1 Powerful, Predictive Drop Of Blood: The Future Of Diagnostic Testing, Epigenetics, DNA Methylation & Machine Learning, With Dr. Matthew Dawson.

I think it's important maybe to talk about why aging is important, because I think some people see it as a vanity thing. We value youth so much in our society, but to me it's not about that. Like, age is the number one risk factor for basically every disease. If you look at the top 10 diseases that we have— cardiovascular disease, cancer, dementia— the things that are going to kill us, age is by far the biggest risk factor for all 10 except for motor vehicle accidents. It's not just by a little bit, it dwarfs smoking, obesity, diabetes. So I want to know what's your pace of aging in biologic age is because of that, because of your risk factors for getting those diseases and just your healthspan in general. So that rate of aging is a really critical measure for me, not just how good do you look.

This episode is brought to you by really the best way to nuke zombie cells. When you look at this buildup of senescent zombie cells in your body, It turns out that although these things churn out inflammation like, well, zombies, there's ways you can fight against them. The good folks at Neurohacker Collective are experts and they've cracked the code at nuking these zombie cells in a safe and effective way with just 2 days a month of using their Qualia Synalytic. You get a pack, take 6 capsules one day, 6 capsules the next day, set it and forget it for the next month. So, if you want 15% off the best way to nuke zombie cells, go to neurohacker.com/BGL15 and use code BGL15 for 15% off any purchase. That's neurohacker.com/BGL15 and use code BGL15 for 15% off of any purchase. Enjoy nuking your zombie cells. Kiss them goodbye. He's back, Dr. Matt Dawson, good friend of mine, uh, one of the most forward-thinking, cutting-edge, so-called precision medicine doctors I know, meaning he uses advanced diagnostics to dig really deep into the body. We've had some fantastic shows in the past about executive screening panels and what he does at his company Wild Health. I've visited him at the Castle down there in Kentucky a few times in Lexington. And man, his understanding of functional medicine and now increasingly genomics, we've had a few chats about everything that he's doing as far as looking at epigenetics and DNA methylation testing. I've just realized I really have to get Dr. Dawson back on the show to talk all things diagnostic testing, epigenetics, DNA methylation, machine testing, Starting to sound like a pretty geeky podcast on that.

It's going to be geeky for sure. I can guarantee that.

Well, if people are concerned and grabbing their Post-it notes, don't worry, I'll put all of the show notes over at BenGreenfieldLife.com/DrDawson. Although inside baseball, Matt, you'll like this. This is how I take notes during a podcast, a little high-tech. Yeah, it's super high-tech. And, for those of you not watching the video, I'm holding up my teeny tiny Post-it note stack that I write furiously on while I'm walking on the treadmill. Matt, I got to ask since I know you're nerded out on bone broths and smoothies and I always eat such good food when I'm down there hanging out with you in Kentucky. Have you had anything to eat yet today?

I have. I've had the same thing that I eat every day for lunch. That's 2 cans of matiz sardines with guacamole, pickle juice, and some pickles. It's delicious.

That sounds a little bit like ketogenic meets Mediterranean, man. Yeah.

No, that's true. Yeah, the 2 cans of sardines a day, an amazing amount of omega-3s. I think you eat a couple grams per can.

Yeah.

No, it's called the KetoMed diet. I actually went outside my boundaries, my normal smoothie boundaries this morning because somebody gifted me a pineapple. And normally, I don't do a lot of fruit in the morning. My smoothie is typically a little bit low-carb kind of keto-ish. But despite me usually blending up pineapple and using it as a meat marinade, because as you know, Matt, all the enzymes in the pineapple really help to break down meat, so it's actually one of the best meat tenderizers out there, I made myself a frozen pineapple vanilla protein coconut water smoothie. And dude, it was like an orange creamsicle. I'm pretty happy about it. I'm going to tuck that one away into the smoothie database.

I thought, honestly, I thought your glucose looked a little high when you first came on video, so that's making sense now.

Yeah. Yeah. You know the hack for lowering glucose from a nutritional standpoint if you're not using blood glucose disposal agents like berberine or or using lifestyle strategies like jumping in a cold bath or lifting weights, a way you can literally plummet it down into 50s or low 60s.

How do you do it?

Take a shot of ketones.

Oh, nice.

It kind of artificially suppresses glucose.

Yeah. So, I actually had ketones today too. I forgot about that. So, Delta-G ketones. I'm really into this company and this potential kind of new type of ketones. It seems to be—

Why is it a new type? I've heard of them before. But, uh, is it— is that like a, um, a monoester? Do you know?

So I'm just digging into the science. Uh, I had somebody recently say, hey, you got to check out this company, they're doing really cool things. And so I got— actually, the company reached out and they sent me some as well. So I am not as deep scientifically as I should be on it, but according to their research, what I did see is there's some really cool research they had on athletes, specifically rowers, and those are really cool when you look at like kind of their power output over time. With these ketones versus no ketones or kind of most ketones you buy, there was a really cool— their power just kept increasing, like, kind of at the 20 to 30 minute mark where normally the same folks as a crossover trial would drop off. And again, I don't know the mechanisms yet, so I shouldn't have brought it up, but just seeing the graph, I'm like, I got to try that. So I did a workout, rowing workout this morning, and I definitely felt great.

It's probably an ester. You know, I, I like the fact that a lot of ketones are becoming more commercially available and affordable. And they're using Ketone IQ, HVMN, they're using 1,3-butanediol. And that's good, it's less expensive and allows you to get into ketosis. But from a pure performance standpoint, even though they're way more expensive, some of these ketone esters where you have that same molecule, the butanediol, bound to beta-hydroxybutyrate, you get really high levels for a longer period of time. And, that might be what Delta-G is doing. I think that's also what KetoneAid does with their ketone.

Yeah. It's on my list to dig into it, but I wanted to just try and see how I felt first.

Yeah. You're making me think I should probably interview Dominic D'Agostino again sometime soon. He's always on the cutting edge with that stuff.

The ketone king.

Yeah, yeah. So, we always have these interesting phone discussions and our last one, You told me about how much you're getting into this idea of methylation testing and how much you kind of think it's the wave of the future as far as diagnostics. So, what got you so interested in it?

Yeah. And so, this is something that I've really changed my mind on. In the past, I think I looked at biologic age testing in general and you get biologic age test through DNA methylation testing. I looked at it as more of kind of a party trick. Like, okay, if you're somebody's 40 chronologically versus 36 versus 44 biologically, I'm still gonna try to optimize them. So kind of what is there to do? And I wasn't that interested until kind of the last 6 months when I started digging into like how you really get these biologic age ages and what you can do about it. There's so much more resolution. And it's just, we'll talk specifically about things like EBPs, epigenetic biomarker proxies, and methylation risk scores. But I think first off, the biologic age tests now are finally good enough to where we can actually tell something from them about what to do and change patient management and outcomes. So, I've just gotten really excited over the last 6 months digging into that science. And, I see so much on social media about biologic age testing. I think there's so much confusion. I think we just were talking and I was like, we really should probably address this with kind of an update into what's out there right now.

Ben Greenfield: I mean, it's absolutely confusing according to the wide range of telomere tests. PACE clock tests, glycation tests, et cetera. My biological age, my chronological age is 42. My biological age according to these tests is 8 years old, 22 years old, 42 years old, 47 years old. And, I think I got one that told me I was 55 years old.

And, you may have sent them all the same day. Yeah. The first-generation clocks, I mean, they were they were not good and not useful in my mind. But I believe in them so much, just about 4 or 5 months ago, we started doing them just included in the Wild Health program for every patient when they start and repeat testing at 6 and 12 months because I really think they can tell some great information that we can actually do something about now.

Ben Greenfield: Okay, cool. I want to know about the kind of tests that you're running and the data that you're getting out of them. But backpedaling just a little bit, how would you define epigenetics and DNA methylation testing. I'm sitting next to you on an airplane and I know you're dripping saliva into a tube and I ask you what you're doing.

Ben Greenfield: For sure. So, I think most people who have listened to your podcast, you've talked about epigenetics before. I mean, the word is kind of around genetics or above genetics. So, it's not actually your genetics. So, at Wild Health, every patient we see, we sequence their DNA. That's your human operating system, but that's only about 20% of your health outcome. When you're going to die, how how much healthspan you're going to have, how you feel and perform. The other 80% of that is based on everything that you do. Epigenetics, how you turn those genes on and off, how you express those genes. So that epigenetics, the 80% of your health outcome, that's what DNA methylation is measuring. Looking at 950,000 CpG sites, for example, like the tests that we use, and you get an incredible amount of information from that. And what they can do is they can correlate these CpG sites, how much they're methylated or not, how much they're turned on or off.

And by the way, Matt, what's a CpG site again?

Sure. So cytosine phosphate guanine is what it stands for. So it's a specific place in the DNA strand that is methylated, which keeps that gene. And how methylated it is depends on— will determine how much you're expressing that gene. So if you have good genes versus bad genes, you talk about APOE4, talk about these other genes, you can turn those down or up based on what you eat. How you sleep, what you're exposed to, like you're walking right now, that's changing how you methylate your genes. Everything that we do turns these on and off. And now we can actually very specifically measure that and we can get incredible information. I mean, there's a couple interesting studies looking at smoking pack year. Like if I ask you how much you've smoked in your life, you tell me. When it's one study where they looked at taking a drop of blood, looking at DNA methylation epigenetics, it's more accurate than what you said about smoking. They can tell what zip code you spend a lot of time in based on pollution markers. So you can impute an incredible amount of data. And biologic age and rate of aging is just one of those things that you can tell. It's really, I think, the holy grail of testing where most diagnostic testing— and actually, I said the word grail. So that grail, the Galleri test, the cancer test, a lot of people familiar with, that's DNA methylation testing. So, I think it's where all diagnostic testing is going to go in the future. What you can tell really depends on the size of the dataset and your imagination.

Okay. You got to explain this to me because if I give— I'm assuming there's a— should I say saliva or blood in the case of the type of testing you're doing?

That's actually a really important question. Blood. So, there are saliva tests, but there's just not as much studies and validation around them. I actually don't really trust any company that does saliva testing. So, blood testing is what I'm talking about specifically.

Ben Greenfield: Okay, okay, good to know. So, I give a drop of blood and let's say I smoke or don't smoke and I send that blood off, you're saying that they could determine how much I'm smoking or not smoking based on that blood test. How does it even work? Are they taking what they see in my blood and comparing it to big data from a large population of people who smoke or don't smoke and seeing how my blood markers match up to that, or is it something else?

No, that's exactly right. So like some of the bigger, some of the best biologic age tests, which we can talk about, like one is out of Harvard and they have 100,000 patient biobank. So you can take those patients, they have the medical records to go with them. They know everything about them, their diseases, when they die, all of their other risk factors and are able to find. So I'll give you one example that kind of makes this clear.

Neurosyphilis, for example, if I'm worried that you have neurosyphilis, I've been worried about that for a while, that I have neurosyphilis.

Yeah, I was going to talk to you about it after the show. But if I was worried about that, it's usually misdiagnosed for a few years. But if I finally want to make the diagnosis, I got to send you to the hospital. You get a lumbar puncture, $5,000, complications, a difficult test. Well, Cornell is doing a study right now where they take several hundred patients, take a population that has neurosyphilis, and they take a control population. They take a drop of blood from each, look at these epigenetics, and they can find the fingerprint by looking at those 950,000 sites and how you're expressing your genes. Everything you do creates these cascades of hormones and things that cause you to express your gene in a certain way. So, it's finding that fingerprint of what is neurosyphilis and then you've got to drop a blood test for neurosyphilis, for example.

Ben Greenfield: Okay. So, would this be similar to what they're using for that Rejuvenation Olympics website, the leaderboard that shows how fast or slow certain people are aging? I just interviewed Julie Gibson Clark, the house mom who plays pickleball. And, I think she takes maybe under $100 worth of supplements and she's towards the top of that leaderboard and beating out a lot of these billionaires and entrepreneurs who are doing all these fancy biohacks. But, the DunedinPACE, I think, is the name of the database or measurement protocol that they use for that. From what I understand, they just followed thousands and thousands of people and measured their grip strength and the risk of mortality and cardiovascular disease and all these parameters. And then, when people send in their blood, they're basically seeing where those people fall into based on this large subset of data that they've collected from a large number of people. Is that very similar to what you're talking about?

It is. And, that's my favorite biologic test. So, that was Duke developed that and it was from the Dunedin cohort. It's out of New Zealand. Really incredible data set. They follow these people over 50 years, and so that gives a rate of aging. So that's exactly right. So Rejuvenation Olympics, Bryan Johnson is behind that, that competition globally. They use the PACE test, uh, for sure. That's, that's the one that they use. And I think it's important maybe to, to step back and then to talk about the specific test, just to talk about why aging is important, because I think some people see it as almost a, a vanity thing. Like, we value youth so much in our society, but to me, it's not about that. Like, age is the number one risk factor for basically every disease. If you look at the top 10 diseases that we have— cardiovascular disease, cancer, dementia— the things that are going to kill us, age is by far the biggest risk factor for all 10 except for motor vehicle accidents. And it, and it is not just by a little bit. It dwarfs smoking, obesity, diabetes. So I want to know what your pace of aging and your biologic age is because of that, because of your risk factors for getting those diseases and just your healthspan in general. So, that rate of aging is a really critical measure for me, not just how good do you look.

Ben Greenfield: Well, I would maybe throw snowboarding accidents in there with another category that might not be associated with aging. The young punks out taking psilocybin and marijuana and blasting down the slopes. But besides that, yeah, I mean, it makes a lot of sense. The idea of what you can get from these tests though, you use smoking as an example and neurosyphilis, which of course everybody as we've established is concerned about. What are some other things you're seeing coming back that you're surprised by as far as the data that people are finding out about themselves?

Yeah. And, I should be really clear, those things I was talking about are kind of a little bit future state. There's not the test for nosyphilis yet, there's just a study going. Just kind of example of the data you can give, can get. But what you could do right now is, is the aging with your biologic age and your pace of aging. The reason why I think it's important to talk about now is in the past, those tests weren't good enough. The first generation clocks like the Horvath and Hannam and a lot of those clocks, they— you already mentioned the reason why it's, it's all over the place. Like the Horvath clock, for example, if you were to send the same test on the same day, the mean absolute error is about 4 years.. So you could have a 4-year difference in the same day. So as a clinician, that's not really useful to me. I, I need something that if I wanna test in 6 months, it may not be 4 years older or younger just because of that mean absolute error. So the new clocks are doing that. And like the Horvath clock, for example, it used 353 CPGs. So that's a pretty small subset when we can look at 950,000 CPGs. So the newer clocks have much better mean absolute error. Like you mentioned, the, the PACE. The PACE is incredible in that that the error on that is about 0.5%. So extremely accurate. So when you get a change there, you know it's from something that actually happened. And when they look at those original clocks too, for example, the, the Horvath clock, if it is increased by 1 standard deviation, that's only about a 2% increased risk of mortality. But a 1 standard deviation increase in the Dunedin PACE, this, this one developed by Duke, increases mortality by 64%. So it really correlates to disease and healthspan. And then you gotta, when you look at these clocks, they have to pass a sniff test as well. Meaning when you look at the Horvath clocks, when they tested them against things like calorie restriction, which we know tends to increase healthspan for pretty much every species, those would actually increase the Horvath clock, calorie restriction. But with these newer clocks, the 3 that I like, one out of Harvard, one out of Yale, one out of Duke, those correlate really well with interventions that we know should decrease. So the new clocks, they have better standard deviations or standard error. They correlate with disease or mortality much better. And the biggest thing too is this kind of so what factor, like what do I do with it? I mentioned earlier, I don't care if you're 44 or 36 biologically, if I don't know what to do about it. And that's what I like about the 3 tests that, that we use now that I think are really really good. I'm happy to dive into those.

Yeah. I'm curious what the 3 tests are, but before you dive into those, you hear people talking about VO2 max, about grip strength, about walking speed. I've even seen some articles about the balance or the ability to get up from the ground with as few limbs as possible efficiently. Do you know how those would compare to one of these blood screenings? Like, is this blood screening kind of like even beating all of those when it comes to predictive data?

I mean, those are great. So I haven't seen head-to-head comparisons. I do know grip strength, VO2 max, those things really correlate really tightly with healthspan and morbidity and mortality. But I think they do because they're kind of an integrator of things like your grip strength and VO2 max. Like, to have a great grip strength and VO2 max, you're doing all the things that you need to get there. You're lifting heavy things, you're doing high-intensity interval training, you're doing all the cardiovascular training. So those things are great predictors and I think they're really important. They're functional things as well. It's not something that I can follow as perfectly and get as easily as from a drop of blood test, for example.

Okay. So you said 3 tests that you guys are doing at Wild Health. What are the 3 tests?

Yeah. And, and again, it has to have really tight error bars that can't be all over the place.. And it needs to, there needs to be some resolution and I need to be able to do something about it. So Yale has one that they developed called Symphony Age. And with it, what I like about it is it actually gives you an age, not just an overall age, but organ system age. So your brain age, your heart age, your liver age, your lung age. And so that gives me some idea as a clinician as to what to work on with you, because it's really kind of your weakest link that's probably gonna get you. So I like that, that Yale test.

Wait, so, so what you're saying is instead of just getting My biological age, it'll break it down and say like, here's how old your heart is, here's how old your liver, your brain, your dick, et cetera.

That's exactly— I'm not sure if there's a dick age, but the other ones, yes.

Dickage.com coming soon.

Yeah. And the way they did it, they trained that against 130 biomarkers. So it's the data set you're training against, which is important too. So they looked at 130 biomarkers they're training it against. To get those. Just like I mentioned the Harvard age, and the Harvard one is called the Omic M age. That's a really interesting one. And we'll save PACE for last because I think it's my favorite. The Omic age from Harvard, this just came out in the last 6 months and they is a massive dataset. Again, they have 100,000 patients in their biobank, but they did proteomics, metabolomics. They looked at all the biomarkers, they looked at the EHR phenotypic data. So all of these omics they trained it against to get a really good age. What is cool about that though is they uncovered this other thing called epigenetic biomarker proxies that they talk about. So they have about 4,000 of these. So because like I mentioned in the, the neurosyphilis, you can kind of impute that and find the fingerprint, but they also kind of find the fingerprint of CRP levels. So they can tell you an imputed triglyceride level, A1C level, CRP level.

So they can give you kind of these 4,000 different biomarkers By the way, without you actually going to the lab and doing a CRP analysis or another blood analysis, you're getting all of that from the single blood spot.

And, that's why this is the future of diagnostic testing. Absolutely. And, when I first started looking at this, and if someone digs into this, what they'll find is your CRP you get is not the same as the CRP you get from LabCorp, for example. It's not going to be perfectly perfect. But, what I found is there's a really cool study where they looked at CRP graphed against cognitive impairment because the reason we care about A1C, CRP, triglycerides is because of its effect on disease and how it correlates to disease. So they looked at CRP and cognitive impairment, they graphed them, and the scatter plot was all over the place. CRP didn't correlate with cognitive impairment, which was confusing because inflammation should. Then they graphed the DNA, or methylated DNA, or the methylated CRP the, the eBP for CRP against cognitive impairment, and it was great. There was a very strong correlation. In fact, the eBP for CRP was 6.2 times more predictive of brain imaging outcomes than the blood CRP. So while these eBPs are not the exact— so if someone gets one, gets this test back and they see an imputed eBP of hemoglobin A1C that is 6, but then their LabCorp one is 6.7 or 5.3, they're gonna think, okay, this is not good enough, it's not useful. However, it may be that it's more predictive of outcomes and it's measuring biologic effect. So just like I don't care about your spot glucose, I want your A1C because it's an average over time. These EBPs may be more an average over time that correlates potentially even better. Now that's not been proven with all of them, but the CRP example makes me think that there's really something there with these EBPs.

Why is it that you think something like, uh, methylation data that shows CRP score versus a direct measurement of CRP would correlate better?

Well, because if we just get your CRP, it's frequently— we see this in patients all the time. They have an elevated CRP and we're worried they have inflammation going on, but it turns out they just did a hard workout right before. Just like if I got a spot glucose, your fasting glucose, you may have had a cup of coffee that spiked your glucose, or you may have gotten scared and the cortisol spiked your glucose.

Yeah, or a pineapple smoothie.

Exactly. But that doesn't mean you're walking around with it with a glucose that's high all the time. So it's more of an average and the biologic effect of CRP than your actual CRP.

Okay. So, it's almost a little bit similar to the hemoglobin A1c 3-month snapshot of blood glucose sometimes being a little bit more, I guess, little bit better analysis of your blood glucose than a single snapshot in time.

For sure. Yeah, A1c is much more predictive of do you have diabetes or not than just a spot glucose is for that, for that exact reason.

Okay.

And so since I can, with the Harvard Omic Image, with that report, I get all these EBPs, there's useful information. Like when I got my report back recently, I was really low in an EBP for carotenoids, and that has a significant effect they found in this Harvard study on biologic age. So I knew then there's something for me to do. So I have the OmicInAge. I don't care that much about it. I then look at the EBPs and there's information there clinically of what I can do. I ordered a carotenoid supplement, I started taking them, and then I'm gonna test my PACE score again and see if it makes a difference.

So you got Symphony, you have this Harvard OmicInAge, and those are the first two.

Those are the first two that I like because there's resolution and there's something to do about it. Okay. And then the PACE, which you already mentioned. That one is my favorite for a very specific reason. One, it's very accurate, it's repeatable. You're going to get a, get a good score if you send the same test two in the same day. You're going to come back very tight, 0.5% error there. But also because what I want to do is N-of-1 studies. So I have patients all the time that say, hey, what's the perfect diet for me? Um, and we'll look at their genetics, their DNA, their blood work. We'll talk to them, we'll try to make an an educated guess, but human physiology is complicated. So now what I do is if someone wants to try a vegan diet or a carnivore diet, or someone says, hey, I want to go get IV stem cells in Panama, well, that's $40 grand or however much it is. So if you're going to do that and you're doing it because you want to reverse aging, let's measure the outcome. So before someone does a big experiment like this, we're going to get a PACE test, see what the rate of aging is. Let them do the intervention for a few months and then repeat it. That you probably saw the Stanford twin study where they randomized these twins to vegan or omnivore diet. Really crappy study, but they used the PACE, PACE test in that, and they did find that switching your diet, they certainly could see differences in their PACE score over time to find which one is slowing down aging versus speeding it up for these individuals.

Which one won out in that study?

So the, the vegan one did, but I don't think it was good science. Like, when you look at the vegan diet, they just ate a lot less calories. So to me, it was a calorie restriction study.

No, they didn't even control for calories.

No. And there was so much they didn't control for that just made it almost useless. The only takeaway from the study with me is like 2 months of dietary change can change your rate of aging. So it led me to experiment more with people and let them try different diets and see which one actually decreases the rate of aging.

So if I send you a drop of my blood, are you using that same blood for all 3 analysis, or do you do 3 different tests on each person?

No. So, so there's 3 different tests. There's one from Harvard, one's Yale, and one's Duke. However, we use the same company that actually gives us report on all 3. Um, it's True Diagnostic is the company. And so I'll look at all 3. I'll look at the Symphony Age, see if there's a specific organ system we need need to work on. I'll look at the Omic MH to look at the EBPs, see what we can do, and then look at the PACE. And the PACE is where we do our N-of-1 studies and compare over time.

Now, couldn't I go to True Diagnostic and just order that test to my house? And if I got the data back, you're saying it would have all 3 of those tests on it?

You can, absolutely. And that's what, and that's what we use. And we use it at Wild Health in conjunction with the genomic data, with the blood work, with wearable data that we're tracking So, to us, the more data we have, the better. But absolutely, if someone did not want to see us, did not want to sign up with us and they just want to order it, yeah, True Diagnostic just sells that test to individuals as well.

And, we actually do have a discount code if you want to use it for True Diagnostics if you just want to order one of those up until Matt rolls out all the powers that he has at Wild Health. And, that's going to be BENVIP. I'll put links in the shownotes where you can get the True Diagnostics test, but Ben VIP if you go to BenGreenfieldLife.com/DrDawson, BenGreenfieldLife.com/DrDawson. I think that one of the issues here kind of like you alluded to how you're combining it with wearable data, with blood data, with other markers and measurements that you're making is how to actually in a scalable way create actionable recommendations based on this data. I mean, I would imagine it's going to take you hours and hours to sit down with all of this and actually generate recommendations for each person. And, you know, this is something I've been scratching my head about for a long time is why can't AI do this yet? Like, why can't a robot just analyze all this stuff and spit out recommendations?

The platform that we built, uh, does that. So we take in 700,000 SNPs with the genomics, the 950,000 CPG sites, the blood work wearables, phenotypic data to give recommendations. So that is, we developed an AI model to do that. We also have now built that on top of a large language model to where you can interact with it so that you can ask questions of it of about your diet. I mean, to give you an example of that is some of the EBPs, these epigenetic biomarker proxies that have been developed are around nutrition. And what is your different vitamins and minerals. So it, it all could be put into the large language model and you could actually then ask it, what is the perfect diet for me? Not the perfect diet even, but design a meal plan for me over the next 7 days. And based on their blood work of their omega-3 or the EBP of their omega-3 and everything else, they take all that into account. And the LLM, just like you could ask ChatGPT, Hey, give me a 7-day meal plan, shopping list for it, and all this stuff, and it could give it to you. This LLM can do that, but with your personal information from the epigenetics, the genetics, the blood work, and all of that. Now, I say we have that. We have not released that into the wild.

That's what I was going to ask you is, where can you get this?

Yeah. So here's what I'll do. We actually have a closed beta, and I'll get you a test flight version of it.

That'd be sick.

That's pretty cool. But how we use it now, because of regulatory reasons, AI can't practice medicine. We have it internally. So when a patient sends a message and asks a question of a health coach or a physician, the LLM gives the doctor or the health coach recommended responses. And then the health coach or doctor, we have a human in the loop. They can edit it, they can reject it, they can just kind of pass it along. So we're using it that way. The model that we'll release into the wild, we'll release this in the next couple of months, probably next few months for our patients. It'll be wellness only. So no medical advice, just about sleep, diet, exercise, those types of things. I used it recently. I got sick. I had a cold. I was so frustrating. I couldn't work out. I'd go to the gym. I remember picking up heavy weights and just putting them down for like 20 minutes. I couldn't do anything. And I finally then started asking questions like, I'm sick, I got to do something, give me a workout. And it designed a workout for me based on my wearable data because my HRV was in the toilet and all of that.

Yeah. Was that from your phone?

Yeah, it's from the phone.

This is kind of a game changer. Now, Last time we talked, you talked about your guys' really impressive executive screening programs and these really fancy concierge services that you have. Is this one of those deals where somebody's got to have deep pockets and sign up for some kind of a $50,000 annual program to have access to your algorithms and your software?

So, this is one of the things that's really bothered me about what we do. What we do is expensive to deliver right now, but I mean, I grew up with with nothing. And where I came from, people can afford that. This is the solution to that. So again, when we initially roll this out in the next few months, it'll be just for our patients initially. But my dream is to get this out to everyone to where it's a scalable version. The marginal costs are very cheap. It's not free because LLMs are expensive with all the calls, but we're talking about tens of dollars per month, not thousands of dollars per month. And this is like— I get excited. Like, I'm getting a little bit of code chills right now just thinking about it. Be able to get this type of precision medicine out to everybody. And it'll be to our patients the next several months and hopefully to anyone and everyone in the next year or so.

By the way, when you say LLM, what's that mean?

Sorry, large language model. Okay.

Okay. So have you actually tried the diet thing? Like, have you fed it through to see what kind of diet would be right for you?

I have. I, and for me, I also have kind of like what I want and what I've liked and developed over time. So it's been kind of spot on. It's been things that I enjoy, and I don't actually ask it. I should say what kind of diet. I just kind of ask it about foods in general and what I enjoy eating. So I haven't used it, like, for the 7-day meal plan or things like that, but just playing with it, I've said, hey, suggest a meal for me tonight, and it's done that. I kind of enjoy doing that myself.

Now, what if you throw it a curveball? What if it's taking all your data in, but it doesn't know that you wanna do an Ironman triathlon? Or you're really hardcore into CrossFit and you need a lot more glycolytic throughput and yet it is predicting that eating a lot of sugars and carbohydrates would potentially be deleterious for your long-term metabolic health or your biological age. Does it take into account your goals, your activity levels, or anything like that?

Yeah. So, it can't read your mind, but it actually can read your activity levels. So, we pull wearable data in as well.. So like I mentioned, the HRV when I was sick would be way down and it can read that. So you can certainly actually just tell it what you want and say, hey, design something for me, but I'm training for an Ironman and I'm doing a 3-hour workout tomorrow. It's not gonna know that. What it will see is if over time, and you can actually log the foods you're eating as well each day. So what it will see if you're logging your food is based on your food log and your activity, cuz it's pulling from your Whoop, your Apple Watch, your Oura, pulls all that data and it'll see your recovery is dropping and it'll suggest, hey, it looks like you didn't have enough calories over the last week because your HRV is going down, your resting heart is going down, you're doing 2 hours of workouts per day and looking at your food log, it's just not enough. So you have to give it that data. It'll pull, pull your wearable data, your activity, but you gotta tell it what you're eating and you gotta tell it what your goals are. But you certainly can when you say develop a 7-day plan, say I'm training for an Ironman or I'm trying to lose weight. And it'll adjust accordingly.

Yeah, that's so cool. Is it really a thing that AI can't practice medicine? Meaning, like, if I go to a chatbot and I say, hey, yeah, I've got the sniffles and a headache and an earache, act like the world's best doctor, which prescription pharmaceutical should I be on? Am I breaking the law if I do that?

No, you're not. Um, and it can practice medicine, it's just not allowed to. So when I say we can't practice medicine, I'm talking from from a regulatory standpoint.

That's what I mean. Like, like, they can't regulate that on a consumer level, but what you're saying is like a physician or medical network isn't allowed to do that.

Yeah. And so like the way we've built it, we put guardrails in. Like, we've taught it what medical advice is and we've told it don't do that. So we had to kind of dumb it down so that, so that we're not going to get in trouble when it's released.

Yeah, that's, that's really neat. So as far as actionable advice, like you mentioned calorie restriction and obviously within reason, that can be a longevity-enhancing strategy. Have you found out anything else as far as ways to slow the pace of aging or simultaneously maximize healthspan from this data that you've been gathering?

For sure. So, and I think the important part here is how individualized it is and how something like the PACE test allows you to do n-of-1 studies. So, we could— because you and I could talk forever about metformin and Synalytix and NAD and spermidine and rapamycin. We could talk about the evidence for those and how good it is or bad it is for populations, but I don't treat populations. I don't really care so much about populations. I'll use that to develop a hypothesis. So I look at the rapamycin data, I'll look at your data, and it will have like a recommendation hypothesis, but the hypothesis is not good enough. What I then will do is say, great, now let's try rapamycin. I think the benefits outweigh the risks for you, Ben Greenfield, or whoever I'm talking about. Now let's do a pace test. Let's do it for 3 months. Let's do another pace test, and then we'll decide if this is gonna be in your stack or not in your stack. So that's the way that I like to approach all of these kind of biohacks and other things. How do you feel? And then what is your rate of aging? And then the other biomarkers. And we, we, we do the other blood tests too. But to me, if we, it just like you talk about VO2 max and grip strength, those are integrators of everything else you're doing. That's what the PACE test is for me. It's a very nice one-number integrator. And even it, it's not, it's not necessarily perfect. Like if you did a bunch of calorie restrictions, probably gotta be improved. But if your goal, like you mentioned, is to get stronger and gain muscle, you're not gonna do that. So it's still a piece of information, but it's a really powerful piece of information for you to figure out exactly what works for you.

Well, speaking of calorie restriction, I don't know if you saw that new study, if you wanna call it that, about intermittent fasting and heart disease. I was curious to hear your opinion on that if you saw it.

I, I did. I think the science on it was so bad, I was honestly hoping you wouldn't bring it up because, because yeah, it's kind of taken the world by storm, like all these media outlets reporting on it. And I just, just similar to the vegan versus omnivore twin study at Stanford, I just don't think it was good, it was good science at all. And even if it was, it's a population-based one too. So it's again going to each person, how it's going to affect them is what I really care about.

Yeah, I mean, well, I think one of the main things I saw was of the population that had the high risk of heart disease in response to intermittent fasting also had a high rate of smoking. They were low income with poor access to high-quality food, and they had high BMIs. And I, and I don't know if it was a food intake questionnaire study, but that can also throw a wrench in things because I don't know about you, Matt, but if you ask me what I ate like last Sunday, I have no clue.

No, anything based on food questionnaires, you just kind of throw away immediately. Immediately. Nobody's good at that. That's been proven over and over.

What about the one that a lot of people are talking about right now regarding niacin? Did you see that, like the link between niacin and heart disease?

No, I didn't.

Okay. Yeah, it was basically showing that there's this inflammatory marker, I believe it's called 4-PY, and when elevated, it can put you at risk for heart disease. And they tested people's levels of 4-PY and showed that to be the case. But then, even though I don't think they actually gave people niacin or NR or NMN, they then went on to hypothesize that high long-term intake of these supplements could contribute to heart disease by elevating this 4-PY. And, it was pretty interesting. I mean, the big takeaway that I got, especially reading Chris Masterjohn's review of it, is that you need to be cautious with high-dose niacin or high-dose NR or NMN intake for long periods of time. In this case, for niacin, it would be like above 250 milligrams daily for months on end. And also that if you wanted to keep NAD levels elevated, which is why a lot of people will use NR or NMN, it might be safer to just use NAD because that does not result in the same increase in 4-PY.

That's interesting. Yeah. And I'll dig into that. It's a really interesting study.

Yeah. Yeah. So what else have you changed about your own lifestyle or diet or exercise program or your approach to patients in general based on what you've learned from this data?

Yeah. So the approach— I'll talk about myself specifically. I think the approach to individuals is pretty straightforward. The niacin paper, like all of these studies are data points and they again can help us with a hypothesis of what we're going to try for someone. But I'm just very committed to having the objective markers of testing before and after. So for myself, for example, um, I test actually once a month on the pace, and that's probably a little too much. Um, but I, I like to do a lot of experiments. Um, and this is very interesting. It, it, this confirmed some beliefs I have, but when I think about how to maximally extend healthspan, I think as much as we could talk about all those biohacks, the number one thing is very unsexy and it's just connection. And, what I mean by that is I think so much of our dis-ease in the developed world has to do with disconnection. And, I'm talking about disconnection from ourselves, our bodies, our food, nature, others, relationships, and something bigger than ourselves. So, for example, my last— a couple PACEs ago, I had a score of 0.88. So, I'm was aging 12% slower than baseline. So great, that's fine. But I was going through a really stressful time in my life and I decided to make a real focus on those things like connection to myself, my relationships, spending more time meditating or praying and really focus on connection. And my next PACE after that without any biohacks or special stuff was 0.76. So that's a 12, that's a really incredible absolute risk reduction, absolute reduction of 12%. Wow. So what that means is If I have 40 more years to live, I'm going to add 5 years to my healthspan. And if you, if you were to cure cancer, so I talked about GRAIL study or the test earlier, they're not trying to cure it, they're trying to diagnose it earlier. If you were to cure cancer, you only add 3 years to your lifespan. And that's not even healthspan because you're still getting old and frail. So just for me, the number one thing I found, and I've believed this for a while, so maybe it was a little bit of a self-fulfilling prophecy, prophecy is that connection. And when I say connection to our bodies, it's like listening to our bodies and doing the exercise, like doing some hard stuff, but also listening when I'm doing too much. From the food, like having a connection to the food, your food and where it comes from, really focusing on your relationships, like backing off of work and kind of this achieving that stuff. And again, the connection to a higher power than you, that stuff. I mean, you, we've known, like there's a great Harvard study in healthspan and relationships were the one thing that made the biggest difference. So, we kind of have this evidence, but to see it for myself was really great. And I love talking about that too because as bad for businesses it is, you don't need a doctor for that. You don't need Wild Health. You don't need anyone else, just focusing on that connection.

Ben Greenfield: That is pretty cool that you can use yourself as an n 1 like that and adapt on the fly. But, it kind of begs the question, about how long it would take to get the results back? Because let's say you make a lifestyle change that could have risk, like you start, I don't know, running or lifting weights too much, or you introduce the wrong kind of oils, or you change up to a diet that actually doesn't really pair well with your genetics. How soon can you actually see that data come back? Like, how long do you have to wait once you send in the blood test?

Yeah, so the twin study, that was an 8-week study, and that showed some changes. In general, most of the studies where we look at PACE, It's a few months. It's 2 to 3 months. So my doing it every once a month is probably a little too much to really see a significant change. But it also comes into play, like, when you, when you do make these changes, again, we're doing them in an educated way. Like, if someone says, hey, I wanna see what smoking does to my pace, we're not gonna do that, that test. Uh, so we wanna be smart about the test we do. There's enough things that show evidence for increasing healthspan. I mean, we probably, we, you and I could probably come up with a list of 100 things So there's enough things with good evidence that we kind of start at the top of the list, the most likely thing to work and go, and go from there.

Yeah, that makes sense. But how, what I mean is how long does it actually take to get your results back once you send the test in?

Yeah, I think when we send them, it's usually 2 to 3 weeks. Okay.

So less than a month basically.

Yeah. So if you do intervention for a couple months, then send the test, you have it back. So it'll be 3 months of doing it, but you certainly could do it for 2 months and the test off and then stop and then start it again or not, depending depending on what your results were.

Are you still using more advanced diagnostics with your screenings at Wild Health, meaning full-body MRI, CT angiography, things like that?

Yeah, absolutely. I mean, we are, and it just depends on the risk factors for the person, um, their lab tests. I mean, if someone has really elevated inflammatory markers and really elevated triglycerides and Lp and ApoB, and we're worried about their heart, then we'll clearly scan. If they have a cancer history, we may do the whole body MRI sooner or a GRAIL test. So, this biological age test that we're talking about is a hot topic. It's not a replacement for everything, it's another piece of data, but it's just a really powerful piece of data. We want as much data as we can get to make these decisions.

Ben Greenfield: And of all these newer diagnostic tests besides biological age, which is obviously super cool, are there any other tests or diagnostic tools that you have your eye on or that you've seen up and coming in some of the research you've been doing?

So I think you named them, and we've named most of them. So just to go through the list quickly, like the genomics, it's great because it gives us all your risk factors, 20% of your health outcomes. So we got to have that, your operating system, all the blood biomarkers that we're all familiar with. I mean, your metabolic health, obviously, and all these things are pretty basic, and all of your listeners understand. Then when we start thinking about what's going to be most likely the thing that kills you, it's going to be cardiovascular disease, cancer, dementia. So for cardiovascular disease, we're getting a Cleerly scan. Usually on most people after they're, after they're 40. And if it's perfect and it's zero, we'll follow them for a long time more before we do another one. But if they have some plaque, we may get another one in 2 years to see their trajectory. So we know how hard to push on cardiovascular risk. The whole body MRI and the GRAIL, the gallery test, I feel really good once I've gotten all the lab tests on someone, a gallery test, a whole body MRI, and a Cleerly scan. I feel really good about them not developing one of these things that we all hear these horrific stories about someone in their 40s dying from.

I don't want to put you on the spot and you don't have to answer this question if you don't want to, but in running a lot of these tests on yourself, is there something that you're really concerned about for yourself that's probably the number one biggest pain point for you or from a genetic standpoint the thing that you want to address the most?

Yeah. I mean, for me, my grandfather died at 48 of cancer. One of my grandfathers, the others died in his 50s as well. And so that's always on my list. So because of that, and I do have some, um, I know that I have increased risk of that, so I will probably a little more often do a Gallaudet test and a whole body MRI. So even though we're talking about all these objective markers, family history is critically important as well. I don't have anything on my labs per se, and on my genetics I don't have an APOE4, for example, even My grandmother had Alzheimer's disease and my mother has an APOE4. I don't have that. So, for me, it's about the biggest risks right now or what I've seen in my family members because all my labs look great and they should. I'm 42 right now.

Would you say that's the biggest advantage of a full-body MRI is early cancer detection?

Yes. However, we pick up other things. Most of the time, it's nothing, but you certainly get other information just about your back, or, uh, which really usually doesn't tell us anything. But you'll find AVMs, arteriovenous malformations, um, in the brain or other things which need to monitor more frequently. So the biggest, biggest one is finding early tumors. I mean, I, I also had a, an aunt who, um, died in her 50s, uh, from, from cancer, from breast cancer, which has a 90% survival rate if caught in stage 1. But in stage 4, it's almost reversed. Like, it's, it's truly hard to survive that. So that early detection is really important. And, and that's what we try to do is to, um, very early detect these. And that brings up an important point actually about, um, methylation risk scores, which is an extension of those EBPs. So when I mentioned that the EBPs, you can see triglycerides, these other, these 4,000 different epigenetic biomarker proxies, you can extrapolate out and realize, just like with the example I gave you of neurosyphilis, why not find the pattern for coronary artery disease for, um, early cancers, which GRAIL has. So Harvard is doing a study right now, which I'm really excited about, where they're looking at the top 8 diagnoses in the world and trying to find those patterns. Um, I've seen some preliminary data on it and it's really incredible. Their methylation risk score for diabetes is actually pretty similar, the AUC for it. To hemoglobin A1C, kind of around 0.84. So I think in the future, this is why I said DNA methylation is the future of testing. I think it was, I can't remember who it was that said software is eating the world like 10 years ago. I think of DNA methylation testing, it's going to eat the world of diagnostics because they'll find this fingerprint for coronary artery disease. So right now, coronary artery disease, you can do Framingham scores, which are really horrible sensitivity specificity. But to make the diagnosis, you do a cath or a stress test. With these methylation risk scores, a drop of blood may have better sensitivity and specificity for these really invasive tests. So I think in the next 2 to 3, maybe 4 years, this drop of blood test is going to potentially replace so much of diagnostic testing, such a large percentage of it.

Do you see the old Disney cartoon Aladdin?

Yeah, yeah. I'm so curious where that's going.

It's like phenomenal cosmic power, teeny tiny living space. It's like phenomenal predictive power, teeny tiny drop of blood. That could be your tagline, man.

Yeah.

We'll take it. Yeah. Well, this is just super interesting. What I'm going to do is put all of this in the shownotes. If you guys really like this idea of precision medicine and you want to understand it better, I would definitely go listen to the other podcasts I've done with Matt because there's 3 or 4 of them. And, the shownotes are going to be at BenGreenfieldLife.com/DrDawson, like Dr. Dawson. And I will also link to Wild Health so you can keep your finger on the pulse of what they're doing over there and also stay up to date on the rate at which they're rolling out this predictive modeling. And Matt, I'll link to your executive health screenings and everything else that you do, but you still have kind of a network of physicians mostly nationally. Is it internationally or nationally?

It's nationally, yes. The practice of medicine is pretty highly regulated. We're in all 50 states. So, yeah, anyone in the, in the US. Once we do have the app, the goal will be to make that international because it'll be a wellness thing. But for right now, we're in all 50 states. We do have a code, Ben20, for 20% off anybody that signs up. But again, I think that I always like to really stress the connection part anytime I'm talking to anyone because you can get, I think, 80 to 90% of the way there just focusing on those natural things and the connection to others and yourself. And, I just think it's so important to stress that instead of trying to sell someone on something very expensive. But, if someone does want an incredible amount of data, they want the AI precision recommendation, yeah, we're in all 50 states and taking patients. Ben Greenfield: Yeah.

And, to wrap your head more around the connection piece, I actually just finished a book a couple months ago called "The Cancer Connection" directly highlighting what you just talked about, the link between lack of relationships, loneliness, and onset of progression of cancer. So, it's It's absolutely important. It's not as sexy to think about, but it certainly is fun to gather your friends and go play some pickleball, right, Matt? Absolutely. Yeah. Alright, folks. Well, I'm Ben Greenfield along with Matt Dawson signing out from BenGreenfieldLife.com. Again, shownotes are BenGreenfieldLife.com/DrDawson. Thanks for tuning in.

Hey, Ben Greenfield here. Quick addendum to today's show. As you can see, it's early morning, day after that podcast. With Dr. Matt Dawson. Uh, he let me know after the show that True Diagnostic is going to give everybody who watches this show access to the same high-level, cutting-edge methylation data test that Dr. Dawson talks about in the show. So if you go to the True Diagnostic website, I think it's truediagnostics.com, but I'll put a link in the show notes if you go to bengreenfieldlife.com/dr-dawson, like Dr. Dawson. That's where all the show notes are for this show, uh, they gave us a code. They gave us a 10% off code. That's Ben VIP. Ben VIP is the code. VIP is all caps. That code is going to stay good for several months after the podcast, and they're going to let you order multiple tests at once if you want to. Dr. Dawson's doing his every month. I'm going to start tracking mine more regularly after everything that I learned in this show. So if you want to a True Diagnostic test, just use code BENVIP on the True Diagnostic website and that'll let you get access to the same kind of testing. So, there you have it. Have fun with all your experimentation.

Do you want free access to comprehensive shownotes, my weekly roundup newsletter, cutting-edge research and articles, top recommendations from me for everything that you need to hack your life, and a whole lot more? More, check out BenGreenfieldLife.com. It's all there, BenGreenfieldLife.com. See you over there. Most of you who listen don't subscribe, like, or rate this show. If you're one of those people who do, then huge thank you. But here's why it's important to subscribe, like, and/or rate this show. If you do that, that means we get more eyeballs, we get higher rankings, and the The bigger the Ben Greenfield Life show gets, the bigger and better the guests get, and the better the content I'm able to deliver to you. So, hit subscribe, leave a ranking, leave a review if you got a little extra time. It means way more than you might think. Thank you so much.