The “Clock” That Can Predict Death

and lots of sausages and they they still live a long life because I eat more salami than anyone you know, you know. short. What if there was a way for you to find out when you're going to die? Basically Final Fantasy without the freaky visuals. Would you want to know? This has been the life's work of Dr. Steve Horvath, an aging researcher, geneticist, and biostatistician at UCLA. In 2013, Dr.

Steve Horvath created the pan-tissue clock, the first of many epigenetic That discovery got a lot of attention because it was so unexpected. Unannounced, unplanned for. Before we continue, this episode is probably reminding you of your own mortality and what's important in life, subscribing to this channel. So please hit subscribe to avoid any deathbed regrets. We all know that every cell in the body contains a code called DNA. But DNA is not static. If it was, then all of the

cells in your body would basically look and act the same. So what differentiates your brain cells from those in your liver? It's a little process called DNA methylation. This involves a methyl group attaching to a specific genetic sequence on a DNA strand, turning that sequence off. So you end up with different cell types determined by which genetic sequences are turned on and off. All seems pretty normal, right? Well, it wouldn't be an episode of American Alchemist if it was normal. So here's where it gets very weird. Basically, by just knowing somebody's DNA methylation

profile, you can tell them how old they are within a very narrow standard deviation. You can do this using epigenetic clocks. Epigenetic clocks are basically algorithms, and when you run somebody's blood through the algorithm, you can tell them what their age is within a fairly narrow error margin. And now the algorithms are getting freakishly close to being able to tell somebody when they're going to die. That gave rise to a new clock which we named after the Grim Reaper because it's our best mortality risk predictor. So,

it's called Grim age clock. I actually have a similar test upstairs called True age, but truthfully I'm a little afraid to take it. So, why is any of this stuff so important? Well, first off, knowing when you're tracking to die might be the first step in prolonging your life. The second and possibly even more important point is that this might be the key to progress in general longevity research. Defeating or slowing aging is really one of the most important tasks of our So, take a listen to the cold, hard truth. Is that death knocking at your

door or just the Amazon delivery guy? Don't wait to find out. Hit subscribe, leave a comment, and be inspired by this week's American alchemist Dr. Steve Maybe you should interview me.

epigenetic clocks, maybe define what they are for an audience that might be unfamiliar. I define an epigenetic clock as a prediction model that uses methylation to estimate age. We had found that using some saliva, you can very accurately estimate the age of a person. But then, 2 years later, I published a paper that describes an epigenetic clock that applies to all human cell types, all tissues, all It's the so-called pan tissue clock.

Then, the next big challenge was to see does it predict lifespan? Do they they predict time to heart disease and many other questions? The short answer is yes to all. So, yes, they do relate to aging phenotypes, one would say, aging conditions. That gave rise to a new clock which we named after the Grim Reaper because it's our best mortality risk predictor. So, it's called Grim age clock. Now, all of the clocks I

described apply to humans. Can't we develop a clock that applies to all species? That word really culminated in a preprint that we just posted. We call it the universal mammalian clock. This feels like a kind of this very important, maybe Archimedes lever in terms of longevity research. We've spent billions on diseases like cancer, Alzheimer's, and Parkinson's, all of which are very important to investigate, but almost nothing on figuring out fundamentally why we age. The reason we can't make progress in longevity is the

fact that we can't run clinical trials on longevity. Previously, you'd have to do an intervention with somebody, and then wait and see, you know, when they when they died. And that would take like a really long time. This seems really effective. You can effectively measure somebody's biological age, or you could predict when they die within a pretty narrow error margin. And you can measure whether an intervention, you know, better sleep or taking DHEA and Metformin, which is an actual study that's been done by Dr. Greg Fahy, can actually affect somebody's biological

age, and you can basically do that in in in pretty short time frames, right? Yeah, that's right. I think people don't appreciate how expensive clinical trials are, especially for aging studies. Let's say you give a 50-year-old this treatment, and then in theory, you need to follow them 40 years to see when did they die, what did they die from. Contrast that with a normal disease like diabetes. You can have a treatment group that all takes a specific medication, and a control group that doesn't. And then you can basically measure the blood glucose of the two groups, and see if

the treatment was effective. With longevity, that's basically impossible. but now your epigenetic profile is a proxy for death. The technical term is surrogate endpoint. Being able to run trials on treatments, interventions, lifestyle changes, and figure out what can make you live longer is an incredibly important thing. The hope is that somebody finds a new anti-aging treatment, and we draw blood samples before the treatment, and let's say 1 year or 2 years after the treatment, and

then we detect that their epigenetic age is reversed by um a certain number of years, you know. Real quick, let's just distinguish between DNA and epigenetics. So, you have a kind of underlying static DNA of epigenetics. Seems somewhat confusing, cuz I feel like a lot of people still have the preconceived notion that genetics don't change over the course of a lifetime, or switch on and off. Genetics is static. You get born with a a certain DNA sequence, you

know? And that sequence is the same in in every cell of your body. However, there are chemical modifications of the And they in certain ways modulate the genes. They tell a gene whether it should be turned on or turned off. They tell your neurons that they're neurons, and your blood cells that are they are blood cells. And these chemical do occur, you know, for example, in response to stress factors, if you smoke

a lot. Epigenetics then uh play a role in terms of cell differentiation and cell formation from sort of a more primordial stem cell state. Yeah, exactly. You know, if you if you don't have methylation, you wouldn't be able to develop, you know? So, a stem cell could never differentiate in other cell types and tissues and organs unless there was methylation present. It It is fascinating though that it's sort of a constant rate of methylation

that maybe you can change at the margins by not smoking or taking DHEA and Metformin or you know, on on on either side. But it sort of makes me speculate that maybe there's some constant force in the universe that's causing methylation over time. Maybe something like gravity or something like that. For whatever reason across almost all mammalian species, DNA seems to methylate at a pretty consistent pace and correlate with age. It's almost as if each species is pre-programmed to die in the same

general age range. So this begets all sorts of important questions. Does DNA methylation cause aging? If it doesn't cause aging, why is it correlated with aging? And if methylation does cause aging, why does this thing that is very important in cell differentiation basically kill us? How is that evolutionarily adaptive? These are all very important questions worthy of investigation. You are right that it's this is constant change, but it's really after development. So let's say in humans after age 20, it's remarkably constant. But yeah, why is it

that let's say after development there's this constant velocity? I call it the clock ticks at a constant rate. Honestly, we don't really know, you know? So um my thinking is there are certain maintenance processes that are possibly quite constant. Um then other people have speculated um that it could be uh tied together with circadian rhythm. But my studies of circadian rhythm don't really support it, you know? So I don't

think that explains it. Whether it's gravity, you know, that's an interesting I um I mean we will analyze astronauts soon. I mean I have some collaborators who want to study astronauts before and their space travel. So, that will be an interesting question. People would think it's increased because they're thinking is they might be exposed to cosmic radiation. Sure. But our studies indicate that radiation like x-rays and so on really do not have

an effect on the methylome. How do we get more people studying interventions and and doing IRBs? It would be good to use these biomarkers in human clinical trials, you know. Ideally, one should really open up the area for crazy ideas, right? A lot has to do with regulatory burden, you know. But in in theory, you know, if there were ways to reduce red tape, lower costs, you know, and then

open up the field. It'd also be good for clinical trials of other drugs cuz if a lot of diseases are sort of downstream of aging or caused by underlying aging factors, you know, it'd be good to know that something that maybe cures depression or uh you know, psychosis or you know, whatever these you're being treated for is also not aging you. It's maybe helping, you know, that the kind of your underlying longevity. Yeah, I couldn't agree more, you know. Like if I

planned a clinical trial, I would always collect some so-called EDTA blood tube, you know, tiny amount of blood and then just run this an epigenetic clock on it. Why? Because maybe I got lucky and this psychotic treatment or whatever treatment it is actually has a beneficial side effect. What is the most interesting work going on in biology outside of what you're doing with epigenetic clocks? I really like, of course, these developments

surrounding cancer therapies, immunotherapies. That's fantastic. When it comes to the aging field, I'm like this idea of using young blood plasma. These kinds of interventions of young plasma also seem to have an effect on the epigenetic clock. I put out a with some collaborators, Harold Katcher and Akshay Sanghavi. They applied it to and we really saw dramatic results. We looked at

blood, heart, and also brain. All of these tissues we found that the young plasma-based treatment um greatly rejuvenated the epigenetic age. In some tissues it was over 50% reduction in age. In other words, this treatment seemed to cut the age in half. But the results are so stunning, so dramatic that I'm actually very nervous about it and but now we will do follow-up studies. Hopefully I get a new data set, you

know, replication data set in about 2 months. Think of it as a pilot study. You know, the pilot study results were dramatic. Let's talk about epigenetic reprogramming a little bit. So, what what makes you excited about that and how does that exactly work? Yeah, so what makes me excited about epigenetic reprogramming is that you have only a handful of genes, really four or five genes or three genes. And if you modify them, it promises to again lead to dramatic age reduction. Um this

certain viruses, you know, to into the eye, and they then stimulate the regrowth of the optical nerve. But you can apply this idea to any other organs, you know, you could think of the sky is the limit of the application, you know. So it's a hot idea, very promising. We will see whether it will pan out. Many people think histone modifications are far superior to methylation because it's more dynamic, you know. Methylation is often more stable, which is a drawback,

you know, you want changes that are more dynamic, for example, responding to healthy diet or so, you know. So why don't I study histone modification? For no good reason other than the technology was never quite developed, you know. So I I used cytosine methylation because the technology was ready for prime time and relatively cheap. If you study histone the costs are much higher, you know. But it's a missed opportunity, you know, because one really should develop

epigenetic clocks based on histone uh, make progress in this field, what would you work on? What would you What would you study? It depends a bit on interest. For the people who like, um, bioinformatics, they are really in a situation where they have data on a silver platter. There's so many opportunities, you know, to

mine the data. What kind of genes or proteins relate to maximum lifespan? What kind of genes change as we administer our favorite anti-aging Do you think we'll ever get to the point in bioinformatics and genetics where we understand exactly which sequences map to which phenotypic traits? Do you think we we we ever get there, or do you think that's just fundamentally elusive thing or misunderstanding of how genetics even works? Yeah, I think it's the latter, actually. I mention it

because I actually got a PhD in statistical genetics, and I was hired as a geneticist at UCLA. I've actually worked in that field. It was a gigantic disappointment to me, right? I mean, you can imagine I spent 10 years on genetic data, and after all that research, I had absolutely nothing to show for. Really nothing, you know? I could have blamed the data. I could have said, "Give me more funding, you know?" But Mhm. if I step back, why did it fail? The answer is actually that that

the signal is so weak. And so, what did I do, you know? I just said, "Let me move to other data types, you know?" Yes. And so, um because my answer is that if you need 5 million people to find an association, what does it tell me? Negligible signal, you know? Yeah. I want to find a signal when I analyze a data set of 100 people. Yes. That's my goal, you know? I I'm looking for low-hanging fruits. My hope was always we we study

exceptional longevity in centenarians or or just generally longevity, and we find them these exciting genes, certain SNP markers. And I think it's fair to say this has been a very disappointing, you know? Do you think some of this I mean, the centenarian stuff I think originally people would assume it's like, you know, it's the Mediterranean diet, or it's, you know, sort of, you know, certain lifestyle factors, or you know, the health of the individuals themselves. And then, I think we started to realize it's it's actually more correlative for

like group usefulness for the people or like the fact that they feel valued by their group and that they feel, you know, part of a community and socially accepted and that sort of thing. So, do you do you think that might affect Yeah, it's a very interesting question actually. I'm not aware of anybody who really studied it, but I I I agree with everything you just said, you know, clearly loneliness is the big killer. People who have lots of friends are very active, you know, they do so much better on many metrics and it would be

interesting to also look at methylation, you know. What's the extent of sort of lifestyle changes we can make now based on our epigenetics if any? Obviously, there's a burgeoning field of like nutrigenomics. Yes. Um, does that affect, you know, what what you eat, when you should sleep, how much you should work out, that sort of thing? I'm kind of a health nut like most people who live in Los Angeles, you know, and so I do everything you just said, but having said that, um, the truth is these, uh, lifestyle choices have

actually a pretty weak effect. It's really shocking how weak the effect is. This explains why all of us know vegans who died fast, early and then you also know the people who smoke and they live till they're 90, you know. Like Trump or Warren Buffett who eat McDonald's every day and they're They just keep going. Yeah, exactly. I don't know. I always like when people eat hamburger and lots of sausages and they they still live a long life. Yeah. Because that's my own, uh,

a sinful behavior. I I eat more salami than anyone you know, you know. All right. Well, thank you for taking the time. I really appreciate Dr. Horvath. This is awesome. Yeah. And uh I enjoyed it. I actually learned quite a lot from you, so. I don't know about that, but that's nice of you to say. I learned a lot. Thank you. Yeah. Yeah. All right. All right. Thank you for having us.