Written by Christopher Kelly
Sept. 16, 2017
Christopher: Hello and welcome to the Nourish Balance Thrive Podcast. My name is Christopher Kelly and today I'm joined by Dr. James McCarter, MD PhD, head of research at Virta Health. Dr. McCarter, I am delighted to have you. I've been waiting very impatiently, I think, for over a year to do this interview so I'm incredibly excited to have you today. Thank you very much for joining me.
James: Well, thank you, Chris.
Christopher: Tell me about -- So, I really want to start at the beginning here and understand how it is you became interested in health in the first place. Could you talk about how you became interested in health?
James: Sure. Let me give you a little bit of my personal story. I'm originally from Chicago, went to Princeton and then onto Washington University in Saint Louis to do an MD and PhD and didn't really have nutrition on the agenda. I was diving deep into things like genetics and developmental biology and eventually infectious disease and ended up starting a company in that space, company Divergence that worked on the control of parasites in plants, animals and humans.
And it was really after that whole experience -- That company was quite successful and was acquired and I was doing corporate venture capital about five years ago and was seeing more and more opportunities to apply biotechnology and technology in general to health and wellness but was particularly interested in this area of nutrition. And perhaps I should have clued into that before. I'd been an athlete growing up and had gone to medical school but nutrition wasn't just never really on the agenda until I started exploring this really in a quantified self way about five years ago.
Christopher: Okay. And I saw your presentation, the quantified self presentation. I think that was how I first became aware of your work and I should link to that video in the show notes. You discovered ketosis and thought it was a really interesting thing. Can you talk about your own personal experience with nutritional ketosis?
James: Sure. So, my first experiments in quantified self were around omega three and omega six-omega three ratios and trying to take supplements and measure that and show that it could have a difference. But it was Thanksgiving of 2012 and I was working for a crop protection company in their corporate venture capital group and my sister made this offhand comment to me at Thanksgiving.
She said, "A lot of that corn that you guys are growing goes into high fructose corn syrup and why is that a good idea?" And I just made a comment like, "Well, I don't think it's any worse than sugar." And that got me reading about sugar. So, I went hunting for references for my sister and ended up landing on Gary Taube's books and eventually Steve Phinney and Jeff Volek's books and that took me down the rabbit hole of really beginning to see that there was this impact of sugar and readily accessible starch on human biochemistry and chronic disease. And that led me to doing quantified self experiments in ketosis.
Christopher: You said something really interesting to me just then. I don't want to get too off track but can you just talk about what you learned about the omega three to six ratios? Because I don't hear people talk about that too much. Did you find anything interesting of beneficial there?
James: So, one of the things about the kind of 20th century is this rise of vegetable oils and which are used in processed foods and in cooking. And they tend to be high in omega six, which is a polyunsaturated fatty acid, actually an essential polyunsaturated fatty acid that we need in the human diet at some dose.
The problem is that the dose that we're getting is manyfold what we would have gotten historically. And there's another essential polyunsaturated fatty acid which are the omega three fatty acids that often you think of as coming from fish and people tend to get not enough of those. So, one of the kind of things that's been thrown out there is that historically people should be getting more like a one to one ratio of omega three to omega six fatty acids and that this has an influence on the composition of our cell membranes.
And whereas the modern American diet is something like a 20 to one overdose of omega six to omega three. And so the two ways that you can address that one is to just cut back on exposure to omega six. So, you'll have some of it in your diet. It's impossible to avoid but perhaps have less of it in a conventional diet. And then to increase the amount of omega three either through eating fish as a regular part of the diet or through supplementation.
So, that's basically, I did just a very simple experiment of just taking a supplement, omega three supplement, an algae derived omega three and measuring before and after to show with the home test kit for blood just to show that the omega three level is rising and the omega six to omega three ratio is changing.
Christopher: And did you see any difference with the supplement?
James: In terms of health effects?
Christopher: Well, in terms of the test result, would be the first thing.
James: Yeah, sure. Sure, yeah. So, the test result was pretty simple. It was pretty night and day.
Where I had rather low omega three level prior to supplementing and then after supplementing for a number of months, and I still continue to take an omega three supplement, showed that the level did indeed rise and the ratio did indeed change. And in terms of effects, that kind of got bound up in what I was doing with ketosis because I had started that experiment in a few months later. But what I saw from both the omega three supplementation together with the ketosis was a decrease in inflammation.
Christopher: Interesting. I can very much relate to what you said about diet. I've been visiting the UK recently and grapeseed oil is all the rage. And most of my US listeners will have never heard of that oil before. It's more commonly known as canola in the US. It's usually cold pressed in the UK so it's not industrially processed seed oil, but it can be in the US.
But it is everywhere and it is in everything. Like if you eat anything that's pre-processed and it's got ingredients, it's almost certainly got grapeseed oil in it. And now we're in Lisbon in Portugal and we're eating sardines at least once a day sometimes twice a day, so it's really interesting to see there's cultural differences between the oil consumption.
So, we tested on our standard blood chemistry that Tommy Wood, my MD PhD, designed and we see that people who eat sardines, these small oily fish, they always rank in the low risk category even if they're not supplementing. So, I thought that was really interesting.
James: Yeah. I do eat sardines regularly as well, usually packed in olive oil.
Christopher: Yeah. They're super good. There's really only one variety I found in the US that I really like, the Wild Planet variety. But here in Portugal they have shops and even museums dedicated to all different types. It's almost like it's a thing, like artisan sardines in Lisbon in Portugal is amazing.
Christopher: Awesome. Okay. So, let's talk about Virta. Tell me about how it all came together. Tell me the story of Virta.
James: Sure. So, my connection to Virta was through a friend in the venture capital community who I was fortunate enough to be introduced to Sami Inkinen in late 2014. And as soon as I read a little bit about what he was doing, I jumped on a plane and I had dinner with him in January of '15 and that was about two and a half years ago. That was my kind of introduction to Virta. But the Virta story really goes back to the late 1970s and to work that Dr. Stephen Phinney was doing. And Steve is an MD from Stanford and PhD from MIT. And you probably are well aware, Chris, of this bike racer story and his study back then.
Christopher: Of course, yeah. And I'll cite it in the show notes as well for people that want to see that study.
James: Yeah. And so, what Steve basically showed was he tried to prove that this idea of fueling on fats was a mistake and that that would be impossible to do. And everybody knows that before going out for a bike ride you want to carb load. So, this idea that you could have athletes fueled on 70% or 80% of their calories coming from fat, that they would still be able to achieve a level of output was just foolish. And they ran that study fortunately for about three weeks and they were able to show that after a few weeks of adaptation that people actually can run on fuel.
So, Steve had this idea of fat for fuel as early as the late '70s, early '80s. And so fast forward to around the year 2000 when he began to do work in Portland, work with Jeff Volek, and they were exploring this idea of nutritional ketosis in much more detail and applying it not just to things like athletics but applying it to chronic disease and to treatment of metabolic disease and type II diabetes.
And Jeff and Steve along with others really laid these ground work of clinical studies over the last decade and half that basically show that if you shift away from a carbohydrate based diet toward a fat based diet you can actually resolve many of the challenges that people face in metabolic disease and type II diabetes.
Christopher: That's incredibly interesting. I've always wondered how is it not the standard of care? So, as somebody without a medical background, when I read about insulin resistant type II diabetes, it reads like a disease of carbohydrate intolerance. Do you have any ideas why it's not already the standard of care for this disease?
James: Yeah, excellent question, Chris. I think more than anything else is this fear of fat. And that really dates back to the 1950s, the 1960s and then indoctrination of that idea in the dietary guidelines in the late '70s with McGovern senate commission and then the USDA and FDA dietary guidelines and the whole kind of creation of the food pyramid, this idea that there would be seven or more servings of grains at the base of the food pyramid and that you needed that for energy, and that fat would be the driver of heart disease, particularly the idea that saturated fat was driving heart disease meant, therefore, that you should cap it in the diet.
The evidence for that which was based on epidemiological studies was never strong and particularly in recent decades it's really fallen apart. If you talk to people about this idea that could you use fat for fuel as opposed to carbohydrates for fuel, the immediate reaction is, "Well, is that safe for cardiovascular health?" And so, I think, it's that this idea that was kind of built into the dietary guidelines back in the '70s and '80s that has people suspicious of this idea.
Christopher: And then so, I just wonder -- So, you talked about some very prestigious academics, Jeff Volek and Steve Phinney, do you think at one time they had aspirations to change the dietary guidelines and now they've given up and gone this different route that is Virta Health?
James: Yeah. So, I think that Jeff and Steve have been influencers on so many levels. They're publishing in the scientific literature, they're conducting studies, they're presenting at conferences, they're writing books. So, they're getting the word out there in so many ways. But in terms of really taking this to scale, Jeff and Steve were fortunate to hook up with Sami Inkinen who is a dynamite entrepreneur and the story -- I can tell you the whole story of how they kind of connected.
But what ended up happening was that Sami had had a successful entrepreneurial career with a company that he co-founded called Trulia in the real estates base and had shown kind of how you can take an idea and kind of build it to scale and where it's impacting millions of people in terms of their real estate purchasing decisions.
And Sami, as an athlete and somebody interested in health, had begun to do health tech investments. And then discovered that he himself, despite being a world champion triathlete, was prediabetic. And that had him asking the question: Why, if I'm exercising three hours a day and eating what I think is the right diet, why am I having this metabolic health issue? And that led him to Steven and Jeff's work. And the thing that's really cool about bringing Steve and Jeff together with Sami is that it brings together this kind of world class science of nutrition and metabolic health together with somebody who really understands software and scaling and the Silicon Valley model of building a company.
Christopher: That's a powerful combination, I'm sure. Talk about the mission. It's extraordinary statement that Virta has a goal to reverse diabetes in a hundred million people. That seems unthinkable. Do you think it's really possible? It's surprising to me that even that many people are affected by the disease.
James: Yeah. When you first hear that there's this kind of gasp of, "Hundred million people? Can you really do that?" Especially for a startup company. But let me tell you why I think it makes a lot of sense. There are 30 million people in the United States with type II diabetes. There are 400 million worldwide. And you might look at that number and just say, "Oh gosh, how do you ever deal with that?" And particularly when the American Diabetes Association makes statements like saying this is a chronic and progressive disease and once you have it you'll always have it and the remission rate is one-tenth of 1%."
You hear statements like that and people just think it's their destiny to carry around this awful disease and perhaps they can manage it and slow it but that it's going to be chronic and progressive and impact the rest of their lives. But think back to where we were in 1980 or think back to where we were even before you and I were born. This was a rare disease. So, it's increased fourfold. Type II diabetes has increased fourfold since 1980 and it increased several fold before that.
So, back at the turn of the century, the 19th century, it was hardly ever seen at all. And when the Joslin Diabetes Center was, when Joslin himself back in the 1920s was practicing medicine, he would talk about type II diabetes as being a rare disease that he rarely encountered. So, it was sort of a special note in his journal when he would see a case of type II diabetes.
The point being though is that we had this huge expansion in this disease, this epidemic of type II diabetes. To me, that actually gives hope that it can be reversed because if these were an underlying genetic defect that we all were destined to get no matter what then it would have been common several generations ago, right?
So, the fact that it has taken off in this way suggests that it's something environmental. It's something about exposure that wasn't occurring several generations ago. That gives hope for the idea that if we can reverse that exposure the disease can actually be -- It self reverses.
Christopher: This is the most interesting question. We've talked about this a lot on the podcast before. We really like to think like engineers. I still think of myself as an engineer and not some sort of health and fitness professional. And Tommy, although he's a medical doctor and a research scientist, he also thinks like an engineer. And so we talk a lot about understanding the root causes of a problem.
So, engineers, you think about when a plane falls out the sky or a bridge collapses or something awful like that -- and this is, we are talking about something awful like that -- the first thing you need to do before you start meddling with potential intervention is understand the root causes of the problem. Why did that plane fall out the sky? And so what do you think it is that's causing this epidemic of type II diabetes?
James: Yeah. Wonderful question, Chris. I think that, indeed, you can fault medicine oftentimes for not having that kind of engineering way of thinking. I think there's a lot for medicine to learn from engineers. I think sometimes we're guilty in the biomedical sciences of kind of modeled thinking. For instance, in epidemiological studies of seeing in association then just assuming that we understand what's going on when, in fact, there's lack of cause and effect between the factors that we see associated with one another.
Here's one idea on why I think it's wrong. There's this idea that if we just eat less and exercise more, everything would be fine. And kind of going along with that is this idea that we all became couch potatoes. So, in the 1970s somehow we were all active and then we've stopped walking and we started watching too much television. We started doing too much on our computers. And the thing about that is that there's not evidence that actually people have all a sudden lost their motivation.
So, there really is not evidence to suggest that people have somehow intrinsically changed. But, I think, that one thing that has changed is what we're exposed to in terms of our food supply. Not only are our calories much more accessible through fast food and less meal preparation, readily obtainable, but also the composition of that diet has radically changed with the promotion of low fat diets in the 1980s.
Dozens of thousands of low fat products were introduced that were high in sugar and starch. The thing is, we have to keep in mind is that sugar was something fairly rare in the diet going back several generations ago. In 1800, for instance, the amount of sugar in the typical American diet was about four pounds a year. Whereas now, it's over 100 pounds a year. And so, I think, that more than anything else is what has spurred the epidemic of obesity and type II diabetes, is first the abundance of food, easily obtainable food, and then the removable of fats from that food and the introduction of this kind of low fat low quality foods that are just full of sugar.
Christopher: And you think that's it? Is there an 80-20 principle here and that by addressing those things that you've just already mentioned that's going to be good enough to solve the problem? Or do you think there are other things that are also important? Or what do you think?
James: Yes. So, one way of addressing that, and I would say do we need to get things fully right in order to right the shift? Or are there ways of kind of jumping in and addressing this in a way that has an impact even if we don't have a 100% understanding of all of the aspects that have gotten us where we are? That's really what we've done with Virta.
So, what Virta does is it takes these insights around nutrition and biochemistry that Stephen and Jeff Volek had driven forward. And then what Sami has brought to the table is being able to digitize that and turn that into something that's scalable in terms of technology support. And what we really help people to do with Virta is to change the way that they're eating, to make changes to their nutrition, and then at the same time manage their medical disease and manage their medicines.
And in concert we find that that in itself has a remarkable impact. So, we can actually reverse type II diabetes in the majority of people with that approach. So, even though we're not altering the food landscape and even though we're not getting people to go out and run marathons, we're still able to reverse their type II diabetes.
Christopher: Okay. And so what are the prerequisites for somebody who's interested in signing up for the program? So, for example, and I'm sure many people listening to this podcast they're already subscribed to the idea of a real food perhaps low carbohydrate or low-ish carbohydrate diet and they're fueling their endurance activity in particular appropriately with the right amount of carbs.
Maybe they're thinking about somebody else in their life who could potentially benefit from the Virta program. For me, personally, I'm thinking about my dad who's in his 70s and he is really just stuck in his ways and he still eats a lot of starch, a lot of cereal, a lot of refined carbohydrates. But he is actually very tech savvy. He's very good. He's a former software engineer, former software salesman. So, he's very good with technology. So, could this Virta program be suitable for my dad?
James: Yeah, absolutely. We're helping people of that type of profile. I'll just make one quick mention just so you don't disappoint anyone and that is that currently we have an age cutoff of 70. Sorry about that, Chris. We hope to change that in the near future. But at the moment, we're taking applicants up to 69. We do expect to expand that.
Virta is available both direct to consumer as well as working through employers and health plans. We're available in 46 states, soon to be all 50 states at virtahealth.com. You can sign up online. So, in terms of who is that intended for, really it's intended for anyone with type II diabetes as well as people who have metabolic health challenges, so people with obesity, prediabetes and metabolic syndrome. But we're particularly focused around the type II diabetes because that's an area where people are suffering so much in terms of the consequences of that disease.
What we provide to people really are five things in terms of what our intervention looks like. First is telemedicine. So, it's a provider, health care provider, physician who really understands endocrinology and metabolic disease and knows how to manage diabetes medications. Second, it's 24/7 access of a health coach and those health coaches are nurses and registered dietitians who are trained in areas like nutritional ketosis.
Third is online nutrition and behavior change education. Fourth is biometric feedback. And fifth is an online peer community of others that are going through the Virta program. So, we find that by providing people with that wrap around intervention together with the right nutritional advice that roughly 70% of people will see that they are achieving glycemic control of most or all of their medication.
Christopher: Interesting. And so would you need a formal diagnosis of type II diabetes before you'd be eligible for the program?
James: For direct to consumer, no. We will also work with people who have really any metabolic challenge. So, if people have pre-diabetes, metabolic syndrome, obesity, those are people who want to lose a significant amount of weight, I would say we're not aiming at optimizers who are perhaps athletes who just want to, have already kind of dialed in very, very well and want to take it to the next level. That's not so much our target. We're really providing a medical service. But what we can do extremely well is we can take somebody who's on insulin or on four or five different oral diabetes meds and get them off those medications and on track for feeling a lot better.
Christopher: Okay. Yeah, I just want to -- Sorry to persist with this but is there any particular biomarker that you're looking for in order to say "yes, you're a good fit" because I'm pretty sure that some people they don't get fat before they get diabetes? So, Sami Inkinen would be good example of this. And I would be another good example of this. And actually, my dad is not terribly overweight. And so, unfortunately, some people, they develop insulin resistant type II diabetes before they've had the pleasure, you might say, of overeating terribly. So, do you look for hemoglobin A1C value or maybe triglycerides or fasting glucose or something else?
James: That's absolutely true, Chris. There's a relationship between obesity and type II diabetes but there's not a full correlation. And so there are people who are not obese who do have type II diabetes and vice versa. So, in terms of making the diagnosis of type II diabetes, really there's a couple of ways of doing that. One is the hemoglobin A1C of 6.5% or greater. And that's kind of a measure of glucose over time. And then the other way is to look directly at glucose itself. And if you're getting a fasting reading in the US scale of 126 or higher that's considered also a way of diagnosing someone as being type II diabetic.
So, in terms of folks that we're accepting into the Virta clinic, many people have a formal diabetes diagnosis but not everyone. And we do work with people who don't have a formal diabetes diagnosis but believe that they can use the program to achieve better metabolic health.
The other thing we do is work with employers. And the announced employer that is our first announced employer is Purdue University in West Lafayette, Indiana. And so we're working with hundreds of people there that have type II diabetes. And so in the case of working with an employer or health plan they'll be more formal about their entry criteria. So, they'll say either we want someone to have a type II diabetes diagnosis or we want them to have a certain body mass index as sort of a cutoff for the employer to pay for the program.
Christopher: That's really interesting that you've gone down that route. And it's always been confusing to me as someone who works previously in the UK and then I come to the US and suddenly I was like scratching my head thinking why is it now my employer's job to look out for my health? Do you have any thoughts about that?
James: How we wound up that way in the United States?
Christopher: Yeah. I mean, yeah, exactly. Why is it my employer's job to look out for my metabolic health?
James: I heard a podcast on this topic. I think it was Planet Money or something like that, or maybe EconTalk. But it was back in the 1930s there was legislation put in place that for some reason made it difficult for employers to provide wage increases but they could provide benefit increases as a way of retaining employees. And so there was this incentive to create a great benefits package separate from the actual take home wage. And so employers started to offer health insurance as part of that package. And then it kind of got enshrined from there. And here we are all these years later still delivering health care that way.
Christopher: All right. I can understand it as an employer. They must be thinking, well, okay, if the standard of care is not working for my employees then maybe I need to take matters into my own hands. Who's going to deliver the best solution? And then, of course, they look to you.
James: Yeah. And so that's something we're working on hard right now with employers is that they are, the self-insured employers. Let me make kind of a distinction. For a small company, and I've been an entrepreneur, they will bundle with other small employers and get insurance. And in that case, the insurer, the insurance company is taking the risk of that pool of employee's independence.
But in the case of large corporations like the Fortune 500, most of the Fortune 500 as well as large universities, their populations are large enough that they can actually take on the risk themselves. They can do what's called self-insure. And so they are paying the full cost of that care and are responsible for the entire population. And so there are savings to be had by achieving better health. Those savings accrue to that employer, not to the insurance company. And the insurance company, in that case, becomes more of an administrator or a pass through.
Christopher: Right. And then, of course, if you save money then that passes down to your bottom line and maybe you can afford to pass on those profits to the employee.
James: Sure. Absolutely. And type II diabetes is extremely expensive. And so if you think about it, even before you get into the blindness from renal damage or amputation following peripheral neuropathy or renal failure and dialysis following nephropathy, these sort of awful sequela of the disease, even before then, just in terms of having, maintaining somebody as a type II diabetic cost roughly $10,000 more per year than maintaining somebody who's metabolically healthy. So, that's the possible savings for whoever is paying the bills whether that's government or employer or insurer or the patient themselves. That's the potential savings that's on the table if you can restore someone to metabolic health.
Christopher: And then what are my chances of getting off meds? Let's say I'm listening to this and I'm taking some sort of medication be it insulin or Metformin or some other medicine. If I do the program, what are the chances of me getting off those drugs?
James: Sure. So, one of the things that we're doing right now, Chris, is that we've been conducting a clinical trial. That's one of the things about Virta that we're trying to really do is to achieve the highest sort of standards of evidence that we can to demonstrate the success of the program. About two years ago, we partnered with Indiana University Health in Lafayette, Indiana and began a clinical trial that enrolled 262 people with type II diabetes. And we're now [0:29:58] [Audio Glitch] year of that experience.
We've actually put together all the first year data and we'll be publishing that soon. And so I look forward to telling you the one year results. But what I can share with you now are the early results that we published this spring. And that's on kind of what happens in the first ten weeks? And that's actually when a lot of the action occurs in terms of medication reduction. Because as you get someone into nutritional ketosis and reduce the amount of carbohydrates in their diet, they actually achieve glycemic control relatively rapidly. And so just kind of go through some of the results.
But at a high level, we're discontinuing about two-thirds the prescriptions. So, we're getting rid of 90% of the [0:30:44] [Indiscernible], we're getting rid of 36% of the insulin, reducing another 51% of insulin prescriptions. We're getting rid of just kind of run through other classes of meds, 86% of the SGLT-2 inhibitors, 57% of the DPP-4 inhibitors.
We're usually leaving people on metformin. We generally don't take away right away. That's a fairly benign medication that's often prescribed to prediabetics. So, it's not as if everybody gets off all their meds immediately but most people get off most of their meds within the first ten weeks.
Christopher: Wow. That's quite incredible. Congratulations. That's fantastic. And so amazing that you're publishing this clinical data.
James: Yeah. And we look forward to publishing a lot more. We are obtaining a great deal of results from this trial and so that's kind of one of my roles at Virta with the research team is to continue to publish this data and get this information out there because it's useful not only to Virta but, I think, to everyone who is interested in figuring out ways of treating this disease.
Christopher: Is this not going to lead to the end of Virta Health as a business? Whoever is the government in this country, the government in some other countries say, "Well, I can see what great results these guys are getting. Let's just change the standard of care." And then why would you pay for Virta when that's what you're going to get if you go to local primary care doctor? Do you think there is a scheduled end to the business model for Virta?
James: That's a fun question, Chris. I think that there is so much headway to be made in terms of treating type II diabetes that we would welcome others to achieve some success with this kind of approach. And one of the things about treating type II diabetes that we're trying to do extremely well is not just getting the nutrition right and through using this approach in nutritional ketosis which I can talk more about.
But also what we call continuous remote care. And that is this idea that people have a hard time doing this on their own or doing this, seeing their doctor for two 15-minute visits a year. And so what we do by providing this relationship with a health coach that's ongoing. And the average number of interactions between our health coach and our patients in the early months of the program is two interactions per day where they would be texting back and forth. And quick question and answer or sometimes longer discussions. It's a very high touch model.
And then individualization. So, it's continuous remote care and individualization so that the type of things that you're concentrating on as a patient and the education that you're being provided is all individualized. And if you think about it this way, this has to work not just for someone who's an elite athlete who has this as their hobby and spends a lot of time in this. But this has to work for somebody who is a busy business traveler or a stay at home parent who's cooking for a large family. This has to work in many, many different contexts. So, it's really our job at Virta to individualize and so in a way that can work for as many people as possible.
Christopher: Do you have any specific things that you use? I guess, you already mentioned some things but I just wanted you to go into more detail about how you achieve behavior change. So, using my dad as example, and I realize now it's not a very good example because he's a little bit too old, but I think the primary barrier to entry for him is he's just kind of stuck in his ways. Like this is what I'd been eating for breakfast the last 20 years and I like to have potatoes with my dinner. How do you get someone like that to change their behavior?
James: Sure. I think that one of the things that physicians will often say is that getting somebody to be, reversing their type II diabetes is impossible because if people would just follow our advice everything would be fine. And they'll make this--
Christopher: But we did that and we still got the disease.
James: Right. And so what we thought a lot about at Virta is what is the behavioral feedback loop that actually gets people to achieve success and to maintain it? And so I would mention really three things. And the first is to be very explicit about what we're asking people to do. And so we'll have, for instance a step in our educational materials where you're actually cleaning out your pantry and actually, even before you start making dietary changes, we'll walk you through, okay, here's the exact things that you need to do to make this work and here is the underlying understanding of your condition that the insight as the why you're making the changes that you're making.
And then the second thing then is to actually measure something that matters. And this is something I saw from the sort of quantified self approach, is that if you measure something that isn't very impactful, it's hard to keep doing it.
For instance, if you're measuring steps and you're Fitbit tells you that you've achieved 10,000 steps that day, that's great. But is it really that different than 8,000 steps or 12,000 steps? Are you going to reverse your diabetes at 10,000 steps but not at 8,000 steps? What we find out is that you really do, if you have type II diabetes, you need to be measuring your glucose and you need to understand where that stands.
And then for our program in particular is that we are asking people to achieve nutritional ketosis and, therefore, we need to be measuring ketones. So, we actually measure blood beta-hydroxybutyrate as a biomarker. And so people know every day in the early stages of the program whether or not they've achieved nutritional ketosis. And so that is kind of like a pathway to say am I on track or am I off track?
And then the third thing -- So, one is very explicit direction as to what to do. Second is explicit measuring. And then the third is this amazing feedback that occurs when you start feeling better. People achieve that relatively quickly. The first thing that happens way before weight loss or even glycemic control is that people just start to feel better and they start to have energy.
People will make the comment in the first couple of weeks of, "I woke up this morning and I don't feel like crap for the first time in ten years." And so it's that kind of positive impact on your health that leads people to say, "I'm going to keep doing this."
Christopher: Yeah, it's addictive.
James: You feel healthy again. I think that was what was missing from so much of the medical advice didn't work and the doctors tended to blame the patients saying, "If you would just follow our advice everything would be fine." And people followed the advice and it didn't work and so they stopped. If you give the right advice it's amazing, people will actually follow it because it works.
Christopher: Right. It's very interesting that you should draw on that Fitbit steps example. And I have tried that. It was just horrible for me. It didn't work at all. The thing that I really noticed and especially traveling in Europe, we've been doing a ton of walking today in Lisbon over the last few days, and Lisbon is like San Francisco. It's very up and very down. And so my three-year old daughter doesn't like walking up all these steps and so I've been carrying her.
I know that what I'm doing is working because my back doesn't hurt and my ankles don't hurt. Who cares how many steps I did today? Those things are very powerful motivators to me. I know that if I don't walk enough as what happens my ankle starts to hurt, my back starts to hurt, and so that's really interesting that you've discovered the same thing.
James: That's one of the things that I didn't have type II diabetes but when I started doing nutritional ketosis four years ago, the things that I did notice, one, my blood pressure came down. I had high blood pressure and I no longer do and now I run about 110/70. And then the other thing that happened is that I've been a swimmer and a runner and I had perpetual pain in my shoulder and pain in my hip and that went away.
I measured C-reactive protein as a marker of inflammation and saw that that went down. And so that was my own personal experience. And then as I've gotten deeper into the science and as more and more science has come out in the last few years, we're beginning to understand more about the mechanism of that. So, for instance, beta-hydroxybutyrate which is one of the ketone bodies, is actually an inhibitor of something called the NLRP3 inflammasome which is a driver of inflammation.
So, this is something we generally are seeing in our patients. For instance, we published in the ten-week data that the white blood cell count came down. And we're now looking at other markers of inflammation. But people have published over the past decade and Jeff Volek has done a lot of this work to show that in ketogenic diets markers of inflammation do decline and people describe that to us. They're saying, "My joint pain went away."
Christopher: Yeah. Very, very interesting. I have read those papers and I will, of course, cite them in the show notes. One thing you mentioned there is that you're getting all your patients to track blood levels of beta-hydroxybutyrate. I think that's interesting because blood levels of beta-hydroxybutyrate don't say anything about weight loss, do they?
James: No, they don't. There's so much misinformation out there on the internet. That's another reason among many for Virta to exist is to be an authoritative provider of information as a medical provider. You can find, for instance, on the internet things saying get into ketosis to lose weight. So, here's what's right about that and here's what's wrong about that. What's wrong about that is that being in nutritional ketosis is not a matter of withholding calories.
Now, you can get into nutritional ketosis through fasting and where you are indeed withholding calories. But I myself had been in ketosis for four years with no weight change, with being weight stable. And that's because what you do actually is you increase, well, holding your carbohydrates low and your level of protein moderate you increase the amount of fat in your diet as fuel where you're no longer in caloric deprivation. So, it's easy actually to be weight stable in nutritional ketosis.
Now, the thing that's right about that is that if you do have excess weight to lose and most of our patients are in that situation, and so we published the average weight loss at ten weeks as 7% and at six months is 12% average weight loss. Oftentimes, people are taking off 30 or 40 pounds. And the reason we think that that's happening is that when you have more fat in the diet and you're in a state of nutritional ketosis, that you're reaching satiety with your calories. And so people are naturally, without being asked to restrict calories, because we don't prescribe caloric restriction, people are naturally consuming less.
Christopher: Interesting. What do you do when the program doesn't work? I listened to an incredible interview on the STEM-Talk Podcast episode number 43 that I'll link to in the show notes as well. STEM-Talk is one of our favorite podcasts. It was a fantastic interview with Jeff Volek. The question came up on what you do when it doesn't work? And Jeff kind of avoided the question some more and brushed it off and he hinted that maybe when it doesn't work you're not doing it right.
I hosted the Keto Summit last year and I heard that too there as well. It's like if it's not working it's because you didn't measure ketones or because you're not doing it right. There's carbohydrate in your broccoli or something like that. I don't accept that as an answer. If it's not working it's probably not because the person is not doing it right. So, what do you do when it doesn't work?
James: I would say that with Virta, that that is really on us to -- We don't blame patients. And it is on us to help you make it work. We've taken that actually to the point where our contracts with employers and insurers are value driven. So, we look at the results at a year and if somebody hasn't hit the agreed to parameters we will refund the money. We're that confident of our ability to achieve the goals of glycemic control with medication reduction in most people.
But the thing that we are working hard on is in the percentage of people that have more challenging cases, how can we help them to achieve success? And we will be publishing some additional results coming up from concentrating on folks that are in our sort of lower quartile of performers. And so we'll have a publication coming on how we actually work in an individualized way with those folks.
And this is actually one of the areas that's really great where we can actually use software to support our coaches and physicians. I know you and I had corresponded about this and you were out of town when this happened but we had this machine learning symposium last week in San Francisco where we invited a couple of other health tech companies to join us and talk a little bit about how they use machine learning in delivering health technology.
One of the things we talked about is we actually have a machine learning algorithm that is modeling retention and is modeling hemoglobin A1C and we can tell then when folks are on track or off track and then prioritize those cases for our health coaches to spend more time and more individualized care for those people.
Christopher: That's interesting. How often are you measuring hemoglobin A1C in order to know that somebody is on track?
James: Yeah. What we talked about it at machine learning symposium is we actually, we're not measuring hemoglobin A1C that often but we're measuring glucose frequently. We can then -- There are kind of common published ways of converting glucose into an A1C value but we actually improved upon those algorithms and developed our own algorithms. We're accurately being able to project based on your last ten or 20 glucose values, what is your likely A1C going to be tomorrow or what's it going to be when we test it at four months or at six months?
What that allows us to do is that there's this whole kind of artificial intelligence or machine learning kind of undergirding a lot of the machinery of Virta at the software level. But what that allows is time for our health coaches to be very compassionate with our patients.
Because these are folks that have had diabetes in our trial. The average is about eight years. And we have people that have had the disease on 20 years. We have people who've been on insulin for 20 years and we have people who have been dealing with obesity their whole lives. The burden of that disease both the physical burden but also the psychological burden is tremendous. And so working with people to overcome this long term disease is not something that happens overnight.
Christopher: Interesting. Can you talk about some of the algorithms that you're using to predict the hemoglobin A1C from glucose, for example?
James: Yeah. And so we had two speakers at this machine learning. I think we'll eventually have clips of that event online. So, Nasir Bhanpuri, who's on my team in research, and then Jackie Lee, who's a data scientist in our engineering team. Both spoke. Jackie talked a little bit about how we predict hemoglobin A1C. I'll just kind of share what she shared. That is that we get a pretty good model by just looking at the glucose values. But then we get even better model by looking at about 20 other parameters.
So, we're looking at people's baseline health characteristics as well as their ketone levels and their weight loss. We have people stepping on a scale every day. And so if you improve the model, and I know you do some of this work too, Chris, but if you look at not just one factor but develop a multi-factorial model you can actually be more accurate in the prediction.
Christopher: Right. This sounds very similar. Did you know if it's a random forest or some other sort of decision tree?
James: Oh, boy. You've gone over my head a little bit.
James: No, I could look it up and find out which method that we're using but I'm not sure off the top of my head.
Christopher: Right. Of course. We've done something very similar by getting athletes to do a health assessment questionnaire. So, 51 questions where you click on radio buttons and then at the same time, in our training data the athletes did the blood chemistry, they did urinary organic acids, they did stool testing. We looked at environmental pollutants. And then we use the health assessment questionnaire data to predict the results of the very expensive test that they did. And I'm just wondering whether you're using some similar algorithms to predict the hemoglobin A1C.
And then I also wanted to know. Do you know if they use some sort of neural network, maybe a recurrent neural network to predict the next value in a sequence? That's a very common algorithm used. Am I still going a little bit over your head? I'll just stop talking.
James: Yeah. I'm not sure what all Jackie included in her talk in terms of how deep she went into and what she's up to. But the other thing that she talked about too is that she actually has a PhD in computer science and linguistics from MIT. And so language is actually her main focus. And so what she is doing actually then also is looking at the text streams and conversation streams going on between our health coaches and patients and being able to mind those.
The thing that she talked about was just something as simple as prioritization. Can you with high confidence label a message as high priority? "I ran out of my medicine" or "I have low blood sugar," or something like that. Can you prioritize a message like that as opposed to "I'm doing fine." So, she's been able to develop algorithms that do that and with high confidence will flag messages as priority.
Christopher: Right. Yeah, they call that a sentiment analysis, I think, is what that's called. Of course, if I can't link to in the show notes, what I'll do is if you go to nourishbalancethrive.com/highlights, I will be sure to link the videos to those presentations in our highlights email that will go out at some point in the future. Do you know if you're going to be doing more meet ups like this in San Francisco? Because I would love to attend.
James: It's pretty amazing, actually, that we had -- I forget how many tickets we gave out. Over a hundred. And then we had a waiting list of 70 people. So, there was huge, huge interest in this and more so than we expected. And, yes, so that was the first question that people said was when is the next one? So, we do plan on, I don't know if it will be exactly machine learning as our topic but we do plan on doing additional events like this to bring the health tech community in San Francisco together.
Christopher: Wow. I'd love to get your thoughts on why you think this is so popular. I'd like to speculate--
Christopher: If I could. When you read the mission statement for Virta and it's a tech company in Silicon Valley of which there are many tech companies in Silicon Valley but you're doing something that's truly meaningful. So that's the problem that people like, computer programmers like me, find now is, okay, so this is a company but are they doing anything really very interesting? Is this the next Tinder for dogs or is it a hedge fund that's going to help somebody who's already stinking rich get even richer?
It's very, very hard to find something. Even Facebook. I mean, let's face it. Maybe to some extent the bread and butter for Google is perhaps the same. Is this something that's really going to get me up at bed in the morning? I just wondered whether you had any thoughts about why you're attracting so many engineers?
James: Absolutely. And I will say that if folks are based in San Francisco or based elsewhere and you want to see what the opportunities are at Virta, virtahealth.com/careers, and it has a full listing of our openings which do include data engineers and full stack and back end and several openings that are available right now.
But we've grown the company from six people when I joined two and a half years ago to approaching 100 either later this year or next year. And what people invariably say is that they're driven by really two things when they decide to come to Virta. People would say this when they first joined but more importantly they'll say it after a year as to why am I still here? Number one is the mission. This is the kind of thing that gets you out of bed in the morning.
I actually live in Saint Louis. I'm an adjunct professor of Genetics at Washington University School of Medicine. I actually am on a three week rotation where I'll spend a week in San Francisco, a week at home in Saint Louis and then a week at our clinical trial site in Indiana. I wouldn't be doing that if it were some mission that didn't matter. I wouldn't be doing that for money. So that absolutely is a mission that gets people incredibly excited. The second thing people mention is the team, that we've really built this team of incredibly talented people that have this shared mission.
Christopher: Awesome. That's absolutely fantastic. Are you using machine learning in any way to change behavior? So, I had this, and maybe some people would cringe at this idea because perhaps it could encourage mindless behavior, but the idea that you could have some sort of app. I've got my mobile phone or some other device somewhere in my person and maybe that app, it doesn't even have a display. So, it's a no UI type app.
It would tell me what to do. It would say something like, "Chris, you need to get out this seat and go for a walk. Chris, don't eat that breakfast cereal. Have bacon and eggs instead." Have you looked at anything like that that might lead to behavior change? Or at the moment is it just like finding the anomalies?
James: No. Fair question. I'd look at it a little bit differently in that there are so much health tech is not built for the end user and does not have user experience in mind. It's built for the insurance company or billing or it's built for the physician network. But people oftentimes have not actually tested it very well on the actual patient.
What we've done at Virta really is to build a full stack company that is not only looking at whoever is paying the bills but is looking with a user experience design approach at the end user, at the customer. And that is the patient. So we do have user experience design team and they are working to make our user experience both the app interface as well as everything that touches the patient enjoyable and delightful.
And so there's a -- I would recommend a book by Nir Eyal kind of the -- I think he calls it the hooked method. It's basically on this idea of what makes an app sticky? Why do you return to Twitter or to Facebook? They found something about the user experience and about human nature that makes people want to return to use that interface commonly and frequently. I would say we're not trying to manipulate people's behavior but we are trying to make it a pleasant experience in a way that it makes them want to come back.
Christopher: Yeah. And why not? So, we understand this dopaminergic system quite well. What is it that makes something addictive and why not use it for good use rather than -- A couple of weeks ago Tommy interviewed James Hewitt about the dangers of staying in this cognitive middle gear. So, you see that, especially when you're out and about being in a lot of airports recently, and the number of people you see now just stood in line just looking around them and being completely in the lowest gear is very rare now.
People, they're looking at their phone, they're scrolling through some sort of newsfeed and this cognitive middle gear. James argued that perhaps that was a mistake and that we should take some down time. And then when we really do need to concentrate we have the ability to shift into that high gear. And, yeah, why not use that for something good rather than being slightly evil Instagram, do something that's actually going to lead to some really positive changes in our health? I think that's fantastic.
James: Yeah. We see that stickiness, for instance, within our peer community where we have this hub of users in Lafayette and West Lafayette, Indiana where we had folks that are part of our clinical trial through Indiana University Health and now we have, over the last several months, this influx of new people coming in from Purdue University.
And so the veterans love to tutor the new users. And that is such a healthy thing. I've started to say, "Hey, I've had this degree of success and I want to help others." That's a very healthy kind of stickiness.
Christopher: Awesome. Are you able to talk about how much the program cost in this interview or is it better to direct people to the website?
James: In general, I think we've disclosed this that it's on order of $400 a month and it is not inexpensive but if you think about the recouping on cost, that can largely be paid for in just the savings on medication alone. In many cases we find that many of our patients are spending $5,000 a year or more on their medication. Whether you're paying out of pocket or whether your employer or insurer is paying, the program pays for itself pretty quickly.
Christopher: Right. And is there a fixed commitment? Is there a one-year program or ongoing?
James: At the moment we do ask people to stick with it for a year. I think that that, in terms of people wanting to see the benefits from it, we find that there are different stages to the learning. There are things that you learn in the first few weeks. There are things that takes several months and longer before you really internalize them as part of your daily routine and part of your life. The other thing that that also does is it gets people through the challenges that are kind of cyclic through the year. So, what does a vacation look like doing this? What do the holidays, the winter holidays look like?
Christopher: And does everyone that sign up for the program, will their data eventually appear in the clinical trial? We're going to see a clinical trial with 100 million people as the end value?
James: Yes. The clinical trial that we're doing with Indiana University Health has a couple of different groups but in the hundreds. We talked about the 262 people with type II diabetes that are part of that trial. There are also some people in the trial that we haven't talked about yet that have pre-diabetes and we'll eventually be publishing that data as well.
Then in terms of having data on everyone participating in Virta, I won't go into too much detail about what we're working on, our version of that now, of how can we get that data out there. And so, I think, there is an answer to that and will stay tuned for more because we do want to be able to publish on thousands and eventually tens and hundreds of thousands of people.
Christopher: Interesting and excellent. I will link to virtahealth.com in the show notes, of course. Is there anything else you'd want people to know about?
James: Yes. So, one thing that comes up is this question of -- You were talking about what leads people to succeed or not succeed in nutritional ketosis. And there are a number of myths out there around challenges that people will see that I think they are worth addressing. One is this idea that people will reintroduce carbohydrates because they think carbohydrates are necessary for thyroid function.
And so I point to a blog that Steve Phinney has put on the Virta website addressing thyroid and looking at TSH and T3 and T4 levels. One of the things we do see is that T3 levels do go down when somebody is in nutritional ketosis. But we think that the reason for that is that actually the sensitivity of the thyroid hormone receptor is increased. And so in the same way that the need for insulin diminishes because of improved insulin sensitivity and perhaps the level of leptin may decrease because of decreased leptin sensitivity.
We think there's a good case to be made that the level of thyroid hormone is actually down because you have increased thyroid hormone sensitivity. One indication that that's the case is that if you were actually fighting a low T3 in a way that your body did not like, you would expect TSH, that's upstream of thyroid hormone, would increase. You would try to increase thyroid stimulating hormone in order to drive more production of T3.
We actually see the opposite, right? So that TSH in ketogenic studies -- TSH is actually either flat or down indicating that it's not trying to drive the production of more thyroid hormone. So, I think there are probably ten sort of myths or fallacies out there about things that people think are true. Keto flu is another. There are all these things that are out there that I look forward to kind of taking those on kind of one at a time and explaining to people why they are perhaps not the barriers that people think they are.
Christopher: I did read that article by Dr. Phinney and I will, of course, link to that in the show notes. Is there somewhere else that you would like to direct people to read more about those myths?
James: I think that we've started to put together a number of those at the Virta Health blog. I think that in addition to the thyroid one there are a handful of others up there.
And so that's probably a pretty good place for people to start. We try to, from the Virta Health Blog, to link to other resources and we'll continue to improve that over time.
Christopher: Yeah. We talked about it just a couple of weeks ago about thyroid function and we linked to a rather excellent article on the deiodinases. It seems like it's immensely complicated thyroid signaling. There can be disconcordance between the amount of conversion that's going on in the brain and that that's going on in the periphery and that could also explain the difference in thyroid signaling. I should link to that article as well. I can send that to you afterwards, Jim. I think it's really interesting. Tommy has been talking about that recently.
Yes, super interesting. I'm fully prepared to admit that what the heck do we know? You measure the blood level of the hormone. What the heck have you really just measured? You don't know what's really happening inside of the cell, how many receptors are there? Even if the receptor was bound, was it activated properly? Did anything happen inside of the nucleus? You really don't know that just from measuring the blood levels of the hormones. I'm sure it's immensely--
James: Yeah. And, I think, the other thing to think about is what is the symptomatology? So, Steve also points out in that blog entry that if you look back at the last decade of low carbohydrate high fat nutrition studies and ketosis studies, that there'd been over 300 people, I think, 350 people that have been looked at in that handful of studies that he cites and there are no cited cases of symptomatic hypothyroid. You would think that if it were a big issue you would be, you know. So, oftentimes I think people are reacting to a number as opposed to actually reacting to symptoms.
Christopher: Exactly. Right. Yeah. Don't treat the number. Treat the [1:01:52] [Indiscernible].
James: Exactly. Yeah.
Christopher: Awesome. Well, this has been fantastic, Jim. Thank you so much. I know I've been all over the shot with this interview and I hope you can forgive me for that but you were just so incredibly interesting to me. I just have to explore these different avenues.
James: Well, thank you, Chris, for everything you're doing. This has been a lot of fun for me as well.
Christopher: And I should thank Douglas Hilbert for making the introduction. Douglas is a former client of ours a long time ago and he's now one of your employees and health coaches. I've had the privilege of spending a lot of time with Doug. He is a fantastic guy. I'm really excited that you have such wonderful people.
James: Yeah, we're thrilled to have Doug. Doug and I are part of this small outpost that we have in Saint Louis. So, most of our employees are in San Francisco but we have folks scattered across the country including Portland, Oregon and Denver, Colorado and Lafayette, Indiana and there are a handful of us here in Saint Louis, Missouri.
Christopher: Excellent. Awesome. Well, thank you so much for your time, Jim. I really appreciate it.
James: Thank you, Chris.
[1:02:47] End of Audio