163: What Your Glucose Monitor Is Missing

Your body may be making extra insulin just to keep your blood sugar in the normal range, and a glucose monitor won't tell you how much insulin that takes. That's why I'm interested in a new sensor that tracks glucose and ketones continuously. The FDA recently authorized Abbott's Libre Duo 10-day for people with diabetes. Here's why those two signals matter: Insulin acts like a brake on ketone production. As insulin falls during fasting, that brake eases up, allowing your liver to make. make more ketones. Watching those changes alongside glucose could help us better understand how the body responds to meals and fasting. The sensor doesn't measure insulin directly, but it adds information that glucose alone cannot provide. For people with diabetes, that information can also help flag rising ketones before a dangerous complication develops. This is Lecture 163 of The Metabolic Classroom. Welcome back to the Metabolic Classroom. I'm Ben Bikman, metabolic scientist and professor of cell biology. Last month, the FDA authorized the first wearable in the United States that continuously measures ketones and the first device anywhere that measures glucose and ketones together in a single sensor. I think that this is a good opportunity to look beyond what we can get here just beyond the numbers. It's, it's more than just more data. There's some, I think, valuable insight we can learn about watching both fuels at the same time. It gives us a much better real-time understanding of fuel use and what's happening to insulin, albeit indirectly. All right, to get started, let me just tell you what the device is. It's Abbott's Libre Duo 10-day continuous dual glucose-ketone monitoring system. It is the first system to meet both the existing integrated continuous glucose monitoring standard and a new integrated continuous glucose and ketone standard. The labeled population for using this, not surprising, is people with diabetes. But of course, like most of you, I'm interested in its use beyond diabetes. The sensor itself works like other continuous monitors do, which is that it samples the interstitial fluid. We'll get into that more in a moment, but it measures it every minute, and it does so for 10 straight days, and then it sends that data, those readings, to your phone app. And then within that app— and I'm sure this is going to iterate and improve over time— but it allows you to see certain trends. It can also send you certain alerts based on how high or low some of these variables may getting. The review that was used to approve this was based on 6 clinical studies, and the agency reported that the device tracked clinically meaningful differences in ketone levels over that 10-day period, including in certain people identifying when ketones were getting too high and warning of the onset of ketoacidosis in people with diabetes. Now, as I noted, the intended users are people with type 1 or type 2 diabetes who take insulin, Certainly if you're type 1, that's needed. In the case of type 2, I would suggest that's not so needed as an aside. But still, there are other glucose-lowering drugs that a type 2 diabetic might take, particularly SGLT2 inhibitors, that are known to dramatically increase ketogenesis. And thus, in both of these instances, there is a risk of ketoacidosis, even if the glucose might not be warning the person of such a trend. All right, that's the population that this is designed for because it's the population for whom an unexpected ketone rise is the most well expected. It's the most relevant in that group. Now there have been, of course, some previous. There are some pre some other existing ketone measuring tools, and I want to review those briefly just so that we can appreciate the value of the continuous monitor because. It, it's these things, these different methods of measuring ketones or measuring different molecules, and they're not always measuring the one that matters most. Speaking of the molecules that matter most, you will remember, I'm sure, that there are 3 ketone bodies. Now, I don't like using the term ketone bodies, so I'm just— we can just call them ketones. But there are a lot of ketones that exist within the body, within the realm of biochemistry, that have nothing to do with fuel use. So usually the term ketone bodies is referring to the 3 ketones that are a part of this ketogenic process in the liver. The first one, beta-hydroxybutyrate, or BHB, that is the dominant circulating ketone that we have in the blood, and it is what we are elevating most significantly when we're talking about ketosis or ketoacidosis. And this is what we measure with the blood strips. When you prick your finger and you put that drop of blood on a little strip in a ketone meter, that's what you're measuring. It is also what the new continuous sensors measure, not from the blood but from the interstitial fluid. As I said a moment ago, more on that in a moment. Acetoacetate is the kind of firstborn ketone, but it's at much lower levels in the blood than BHB is. This is what you measure in the Classic urine strips. So you get those cheap little strips, you pee on them, and that's what it's telling you. That's the ketone it's measuring. And then lastly, you may be familiar with the fact that you can measure your ketones in your breath. That measures acetone, the volatile leftover, the, the sibling here. That's what's coming out in the breath. All right. So classically, the blood BHB strips are the best available snapshot. Since that little drop of capillary blood can give you a number of BHB, an amount in, or a concentration in your blood in a good range in millimoles per liter. That's the, that's the, the, the formula we want, or the units that we want, millimoles per liter. And so then you— I use the word snapshot deliberately because that tells you right then in that moment where you're at. But the problem here is that that snapshot only exists in the person who— or the snapshot is only known in the person who has the glucose strip, has the finger lance to, to poke their finger, and then they have the glucose meter ready to go and they remember to actually do it. So as valuable as that BHB measurement is from the blood, the main hurdle is just actually getting that number. Especially in the case of, say, a diabetic who may not know how high they're getting with their ketones, and they might not have brought their meter along with them. You can't act on a number if you don't know it. Now, the urine strips are a problem for a different reason, just largely because they measure the wrong ketone when it comes to early detection of someone who might be trending towards ketoacidosis. When a person's developing ketoacidosis, it's BHB that is dominating this process. So the strip that measures acetoacetate is kind of lagging behind the measurement, the BHB that you actually want to get. But then, as the ketoacidosis is treated or resolved and BHB is converted back toward acetoacetate, which is the biochemistry here, the urine strip can stay dark. Because it's measuring acetoacetate, even though the one ketone of concern, BHB, is actually falling. So you can miss the rise and then you end up kind of overestimating the, the recovery. Now lastly, breath acetone is appealing because it's totally non-invasive. It's the only one. Well, it, it's the most convenient, let's say. You don't have to take it into the bathroom with you. And the— there is a good correlation of breath acetone with capillary BHB, but the timing once again can be a bit off. In one documented episode of, of, of just ketosis, so not ketoacidosis, in a person with type 1 diabetes, blood BHB normalized more than 12 hours before breath acetone did. So if you're chasing the breath acetone numbers, if you're measuring that one, you can believe that you're still in trouble. Long after the blood ketone has settled. Now, none of these tools shows a rise or a fall as it happens unless you're, you know, really trying to track it in the moment. But none of them can monitor it all, all the time in the background. And, you know, for example, could wake you up at 2:00 AM if there is a problem going on. They all require a deliberate decision to test. Often, unfortunately, at least in someone with diabetes who needs to be paying attention to this, it may— they might not do so until they already are starting to feel symptoms of it. So they already feel sick, and so they've missed an opportunity, a window of time where they could act before things started to get complicated. Now, we already know what happens when a metabolic signal moves from intermittent to continuous. Because that experiment, those observations have been done with the, with glucose and the use of continuous glucose monitors in people with diabetes. In adults with type 1 diabetes on, uh, who are of course, uh, needing to have frequent or daily, multiple times a day insulin injections, continuous glucose monitoring lowered the A1C significantly more than those who were monitoring their glucose with just finger sticks. Over a multiple-month period of time. And they also had significantly greater improvements in time in range. So when you think about the glycemic variability, if they were able to continuously monitor it, what was big variability swings up, swings down and up and down, hyperglycemia, hypoglycemia. With the use of the CGM, they were able to tighten that a lot. No surprise, right? When you're getting real-time measurements, you can get real-time intervention separately in adults with type 2 diabetes who were taking insulin as a primary point of care or intervention of care, uh, when they were continuously monitoring glucose. Once again, A1C improved. Once again, time in range improved. In fact, by a lot. Um, it went up almost like 60%, um, over the— in those who were not measuring it in real time. So the value of the continuous data. More than the number itself is the trend that you're seeing. And, and that is— it's just, that's not really possible when you have to measure the marker only when you remember to do it. Okay, now I want to take a moment to step away from the, the, the direct use of the— these continuous monitors in disease states and just use it as an opportunity to review some of the metabolic processes here, the, the, the metabolic milieu in which these things can start to make some more sense. I think that it's helpful for people who are using these sensors, um, to understand what the ketone number is reporting and what it suggests is happening with insulin, because as you know, insulin is the principal inhibitor on ketone production. And it Does so. This is where I'm reviewing some of the biochemistry with you at 3 important points. Now, no surprise, I, I want to discuss this regulation because it's further evidence of how critical insulin is to metabolic regulation in the body and at the level of the cell. Remember, remember, when it comes to metabolism, nothing matches the relevance of insulin. It is the master metabolic hormone. All right. That first point of 3 that I said, the first is at the fat cell. Insulin suppresses the enzyme that breaks down stored triglycerides, the stored form of fat, into free fatty acids. That is the released form of fat that's coming from the fat cell as it's breaking fat down. When insulin is elevated, and even elevated modestly, this is something that this fat cell responds to very, very readily. Insulin goes up, free fatty acids come down. There are fewer fatty acids coming out of that fat cell, so fewer of those fatty acids get to the liver. as substrate or building blocks, because it's when liver— when insulin is down and the fats are coming from the fat cell coming to the liver, that low insulin allows the liver to burn those fats very readily. So readily, in fact, that it's burning fat to the point that that is basically being converted into ketones. Remember, Ketones are essentially just products of fat burning when there's a lot of fat burning going on. And when insulin falls, all of the inhibition of those processes go down. So the processes are disinhibited. In other words, they're running. I don't want to use a double negative there. They're, they're, it's, it's ramped up to an 11 out of 10. All right. Now the second point here is what happens at the mitochondrial membrane inside the liver cell. I actually kind of touched on this in the first point. So the first point was intended to just be that the fat cell is able to release its fats. Now we get to the mitochondria within the liver because here we have insulin will block— insulin blocking the transporter that will carry the fats into the mitochondria. If you block that step and fat cannot enter the mitochondria, in that case you cannot activate ketogenesis no matter how much of The, the amount, no matter how much of the fat may be sitting outside the mitochondria. Lower insulin. Once insulin is lowered, insulin comes down, and then that transporter process can occur. Now the mitochondria are pulling in fats very readily and happily burning them. Okay. The 3rd and final step is, is this committed step of the ketogenesis or the ketone synthesis itself. This happens at the point where the liver directs acetyl-CoA, which is the, the entry product into the, into the mitochondria, and it shifts it out toward ketone production. When you get to acetyl-CoA is an interesting molecule because it kind of exists at a metabolic crossroads. It can go a lot of different directions, and insulin largely determines which direction it goes. If insulin is up, Then the acetyl-CoA is going into the pathway of lipogenesis. It's taking that acetyl-CoA and turning it into fat. If insulin is down— and at the same time, insulin is inhibiting ketogenesis if it's elevated. So if insulin is up, it's inhibiting ketogenesis and activating lipogenesis. If insulin is down, then you have no lipogenesis happening, and that acetyl-CoA Is going to be— you can burn it in the citrate cycle or the Krebs cycle, but the Krebs cycle is handling so much of it because you can't stop burning if insulin's low that the acetyl— the excess acetyl-CoA starts getting diverted or shunted out into this ketogenic pathway. But again, it can only happen when insulin comes down. So back to the relevance of the continuous monitor here, a rising ketone then is not this unexpected fuel that's appearing from nowhere in a, in a very real way. It's a readout, it's a measurement that insulin is no longer blocking fat burning and ketogenesis. That can be physiologic, in other words, people who want to monitor to know how well they're doing with fat burning state. Or of course, you have a pathological relevance, which is the people who may have no insulin in the untreated diabetic, type 1 diabetic, or they're on an SGLT2 inhibitor that's activating ketogenesis too much. And now you know that they're going a little too far with their ketone production. Okay. Nutritional ketosis as a term, it's a, it's the process of a, a very well-regulated ketone production. So there's enough insulin to prevent uncontrolled lipolysis or uncontrolled ketogenesis. and the acidosis that would come from that excess production of ketones. In the average person who's eating a normal standard global diet, usually the BHB levels are going to be below, you know, 0.2, 0.1. I mean, often depending on the device, it's below a range of detection. But if a person starts restricting carbohydrates, um, or they're fasting, then BHB will start to get above 0.5, which is classically considered ketosis, and it can go up to around 2 or so millimolars per liter. Of course, it can go higher, um, and, and it won't affect pH. The, the precise level of, of BHB that matches with the onset of acidosis is, is variable in people, um, but it's usually not something that a person can low-carb their way into. Um, with diabetic ketoacidosis, that uncontrolled ketone production is a result of Essentially no insulin. I mean, or, or just not enough to turn it off unless a person has type 1 diabetes. Um, then they're— it's essentially impossible. Of course, medications can skew that. Heavy alcohol can as well. The, the consensus view when it comes to the criteria, um, and, and BHB, they look at a BHB of 3 millimoles per liter or higher. as the level that correlates strongly with acid-base change. Now, again, I say that as the consensus view because I actually don't agree with that. I think 3 millimolar is still an amount that I know people who get into, they can get to that level of ketosis and there's no signs or symptoms of acidosis at all. So I add this next part very deliberately. To diagnose someone in ketoacidosis, you need much more information, including measuring pH and looking at things like urinary bicarbonate. I don't want to get into the— Yeah. Buffering system here and why the bicarbonate matters. But suffice to say, there are more variables that need to be confirmed before you would say that a person's in ketoacidosis. Now, with ketosis and ketoacidosis, the 2 states can overlap. That's why I said there's not a clear number that distinguishes one from the other, because someone could feasibly be at a BHB level of 3 or 4 and be in significant ketoacidosis, perhaps, but someone else could be at 3 or 4 and not have any pH change at all. Anyone who's experimented with ketogenic diets or some long-term fasts, some of you've probably gotten to 3 millimolar with your BHB levels, and it's not something that the clinician would need to react to at all. So back to the consensus. An international panel published the first consensus framework for how the continuous ketone monitoring should be used. And those— and some of the numbers they have in mind Well, are, are, are these ones. So normal in based on this consensus is below 0.6 millimoles per liter. They call that normal. Of course, I mean, we could chuckle at that, those of us who at least defend ketogenic diets. But that was what they said. Elevated ketones, elevated was considered 0.6 to 1.5. Then high is above 1.5 up until 3, and then urgent is 3 or above. Again, I think those are pretty strict and unnecessarily low, but we all know that general consensus statements are typically not very friendly to low-carb diets in the first place. But nevertheless, the panel recommended a trend scale of about 0.4 millimoles per liter per hour. And, and if someone was right, had ketones increasing at that rate, again, 0.4 millimoles per liter per hour, then there would have to be an alarm in order to help them react and intervene before things got too high. I actually think that's fine. So as much as I kind of wince a little bit at the hard levels they're outlining here, I think that trend, if someone's accelerating their ketogenesis that quickly, that probably is suspicious and likely indicative of a true insulin deficiency or the SGLT2 inhibitor doing a little too much. Typical low carb, even fasting, isn't going to increase ketones quite that quickly. But with the rise of continuous ketone monitors, we'll find out. But all of this to say, it's part of why glucose alone can be misleading, because ketones can rise independently of glucose. Remember, glucose itself does not have anything to do with ketogenesis. It's the insulin. Ketoacidosis, including the euglycemic form, so the form where someone has They're in ketoacidosis, but their glucose levels are in a normal range. That can develop again in range. That's why we're saying euglycemia, or the glucose may be only modestly elevated. And an SGLT2 inhibitor is particularly relevant here, but there are some other ones too, like certain illnesses can kind of complicate that too. If it's a genuine insulin absence, that's going to be less likely where you'd see the ketones rising. You're also going to see the glucose rising. Unless there were some other extenuating circumstances, including, you know, maybe the use of a medication, um, that was also kind of muddying the water. All right, but the manufacturer Abbott, they had actually a good description here. They said that a glucose monitor without ketones is a speedometer without a check engine light. I actually like that. In metabolic terms, glucose shows you how carbohydrate is being handled. And ketones show you whether insulin is working well enough or at a high enough level. I shouldn't say working well enough, but at a high enough level to inhibit fat breakdown and ketogenesis. Now, a moment ago I mentioned SGLT2 inhibitors. They are known to carry a ketoacidosis risk, including when glucose levels are in a somewhat normal, although I would say on the high end of normal range. So this is one of the main utilities, I think. Or what could have helped the FDA panel approve this for the use of diabetes. The SGLT2 inhibitors are a quirky class of medication. Okay, now another important feature with these continuous monitors is that the new— these sensors measure BHB in the interstitial fluid rather than in the blood. And the blood, of course, is what you measure with the finger stick strip. So this is a different compartment. The interstitial fluid is the space outside of your blood vessels. That's that interstitial compartment. There is a physiological lag from blood into interstitial fluid for both glucose and ketones, and the, the degree of that lag varies by person and by metabolic state. There's some interesting science to review here. A first-in-human feasibility study published 5 years ago used an enzyme-based sensor to track interstitial BHB across 14 days of wearing this device, but based on a single calibration and in 12 healthy adults eating low-carb diets, the interstitial BHB tracked capillary BHB well enough to ultimately justify the approach. Now, a separate emergency room study compared interstitial BHB with venous BHB. in 34 adults, or, or blood BHB being treated for ketoacidosis. And they, they on average, they came to the emergency room with a BHB of 7 millimoles. That's in that high enough range where pH is getting affected. The interstitial values did run slightly lower with the mean difference of about 0.4 millimoles. So it's pretty close. So pretty strong correlation. In that study, the continuous sensor actually detected the resolution of ketoacidosis about an hour earlier than standard care did. So once again, we see the value of these— of the continuous monitoring, because naturally you don't want to continue to give the person insulin to try to slow down their ketones if their ketones are already in a normal range. Otherwise, you run the risk of pushing them too low. So interstitial BHB is a very real And an actionable signal here that allows the in-person to react much more promptly. Okay, the people that this changes the approach or life for, I think it can be life-changing, especially in those for whom the device was explicitly approved. People with type 1 diabetes, man, you miss an injection or your pump is failing, you can develop ketoacidosis in a pretty short order. People with type 2 diabetes who are on an SGLT2 inhibitor, once again, these are people who really want to pay attention to what their ketones may be doing because the glucose might not be shifting quite as much. With the— as the— as this continuous glucose and ketone monitor develops further, naturally it's going to kind of plug in with the insulin delivery mechanisms, the insulin pumps, and then the pump can be responding maybe to both of these signals in various ways. Now, I know that the Metabolic Classroom has an audience of people who like to wear these sorts of devices, even though they may not have diabetes. They want to be looking at their ketones as a mark of fat burning rather than a warning that their insulin's too low. So I wanted to just go through that very quickly as we near the end here. The glucose sensor alone answers the obvious question, which is how well glucose is being cleared from the blood after you've eaten some carbohydrates. You can see how high the meal pushes the glucose, how long it's staying up, And then how quickly it comes back down, all of which can give you a very good idea of your own metabolic response to glucose. What it cannot tell you is how much insulin the body had to spend to produce that curve. And that is the blind spot, because someone with substantial insulin resistance may still have a decent glucose curve, but it needs to flood the system with insulin in order to keep it decent. So A smaller curve is not necessarily the same thing as a healthy curve. Ketones can bring in the part of this that glucose can't. Since insulin is what suppresses ketogenesis through the mechanisms I mentioned earlier, measuring ketones can end up being an indirect measurement of what insulin is doing. An overnight or fasting BHB is, is this indirect statement or indicator of how insulin is, is behaving, how quickly it dropped, how quickly ketogenesis was able to turn back on. So that's something that's important for people who are looking at the ketones for a reason beyond diabetes and the concern of, say, insufficient insulin or a certain medication. But when you put these things together, you can actually start to get a good idea of metabolic flexibility, that capacity to move between fuels, to move between sugar burning and fat burning. So over an overnight fast, you'd expect that glucose ought to be coming down to a nice, good fasting range, and ketones ought to be drifting up. Then after you eat or drink some carb-heavy meal, then you're— you'd expect the reverse. Ketones should be coming down as insulin goes up. And then over a few hours later, you'd start to see the opposite. But how far the ketone goes up or goes down ends up being again that inverse indirect indicator of how high or low the insulin may be. Now, we don't have some good data on the normal daily rhythm of ketones in people yet, so that— but we're going to shortly, of course. So this technology kind of opens up that territory for us to be able to learn more about ketone dynamics. The sensor, of course, is approved for people with diabetes. But the physiology here, I think, starts to make sense for everyone to use. So as much as I'm an advocate of people using CGMs, at least for some period of time, to learn how their body responds to foods and habits, I just as strongly advocate the use of continuous ketone monitors. And there is a particular power when it comes to measuring both at the same time. Glucose, in closing, helps us understand how the body handles carbohydrates and our glucose tolerance. Ketones help us understand, at least to a degree, what insulin's doing. But again, albeit directly. Together, I think they help us paint a good picture of metabolic control, definitely more than either one of them can do alone. Class dismissed. Until next time, more knowledge, better health.