"Sleep Age" – A NEW Predictor For YOUR Health & Longevity

Health Longevity
19 followers
5 Video Views·Sep 29, 2024
The good thing about sleep is that it's involved in everything.
In fact, if you don't sleep, you look at yourself in a mirror
and you see that your skin is not good looking and that you have aged.
So there's absolutely no doubt that sleep is also a very important aspect
of our health and also of longevity.
I want to start by mentioning that a lot of people wonder why we study sleep.
Why is that important?
I will say there are three reasons that I usually give.
The first one is still a very interesting biological process
we don't know much about.
You have heard a lot about circadian rhythm,
which is really like part of sleep as well.
We like to say that we are sleep and circadian researchers,
that there is a lot of genes that fluctuate during the night and day
and that's why you experience jet lag.
It's an enormously important phenomenon.
It affects all the biology.
There is about one-third of the genome that changes,
of the proteomes that changes with circadian time.
And we know quite a bit about that.
But for sleep itself, why you feel better when you sleep
and why you're more rested and you look younger,
we don't really know much about it.
So it's really one biological mystery that we're trying to answer.
The second one is that we certainly know
that there is tons and tons of literature
that shows that if you don't sleep well,
it's definitely a predictor of a bad outcome in general.
And then finally, there are a lot of people with sleep disorders
and it's very common.
You all know sleep apnea.
You've probably heard of insomnia, etc.
And even shift workers, bad sleep habits and sleep deprivation.
And all those have been generally found to be a predictor
or contribute to reduce lifespan or a lot of diseases.
And finally, it's very important to realize that sleep, again,
doesn't affect just the brain, but the brain and the body.
What I like to say about sleep that I like especially
is that it's absolutely ideal for the biological
and technological evolution that's coming today.
Indeed, when you really look at what's happening,
we have a revolution of genomics.
It has been for a while.
We know all the genes.
The human genome was finally completed.
And we really have all this information available.
Now we are starting to study all the proteins of the body,
which is what the genes are doing.
And this is, of course, very important
because it's closer to the physiology.
And we can study all this at once.
But there's also a revolution in hardware.
More and more, we can have smaller and smaller devices.
A lot of you wear devices that can monitor yourself.
And there's a possibility now to monitor your sleep at home
and more and more conveniently, very easily.
And we'll talk a little bit about this.
And finally, there is a revolution in statistics.
We often now, the time of looking at one factor at a time is finished.
We don't want to look at one.
We still measure glomerular filtration with creatinine.
I'm sure a lot of you have doctors.
But in fact, if you measure 10,000 proteins,
you find 500 creatinines.
So it doesn't really make any sense to measure creatinine.
We could do a score with thousands of proteins
and it would be more effective.
So I like to say that studying genetics
and proteomics and metabolomics is really important
because the genetics is always at the root cause.
You cannot modify your genes.
So you know by definition that if you find a gene
associated with something, it's causal.
So proteins are actually more interesting
because they are closer to the physiology,
but they can be modified by both the environment and gene.
But when you find the magic formula
is when you find a gene that modifies something
and the protein that's encoded by the genes
that is even closer to the modification.
An example would be cholesterol, even so it's not a protein.
But we know that cholesterol can be modified by genes
and the genes like the HMG coreductase
that modulates cholesterol are associated with heart attack.
But of course cholesterol depends on the hamburger you just ate
and it's closer to heart attack.
So when you have that chain like from the protein
to the gene, to the cause,
it's the ideal to try to see what you can modify.
So we are doing a lot of work with proteins.
Actually with proteins you can measure 5,000 proteins.
You can very easily predict a protein age actually.
But similarly one of the things we have a lot of trouble to measure right now
is when people are tired, we don't know if it's because they are in permanent jet lag,
that they have this circadian abnormality,
or if it's because they don't sleep well,
or they don't have a lot of wakefulness, enough wakefulness.
And one of the ways we are trying to address that
is by measuring these thousands of proteins.
And we actually find proteins that, not one protein,
but a combination of proteins, a signature with machine learning
that can predict what is your internal body time.
For example, if some people are very tired,
don't do well because they are in Tokyo time,
even so they are here in California,
or because you have not slept.
And I think it's going to help a lot designing better medication,
either shifting your biological clock,
or making you more awake, or making you sleep better,
because we'll know what's the cause of your problem
in terms of being tired or not sleeping enough.
And of course the next thing I like to joke,
the future is not me, it's the device that I'm wearing.
As you see right now, measuring sleep is really inconvenient.
It just doesn't work well, we have all kinds of electrodes,
it just doesn't work.
But now there's this technology that's quite amazing,
where you can actually paste some stickers,
and you already have the electrodes,
and a very small battery, and you can sleep at home,
and you get the same information.
And of course the goal now is to be able to do that regularly.
And what I like to say is that sleep is not only to detect sleep disorders,
but also it's kind of a picture of your brain health,
when you sleep, and your body health.
When you sleep, you have kind of a pre-programmed process that occurs,
and then you just go through these sleep stages,
your heart rate changes in a certain way,
all your body physiology changes,
and we're captive, we're not doing anything.
We might as well measure yourself,
and then try to measure yourself regularly,
and see if something changes, and if we can predict disease.
And we all know that when you develop depression,
you have change in sleep.
When you develop Alzheimer's, you have change in sleep.
With Parkinson's, very often people act out their dreams.
So there's a trove of biomarkers in measuring sleep,
in isolation of all the noise of our life,
for biology, and try to understand how it's predicting disease.
And in fact, that is really facilitated by machine learning,
which now we do a lot of deep learning,
we can pretty much recapitulate how we score sleep now,
with machine learning.
We don't need technicians anymore.
It works better than humans to use machine learning,
because it's so powerful to recapitulate human annotations.
And now we have a way to study sleep pretty much automatically,
without humans.
But the most important for me,
is not only to be able to detect all the sleep disorders,
and look at the impact on health automatically,
but it's to use this device, and maybe predict something more.
And for example, we have a study,
where we did exactly, you have heard,
there's a lot of different biological age.
You can have protein age, you can have skin age,
we also did sleep age.
And as you see, if you study your sleep,
some people have a sleep that's 50 years old,
and they have a sleep of 30 years old.
Some people are 50 years old, and they have a sleep of 80 years old.
And when you measure this difference in age,
that a machine learning will tell you,
studying your sleep,
it will tell you, if you are younger,
and your sleep age is younger,
you know for sure that you are actually going to survive longer.
So we know for sure that having a better sleep,
looking younger, is actually a predictor of mortality.
But what we really want to do now,
is much more ambitious.
We have a study funded by NIH,
where we are looking at 250,000 people,
that have been recorded for sleep studies.
And we have all these physiologies,
heart rate, sleep, not only sleep,
but all the body physiologies that we record at the same time.
And then we are going to look,
we have these patients sometimes 10 or 20 years before,
and now we are going to look at the people who develop MI,
that develop stroke, that develop Alzheimer,
and we are going to use machine learning
to try to see inside the signal,
what could predict some of these diseases.
Now just an example for some people,
you have probably heard of a stress test,
when you want to see if your coronary bed is doing well,
sometimes you do an exercise,
and you look at how the EKG changes.
With sleep study, we have the EKG continuously for 8 hours,
and it changes with different sleep stages.
So we suspect that by studying
how the heart varies across sleep stages,
could give us some new biology to predict maybe a heart attack.
It's a very rich physiology,
and being able to measure it at home,
is really exciting,
because we'll be able to follow people on,
without disturbing them,
and maybe when we see something changes,
we'll say, ah, this person is going to develop depression,
or this person is at risk of a heart attack,
and be able to intervene.
So I think the future is really exciting,
because this combination of using protomics, genomics,
and biomarkers that you have at home,
and measuring outcome,
I think is going to be a completely new way
of monitoring your health,
and a new way to really understand
also the molecular component of sleep.
So that's my curiosity as a scientist,
but most importantly to try to predict
and prevent all these diseases that are occurring.
So I think I'd like to say that sleep,
I think, is a window to health in general,
and I think is also the result of the confluence
of all this revolution, technological revolution,
is perfect to be applied to sleep.
That's why I'm so excited to share this story with you.
Thank you.
