point - start with the design… one day!