I think we already know why aging happens
2avturchin
1alex_dekan
1avturchin
1Sergei Vatolin
1Alexander Wolf
1alex_dekan
1PeterLidsky
1alex_dekan
1PeterLidsky
This is close to what I discussed in No Theory for Old Man. Evolution led to an Equal Contribution of Various Aging Mechanisms
Besides the main claim of equal contribution, I discussed that the theres is three levels: organism, repair and repair of repair. The main source of aging is damage to the repair systems and we do not have 'repair of repair" systems but most suggested therapies of aging are those second order repair systems.
The definition of "damage" is questionable. Could what we call "damage" represent a different physiological state?
Damage to a function implies a decline or complete loss of that function. But what if the observed functional decline in a living organism is not actually damage? What if this decline, a deviation from established parameters is required for adaptation?
Here is a very good, most recent example: Genomic instability (damage?) triggered ...pro-survival phenotype.
•Genomic instability generates micronuclei and chromosome fragments in the cytoplasm
•Direct cell-cell contact triggers intercellular transfer of genomic DNA
•Transferred DNA fragments are functional and maintained within recipient cell genomes
•Intercellular DNA transfer can confer heritable phenotypic changes
And human factor: In many published papares, molecular "damages" must be shown to have a manucript publishe.
One important weakness of this theory: if complex living system accumulate damage, why isn't that damage identifiable? Why can nobody pin down the damage? Is it invisible? Why has no experiment succeeded in connecting some kind of accumulating damage to lifespan or death?
And why don't we solve accumulating damage by increasing turnover? Many structures renew themselves, like epithelia. Why does skin age, if we can continuously make new skin?
1) "if complex living system accumulate damage, why isn't that damage identifiable? Why can nobody pin down the damage? Is it invisible?"
On the contrary, we know almost all kinds of damage that happens and able to pin it down. Starting from the hallmarks of aging and ending with molecular degradation, especially of long-lived molecules like elastin.
2) "Why has no experiment succeeded in connecting some kind of accumulating damage to lifespan or death?"
Thousands of experiments and even human diseases showing this. Take any DNA repair and replication enzyme mutation that are known to lead to progeria and accelerated aging,
3) "And why don't we solve accumulating damage by increasing turnover?"
That's actually the point. However, as avturchin pointed out in his post, the turnover is required on all levels of an organism, from molecules to whole tissues. That is a hard problem.
4) This is a good one:
"Many structures renew themselves, like epithelia. Why does skin age, if we can continuously make new skin?"
Mostly because elastin lives around 70 years, it degrades, accumulates damage. Not only elastin, but that is one of the major factors. We do not have in-built mechanisms of full replacement from the inside. Old elastin should be degraded and the new one synthesised, but that is not happening because we do not have such a mechanism. We never evolved one.
1) The sword intuition is confusing: living organisms are open systems that cannot be compared to inanimate objects.
2) The model fails to answer basic evolutionary questions: why did long lifespan evolve in humans, birds, naked mole rats, etc., and not in other species?
3) It may be of interest to you:
https://www.nature.com/articles/s41514-026-00365-x
in our recent paper, we show that some forms of aging plasticity and rejuvenation are mathematically incompatible with damage accumulation and trade-off classes of theories
1) Are not we made from the same molecules that degrade on their own?
2) I agree, it does not. It is where your explanations do work. But the point was that we do not actually have truly long-lived species. Just some species with better control / stability that allows them to live longer.
3)
a) I think Model II makes repair too reversible.
The model assumes damage is produced every day, and that maintenance can be shifted to later life.
But in real biology, under-repairing early does not just mean:
damage now → repair later → all fine
It often means:
damage now → worse tissue state → worse repair later
b) Old organisms already try to protect themselves.
Aged organisms are not simply “choosing low maintenance.”
Old tissues show elevated stress-response programs:
DNA damage response
antioxidant response
inflammation
senescence programs
immune activation
repair and compensation signals
Aging clocks also capture this kind of changed state.
But this is not rejuvenation.
c) Animal examples fit this better.
Long-lived animals are not examples of damage being irrelevant.
They are examples of better damage control.
Naked mole rats, bats, whales, ant queens, hydra, and other long-lived systems are interesting because they maintain stability better: better DNA repair, proteostasis, cancer suppression, stress resistance, regeneration, or tissue renewal.
Hydra is the cleanest example:
replace cells well enough → aging almost disappears
Aged mammalian tissue is the opposite example:
senescent cells + inflammation + fibrosis → regeneration becomes worse
So I do not think the paper refutes damage accumulation.
It refutes a too-simple version of damage accumulation where damage is just an additive pile that can be repaired later.
My view is different:
damage is continuous
repair is imperfect
unrepaired damage changes the system
changed systems repair themselves worse
1) the molecules are different, organic and inorganic chemistry are quite apart. But most importantly we can replace molecules with normal ones, while the inanimate objects cannot.
2) this is exactly the manipulation that justifies cherry-picking. We have lifespan differences of hundreds folds , but ignore it saying that as there is some aging, the phenomenon does not merit discussing.
3) rejuvenation and pausing aging are real-world phenomena, not modeling abstractions. Pausing is very widespread. Both are incompatible with damage accumulation framework. If there is a way to extend lifespan in adulthood, it is predicted to be used ALWAYS, not only when the environment is harsh. It is a fundamental challenge to damage accumulation model.
Status: conceptual argument. I am not claiming that every mechanism is known. I am not claiming that we already have a cheap clinical protocol. I am claiming that we understand the underlying process behind aging, and the shape of the solution.
Today I went to a museum and saw a sword from about 1000 years ago.
Only 1000 years.
And the sword was destroyed.
Black. Rusted. Thin. Full of holes. You could still see that it used to be a weapon, but physically it was almost eaten away by time.
That is the intuition I want to start from.
Matter does not stay the same by default.
Iron rusts. Stone cracks. Proteins misfold. DNA mutates. Cells drift.
Living organisms are special not because they avoid physics, but because they constantly fight it.
So the question is not:
“Does damage happen?”
Of course damage happens.
The question is:
“Why can’t biology keep repairing it forever?”
I want to make the argument as a sequence of questions.
If you disagree, this should make it easier to point to the exact step where your answer differs.
1. What kind of system is a living organism?
My answer:
A living organism is not a stable object.
It is chemistry held away from equilibrium. It stays alive by spending energy on maintenance: repairing, replacing, controlling, and preserving information.
So the first principle is simple:
life is maintenance.
Without maintenance, the body does not stay young. It does not even stay alive.
2. What is aging, at the highest level?
My answer:
Aging is what happens when a complex living system accumulates damage and fails to fully repair, remove, replace, or compensate for it.
In some cases, biology may even accelerate decline on purpose. But that does not change the basic structure: the system is moving away from the young functional state, and maintenance is not stopping it.
People sometimes say “damage explains nothing.” I disagree. Saying “damage explains nothing” is like saying “wear explains nothing about why cars fail.” Of course, after that you still need to ask which part failed and how to fix it. But wear is still explanatory.
Same in biology.
Damage is not one molecule. It can be a mutation, wrong cell state, matrix damage, or tissue misrepair. The substrate matters. But the underlying process is still damage plus imperfect renewal.
3. Does this mean oxygen or ROS explain aging?
My answer:
No.
This was probably the biggest misunderstanding of the first version of this post.
When I used oxygen as an example, I did not mean “oxygen is the cause of aging.” I meant oxygen as an example of something you cannot simply remove from a living human body.
You cannot remove oxygen. You cannot remove water chemistry. You cannot remove thermal noise. You cannot remove replication errors.
The point is that damage is unavoidable because we are physical systems.
4. If damage is unavoidable, why do species age at different rates?
My answer:
Because biology controls damage.
This is where the interesting part begins.
The same physics gives us ants, humans, whales, hydra, and turtles. Physics is not changing between them. What changes is the repair-control architecture.
This is also how I read the work of Ben Shenhar / Uri Alon and Peter Fedichev / Jan Gruber. The point is not that there is one molecule called “damage.” The point is that aging can be modeled as a recovery problem in a system under continuous damage production.
Species differ because production, repair, removal, and resilience differ.
So my short version is:
physics gives damage.
biology controls damage.
aging appears when control fails.
5. Are ant queens a counterexample?
Before answering this, I should explain why ants matter here.
In many ant species, workers and queens can have almost the same genome, but radically different lifespans. Workers may live months or a few years. Queens can live for decades. In some species, queens live close to 30 years.
So ants are a strong example that lifespan is not fixed by “being an insect.” Biology can tune it.
My answer:
No. Ant queens are not a counterexample. They are actually a good example.
They show that lifespan is strongly controllable by biology. Same genome, different caste, huge lifespan difference.
That is amazing.
But even with this control, ant queens are still short-lived compared with some human cells.
Human neurons have very high metabolic demand. Much higher than an ant queen’s tissues, if we think in energetic terms. Yet many neurons survive for a human lifetime, and some mutation-based models suggest that their maximum lifespan is ~140 years.
So what does this comparison show?
Not that damage does not matter.
It shows that raw damage rate is not enough.
Architecture matters.
An ant queen can suppress aging much better than a worker ant, but still dies in decades. A human neuron can run an expensive program for many decades because its maintenance architecture is different.
So lifespan is not explained by metabolism alone, body size alone, or oxygen alone.
It is explained by damage control.
6. What about parasites, plaques, ROS, metabolism, mutations?
My answer:
They matter.
But I do not think they are the deepest explanation.
Parasite theory, for example, may explain part of the evolutionary story. Older individuals accumulate viruses, and, therefore, eliminated by evolution to prevent diseases spreading. It may explain why some maintenance strategies were costly or not selected.
But even in a parasite-free world, the body still needs renewal.
So parasite theory is not wrong because parasites do not matter.
It is incomplete because damage and imperfect renewal remain underneath.
7. Why does biological aging accelerate?
My answer:
Because damage feeds back.
This is where organisms differ from simple machines.
A car ages because parts wear out. To keep it young, you replace parts. You do not remove friction from the universe.
But a body is harder.
Aged tires do not usually damage the engine. But aged mitochondria affect the cell. Aged immune cells affect tissues. A damaged liver affects the whole organism.
So biological aging is not just damage accumulation.
It is damage damaging repair.
That is the central loop.
8. Why does Hydra matter?
Hydra matters because it is one of the cleanest examples of extreme biological renewal.
It is a very simple animal, and in favorable conditions it can replace its cells so effectively that it shows little or no ordinary aging. This does not mean Hydra is magic. It means its body plan makes renewal much easier than in a human.
My answer:
Hydra is useful because it shows the opposite direction.
It can replace so much of itself that aging almost disappears. It does not violate physics. It just has a body plan where renewal works unusually well.
Humans cannot do that easily.
We are not just a bag of replaceable cells. We are a brain, immune system, vasculature, and memory-bearing organism. Full renewal is hard because identity and integration must be preserved.
But hard is not the same as mysterious.
9. So do we know why aging happens?
My answer:
Yes, at the level that matters most.
Aging emerges from the interaction of unavoidable damage and incomplete repair.
It is not one plaque, one parasite, one pathway, or one evolutionary mistake.
It is a failure mode of a repair-control system made of unstable matter.
That is why I think “damage” is explanatory.
Not because it names every mechanism.
But because it identifies the underlying process.
10. Do we know the shape of the solution?
My answer:
Yes.
Replace the failing substrate.
Or renew it so well that replacement becomes unnecessary.
This can mean organ replacement, engineered tissues, cell replacement, artificial organs, body-wide rejuvenation, or eventually a mostly artificial body. The route is open. But the endpoint is not mysterious.
This is not cheap.
This is not a current clinical protocol.
But conceptually it is clear.
The hard problem is not:
“What causes aging?”
The hard problem is:
“How do we replace or renew enough of the body to keep a person alive?”
11. Where does genetics fit?
I bring up genetics because it is the clearest proof that lifespan is controllable by biological architecture.
Different species, castes, and cell types do not age differently because physics changes for them. They age differently because their programs for repair, replacement, cancer suppression, and damage control are different.
My answer:
Genetics shows that lifespan control is real.
Ant queens show it. Hydra shows it. Long-lived mammals show it. Human neurons show it.
12. What would change my mind?
I would update strongly if one specific cause, such as parasites or plaques, could be removed and the organism then stayed young indefinitely without broad renewal.
I would update if complex mammals already had enough renewal capacity, and aging turned out to be mostly one removable lesion.
I would update if long-lived species did not show better maintenance, repair, replacement, or damage control after proper comparison.
My current position
We know why aging happens.
It is the result of a complex living system accumulating damage and failing to fully repair, remove, replace, or compensate for it.
We also know the shape of the solution.
Damage must be repaired, removed, replaced, or made irrelevant.
The difficulty is cost, scale, safety, and integration.
That is very different from saying “we have no idea.”
We do have an idea.
It is just not a small one.
References and what I use them for
Ben Shenhar / Uri Alon group damage accumulation model: I use this for the damage production/removal framing. Not as proof that rejuvenation is solved.
https://www.biorxiv.org/content/10.64898/2025.12.22.695887v1
Peter Fedichev / Jan Gruber minimal model: I use this for the dynamic instability / resilience framing.
https://www.biorxiv.org/content/10.1101/2025.08.25.671954v3
Ant queen lifespan: I use this as evidence that lifespan can be strongly controlled by biology.
https://pmc.ncbi.nlm.nih.gov/articles/PMC3458492/
https://pmc.ncbi.nlm.nih.gov/articles/PMC5076446/
Human neurons / mutation-bound model: I use this cautiously, as model-level support for the idea that high metabolic demand does not automatically mean short lifespan.
https://www.biorxiv.org/content/10.1101/2025.11.23.689982v1
Bowhead whale CIRBP paper: I use this as an example that long-lived species may contain specific repair tricks worth extracting.
https://www.nature.com/articles/s41586-025-09694-5
Human somatic mutation landscape paper: I use this as a correction against an oxygen-only view.
https://www.nature.com/articles/s41586-021-03822-7
AI use: I used ChatGPT for editing and structure. The claims and final responsibility are mine.