Latest / Two PhDs on a Pod / Why "Healthy" Foods Spike Your Sugar
Transcript
- speaker-0: you speaker-1: Welcome to another episode of our podcast and Guru, I have a joke to start with. So I made this up yesterday while walking nearby. So here's the joke. Two apple cores are chatting inside a compost bin. And one asks the other, how did you do today? And one says, oh, I did very well. I pumped the blood glucose up to 175 milligrams per deciliter. speaker-0: Okay. speaker-1: for the guy. What about you? And the other says, ⁓ not so great. I only got him up to 120. His microbiome shut me down. speaker-0: Alright, alright, I got it. speaker-1: You got it. I wonder what percentage of our audience will get it, but we are going to explain this joke throughout this episode and people will get it at the end for those who didn't get it upfront. speaker-0: I agree with you. think you need to watch the episode to understand what the joke really means. that joke actually brings us to the core. The core. The core thesis of this show, which is that traditional medicine looks at a one size fits all kind of an approach. They look at a food and they say this food has a glycemic index and it's going to have the same effect on everybody. That's traditional medicine. But that's really not how it is. One site does not fit all. Every food has a different sugar response on different individuals. It depends on your biology. It depends on your microbiome. And we're to talk about how does your microbiome and your full individual biology impact your blood sugar response today. A banana might spike my blood sugar, but keep yours perfectly calm. And a piece of bread might spike your blood sugar, but might keep mine perfectly flat. It's not about the food. It's about your individual biology. speaker-1: So keep listening to learn more. We're going to dive deep into this topic. I'm Momo, a biochemist. speaker-0: And I'm Guru, an AI and data expert. We're two PhDs on a pod. speaker-1: All right, well, after that, now we're gonna go into what's on your radar, Guru. speaker-0: All right, well, you know what? I have something on my radar that I think you will enjoy for sure. And I hope a lot of the people in our audience enjoys as well. I made a really delicious South Indian recipe for a dish called Uttapam, which is a very famous dish from where I come from. And my mom, who is probably going to be listening, used to make this for us and still does whenever we ask her to. And everybody that I have in my family used to make it. And the thing is, everybody made it differently. People had their own little tricks and ingredients and everything. It is a platform, exactly. It's like a pizza. speaker-1: It's a platform, right? It's like a pizza. You can have as many different kinds as you want. speaker-0: Exactly. actually it's not too dissimilar from the idea of a pancake in many ways, but there is some very critical difference. I'm going to talk about it just a minute. So I ended up using as the base, ragi, which is finger millet. Ragi is very common in India. It used to be the staple food of my grandparents, but it's reduced now over time. My parents did not use it a whole lot. speaker-1: For those who are not familiar, please tell us what Raghi is. speaker-0: Raghi is finger millet. speaker-1: Right, what does it look like? Is it a green? it a- speaker-0: It's a grain and it's a tiny little black particle and you powder it and you can use it for many things. mean, we used to use ragi for just making a steamed ball, or you can make it a porridge and you can eat it for breakfast, or you can make, or you can make rotis with it. You can make anything with it. So what I decided to, see here on my screen, this little bowl has about a cup of ragi and I have added to it. a bunch of other veggies. I simple onions diced. have carrots, which are grated. And I have some coriander, which is one of my favorite little additions. I would call it chopped. And I like to put yogurt. I just take plain yogurt. You can take Greek yogurt. You can take other types of yogurt. I put some yogurt in there as well. And then I put everything together. I put some water and I mix it up. and it looks like this in the bowl. then I just pour, the consistency should be a little bit like a pancake consistency, I would say. And you can just pour that onto your pan. You see on the right hand side over here, that's what it looks like when it's cooking. And it just takes about, I would say two minutes, three minutes to finish one of these things. And once you put it on the plate, It looks like that, but because it's ragi, it's a little bit dark brown in color. And you can see the carrots and the onions and the coriander showing up over there. It's very, very tasty. Now you can have it with many accompaniments. I made something that I thought would be super easy for American households to prepare, which is a peanut sauce. I took some peanut and I added a bunch of other very interesting dimensions to it, like soy sauce, some honey. some vinegar and some lemon here and some some chili crisps. So when you put all the stuff together and mash it up, it's a great paste. You can actually use it with many dishes, but I like eating that with my uttapam. So I just made that over here on the side. You can see and I enjoyed it a few days ago and I want to do it again. speaker-1: Love it. I want to try this version for sure. I've never tried. I've never had this version. speaker-0: In bharagi or you mean with the peanut sauce? Try it, try it. speaker-1: peanut sauce. also your combination, it's not just ragi, it's all the other vegetables you added to it. It makes it very unique. speaker-0: Exactly. And I know Momo that you make Utapam too, so how do you make it? speaker-1: Yeah, I make uttapams pretty regularly. I fell in love with these things and it's again, a platform that you can play with. So I've experimented making many, many different kinds of uttapams. And the one that I've settled on that I absolutely love is this one. This is my favorite one, both in terms of the base composition and also in terms of what's embedded in the batter and also what I put on it. So I'll just go over that fairly quickly here. So... We actually grind our own flowers. just use Vitamix, which does a fantastic job. So in this particular recipe, I will grind one third ⁓ lentils, pick any lentils you want. doesn't really matter. Lentils, rice. And in this case, you need to use white rice because you need it to bind things together. You need that starch to bind things together. You can try to get away with brown rice, but it's, it's not really going to work so well. And then one third, we use wheat, whole wheat, just all grains. And so we put that in the Vitamix, blend it well, ⁓ add water to it and some salt. And then my favorite type of uttapam is with purple onions and cilantro. And so you just make the batter and then you make uttapams just like Guru described, very simple. And then what I really like is this veggie masala and we add tofu. So the veggie masala is the traditional potato pea masala, veggie masala that is made billions of times in India, but to increase the protein content. We also add in the super firm tofu from Trader Joe's. And so this is easily 30 grams of protein per serving. And so, yeah, we make that veggie masala and then you can make it into a burrito if you'd like, or you can eat it on the side. ⁓ I mean, it's just a phenomenal dish that's super easy to make. Two people, 30 minutes, you're done. It's served. Love it. speaker-0: I know, I know Momo and you know, there are so many variations. Like you said, it's a platform. So I use millet, you use lentils and rice and wheat. I sometimes put all kinds of different flowers in it. To be honest with you, I use millet because it's actually an enjoy food for me. Well, right. Sure. It doesn't it doesn't spike my blood sugar. Absolutely. I could have used something else. Right. And you should probably use something else. If millet is spiking your blood sugar, you should probably use something else. And that is the actual topic of this show. How do you make choices when you know that your glycemic response reacts one way to certain foods, different way to certain other foods, and then how do you make the best healthy food for yourself? So Momo, we all know that metabolic disease is such a big epidemic around the world. And tell me what is your perspective on metabolic disease and the way it's evolving and everything. speaker-1: I mean, we're going to have a, ⁓ like a top top endocrinologist as a guest on this episode. So we'll dive deeper, but I just want to mention at this point in time that metabolic disease, is one of the characteristics is elevated blood sugar is a risk factor for essentially all chronic diseases, but let's just highlight the big ones. So cardiovascular disease, which is the number one killer Alzheimer's disease, which is becoming the number one killer in many populations. And then fatty liver disease, kidney disease, eye disease, neuropathies, and very importantly, several cancers. And so what we're going to discuss today is how do we leverage each person's biology to make informed food choices where we still have essentially unlimited number of options to cook, yet in an informed way, we can actually reduce that glycemic spike, which we will see how dangerous it is, not just for these chronic diseases. ⁓ but in general. So very excited about this episode. speaker-0: Yeah. So the thing that always surprises me when I see is that the same food can have such different impacts on different individuals. And here's a chart that we saw in our own data. Here are two people who ate two different foods, bananas and bread. The person on the left had a sugar spike for the banana, but they had no real spike for the bread, the wheat bread. The person on the right had a spike for the bread, but had no spike for the banana. So it's like two individuals that have exactly the opposite reaction to specific foods because of their biology. So the question is, why is this happening? And how do we understand it? And how do we do something about it? And that's why we're going to have our two guests who we're going to invite on the show right now. speaker-1: To further discuss this very exciting and very important topic, we have invited two very, very special guests today. ⁓ Dr. Eric Patridge and Dr. Damon Tanton. are different types of doctors. Dr. Patridge is a PhD in just about any molecular or chemical science you can imagine, and also codes and does all kinds of things. And he's the head of our systems biology, computational systems biology team. And then Dr. Damon Tanton. speaker-2: We have. speaker-1: He's a big chief at Advent Health in Orlando, Florida. And he is an endocrinologist and a fantastic human being. So welcome, Eric and Damon. speaker-3: Thanks. speaker-1: All right. So I see you're very excited to be here. All right. So let's start with Damon. So please tell us about this metabolic disease and global epidemic and what are your observations from traveling the world and talking to- speaker-2: You speaker-0: I'm going to in here because since you're talking traveling the world, I just wanted to mention here that Damon has just come back from a very special experience that is supposed to be a darkness retreat, which I'm, actually this was the first time I heard about it when Damon just came back. So Damon, you look like you're still not completely out of it because there's a dark background behind you, which is, which is awesome. But can you just tell us where were you? What were you doing there? I think it's going to be very interesting to our audience. speaker-3: Yeah, so my wife and I went to Poland and there is a retreat there that is specifically focused on darkness. So we spent four days in separate cabins, four days without an ounce of light. So completely dark. The retreat is actually called within. That's the title of the retreat. And the whole idea is if you remove every distraction from without or external. then you can focus on within. ⁓ And you know, the theory is you have everything that you need within yourself. You don't need to have external teaching. So I told the group, I've probably said the word insight 500 times in my life, but I never quite realized that the word means sight within, right? And so true insights come from spending time alone in thought. speaker-1: Cool. Did you discover some new things that you didn't think or know about before? speaker-3: Well, my intention before starting this retreat was to decide how I wanted to live the second half of my life. Right. So after I turned 50, I mean, I can't, I'm not guaranteed to live a hundred, but let's say I do. ⁓ When I turned 50, that's kind of the second half of life. And I wanted to figure out how to go from success to significance. I, you know, it was more about. not how to make more or acquire more. It was really about how to take care and appreciate. speaker-0: That is so cool. That is so cool. we to come back to that point maybe later in the episode as well. But I think it's a great thing for you to go back to Momo's main question about the metabolic disease being a global epidemic. yeah, how would you explain that? speaker-3: Well, you know, I'm biased, but I believe that metabolic disease and metabolic health is the most important disease process that is impacting the majority of people on earth. And so we probably first should define what metabolic health or metabolic disease is. Usually I equate metabolic health with how the body handles macronutrients. and how it manages nutrition with insulin and glucose. So insulin resistance to me is almost a synonym for metabolic disease. And if you think about all the diseases that metabolic health encompasses, they're all related to insulin resistance and inflammation. so ⁓ fatty liver, otherwise known as mash, Cardio renal disease, inflammatory disease, these are all, they all have the undercurrent of insulin resistance. And then you say, well, what is insulin resistance? speaker-1: Let's please dive deeper into that and explain it very thoroughly. speaker-3: Yes. So everybody, I think, knows about diabetes, right? But the problem is that diabetes is an arbitrary diagnosis. And when someone has glucose or sugar that gets out of control to a certain level, we call it diabetes, right? But that is the manifestation of 10 to 15 years of worsening metabolic disease. So if you look at ⁓ insulin levels instead of sugar levels, ⁓ insulin levels rise about 15 years before when your body starts developing resistance. so the pancreas, because it's a team player, it says, I can make more insulin. It's okay if you're resistant, I'll just make more insulin. But the problem is insulin is the most obesogenic and inflammatory. hormone that we have. And since most of disease, chronic disease is inflammatory in nature, you can see how insulin plays a major role. speaker-1: Yeah, let me just mention something that I saw in a paper that states this obesogenic factor that you just mentioned. And the way I interpret that is that if you look at the fat storage, so after you consume basically fats in your food, the amount of fat stored is proportional linearly to the amount of insulin released. And so it's literally like a signal to your body to store fat. And that's obesogenic, right? speaker-3: Absolutely, and it's not just, it goes farther than that. ⁓ It's not just that you're storing fat, it's the place that you're storing it. So insulin preferentially causes you to store fat in the viscera or the organs, right? And the one that's most famous or most consequential is the liver. So everybody knows about ⁓ fatty liver, ⁓ but you might not realize talking about the scale and scope of this, that a hundred million people in the United States have some level of fatty liver. A hundred million people in the United States have some level of insulin resistance because they're the same people. Insulin resistance and fatty liver are the same disease, right? And so ⁓ there's arguments ⁓ as to what comes first. Do you develop fat in the liver that leads to insulin resistance or do you develop hyperinsulinemia that leads to fat in the liver, it's probably a cycle. But I used to say that fatty liver was the first domino to fall in the cascade. But now I say that it's not even the liver that's the first domino, it's the gut. ⁓ yeah, yeah. There's good evidence that insulin resistance starts at the gut lining. speaker-2: Woo. speaker-1: We love to hear that. speaker-3: way before it involves the peripheral tissues or the pancreas. speaker-1: Okay. Let's, let's, ⁓ want to mention a couple of things here that I'm excited to share. One is that, insulin resistance basically means that our cells have insulin receptors displayed on the surface and these insulin receptors enable, enable the cells to store sugar from the blood. And when these receptors malfunction, then the cells like muscle cells and liver cells that are storing glucose from blood are, are less capable of doing that. And therefore the blood sugar spikes. Now. I also want to mention something that most people don't know, which is that one of the potential causative factors for insulin resistance is that gingipanes, which are proteins or enzymes released by an oral bacterium called porphyromonous gingivalis, they have been shown to be able to cleave or cut in half these insulin receptors. So one of the possible ways, and maybe the main way we don't know yet, for insulin resistance is that actually ginger pains are released from porphyromone's gingivalis in saliva. They translocate into the bloodstream via leaky gums, and then they literally cut off the receptors. And now those muscle cells and liver cells cannot sense insulin because they no longer have the receptor. Did you know that, Damon? speaker-3: I did not, you taught me something today, Momo. speaker-1: This is pretty cool, right? That oral microbiome speaker-0: And we're going to actually talk about both the microbiomes and how that's going to be used in delivering a service to people that can actually help them decide what to eat, what not to eat. So we're going to get to that. But I want to step back for a minute before we go super deep. I want to talk about the global impact. I think you said more than 100 million people in the US. But if you look at the world, Damon, I have a chart here that basically shows that It is increasing dramatically all around the world. And it's not just a developed country phenomenon. It is a developing country phenomenon. You can see all the red spots here. There's a lot of developing countries and it turns out there is a lot more younger people who are getting more and more diabetic or pre-diabetic. You see the incidence curve that is to the left of the prevalence curve on the right-hand side. That's basically telling us that the incidence is increasing as you get younger and younger. speaker-1: And incidence means new cases and the prevalence means... speaker-0: And here's another interesting experience I had when I was in India in October of last year. I was in this conference called Treat DM, which is an annual conference. And the reason they do this annual conference is because diabetes has become the major disease epidemic, I should say, in India. And you can see on the left-hand side, you can see the population that is the highest. in diabetes are all the more slightly more developed states like in the southern region, which actually where I come from, the GDP is higher. There's more people who moved from the lower middle class to the upper middle class and a lot of people who moved into the lower middle class. And so now they're able to afford all of these processed foods. Now they're able to suddenly go out and get a lot of things that they were not previously able to get. And so they have probably not been very discriminating about what they're eating and they have not been taught what to eat and what not to eat. Back in the village, they were eating really great, healthy, whole grain foods. And now they're in the city with processed foods. And you can see in the urban versus the rural divide in the urban areas, it's literally the whole country of India is all red. And in the rural parks, it is a lot more green. People are still not reached that part, but I learned that there's a lot of interventions that the Indian community is doing in those regions. I'm just using this as an example of one of the most severe changes in the demographics and the disease population that's happened in the world, which is in India, where people moved in the last 20 to 30 years, they moved into middle-class, they moved into more urban areas, amount of metabolic disease has basically gone through the roof. I was just telling Momo this, the number of people with metabolic disease in India is more than the population of the US. It's probably like half a billion people. And the number of people with diabetes is about 150 million people. the population with pre-diabetes and all the other diseases that you talked about put together is probably a few hundred million more. So that's a half a billion people in India. speaker-1: crazy. speaker-3: And I think this is only the tip of the iceberg group because these are graphics based on diabetes, which does not capture the full scope of metabolic disease. Right. Yep. And if insulin resistance predates diabetes by 10 to 15 years, then you can go ahead and multiply the people with diabetes by four or five times. Right. speaker-0: Yes. speaker-3: And I do think it's important to recognize that in different parts of the world, especially like India, Pakistan, China, the distribution of fat is completely different than in the West. And so what happens is for some reason, people from that part of the world for that heritage typically store their fat in the organs first. So it's phenotypically from the outside, invisible. And so what happens is they present with out of control diabetes type two, and you look at them and you say, you don't have a big belly. You don't look that off, right? ⁓ And it's because the fat has been stored in the liver first. Yeah, whereas in the US we have to get a big fat belly and then we see it in the mirror and like, ⁓ we probably should do something about that. And then it slowly becomes an organ thing. So I think that's really important. When I have patients from Pakistan or from India or from China, I'm always super suspicious for insulin resistance. And I will check an insulin level rather than just a glucose level. speaker-1: Love that. Love that. And that really brings us to the next topic, which is, Damon, we would really now like you to tease apart this 10 to 15 year process that you mentioned. How does a person go from completely normal, healthy glucose metabolism to type 2 diabetes? Walk us through that and please definitely mention when they should start testing, what they should start testing for, and what are the warning signs. speaker-3: Well, I do think, and this is pertinent to this podcast and biome, that it really does begin with the gut from the beginning. And at least I can speak to the Western diet. I don't know much about some of the other diets, but the Western diet is low fiber, it's high saturated fats, it's refined carbohydrates, it's low plant diversity, right? And that obviously starts the process of dysbiosis, right? And you guys know this better than I, as far as the microbiome, but when you lose diversity from a microbiome standpoint, you start to have breakdown and barriers in the gut. And then that leads to an inflammatory response because we call it a metabolic endotoxemia. So basically. there are substances that get through that shouldn't, right? And starts to set up an inflammatory cascade, right? And that stresses the whole body out. And that's the point when you start developing insulin resistance, fatty liver, atherosclerosis, chronic kidney disease, and that all still predates diabetes. And so the person has gone through these kind of stages over 10 to 15 years. and they still don't have a diagnosis of diabetes. So the whole diagnosis of diabetes is a problem. We are basically giving patients a false sense of security that as long as they're not labeled as diabetic, that they're metabolically okay. That couldn't be any farther from the truth. speaker-1: Love that. Love that highlight. That's really one of the major takeaways is just because you don't have diabetes, it doesn't mean you shouldn't really pay attention to your body. speaker-3: And if you want to get as upstream as possible, yeah, because the farther in my career I go, I'm looking for upstream interventions, Not treating the trunk or the roots of the tree rather than all the branches, right? And so if you want to get as upstream as possible, and I'm not doing that just because it's this podcast, speaker-1: because our audience does. speaker-3: you really need to focus on the gut first because that's the entryway and the absolute first domino to fall in this 10 to 15 years. speaker-0: Thank you, Damon. That's a great way to sort of move into one of the major studies that we did together. speaker-1: Guru, let's just highlight please the rise and fall of fasting insulin, fasting glucose, and the tests that people should order and should insist to order, even though their doctor may not order it until they have symptoms. So yeah, Damon, please talk to us about fasting insulin, fasting glucose, HbA1c, the succession of those, and what should people insist on, because I've just had to do this recently for a family member. speaker-3: Yeah, it's a problem. It's ⁓ a lot of my now a lot of my colleagues have gotten better over time. I've been preaching this for 15 years and at the beginning they're like didn't want to hear it and now it's cool to draw insulin levels. So just like it's cool to be on a GLP-1 right even though they're not new and the first one I wrote for was in 2007 right but now it's a new thing. If you look at the progression and you want something that you can measure that's in the blood, I'm not talking about the Viome test, I'm talking about blood test. The first thing that you can realistically measure that's going to give you the earliest sign is going to be your fasting insulin. Now, that's not the first thing to go up. The post-prandial or after meal insulin is the first thing to go up. But that's a very difficult one to analyze because you have to then normalize it by what you ate. So that's not usually clinically something that we do. So we rely on the fasting insulin as a very early measurement of metabolic disease because that predates the glucose, right? And because remember the whole reason that the insulin is high is to compensate to bring the sugar down. And the whole reason that the diagnosis of diabetes is delayed is because you're hyper producing insulin. it just so happens that the hyper production of insulin is actually leading to disease in and of itself. So ⁓ fasting insulin is probably the best, in my opinion, the best one to rely on if you want to be early and speaker-1: And I want to mention, Damon, that as far as I know, all insurance plans will cover that test at least once a year. But even if they don't cover it, it's very inexpensive. It's like on the order of $10 to $20, I think the challenge is that general practitioners don't order it. That's the challenge. And so you have to basically tell your practitioner you want it. speaker-3: It is because we have been overwed to A1C for 35, 40 years, right? Right. And so I can't stand when I hear studies based on A1C that that's the only thing to look at because you're looking at a very late marker. So to get back to your question, Momo, the A1C is the last thing to go. Right. Because the A1C is an average and you're putting high sugars, low sugars into it. it only rises when the high sugar periods outweigh the low sugar periods, right? And remember, in early metabolic disease, you're dominated by insulin, so your sugar is actually on the low side. So the first thing we see is a high insulin, low A1C. So when I see a patient that has an A1C of like 4.7, that's a huge red flag. Yeah, because as long as I mean, unless that person is a professional soccer player or someone that has no visceral fat whatsoever, that would make sense. But if you're looking at the person, the patient has maybe 20, 30 pounds of excess belly fat and they have an A1C of 4.7, that screams that their insulin is dominating their sugar and that's why the average is speaker-0: That is speaker-1: Okay. So insulin rises first. ⁓ And so people should obviously pay attention to that. What is the next one? HbA1c or do we want to discuss fasting glucose? speaker-3: First thing to rise is the after meal insulin. The second thing to rise is the after meal sugar. The third thing is the fasting insulin and the fourth is the fasting sugar. And then finally, ⁓ bringing up the rear is the A1. speaker-1: Yeah. speaker-0: Actually, I think the way you just explained it is, the first thing to break is your gut barrier, right? And then there is the inflammation that comes from that. And then there is the insulin resistance and then so on and so forth, right? And then it all goes all the way down to the HVMC. speaker-3: Absolutely. Right. So if you were looking at this in stages, the gut is like stage one, insulin resistance is about stage five, right? And then diabetes is like stage 72. In other words, we're waiting way too long to address metabolic disease and it's giving patients a false sense of security that everything's okay, right? And so the last thing I'll say to this is we can get fancy with tests as far as you know, blood tests and so on. But the easiest thing to do, and I tell my patients this during our visits, is just to look down. Look at your belly. If your belly is not flat, then there is a super high chance that you have insulin resistance. It's that simple. Now you can go and get further analysis. You can get biome testing. You can get blood testing. You can do all that stuff. But the first thing that you should realize is that if your belly is not flat, you have metabolic disease until proven otherwise. speaker-0: Yeah, so what I was going to say was, know, this whole progression that you're talking about, Damon, is very, very educational. So if you want to get to the root cause, you want to understand what's going on in the gut. And then you want to see all the step-by-step downstream effects. as you know, in VIO, we are looking at all of that progression, all the steps in the elements of the progression, starting from the gut. You call it diversity. I'll call it function, the functional diversity, which has all the anti-inflammatory aspects to it. has all the reduction of oxidative stress, for example, the reduction of everything that creates, that breaks the gut barrier. And then we look at what is the inflammatory properties that are in the gut, also in the blood. then we, and then the saliva, and the saliva, Momo brought that up. That's also important to, you know, it's a contributor to the, to the whole system here. And then we ask ourselves, what can that help people do in their real, in their actual life? Every time you make a decision about what to eat and what not to eat, can we help you make that decision a little bit better? Right? So in order to go there, we designed the study. Absolutely. And I would like to get into the details of how we designed that study. So Eric was part of the team that designed that study. we'll switch to you, Eric. And I want you to help us understand what are the basic markers that we could measure in order to even study this whole idea of glycemic response, insulin resistance, et cetera, and tell us what we did in the study and tell us how we did it. speaker-2: Sounds great. Yes, we had 1,100 people come through our doors over several months to years when we were collecting a lot of data. And for every individual, I believe we were collecting about two weeks of data in various capacities. And over those two weeks, we designed a very specific diet, which I'll get into with MoMA at later point in this conversation. In terms of the metrics, there was a little bit of setup. So what would happen is each individual came into the office and we collected some anthropometrics and some of those anthropometrics are straightforward. We collected high weight, calculated BMI. We also did get HbO1C a baseline. And then in order to follow them over two weeks, we had apps that were set up and we had permissions on their apps to track them and their compliance. And before they left, we had a continuous glucose monitor put on their arm. And that monitor tracked their blood sugar in, I believe, either five or 15 minute intervals continuously for two weeks. And over that time period, we would then have them take photos of what they ate and their scheduled meals. And some of the other metrics that we were able to collect, we did give them an activity tracker and we had their time periods for light and deep sleep, as well as the number of steps that they took. So what we're taking a look at here is that data fusion. The blue line is the continuous glucose monitor. The dark grays and the red bar, red lines at the bottom is their activity and their sleep levels. And then there's a couple of instances that you can see in this data fusion visualization, which includes the food and the meal time that they had. So we were able to take all of this data and do some heavy curation and we were able put together this glycemic response model to be able to predict what we should be feeding to people. Any other questions about the metrics here? speaker-1: Yeah, I have a couple of things to add from my experience here. First of all, we bought 40,000 meals total for these participants. And I remember going to Whole Foods and ordering a truckload of food. And they were just like, what is this for? What kind of restaurant are you running? And I'm like, we're not. We got 550 hungry people ready to eat. So that was an interesting vignette from running the trial. And then the other thing that I remember is that Damon spent countless hours looking at all the CGM curves and aligning them and grouping them. So maybe he can give us a bit of an insight of that analysis and how he sorted it and what he learned. speaker-3: It was pain, it was painful. speaker-0: I mean, look, let me just point out something here. So first of all, I gotta also say that Eric was running around the New York City area. Sometimes when people forget to take their food, he would take their groceries and drive all the way to Staten Island from Manhattan and drop you off at their house when they couldn't do it themselves. most of the time it was fine. We did the same thing in the Whole Foods here in New York City as well. coming back to the curves question, right? So what I wanted our audience to notice here is if you can see my... mouse, this little bar, vertical bar here on the chart is a meal. And when this person had this meal, which by the way, shows their carbs, their fat and their protein levels as well, those are three different colors, the blue, the yellow and the pink, you can see that the blue continuous glucose monitor line shoots up after a little bit of time and then it comes back down. You can see every meal, like down here, there's another meal. They had this meal, presumably that was breakfast because they just got up in the morning after sleeping, as you can see from their sleep tracker, the curve goes up and then comes back down. So the question we asked Damon was, when does the shape of this curve indicate that this person is okay or is not okay? You know, does the shape of the curve tell us something? It turns out there's a Yes is the answer. There's a lot of literature on how to study the curve shapes to figure out what is a healthy shape, what is not a healthy shape. It has a lot of factors to it. It's not just about how high it goes. It's also about how long it is, but it's some combination of the two things. So, Damon, can you explain to us how you think about the shape of that curve? speaker-3: Yeah, it's a great question because the curve represents sugar, but implicitly it is representing insulin behind the scenes, right? And so if you have ⁓ a slow, steady increase of sugar, glucose, then you can assume that in the background, the insulin is gating that response, right? So people that are healthy will have slow increase followed by a slow decrease. So it'll be like ⁓ a nice rolling hill, right? Whereas people that have insulin resistance will have a quick upswing of sugar because the insulin is not quite doing its job yet because the body is resistant to it, right? And so you'll get this spike that's more pointed. And then as long as you have a healthy pancreas, you then get a flood of insulin. So you get a rapid decrease down. So it's a tall pointy mountain, narrow mountain versus a lower amplitude hip. speaker-1: Okay, it's like the Rockies versus the Appalachians. You to be hiking the Appalachians here, not the Rockies. speaker-3: Yes. ⁓ You do, but even more than that, sometimes if you have an insulin response that's over robust, you will actually have an over correction and you will fall lower than the sugar that you originally started with prior to the meal. so that's that is now, and remember, if you're looking at metabolic health, this is representing very high inflammation. Not only is it pushing all this fat into the organs, through the insulin, but the insulin being the most inflammatory hormone that we have. When you have huge surges of insulin, your inflammatory tone is... speaker-0: off the charts. That's very interesting. speaker-1: right, so one takeaway here for the audience is that if they are wearing a CGM and they see very low level or hypoglycemia, that's also a signal that they may have insulin resistance, correct? speaker-3: they absolutely have insulin resistance because you shouldn't have, your body should be tuned to the point where it doesn't allow you to have prolonged high. speaker-0: So the one thing that I wanted to point out from a data perspective here is that rather than looking at all of these different shapes that Damon is talking about, there's literally like hundreds of different shapes that you can imagine. And you know what, one of the things he did for us is to cluster those shapes in such a way that we could start analyzing them and labeling them as good or not good and so on. But more importantly, what we could do is mathematically we could calculate the area under the curve. It's also known as the integrated area under the curve, IAUC. That's the term that I'm gonna use later on, the integrated area under the curve. And when you calculate the area under the curve, you can get a first level approximation of whether that's a decent response or it's not a decent response. But I say first level because you can have... the same area under the curve, but you can have different shapes and those shapes matter too. So that's the reason why we had Damon take the curves that had more or less the same area under the curve, but classify them and say, okay, this shape is good. This shape is not great. And so we were able to incorporate that into our downstream analysis, which I'll get into in a minute, but I want to switch the topic slightly to the types of foods. see here a picture of one of the meals that this individual ate during the study, they took a picture on their phone and we had a system set up like Eric just explained that we would capture it on the data end and we can show that in this data fusion. And this particular meal turns out it has this meal ID, it has this amount of calories, it has this much fat, protein, complex carbs, et cetera. The whole breakdown was right here on the visualization. But more importantly, Eric, I want to come back to you and ask the question, how did we even decide what meals to give these people. speaker-2: Right, thank you for that question. ⁓ We spent a lot of time with several nutritionists, clinical nutritionists, to sit down and think about what kinds of foods and meals do we want to give people both to measure this broad space of carbs, fiber, protein, fat of various levels within a meal, as well as how do we collect this data in a way that makes it easy for the individual as well as spreads out that data set over a two week period. So we're not necessarily biasing the data by putting the same meal right after another, know, two of the same types of meals right after each other. So we sat down and we said, we started with a diet that was proposed by the clinicians and we said, let's put together 13 different diet types. And we calculated that out over ⁓ high and low values for carbs, fiber, protein, fat. And the nutritionist laid out some possible sets of meals. And we spent probably several days, if not weeks, going back and forth and saying, okay, here is a stick of butter or half a stick of butter for the fat and here's the rice. Let's calculate that. Let's figure out how to revise it. And the goal really was to figure out 13 different diet types that we gave them. I believe it was twice each over that period of time. And we also had in there nine specific meals that we repeated 100 % at randomized times throughout that two week period to have a touchstone data point to come back to. So it was a lot of collaboration and planning and computation on the fly to say, here's the high, high, high, low. right? When we're thinking about high carb, high fiber, high protein, low fat. And so we did that for each of those food and meal types. speaker-1: Okay, is it fair to summarize this as these are 104 different kinds of meals that capture a very broad omnivorous diversity and you built in meals that were repeated at least once so that you can capture the glycemic response to the exact same meal on different days to see what the variance is. speaker-2: I think you've summarized it. speaker-0: Again, from a data perspective, this is how the data looks like, right? So the blue dots on this picture here are the specific meals that we designed, the 104 meals that we designed when we started the study. And you can see any dot here has a particular proportion of fats, particular proportion of carbs, particular proportion of protein. And the red dots are the meals that were actually consumed by the population, the study population. So that's obviously a lot more data points here. So we collected tens of thousands of these red dots, which correspond to all the different standardized meals that we provided, but also some free meals that people had. The second half of the day, we let them have free meals after lunch. They can have free meals, but they can just record the contents of it. And we make a best guess as to whether or not it fits within the framework that we want it to study. Right. So we collected all this data and then We had to understand and analyze the data to create a model for predicting glycemic response. speaker-1: And I'd like to say a couple of things here that are foundational. One is that, first of all, this is a publication that people can go and read. So if they want to geek out and read all the fine details of this, it's published and we'll list the publication of the show notes. And the second thing I want to mention is that while this type of an effort, similar style, but a smaller scale had been done before by previous groups, I want to highlight that ours is the only effort that uses RNA sequencing for this approach. speaker-0: Exactly. So before I go into the analysis, I want to do a small primer on machine learning. I think if you have to understand how to analyze this large data set. Again, remember when I say large data set, I'm talking about here, let's say a hundred thousand data points, hundred thousand meals that we can get from all of these individuals. Now you have to understand all the different dimensions of that data, but There's another crucial thing. Each one of these participants also got their gut microbiome tested at the very beginning at a baseline. And so we have a high dimensional data set from the gut microbiome, which is all the functions, every single gene that's expressed by every single organism that's in your gut. So we had to extract all of that data and we have to understand how the gut microbiome impacts the glycemic response because now we have all this other data from the study. We have the gut micro one. Now we have to combine the two things. Right. So that's the problem that we had in front of us. So in order to understand how we did the machine learning, I just want to give you a very simple example. So let's start with a machine that has been given a lot of examples of a certain object. OK. And it's very, very popular to think about cat pictures. in machine learning literature. imagine, and there's of So imagine a machine is shown like a million cat pictures and a million dog pictures. Okay. That's called training, which is it looks at these pictures, it analyzes pixel by pixel, and it creates an internal data structure. As we call it, it's actually happens to be a neural network data structure. I won't go into the details of that right now, but speaker-1: And there are plenty of them on the internet. speaker-0: Just imagine that as it's looking at each picture, it is creating an internal representation of that picture and understanding what are the key differences between cats and dogs? What are the features? It first looks at each individual pixel, then it looks at each individual line, then it looks at each individual segment, and then it looks at the colors of the segments, then it looks at the geometry of the segments, then it puts everything together. So each of those levels in the neural network understands a different level of abstraction in the picture, which is the pixels, the lines, the segments, the geometry. Every single one of them is understood by a new layer in the neural network. So that neural network is called the model of the classifier between cats and dogs. speaker-1: And Guru, one important feature here is that you tell the machine which ones are... speaker-0: Beginning you say, here's a cat, you know, figure out what this does. Here's a dog. Here figure out how it looks. And so you keep doing that. This is called supervised machine learning. And that's just one type of machine learning. There's multiple other types. There's another type called unsupervised. We don't tell the machine what the picture was originally. Okay. And then there's a new thing. This is the latest, greatest thing. It's called self-supervised, meaning that no one externally has to tell the machine. This is a cat or a dog. Literally there are models today. that take a cat or a dog and they can self-supervise, meaning they say, if I were to see a series of pixels that look like this, what is the next most likely pixel? And it learns based on the history of the order of the pixels. That's actually the basis of all the transformers and the large language models and everything else that we're talking about. But that's for a different episode for me to explain all that stuff. I'll go back to the basic method. which is the supervised machine learning, which is where you say, here's a picture of a cat, here's a picture of a dog, understand it in your neural network, and then now I'm going to show you a new picture, and you have to tell me whether it's a cat or a dog. I didn't tell you whether it was a cat or a dog, you have to tell me whether it's a cat or a dog, and the machine takes that internal representation of the neural network, and it goes through the same analysis that it did while it was understanding the image of cats and dogs, and it says, ⁓ Does this new image that I don't know what it is look more like a cat or more like a dog? And that is built into that neural network now. And then it comes back with an answer and says, ⁓ this one is a dog. And so that is known as a classifier in machine learning. And there's many other ways of doing things. And, and you can also do not two different animals. You can do a million different objects. You can do a completely arbitrary language process. You know, you, you predict. and generate what kind of language comes out of that picture. You can do all sorts of things, but I'm going to stick with the simple supervised machine learning with just two answers, cat or dog. That's what I just explained to you. Now, how does that apply to our glycemic response study? What we are trying to determine here is just two things. If I know this person's gut microbiome, and if I know how a whole bunch of people responded to a particular type of food, then can I tell this person whether this food is going to spike their glucose in an unhealthy way or it's not going to spike their glucose in an unhealthy way? Those are two answers. That's the cat dog, you know, analogy here, right? So in order for the machine to be able to say, yes, this is okay for you to eat or no, it's not okay for you to eat, you to first train the machine with many, many, many data points of why something could be good. and why something could be not good. What kind of gut microbiome results in a good response? What type of gut microbiome results in a bad response? What types of glycemic responses from a CGM is considered good? What is not good, right? That's what Damon and Eric and Momo and all of us wanted to do in the study. in the end, after we collected all the data from the study, we ended up having a huge data set for training a machine. Turns out... The model looks like this. The training data came from about 1,100 people. They were both in the US and in Japan. We started off with something like a hundred thousand meals or something, but when we figured out which are the specific meals that we can use to reliably train this machine, thanks to Damon's classification clustering, where we had to whittle it down to highly reliable data that was 60,000 meals. And again, this was collected over two weeks of all of those people, 1,100 people with the CGM data, with the microbiome testing and all that stuff. We took all that data and we fit that into a machine and made it learn the patterns in that data. And now we asked the question for a brand new set. said, if we were to give you a new person's data, their microbiome, and a set of foods that have the same characteristics in terms of the carbohydrates, the proteins, the fats, and the fibers and everything else that we explained before, predict for me what is the likely glycemic response for this individual. And so we took those predictions and that's on the x-axis over here. We took the predictions of the machine and then we actually compared it to the actuals of what we had actually measured previously. And we saw whether there was a correspondence. And the good news here is, you see this red line, the red line being across close to the diagonal, basically means yes, this agreement between the actual and the predicted. And once you can see the diagonal line like that, you can tell somebody, ⁓ listen, we are able to pretty reliably predict your glycemic response to a banana or a piece of bread. So What we had to do then is to take this glycemic response predictor. We call that a GRM, glycemic response model. We can take that predictor and embed that in the Viome recommendation engine. speaker-1: I just want to comment here, that this is so far everything you've explained would have been done in science like 99.9 % of the time, which is scientists study something, they conclude something, and they publish something. And then that's it. It never makes it into the hands of people to actually use in real life. And so this is now- speaker-0: the translation exactly. Thank you, for highlighting that. So the point is, it's not just a study, it's now turning into a production system that can help every customer of Viya. Damon, did you want to say something? speaker-3: I was just going to say is highly practical and I agree with Momo, 99 % of traditional research is not practical at all. ⁓ So this gives us a way to actually have a personalized intervention based on highly technical science. speaker-0: So what I wanted people to appreciate here is that we have a very large food database, hundreds and hundreds of foods. And every single food, we know the macronutrients of that food. We know the micronutrients of that food. Just based on the macronutrients, we can run the level of carbs, the fats, the proteins, and the fibers through this glycemic response prediction model that I just described. And then you get An answer you say, ⁓ does that fall into this blue enjoy foods bucket or does it fall into the orange minimize foods bucket? And so if a banana could be. An enjoy food for you, but it could be a minimize food for me because my glycemic response is going to be very different than yours based on my microbiome, based on how my body works and based on all the functions that we've talked about so far. So we do this now for every. customer vial for every food in the database and we tell them if something is going to spike their blood glucose, minimize it. And we also do further analysis in the microbiome pathway, which we're not going to go into today because that's another topic and it's a lot deeper about the inflammation, about the oxidative stress, about oxalate elimination, et cetera, et cetera. Those are all analyses that we do based on the micronutrients that are available in your food. We do that too. and then decide ultimately whether something should be a super food and enjoy food or minimize food or avoid food. But that's a topic for another conversation. I just want to say that the glycemic response prediction model is step one in the Viome recommendation engine. speaker-3: And let me make it real quick comment. Traditionally, we used to think that the glycemic index or the response to a food was inherent in the food itself. And I was taught this in medical school and I was taught it in endocrinology fellowship that certain foods have static glycemic indices. It doesn't matter the person, it's the food. And I think it's important and maybe. Maybe people already realize this, but it's important to focus on the fact that that is untrue. speaker-1: super important. speaker-3: Yeah, the glycemic response to a certain food is individualized and it can change as your microbiome changes. So if you have a problem with bananas, it doesn't mean that you're necessarily going to have a problem with bananas forever. If you improve the health of your microbiome and you're able to process bananas in a more physiologic way, then you can enjoy bananas. But at a current state when your microbiome is unhealthy, bananas cause inflammation through by way of insulin response. speaker-0: Super, super important point, Damon. And you're right. So when we see the data for the Viome customer population, we can show that the food is not the determinant by itself of the glycemic response. It's the food and the individual biology. You have these two people here. The person on the left, who we are calling participant P3, has a spike. For banana, that's the green spike. And another participant, P4, has a flat response to banana. speaker-2: So, I want to also offer a little bit of additional info here while we're talking about this particular slide, because I know we're going to have some audience when we're talking about food composition being equally important. And I think it's important to just mention that we understand that the formulation of the food, if it's mashed potatoes, baked potatoes, that also changes some of this. And in the image we're looking at here, we always made sure to tell our participants to aim for a green to yellowish banana and to avoid the brown bananas because that also speaker-1: Great point. speaker-0: So, ⁓ absolutely. So, it's quite possible that on the one hand, the left person is spiking for banana and the right person is flat for banana, but the right person is spiking for bread and the left person is flat for bread. So, I mean, it can be completely opposite. So, it's about you as an individual. It's not about the food itself. speaker-1: And to add to Eric's point, that bread was actually sliced bread from Whole Foods that was super standardized. There was no difference in that bread. And look at the difference in the response. speaker-0: So how does this translate to a population is what I was going to say as well. Right. So I pulled up some data from the whole from about 10,000 Viome customers that we've given recommendations to. And this is what it looks like. So if your individual biology determines whether your blood sugar spikes or not. if you're on the right hand side of this chart, that means your blood sugar is not spiking. So you're in the safe zone. If you're on the left side of this chart, that means your sugar is spiking. You're not in the safe zone. You got to watch out. And I took a few examples because early on in this episode, we talked about millet and how you can make uttapam from many different grains. happen to use millet. Turns out millet right here in the middle, most people are going to be in the green zone, 83 % of the people, but 17 % of the people are in the red zone. But if you look at bananas, you have more or less half and half people. on the left and the right. So do you know where you are? Are you on the right or on left? If you want to know, you have to test something like Viome so you can tell whether or not bananas are going to spike your blood sugar. But equally importantly, if for example, look at this white rice here, 90 % of our population is in the red zone, only 10 % of the green zone. I keep telling my mom and I know my mom is going to be watching this. keep telling her to not use white rice and use brown rice. So look at the difference in the brown rice response. Brown rice, it's the opposite. Most people, almost 90 % of the people are in the green zone and only 10, 12 % of the people are in the red zone. So it just flips the whole story upside down. So you can make choices. In order to make choices, you have to know your own personal response. Now, if you happen to be in this 10 % green of the population for whom white rice is great, go ahead and eat white rice, but you're more likely to be in this green bar here. But I've switched to quinoa and I can make quinoa flour. Quinoa is generally speaking from a glycemic response perspective is great. It's almost a hundred percent of the people have no spike according to our data. Tapioca, I take. great care. know, tapioca is almost 100 % of the people in the red zone, so I try not to use tapioca too much. But these are the kinds of things that can help you personalize your individual diet on a daily basis. speaker-1: Yeah, but this is just a sample of carbs. There are so many carb containing foods and really the Viome app will show you for every single one of those foods what your response is. So you can guide your cooking a hundred percent based on your individual biology. And that's really the key concept that we addressed here with this. speaker-3: I think it's important to realize that in this case and in all cases, glucose is just a surrogate marker. So when you look at, say, white rice and you say it's spiking the glucose and that's a problem, it is, but not so much from a standpoint of diabetes, but from a standpoint of the insulin that's released that has to handle the white rice. That's people that are listening may say, well, I don't have a problem with sugar. don't, what does it matter? Well, it does because it matters how much work your pancreas has to make to keep your sugar down. And during the process of releasing a bunch of insulin, you're causing inflammatory and detrimental effects to your body. speaker-1: And you're storing fat in the places that are going to, you know, that are going to spike all of the above that you mentioned inflammation and so on. So it's a vicious cycle. So if you're going to eat white rice, you can eat it, for example, with mango, like mango sticky rice, it's all carbs. There's no fat to store. So if your insulin spikes, you're okay because there's no fat to store. But if you eat white rice and it spikes your blood sugar very high and you eat it with like a steak, that's going to have a lot of saturated fat. that's gonna be a problem because the insulin is gonna go high and your fat's gonna be stored. speaker-3: Yeah. And I think the weight, ⁓ like it depends on what people are, are wanting to get out of this, but a lot of people are wanting to, ⁓ to optimize their weight. Let's just be, let's be clear, right? the reason that the GLP ones are, have become such ⁓ a massive discussion topic is not because of sugar and glucose control. ⁓ you know, ⁓ we were treating glucose and sugar control. 15 years ago with GLP-1s, they only became a national discussion when people realized that they could control weight. Why do they control weight? Because they lower insulin. That's why. They are by nature insulin sensitizers. And because you're lowering the overall insulin or the hyperinsulinemia that is prevalent, pervasive, ⁓ you lose weight. And so I don't really care that my patients only care about weight because it's make them want to also improve their metabolic health because you can't improve your weight with a GLP-1 or a diet like a Viome suggests without also improving your metabolic health. speaker-0: hot. Excellent, excellent points. the one critical point that we have to discuss is to bring this all home and say, okay, now that everything has been analyzed and somebody gets a recommendation, does it actually make a difference? I mean, if you were to follow somebody's recommendation and say, eat this, don't eat this, try this, minimize this and so on, does it really make a difference? So, Momo, I think I'd like to ask you to just walk us through how do we show that this kind of recommendation actually matters for you. speaker-1: Yeah. So you spoke about how the Viome recommendations based on the glycemic response model should reduce the overall glycemic burden in the blood. So that's one thing. But we have many other scores and we measure many other physiological phenomena that reduce the inflammation and that improve functional richness of the gut and so on. So when we combine all that, we would hope to see ⁓ positive impacts on people's insulin, people's glucose, as well as people's HbA1c, because HbA1c is sort of the average glucose burden over a three month period. So we were hoping to see that, but you don't know until you actually test it. So to test how well our recommendations work, we launched a blinded randomized placebo control trial for people with mostly pre-diabetes. And now if you can share the slides, we can just show you some graphics that we prepared. This is for... This is all published in a pre-print and it's currently under review. So the highlights here are that in three months, the interventional arm participants who followed the Vion diet were able to reduce their HbA1c by 0.42%. These were pre-diabetics. So that's a clinically meaningful improvement with a statistically significant result. And then if you go to the next slide, while we did not reach the statistical significance for fasting insulin and fasting glucose, we did see very dramatic clinically meaningful improvements. so on the left side, we're showing that fasting glucose was reduced by 26 points, 26 milligrams per deciliter. ⁓ I'm going to let Damon comment on how significant that is clinically. And then on the right side, we can see fasting insulin improved by five million international units per liter in just three months. Damon, do you want to comment on these clinically? speaker-3: Well, to be really simple, these are highly clinically relevant. To lower your fasting glucose by 25 points is on par with what we see with some medications. And to lower the fasting insulin by five, when the four insulin wasn't that high anyway, so percentage-wise, you're lowering the insulin probably 20, 30%. Yep. That is... highly relevant and clinically very important. Statistics are one thing, right? We all know that and whether a trial is powered to show an effect, but I can tell you from seeing 20,000 patients in my career, a reduction of 25 in five is highly significant. speaker-1: Yeah, and for the context on the right side, starting baseline fasting insulin was 14. speaker-3: Yeah, so that's 33 % drop. speaker-1: Right, right. So basically on average, these people were pre-diabetic because 14 is pre-diabetes, right? And then they went to average down to nine and nine is healthy, speaker-3: It's better. ⁓ So, you know, ⁓ it's better. My professional athletes have insulin of two and the so-called normal insulin is supposed to be five, but it doesn't matter. You're moving in the right direction by a magnitude of 33 % after only three months with no medication. speaker-0: you speaker-1: Yeah, yeah. No, we're very proud of these data. And these are basically class leading data for showing that nutrition can make a very significant impact in just three months. And I also want to highlight something that we need to highlight, I think, in every episode is that when we talk about bio-nutrition, we are not talking about fad diets that are so restrictive and so awful that they're basically impossible to follow. Many of these diets are so restrictive that... Sure, people can stand them for a month or two or three, but not for life. Whereas the key here is that every Viome diet is very omnivorous. People still get choice of hundreds of foods to eat. they can eat... I mean, I have a more restrictive diet than most Viome customers, and yet I can eat a different meal for a million years. That's how much diversity there is, right? And so ⁓ that's really a key that these are diets that... you that allow you to establish a very positive relationship with food and cooking and really do this. speaker-0: I gotta bring it back to the uttapam that we talked about, In this episode, is... Sure. You can make uttapam so many ways. It's a very, very flexible dish and you just have to pick the right ingredients for yourself. And you can make it super tasty, you can make it super healthy, and it's very fast too. You can do it in like 20 minutes. ⁓ So it's one of the best foods. It's a great sort of a carrier for taking all of these messages to your So with that, think we're to thank our guest, Damon. It was a great pleasure having you on the show. It's been a little while since we talked and it's great to catch up with you and wonderful comments. Eric, we always talk about many, many things every single day, working on various many topics, but this is one of the projects that I will remember for a long time. did this project with you. Thank you for leading it. Thank you for doing a great job on it. And I think the results speak for themselves. So thank you both very much and we will see you around. speaker-1: Thank you guys very much, very grateful for you. All right, great. Now we are just going to give you some reminders. Please subscribe to the podcast. Please like it, please share it with all of your social cloud. And ⁓ a reminder of our monthly challenge for the month of April, ⁓ we are asking you to make a social media post about personalized nutrition, your experience or a publication or any aspect of personalizing nutrition. Include our podcast hashtag in your post. So that's at or ⁓ hashtag two PhDs on a pod in your post, and then share the URL of your post as a comment on one of our episodes or send it to us via email at podcast at my own.com. And then we will pick the winner based on the content and the engagement of your post. That's it. That's all you have to do. The winner will receive our premium consumer kit, which is the full body intelligence, the stool, blood and saliva kit with tons of results and recommendations for. personalized nutrition. speaker-0: And specifically, it includes the recommendations based on your glycemic response, which is what we talked about in this entire show, which is awesome. And that also brings us to the topic for the next episode on our series, which is going to get even deeper into how does Viome understand what's in the samples that you send, right? So how does the little... know, speck of poop that you send turn into data that helps you improve your life, improve your health. How does the finger prick blood, the spit that you send, how does that help us, first of all, understand your biology and then explain that to you in the app so that you can then use that insight to make a change in your life, in the food that you eat, in the supplements that you take, in the lifestyle that you follow. So that's going to be the next episode. We're going to go from the wet lab. Momo, you're going to... show us some videos of the wet lab and you're to explain to us what you do in the wet lab. then we're to have a guest who is going to help us understand a significant part of the dry lab. And I will contribute into that as well and talk about what we do in the analysis. So I think we're to have a wonderful episode. So I hope you'll all join us. Subscribe, like, and come back to the next episode. speaker-1: Okay, that's it for today's episode. I'm Momo. speaker-0: I'm Guru. We're two PhDs on a paw. Woohoo! speaker-3: Two PhDs on a pod is a Viome Health Science podcast production written and hosted by Viome CTO Guru Banavar and Viome CSO Momo Vujicic. Produced and edited by Lisa Shomo. Podcast art by Kevin Wu. If you enjoyed today's episode, like and subscribe for new episodes every other week. We'll see you next time.