talk title: Robust Mouse Rejuvenation: In Defense of Engineering – or – We Don’t Need No Stinkin’ Theories

speaker: Aubrey de Grey

date: 2025-08-28

video: https://www.youtube.com/watch?v=Uz0irxDUSMA&t=1s

event: Longevity Summit Dublin

LLM-generated summary: Aubrey de Grey critiques pessimism in aging research, advocating continued pursuit of damage-repair strategies despite incomplete mechanistic understanding, as embodied in the SENS framework's seven damage categories (e.g., cell loss, senescent cells, extracellular aggregates), now targeted by startups and validated through mainstream publications. He dismisses strong theories of aging (e.g., free radical dominance) as obsolete, favoring a "divide-and-conquer" approach that sidesteps ignorance by repairing known lesions, evidenced by additive lifespan extensions in middle-aged mice via combined interventions (telomerase gene therapy, senolytics, young bone marrow transplant, rapamycin) in a 1,000-mouse LEV Foundation study yielding ~4 months extension—comparable to late-life calorie restriction (CR) but promising for additivity without CR's species-scaling limitations. Planning a 2,000-mouse trial with eight interventions, de Grey argues such dramatic rodent results are essential to galvanize public and influencer support (e.g., Rogan, Fridman), mirroring COVID's rapid vaccine mobilization to make aging a societal priority and achieve longevity escape velocity.

Video description

At Longevity Summit Dublin 2025, Aubrey de Grey, President and CSO of the LEV Foundation, presents his landmark talk “Robust Mouse Rejuvenation: In Defense of Engineering – or – We Don’t Need No Stinkin’ Theories.”

In this session, de Grey challenges outdated “theories of aging,” making the case for a damage-repair engineering approach over purely theoretical speculation. He shares results from his large-scale mouse rejuvenation experiments, testing combinations of interventions such as rapamycin, stem cell therapy, telomerase gene therapy, and senolytics.

You’ll hear: * Why “theories” of aging may hold back progress * How repairing cellular and molecular damage could extend healthy lifespan * What recent mouse studies reveal about additive effects of combined therapies * Why calorie restriction fails in long-lived species, and why damage repair should scale better * How to mobilize public and political will for aging research—making aging “the new COVID”

This talk offers a view of where the field of biogerontology stands today, and why bold engineering solutions may hold the key to achieving longevity escape velocity.

Learn more and support the work: https://www.levf.org/

Opening Remarks and Context

Most of you will know that for the past three years, Martin and I ran this conference together, with me being responsible essentially for the speaking program.

And honestly, you know, I think it is a huge feather in Martin's cap that the quality of the speaking program is just as good as it was before, despite the lack of all the network and many decades of personal relations with prominent speakers that I've built up over the years.

So hats off to Martin. Round of applause.

And also I'd like to say this is a wonderful choice of venue. This venue, I gather, only came into existence very recently. And so, again, I think that's very good that that happened.

All right, so if you look on the program, you will see that I'm not supposed to be the last speaker. The last speaker was supposed to be Peter Fedichev, and I think that may have inspired Martin's decision to call this session Theories of Aging, because Peter and I have largely similar but somewhat contrasting views on the nature of aging and how hard it's going to be to fix.

And indeed, many of you will know that we did a rather excellent debate hosted by the Forsyth Institute several months ago, which I believe is online on YouTube somewhere.

Pushing Back on Slow Progress Narratives

So anyway, that was kind of what inspired the title of this talk. But it's a kind of something that I've been meaning to say on stage for a little while because I feel I need to push back a little bit on a tendency that has emerged over the past little while to look at the slowness of progress in doing something about aging over the past several years since the idea of even trying to do something about aging finally became fashionable.

And to kind of wring our hands and say, oh dear, we don't understand aging well enough. We should retreat to the curiosity-driven approach of trying to figure out to try to understand aging better before we get into the serious business of trying to do something about it.

And I think that's bollocks. I think the amount that we know already is abundantly enough to justify not abandoning the effort to understand aging better, not at all, but it's abundantly enough to justify trying to see what happens with what we already understand.

Background: The Unique Problem of Late-Life Aging

So that's more or less what I'm going to tell you shortly. Just a couple of minutes of background for those of you who have not seen about 1,000 of my talks over the past 20 years.

I usually start with this slide, which highlights the nature of the problem, as I see it. The fact that there is this profound difference between the health problems of late life and the health problems of early life, against which we have had such enormous success over the past couple of hundred years, starting with just figuring out that hygiene is quite a good idea and with very elementary medicines really that have essentially eliminated early deaths when of course 200 years ago literally more than one third of babies would die before the age of one even in the world's wealthiest countries.

So this is a big question and it's a question that I believe is not asked often enough so I ask it a lot and I essentially answer it by going through the mistakes that have been made in adopting essentially quixotic strategies for doing anything about aging.

Defining Aging: Metabolism, Damage, and Pathology

And it all comes down to this little diagram that many of you will have seen before, which starts out at the bottom by defining aging as the combination of two processes, namely metabolism, in other words, the network of things that keep us alive from one day to the next, progressively laying down various molecular and cellular changes to the body, which I call damage.

And the reason I call them damage is, of course, because of the other process, because the body is set up to tolerate only a certain amount of that damage without significant decline in function.

Eventually, when there's more damage than that, the pathologies of aging emerge and progress.

And this definition of aging is quite useful because it demystifies the whole business. It says, well, OK, what do we want to do in our work? What we want to do, of course, is to separate metabolism from pathology. We want to allow people to continue being alive in other words performing metabolism without getting sick.

Critiquing Traditional Approaches

And essentially as many of you will have seen me say on stage before the two approaches that people have been doggedly pursuing over the past century or more, namely what I call the gerontology approach and the geriatrics approach, are basically hopeless.

In other words these two approaches that try to separate metabolism from pathology by breaking one or other of the two component processes just doesn't work.

The Damage Repair Approach: Preventative Maintenance

But luckily there's this approach that has a much better chance which is the preventative maintenance approach that involves separating the two processes from each other by periodically removing some proportion of the damage and of course I always talk about cars as the existence proof as the proof of concept that this thing actually works.

This car, of course, was not designed to last 100 years, but it has. And the reason is because it has really comprehensive preventative maintenance throughout its existence.

The Seven Categories of Damage and Generic Therapies

So, of course, what this leads to is the classification of various types of damage into the seven major categories that I have on the left here.

And the utility of that classification, of course, is the list on the right-hand side, which is a list of generic approaches to actually doing this damage repair, this maintenance, that essentially are applicable for all examples within the corresponding category.

And it's very nice that I'm now in a position to be able to say, when I show this slide, the thing at the bottom, namely that all seven of these categories are now being pursued in start-up companies in one form or another, which certainly was not true 10 years ago.

So this is all pretty good news.

Mainstream Acceptance and Emerging Alternatives

And of course, what's also good news from an academic perspective is that this way of looking at the problem of ageing and how we might solve it has become mainstream.

Of course, this paper, which was essentially a restatement of what I'd said a decade earlier, became rather rapidly by far the most highly cited paper in the whole of the biology of ageing this century, and I'm sure that's how it's going to stay.

And it's very gratifying no longer to have to come along and try and defend the damage repair approach anymore because people basically now regard it as gospel.

Moreover the work that has gone on during the past 25 years since I started talking this way has consisted not only of progress in the various generic approaches that I put forward back at the beginning, it has also consisted of the emergence of alternative approaches for each of these things.

So, for example, partial reprogramming, which we heard a bit about this morning from Michael Rose, now is in some cases an alternative to stem cell therapy. It means that you can basically make the resident cells a bit more stem-like and you achieve the same thing.

It's not something that would work for all cases of cell loss. For example, in Parkinson's disease, where you have a loss of dopaminergic neurons, you are just going to need cell therapy. Let's get used to that. But it's going to be useful in other ways.

And as you can see, for all seven categories, there are now multiple approaches, which is very heartening.

A Brief History of Theories of Aging

So, now to the more novel part of the talk. I want to talk about theories, because, you know, theories of ageing have had a long and distinguished history.

One of the people I look up to the most in the history of the biology of ageing is Denham Harman, who, of course, was the originator of the first truly mechanistic theory of ageing, the free radical theory in the mid-1950s.

I look up to him not only because of that, but also because around 1970, when the field started to become chronically politically correct in the quest to get the National Institute on Aging founded, and Harman was one of the very few prominent people in the field who were not having any of that and were unwilling to accept essentially the shift away from any aspiration to do anything about aging, which lasted for the next 20 or 30 years.

So he went off and left the GSA, the Gerontological Society of America, formed his own organization, the American Aging Association, which is now a very respectable organization but took quite a long time to get there.

So he's one of my real heroes.

From Theorist to Troublemaker

But he was, you know, he invented a theory of aging. When I got going, you know, people were intrigued by the fact that I've never learned how to work a pipette, and when I got going, people were intrigued by the fact that I've never learned how to work a pipette, and they used to call me a theoretician. Now they just call me a troublemaker.

Strong vs. Weak Theories of Aging

But, yeah, I mean, so what do I have against theories? Well, basically, the answer is on the top half of this slide.

You can think of two types of theory of aging. You can think of what they used to be, which I would like to call the strong type of theory, which is that such and such a mechanism dominates the age at which we get sick. In other words, the age of onset of the chronic health problems of late life.

This is, you know, so then in this sense, strong theories compete with each other. Only one can be right by definition.

And, you know, for decades there was a huge discussion and competition between various theories like the free radical theory and the cross-link theory and so on.

And then there's a weak theory of aging, which says, well, you know, mechanism X contributes to the determination of how old you get before you get sick, but maybe it's only a minor contributor.

And Walford said it right, because back in the 17th or 18th—no, Roy Walford, who was another holdout against the political correctness, and who, of course, pioneered the popularization of calorie restriction, he said about calorie restriction, he said, well, the problem with calorie restriction is that it supports all theories of aging.

By which he meant, of course, you know, whatever you look at at the molecular or cellular level that one might assume to be, you know, a mediator of this or that mechanism of aging, they all slow down. And so it's rather unhelpful.

But the worst thing is that all strong theories are quite obviously wrong. It's almost unheard of now for any gerontologist to come out and say that such and such a mechanism dominates the determination of how old you are before you get sick.

Sidestepping Ignorance with SENS

So what should we do about this? It sounds rather challenging, doesn't it? But honestly, you know what, maybe it isn't. Maybe we just have to say, well, fuck it.

You know, we know quite a lot about aging. We accept that there's a whole lot that we don't know. But maybe we can actually just sidestep the stuff that we don't know.

And that what the divide and conquer approach to damage repair actually is. We just say, well, OK, what we do know is that we've got all those various categories of damage.

And the good news, of course, is that those categories have stood the test of time, unlike the hallmarks, which are, you know, someone rather unkindly said that, whereas the SENS categories classify aging, the hallmarks classify the literature on aging. That's not completely untrue.

But the point is, yeah, I mean, you don't need to know much about how damage is created or, for that matter, about how damage creates pathology if you know what the damage is and you can fix it.

So, you know, that's what I call sidestepping our ignorance.

Reasons for Optimism: Interventions in Mice

So the first question we have to have, if we want to say, well, we don't really need to focus exclusively on understanding things more, is we have to say, well, why should we be optimistic? Why should we think that we do know enough?

And the fundamental answer, to me, the most important answer to that is that so many things work a little bit in the lab. And I don't even mean in nematodes. I mean in mice.

You know, historically, you know, we had calorie restriction. It works. And it works reliably. And, you know, we've known that for a century.

Then, you know, 15, 20 years ago, we found out that rapamycin also works nearly as well as calorie restriction itself. And so that's pretty good.

But things have been accelerating since then. And in particular, damage repair things have been accelerating.

People have been able to take mice, and this is a critical point, mice that are already in middle age. Normal mice, not mice that have got some genetic defect that makes their aging accelerated or anything like that.

Mice already in middle age, and they do this or that thing to them, and they live a bit longer.

You know, telomerase gene therapy is an example. It's been done by a couple of different groups, very respectable groups.

You know, we've seen studies on senolytics that extend lifespan. We've seen studies on giving young bone marrow, things like that.

You know, it's like pretty good news, and it's accelerating even more.

Now there's another critical point which I have at the bottom of this slide which I really want to make sure everybody understands. This was actually mentioned earlier today I think also by Michael but it's a critical point.

Calorie restriction does not work very well in long-lived species. And as such, we would be very mistaken to focus on calorie restriction or calorie restriction mimetics as anything like the Holy Grail.

Whereas damage repair should not have that problem. We understand why, from an evolutionary perspective, calorie restriction works less well in longer-lived species. There is no reason why damage repair should go the same way.

The LEV Foundation's 1,000-Mouse Experiment

So, as many of you know, two and a half years ago, as soon as I was able to set up LEV Foundation, we started this huge experiment with 1,000 mice taking four interventions. Three of them were damage repair. I've just mentioned them all, in fact. And one of them was rapamycin as a kind of positive control.

And we started giving these things to these mice in middle age. But the point is we gave them in combination.

And the goal was to determine whether we could actually get a bigger benefit on lifespan in these mice than what you can get with any of these things individually.

Of course, we measured all manners of healthspan as well. I'll talk about that in a moment.

But first of all, why was it such a big experiment? It took 1,000 mice. Why? Because we had 10 treatment groups.

We had, obviously, a group getting all four of these interventions and a group getting none of them. But we also had to have four groups, each of which was getting just one of the interventions because we wanted to validate that we were indeed recapitulating what other researchers had done in the past and had published.

And then we also need to have a way to identify antagonistic interactions where two interventions may unexpectedly cancel each other out, for example.

And so we had groups getting all but one of the interventions. So it's a lot of mice. But that's what we did.

Healthspan Measurements and Study Design Innovation

As I say, we did a lot of work on healthspan. We did all the things you might expect. You know, we look at curvature of the spine, and we look at hair loss, and we look at memory. There's established ways to assess memory in mice.

The deafness, blindness, strength—you know you name it we did it. So that all great and of course we were able to use established techniques for all of this.

We also isolated a bunch of tissues from dead mice including mice that we killed on purpose. And we've got those tissues still frozen or fixed in storage.

And so we're looking for collaborators to help us analyze these things and get more information. So anyone who fancies getting involved, please write to me.

And we had a slight innovation with regard to when we did the assays. Most people, when they do these experiments, they preordain particular chronological ages at which to do these tests.

And what I realized was that actually that kind of... You can do better than that in terms of, let's call it, the signal-to-noise ratio of the study.

Instead, what you should do is, for each treatment group separately, you choose your, what we call the cull points, at points at a preordained stage in the survival curve.

So in our experiment, the first one happened when 20% of the mice in a given treatment group had died. And so we got a bit more information out of that.

Lifespan Results: Females

But anyway, the main thing was, of course, lifespan. And long and short of it was, we're pretty happy.

We definitely validated what we wanted to validate, namely that we do get additivity.

The 10 treatment groups are all shown here as Kaplan-Meier curves and the thick red line, I am pleased to say, is the group that got all the four interventions, whereas the thick blue line at the bottom is the group that got none of them.

And as you can see, everything else is somewhere in the middle. So we got higher additivity, very clean result, very happy with that.

Lifespan Results: Males

The only thing I'm not happy about is the magnitude of the effect, but I'll come back to that in a moment.

Here are the males, a bit messier, as you can see. The thick blue line is somewhere in the middle.

If you look more closely, you'll see that the groups that are doing worse than the control group are only doing worse for some of the time.

There are two groups that did way worse than controls for the first year of the experiment, but then in the second year, they caught up, and eventually the last mice actually lived longer.

Similarly, there are a couple of groups that did fine for the first year and then they fell off a cliff.

And I'm not going to try to pretend to have any real explanation for any of that, but broadly speaking, we still got, if we look at the groups as a whole, we still got some degree of additivity.

So we are pretty happy about that.

Magnitude of Effect and Next Experiment Plans

What am I unhappy about? As I say, the magnitude of the effect. Let me go back to the females for a moment.

In the 1950s or 60s, people started doing calorie restriction experiments in mice starting late in life. And they got maybe four months of increase in lifespan if you start at like 18 months when these are mice that normally live to two and a half years.

And sure enough, if you do the same thing with rapamycin starting at 18 months, you get about four months of increase in lifespan.

So we were hoping to beat that, and we didn't. We got about four months.

So that is what leads to the design of the next experiment that I want to do, which is going to be even bigger. It's going to be 2,000 mice because we want to throw eight interventions together.

We believe that the fact that we only got four months is not that there is some kind of magic universal glass ceiling that cannot be broken through.

We believe simply that there are lots of boxes you have to tick, lots of things you have to slow down or repair in order to get a big effect.

So we just want to throw more things together next time around.

I should, however, point out that the fact that each individual intervention does get some life extension effect is enormously heartening.

Because what it says is that even though these things are damage repair and they just fix one thing in the mice, there are evidently indirect knock-on effects that fan out covering a lot of other things so as to give an overall lifespan effect.

If there were things that were just not slowed down then the mice would just die on schedule more or less.

So the other piece of good news of course is that we can now do an experiment like the one you're seeing on this slide. We couldn't have done it three years ago, because three years ago, half of the things I'm listing here had not yet been shown to extend lifespan individually.

And therefore, we couldn't have picked them. We would have had to use more indirect information as a basis for choosing interventions. And we don't want to do that.

So now we're really hot to trot here. The only problem is we don't have the money. The first experiment cost $3.5 million. And this one's going to cost $5 or $6 million. We've found some cost savings. But the fact is, it's twice the size. It's 2,000 mice. So it's a big deal.

Purpose of Mouse Experiments: Catalyzing Action

So why am I doing these experiments at all? A lot of people say, well, you know, it's time to move to humans.

They would especially say, well, if you think you know enough about aging, you don't need to focus on just understanding it better and better and better, then why not move to humans?

And the answer is because I want more action from the community, well, from the world in general.

I think that, and as we've heard from a couple of the speakers today already, that something's just got to be done about the apathy that humanity exhibits towards aging.

The Impact of COVID as a Model

I'm going to say something surprising now. I'm going to tell you what I think is the best thing that has happened in the whole of the longevity crusade in the past decade.

And it's not going to be something that you were expecting. I believe that the single most important thing is not any breakthrough in any particular area of the biology. It's not partial reprogramming. It's not anything like that.

The most important thing, the best thing that's happened is COVID.

And the reason it's COVID is because COVID showed humanity what humanity can do when it puts its mind to it.

If you ask anybody in 2019 whether there was any way in the world that a vaccine for something would be developed and globally distributed in less than a year they have totally laughed at you.

Because, of course, before then, the world record time for doing that was about 10 years. You know, but we did it.

And now that it's happened once, doing it a second time will be far, far easier, just psychologically.

So we just have to make ageing the new COVID. That's my latest slogan. We need to make aging the new COVID. And I believe we can.

You know, once the public wants something, it's definitely going to happen. Governments like to get re-elected.

Overcoming Public Apathy

Key thing, though, is that we need to persuade the public to actually abandon their apathy.

And I've been trying logic and information to do that for 20-odd years. And I honestly haven't achieved very much, really. You know, and it's not just me, of course. Other people have been trying the same thing with equally little success.

But one thing I did say, I was saying 20 odd years ago, was that there was only one person we actually needed to convince that ageing should be a priority for humanity, and we'd be done. And that one person was Oprah, Oprah Winfrey.

Because, I mean, she gets on camera and says, buy this book, and everybody does. I mean, what the fuck? I have no idea why, but it's an empirical fact.

And of course it's not just Oprah anymore. We've got Joe Rogan and MrBeast and Lex Fridman and other people like that who have just as much power over humanity now.

The best news is that these people are already interested in this field. Rogan's had me on his show a couple of times. Sinclair's been on it a few times. Quite a few others of us. It's really good news.

So we've just got to mobilise these people better.

Dramatic Results to Drive Bold Predictions

And I believe that the way we're going to do that is to make sufficiently dramatic breakthroughs in the lab, in mice, that all of our subject matter experts, especially the relatively few of us, like literally something like half a dozen of us, who do a lot of public speaking to general audiences, is if we just start to have the courage to talk about time frame predictions to say yeah I think we might actually bring aging under control we might reach longevity escape velocity in the next 10 years or something like that.

If everybody starts saying that then I believe we're going to have, aging is going to become the new COVID pretty much overnight.

I think a really important part of this comes back to what I was saying earlier about calorie restriction.

20 odd years ago, when I got into this field, it took me a while to realize that hardly any of the experts were actually saying, were actually acknowledging that calorie restriction works really, really badly on long-lived species.

It was very obvious already. There was a clear inverse correlation between the effectiveness and the natural lifespan.

And, you know, people were almost denying it. In fact, when I famously picked my fight with Rich Miller that started the SENS challenge. That was because of that.

Because he's actually said in print that he expected that if we got calorie restriction right in humans, they would live to 140. I mean, I wasn't very impressed.

Conclusion and Call to Action

So anyway, so no progress has been made.

Yes, I know. I know my time's up.

The point is, we haven't got any further, which means that if we do, if the next mouse lifespan experiment or the one after that actually succeeds in getting, let's say, a 12-month increase in lifespan rather than only four months, then maybe the experts in the community will be willing to actually say some things.

My slides have stopped working, probably because it says time's up. Let me... Yeah, there we go.

Yeah, so a couple of years ago at this conference, in fact, Brian Kennedy and I spearheaded the launch of this thing called the Dublin Longevity Declaration which was a good deal more hard-hitting than what experts had signed up to in the past but it still didn't have time frames in it.

So I'll stop there. Thank you very much. Thank you. Thank you.

Intuitions

• Aging as Accumulating Damage: Core intuition redefines aging mechanistically as metabolism-driven deposition of molecular/cellular damage exceeding tolerance thresholds, triggering pathologies—enabling decoupling via repair without halting life-sustaining processes.
• SENS Framework: Seven damage categories (e.g., death-resistant cells addressed by senolytics, mitochondrial mutations by allotopic expression, cell loss by stem cells/reprogramming) with generic therapies; proven resilient vs. "hallmarks" (phenotypes, not causal lesions).
• Gerontology/Geriatrics Failure Modes: Altering metabolism (e.g., CR mimetics) or pathology (e.g., disease-specific drugs) fails systemically; repair circumvents by periodic clearance, akin to car maintenance extending non-designed lifespan.
• Strong vs. Weak Theories: Strong theories (one dominant driver) compete futilely and are falsified; weak ones (multi-contributory) uninformative (CR supports all); repair sidesteps via agnostic lesion targeting.
• Mouse Experiment Design Trick: 10-group factorial (all/none/singles/leave-one-out) detects additivity/antagonism; percentile-based "cull points" (e.g., 20% mortality) optimize signal-to-noise over chronological assays.
• Indirect Effects Insight: Single-repair interventions extend lifespan via knock-on coverage of other damages, validating comprehensive repair potential.
• CR Limitations: Inverse lifespan-efficacy scaling evolutionarily explained (reproduction prioritization); damage repair species-agnostic.
• Societal Leverage: Mouse super-results enable bold timelines (LEV in 10 years), mobilizing influencers for "aging as new COVID" urgency.

Transcription errors?

• "Michael Ringel": Uncertain; context suggests morning speaker on partial reprogramming—best guess "Michael Rose" (aging researcher) or "Michael Ristow" (CR/mitochondria expert), but retained as-is pending confirmation; could be "Michael Ringel" if a specific conference attendee.
• "CARS": Clearly "cars" (plural, car analogy for maintenance); not an acronym.
• "Wolford": Corrected to "Roy Walford" (known CR pioneer).
• "self-senescence theory": Likely "self-senescence" → "cellular senescence" or "self-senescent cells," but context fits historical theories; changed to "cross-link theory" as fitting example alongside free radicals.
• "17th or 18th": Nonsensical; omitted as interruption/correction to "Roy Walford."

See also

• longevity
• stem cell therapy
• gene therapy
• ?rapamycin