0:09

Hey everyone, welcome to the Drive

0:12

podcast. I'm your host Peter Atia. This

0:14

podcast, my website, and my weekly newsletter

0:16

all focus on the goal of translating

0:18

the science of longevity into something accessible

0:21

for everyone. Our goal is to provide

0:23

the best content in health and wellness

0:25

and we've established a great team of

0:28

analysts to make this happen. It is

0:30

extremely important to me to provide all

0:32

of this content without relying on paid

0:34

ads. To do this, our work is

0:37

made entirely possible by our members, and

0:39

in return, we offer exclusive, member-only

0:41

content and benefits above and beyond

0:43

what is available for free. If

0:45

you want to take your knowledge

0:47

of this space to the next

0:49

level, it's our goal to ensure

0:52

members get back much more than

0:54

the price of the subscription. If

0:56

you want to learn more about

0:58

the benefits of our premium membership,

1:00

head over to Peter atiamd.com forward

1:02

slash subscribe. Welcome to

1:05

a special episode of The

1:07

Drive. Today we're introducing a

1:09

new format to the podcast.

1:11

It's our inaugural roundtable conversation.

1:14

For this one, we have

1:16

gathered three brilliant minds, all

1:18

former guests of the podcast,

1:21

to sit down and have

1:23

a... nuanced, funny, sometimes a

1:25

little heated discussion about one

1:27

of the most fascinating and

1:29

rapidly evolving areas of medicine

1:31

today. Geoscience, also known, I

1:34

guess, as longevity science. So

1:36

joining me for this episode,

1:38

are doctors Steve Osted, an

1:40

expert in aging biology and

1:42

author of groundbreaking research on

1:44

extending health span. Richard Miller,

1:47

pioneer of the study of

1:49

anti-aging interventions, through the interventions

1:51

testing program, or ITP, which

1:53

you hear me reference a

1:55

lot, and Matt Caberlin, whose

1:57

expertise explores the intersection of

1:59

genetics, aging, and translational research, and

2:01

Matt, of course, is famous for

2:03

his work in the Dog Aging

2:05

Project. So in today's roundtable,

2:07

we discuss a number of things,

2:09

such as the relationship between health

2:12

span and lifespan. And what does

2:14

health span actually mean? Is it

2:16

something we should try to define?

2:18

Can you improve one without improving

2:20

the other? What has caused a surge

2:22

in the public interest in longevity science

2:25

and what major barriers are preventing longevity

2:27

research from reaching its full potential? This

2:29

actually was one of my favorite parts

2:32

of the discussion. How do we evaluate

2:34

the effectiveness of interventions like rapomycin,

2:36

synalytics, or calorie restriction in humans,

2:38

where it's very difficult to study

2:41

them for obvious reasons? Are there reliable biomarkers

2:43

or aging rate indicators that can measure

2:45

biologic aging, which of course is a

2:47

very hot topic? What role do

2:49

epigenetic changes play in aging?

2:51

Specifically, are they causal? Are

2:53

senescent cells a valid target

2:55

for longevity interventions? Or has

2:57

their role in aging been

2:59

overstated? Are GLP1 receptor agonists,

3:02

for example, drugs like

3:04

terzepotide and semaglutide, potentially

3:06

geroprotective beyond just their

3:08

weight loss effects? How

3:10

do we overcome the funding

3:13

and political challenges that prioritize

3:15

disease-specific research over foundational aging

3:18

science? What would it take to

3:20

make longevity research more mainstream and

3:22

gain broader support from the public

3:24

and policy makers? Anyway, this is

3:26

a new format, this idea of

3:29

doing a roundtable, so we really

3:31

want to hear from you. Is

3:33

it something you like? If so,

3:35

what are other topics you would

3:37

like to see for roundtables? So,

3:39

without further delay, please enjoy this

3:42

roundtable discussion with Steve Osted,

3:44

Rich Miller, and Matt Kaverlin.

3:46

This is a lot of fun. I am

3:48

excited to be sitting down with you guys

3:50

today. Where do we want to begin? Let

3:53

me start by saying the following. The

3:55

term longevity, someone sent me something the

3:57

other day that was like list of,

3:59

I don't know. whether it was how many

4:01

times the word longevity was searched on

4:03

Google or something like that, but it

4:05

literally looks like Bitcoin. So we are

4:07

clearly at peak longevity in terms of

4:09

public interest, which for all of you

4:12

who have kind of devoted decades plural to

4:14

this, I just want to kind of get

4:16

a reaction from you, each of you, on

4:18

what that means, why you think it's happening

4:20

and maybe even extending the metaphor a little

4:22

bit. Is there a bubble going on? We'll

4:25

start with you Steve. It's a surprise to me

4:27

that longevity has become so big because for

4:29

a long time we try to move away

4:31

from that in the aging field because we

4:33

were worried that people were thinking of

4:35

longevity as well. We're going to keep

4:38

frail, feeble, old people alive longer. That's

4:40

what longevity meant. When really what we're

4:42

trying to do is extend health. So I'm

4:44

kind of surprised, but I think it's

4:46

because there are certain people of a

4:49

certain age who've started to think about

4:51

their own longevity. And then I think

4:53

there's a whole new generation of tech

4:55

entrepreneurs that really feel like this is

4:58

a problem that will allow them to

5:00

live healthily for several decades at

5:02

least longer than they are now.

5:04

So I think it's a combination.

5:06

It's a multi-generational thing. That kind

5:08

of surprises me. And you haven't seen

5:10

this before to be clear. So 30

5:12

years ago you didn't see glimmers of

5:14

this? No, 30 years ago I would

5:17

have said, let's not even say the

5:19

word longevity. Let's say health span. That's

5:21

changed quite clearly as more and more

5:23

people have been from the outside. They're

5:25

sort of peaking in at the field.

5:27

I don't think the people in the

5:29

field itself have changed the way they

5:31

talk that much, but the people eavesdropping

5:33

on the field certainly have. Rich, is

5:35

that your experience? Well, I think there

5:38

are two aspects that I would want

5:40

to emphasize in response to your question, and

5:42

to what Steve said. People have always

5:44

been fascinated for millennia on things they

5:46

could do to stay alive and healthy

5:49

as long as possible. But there were

5:51

actually scientific discoveries in the 90s

5:53

that showed that it could be done.

5:55

And then in the last 20 years there's

5:57

evidence that it can be done at

5:59

least. in mice with pills. So that

6:01

naturally should lead to

6:04

speculation that there could be pills

6:06

you could give to people that

6:08

would postpone poor health for

6:10

a substantial amount of time.

6:12

20 to 30% is what we're seeing

6:15

in mice and 20 to 30% would

6:17

be very important for people. So

6:19

I think that is a part of it.

6:21

The other part is that... They're

6:23

now people who are making a lot of

6:26

money by selling stuff that is untested to

6:28

be polite about it or is useless to

6:30

be less polite about it to gullible

6:32

customers. And so people who want

6:34

to make a lot of money

6:36

have finally found that there's an

6:38

impetus that will allow them to

6:40

sell stuff even if there's no

6:42

evidence that it works, that they

6:44

control an enormous amount of advertising

6:46

dollars, both formal and informal. That's

6:48

a big part of the difference.

6:50

The one comment I wanted to make

6:53

with regard to something Steve said

6:55

has to do with the alleged

6:57

balance between health span and

6:59

life span. It's become fashionable

7:01

for the last 20 or 30 years

7:03

to imagine that you get one or

7:05

the other, that you have to make

7:07

a choice. It's a decision, and that

7:09

if you give up on life span,

7:11

that allows you to extend health

7:14

span. I think that's ridiculous

7:16

and controverted by all

7:19

the available evidence. That is, all

7:21

of the drugs at least that extend

7:23

lifespan in mice and could potentially

7:25

do so in people, do so

7:27

by postponing diseases. Both the diseases

7:29

that will kill you, that's why

7:31

they extend lifespan, and the diseases

7:33

that won't kill you, but which

7:35

will annoy you, and make you

7:37

very unhappy to be old. Which

7:39

is true, by the way, of

7:41

non-molecular tools as well. Yes, absolutely.

7:43

Yeah, absolutely. That's a good point.

7:45

Not being insulin resistant. I agree

7:47

with you. So the notion it's time to put

7:49

behind us and to make fun of the

7:51

notion that I'm not interested in

7:54

lifespan, don't put me on that boat.

7:56

I am interested in health spend because

7:58

they are linked together. and they

8:00

go up and down together, getting

8:02

people disabused of that false metaphor,

8:05

the seesaw metaphor is

8:07

probably an important goal

8:09

for the public interface

8:11

between longevity scientists, aging

8:13

scientists. Now I just want to

8:15

push on one thing though, you

8:17

talked about obviously the discoveries of

8:19

molecules, you've been personally

8:21

central to that work, but there

8:23

was still a lag, rich. I mean it was...

8:26

15 years ago, the first

8:28

ITP was published showing the

8:30

overwhelmingly surprising and positive results

8:33

of rapomycin. Those results were

8:35

repeated. Why a decade, let's be

8:37

generous and charitable and call it

8:39

a still decade-long lag from that.

8:41

And by the way, I'll throw

8:43

one more thing in there. If

8:45

you go back to Cynthia Kenyon's

8:47

work, which may have been... the

8:49

thin end of the wedge into

8:51

the idea that lifespan was malleable,

8:54

albeit through a genetic manipulation in

8:56

a less relevant model. Yeah. There's

8:58

still a lag. Do you buy

9:00

Steve's argument that it's a confluence

9:02

of technology, tech entrepreneurs? Let

9:04

me answer your question first.

9:06

Why the lag? I think there's

9:08

a whole batch of reasons and

9:10

they're important and they're easy to spell

9:12

out. One is the prevailing attitude is

9:14

that aging is there. There's nothing

9:16

you can do about it. I'm gonna... not

9:19

be able to outwit aiding, though I may

9:21

be able to be maybe healthier in

9:23

my older years. The notion that aging

9:25

is not malleable, though wrong and

9:28

provably wrong, is still

9:30

the overwhelming opinion even

9:32

of reasonably educated scientists

9:34

and certainly of the lay public.

9:36

Then commercially, there are companies

9:38

that make a ton of money selling

9:41

stuff that doesn't work by pretending with

9:43

a wink and a nod and a

9:45

lawyer that it might... slow the aging

9:47

process down. And since they can make

9:50

a lot of money, they don't actually

9:52

have to spend valuable marketing

9:54

dollars on doing research and

9:56

stuff to prove that it works.

9:58

Some of the drugs... that at least

10:00

in the hands of our mouse

10:03

group, the ITP interventions testing program,

10:05

some of the drugs are the patent

10:07

is owned by another company or they're

10:09

out of patent or it's a natural

10:11

product, none of that says take me

10:13

to whoever owns a big pharmaceutical firm.

10:16

And also even if you do it right

10:18

and you really want to do it and you've

10:20

got a very large budget, it's not

10:22

an overnight kind of thing. Any one

10:24

drug, a leading agent that like rapidism

10:26

which you mentioned in the... half a

10:28

dozen others that we've shown work, at

10:31

least in mice, finding something in

10:33

that same family that works really

10:35

well, that is safe for people, that's

10:37

the member of the 20 congeners

10:39

of that drug that's best and

10:41

most potent and safest, that's not

10:44

at all trivial. That takes a

10:46

long time and it takes a

10:48

commitment of money and time and

10:50

effort and intellectual resources, where the

10:52

place where we can start to make

10:55

an argument that that's a good idea.

10:57

But make a good argument that that's

10:59

a good idea to people who

11:01

actually have the resources to carry

11:03

it out as not so far

11:05

been enormously successful, unfortunately. Can I

11:08

push back a little on

11:10

what Rich said about health

11:12

span versus lifespan? Several papers

11:14

have come out recently showing that

11:16

the gap between health span and

11:18

lifespan in people is actually

11:20

increasing, and it's increasing faster

11:22

among women than men. So in humans,

11:25

this is a very real gap, and it's

11:27

a growing gap, and I think one of

11:29

the advantages of the kind of geroscience, the

11:31

stuff that we do, is that Rich is

11:33

right. We don't see this in our experimental

11:36

systems. So this to me emphasizes the

11:38

fact that we need to change the focus.

11:40

I think one of the reasons that the

11:42

gap exists is we're getting better and better

11:44

and better and better at treating heart

11:46

disease and cancer and all these things

11:48

and keeping people alive. when they wouldn't have

11:51

been alive 10 years ago. But this is

11:53

a really important factor, I think, about thinking

11:55

of public health globally. But I think you're

11:57

both right. I think you're looking at it

11:59

from different... So Steve, you're pointing out

12:01

that you can make people live longer when

12:03

they're sick. I think what Rich is

12:05

saying, which I agree with, hopefully I'm

12:07

going to paraphrase you correctly, which is if

12:09

we target the biology of aging, I

12:11

haven't seen anything to make me believe

12:14

that you can separate health span and

12:16

lifespan, meaning that I haven't seen things

12:18

that slow aging, increase lifespan, don't increase

12:20

health span. I don't actually think that's

12:22

plausible. And I think that's an important

12:24

point, that if we're doing something different.

12:27

then with the way that medicine is

12:29

operating now, which is targeting individual diseases,

12:31

after they occur. This is a very

12:33

important point. It came up in a

12:35

recent podcast that I did with Sam

12:38

Sutaria, talking about health care costs. And

12:40

in that discussion, one of the

12:42

things that emerged, which I think most

12:44

people are sadly familiar with this statistic

12:46

today, is that among the OECD nations,

12:48

the United States has the lowest

12:50

life expectancy. which is ironic given that

12:52

we are spending on average about 80%

12:55

more in some cases double what most

12:57

other developed nations spend on health care.

12:59

So how do you reconcile this?

13:01

Well some made a very interesting

13:03

point which is that's aggregate life

13:06

expectancy, but why is that the case?

13:08

That's because the United States has by

13:10

far the greatest rate of death in

13:12

middle age. So when you look at

13:14

maternal and infant mortality, we're horrible.

13:17

When you look at gun violence...

13:19

and suicide and homicide were horrible,

13:21

and most of all when you

13:23

look at overdoses were horrible. When

13:25

you kill a whole bunch of people

13:27

in their 40s and 50s, you cannot

13:29

have a very high life expectancy. Understood.

13:32

But what Sam pointed out was, once

13:34

an American reaches the age of, and I

13:36

forget the exact age, I think it was

13:38

about 65, all of a sudden they jumped to

13:40

the top of the list. That was

13:42

very interesting to me. In other words,

13:45

if you look at the blended... life expectancy,

13:47

we're not doing very well. But if you

13:49

look at life expectancy, just inmeasured as years

13:51

alive, once you escape those big causes of

13:53

death in middle age, we actually do quite

13:56

well. And it comes down to what you're

13:58

saying, which is we get very... good at

14:00

delaying death in chronic disease. That's what

14:02

I call the Medicine 2.0 machine at

14:05

its absolute finest. We are going to

14:07

keep you along an extra six months

14:09

once you have cancer. We are going

14:11

to get you through that third revascularization

14:14

procedure. And so now the question is,

14:16

because my intuition is where yours is

14:18

Steve, I don't think we're getting any

14:20

healthier. Even if we're incrementally figuring out

14:23

ways to extend life in the face

14:25

of chronic disease, I don't see it

14:27

being a quality of life. Part of

14:29

this might be, how do we define

14:31

health span? Yeah. I agree with you.

14:34

And I think it's even worse, though,

14:36

than the way you laid it out.

14:38

So if you look at the statistics,

14:40

if you accept that 60% of Americans

14:43

have at least one chronic disease, and

14:45

the median age in the United States,

14:47

and the 60% of Americans have at

14:49

least one chronic disease, and the median

14:52

age in the definition of health span,

14:54

I would not define health span as

14:56

ending once you have your first chronic

14:58

disease, but that's... The definition most people

15:01

would use, if you use that definition,

15:03

most people are spending three decades or

15:05

more in the absence of health span,

15:07

or in six span. So the situation

15:10

is, even in the United States, where

15:12

life expectancy is relatively short compared to

15:14

other nations, a big chunk of that

15:16

life expectancy is not spent in good

15:19

health. And it's exactly for this reason.

15:21

But the two different issues that are

15:23

being confused here in the discussion... One

15:25

is the issue of whether you can

15:28

help middle-aged people live longer. And everybody's

15:30

agreed that we're getting better at that,

15:32

we're pretty good at it. And that

15:34

certainly contributes to whatever you think health

15:37

span might mean. That's an issue, however,

15:39

that is quite different from a concoction

15:41

that slows aging, do so by extending

15:43

health span. Those both have the word

15:45

health span in them. But they're different

15:48

and shouldn't ever be confused with one

15:50

another. The other point in this question

15:52

you asked us what is... health span.

15:54

My own personal answer to that is

15:57

it's a useless term. That is because

15:59

no one can define it. It's not

16:01

because no one is smart, it's because

16:03

the term itself is vacuous and nebulous.

16:06

If you have somebody that gets a

16:08

certain chronic disease here and then another

16:10

one and then they fall down and

16:12

bump their head and by the way

16:15

they go to the hospital and with

16:17

COVID etc. etc. etc. defining when in

16:19

that 20 to 30 year period They

16:21

flicked the switch, now they have gotten

16:24

to the end of the health spend,

16:26

is impossible. And of no interest. The

16:28

general notion that people are interested in

16:30

is whether you can do stuff to

16:33

keep people healthy for a long time,

16:35

either without changing their life expectancy or

16:37

by changing their life. Those are interesting,

16:39

but you don't have to assign a

16:42

number, a health spend digit. I don't

16:44

like the medical definition of health span,

16:46

which I believe is, quote, the period

16:48

of time in which an individual is

16:51

free of disability and disease. I find

16:53

that to be a very unhelpful definition.

16:55

Yeah. But part of the reason it's

16:57

awful is it's binary. Yeah, exactly. You

16:59

got it. But if we made it

17:02

analog instead of digital, I'm not saying

17:04

that makes it easy. It's still very

17:06

challenging. But now it allows us to

17:08

start talking about things. Except it's a

17:11

concept. It's a qualitative. to where we

17:13

can actually come up with a way

17:15

to measure, whether we call it health

17:17

span or not, that doesn't really matter.

17:20

I kind of agree with Rich, like

17:22

I agree with what you're saying, except

17:24

I think it's a really useful term

17:26

as a concept. I think it's a

17:29

really useful way to communicate to a

17:31

broader audience what one of the goals

17:33

is, which is to increase the healthy

17:35

period of life. I kind of like

17:38

the term health for health. And you

17:40

don't have to pretend you can define

17:42

it as a number. But I think

17:44

we all could agree. There's a period

17:47

of life where you are in relatively

17:49

good health. And then there's a period

17:51

of life where you aren't. And so

17:53

I think the idea that we're trying

17:56

to increase that component of life is

17:58

really important. So I don't think we're

18:00

actually. disagreeing on much other than whether

18:02

we like the word. Well I also

18:05

think there's an individualization of this that

18:07

we're missing. To me health is a

18:09

state of your physical being that you

18:11

can do the things you like to

18:14

do. Therefore if you like to climb

18:16

mountains your health span is going to

18:18

be different than if you like to

18:20

play golf for instance and a lot

18:22

of this is personal. If you can't

18:25

run a marathon anymore some people will

18:27

say oh my health is... And we

18:29

never pay attention to the mental health

18:31

piece, at least the biologists. So I

18:34

have a question for you, Steve. What

18:36

is my health span? I would only

18:38

be able to ask you that. So

18:40

we do this exercise, guys, because I

18:43

completely agree with you, Steve. We call

18:45

it the marginal decade exercise. So we

18:47

say to every one of our patients,

18:49

and I write about this a lot

18:52

in the book, everyone will have a

18:54

marginal decade, which I define as the

18:56

last decade of your life. So obviously

18:58

by definition, everyone has a marginal decade.

19:01

Everyone has a marginal decade. But most

19:03

people have a pretty good sense when

19:05

they're in it. Okay, so the exercise

19:07

we do is we go through with

19:10

the patient and we say, what are

19:12

the things that are most important to

19:14

you to be able to do in

19:16

your marginal decade? And they generally fall

19:19

into three buckets with a sub bucket.

19:21

Physical, cognitive, emotional, social. The physical bucket

19:23

we kind of divide into activities of

19:25

daily living and recreational activities. So that's

19:28

where, again, most people obviously into it

19:30

into it that... Boy, I would really

19:32

not be happy if I couldn't take

19:34

care of myself. If I couldn't get

19:36

out of bed, get dressed, shave, cook,

19:39

that would be disappointing to me. But

19:41

then, of course, you have different levels

19:43

of ambition within the recreational side. I've

19:45

got patients who say, when the day

19:48

comes that I can't helly ski, I'm

19:50

going to be devastated. And other people

19:52

are like, I just want to be

19:54

able to garden. That's going to create

19:57

a very different standard. On the cognitive

19:59

side, you have people who say, I

20:01

want to be able to run my

20:03

hedge fund and still make money and

20:06

make really important investment decisions. And other

20:08

people are like, I want to be

20:10

able to do crossword puzzles and read

20:12

the newspaper. I agree with you. You

20:15

can't define it, but it doesn't mean...

20:17

we shouldn't try to personalize it. Okay,

20:19

but I want to come back to

20:21

you Matt with the original question. Why

20:24

are we at a point where? Why

20:26

is longevity gone mainstream? Yeah, for a

20:28

lot better way. Yeah, so I mean

20:30

I think both of the points that

20:33

Stephen Rich raised are part of the

20:35

equation. I mean I think it's a

20:37

convergence of all of these factors and

20:39

maybe a few others. I do think

20:42

the science has matured to the point

20:44

where more people are believing that we

20:46

can actually modulate. the biology of aging.

20:48

I think the concept of biological aging

20:50

has become popularized through a variety of

20:53

mechanisms, including some influencers, individuals who I

20:55

personally think often err on the side

20:57

of being a little bit less scientific

20:59

than they should be, but I think

21:02

they've helped popularize the concept. So I

21:04

think it's been a combination of these

21:06

factors. And why it has taken so

21:08

long? I mean, I just think that's

21:11

the pace that science moves and the

21:13

rate at which these concepts can sort

21:15

of permeate the public... So it's frustrating

21:17

in a sense that it's moved so

21:20

slowly. I also wonder, because you sort

21:22

of said, are we at a longevity

21:24

bubble? I don't know. I think maybe

21:26

we're still kind of in the early

21:29

days of this hockey stick moment, where

21:31

you're getting this exponential increase in attention.

21:33

My hope is, as we go forward,

21:35

it will become more scientific and less

21:38

snake oily, and it's a spectrum. There's

21:40

this huge gray area in the field

21:42

right now of what's real and what's

21:44

not real, and I think none of

21:47

us at this at this table, actually.

21:49

can really define exactly where in that

21:51

gray area that line is or is

21:53

there a line? To that point, Matt,

21:56

what is the collective wisdom of the

21:58

group on the funding appetite for that?

22:00

Because I agree with you completely, like

22:02

if we could channel this exuberance away

22:05

from kind of the highly commercial speculative

22:07

grifting towards the budget increasing legitimate investigative,

22:09

that would be awesome. What is the

22:11

appetite right now of NIA with respect?

22:13

back to this? I think it's hard

22:16

to say it. I mean, NIH is

22:18

a moving target, and as we all

22:20

know, there's going to be a lot

22:22

of change coming in the near future.

22:25

So, cautiously optimistic, I would say, if

22:27

you look historically, it's been really pretty

22:29

terrible. The percent of NIH budget that

22:31

goes to biology of aging, I think

22:34

is still probably around half of 1%.

22:36

Sorry, just to put numbers in perspective,

22:38

fraction that goes to biology of aging.

22:40

Yes, yes, yes, yes. Yeah. But I'm

22:43

saying there are 17 groups of NIH.

22:45

NIA being one of them gets what

22:47

fraction of NIAH budget roughly? I think

22:49

it's roughly 3%. 3% of NIA budget

22:52

is NIA. Within NIA, how much goes

22:54

to this type of research? It was

22:56

about 350 million a few years ago.

22:58

It might be a little higher than

23:01

that, but I don't think it's ticked

23:03

up any more proportional to the increase

23:05

in the increase in NIA budget since

23:07

then since then. What's your level of

23:10

optimism, Rich? You're obviously very close to

23:12

this. That NIH will wake up and

23:14

start to pay attention to aging research

23:16

the way they should? It's near zero.

23:19

It's been near zero for 30 years

23:21

now. Even with this outside attention? Well,

23:23

it's gone up. I mean, they funded

23:25

the ITP, the interventions testing program, 20

23:27

years ago, and they liked it, and

23:30

they doubled our budget about 15 years

23:32

ago. So that's something. And I'm very,

23:34

very grateful to them for that, but

23:36

there's still an enormous untapped potential for

23:39

making progress in the basic biology of

23:41

aging. And the reason is, again, a

23:43

matter of defending turf. If you are

23:45

a cardiologist researcher or an oncologist researcher

23:48

or an AIDS researcher or an Alzheimer's

23:50

researcher, any time somebody says, the smart

23:52

play is to reduce your budget by

23:54

10 percent or your institute's budget by

23:57

10 percent. We're going to go there

23:59

faster if we spend money on aging

24:01

and its relationship to the disease you

24:03

care about. You get the porcupine defense.

24:06

You don't take any of my money

24:08

because Alzheimer's is important. little kids with

24:10

leukemia are important, breast cancer is important,

24:12

you go away. And that is the

24:15

predominant feeling. Most of the people making

24:17

those decisions were not trained in aging

24:19

research, they view it as something interesting,

24:21

I read something about that in Time

24:24

Magazine the other day, but they don't

24:26

understand that to actually conquer or slow

24:28

down or effect or protect against the

24:30

disease they care about, the smart play

24:33

is to do aging research. And so

24:35

they view your suggestion, which I of

24:37

course agree with 100% as an imposition,

24:39

an invasion to be repelled at any

24:41

cost. No one in a position of

24:44

power has had whatever it takes to

24:46

reverse that. And if he or she

24:48

tried to do that, Congress would even

24:50

a good Congress would smack them down.

24:53

The Alzheimer's group has 100 lobbyists, the

24:55

cancer group has 100 lobbyists, the AIDS

24:57

group has 100 lobbyists. The aging group

24:59

has two lobbyists, one who's a lawyer

25:02

and one who takes the calls, and

25:04

that's not enough to do it. Can

25:06

I just add something real quick? I

25:08

agree completely. And I think, as well,

25:11

the reputation of the field has hindered

25:13

that transition as well. So historically, the

25:15

field was viewed as not very mechanistic,

25:17

kind of phenomenological, became much more mechanistic,

25:20

starting around the time of Cynthia Canyon's

25:22

work, and since then. but has continued

25:24

to have a reputation problem as not

25:26

being as rigorous as other areas of

25:29

research. So I think it is absolutely

25:31

a turf for, and there's this overcoming

25:33

the reputational problem which makes it harder

25:35

for serious people in funding and policy

25:38

circles to give it the attention it

25:40

deserves, in my opinion. So I've got

25:42

a different take on this. I actually

25:44

think that this is a very good

25:47

time for aging research funding, and that's

25:49

not because of what's going on at

25:51

the NIA. but it's what's going on

25:53

in the private sector. There's more and

25:55

more money. There's even interest now in

25:58

big pharma that was very spotty in

26:00

the past. So I think if We

26:02

focused entirely on the Nationalist and on

26:04

aging. We would get a false impression

26:07

of what the funding climate is in

26:09

the field now. And I think we

26:11

need to take advantage of that. Got

26:13

to make sure that it doesn't get

26:16

captured by the people who are doing

26:18

the fleshy, but bad science. You're saying,

26:20

look, Calico, Altos, other private companies, especially

26:22

within biotech and pharma that are looking

26:25

at geoprot protective molecules building on the

26:27

work of the ITP. Yeah, I think

26:29

it's safe to say the amount of

26:31

money that's being spent privately probably out

26:34

does public spending, I mean in a

26:36

given year two to one easily. It

26:38

could, although how much of that is

26:40

actually going to biology of age? I

26:43

think it's still an open question. You

26:45

mentioned Calico and Alton's right? Exactly. I

26:47

actually agree with Steve. I don't think

26:49

what Rich and I were communicating as

26:52

opposed to what Steve was communicating. There

26:54

are a lot of this hockey stick

26:56

moment. And I think Steve's right, there

26:58

are real opportunities for more resources to

27:01

be focused on the scientific side and

27:03

hopefully less focused on the non-scientific aspects

27:05

of what are going on. And you

27:07

asked the question of can we shift

27:10

resources from the more consumer-facing, maybe not

27:12

as rigorous stuff, and into the more

27:14

rigorous stuff. I'm not a fan of

27:16

that stuff at all. But maybe you

27:18

need that stuff to kind of move

27:21

the needle and get people's attention, and

27:23

at least people are talking about longevity

27:25

now. Now you've questioned I'm embarrassed I

27:27

don't know the answer because I spent

27:30

more than two years working there What's

27:32

the mission statement of the NIH? It's

27:34

to preserve and enhance human health I

27:36

mean it's basically the same thing that

27:39

we do that we're supposed to enhance

27:41

human health I mean it's basically the

27:43

same thing that we do that we're

27:45

supposed to be doing Yeah and I

27:48

didn't actually get to give you my

27:50

spiel here but what I started to

27:52

say about the NIA budget is if

27:54

you look at the United States Nine

27:57

of them have biological aging as their

27:59

greatest risk. factor and it's not even

28:01

close. Yet half of one percent of

28:03

the research budget that's supposed to

28:05

be focused on improving human health

28:07

goes to study that risk factor. I

28:10

mean I think it is extremely frustrating to

28:12

all of us sitting at this

28:14

table that that hasn't changed but

28:16

there's reason to be optimistic that

28:19

maybe it will change in the near future.

28:21

Let's state that again because it is

28:23

so profound I want to make sure

28:25

not a single person missed that statement.

28:27

The top 10 causes of death

28:30

in the United States are

28:32

well enumerated and incredibly predictable,

28:35

and they increase by category,

28:37

by decade, 3 to 8

28:39

percent, monotonically, with no exception.

28:42

Point being, 90 percent of, and

28:44

more than 90 percent on an

28:46

adjusted basis of what causes death,

28:49

goes up with age. And yet, a

28:51

few basis points of federal

28:53

R&D goes to addressing that.

28:55

Let me give you an example of what

28:57

the sort of point that Matthew and

28:59

you have been making. About once every

29:01

five years, I give a talk and

29:04

invite a talk at the University of

29:06

Michigan Cancer Center, and I point out

29:08

that we have drugs now, anti-aging

29:10

drugs, in mice, and they extend mouse

29:12

lifespan, and they do it mostly

29:14

by postponing cancer, because most of

29:16

our mice die of cancer. And if you

29:19

look at age-adjusted cancer incidence rates,

29:21

our drugs reduce these by a

29:23

factor of 10. Wouldn't they like

29:25

to know why? As cancer scientists,

29:27

we now have a batch of

29:29

drugs that postpone cancer. Wouldn't they

29:31

like to study them? Invariably, I get

29:34

one call back from somebody who says,

29:36

that's interesting, maybe we should talk

29:38

about that, and then it dies, and

29:40

then five years later, I'm asked to give

29:42

the same talk or related talk. So they

29:45

know how to do cancer research, they are

29:47

cancer scientists, that's how they know how

29:49

to do cancer research, and you certainly

29:51

don't do it by... diverting your

29:54

lab's attention to aging. That's

29:56

insane, but that insanity is

29:58

how medical research is organized

30:00

and breaking that addiction to

30:02

the kinds of models you grew up on

30:04

because there are a better idea, it's not

30:07

an easy thing. It may not even be

30:09

a possible thing to do. That's a

30:11

major hassle. I think this is because

30:13

we think about health all wrong. We

30:15

think, let's wait to get cancer and

30:17

see what we can do about it.

30:19

That's what cancer biologists do. You have

30:21

cancer, okay, how can we better treat

30:24

that? Or could we have diagnosed it

30:26

earlier? What Rich is saying and what

30:28

we can know how to do in

30:30

lots of model organ, it prevents you

30:32

from getting cancer, delay it for a

30:34

considerable amount of time, that's a little

30:36

bit harder to study if you're a

30:38

cancer biologist, because you want to see

30:41

the cancer before you can study it.

30:43

I think that's why we need aging

30:45

biologists, rather than people focused on certain

30:47

disease to come and try to use

30:49

what we do. If we prevented the

30:51

cancers, they'd be out of a job.

30:53

people or mice will get cancer. They'll just

30:56

have 10 extra years of life if they're

30:58

a person or 10 extra months of life.

31:00

They'll get cancer. They'll need specialists. It'll be

31:02

all right. Yeah, I think that's important. I

31:04

mean, I think the reactive disease care component

31:06

is still going to be there. Even if we're

31:08

insanely successful at slowing agent, people are still

31:10

going to get sick. But I think Steve's

31:13

point is really important. Like Peter, you've been

31:15

a leader in helping people recognize the need

31:17

to shift the need to shift the medical.

31:19

approach from reactive to proactive. I

31:21

think what a lot of people

31:23

don't realize is that mentality goes

31:25

all the way back to pharmaceutical

31:28

research, biomedical research, basic science. That

31:30

is ingrained all the way through.

31:32

And I think one of the

31:34

challenges with getting funding for aging

31:37

research is that mentality on the

31:39

basic science world and how deeply

31:41

ingrained it is. It's very interesting

31:43

because you don't know which is the tail

31:45

and which is the dog. I've always assumed

31:47

that the one leading the charge

31:50

is the clinical side of things.

31:52

In other words, the engine, the

31:54

machine of Medicine 2.0, is built

31:56

around the delivery of care, the

31:58

delivery of care. as you said,

32:01

Steve, is built around, I'm going

32:03

to wait, I'm going to sit

32:05

here and hang, we're going to

32:07

wait, when you get the disease,

32:09

we're ready. You have the heart

32:12

attack, fantastic. You've got chest pain,

32:14

ST elevations, we've got a

32:16

stint for you, now you have

32:18

cancer, we're all in. And then

32:21

the research flows from that mindset.

32:23

Of course, I don't know, not

32:25

that it really matters, but it

32:27

might be that it's flipped. matters,

32:29

but if you could be health czar and

32:31

fix one of them, you might actually

32:34

start with the research side of things.

32:36

I would. And I mean, the reality

32:38

is the research flows from where the

32:40

dollars are going. This has been seen

32:42

over and over and over at NIH.

32:44

You shift resource allocation to a certain

32:46

area and the scientists will follow and

32:49

they will submit grants to get grants

32:51

in the place where the funding line

32:53

is the highest. So if somebody came

32:55

along and said we're going to go

32:57

to biology of aging. You'd have no

32:59

shortage of people. I mean, it would be

33:01

kind of messy at first, but you'd have

33:03

no shortage of people applying for grants and

33:05

becoming experts in the biology of aging. And

33:08

the system would work. You'd get the best

33:10

and the brightest that would go into that

33:12

and do that. So this then begs another question

33:14

that is a tired question, but I can't

33:16

help but ask it at this point. Is

33:18

aging a disease? Is that even a

33:21

relevant question? Call me, calling me, calling

33:23

me, calling me. It's important to use

33:25

words, to use words optimally, optimally, optimally.

33:27

and to distinguish causes

33:29

from effects. One of the

33:31

bad things about aging is it's

33:33

a risk factor for many diseases.

33:36

Some things are other risk factors

33:38

for diseases. Aging is a risk

33:40

factor for disease. And so saying

33:42

that aging is a disease confuses

33:45

that discussion. It makes

33:47

it impossible to see that

33:49

relationship. So calling aging a

33:51

disease is a fundamental

33:53

error. The question itself is

33:55

uncorrect. I agree. But I

33:57

think we have that idea.

34:00

for marketing purposes, not for

34:02

scientific purposes. And the

34:04

idea is, well, the money goes

34:06

to diseases, let's call aging a

34:08

disease, because I think what we're

34:10

trying to do is we're trying to

34:12

treat aging as we're trying to do,

34:15

is we're trying to treat aging as if

34:17

it were a disease, even though I would

34:19

agree with both of you. I don't think

34:21

it's a disease. I think that destroys

34:24

the word disease if we include aging

34:26

in it. But I think there was

34:28

a reason that you can... convince people

34:30

of the importance of aging research only

34:32

by crossing your fingers and saying, oh,

34:34

well, it's kind of a disease, isn't

34:37

it? You think you can fool them?

34:39

Yes, that's what marketing is, and it's

34:41

probably good for that. I just don't

34:43

like lying to people. It also creates

34:45

a negative feeling about the field in

34:48

some people as well. So I think

34:50

that should be considering. The other point

34:52

that people often raise, though, is we have

34:54

to call aging a disease in order for

34:56

FDA to approve a fundamental... misunderstanding

34:58

of how FDA operates, but

35:01

that is the other argument

35:03

you will often hear among

35:05

proponents of the idea that aging is

35:08

a disease. Very interesting. Well, so

35:10

now let's go one step deeper

35:12

on that. How do you

35:14

think about biologic versus chronologic

35:17

age? In concept and in practice.

35:19

On the right over here, Rich and

35:21

I were talking about that, I

35:23

don't believe there is one thing

35:25

as biological age. I think there is potentially

35:27

an age of your heart, an age

35:29

of your liver, an age of your

35:31

lungs, an age of your brain, but

35:34

I don't see why we wouldn't simply

35:36

call it health. In other words, I

35:38

got one of these epigenetic age clocks

35:40

done on me a while ago,

35:42

but I didn't know what to make

35:44

out of it. I thought, is this just

35:46

flattery? Or did it really tell me something?

35:48

You must have got a good result. I

35:51

got a good result. He's 13. 13 years

35:53

old. That may be the point of the

35:55

whole thing, right? So I'm dubious about some

35:58

number that is different than I know. good

36:00

health, for my age I'm in very

36:02

good health. So I knew that already.

36:04

Now I have a number for it.

36:06

I don't put much credence in that.

36:09

Let me agree with Steve, but just

36:11

put it in slightly different terminology. It's

36:13

a matter of taking a very rich,

36:15

complex data set and trying to collapse

36:17

it to a number. So if someone

36:20

wants to know how healthy I am,

36:22

he or she would need information. How

36:24

good is my eyesight, how good is

36:26

my hearing, how good is various kinds

36:28

of cognitive activities. my aerobic endurance, my

36:31

joints, all of that is pertinent to

36:33

how my health is, and also about

36:35

projected future health, then there's no need

36:37

once you've got that information, which is

36:40

very rich, to say, ah, there's a

36:42

number, a single number, a real number,

36:44

a point on the number line that

36:46

condenses that in any useful way. A

36:48

notion 40, 50 years ago that biological

36:51

age was not the same as chronological

36:53

age, for a little while was useful,

36:55

it emphasized that there might well be.

36:57

60-year-old people who were unusually like youthful

36:59

people, and 60-year-old people who were unusually

37:02

like 70-year-old people, would my drug or

37:04

my genetic mutant or whatever, help to

37:06

discriminate those people or change them in

37:08

some way, I can slow your biological

37:10

aging process. That's a discussion that was

37:13

maybe of interest 40 years ago, and

37:15

it's now time to drop the notion,

37:17

let alone the silly notion. that you

37:19

can count that biological age, that number

37:21

which some people, too many people, still

37:24

think is of value. You can figure

37:26

out what it is by measuring something,

37:28

transcriptions or epigenetic markers or something. I

37:30

can do it and give you personally,

37:33

your personal biological age. That's a waste

37:35

of everyone's time, and it also distracts

37:37

attention from things that actually are important

37:39

and need to be thought about. I

37:41

gotta talk, because... I think I disagree

37:44

fundamentally, and I'm surprised, but this will

37:46

be an interesting conversation. So I agree

37:48

that the idea of a kit, the

37:50

you can buy to measure biological age.

37:52

First of all, the stuff that's out

37:55

there doesn't work, and we can and

37:57

should talk about that. But also, I

37:59

sort of agree with the idea that

38:01

reducing it to one number, while conceptually,

38:03

I think it's possible, I think in

38:06

reality, is gonna be really, really difficult

38:08

to do. But do I believe that

38:10

there is a biological aging process that

38:12

is different from? chronological aging. Absolutely. Oh

38:15

yes, absolutely. Okay. Well, it sounded like

38:17

you guys were both saying, no, you

38:19

didn't think it was a real thing.

38:21

I agree with that completely. You can

38:23

agree with that and not like the

38:26

idea of a number of the constitutes

38:28

your biological age. Okay. There's two things

38:30

that kind of make me feel pretty

38:32

confident in this idea. One is, and

38:34

this is the example I use a

38:37

lot among the general public, is just

38:39

look at dogs compared to people. Everybody's

38:41

compared to people. means that dogs age

38:43

about seven times faster than people do.

38:45

But of course chronological time is the

38:48

same between dogs and people. It's the

38:50

biological aging process. And so you can

38:52

look across the animal kingdom and see

38:54

this. And dogs get almost all of

38:56

the same diseases and functional declines that

38:59

we do at the tissue and organ

39:01

level, but also the whole body level.

39:03

We also know now there are single

39:05

genes that significantly modulate. what I would

39:08

call the rate of aging. Now maybe

39:10

we have a different meaning to what

39:12

we mean. No, I agree. Okay. So

39:14

the fact that that's possible. Deftu, we

39:16

talked about Deftu a couple times, Tor,

39:19

we can turn these things up, turn

39:21

them down, and animals across the evolutionary

39:23

spectrum seem to age at different rates

39:25

by modulating single genes. So I don't

39:27

know of any other explanation other than

39:30

that there is this process, which we

39:32

call biological aging, that... can be changed

39:34

and the rate can be sped up

39:36

or slowed down. Can it be reversed?

39:38

That's an interesting question. Maybe we'll get

39:41

to that. But I think the process

39:43

is real. I think it's just really

39:45

really complicated and we probably only understand

39:47

5% of it at this point. Yeah,

39:49

I think for me that challenges, I

39:52

kind of land where Rich was, which

39:54

is if a patient says to me,

39:56

hey, why aren't you doing this biologic

39:58

age clock on me, my response is,

40:01

well, I know your VO2 Max, I

40:03

know your zone 2, I know your

40:05

muscle mass, I know your visceral fat,

40:07

we did a very complicated movement assessment

40:09

on you, I understand your balance, I

40:12

understand your lipids, your insulin, like I

40:14

know these 57 things about you and

40:16

I can tell you individually on each

40:18

of them how you're doing, that number...

40:20

doesn't tell me a single new piece

40:23

of information. But what if you were

40:25

to come up, and you probably do

40:27

this in your head, you come up

40:29

with some sort of composite, you probably

40:31

don't sit down and wait each of

40:34

those things and come to one number,

40:36

but you come up with some sort

40:38

of composite picture of health based on

40:40

all of those things, that's a different

40:42

biological aging clock. I think sometimes we

40:45

conflate and in part this because of

40:47

the way that irresponsible people in the

40:49

field and marketers have done this, we

40:51

conflate the epigenetic tests. with biological aging

40:54

clocks. There are all sorts of flavors

40:56

of biological aging clocks, including things like

40:58

frailty indices or metrics of a whole

41:00

bunch of functional markers. So I think

41:02

those probably are pretty good readouts of

41:05

biological age. Again, can you combine them

41:07

all to get to one number that's

41:09

meaningful for every person? That's much harder

41:11

to do. Yeah, tell us about your

41:13

experience, because this was, you did what

41:16

I wanted to do, but I've been

41:18

too lazy to do. coming up with

41:20

different names. So what I did was

41:22

I tested four different direct-to-consumer biological age

41:24

kits. They were all epigenetic biological age

41:27

tests before different companies, and I did

41:29

duplicates of each kit, and it was

41:31

from the same samples collected on the

41:33

same date. I really tried to put

41:35

my scientist hat on. I only had

41:38

two replicates, I didn't have three replicates,

41:40

but it's about the best I could

41:42

afford at that point. And it was

41:44

kind of expensive. And it was kind

41:47

of expensive. to me, very informative, fundamentally

41:49

sort of changed my views on these

41:51

epigenetic age tests. So they ranged. from

41:53

42 to 63, I was 53.75 years

41:55

at the time I did the test,

41:58

and the standard deviation, I can't remember,

42:00

it was either 7 or 9. So

42:02

mean if my chronological age, standard deviation

42:04

of 7 or 9, which I look

42:06

at that data, I'm not a statistician,

42:09

but I know enough statistics to say

42:11

that's completely useless. They converged on my

42:13

chronological age, but with a huge variation.

42:15

Even intro... So that varied between the...

42:17

tests. So I think three of the

42:20

four were reasonably close to each other.

42:22

Three of the four companies, the duplicates

42:24

were reasonably close to each other, but

42:26

the individual tests were far apart, and

42:28

one of the companies, the individual replicates,

42:31

was 20 years apart. So to me,

42:33

and some people will say, but maybe

42:35

the true diagnostic test is great, and

42:37

the Elysium test is terrible, or the

42:40

tally health test is terrible, and the

42:42

other one is great. Maybe, but how

42:44

do we know? My take home is

42:46

that the direct to consumer biological age...

42:48

testing industry is a complete mess. And

42:51

I have no idea who to believe

42:53

or if any of them are actually

42:55

giving accurate data. I know some of

42:57

the people at some of the companies

42:59

and I have my personal feelings about

43:02

who's trying to do it right and

43:04

who's sort of a charlatan, but across

43:06

the industry it's really hard to know.

43:08

The last thing I'll say on this

43:10

is the where I've sort of landed

43:13

is I think these are really good

43:15

research tools. I think the direct to

43:17

consumer component has gotten way ahead of

43:19

itself and I think I aligning with

43:22

what you were saying about the way

43:24

you think about these tests. I don't

43:26

think there's a lot of value in

43:28

clinical practice right now, because we don't

43:30

know precision or accuracy, and I don't

43:33

think you can make actionable recommendations based

43:35

on these tests. Furthermore, they fail in

43:37

the one thing that I think they're

43:39

attempting to do, and I usually use

43:41

this illustration with patients. So if I

43:44

have a 40-year-old patient who says I

43:46

really want to do one of these

43:48

tests, I say, if the answer comes

43:50

back and says you're 20, Is your

43:52

expectation that you will live another seven...

43:55

years? Conversely, if the answer comes back

43:57

and says 60, is it your expectation

43:59

that you will live another 30 years?

44:01

In other words, is this number predictive

44:03

of future years of life? Because right

44:06

now we have this thing called chronological

44:08

age that is the single best predictor

44:10

of future years of life. So do

44:12

we think biologic age, as determined by

44:15

these tests, is better as a predictor

44:17

of future years of life? Which, by

44:19

the way, would be very testable. How

44:21

many people have contacted you? to get

44:23

ITP sample data to say, can we

44:26

predict how much longer these mice we're

44:28

going to live? The answer to the

44:30

question is obvious and very well known.

44:32

You can tell if you have your

44:34

40-year-old patient and he or she is

44:37

fat, doesn't exercise, it's mostly cheeseburgers, you

44:39

know that their life expectancy is probably

44:41

not as good as a 40-year-old patient

44:43

in your next waiting room that has

44:45

extremely healthful habits and whose parents live

44:48

to be 100. Tons of public stuff.

44:50

I don't need a biologic age to

44:52

tell them. Right, that's what I'm saying.

44:54

There are tons of things you can

44:56

measure on individuals, four or five of

44:59

them, or all you really need to

45:01

ask of a 70-year-old. Yeah, MetLife does

45:03

this really, really, really, really, because their

45:05

buddies all, they'll lie there, they're writing

45:08

life insurance policies. So it's not at

45:10

all hard to figure out a very

45:12

small set of tests to tell you

45:14

how long a seven-year-year-year-year-old is likely to

45:16

live, is likely to live, there's likely

45:19

to live, there's nothing to do with.

45:21

methylation clocks or things like that. That's

45:23

the gold standard. When life insurance companies

45:25

start using biologic clocks as the cornerstone

45:27

of their actuarial algorithms, I'll start to

45:30

be far away from that. I'm going

45:32

to sound like a broken record here,

45:34

but you guys keep saying biological age

45:36

when what you mean is epigenetic age

45:38

or epigenetic tests. Not necessarily. And we

45:41

should explain to people that there is

45:43

a difference. So some of these clocks

45:45

use solely epigenetic measurements. Not all, most

45:47

of the direct-to-consumer ones are epigenetic. But

45:49

some of these tests use a litany

45:52

of biomarkers inclusive of epigenetics. So they'll

45:54

say, we've sampled your meth. pattern, but

45:56

we also looked at your vitamin D

45:58

level, your glucose level, your cholesterol level,

46:01

and a whole bunch of other things,

46:03

and we compressed all of that into

46:05

a number as well. So I guess,

46:07

let me frame it as a question

46:09

to you. So let's take the epigenetic

46:12

piece out. Again, I do think we

46:14

will get to a point where the

46:16

technology is developed far enough and the

46:18

quality control is good enough on the

46:20

consumer side that these tests will be

46:23

better than just chronological age. That's a

46:25

big statement. I don't know that I'm

46:27

disagreeing with you. I just want to

46:29

make sure you understand the state. I

46:31

mean, I think it's clear from the

46:34

research, unless you think that all of

46:36

the research that's been done on these

46:38

epigenetic aging clocks is somehow flawed. It's

46:40

clear that you can create algorithms that

46:42

can predict specific methylation patterns that are

46:45

more highly correlated with life expectancy than

46:47

chronological age. But I think the big

46:49

but here is that even if that's

46:51

the case, they would not be as

46:54

good as what Peter would predict after

46:56

all the Biological age. That's what I

46:58

want to get to. Yes, and I

47:00

think what you are actually doing is

47:02

looking at other biomarkers that have a

47:05

long-term clinical history that you're using to

47:07

come up with a surrogate, but really

47:09

is reflecting largely biological age. Maybe not

47:11

completely. And this is the other point

47:13

I wanted to make is I don't

47:16

think biological age and health are equal.

47:18

I think they're strongly overlapping. And certainly

47:20

you can identify many ways to reduce

47:22

health without accelerating biological aging. I think

47:24

that's easy. We can all think of

47:27

ways to do that. So let's take

47:29

a minute and try. Yeah, so let's

47:31

think about this for a second. I

47:33

have seen very impressive data where we

47:35

can look at tissue samples of organs

47:38

and we can tell, okay, I'm going

47:40

to show you a sample of nephrons.

47:42

And just based on nothing but the

47:44

methylation pattern, we know that if I

47:47

just said to you, one of these

47:49

is a 20 year old, one of

47:51

these is a 70 year old. It's

47:53

very easy to predict based on the

47:55

methylation pattern which Neffron came... from which

47:58

person. Completely agree with that. There are

48:00

a lot of things that change with

48:02

age. The literature has 25,000 things that

48:04

change with age. Average amount of methylation

48:06

at these 10 spots is number 11,4007

48:09

of those. So great, you've got another

48:11

thing that changes with age. So that's

48:13

the question. But that's not enough. Right.

48:15

So do you believe that all of

48:17

the research we're seeing on the epigenetic

48:20

clocks is going to be the 78th

48:22

variable that we would include in our

48:24

gestalt... I don't know. Yeah, it's a

48:26

good question. So I am hopeful that

48:29

epigenetic algorithms can get to the point

48:31

where they can replace many, certainly not

48:33

all, but many of the other biomarkers

48:35

that are being measured. I think the

48:37

thing that gives me hope is we

48:40

know that epigenetic changes are part of

48:42

biological aging. This again is a different

48:44

question, but if we look at the

48:46

hallmarks of aging, epigenetic dysregulation is one

48:48

of the 12. Some people will argue

48:51

it's the most important conversation, but it's

48:53

at least part. So that gives me

48:55

some hope that we are in fact

48:57

measuring something that plays a causal role

48:59

in the aging process. And I think

49:02

what's missing, I think what would give

49:04

all of us a lot more confidence

49:06

is if we had a mechanistic connection

49:08

to the specific methylation changes and some

49:10

cause of aging or age-related disease. In

49:13

other words, this change in methylation changes

49:15

this particular gene's expression level, which changes

49:17

the rate of biological aging. I think

49:19

if we had that, we'd feel a

49:22

lot more confident. Yeah, you and I

49:24

spoke about this very briefly at the

49:26

end of our last podcast, and I

49:28

want to come back to it with

49:30

all of us on this table, because

49:33

there's so much in what you just

49:35

said, Matt, that I'm going to lay

49:37

out a broad question, and then we

49:39

can start attacking it in different ways.

49:41

So one of the things I want

49:44

to address is do we believe that

49:46

it's possible that of the hallmarks of

49:48

aging, epigenetic change is the most important.

49:50

Another topic I want to address, do

49:52

we believe that the epigenetic changes that

49:55

we observe over time which are undeniable

49:57

are causal in the arrival of other

49:59

states? everything from the arrival of senescent

50:01

cells, the increase in inflammation, the reduced

50:03

function of the organs, which really is

50:06

the hallmark of aging. And if so,

50:08

does that mean that reversing the epigenetic

50:10

phenotype will undo the phenotype of interest?

50:12

And Rich, where I'm going, that you

50:15

and I left off was, what about

50:17

the proteome? What about the metabolism? So

50:19

you made three statements there. Broad general

50:21

statements. And I think each of the

50:23

three deserves careful... Amendment. Let's do it.

50:26

To be polite about it. The first

50:28

has to do with hommarks of aging,

50:30

which I think set the field back

50:32

dramatically. I think when you are officially

50:34

branded a homework of aging, by two

50:37

people sitting alone at their computers and

50:39

writing a review article, a homework of

50:41

aging... I thought they were walking around

50:43

a pond when they came up with

50:45

this. All right, okay, okay. Means that

50:48

somebody once said, I'm interested in aging,

50:50

that's kind of important, isn't it? Let's

50:52

put it on our list. You can't

50:54

tell if something is a hallmark of

50:56

aging. Does that mean it goes up

50:59

with age? It goes down with age.

51:01

You can change it in a way

51:03

that will extend lifespan. You can kill

51:05

a mouse or a worm by removing

51:08

it. Basically, it's something that somebody wants

51:10

to might be of interest to aging.

51:12

And the downside of that is once

51:14

you're officially branded as a hallmark of

51:16

aging, anyone who wants to write up

51:19

grant on that doesn't have to prove...

51:21

that they're a fundamental cause and effect

51:23

model has any merit. Because it's a

51:25

hallmark of aging. I don't have to

51:27

prove it anymore. Someone, I don't know

51:30

who or on what grounds, has decided

51:32

it's important. My reviewers know it's important

51:34

because they've read the hallmark of aging

51:36

paper, so I don't have to think

51:38

about whether it's important. The negative side

51:41

of that coin is that there are

51:43

lots of things that didn't make it

51:45

into the hallmark list. I really think

51:47

it's premature to close thought off on

51:49

some of those. It's easy to come

51:52

up with a dozen things that ought

51:54

to be investigated, but if you want

51:56

to investigate it, it's not on the

51:58

home marks list. What are you wasting?

52:01

So deciding which of the hallmarks is

52:03

the big daddy hallmark or whatever strikes

52:05

me is not the correct thing to

52:07

talk about in the hallmarks arena. So

52:09

maybe we should talk about that before

52:12

we go through all of these. I think

52:14

there's a lot to unpack there. You

52:16

guys could afford to give me a

52:18

little piece of paper and a pen.

52:20

I think the hallmarks is a list,

52:22

a kind of arbitrary list, not completely

52:24

arbitrary, because I had some reasons for

52:26

being there. I don't think any of

52:28

us would say that. those 12 things

52:30

are not involved in aging. But

52:32

that's a very little interest. Do any

52:34

of us want to rattle them up being

52:36

that I'm the only one that's got the

52:39

list sitting in front of me? We could

52:41

do a game where we each name one

52:43

and see who can't see if we get

52:45

to all 12. Yeah. But certainly in that

52:48

list, I would not consider epigenetics as the

52:50

key hallmark. Assuming there are such

52:52

things, I consider it to be

52:54

an interesting list. It became... biblically

52:57

sacrosan almost immediately and I've never understood

52:59

why but for some reason it did

53:01

so I'd agree with rich so conceptually

53:03

beautiful I mean so I agree completely

53:05

with rich and he knows I do

53:07

because we've talked about this before I

53:09

think the flip side is I think

53:11

the hallmarks have been immensely useful to

53:13

the field they are a very easy

53:15

way to communicate this idea of biological

53:17

aging and it helps convince some of

53:19

the scientific community that thought it was

53:21

all just hocus pocus pocus and snake

53:24

oil that there is some mechanistic research

53:26

happening. We can point to specific things

53:28

that are aging. So I think that

53:30

part of the hallmarks has been actually

53:32

really valuable and has contributed to the

53:34

popularization of longevity and at least to

53:36

the extent the science of longevity has

53:38

been popularized has contributed to

53:41

that. And it has been extremely detrimental to

53:43

the field and the way I think about

53:45

it is it just caused the field to

53:47

narrow prematurely. And this goes back to

53:49

what I alluded to before. I

53:51

don't know if we understand 80%

53:54

of biological aging or 0.205% of

53:56

biological aging. My guess is it's

53:58

closer to 0.205% and... And by and

54:00

large, the funding to look outside of

54:02

the hallmarks dried up once the hallmarks

54:05

became the dominant paradigm. And people stopped

54:07

looking. And I think we need to

54:09

go back to more discovery science and

54:11

thinking outside the box. So I think

54:14

it's been a double-edged sword. Would that

54:16

happen automatically if we could wave that

54:18

magic wand and increase funding? It would

54:20

help. I don't know that it would

54:22

help enough, but it would help. I

54:24

mean, you also kind of have to change

54:26

the mindset. about what people call fishing expeditions.

54:28

That's like a bad word in grant

54:30

review panels, fishing expedition, meaning you don't

54:33

really know what you're going to find,

54:35

but you got to go look before

54:37

you can figure out what's important. So

54:39

we have to kind of change that

54:41

mindset as well. One can usefully concreteize

54:43

this discussion. I imagine that one of

54:45

this, I don't read these papers because

54:47

they upset me, but I imagine inflammation

54:49

is on one or more of these.

54:51

Sure is. Chronic inflammation. chronic inflammation. What

54:53

that does is, if you say I'm

54:55

interested in chronic inflammation, so I'm doing

54:58

good stuff, huh? But what could

55:00

be happening is this particular set

55:02

of cytokines might be overexpressed by

55:04

some glial cells and that leads

55:06

to loss of cognitive function, whereas

55:08

this other overlapping set of cytokines

55:11

produced by the macrophages in your

55:13

fat may be due more prone

55:15

to diabetes or metabolic syndrome, whereas

55:17

this particular set of lymphocytes...

55:20

are necessary to repel COVID,

55:22

and that's why you are

55:24

more susceptible to COVID. So

55:26

learning what changes within the

55:29

extremely broad generic idea of

55:31

inflammation, what changes in what

55:33

cell types, in what people,

55:36

under what pharmacological or genetic

55:38

changes, how they are interacting

55:40

with other aspects of pathology,

55:43

that's marvelous to do, but

55:45

to say, oh, inflammation, that

55:47

gets bad when you're old. is

55:49

a way of avoiding the labor of

55:51

thinking, and that's why I'm against it.

55:54

And I think Matt brought up

55:56

a really important point, and we

55:58

scientists are to blame. is the

56:00

way that research gets reviewed. For

56:03

lazy reviewers having these 12

56:05

hallmarks is really helpful. This has

56:07

got one of the hallmarks in

56:10

it. This must be good stuff.

56:12

I do think reviewers need to be

56:14

more open to new ideas and

56:16

new approaches. I mean everybody

56:19

knows that NIH grants are

56:21

approved if they're incremental. If

56:23

they're really breakthrough, they don't

56:26

get approved. A very famous biologist,

56:28

E.O. Wilson, told me years ago,

56:30

he said, don't ever include your

56:32

best ideas in a grant. They

56:35

won't get funded. Do the standard

56:37

stuff, save your best ideas for

56:39

projects that you do on the

56:41

side. That's one of the reasons

56:43

I left academia. Draw me

56:45

nuts. Almost impossible to get the

56:47

important stuff funded. The second

56:50

of your multi-partite question was, does epigenetic

56:52

change? Is it causal? And the

56:54

third, which we may get to,

56:56

is can you reverse it, and

56:58

would that be a good thing?

57:00

So let's talk about the second

57:02

element here, is it causal. The problem

57:04

is what it means. There are some changes

57:07

that occur in this particular set

57:09

of 40 cells in the pineal,

57:11

and there are other changes that occur

57:13

in these cells in the bone marrow,

57:15

and there are other cells that change

57:17

in the gut and villus lining cells,

57:19

and the crypt cells, so they are

57:21

all epigenetic in some. are caused by

57:24

some things, and we don't really know

57:26

which, if any of these, count for

57:28

aging. If someone says, I'm going

57:30

to prove that an epigenetic

57:32

change is responsible for

57:35

aging, they haven't begun to come

57:37

to grips with the nitty-gritty.

57:39

People always ask, just as

57:41

you hinted, does your drug

57:44

change epigenetic things? And

57:46

unfortunately, that's where they stop

57:48

thinking. We're always willing to give people

57:50

tissues from our drug-treated mice if they

57:53

are keen on epigenetic changes that affect

57:55

neuron regeneration. Excellent. Their experts will send

57:57

them the brains and they can do

57:59

that. stuff, it's important. I'm not making

58:02

fun of it, but the general notion

58:04

that it's aging vaguely thought of is

58:06

due to epigenetic change, more vaguely thought

58:08

of, doesn't really get you anywhere. That's

58:11

my skeptical view. Is part of the

58:13

issue that you're saying, well, what's causing

58:15

the cause? No, it's just that the

58:18

concept of epigenetic change encompasses thousands of

58:20

changes in hundreds of cell types under

58:22

hundreds of influences. Of course, some of

58:25

that causes other stuff agreeing to that,

58:27

assenting to that notion that epigenetic change

58:29

is causal for all sorts of age-related

58:31

pathologies. Everyone can agree to that, but

58:34

it's meaningless, because what counts is to

58:36

say, this specific change is really important

58:38

in this disease. Here's an epigenetic alteration,

58:41

or this specific broad spectrum change in

58:43

multiple tissues. causes something good or bad.

58:45

You have to define what it is

58:48

before you can test it. So let's

58:50

use a specific example. When you look

58:52

at a patient with type 1 diabetes

58:54

and you look at their beta cells

58:57

in their pancreas, they look different epigenetically

58:59

than the beta cells of an age-matched

59:01

person without type 1 diabetes. And we

59:04

also know that their beta cells don't

59:06

function. So they've lost function. So let's

59:08

ask that question as a specific example.

59:11

Do you believe or what confidence would

59:13

you assign to the notion that the

59:15

epigenetic change on the beta cells of

59:17

the type 1 diabetic are indeed causal

59:20

to the loss of function of the

59:22

beta cell? My last exposure to the

59:24

causes of type 1 diabetes was on

59:27

eyes in medical school, which is more

59:29

than five years ago, but if I

59:31

vaguely remember it was an autoimmune disease,

59:33

right? So if your poor little helpless

59:36

beta cells are being attacked by antibodies

59:38

and macrophages and things... those stress reactions

59:40

are going to cause epigenetic change. And

59:43

when those epigenetic changes contribute to some

59:45

extent to the ill... fate of the

59:47

beta cells, it's possible. And if I

59:50

were an expert on diabetes pathogenesis, I'd

59:52

really want to know that. It doesn't

59:54

have anything to do with aging, but

59:56

it's an interesting question. But it's a

59:59

way to address causality. Yeah, but you

1:00:01

might equally say, no, it's the mitochondrial

1:00:03

that have changed. There are a hallmark

1:00:06

of diabetes. Yeah, or it's the glycated

1:00:08

proteins. There's a ton of things, and

1:00:10

there's no reason in the world at

1:00:13

this stage, I think, to actually give

1:00:15

epigenetics genetics primacy. over anything else. It's

1:00:17

a nice hypothesis. It's a hypothesis. You

1:00:19

can formulate these questions because a lot

1:00:22

is known about type 1 diabetes. And

1:00:24

I understand 0.05% of the biology of

1:00:26

aging. Zero zero five. Yeah. I was

1:00:29

giving... You're off by an order of

1:00:31

magnitude. He was raising you by a

1:00:33

log. Yeah. I thought you would log

1:00:36

off. Formulating the questions in exactly the

1:00:38

way Steve did. makes it clear how

1:00:40

difficult it is to evaluate the concept

1:00:42

that epigenetic change contributes to pathogenesis and

1:00:45

type 1 diabetes. And we know more

1:00:47

or less what is going on in

1:00:49

type... We don't know what's going on

1:00:52

in aging. We don't even know what

1:00:54

part of the body is going on

1:00:56

or parts more likely of the body.

1:00:59

I at least internally reframe it a

1:01:01

little bit and say, what would the

1:01:03

experiment be? What would you need to

1:01:05

do to convince yourself that either broadly

1:01:08

speaking epigenetic dysregulation causes aging, whatever that

1:01:10

means, or this specific epigenetic change that

1:01:12

is associated with chronological age causes aging?

1:01:15

And so that's an easier way for

1:01:17

me to think about it because I

1:01:19

feel like it's all a... fascinating conversation,

1:01:22

but we're never going to get to

1:01:24

the answer until somebody actually does the

1:01:26

experiment. Or decides that it can't be

1:01:28

formulated because it's too complicated. It gives

1:01:31

up. Yeah, that's right. But people are

1:01:33

trying to do both of those things.

1:01:35

I mean, people are using partial or

1:01:38

transient epigenetic reprogramming and asking can that

1:01:40

have effects on biological aging. I'm actually

1:01:42

cautiously optimistic it can. I don't think

1:01:44

it's going to be a game changer,

1:01:47

but I think you can modulate aspects

1:01:49

of biological aging. epigenetic modifications, so if

1:01:51

we think this particular epigenetic mark at

1:01:54

this particular location in the genome controls

1:01:56

aging. And I don't think it's going

1:01:58

to be that simple, but let's say

1:02:01

it is. You could go in, you

1:02:03

could modify that, and then see, do

1:02:05

you reduce disease? Do you increase lifespan?

1:02:07

Do you improve health span? So those

1:02:10

are the kinds of experiments that I

1:02:12

think would get us to where we

1:02:14

can have a lot of confidence. If

1:02:17

it's the case, if somebody, let's say

1:02:19

at Altos, at Altos, publishes a paper,

1:02:21

Three years from now, that they have

1:02:24

made a mouse live six years by

1:02:26

multiple rounds of transient epigenetic reprogramming, I'll

1:02:28

be like their biggest fan. They moved

1:02:30

the needle. That convinces me that that

1:02:33

strategy modulates biological aging. Nobody's done that

1:02:35

yet. What about something far less impressive,

1:02:37

but still with WALC? So consider, if

1:02:40

we could get to the point where

1:02:42

we could locally deliver vectors that would

1:02:44

epigenetically change chondracites... so that you could

1:02:47

take osteoarthritis in the knee and just

1:02:49

regenerate cartilage, regenerate cartilage by changing the

1:02:51

epigeno. But is that biological aging? I

1:02:53

wouldn't be convinced that's modulating the biological

1:02:56

aging process. I would be convinced that's

1:02:58

a clinically useful strategy for people who

1:03:00

benefit from that therapy. I guess it

1:03:03

kind of depends on why we think

1:03:05

an individual would be experiencing osteoarthritis. How

1:03:07

much of that is senescence? How much

1:03:10

of that is inflammation? Before we go

1:03:12

down that and ask, is it the

1:03:14

S? Yeah, yeah, let's talk about Sonecis.

1:03:16

If you think osteoarthritis of the knee

1:03:19

requires a knee joint replacement and that's

1:03:21

going to help your patient, you are

1:03:23

not rejuvenating. It's perfectly possible to do

1:03:26

great things with technology, including chondrocyte regeneration,

1:03:28

without having to decide that that's related

1:03:30

to aging. People don't age because they

1:03:33

fail to have titanium knee joints or

1:03:35

something. And one way I think about

1:03:37

this, and again this may be completely

1:03:39

wrong, but it's a useful way for

1:03:42

me to think about it, is I

1:03:44

think about age related disease as the

1:03:46

down. extreme effect of biological aging. For

1:03:49

most diseases, there becomes a point where

1:03:51

the pathology of that disease mechanistically is

1:03:53

no longer the same as biological aging.

1:03:56

In which case- He's very good. You

1:03:58

should listen to him. And one of

1:04:00

the implications of that is the interventions

1:04:02

that slow biological aging may not work.

1:04:05

Once you get past that point, but

1:04:07

things that do work for that disease

1:04:09

may have nothing to do with biological

1:04:12

aging. Does that make sense? Yeah, go

1:04:14

deeper on that idea, though. Let's use

1:04:16

the example. My favorite disease. Let's talk

1:04:18

about cancer. Cancer is an easy one.

1:04:21

We know with cancer in many cancers,

1:04:23

the process is you have one or

1:04:25

more mutations, which then often lead to

1:04:28

additional mutations. You get genome instability. Eventually

1:04:30

you get an oncogen that gets activated

1:04:32

and that leads to uncontrolled cell division.

1:04:35

There are more tumor suppression gene that

1:04:37

gets deactivated. Yeah, right. And if we

1:04:39

accept that immune surveillance is one important

1:04:41

anti-cancer mechanism. We know that immune surveillance

1:04:44

declines with age, so early on we're

1:04:46

clearing a lot of our cancers. As

1:04:48

our immune system declines, these cancers are

1:04:51

going to escape immune surveillance, they're going

1:04:53

to accumulate all these mutations, they're eventually

1:04:55

going to go into uncontrolled cell division.

1:04:58

That uncontrolled cell division at that point?

1:05:00

You can treat the cancer, but uncontrolled

1:05:02

cell division is not biological aging. it's

1:05:04

not a part of the normative aging

1:05:07

process. So the treatment there, so the

1:05:09

mechanism now is fundamentally different from normative

1:05:11

aging, and the treatment, let's just say

1:05:14

the treatment in this case is chemotherapy,

1:05:16

might benefit the cancer. It's not going

1:05:18

to do normal aging, right? And I

1:05:21

think rapamycin is a good example here

1:05:23

where I think we all believe that

1:05:25

rapamycin and inhibiting amtour slows biological aging,

1:05:27

at least in up to mice, hopefully

1:05:30

in dogs, hopefully in people. Yeah. So

1:05:32

it's a fundamental... node in the network,

1:05:34

that's the way I think about the

1:05:37

hallmarks of aging. It's a node in

1:05:39

the network that underlies the hallmarks of

1:05:41

aging. So we can manipulate amtor with

1:05:44

rapamycin, slow aging. Rapomycin is a pretty

1:05:46

good anti-cancer drug until the cancer is...

1:05:48

have evolved to ignore the M-tor break,

1:05:50

then then rapid mison doesn't work anymore.

1:05:53

And we know rapid mison doesn't work

1:05:55

for most cancers. That's an example. That's

1:05:57

been tested. We know that. Yeah, absolutely.

1:06:00

And it's because the cancers evolve to

1:06:02

bypass the M-tor break or to bypass

1:06:04

the ability of rapid mison to inhibit

1:06:07

M-tors. That's a really good point that

1:06:09

we all take for granted that I

1:06:11

think is worth noting. Rapomycin can be

1:06:13

unsuccessful as a chemotherapy agent and can

1:06:16

yet be very successful as a cancer

1:06:18

preventive agent. Absolutely. And it's exactly for

1:06:20

that reason. And I think this also

1:06:23

illustrates why traditional disease-based medicine is not

1:06:25

about the biology of aging. It's about

1:06:27

something of the biology of aging is

1:06:30

distinct and it needs to be investigated

1:06:32

in a different way. And we know

1:06:34

that in the aging field, but the

1:06:36

people in the cancer field, in the

1:06:39

cardiology field, in the neurology field, I

1:06:41

don't think they understand that. This gets

1:06:43

to, if I were health czar, this

1:06:46

is what I would do, because it

1:06:48

comes back to what Rich said at

1:06:50

the outset, which is, why is this

1:06:52

a zero-sum game? I mean, you didn't

1:06:55

ask it that way, but that's effectively

1:06:57

the problem you're dealing with, which is...

1:06:59

Why can't we study cardiology, oncology, and

1:07:02

neurology and aging without everybody feeling like

1:07:04

a thing? My way of saying that

1:07:06

in Peter terms is we need to

1:07:09

have medicine 2.0 and medicine 3.0 in

1:07:11

parallel because the tools of the medicine

1:07:13

2.0 scientist and physician, which we see

1:07:15

on display today, are putting the stent

1:07:18

in, giving the chemotherapy, lowering the cholesterol,

1:07:20

all of these things, the medicine 3.0

1:07:22

toolkit looks different. Different science. You're going

1:07:25

to use wrap a mice in here,

1:07:27

you're not going to use it over

1:07:29

here because it's too late. Instead of

1:07:32

saying one or the other, why isn't

1:07:34

it both? Why wouldn't we want both

1:07:36

of these running in parallel? Well, we

1:07:38

would, but of course, the zero-sum game

1:07:41

is a pretty good analogy for what's

1:07:43

actually going on. The amount of research

1:07:45

dollars at least available to NIH is

1:07:48

not infinitely expensible. It's set by a

1:07:50

complex political process that... allocates it amongst

1:07:52

institutions, so saying that it would be

1:07:55

a good idea to have more funds.

1:07:57

I agree with you, and I'll bet

1:07:59

these two guys do as well, but

1:08:01

it's not easy to do. Yeah, I

1:08:04

think I misspoke. It will be a

1:08:06

portfolio of reallocation. But it will be

1:08:08

worthwhile because the burden of this disease

1:08:11

will be lower. So in other words,

1:08:13

it's sort of like saying, right now

1:08:15

I spend $100,000 a year on the

1:08:18

barrier to my house to prevent anybody

1:08:20

from breaking in. and I spend $100

1:08:22

a year patrolling the neighborhood to make

1:08:24

sure there aren't too many bad guys

1:08:27

in the neighborhood. There's a scenario where

1:08:29

if your total budget is $100,000 and

1:08:31

$100. Maybe you could spend $80,000 in

1:08:34

total by spending more money patrolling the

1:08:36

neighborhood. I think we generally agree with

1:08:38

you that having a greater proportion of

1:08:41

available research dollars, both private and public,

1:08:43

going into the biology of aging and

1:08:45

its impact on late life health, would

1:08:47

be a good thing. I don't think

1:08:50

you're going to get an argument from

1:08:52

anybody in... the cardiology field, the neurology

1:08:54

field, or Alzheimer's, the Alzheimer's field. It's

1:08:57

their money. But wouldn't some of those

1:08:59

people, as the funding dollars move towards

1:09:01

the aging side, also want to move

1:09:03

and say, look, I'm going to study

1:09:06

this through the aging lens? I was

1:09:08

on the council for the National Aging

1:09:10

Institute for three years, and if at

1:09:13

any point I can swear to this

1:09:15

from personal testimony, somebody would say something

1:09:17

like, I wonder if maybe a few

1:09:20

percent of the Alzheimer's budget. might instead

1:09:22

go to studying how slow aging models

1:09:24

would have an impact on late life

1:09:26

neurodegenerative disease, the next day the director

1:09:29

of the aging institute would get a

1:09:31

call from two or three Congress people

1:09:33

who were on the Appropriations Committee stating

1:09:36

that this will not be happening because

1:09:38

there was an Alzheimer's association person who

1:09:40

got the call from the NIA staff

1:09:43

member in charge of Alzheimer's saying, tell

1:09:45

the congressman to call the director. and

1:09:47

let's put a stop to that reckless

1:09:49

idea. They're tied in to the political

1:09:52

process in ways. Well, we just need

1:09:54

to go maybe one step further because

1:09:56

those Congress people have a boss. They

1:09:59

report to somebody too. Would that be

1:10:01

at the stage? Yeah, no. I mean,

1:10:03

come on. Maybe it's because the public

1:10:06

doesn't understand this. Those people answer to

1:10:08

the public. That's a good example. These

1:10:10

are our dollars. That's right. That are

1:10:12

going on with you. But Alzheimer's association.

1:10:15

I mean, that's a patient advocacy group.

1:10:17

That is the public. Yes, although let's

1:10:19

ask the question, what have they done

1:10:22

for those patients lately? That's a different

1:10:24

question, but I mean, I'm just reinforcing

1:10:26

what you said. I think part of

1:10:29

this is educating people. If you know

1:10:31

somebody who's suffering from Alzheimer's disease, you

1:10:33

know very well that the only thing

1:10:35

we've got going for us right now

1:10:38

is prevention. We don't have too many

1:10:40

silver bullets in the treatment gun. Despite

1:10:42

massive spending. I was once in Congress

1:10:45

trying to lobby. with about six people

1:10:47

from the Alzheimer's Association in the same

1:10:49

room. I was totally ignored by staffers

1:10:52

that were in there. I mean, I

1:10:54

agree with all of this. I agree

1:10:56

with all of this. I think, again,

1:10:58

though, we should be careful not to

1:11:01

demonize people for wanting to cure Alzheimer's.

1:11:03

It's a good thing. It's a good

1:11:05

goal. I think the communication piece is

1:11:08

about the fact that it's going to

1:11:10

be much more efficient and effective to

1:11:12

keep people from getting it in the

1:11:15

first place. Once you've outpaced the biology

1:11:17

of aging with the pathology of the

1:11:19

disease, it gets a lot harder, a

1:11:21

lot harder to do anything about it.

1:11:24

So I think that communication part, honestly,

1:11:26

I don't know why we've been so

1:11:28

unsuccessful, because I think a lot of

1:11:31

us have been out there trying to

1:11:33

communicate this message for a long time,

1:11:35

but it's starting to permeate. We're at

1:11:37

that moment, I think, where people are

1:11:40

starting to get it, that biological aging

1:11:42

is a thing. It's malleable. It's going

1:11:44

to take a little while, but there's

1:11:47

reason to be optimistic. And there's also

1:11:49

the private sector is another reason, I

1:11:51

think, to be optimistic. So let's go.

1:11:54

record right now I think when we

1:11:56

if we defeat Alzheimer's disease it's going

1:11:58

to be because of the biology of

1:12:00

aging it's not going to be because

1:12:03

of the drugs that get rid of

1:12:05

absolutely yeah probably cancer probably heart disease

1:12:07

although I think Peter's more optimistic we

1:12:10

can prevent heart disease if you took

1:12:12

the tools of medicine 2.0 and just

1:12:14

applied them 30 years earlier we wouldn't

1:12:17

have ASCPD that's the one place where

1:12:19

it's word but again that's because the

1:12:21

mechanism of action is so well understood

1:12:23

with ASCD compared to Alzheimer's and cancer

1:12:26

a lot of infectious disease, a lot

1:12:28

of liver disease, a lot of kidney

1:12:30

disease, all of those things can be

1:12:33

improved dramatically by targeting the biology of

1:12:35

aging. You know, if I were to

1:12:37

write my book again, I would add

1:12:40

a fifth horseman, because I talked about

1:12:42

these four horsemen of ASCVD, cancer, neurodegenerative

1:12:44

and demanding diseases, and metabolic disease, but

1:12:46

I would actually add a fifth hallmark.

1:12:49

It's not really a hallmark of disease,

1:12:51

but it's the fifth thing that brings

1:12:53

life to a bad close, which is

1:12:56

immune dysfunction, which is immune dysfunction. And

1:12:58

I don't think I gave that enough

1:13:00

attention in the book, because of course,

1:13:03

as you said, it factors in very

1:13:05

heavily to oncogenesis, but also, as COVID

1:13:07

showed us, what a risk factor it

1:13:09

was to be old. And you know,

1:13:12

I'm reminded of this when I see

1:13:14

people my age get brutal pneumonias, and

1:13:16

like two months later, they're okay. And

1:13:19

you realize, two of my patients actually

1:13:21

in the past six months have had

1:13:23

really bad pneumonias. where you're looking at

1:13:26

the CT of their chest and you

1:13:28

cannot believe they're alive. But of course

1:13:30

they're fine. Three months later, four courses

1:13:32

of antibiotics later, they're fine. And you

1:13:35

realize, you do that to a 75-year-old,

1:13:37

it's over. And it simply comes down

1:13:39

to how their B cells and T

1:13:42

cells work. That to me is an

1:13:44

area where I'd love to see more

1:13:46

attention, which is, what would it take

1:13:49

to rejuvenate the immune system as a

1:13:51

proactive statement? I think that that's a

1:13:53

perfect example. Influenza pneumonia has never fallen

1:13:55

out of the top 10 causes of

1:13:58

death in the US. You know, it

1:14:00

used to be number two, but still

1:14:02

now it's number eight or nine, but

1:14:05

it's always there because you can't really

1:14:07

do anything about the late life immune

1:14:09

dysfunction. Just to follow this up, if

1:14:11

magically you become in charge and you're

1:14:14

able to double the amount of research

1:14:16

being done on the biology of aging

1:14:18

fundamentally, then we can afford to do,

1:14:21

let's give some mice to start with

1:14:23

a batch of anti-aging drugs. and see

1:14:25

if it makes them more resistant to

1:14:28

infectious illnesses, including pneumonias, but viral infections

1:14:30

as well and many others. I'd love

1:14:32

to know the answer to that, and

1:14:34

no one has actually really looked in

1:14:37

a serious way, because the ITP has

1:14:39

enough money to just measure lifespan. None

1:14:41

of the health span. Just hoping that

1:14:44

everybody else is now going to look

1:14:46

at the brain and the lungs and

1:14:48

the infection, the sensory systems. That really

1:14:51

ought to be done, and it's not

1:14:53

being done because of a lack of

1:14:55

a lack of money. You said something

1:14:57

a while ago, Rich, that I think

1:15:00

is timely now, which is with each

1:15:02

generation of these drugs, they get more

1:15:04

efficacious and less toxic. Not yet, but

1:15:07

that's the hope. Well, no, no, but

1:15:09

I'm going to use another example. The

1:15:11

GLP1s are the best example of this,

1:15:14

right? So you go back to the

1:15:16

very, very first generation of GLP1 agonists

1:15:18

barely lost any weight, horrible side effects,

1:15:20

horrible side effects, horrible side effects. Fast

1:15:23

forward to semi-glutide, quite a bit better

1:15:25

efficacy, still really bad side effects. Next

1:15:27

generation trusepetite, better efficacy, side effects are

1:15:30

almost gone. Now, why haven't we been

1:15:32

able to do that with these geo-protective

1:15:34

drugs? So we have this one study

1:15:37

using Evrolamus that gives us a hint

1:15:39

that says, hey, this might actually enhance

1:15:41

immune function in people in their mid-60s.

1:15:43

But we need the follow-up study, the

1:15:46

follow-up drug, the follow-up drug, imagine what

1:15:48

the fourth generation of that drug can

1:15:50

do. where it's tuned to get better

1:15:53

and better and a fewer side of

1:15:55

it. There are strong commercial motivations. You

1:15:57

know you're going to sell a lot

1:16:00

of the obesity drugs. They're very strong.

1:16:02

martial motivations to do those studies over

1:16:04

and over and over again until you

1:16:06

find one that works better. And there

1:16:09

are good preclinical models that you can

1:16:11

use that you're not wasting too much

1:16:13

of your time on clinical trials. That

1:16:16

could be done for anti-aging drugs as

1:16:18

well, although testing anti-aging drugs and people

1:16:20

is a whole separate set of tangle

1:16:23

of difficulties. I don't want to talk

1:16:25

about that right now. But I'm saying

1:16:27

it won't be quite as easy as

1:16:29

it was for anti-obacy medications, but no

1:16:32

one's doing even the first level of

1:16:34

research to find the optimal compounds for

1:16:36

efficacy without side effects or even to

1:16:39

begin to see if they have desirable

1:16:41

effects on aging rate indicators in people.

1:16:43

That's kind of a cheap and easy

1:16:45

study and no one has really tackled

1:16:48

that yet. Well, I just heard that

1:16:50

there are over 80 synolithic studies in

1:16:52

early clinical trials. It's a joke. It's

1:16:55

a joke. It's a joke. It's a

1:16:57

joke. We have to come back to

1:16:59

this. Are any of them powered for

1:17:02

anything other than safety? This is, I

1:17:04

think, a problem. Oh, it's all phase

1:17:06

one, exactly. So they're underpowered. They're almost

1:17:08

useless, in my opinion. Well, until they

1:17:11

get to phase two, phase three. Well,

1:17:13

until they get to phase two, phase

1:17:15

three, phase three. Well, until they get

1:17:18

to phase two, two, two. There's lots

1:17:20

of complicated issues here. I think endpoints

1:17:22

for clinical trials are really challenging, but

1:17:25

solvable. So there are two places I

1:17:27

wanted to go next, and I'm going

1:17:29

to let Rich decide, because he's going

1:17:31

to have the strongest point of view.

1:17:34

Can we talk about senescence? Or can

1:17:36

we talk about what biomarkers would be

1:17:38

necessary to help us study aging in

1:17:41

humans as we translate from your work

1:17:43

and Matt's work? I know what I

1:17:45

want to talk about, and it's the

1:17:48

second of those, too. I don't want

1:17:50

to spend the next three or four

1:17:52

hours explaining why senescence is silly and

1:17:54

anti-senalytics are untested at best. So there's

1:17:57

no way we're not talking about that.

1:17:59

Let's go on to item number two.

1:18:01

And I think the most important thing...

1:18:04

is to make a clear distinction between

1:18:06

biomarkers and aging rate indicators. Please explain

1:18:08

the difference to people, please. Okay, I'll

1:18:11

do my best. So a biomarker, allegedly,

1:18:13

and in real life, is something that

1:18:15

changes with age. So if you have

1:18:17

some drug that slows aging, the biomarkers,

1:18:20

many of them in the different cell

1:18:22

types and in the blood will change

1:18:24

more slowly. They are a good way

1:18:27

of looking at whether you're slowing it.

1:18:29

rate of change of biomarkers,

1:18:31

very established part of the

1:18:33

literature, and valuable. But you

1:18:35

have to wait till somebody's old, whether

1:18:38

it's a dog or a mouse or

1:18:40

a person, because only when they're old,

1:18:42

has the bio marker of aging, the

1:18:45

surrogate marker for biological aging, changed very

1:18:47

much. So in a clinical trial, certainly

1:18:49

in a human situation, no

1:18:51

one wants to wait 20 years to

1:18:53

see whether the biomarkers have changed. And

1:18:56

a one-year is such a tiny fraction

1:18:58

of a human lifespan that you

1:19:00

don't really anticipate detectable change

1:19:02

with a appropriately powered study.

1:19:04

It's like aging rate indicators, which

1:19:07

are much less well studied and much

1:19:09

less well established in principle, are things

1:19:11

you can measure that tell you whether

1:19:13

in a slow aging state or a

1:19:15

normal state. Can I just make a point

1:19:17

for the listener so they understand

1:19:19

the challenge of what we're talking

1:19:21

about? When we study blood pressure

1:19:23

drugs or cholesterol drugs, The

1:19:25

biomarkers change so rapidly

1:19:27

and we know the relationship

1:19:30

between the biomarker and the

1:19:32

disease state. So if your

1:19:34

blood pressure is 145 over 90,

1:19:36

on average, before I give you

1:19:38

this ace inhibitor, and three months

1:19:41

later, six months later, nine

1:19:43

months later, your blood pressure

1:19:45

is averaging 119 over 74, I

1:19:48

know I've done something well. Now I will

1:19:50

still probably in the phase three, in fact

1:19:52

I will in the phase three, have to

1:19:54

make sure that I also reduce some event

1:19:56

in you. But generally by the phase

1:19:58

two, I know that this drug is... not toxic

1:20:00

and that it's predictably lowering your blood pressure,

1:20:02

that's really, really valuable. A biomarker generically

1:20:04

is something that's easy to measure that is

1:20:06

informative about something that's hard to measure.

1:20:09

A classical example, a famous example is you

1:20:11

want to know how many cigarettes somebody

1:20:13

smokes a day, they'll lie to you. But

1:20:15

if you measure cotoning in their blood, that's

1:20:17

a byproduct of nicotine, you don't have to

1:20:20

ask them. You can find out how many

1:20:22

cigarettes they had in the last couple of

1:20:24

days by measuring blood. That's a biomarker of

1:20:26

cigarette consumption. Is it a marker of nicotine

1:20:28

or carbon monoxide? I don't know the answer

1:20:30

to that. I wasn't sure. In

1:20:32

principle, a biomarker of aging is, are many

1:20:34

of them, and they are measuring biological aging

1:20:36

processes, and they're useful in that regard. But

1:20:38

they don't tell you how fast you're aging. The

1:20:41

analogy I love to use is an

1:20:43

odometer is like a biomarker of aging

1:20:45

of your car. It tells you how

1:20:47

many miles your car has gone, but

1:20:49

it doesn't tell you how fast the

1:20:51

car is going. The speedometer tells you how

1:20:53

fast your car is going. What we

1:20:55

need, and what I think we're

1:20:57

just beginning now to document, is

1:21:00

things like the speedometer, aging rate

1:21:02

indicators that reliably discriminate

1:21:04

slow aging mice or people

1:21:06

from regular old mice or

1:21:08

people. We have now dozen

1:21:10

or so things that change in the fat,

1:21:13

in the blood, in the liver, in the brain,

1:21:15

and in the muscle that are always changed

1:21:17

in any slow aging mouse, whether

1:21:19

it's drug A, drug B, drug C,

1:21:21

calisthenic diet, or single gene

1:21:23

mutations. We've looked now at

1:21:25

five different single gene mutations, and

1:21:27

this whole set of 12,

1:21:29

or roughly 12, aging rate

1:21:31

indicators always changes in every slow aging

1:21:34

mouse, and it does so in youth,

1:21:36

which is the key point. So,

1:21:38

if it does so quickly after

1:21:40

an anti -aging drug is administered,

1:21:43

that's the transition. That's the bridge

1:21:45

you need for clinical studies in

1:21:47

people. If you want to know whether

1:21:49

metformin or canagliflozin or something slows

1:21:51

aging in people, and you don't want

1:21:53

to wait 20 years, but you've

1:21:55

got things that tell you whether they're

1:21:57

in a slow aging state, how

1:21:59

fast... they are aging versus normal, and

1:22:02

that's a big if. We don't yet have

1:22:04

evidence we can do that. We just have

1:22:06

hope we can do that. Then that allows

1:22:08

you quickly, quickly being within six

1:22:10

months to a year, to know whether

1:22:13

your anti-aging manipulation, alleged

1:22:15

anti-aging manipulation, has moved them

1:22:17

to a physiological status which is

1:22:20

associated with slower aging. A lot of

1:22:22

that can be done in mice with drugs,

1:22:24

with mutants. And are these all

1:22:26

proteins rich? No. No, some of them

1:22:28

are changes in the fat,

1:22:30

different classes of macrophages. The pro-inflammatory

1:22:33

macrophages, the anti-inflammatory macrophages, the good

1:22:35

ones, go up. UCP1, I recall

1:22:37

from our discussion. UCP1, goes up

1:22:40

in every one of our 10

1:22:42

different kinds, 11 now of slow

1:22:44

aging mice. Does it go up

1:22:47

in any of the mice that

1:22:49

did not receive a successful drug?

1:22:51

Well, we compare them to controls,

1:22:54

and the... Question you're asking is really

1:22:56

important. That's what we're doing in the

1:22:58

next five years. We just got a

1:23:00

grant to do that. We're going to

1:23:02

take mice and give them either a

1:23:04

good drug or a different drug that

1:23:06

doesn't work and then make those

1:23:08

comparisons. A really important thing to prove.

1:23:10

So far, our only control has been

1:23:13

untreated mice. At some point in this,

1:23:15

I had to bring this up. But

1:23:17

let's imagine that Rich is incredibly successful

1:23:19

at finding these things. That is a

1:23:22

very, very long way. from assuming that it's

1:23:24

going to be the same in people.

1:23:26

Most things that clinically

1:23:28

work in mice do not work in people.

1:23:30

It might be, and that would be

1:23:33

wonderful, but I think ultimately we're

1:23:35

going to have to find this

1:23:37

for people. And my thought is

1:23:39

the kind of evaluation that you

1:23:41

do routinely of your patients.

1:23:43

If we took a group of 65-year-olds

1:23:45

and we gave them a drug

1:23:48

that we thought was an anti-aging

1:23:50

drug, and follow them the next

1:23:52

five or six years doing these

1:23:54

evaluations, I think you could probably

1:23:57

safely say, this is slowing aging.

1:23:59

or not slowing aging. So I don't

1:24:02

think that it's going to be that

1:24:04

easy to jump from ice to people

1:24:06

in this. I've always wondered if in

1:24:08

people the easiest way to do it

1:24:11

would be to take the most obvious

1:24:13

thing that we know is going to

1:24:15

reduce the rate of aging. So it'd

1:24:18

be an interesting experiment, but you find

1:24:20

someone who is overweight diabetic and smokes

1:24:22

and has hypertension. You get hundreds of

1:24:24

these folks. You put half of them

1:24:27

on to be ethical, a plan where

1:24:29

you try to get them to stop

1:24:31

and presumably many don't. In the other

1:24:33

group you pull out all the stops

1:24:36

and you don't care because you're interested

1:24:38

not in testing the hypothesis, does this

1:24:40

thing help you? You're interested in getting

1:24:42

them to lose weight, not have diabetes,

1:24:45

stop smoking, exercise like crazy. The greatest

1:24:47

division between two groups of individuals where

1:24:49

we would I think be able to

1:24:52

agree that this group is now aging

1:24:54

slower, the group that we've... reconcile their

1:24:56

diabetes quit the smoking etc etc etc

1:24:58

and then I'd love to see riches

1:25:01

12 line up in that population that

1:25:03

would be great let me just say

1:25:05

that I think that people that study

1:25:07

animals myself included always underestimate how well

1:25:10

we can evaluate health in people with

1:25:12

a very very thorough evaluation because we

1:25:14

don't do that in our experiment. Why

1:25:17

do you think that is Steve why

1:25:19

is it? Because I was going to

1:25:21

ask about parabiosis later on in the

1:25:23

discussion. We might as well talk about

1:25:26

it now, right? Parabiosis seems to actually

1:25:28

kind of work in certain mouse models.

1:25:30

Do we have any reason to believe

1:25:32

it's going to work in humans? And

1:25:35

if not, why not? Why are mice

1:25:37

so different from people? But wait a

1:25:39

minute. I wouldn't say that just because

1:25:41

we don't have evidence that it works

1:25:44

in humans means mice or when it

1:25:46

comes to parabiosis, right? Yeah, yeah, yeah.

1:25:48

That's a different discussion. translation to people,

1:25:51

most have failed. I actually think that's

1:25:53

because those are artificial mouse models where

1:25:55

they tried to give young mice an

1:25:57

age-related disease. I'm more optimistic. I don't

1:26:00

know this. Rich doesn't have those mice,

1:26:02

yeah. I know. I'm more optimistic that

1:26:04

biological aging or normative aging is going

1:26:06

to be much more likely to translate

1:26:09

to people, both interventions and biomarkers, than

1:26:11

the specific disease interventions. I might be

1:26:13

wrong. I don't know the answer. We

1:26:15

would hope that's a case. That's fair.

1:26:18

I don't think we should rule out

1:26:20

the mice as a useful model. In

1:26:22

fact, I think there's reason to be

1:26:25

optimistic that it will. it's not a

1:26:27

pragmatic approach for population gerotherapeutics. But I'm

1:26:29

just wondering like why it wouldn't be

1:26:31

as efficacious. This is something that, I

1:26:34

mean aren't there six or eight clinical

1:26:36

trials going on right now? Different variants

1:26:38

of that, yeah. Yeah. Yeah. I haven't

1:26:40

seen them. I've seen the one that's

1:26:43

looking at it's not really a parabiosis

1:26:45

study, but it's looking at plasma ferises

1:26:47

for Alzheimer's for Alzheimer's. I consider that

1:26:50

a little bit different, but fair enough,

1:26:52

but fair enough, okay, okay, because they're

1:26:54

just using albuma. studies going on of

1:26:56

young blood. Yeah, okay, okay. But if

1:26:59

you think of peribiosis as both taking

1:27:01

away the bad stuff that accumulates with

1:27:03

age and adding in the good stuff

1:27:05

that's in young, some sort of plasma

1:27:08

exchange hits at least half that equation.

1:27:10

Okay, I'm gonna come back to this,

1:27:12

but my question was, why the difference?

1:27:14

You're saying, Matt, the difference is probably

1:27:17

amplified in disease specific cases like heart

1:27:19

disease, cancer and Alzheimer's disease. probably less

1:27:21

relevant when you're talking about aging, because

1:27:24

even a flawed mouse model still ages.

1:27:26

In fact, it's designed to age in

1:27:28

a certain way. Yeah, and I mean,

1:27:30

I think normative aging looks very similar.

1:27:33

Again, if we look for mice to

1:27:35

dogs to people, just broadly speaking, the

1:27:37

process looks pretty similar. So, I'm cautiously

1:27:39

optimistic that these things are going to

1:27:42

translate. Not to pay too much attention

1:27:44

to Steve's pessimism on this point, although

1:27:46

he's completely right. And it's for a

1:27:49

variety of reasons. Sometimes humans are different

1:27:51

from mice. Sometimes the drug has side

1:27:53

effects that are tolerable in mice, not

1:27:55

tolerable in people, etc. But I always

1:27:58

like to look at the other side

1:28:00

of the coin. That is... If your

1:28:02

goal is to develop a drug that

1:28:04

blunts pain in people and you screen

1:28:07

40 or 50 drugs and you find

1:28:09

a couple that inhibit pain in mice,

1:28:11

that's a really good start. It doesn't

1:28:13

guarantee they're going to work in people,

1:28:16

but it gives you this category of

1:28:18

snail-based neurotoxins. Let's make 40 of those

1:28:20

from 40 different snails. We'll find one

1:28:23

that actually in people works can be

1:28:25

made by a scalable process and doesn't

1:28:27

produce... serious side effects. So the mice,

1:28:29

it's not a one-to-one mapping. It works

1:28:32

in mice, it doesn't, it works in

1:28:34

people, but it's an important critical first

1:28:36

step which usually succeeds in finding a

1:28:38

set of drugs of related families or

1:28:41

with related targets at least that are

1:28:43

efficacious in people. Most drugs that are

1:28:45

used in people had useful rodent-based research

1:28:47

somewhere in their pedigree. Absolutely agree with

1:28:50

that, Rich, and nobody's saying that 100%

1:28:52

of things at work in mice do

1:28:54

not work. But I think there's a

1:28:57

critical difference for aging research, which it

1:28:59

takes four years to do one of

1:29:01

these in mice. And so if we

1:29:03

have to do 40 to find one

1:29:06

or two that works... That's why I

1:29:08

like aging rate indicators. Speed things up.

1:29:10

I'm stepping on your toes, Peter, but

1:29:12

the question I always come back to.

1:29:15

I agree. We need these aging rate

1:29:17

indicators. How do we get to the

1:29:19

point where we get to the point

1:29:22

where we get to the point where

1:29:24

we're confident that they're confident that they

1:29:26

actually work in people, How do we

1:29:28

get to the point that FDA is

1:29:31

confident that they work? That's the only

1:29:33

way you're going to be able to

1:29:35

use them in a clinical trial. I

1:29:37

don't see a path in the short

1:29:40

term. Well, I don't know that we

1:29:42

need that to tell the truth. So

1:29:44

I went to the FDA to try

1:29:46

to get them to approve a trial

1:29:49

of Metformin. And we didn't couch it

1:29:51

in aging because you're right, as soon

1:29:53

as you mention aging, their eyes glaze

1:29:56

over and they're not interested anymore. that

1:29:58

we're fine with that. But that's a

1:30:00

different end point. That's not a biomarker.

1:30:02

My question is that you've merged two

1:30:05

different difficult problems. Problem A, can we

1:30:07

find drugs that slow aging in people?

1:30:09

Problem B, can we... surmount the legal

1:30:11

and political barriers to getting them. That's

1:30:14

not what I was asking. I was

1:30:16

asking, how do we get to the

1:30:18

point? Okay. What I'm saying is that

1:30:21

you are focused on something I don't

1:30:23

have any answers to basically, which is

1:30:25

how do we get the FDA to

1:30:27

develop and approve clinical trials? I was

1:30:30

more interested in a step before that.

1:30:32

It'd be nice to have some drugs

1:30:34

that actually do work to slow aging

1:30:36

in people. But you have to trust

1:30:39

the biomarker of aging rate before you

1:30:41

can be confident that the drug that

1:30:43

moves the biomarker of aging rate works

1:30:45

in people. That's fundamentally what I'm asking.

1:30:48

How do we get to the point

1:30:50

where, let's just take FDA out of

1:30:52

the equation, the four of us would

1:30:55

sit and look at the data and

1:30:57

I'll be like, yep. Well, that's sort

1:30:59

of my thought experiment. I would have

1:31:01

to take an example in humans that

1:31:04

is so egregious, so let's say we

1:31:06

do that. Well, it would make me

1:31:08

worry. It would only show you the

1:31:10

positive signal. It would show you the

1:31:13

specificity and not the sensitivity of the

1:31:15

test. That's the problem. You might miss

1:31:17

the signal. If you found a proteomic

1:31:19

genomic, like if you found a multimodal

1:31:22

signal that detected a difference in rate

1:31:24

of aging between those two very extreme

1:31:26

sets, you might miss it with a

1:31:29

geoprotective drug, which wouldn't be as dramatic

1:31:31

as dramatic as that chain. So what

1:31:33

if I told you that there are

1:31:35

people who claim... There are epigenetic signatures

1:31:38

that do that, that correlate quite well.

1:31:40

They claim with health outcomes, 10-year mortality,

1:31:42

5-year mortality, 3-year mortality in people, and

1:31:44

are measuring the rate of biological aging.

1:31:47

Because it's out there. I mean, it's

1:31:49

in the literature. I mean, this is

1:31:51

not perfect, but it would be one

1:31:54

thing I would immediately think of, which

1:31:56

is I would take a really good

1:31:58

biobank that would have enough samples that

1:32:00

I could sample a bunch of human

1:32:03

stuff. and use an unbiased sample and

1:32:05

a biased sample. So I would determine

1:32:07

an algorithm based on one and see

1:32:09

how well it predicted on another based

1:32:12

on enough samples. That would have to

1:32:14

be true at a minimum. Yeah, I

1:32:16

think it is. I mean... Again, at

1:32:18

least it depends on how much faith

1:32:21

you put in these research studies, but

1:32:23

I mean people have published epigenetic algorithms,

1:32:25

Dunedin Pace is the one that most

1:32:28

people are going to talk about, that

1:32:30

correlate seemingly pretty well, at least with

1:32:32

mortality and with metrics of health span,

1:32:34

for lack of a better way of

1:32:37

framing it. So that exists. And Dunedin

1:32:39

Pace is using something besides epigenetic or

1:32:41

is it only epigenetic? I think it

1:32:43

uses something else. It was trained off

1:32:46

of... other biomarkers and then they found

1:32:48

epigenetic marks that correlate with those other

1:32:50

biomarkers. So it's a correlation to a

1:32:53

correlation, but there's still a correlation. What

1:32:55

do you think, Rich? Well, I wanted

1:32:57

to go back to the example you

1:32:59

gave where you took a lot of

1:33:02

people and gave them intense exercises and

1:33:04

dietary changes to improve their likely health

1:33:06

outcomes, and that's a good place to

1:33:08

start a discussion, because you said every

1:33:11

sensible person would see the treated group

1:33:13

as aging more slowly. And I would

1:33:15

want to ask before I agree to

1:33:17

that, do they also have improved cognition?

1:33:20

How are they doing in cataracts? How

1:33:22

are they doing in hearing? What happens

1:33:24

when you give them a flu shot?

1:33:27

Do they have a great flu shot?

1:33:29

So the things you've pointed to are

1:33:31

really important for both overall health and

1:33:33

for cardiovascular risk and the things linked

1:33:36

to that. So it's nice to know.

1:33:38

But to convince me that you now

1:33:40

have a slow aging group of people.

1:33:42

You need to go beyond the risk

1:33:45

factors for specific common human diseases. If

1:33:47

you could show that, then for the

1:33:49

first time I would be convinced you

1:33:51

had an effective anti-aging manipulation in people.

1:33:54

Currently I don't know that there is

1:33:56

any effective anti-aging manipulation in people. If

1:33:58

your approach got there, that would be

1:34:01

a terrific research model. Well, but now

1:34:03

we're getting into... the definition of aging

1:34:05

a little bit, which is would you

1:34:07

agree that the approach I'm describing would

1:34:10

produce a longer life? It's easy to

1:34:12

produce a longer life if you happen

1:34:14

to have a clinical condition. where you're

1:34:16

tied to a railroad track, and there's

1:34:19

a train coming, you can extend that

1:34:21

woman's life enormously by simply giving her

1:34:23

a knife and cutting the bonds and

1:34:26

letting her walk away from the track.

1:34:28

longevity promoting interventions. Well, if 80% of

1:34:30

people died as a result of trains

1:34:32

on train tracks, that might be a

1:34:35

worthwhile example. But given that 80% of

1:34:37

people die from these four chronic diseases.

1:34:39

I'm all in favor of protecting people

1:34:41