Rhonda: So, I kind of wanted to shift gears and talk a little bit about this epigenetic age acceleration. I mean, we've been sort of talking about how people age at different rates. I think you were a co-author on one of the studies like a few years back, that was one of the big ones that came out where it was like, "People age at different rates," and there was like 18 biomarkers that were looked at. And I think it was PNAS or something, a PNAS paper.

Dr. Levine: Oh yeah, the Belsky paper.

Rhonda: Yes, yes. And it was like, "Look, people are aging at different rates, and you can even look at their faces and it correlates with their, you know, biological age more than their chronological age." And so, to me, you know, clearly there's lifestyle factors, environmental factors that play a role in the way you age. Can you explain to people what epigenetic age acceleration is and what some of the most robust biological, environmental, perhaps social causes of epigenetic age acceleration are?

Dr. Levine: Yeah. So, we usually use this term "age acceleration" to just mean kind of the discordance between your chronological age, so, the age you know that you are and your predicted age based on whether it's GrimAge or PhenoAge or any of these epigenetic clocks. And that, again, we think it's biologically meaningful. So, someone who's predicted much older than they are chronologically are people who are higher risk for disease or mortality.

And so, you know, the next question is why are some people predicted older and other people are predicted younger? And a lot of people think, "Oh, it's just genetic, you know, maybe my family is just high-risk." But actually it seems to have very little impact on your epigenetic age, so, I think they estimate like 10%, maybe at the upper most 20% impact, your genes have that kind of impact on your epigenetic aging rate. And actually probably the majority of it is environment and lifestyle.

And when we look, again, these are not clinical trials, it's looking at epidemiological data, so, just saying, "In the population, the people who are predicted to be older versus people who are predicted to be younger, what are their characteristics?" We find things that are not surprising, so, socioeconomic status is a big thing in terms of differences in epigenetic age but also behavior. So, smoking really accelerates your epigenetic age. Generally, exercise will tend to decrease epigenetic age. Eating we think probably plant-based diet is going to decrease epigenetic age. And then, yeah, a lot of the things, don't drink heavily, get, you know, good-quality sleep, minimize stress, all the things that everyone's mother and grandmother told them to do in life.

Rhonda: How does being a male affect epigenetic age? Because males live on average, what, four years...like, their life span's like four or so years shorter than females', right? Is that reflected in...

Dr. Levine: Yeah, it is reflected in epigenetic. So, on average, again, not across the board but, if you look at the distributions, females on average will have lower epigenetic age than same chronological age males.

Rhonda: Similar question, females undergo menopause when they reach like 50s or something like that, plus or minus, I don't know how many years, but how does menopause affect epigenetic aging?

Dr. Levine: Yeah, so, this is actually a study I did while I was in Steve Horvath's Lab. So, we looked at women who undergone menopause and how long since they'd undergone menopause. And it seems to be that menopause is actually an epigenetic-aging-accelerated event. So, before menopause, women are doing pretty well and then, when they go through menopause, it seems to accelerate their epigenetic age. And we didn't have the kind of data you would want where we'd have the same women pre and post but we can even look at surgical menopause and that seems to also show this kind of accelerated epigenetic aging kind of manifestation.

Rhonda: That kind of leads me into another question, which is do the changes in these methylation patterns, that you and others are measuring, is it pretty stable over the lifespan or are they, like, you know, like, once you hit mid-life...because, like, there's functional aging that really starts to hit you, you start to get, like, late to midlife and then it starts to go down, right? So, does the epigenetic clock mirror that or is it pretty stable?

Dr. Levine: So, it's not stable but it actually doesn't mirror what we think of in terms of functional aging. So, if you think of a frailty index or even mortality risk, it increases exponentially after, let's say, age 30. The epigenetic clocks show a totally different pattern, it's still not linear but actually most of the changes happen during development. So, you have this huge increase in epigenetic age between...we can even measure it in fetal samples...and then it kind of starts becoming more linear and steady around age 20 and then, interestingly, actually slows down again, you know, very late, so, after, let's say, age 80. We don't know why these patterns look this way but, yeah, it's not perfectly stable across the life course.

Rhonda: Okay, this leads to a couple of other questions that, you know, sort of came in my mind. One is then what about, when people get, you know, disease states, so, like, they do get type-2 diabetes or cardiovascular disease, you know, does that then, being in that disease state, in that functional decline, like, state, does that accelerate the aging clock?

Dr. Levine: So, I don't know if we actually know that because we don't have very good...the problem with epigenetic data right now is we don't have good kind of time course data. We're not following people longitudinally, there are very few studies that do this. There are more studies that are starting to but I don't think we've reached the point to say, "I can look at someone's epigenetic aging pre-disease state and then see what happens after they've developed some disease," but I would imagine that it would probably kind of snowball and accelerate. And we do know, not looking at epigenetics, that, once you get a disease, it's actually shorter time to each subsequent disease. So, there does seem to be this kind of accelerating event in aging that occurs.

Rhonda: Biobank data might be a good source, they do a lot of, you know, like, they've got just tons and tons of samples, you know, because they have people come in for, like, routine...

Dr. Levine: Yeah. So, we've talked to them, the problem is that the epigenetic data is not super cheap. So, you know, to do that many people that many times, yeah, you have to come up with quite a bit of funding to be able to do that.

Rhonda: It would be interesting, it would definitely be...

Dr. Levine: Yeah, I know. I'm all for this. The more data samples we can get, I think the better we'll be able to figure this all out.

Rhonda: Right. You also mentioned that the most of the changes are happening during development. And this is also really interesting, mostly...Steve, when Dr. Steve Horvath was on the podcast last time, and this kind of gets into the next section, which is, like, underlying mechanisms causing these changes in the epigenetic patterns, but he had mentioned that, like, he had developed at least, you know, and maybe there's something new now, but back then, a few years ago, there was a few clocks that he had used that could really beautifully measure gestational age.

Which was interesting because measuring the aging process during gestation where you've got this really coordinated program that has very little background noise, right, inflammatory processes aren't going off and all this damage and, you know, this stuff, I mean, it's just a very clean place to, like, you know, measure aging.

So, and then he's also...I think there's a preprint I saw pretty recently where he had developed a universal aging clock.

Dr. Levine: Pan-mammalian.

Rhonda: Yeah, there was like I don't know how many different mammalian samples that were, you know, used to generate the clock, and I don't understand everything that goes into generating it. But I just looked, you know, skimmed it and it was really talking about...this universal clock was also really coordinated with development. And so, you're mentioning development, it poses the question, like, do you think that the epigenetic aging could be sort of like a program, like a program that's regulating aging? Like, is that a possibility?

Dr. Levine: Yeah, I don't think it was a program designed...you know, some people would argue that it is, you know, a program designed to drive aging. Because I think, you know, for species selection, you need things to age and die so that the species [inaudible 00:28:16]. I think it's a developmental program that just doesn't really get turned off and maybe goes a little bit awry as all these other changes start to accumulate in our bodies. But yes, the epigenetic clocks are really tracking something very central to development because most of these changes we can see during development and a lot of the genes that seem to be involved are these developmental genes. We're still, again, not sure what this means or what this program actually is but it definitely is tied to development. But people would also argue that aging is very tied to development. So, there are beautiful experiments in flies where, if you can extend kind of the developmental period, it extends the lifespan of these animals.

And so, development and aging are kind of two dichotomous things that we usually think of, right, that you're going through development, you hit age 20 and...okay, maybe age 20 or 30 is when your aging starts. But there's a lot of great work, even one of my colleagues at Harvard, Vadim Gladyshev, showing kind of when he thinks this ground zero when aging starts, which is, according to him, day eight of gestation. So, there's...

Rhonda: In humans?

Dr. Levine: In humans, yes. So...

Rhonda: When you said, "In flies and fruit flies," Drosophila probably...

Dr. Levine: Yeah, yeah.

Rhonda: "When they extend the development," can you explain that, what do you mean by that?

Dr. Levine: Yeah, so, I think, in this case, actually, the study I'm thinking of, they extended kind of the reproductive kind of age of the fly. So, they can push flies to, what we would consider, like a late fecundity, so, they can develop a little bit longer and don't reproduce until slightly later.

Rhonda: And how do they do that?

Dr. Levine: It was more like selection. So, they're selecting for flies over generations that are going to be these later-fecundity flies, and they show that they also live longer in the end.

Rhonda: Like, is it a genetic...genes that are controlling it?

Dr. Levine: Yeah.

Rhonda: Okay. So, a couple of things here, then back to this...it's very interesting, the development thing. And another thing came to mind from the conversation I had with Steve was he had mentioned, like, if you take a cell that has not been immortalized in tissue culture and then you immortalize it with a component of telomerase, TERT, and you, essentially, overcome cellular senescence, which is one of the hallmarks of aging. Right? When a cell undergoes senescence, I mean, it's pretty much not...I mean, it's still metabolically active but, you know, it's considered sort of the end, right?

Dr. Levine: Yeah.

Rhonda: And these cells, if you continue culturing them and tissue culture, their epigenetic age just keeps going and going and going.

Dr. Levine: Yeah, we're actually doing this exact thing in my lab right now where we use hTERT to immortalize cells and we are just seeing how long...like, there has to be some...like, eventually, can it just continue to change forever? I think two people have looked at things like HeLa cells which have just been changing, you know, they've been evolving for decades. And like at what point does the epigenetic age kind of reach a saturation point? And again, I don't think we know but, yeah, definitely with these immortalized cells, every time you passage them, their epigenetic age keeps...at least it seems to continue to increase over time.

Rhonda: It's interesting. And sort of, on the flip side of that, would be, like you mentioned, like, are the epigenetic aging clocks biomarking something, like, something else that's causing aging? And a study that you were a co-author on, as I was preparing for this podcast, I was thinking about it and I was, you know, in my mind, I was like, "Well, how could you, like, cause something like that would be massive damage to accelerate aging?" And so, I googled "cancer chemotherapy epigenetic clock" and, like, the paper you were a co-author came up on it. I was like, "Oh, this is Morgan's."

Okay, so, I was reading the paper and these patients that had head and neck cancer and they were getting treated for it, radiotherapy, chemotherapy, you know, which causes massive damage, inflammation, these patients, their epigenetic age was measured before the treatment, after the treatment, and then six months later and a year later. And it was so interesting to me because they had aged...like, their epigenetic age had accelerated by 4.9 years right after the treatment. But then, six months later and a year later, like, their epigenetic age had, like, normalized back to baseline. And sub-analysis then showed actually not only did the epigenetic age acceleration of almost five years correlate with inflammatory biomarkers but people that had extremely high inflammatory biomarkers one year later did still experience the age acceleration. So, I'm curious as to what your thoughts are on what that means? Like, to me, I look at that and I go, "Wow, inflammation is causing epigenetic age acceleration," because you see this like graph, right, I mean... So...

Dr. Levine: Yeah. I think, definitely, when we measure aging in blood, we need to think, you know, what is, you know, probably driving these signals that we see. And I would guess that epigenetic age acceleration in blood is mostly reflective of inflammation. Unless, again, you're developing a clock that's specifically tuned to some other thing. Although inflammation seems so, you know, vast and systemic it affects so many different things.

But I don't think everything that epigenetic clocks are capturing is inflammation. Because again, when you look at immortalized cells, it's not because they're becoming more inflammatory every time you're passaging them per se but definitely, I think, epigenetic aging measured in blood is very much tied to inflammation. Which again, is probably why it's highly predictive of a number of diseases which we know inflammation can be a major driver of.

Rhonda: Is that where the extrinsic and intrinsic aging clock...or, I don't know exactly, one of them considers the external factors in blood and one doesn't or something, is inflammation calculated in that or not really? Is it sort of...

Dr. Levine: Yeah. So, these are two of the first-generation clocks, so, I think, you know, Steve kind of called them intrinsic-extrinsic aging. I think he called the original Horvath pan-issue clock was the intrinsic aging, it wasn't that tuned to differences in kind of cell turnover or inflammation. Whereas a clock, that was developed by [inaudible 00:35:08], he kind of added these different kind of cell composition measures that actually ended up picking up inflammation a little bit better. But this was before these second-generation clocks came into being. And then I think, once they came into being, they're probably picking up inflammation a lot more than even the first-generation clocks.

And again, we can make these kind of systems clocks and one of our systems is inflammation, and we can show that it's highly predictive of outcomes, it's definitely capturing things related to inflammation. Preliminarily, I can say we have data from individuals with COVID, and we can look at the inflammation measure and we find that people with severe symptoms have much more accelerated inflammation, epigenetic clock, than people with basically asymptomatic or mild symptoms.

Rhonda: It'd be interesting to see when those symptoms resolve and how long it takes for a person to like go back to more of their baseline ever, hopefully. But so, that's definitely...are you guys going to continue looking at that or...

Dr. Levine: Yeah, I mean, we don't have the ability to track the same people over time but I think this is an important thing. And I think this is important when people start to look at applications of the clocks for intervention testing. Because you can do an intervention that's going to change kind of your blood cell composition and it might be reflective of inflammation but, you know, it could be this acute event. Right? And whether that really means you change your aging I think still needs to be kind of considered.

Rhonda: Yeah. And I think that was the big eye-opener for me when I read this study, I don't know, a couple days ago. And it wasn't a new study but, you know, it was like, "Oh, well, this changed really dramatically." But then it wasn't, like, a permanent thing, I mean, it went back. And so, yeah, it's almost like you're saying with interventions, it's like, "Well, I mean, make sure you didn't get sick," or, like, you know, sick, like, too early before, you know, measuring. And we can talk about that in a little bit, a little bit later, but to get sort of just back into the cause and effect of aging and if the epigenetic clock changes are really causal, I mean, of course, you're, obviously, trying to figure that out, but, like, even if it was, let's say, downstream of something, if it was biomarking aging, what...the epigenetic changes that are happening with aging, you kind of mentioned this early in the podcast about how, you know, they're clustering in gene regulatory regions, and, so, they're changing the way genes are turned on or turned off, is there like a feed-forward loop in aging where it's like, "Okay, now these epigenetic changes are turning off genes that we want on to repair damage and they're turning on genes that are cellular senescence?" you know, so, it's accelerating this feed forward loop...

Dr. Levine: Yeah. I mean it's definitely possible. I think, yeah, it's really hard to figure out causality here, right? Like, and it could be that, you know...I mean my perspective is not, "There's a cause of aging," right, and, you know, "it's this thing and, once you fix that, everything else will go away." I mean so many things go wrong and your system can change, it can diverge in some...going back to kind of Mike Snyder saying, "Even if you bring that down to the molecular level, there's so many different ways that someone's system can kind of change over time." And I don't think it's just this one thing that's going to then drive all of aging.

And yeah, you know, our systems are responsive, right, so, one thing changes, something else is going to respond. And that can be maladaptive, which would, you know, snowball things. So yeah, I think it's going to be hard to figure out like what's causal, what's correlative, but I would say, even if it's not what some people might consider the central driver, as long as it's picking up things that are critical to aging and you can use that to track aging or understand it a little bit better, I think it still has utility. I don't know if it needs to be kind of the central cause of aging for it to be useful.