#139 – Our large-scale research study to establish a baseline for health | Dr. Taylor Sittler, Maz Brumand, & Dom D’Agostino, PhD

We launched an observational research study in order to gather CGM data. Data from our members will help us better understand what blood glucose patterns look like in the general wellness population. In this episode, Levels’ Head of Research and Development Dr. Taylor Sittler, Levels’ Head of Business Maz Brumand, and one of the Levels’ advisors Dom D’Agostino, PhD, sit down to discuss:

• The Levels research study and how we will collect data from our members
• What this large study and data collection mean for the future of metabolic health
• How we work to be part of health optimization medicine to get ahead of disease

Follow our podcast, A Whole New Level, for more conversations about metabolic health. We talk about our mission to solve the metabolic health crisis, our fully remote and async startup culture, and the process of building a health and wellness movement from the ground up.

02:36 – The Levels research study

Dr. Taylor introduces the significance of why Levels is researching glucose patterns.

So really, the background and the significance of this study is there really aren’t large studies available. So while we have a lot of data now on how glucose patterns can impact people with diabetes and what those look like, and we’re even learning how to manage diabetes with some of those patterns, right? It can be used to determine what therapeutic doses of insulin, they can be used to raise or lower doses of other oral medications and things like that. But what we really don’t know is what are the patterns of glucose that are present in prediabetic folks and folks who don’t have these diseases? And understanding that can lead to some major breakthroughs if we do that well.

04:04 – A look at the data behind Levels

Dr. Taylor explained that Levels is conducting an observational study to better understand the changes in people’s glucose patterns.

What we’re doing is we’re conducting an observational study. So we’re just collecting information from people who are using Levels and basically from the data that are collected in that process, what we expect to see are changes in glucose, and some people, about 60% to 70% of people that use our application, also log the food that they eat, the exercise that they’re doing. So we can see in time how these glucose patterns are correlated with particular events that they log for us.

05:54 – The importance of baseline and understanding

Dom said baseline data and understanding is important because we do not have a good understanding of what normal glucose is.

Oral glucose tolerance test is going to give you a false response for someone who’s on a low carb diet, because their body is adapted to burning and metabolizing fat for fuel, so they’re not going to tolerate glucose. So I think it’s just, we need to have, in many fields of science, we have this baseline data and understanding. But interestingly, and this was even discussed at the American Diabetes Association Conference that I attended, the field generally agrees that we do not have a good understanding of what normal glucose is and normal glucose response is to dietary patterns too.

07:13 – What is Color?

Dr. Taylor said Color can help describe a lot of genetic variants that people don’t understand. Understanding this may help us understand who is most at risk for cancer and other diseases.

Color was really dedicated to making genetic testing a part of primary care, at least in its early days. And we eventually launched this. Initially, it was 10,000 samples that we ran across a genetically diverse population. And we were able to describe a lot of these variants that people didn’t understand ahead of time. And being able to see what the background frequency is of all these variants, it was extremely powerful to helping us understand which variants would put someone at risk for cancer, what some of the potential mechanisms were, things like that.

09:19 – There is no one-size-fits-all solution

Dom said different diets or dietary patterns may or may not work for people, depending on how their body handles certain foods.

There’s a variety of snips throughout the population that could make a dietary pattern favorable or unfavorable. And we published a review on this in the context of a low carb diet with Jeff Volek and Lucia Aronica. It was published a few months back, and you’re not going to know that if you don’t have CGM data. And then of course, the foods that people are eating to identify this. So occasionally you come across people who are completely utterly intolerant to a high fat diet, a ketogenic diet, which is what we study. And undoubtedly, they have a deficiency in a fatty acid oxidation enzyme, or maybe a lipase enzyme or something like that.

17:18 – Elevated insulin versus insulin resistance

Dom said insulin increases exponentially as we age, and measuring insulin levels can show you whether you’ve developed a resistance to insulin.

My colleague, Dr. Barbara Hansen, has done some work on insulin and how insulin increases exponentially as we age. And you need to pump out more insulin to maintain certain blood glucose levels but that would show up in the glycemic response to a meal would basically show you that insulin resistance. Elevated insulin is always indicative of insulin resistance. I think Ben Bikman would confirm that statement. And you could measure insulin, or you could measure the glycemic response to a meal, which shows insulin is not working. So you’ll glean really informative information from this observational study that just has not been done before.

19:36 – The functional biomarkers

Dom said biomarkers are measurable features of an individual that represent indicators of a disease state or outcome with treatment.

I think it can be part of that overall framework, and I think thinking about it from a low tech perspective, there are I guess maybe what you would call functional biomarkers. If you’re 40 years old, you can do X amount of pushups. And if you could do 20 pushups at 80, you’re going to be in that one percentage of people. And if you take that person that can do 20 pushups at the age of 80, and then go back and look at, if they were to wear a CGM, that CGM profile would look really, really good. He would maintain that level of metabolic health, whereas I mean metabolic health is to a large extent tied to skeletal muscle mass.

23:16 – Health is on a spectrum

Dom said our conventional standards for healthy need to be modified so we can move toward optimization when we’re young.

The study we’re doing now is showing that healthy, or the definition of healthy that’s used now, is not really healthy. I mean, we have subjects, non-diabetic subjects entering the clinical trial we’re running with Dr. Allison Hull and 80% of them have a hepatic steatosis. And that’s reversing quick and correlating very nicely with improved CGM data glycemic responses. So, I think our conventional standards for healthy need to be modified where we can have optimal health and in a spectrum. Health is on a spectrum, and suboptimal health, and then to a pathological condition. And I think we need to move towards optimizing, especially when we’re young and we have better control over that because if we don’t get ahead of that, then if we do get ahead of that, that’s going to pay big dividends in the future.

25:23 – How body weight affects glycemic control

Dom said body weight factors into glycemic control, but they want to do this study to see if changes in body weight affect glycemic control.

I think it’s going to be important to understand if subjects will be reporting their body composition changes, not body composition but just their weight. So there’s a big discussion about if you are closer to your ideal weight, then you have better glycemic control whereas some people can maintain a quite higher body weight and still have proper control. So I think this may add some insight into how important body weight is. If you have normal subjects, how important body weight is and changes over time and how that will reflect glycemic control. And I think that’s sort of an important question that talked about, at least in our class at the med school, how important body weight is and achieving a caloric deficit to improve insulin sensitivity.

Dom D’Agostino: (00:06) If you have a dashboard in a car, and if your body is an instrument and you need a dashboard, the glucose level is probably the most important instrument to look at for your fuel. And I think knowing how we use that fuel, if we have excess surplus calories, that’s going to produce a very predictable rise in glucose and persistent elevation of glucose and insulin that’s going to impact our overall longevity in health over time. And we would not know that if we didn’t have a CGM device.

Ben Grynol: (00:45)

I’m Ben Grynol, part of the early startup team here at Levels. We’re building tech that helps people to understand their metabolic health. And this is your front row seat to everything we do. This is A Whole New Level.

Ben Grynol: (01:11)

One of the coolest things that has been a byproduct of the number of members using Levels is the amount of data, granted it’s anonymized, but the amount of data that has been aggregated outside of the type one and type two community. And so Dr. Taylor Sittler, head of research and development, Maz Bruman, head of business, Dom D’Agostino, one of our medical advisors, the three of them sat down and discussed this idea of research. What we can do with all of this data, why it’s important in general outside of the type one and type two community and what it means to see glucose data and how different food affects people’s health at different ages, different genders, different ethnicities, and different stages of life?

Ben Grynol: (01:56)

What are the different dietary patterns as they pertain to metabolic response in the general population? It’s still something that we are looking at on a deeper level, but it’s very cool to think about what the implications for this data are and how we can think about it moving forward. What are some of the insights that we can start to glean when we dig into this data?

Ben Grynol: (02:17)

It’s very cool to think about all the possibilities moving forward as we scale and as more people start to use CGM as part of their health wellness routines. It’s a great conversation. Here’s where they kick things off.

Dr. Taylor Sittler: (02:35)

So really the background and the significance of this study is there really aren’t large studies available. So while we have a lot of data now on how glucose patterns can impact people with diabetes and what those look like, and we’re even learning how to manage diabetes with some of those patterns, right?

Dr. Taylor Sittler: (02:54)

It can be used to determine what therapeutic doses of insulin, they can be used to raise or lower doses of other oral medications and things like that. But what we really don’t know is what are the patterns of glucose that are present in prediabetic folks and folks who don’t have these diseases? And understanding that can lead to some major breakthroughs if we do that well.

Dom D’Agostino: (03:19)

Yeah, someone could have normal glucose, but their postprandial response could be way out of line. I mean, their insulin levels could be 10 times higher. They could be pumping out 10 times more insulin to maintain that level of glucose. And that would show up in the CGM data as you know, because it had been paired insulin resistance. So you’d see the bigger postprandial excursion in glucose.

Dom D’Agostino: (03:43)

And yeah, and the biggest thing my colleagues harp on is that unfortunately, we just do not have data on non-diabetic subjects on what normal glucose is and normal glucose response. So that’s why this is super important, this study.

Dr. Taylor Sittler: (04:00)

Awesome. Let me provide a little bit more detail there. So what we’re doing is we’re conducting an observational study. So we’re just collecting information from people who are using Levels and basically from the data that are collected in that process, what we expect to see are changes in glucose and with some people, about 60% to 70% of people that use our application also log the food that they eat, the exercise that they’re doing.

Dr. Taylor Sittler: (04:28)

So we can see in time how these glucose patterns are correlated with particular events that they log for us. What do you think would be the biggest impact of this kind of data on people? If we’re able to see that, say, in 50,000 or a hundred thousand people, what do you anticipate will be the results of that or what patterns you think will emerge?

Dom D’Agostino: (04:53)

Yeah, well, in addition to establishing what normal glucose is, assuming the fasting level being under 126 milligrams per deciliter and things like that, then we’ll start to get some understanding of, and I think this is super important for the use of CGMs, the personalized response that people may have to different types of food.

Dom D’Agostino: (05:18)

And especially, if this is coupled with other things like changes in weight, or perhaps a particular type of diet that they’re following, you could start to garner insight into how different people’s metabolisms are, and that we need to acknowledge that we can’t apply a cookie cutter type diet to everybody and expect them to maintain glycemic control, where some people may be much more carb tolerant than others.

Dom D’Agostino: (05:49)

And there’s a lot of caveats to using other means to gain information on this, like oral glucose tolerance test is going to give you a false response for someone who’s on a low carb diet, whereas because their body is adapted to burning and metabolizing fat for fuel, so they’re not going to tolerate glucose.

Dom D’Agostino: (06:09)

So I think it’s just, we need to have, in many fields of science, we have this baseline data and understanding. But interestingly, and this was even discussed at the American Diabetes Association Conference that I attended, the field generally agrees that we do not have a good understanding of what normal glucose is and normal glucose response is to dietary patterns too.

Dr. Taylor Sittler: (06:34)

No, it’s interesting. There are some parallels for me to some of my previous work, which was in genetics. When we started Color Genomics back in 2013, 2014, there really wasn’t any baseline data on the population in specifically around the genes that put you at risk for cancer.

Dr. Taylor Sittler: (06:54)

And so a lot of people would get test results back with these variants that were unclear. People weren’t sure. Oh, this could be … A medical geneticist might postulate that, “Oh, this could actually be very risky for somebody to get cancer.” And another might say no.

Dr. Taylor Sittler: (07:11)

And what we found was after, so Color was really dedicated to making genetic testing a part of primary care, at least in its early days. And we eventually launched this. Initially, it was 10,000 samples that we ran across a genetically diverse population. And we were able to describe a lot of these variants that people didn’t understand ahead of time. And being able to see what the background frequency is of all these variants, it was extremely powerful to helping us understand which variants would put someone at risk for cancer, what some of the potential mechanisms were, things like that.

Dr. Taylor Sittler: (07:48)

And so I think whenever you have a technology that’s grown up in the context of a particular disease, you have this problem. In the case of CGM, it’s always been used for monitoring diabetes. And when you broaden that, there can be so much more power to it, but you really need to be able to understand what the baseline looks like before you can start attaching any kind of meaning to the patterns to your point, to postprandial patterns for instance, or even understanding what baseline fasting glucose looks like for people overnight, or waking glucose in the morning.

Dr. Taylor Sittler: (08:26)

I mean, for a lot of people, glucose will jump up a little bit in the morning once they get up. And what does that look like? And to me, I think to everybody on this call, the value is super clear. But I think if you haven’t been involved in one of these projects, it’s hard to understand why we want to measure it on 50,000 people.

Maz Bruman: (08:45)

About summarize what you guys are saying specifically, we don’t know what’s normal. What’s the baseline of what’s normal, especially when you think about different genders, different ethnicities and understanding that is the place to start.

Maz Bruman: (09:02)

When I see this rise, when people wake up, is that normal or is their metabolic system is dysfunctional? And that seems to be more to start and then build on that to figure out then what is abnormal and how do we think about that in terms of the future?

Dr. Taylor Sittler: (09:17)

Yeah, a hundred percent.

Dom D’Agostino: (09:19)

There’s a variety of snips throughout the population that could make a dietary pattern favorable or unfavorable. And we published a review on this in the context of a low carb diet with Jeff Volek and Lucia Aronica. It was published a few months back, and you’re not going to know that if you don’t have CGM data.

Dom D’Agostino: (09:40)

And then of course, the foods that people are eating to identify this. So occasionally you come across people who are completely utterly intolerant to a high fat diet, a ketogenic diet, which is what we study. And undoubtedly, they have a deficiency in a fatty acid oxidation enzyme, or maybe a lipase enzyme or something like that.

Dom D’Agostino: (10:00)

But that would probably show up with genetic testing, but not everybody is going to do that. But a CGM would really give you insight, especially when it’s coupled with other biomarkers levels is doing especially insulin. And we don’t have any really good standardized reference ranges for insulin. And it’s not even part of the metabolic panel, which is amazing.

Dom D’Agostino: (10:23)

And I know if we go into the direction of get CGM data and couple it with other biomarkers, which you’re doing it, then it becomes exponentially more informative.

Dr. Taylor Sittler: (10:33)

Yeah, no, it’s really exciting. I’m also excited to see once we do start to have this data, how we’re able to get it out to folks, because I think as Color was able to do, eventually it was once you can publish that data and give the whole field a reference range, it really helps a bunch of other folks, a bunch of other investigators be able to then look into specific patterns. Once they start to understand, to Maz’s point earlier, what’s normal and what’s abnormal.

Dr. Taylor Sittler: (11:01)

So, it’s going to be really exciting not just to collect this data, but to then to be able to collate it and get it out there for people to be able to understand and make future contributions to it.

Maz Bruman: (11:12)

I think where we are today is the data on the healthy people is very scant. And first thing we need to do is understand what is the actual baseline for people that are health, right? But can you talk through, what is the arc of this? If we fast forward five years, how has this research contribute to end goal, which obviously want to help people improve or maintain their metabolic health?

Maz Bruman: (11:35)

Can you talk through a little bit of that, Taylor or Dom, how does this create the stepping stone to get to where we all want go, which is to really help people improve their metabolic health?

Dom D’Agostino: (11:46)

Yeah, well, I think as we collect data and publish it, then it becomes a framework because CGMs will be more accessible for people. And when they’re using CGMs, they will have an existing framework to titrate and adjust their diets, whether it’s the macronutrient composition or the amount that they’re eating per meal to stay within a particular reference range that the field generally speaking, in metabolism would agree that a postprandial excursion to 120 is much more favorable than 220.

Dom D’Agostino: (12:22)

And you’re not going to know that if you’re not using a CGM and what a CGM and Levels software do is that it makes it actionable. So having metabolic awareness, then the software then becomes a behavioral tool to modify the types of food you’re eating, the amount you’re eating to optimize within reference ranges that may take some time to get people in metabolism and endocrinology to agree on what’s optimal.

Dom D’Agostino: (12:52)

But I think generally speaking, one could say lower is better excluding hypoglycemic ranges. Pretty much every endocrinologist or metabolic expert is going to agree that a lower average glucose and a lower postprandial glucose is going to be better and more indicative of metabolic health.

Maz Bruman: (13:14)

One of the things we’re trying to do at Levels is try to not only tell people things that work, like for example, an excursion is bad, but also things that people can actually do and adhere to in a way that it’s easy to do and it’s sustainable to do.

Maz Bruman: (13:29)

For example, it may be easier for somebody to keep the area on the curve small. And if research proves that that’s the thing that we should go tell people to keep the area on the curve small versus try to avoid spikes. Do you think some of this research will actually not only help us understand, like where’s the ranges that we’re asking people to do? Because it comes with cost, right?

Maz Bruman: (13:50)

When you tell people to keep your glucose within this reference range, they are thinking of it as I have to give up my ice cream at night, or I have to give up the morning breakfast that I enjoyed with orange juice. And so the question that we try to ask ourselves is what is the lowest burden thing that we can tell people to do that they will continue to engage with and will have the highest efficacy?

Maz Bruman: (14:14)

And really this research help us to get there. So we ask people to do the things that’s their minimum that they have to do, or the effort that they have to do in order to achieve most effective outcome.

Dr. Taylor Sittler: (14:25)

To sort of talk through that arc, we’ll start with these very large trials where we’re just collecting observational data. Over time, that allows us to start to publish this baseline and some of these reference ranges in these different variables, right? So we probably won’t publish the whole traces, but we’ll publish what the windows, the normal fasting windows might be, or the normal windows of postprandial response might be.

Dr. Taylor Sittler: (14:50)

Or looking at different patterns, we’ll understand what those normal windows are, and hopefully to some extent how they correlate with some genetic markers. Because I think what I’ve certainly seen in other situations is that people are different, right? There’s a range that you need to get and understanding what group you belong to is super helpful in figuring out what those ranges are. We’re not all the same.

Dr. Taylor Sittler: (15:13)

So I would see us basically collecting this data, establishing some of those variables and ranges. And once that starts to get out into the public domain, that enables a bunch of researchers to use that data, refine it, and really establish for the field what those reference ranges are. Simultaneously then, we can start to internalize some of that and start to tell our members, “Hey, this looks abnormal, this looks abnormal. You may want to change this. You may want to fix that.”

Dr. Taylor Sittler: (15:43)

Once those baseline numbers start to be agreed upon, then we can really use that to help people modify their behaviors to improve their health. And I think hopefully within the first couple of years, we’ll be able to start doing that. And then after that, I think once we get some data back on which of those interventions is effective based on these variables that we’ve established, then we can start to say with some certainty what’s working therapeutically.

Dr. Taylor Sittler: (16:12)

So what I see in the more distant future based on what I heard from you two is there might actually be almost approved therapies where we can say this diet is really going to help you out or this fasting regimen or this exercise regimen, or looking at these particular things could be really helpful for you.

Dom D’Agostino: (16:33)

Yeah, that makes great sense. So instead of treating a disease is what modern medicine is, but health optimization medicine, if you want to call it that, can get ahead of the problem and start to prevent the problem from occurring in the first place. So if we know what healthy is, and we can start, as people are on a trend towards getting to be prediabetic and maybe understanding the CGM curve, maybe they are in the context of the postprandial response, like approaching levels.

Dom D’Agostino: (17:08)

And I think maybe this study could further define what it means to have prediabetes or diabetes. Like I said, my colleague, Dr. Barbara Hansen, has done some work on insulin and how insulin increases exponentially as we age. And you need to pump out more insulin to maintain a certain blood glucose levels but that would show up in the glycemic response to a meal would basically show you that insulin resistance.

Dom D’Agostino: (17:41)

So, elevated insulin is always indicative of insulin resistance. I think Ben Bikman would confirm that statement. And you could measure insulin, or you could measure the glycemic response to a meal, which shows insulin is not working. So you’ll glean really informative information from this observational study that just has not been done before.

Maz Bruman: (18:07)

One thing you just said that really resonates I think with us at Levels, which is what is this progression of health? So we start in our twenties and we’re super healthy and we’re super resilient. We can just take on anything and feel invincible. And as we kind of progress into our middle ages, it’s kind of you’re not sick yet, but you just don’t feel as much as good as you did before. You probably can’t endure as many hours of intense, whatever you’re doing in that progression. At some point, we categorize it as disease if there’s dysfunction happening in our physiology.

Maz Bruman: (18:44)

And I guess when you look at what you said that triggered, the thought is the baseline is, hey, anybody that’s in his 20, maybe this is what their glucose response looks like. That also doesn’t have a disease. When you turn from that 20 healthy person to still a healthy 40-year-old, there is some degradation that’s happening naturally. And then obviously as you go from that healthy 40-year-old to maybe somebody that has disease 50-year-old, there’s again another change.

Maz Bruman: (19:11)

But what does this spectrum actually look like? And can we identify, A, this is happening. It’s not your imagination, we’re getting old. And B, one of the things we can do to maybe slow that down so that you don’t either progress into that disease state or slow down that provision to the extent. Is that something that this research can at least set the groundwork to achieve?

Dom D’Agostino: (19:36)

I think it can be part of that overall framework and I think thinking about it from a low tech perspective, there are I guess maybe what you would call functional biomarkers. If you’re 40 years old, you can do X amount of pushups. And if you could do 20 pushups at 80, you’re going to be in that one percentage of people. And if you take that person that can do 20 pushups at the age of 80, and then go back and look at, if they were to wear a CGM, that CGM profile would look really, really good. He would maintain that level of metabolic health, whereas I mean metabolic health is to a large extent tied to skeletal muscle mass.

Dom D’Agostino: (20:20)

So, one low tech way could be to focus on optimizing that and then to pay attention. And then the CGM is a level of, there are blood-born biomarkers that can inform you to optimize your function, whether that’d be endurance activity or strength training or whatever your sport is. But I look at it at various ways. There’s like functional biomarkers. And then there’s physiological things like blood pressure. And I think you have to look at the complete picture, but the continuous glucose monitoring is probably the best.

Dom D’Agostino: (21:03)

If you have a dashboard in a car, it’s probably going to be, and if your body is sort of an instrument and you need a dashboard, the glucose level is probably the most important instrument to look at for your fuel flow. And I think knowing how we use that fuel, if we have excess surplus calories, that’s going to produce a very predictable rise in glucose and persistent elevation of glucose and insulin that’s going to impact our overall longevity in health over time. And we would not know that if we didn’t have a CGM device.

Dr. Taylor Sittler: (21:40)

Yeah. What I love about this discussion is that it takes the traditional approach to doing clinical studies and really flips it on its head. We’re talking about establishing a baseline for health rather than a baseline for disease. And I think that’s really what’s so powerful. We’re moving into an era where we’re going to be able to define what it means to be healthy rather than what it means to be sick. And then we can start to optimize toward that.

Dr. Taylor Sittler: (22:07)

And if you think about it from whether it’s from a statistical and machine learning perspective, or even from a basic clinical perspective, if you’ve got a single goal that you’re going toward that being health, it’s much easier than optimizing away from diabetes, away from cardiovascular disease, away from Alzheimer’s which is unfortunately what every primary care doc has to do in the 15 minutes that they have with patients who come in.

Dr. Taylor Sittler: (22:31)

So that’s what I think is so powerful. And that really the way you described it, Dom, really triggered that for me, was this, if we do our job correctly with this observational study, we’re helping to develop, as you mentioned, that bigger framework for us to be able to really understand what it means to be healthy. And once we understand that, then we can start identifying patterns that will lead back to health.

Dr. Taylor Sittler: (22:56)

And I think that’s a new way of doing medicine. It’s a new way of optimizing for longevity and for optimizing for prevention of disease. So I’m pretty excited about it. And we’re lucky to have folks like you helping us out because you’ve been building this framework for 10 years already, or 20 years.

Dom D’Agostino: (23:15)

Yeah. I mean, the study we’re doing now is showing that healthy, or the definition of healthy that’s used now, is not really healthy. I mean, we have subjects, non-diabetic subjects entering the clinical trial we’re running with Dr. Allison Hull and 80% of them have a hepatic steatosis. And that’s reversing quick and correlating very nicely with improved CGM data glycemic responses.

Dom D’Agostino: (23:42)

So, I think our conventional standards for healthy need to be modified where we can have optimal health and in a spectrum. Health is on a spectrum, and suboptimal health, and then to a pathological condition. And I think we need to move towards optimizing, especially when we’re young and we have better control over that because if we don’t get ahead of that, then if we do get ahead of that, that’s going to pay big dividends in the future.

Dom D’Agostino: (24:12)

And then I think that’s the major contribution that this study can have is sort of defining what’s normal and what can be optimal. And we just don’t have that data now.

Maz Bruman: (24:22)

So it seems like the problem starts from measurement because it’s very hard to know where you are on that spectrum. Before, it’s something that you can measure, which is like obviously when you’re sick, you’re in the hospital, you measure at that point. But before that point, it’s really hard to measure. This step is starting to measure that progression of super healthy to kind of healthy to, “Oh boy, you’re about to hit that brick wall,” which is the disease.

Dr. Taylor Sittler: (24:47)

Yeah. And if you’re coming in with fatty liver disease, you’re close. It’s not far away anymore.

Dom D’Agostino: (24:55)

That’s like a silent killer and it could just probably from, Dr. Robert Lustig is saying from the excess sugar and the fructose and things like that, that probably show up in a nice CGM profile. But these things creep up on you in ways that are kind of sneaky. So, it’s super important to get ahead of that.

Dr. Taylor Sittler: (25:17)

Awesome. Well, anything else you all feel like we should talk about before we wrap up here?

Dom D’Agostino: (25:23)

I think it’s going to be important to understand if subjects will be reporting their body composition changes, not body composition but just their weight. So there’s a big discussion about if you are closer to your ideal weight, then you have better glycemic control whereas some people can maintain a quite higher body weight and still have proper control.

Dom D’Agostino: (25:49)

So I think this may add some insight into how important body weight is. If you have normal subjects, how important body weight is and changes over time and how that will reflect glycemic control. And I think that’s sort of an important question that talked about, at least in our class at the med school, how important body weight is and achieving a caloric deficit to improve insulin sensitivity.

Dr. Taylor Sittler: (26:15)

Well, so the study is observational. So we’re really just putting whatever data people are offering to us. As part of being a Levels member, you can connect your account to your Apple health data. And some folks are collecting their weight and measuring it and keeping it in there. And for those folks who are willing to share, I think we will learn a lot.

Dr. Taylor Sittler: (26:38)

But it’s also the nature of this observational study that it’s not completely consistent, so we’re not going to get everything from everyone. And we wouldn’t demand that. I think that would be too difficult. But from those that are willing to share, I think we’ll be able to hopefully learn a lot. And I think what this will definitely do is enable us to pose hypotheses for future studies that are really helpful.

Dr. Taylor Sittler: (27:03)

So it may be that that’s the second Levels study that we do next year based on some of the data that we’re collecting and that we understand that better deadline.

Maz Bruman: (27:15)

And I think it speaks to why this study is large, because one, I think we designed the study to be a minimal burden on people so we can get more people to participate. And so for that, we decided to make the observation so that you’re doing what you’re doing, we’re just collecting.

Maz Bruman: (27:26)

And I think the size of it will speak to the diversity, right? And the personalization that we’ve been talking about by being this large, fully we can capture a lot of different types of people, the way they live, whether they come from different backgrounds, whether they come from different genders. And I think the size and the low friction kind have to go hand in hand for it to work. You can’t have one or the other.

Maz Bruman: (27:51)

And so I think the exciting thing about this is we will get a lot of information about a lot of things that can say, “Okay, this is really interesting.” I think what you, Dom, said is the correlation between body weight, glycemic response or metabolic function. Is there a one-to-one correlation or it’s a one-to-one correlation only if you have these things, like if you are this age or if you are this genetic background?

Maz Bruman: (28:16)

So I think it’s really exciting to see all those and then design personalized intervention based on those confounders that can actually work for people, because we hear it all the time. It’s like, “Hey, why does this work for so and so and doesn’t work for me?” And hopefully, we can have enough information and say, “The reason it works for so and so is because of X and the reason it doesn’t work for you is Y. And now we have something that actually will work for you, but it will be different.” So, pretty excited about that.

Dom D’Agostino: (28:41)

That’s a great point. Yeah. I mean, there’s huge power in numbers, I mean literally, statistical power when you have such a robust and large data set that you can glean a lot of insight into that.

Maz Bruman: (28:55)

They always get the saying, it’s like, “Hey, why is my friend eating really terribly, but looks great? And I’m doing everything right and I’m overweight?” It’s like, “Well, it doesn’t mean like there’s so many layers into that that you need to unpack,” to answer that question versus saying, “Okay, it’s okay to eat terribly because it worked for somebody else or at least it appeared to work for somebody else.”

Dr. Taylor Sittler: (29:17)

Yeah. I mean, I think one thing we can definitively say at this point is it’s not simple.

Dom D’Agostino: (29:22)

And the best way to go about it is just to cast a wide net and collect this observational data and then figure out what the important testable hypotheses are after getting this observational data, which will be really the first of its kind. So, it’s super exciting at Levels of spearheading this. And I’m just excited to be part of it.