FULL TRANSCRIPT:

Glucose Metabolism: HbA1c, Insulin Resistance & Better Energy | Dr. Ben Bikman & Mike Haney

In a recent episode of A Whole New Level, Levels editorial director Mike Haney sits down with Dr. Ben Bikman, professor of cell biology and physiology at Brigham Young University and one of the world's foremost researchers on insulin resistance and metabolic disease. Bikman spent his early career studying how fat cells drive chronic inflammation and insulin resistance — first as a doctoral student in bioenergetics, then as a postdoctoral fellow at Duke-NUS Medical School in Singapore, where he investigated why some ethnicities develop type 2 diabetes at far lower body weights than others. He is the author of Why We Get Sick, which makes the case that insulin resistance is the common metabolic thread running through virtually every major chronic disease.

The conversation covers what insulin resistance actually is at the cellular and tissue level, why hyperinsulinemia and insulin resistance are inseparable, which markers matter most and why fasting insulin is almost never tested, how to read glucose curves on a CGM, and what it actually takes to reverse metabolic decline.

"Fasting insulin is the metabolic canary in the coal mine. That is going to be, in most people, the earliest signal that we can look at — projecting them down the metabolic road into metabolic decay." — Dr. Ben Bikman

From exercise physiology to insulin: how a fat cell paper changed everything

Mike Haney: Dr. Ben Bikman, thanks for joining us today.

Ben Bikman: Oh, yeah. My pleasure, Mike. Thanks so much.

Mike Haney: This is exciting for me because when I joined Levels — which was really my introduction to glucose and insulin as molecules that we want to pay attention to — the first book that Levels gave me to read was Why We Get Sick, your book on insulin. And it is still the book that I make all of our writers read and that I point everybody to if you want to get a really simple understanding of why these molecules matter — molecules you probably haven't heard of unless you've been diagnosed with diabetes or know somebody who has, and why they are so important. So I'm glad to finally get a chance to sit down and chat today.

Ben Bikman: My pleasure. That's great. I'm delighted to hear that the book has had some legs to it.

Mike Haney: Yeah, absolutely. So even though I've read the book and you've worked with Levels for a long time, I don't know a lot about your background. I just want to start there. How did you get to where you are? How did you come to focus on insulin?

Ben Bikman: Yeah, thanks. That's a fun question for me to answer. So I'm born and raised in a little town in Alberta, Canada. My upbringing was a typical Gen X type kid where you're just out all the time. I loved physical activity, I loved exercise. So my undergraduate degree and my master's degree was in exercise physiology, where I imagined being a scientist who studied muscle. That was my main interest, in part because of my past in some athletics.

But during the course of my master's degree, when I was looking to the future and anticipating what I wanted to become an expert in, I stumbled on a paper written by a scientist — last name Hotamisligil — from a Harvard lab, where his lab had just identified what was a big topic in the late '90s: the beginning of what was described as subclinical chronic inflammation. It wasn't inflammation to the point of septic shock or sepsis, but there were detectable elevations in inflammatory cytokines. So it was subclinical because it wasn't causing overt disease, and yet it was persistent. They identified that the fat cell was the source of this subclinical chronic inflammation that always seemed to be creeping along with obesity. And that was one revelation for me. It was an absolute mind-blowing realization to understand that the fat cell was more than just a passive storage organ for energy — that it was a very active part of the endocrine system.

I was at the time taking what turned out to be my favorite class of all time, endocrinology, from a wonderful professor whom I love to this day. And when I later got hired at my same undergraduate institution at BYU, the one thing I wanted was to take over that class — a graduate-level endocrinology class. I was absolutely dumbfounded to learn that the fat cell plays a very active part of the endocrine system. But further, it was that inflammation caused by the fat tissue that connected fat gain to type 2 diabetes. My interest in insulin came from wanting to understand what was the connection between too much fat and too much blood sugar — between obesity and type 2 diabetes — and insulin resistance was the great mediator between those two.

Over the course of my doctoral studies, which was in a field of bioenergetics with a wonderful scientist named Linus Dome, I began studying lipid-induced insulin resistance and insulin resistance in obesity. Then my postdoctoral work was with another incredible scientist, this time in the beautiful country of Singapore, where the Singapore government was interested in understanding why the average East Asian and Southeast Asian Singaporean had such high levels of diabetes compared to fatter Caucasian Europeans who were also in Singapore — and why South Asians were even worse than the Southeast Asians with their diabetes risk. A large part of Duke-NUS Medical School in Singapore was focused on metabolic disorders. I did my fellowship there with a man named Scott Summers — incredible scientist.

During the course of my postdoctoral work, I attended an American Diabetes Association meeting in the US. I was presenting some of our work looking at this class of highly active lipids called ceramides and how ceramides would disrupt the insulin signal. But I was struck by a session being held at the same time — something like "Is Alzheimer's Disease the Newest Type of Diabetes?" It was a very provocative title, but I was struck by the fact that perhaps insulin resistance has a role in health or disease that goes beyond just type 2 diabetes. And that was the beginning of what ultimately became my book Why We Get Sick — to better understand and then share the fact that insulin resistance is actually a common pathology that connects virtually every chronic disease. All of these plagues of prosperity, despite their slight differences in origins, will share a common core at least to some degree.

That became much of my interest, going beyond my university efforts where I'm a professor and I love it. But I didn't want my career to be defined by how many papers I publish in peer-reviewed journals. Even as an academic, I appreciate the utility in that, but I also as a non-academic appreciate the limitations in actually relying on that as a mechanism of sharing information. And so I thought, what would I want my career to be defined by outside of academia? And it would be these efforts to promote metabolic awareness — an appreciation of the metabolic origins of chronic disease. Which is to say, insulin became a professional obsession.

Mike Haney: So interesting that those insights were coming about in the '90s — the idea of adipose tissue being an endocrine organ. I guess I would have assumed we knew that in the '50s.

Ben Bikman: No. Let me clarify, because that's a good point. It was a revelation to me at the time, but we had known that hormones like leptin and others were coming from the fat cell — I just hadn't learned that. So for me it was sort of a two-part revelation: learning that there were any hormones coming from the fat cell, having never heard that before, and then learning that even now you'd be very hard-pressed to find a classic or any modern endocrinology textbook that actually has a chapter devoted to adipose as an endocrine organ. When I created my graduate course, I don't rely on textbooks anyway — I rely just on peer-reviewed, first-source articles for the content of my lectures. But there was no textbook I could even go to that would have a chapter on the thyroid, a chapter on the gonads, the adrenal glands, and then adipose tissue. So even now it's not one of the canonical endocrine organs. And not only does it release some protein hormones, but a lot of them are related to inflammation.

Defining insulin resistance — and why it always comes with hyperinsulinemia

Mike Haney: Well, I think that's a good lead-in to where we want to start, which is really defining insulin resistance. This is part of a series we're doing really inspired by blood markers and how we measure the health of different systems within the body. Today the system we're talking about is metabolic — and that's ultimately going to come down to glucose and insulin, though we can talk about other markers that might be relevant. And the dysfunction at the core, for all the reasons you just said, is insulin resistance. It's a topic we've talked about a lot, but I think it is absolutely worthwhile for you to define it in a way that will help set up the rest of this conversation.

Ben Bikman: There is the one part of insulin resistance that the name itself evokes, which is that insulin isn't working particularly well. Even that's not so simple, where we may describe a body as being insulin resistant and the assumption would be that every tissue has an equal degree — that they're all 80% insulin resistant or something like that — which is absolutely not the case. There can be not only a sequence to this, where this tissue will become insulin resistant and then this one may follow, but there will be other tissues or cells that may never manifest with a degree of insulin resistance. Nevertheless, if we were to say that a cell or a tissue is insulin resistant, that would simply mean it's not responding very well to the signal that insulin is attempting to turn on. Insulin's coming and knocking on the door of the cell, and the cell once upon a time would quickly open the door and respond. Now, however, the cell isn't opening the door particularly well.

But that definition alone is insufficient when we actually step back and view the entire body. What we're describing at the cellular level is that the cell isn't responding well to insulin — that is the insulin resistance. But at a whole-body level, we need to understand the other part of this two-part pathology, the other side of the coin, which is chronically elevated insulin or hyperinsulinemia. I will make this a declarative statement: there is no such thing as insulin resistance in the human body without accompanying hyperinsulinemia. The hyperinsulinemia can be both cause and consequence of the insulin resistance. But suffice it to say, they will always come together. If you want to invoke the term insulin resistance, you are at the same time describing an overall state of hyperinsulinemia. If you remove the hyperinsulinemia, what you think you're describing as insulin resistance is not — it will be something else metabolic that is tangentially connected to insulin, but it's not going to be insulin resistance.

Mike Haney: That cellular story — insulin's knocking on the door, the receptors that are supposed to respond are not — can that happen in the absence of hyperinsulinemia, especially if the causal direction is both ways? What do those two scenarios look like?

Ben Bikman: Yeah, that's a great question. If you start at the level of the cell, there's going to be some intracellular mediator that is somehow blocking the insulin signals. So when insulin comes and knocks on the door, there would be a series of events — like in my home, someone knocks on the door, the dog barks, a kid answers the door, a kid yells for someone else to come. There's going to be some intracellular mediator that blocks it. A lot of my work has focused on ceramides. Other scientists have focused on other active lipids in that process, but regardless, there will be some signal. So whatever is going to cause a decayed insulin response at the cell must somehow increase the level of that intracellular mediator.

To me there are three primary causes of insulin resistance. My own criteria for what would meet the metrics of qualifying something as a primary cause would be that it is capable of causing insulin resistance on its own without any other stimulus, and that it has been validated in all three commonly used biomedical models — isolated cells, rodents, and humans.

One of the three is insulin itself. Too much insulin will cause insulin resistance, so in that sense it is a direct cause. Whether it's cells, animals, or humans, you give any of those things a little more insulin and they'll become demonstrably less sensitive to the insulin. That itself is just reflective of a fundamental biological principle: too much of a signal will cause a resistance to that signal.

The other two can occur in the absence of hyperinsulinemia, which is stress and inflammation. I have published papers on both of them. With stress — perhaps it's the endocrinology professor in me — stress is best defined as an endocrine state where you have the two primary stress hormones elevated, or one or both of them: cortisol and epinephrine. Anytime cortisol or epinephrine are elevated in cells, rodents, or humans, it will become insulin resistant even in the absence of elevated insulin. And then the last one being inflammation. This was a big part of my postdoctoral work — to understand the degree to which inflammation could hijack this system, promote ceramide accrual within the cell, and then cause insulin resistance. But anytime a cell is getting a lot of pro-inflammatory cytokines coming to it — TNF alpha, or the interleukins, or C-reactive protein, which is commonly measured — the cell will become insulin resistant quite quickly.

So of the three primary causes, one is insulin itself, but there are two others that represent an insulin-independent cause. But then when we put it into the whole body, as that insulin resistance starts to progress, the hyperinsulinemia will now be a consequence to try to overcome the insulin resistance. So if a cell has become insulin resistant because of, say, cortisol — maybe a person is on immune therapy, taking a corticosteroid — the body is becoming insulin resistant. Within a few weeks, they will be very insulin resistant. If you measure their blood, their insulin levels are high. It's because now the insulin is trying to overcome the lack of response. So one molecule of insulin was coming and knocking on the door of the cell and it was insufficient to get the response that insulin was seeking. So now it's a mob of insulin coming and banging on the door of the cell.

Now, Mike, even as I've described all of this, I have to say that the three primary causes are what I call increasingly fast insulin resistance — those are stimuli capable of causing insulin resistance within hours. And as the stimulus goes away, so too does the insulin resistance almost that quickly. But there is a more subtle or slow insulin resistance that is more tissue specific. And that's actually an idea sort of born from my experience looking across ethnicities from my postdoctoral time in Singapore — and to this day I collaborate with clinics in Singapore just to understand why it is that some ethnicities are further on that path of metabolic derangement despite looking quite lean.

Mike Haney: Just on that cellular story — if hyperinsulinemia is the cause, does that drive the other two? Will it create states in which you start to get, in other cells or other tissues, stress-mediated or inflammation-mediated insulin resistance?

Ben Bikman: That's a great question. It's not something I've ever considered. But I think it could — and it would be indirect. The hyperinsulinemia would be causing — and I'm getting now to the tissue, slow version of insulin resistance — hyperinsulinemia promotes hypertrophy of fat cells. And when fat cells undergo hypertrophy, they become very pro-inflammatory, thereby vomiting out a host of pro-inflammatory cytokines, which will then promote systemic inflammation. So that would be an indirect mechanism via the fat cell whereby hyperinsulinemia could promote it.

And then through the stress system — one interesting aspect is the dynamics with glucose. I could imagine a situation in which hyperinsulinemia-induced insulin resistance could be creating disruption in glycemia, and that disrupted glycemia will stimulate — anytime the body starts to go hypoglycemic, that is actually the one thing the two stress hormones have in common. As much as we look at cortisol and epinephrine as walking hand in hand through the body as part of the stress response, they actually are fantastically different. They come from different cell types. They have different origins and how they're produced — one being an amine hormone, one being a cholesterol-based hormone. They move through the blood differently. They act on their target cells differently. They have nothing in common except they both will spike blood glucose. And so if your insulin resistance and hyperinsulinemia is resulting in periods of hypoglycemia, you would be activating that stress response, which could be compounding the hyperinsulinemia-induced insulin resistance at the same time.

What insulin actually does — and why it affects every cell in the body

Mike Haney: Got it. Before we keep on this dive into insulin resistance, it strikes me it's worth taking a moment to talk more about what insulin does. The very high level, almost cartoon narrative, is: insulin tells the cells to take up glucose, our cells need glucose for energy — high glucose will trigger too much insulin, leading to hyperinsulinemia, which then triggers insulin resistance. But your book does really well in telling that glucose story and the million other things that insulin does within the body. So maybe just a few minutes on what insulin does, to set up why do we care about insulin resistance? Why is that problematic?

Ben Bikman: In fact, that would probably be the best description of my book. The most accurate title of Why We Get Sick would have been something like Why You Should Care About Insulin Resistance — but of course that would be a terrible title and no one would buy it, because the only people who would be interested — even the way we started this conversation, you'd mentioned how most people hear the word insulin and just think about diabetes — that would be the only people who would buy that book. Because most people, even very well-trained clinicians — in fact, I submit this is part of the problem with modern medicine — can't disconnect insulin from glucose, which is fantastically unfair to insulin. This humble little hormone does far more than just regulate blood glucose. In fact, that is a very small part of what it does.

There are only a few tissues that actually require an insulin stimulus to tell them when they should take in glucose — like the muscle and the fat tissue. Now they're big tissues, so they matter just by mass, but in the grand scale of the human body and all the myriad cells, all the rest of them don't need insulin to tell them to take in glucose. But they still need insulin to tell them what to do with energy and how to regulate and play nicely in the entire symphony of the human body. So insulin ends up being like a metabolic conductor, where insulin is telling which part of the body to take in energy or to store it or to burn it.

One of the best examples at helping people understand the role of insulin — and why insulin resistance is so relevant from top to bottom in the body — is looking at infertility. The two most common forms: in men, erectile dysfunction; in women, polycystic ovary syndrome. Both actually have an intimate connection to metabolism and metabolic health, which is to say to insulin, the ultimate metabolic hormone.

But before I even mention that, I should just say that insulin literally affects every single cell of the body. And I'm not using the word literally too liberally like my college students do, where every other word is literally when they don't mean it at all. I mean literally. Every cell of the body will respond to insulin. That is uncommon for a peptide, for a protein-based hormone, to be able to do. So insulin is in an elite class within the grand scale of hormones in that it will affect every cell.

So the forms of infertility are a fascinating example of just how wide or broad insulin's effects are, and also how insulin resistance is that pathology with two parts — both decayed signal and hyperinsulinemia. In the case of erectile dysfunction, there was a paper published a few years ago with a title something like "Is Erectile Dysfunction the Earliest Manifestation of Insulin Resistance in Otherwise Healthy Men?" This group of clinicians was positing the idea that erectile dysfunction could be an early diagnosis for insulin resistance — a sentiment that I appreciate.

Within the endothelium of the blood vessel — all blood vessels at the capillaries will have this endothelial layer — the endothelial layer will need a signal to tell it when it ought to vasodilate to increase blood flow to a certain area, or to vasoconstrict to reduce blood flow. Suffice it to say, normal erectile function requires a significant and specific vasodilation. Most people don't appreciate the fact that insulin is an essential signal that's stimulating the synthesis of nitric oxide in the endothelium. And so when insulin is working well, vasodilation will occur appropriately and then the man will have normal function. However, if the endothelium has become insulin resistant, insulin is attempting to facilitate the dilation but it can't — the endothelium has become insulin resistant, the signal is decayed, vasoconstriction remains the main situation, and thus there's no erectile function.

On the other hand, in women, the most common form of infertility is a problem within the theca and granulosa cells of the ovary. They don't become insulin resistant — and once again, most people don't appreciate the fact that one of insulin's many effects is to regulate the actions of the enzyme aromatase, which is responsible for converting testosterone into the estrogens. All estrogens in men and women were once testosterone. The gonads — ovaries or testes — will convert that testosterone into a certain degree of estrogens. Now, in women, there must be a robust estrogen signal throughout the course of the month in order to have ovulation. If there isn't this huge estrogen spike, there will not be actual ovulation. So the ovaries will have had multiple eggs maturing and developing, but in the absence of the big estrogen spike, you don't get the overall endocrine dynamic and you won't have ovulation. And so all of those eggs just stick around and they become cysts. That's the polycysts of the ovaries.

But all of it came back to insulin mediating the conversion of testosterone to estrogens. And that is a signal that does not decay. So in the midst of her hyperinsulinemic body, that high insulin is resulting in too robust an inhibition of aromatase — leading to a state where she not only isn't making sufficient estrogens to get the ovulation, but she also now has too much testosterone. This gives her some of the less desirable side effects of PCOS, like the hirsutism, the higher body hair or facial hair, acne, male pattern baldness — all of these testosterone-related things she now also has to suffer from. And at the heart of it all was insulin just being too high. Not the insulin resistance per se.

Mike Haney: As much as we've written about PCOS, I don't think I've ever heard that connection. I've heard that insulin matters and it influences those hormones, but not to that sort of degree. Coming back to that — maybe this is applicable to both the cellular and the tissue level — I want to revisit that very basic concept you mentioned that just sort of applies to all biology: the idea that too much of a thing will cause resistance to it. Why is that? And what's the scale at play — is there a tipping point, or is it a linear kind of response? And particular to insulin — how does that work? And I guess another part of this: is everybody a little bit insulin resistant? Is there a progression where individual cells or parts of tissue can be more or less resistant, or is it more of a tipping point?

Ben Bikman: Well, that's a great question. I wish I could cite a specific threshold, but I have to speculate. I'm pretty comfortable saying it's going to be a gradient — that there will be this mounting movement where as insulin's going up, it's going to be subtly decaying the signal, subtly decaying the signal, and as it keeps going, the signal continues to decay. I'm pretty comfortable stating it like that, but there probably is a range where — just for the sake of clarity for the audience — it's probably the mid-teens of insulin in microunits per milliliter. When I wrote Why We Get Sick, I wanted people to have actionable information. That's where I'm a bit of an uncommon academic — I'm really driven by the desire to translate the science into something that a person could say, "Okay, now I know more and I know what to do." And so I actually scoured the literature to try to find what were the cutoffs, because you cannot rely on modern medicine where insulin resistance is the most common health disorder worldwide.

So you can't only not find a consensus for what are good insulin levels, but if you're looking at your blood test and it's showing that insulin level of 25 and below is good — remember that's based on the average population, and the average person is insulin resistant. So that's already going to be too high. So for me, I'll answer this in two parts. One, it's absolutely going to be a gradient — as you're cranking up the volume at a hard rock concert, standing too close to the speakers, as the volume keeps getting cranked up, you're getting increasingly deaf to it. Too much of a signal resulting in a disruption of the signal. But at the same time, I think the average person who's going to go get their blood test — I have sort of different degrees, but just to make it simple — I think if you're seeing your insulin in the mid-teens consistently, that's a warning. I actually put that in a yellow light category. If you're below that, you're pretty good — that's a good sign that you're insulin sensitive. If you're above that, that's a pretty good sign that you're insulin resistant consistently. But around the teens is you kind of being on the threshold of tipping into it.

But then you'd mentioned whether everyone is to some degree insulin resistant. Yeah, just like how everybody wouldn't have an equal degree of hearing — what would be perfect hearing? What would perfect total insulin sensitivity look like? I don't know how that would look, but it wouldn't look the same at all tissues, even within the same body. Someone's fat cells — which is the tissue that I believe becomes insulin resistant first in the slow version of insulin resistance — someone may have fat tissue that's becoming insulin resistant first. But someone else may have a liver that's becoming insulin resistant first because of the dripping of microbial content from the gut, because of leaky gut, and the liver being on the front lines of receiving that.

Unfortunately, everyone's going to be somewhere on that spectrum. And that's not even necessarily a bad thing, Mike. As much as my work focuses on pathological insulin resistance — the insulin resistance that conveys disease risk, whether it's Alzheimer's or infertility or hypertension or whatever — there is a realm in which insulin resistance is supposed to happen, which is physiological insulin resistance. There are the two P's of physiological insulin resistance: pregnancy and puberty. Everyone who goes through puberty is experiencing demonstrable insulin resistance. It's supposed to happen, and it's selective, and it's enhancing growth of certain tissues — it's happening by design.

Similarly, when a woman is pregnant, she needs certain tissues to be growing fantastically faster than anywhere else in the body. This selective but still manifested at a whole-body-level insulin resistance will facilitate her body preparing for the metabolic marathon of pregnancy and helping that baby grow appropriately. Pregnancy is a fascinating one where you can track it in almost a perfectly linear fashion, going up steadily from day zero to month nine. The moment the baby is born, insulin sensitivity returns quite quickly — all the more quickly if she is able to breastfeed.

Mike Haney: Right. And thus the relationship with gestational diabetes — when it gets a little out of hand.

Ben Bikman: Yeah, that's right. When the insulin resistance has gone so far that now she can't keep her glucose in check. But that brings us back — if you'll allow me to very quickly revisit part of how we started the conversation — which is that modern medicine has coupled insulin and glucose so tightly together that they can't conceive a scenario in which, well, if glucose is high, of course your insulin is high, but if glucose is normal, of course your insulin is normal. I've had bizarrely circular conversations with very smart clinicians where I will say at the end of my talk, "We need to be measuring fasting insulin to get an idea of insulin resistance status." And they will say, "Well, why do we need to? We've already measured glucose." Then I will say, "But glucose and insulin levels aren't going to be the same." And they'll say, "Well, they're always the same. We don't need to measure insulin if we're measuring glucose." And we just keep sort of talking in this bizarre circle.

But to make it very clear: they're not the same. And if we want to understand insulin resistance — whether it's in pregnancy, puberty, or the pathological version — we need to appreciate that insulin resistance, at least in its earliest stages and depending on which tissues have become insulin resistant in what order, has insulin elevated but that insulin is sufficiently capable of keeping blood glucose in check. But in the midst of this, because we have such a glucose-centric paradigm of modern medicine, not only are we missing the diagnosis, but we're also then diagnosing people with all of the problems of insulin resistance — not appreciating that they could have this metabolic foundation. We're giving them a blood pressure medication or a medication for their migraine headaches or their fatty liver disease or whatever, not knowing that this body is actually insulin resistant. We were just looking at a marker that in the fasted state isn't going to tell us as much. Now, in the dynamic state with CGMs, of course, that becomes much more useful than just going in once a year and getting a fasted blood glucose measurement.

Slow insulin resistance: why fat cells fall first — and what that means across ethnicities

Mike Haney: I want to talk more about this tissue level, this slow insulin resistance. We've talked about insulin resistance in the muscle, in the brain, in the liver, in the endothelial cells. What does that mean for the average person thinking about their health and the development of insulin resistance? Do those have unique pathologies? Are they related to each other? What's the utility of understanding slow insulin resistance?

Ben Bikman: So before I answer the question, just to sort of set the stage — why is it that you can go to Singapore, which is a beautiful kind of melting pot? You have a lot of East Asian — primarily Chinese — ethnicity. You have a lot of South Asian or Indian ethnicities, Malay, and then white European, all mingled together. On the extreme ends, you could have an East Asian guy and a Caucasian European guy. Why is it that the average white American will be much fatter than the average East Asian Japanese individual and yet be less likely to have type 2 diabetes? Why could some ethnicities be fat and metabolically okay, and then other ethnicities be relatively lean, maybe a little chubby, and yet already have deep type 2 diabetes?

That comes back to the slow version of insulin resistance, which I believe starts with the fat cell. Now, there will be scientists who say it's the liver that becomes insulin resistant first. No, it's the muscle that becomes insulin resistant first. I'm putting my stamp on this — they're all wrong. It's the fat cell that will become insulin resistant first. Now, you can have those unique pathologies — like if it's leaky gut and the liver is at the front lines of receiving what's coming from the gut and getting a lot of pro-inflammatory material — the liver is going to become insulin resistant. But if the insulin resistance is starting at the liver or the muscle, you will not maintain normal glycemia. The muscle and the liver in particular are so powerful at regulating blood glucose that if they start to become insulin resistant, yes, you're going to have elevated insulin, but you will have elevated glucose. So we've already gone into formal pre-diabetes or type 2 diabetes.

But why is it that you can have people with elevated insulin for 10 or 20 years before the glucose ever changes? If it were those tissues that were falling first, you wouldn't have that — you would have more rapidly bumped up the glucose. So I firmly believe that the slow march of insulin resistance starts in the fat cell, and that explains why some ethnicities are further down that path at only modest elevations in body fat levels. Why is it that you put 10 pounds of fat on a Chinese or Japanese guy and he's already type 2 diabetic? You put 10 pounds of fat on the white guy or the black guy — actually whites and blacks kind of have this in common — they just are a little chubbier and don't look as good in their speedo. What is the difference? It's the size of the fat cells.

Our view on adipose tissue and obesity is incorrect — we think it's all a matter of fat mass. No, the size of the fat cell matters more than the mass of fat that the body has. This also explains even within ethnicities why women are invariably fatter than their male counterparts and yet healthier in every metabolic marker. If it were just a matter of fat mass, that should be impossible. It's not a matter of fat mass — it's a matter of fat cell size.

Some ethnicities and women across all ethnicities can do this more than men: as there is a pressure to store more fat, they will make new fat cells. They will activate a process called adipose hyperplasia, where the fat cells are able to multiply, if you will. Thus this body type will have more fat cells but they're all smaller. Small fat cells are very insulin sensitive and anti-inflammatory, thereby promoting an even further insulin-sensitive state.

In contrast, there was one study that took adipose biopsies from South Asian men and compared them to Caucasian men — even though they were at the same body fat level, the South Asian men had fat cells that were about four or five times larger in volume than the fat cells from the same depot of fat in the Caucasian men. So in this other group, they don't have the proper genotype to switch on the synthesis of new fat cells. Thus any pressure to store fat is stored up in larger and larger fat cells — a process called adipocyte hypertrophy.

Hypertrophy, not hyperplasia, will create a very disadvantageous metabolic milieu where the fat cell will get so big that it will need to start limiting its growth, and in order to do that, it becomes insulin resistant — because insulin is a signal that tells the fat cell what to do with its calories, to store them as triglycerides. But as the fat cell is reaching a point of maximum dimension, it must become insulin resistant to stop its growth, lest it literally erode its own membrane and fall apart, which would be a very messy death. At the same time, as the fat cells are getting so big, they're pushing each other further and further from capillaries and their life-giving blood. And so, in order to rectify the hypoxia or the suffocation, they will start releasing pro-inflammatory cytokines — because some of them will act like a trail of breadcrumbs for the capillary to sense and then start to follow, growing off or budding out a new capillary to feed the hypoxic, hypertrophic adipocytes. But at the same time, as that pro-inflammatory signal is moving through the body, it's further compounding the insulin resistance.

Now, having said all of this, Mike, I would just say that here we actually have an opportunity to identify insulin resistance at what I believe is the very first place — the fat cell itself. And that is not only measuring the insulin but then also measuring the product of what's coming from the insulin-resistant fat cell, which is free fatty acids. That's a very uncommon measurement — when we look at the lipid profile, none of those will include free fatty acids. But it's such a beautiful example of the balance between endocrinology and metabolism that I can't help but mention it.

In the normal human body, when insulin goes up, part of the mechanism whereby it promotes the growth of the fat cell is by inhibiting lipolysis or the breakdown of fat. So if insulin is up, you would expect free fatty acids — the product of lipolysis — to be down. That's in the postprandial or fed state. We eat a lunch, we get some carbs, insulin goes up, free fatty acids would be low. In contrast, we're fasting or eating low carb or exercising — insulin will go down, and with insulin being down, lipolysis at the fat cell is disinhibited. So we'd expect free fatty acids to be up. It's always one or the other.

Now, however, coming back to the insulin resistance of the fat cell as it's undergone hypertrophy — both of them are high. You have high insulin and you paradoxically have this metabolic oddity of also high free fatty acids. That's something called the adipose insulin resistance index. In men, ideally it's around six — that's an arbitrary unit. Women, it's ideally around nine. Women naturally have higher free fatty acids than men do — they're more lipolytic, burning a little more fat than men are at any moment. That is one marker that's one of my favorites. It's a reflection of insulin resistance in that slow progression at its earliest tissue, namely the fat cells.

Mike Haney: Why does all of that make you think that the fat cell is where insulin resistance starts?

Ben Bikman: Because its metabolic demands are sufficiently modest that it is capable of spreading the insulin resistance through the body — and it can become insulin resistant itself to prevent its bursting like an overfilled water balloon — but if it becomes insulin resistant, it will have minimal to no impact on underlying glucose levels. The reason I put the fat first and then other tissues like the liver and muscle — which are other contenders for being the first to fall — is that if those become insulin resistant, you will see changes in glucose. They're not going to have a normal glucose level anymore. The glucose is going to have been creeping up almost lock-step with the insulin. But we don't commonly see that.

Not that it can't happen — I've already mentioned a situation where I could imagine the liver becoming insulin resistant with something like a specific liver infection, or muscle disuse and sarcopenia. I have some friends at the University of Utah who have really studied that well. But for the average individual who's marching through life towards insulin resistance, I firmly say it's probably the fat cell. Those other scenarios would be very specific instances of a particular liver problem or pathology, or leaky gut, or very specific muscle wasting and sarcopenia.

Mike Haney: Just to play out that pathology — is there then a path where insulin resistance begins in the fat cells, and the degree to which you are susceptible has a significant genetic component, and then that insulin resistance spreads to the other tissues?

Ben Bikman: Yeah, absolutely. Fat falls first — implying there are others to fall, where it's the first domino to bump into the next ones. If I were to imagine an average person, I bet it would be fat and then liver, because of where the liver is with regards to blood flow from visceral fat — it is right downstream from visceral fat. The liver is at the first line of response in a lot of remarkable ways that should make us very grateful for the role of the liver. But if you have a lot of visceral fat — and visceral fat will almost exclusively grow through hypertrophy, which is why visceral fat is so problematic, because it does not have the potential for hyperplasia, it will only grow through hypertrophy — then it's going to be belching out not only all of its free fatty acids but also the pro-inflammatory cytokines, and the liver is going to be literally the first one to receive all of that.

So I would suspect it would be fat, then spreading to the liver, which is then spreading it to other tissues to varying degrees. And that I think would have a degree of individuality to it — where if a person starts to manifest with hypertension, then it was probably his endothelial cells that started to become insulin resistant systemically, staying constricted when they ought to be dilated. In someone else, it may be manifesting as early cognitive decline, where specific neurons are becoming insulin resistant and they're going hungry — not functioning very well, and thus you start to manifest with memory or cognitive deficits. And in someone else, it could be mood deficits, because of other neurons becoming insulin resistant. So there would be absolutely some individuality at that point — sort of choose your own adventure. Where do you have a genetic susceptibility? Is it your blood vessels? Is it your infertility? Is it your brain? That could be quite individual.

Fat storage, insulin's role, and what actually makes a cell get big

Mike Haney: So the insulin resistance developing in the fat cells — would you put insulin first among the causes?

Ben Bikman: For me, if you even want to understand how the body is storing fat in the first place — as much as we still predominantly have a calorie-centric view of what would tell the fat cell to get big in the first place — the calories matter, but only insofar as the fat cell knows what to do with those calories. And that is an absolute exclamation mark statement that I'm not wavering on: in the absence of insulin, it is utterly and totally impossible for the fat cell to get big.

And the proof positive of this is a condition in type 1 diabetes called diabulimia, where the type 1 diabetic — maybe having learned in teenage years, most susceptible to these sorts of pressures — has figured out that all they need to do is deliberately underdose their insulin. In the absence of that one single signal, if you turn it down, it doesn't matter what else is happening in the body. It is totally impossible for that body to hold on to its body fat. They can be eating a caloric surplus. It can be a petite little 14-year-old girl eating 6,000 calories a day, and by deliberately underdosing her insulin, it creates a metabolic impossibility — she cannot store that energy as fat.

Now, for reasons I won't get into, we're not breaking the laws of thermodynamics. My PhD is in bioenergetics, and so while I'm called a heretic for denying thermodynamics, I actually, ironically, probably have a greater appreciation of thermodynamics and biology than anyone who's accusing me of being a denier. But I'm also smart enough and informed enough to appreciate the endocrinology and the essential endocrine signal, which is insulin.

So all of that having been said: to understand why a body is storing more fat mass, you have to have a signal to tell the fat cell to store that energy — which is insulin. Then you have to have sufficient energy to fuel that storage, which is where the calories come in. So you do need both. I just end up focusing more on the insulin because it's the one that people either aren't aware of or are aware of but are doing their best to try to deny.

The pancreas side: what happens to beta cells as insulin resistance progresses

Mike Haney: Before we leave the science lesson here and get into the practical testing and what to do about this — the one other area I want to talk about is on the pancreas side. The actual beta cells that are sending out the insulin. The sort of common high-level narrative is that as the body requires more insulin to respond, it's stressing the beta cells, and eventually they just get tired — maybe not crap out, but sort of go into a kind of offline state. Talk about what's happening in the pancreas as all this is going on.

Ben Bikman: I'll need to just preface all of my answer by clarifying the state, which is type 2 diabetes — as opposed to type 1, which is a totally different situation where it's autoimmune, where you are targeting those cells for destruction, and that's just the genetic predisposition combined with an environmental trigger. So putting that to the side: with type 2 diabetes, I appreciate being able to discuss this because once again we have a fantastic misunderstanding of the progression of type 2 diabetes. Within the biomedical realm, we will commonly say insulin becomes insufficient to control blood glucose. Insufficient is a relative term unfortunately, and it leads to a lot of confusion, because the individual may hear that and think, okay, insulin levels are too low. Too low relative to what?

The problem with that view is that in true type 2 diabetes, insulin will follow a bit of a journey. Let's say it started right here at my knee level, at the ideal insulin-sensitive state. But then over time as the person's progressing through insulin resistance to type 2 diabetes, insulin is climbing, climbing, climbing, climbing, climbing. And then in some people, insulin can stay high in a person with type 2 diabetes — because type 2 diabetes is when the glucose levels start to climb. We don't define it based on insulin. If only we would — we'd detect it a lot sooner if we did. But we define it by the glucose.

So when glucose levels start to climb, that can happen when insulin is still really, really high. The muscle is the main consumer of glucose — 80% of what brings the glucose down when someone's wearing their CGM is what's coming into the muscle. So the muscle is the great lion's share consumer of glucose. If it's becoming insulin resistant, that glucose is going to stay high. Similarly, if the liver becomes insulin resistant, now insulin can't signal the liver to hold on to all the glucose — the liver is too liberally breaking down glucose to dump it into the blood, even if there's already a lot there. So as those two tissues become insulin resistant, then the glucose levels will start to climb.

In other people, the beta cells will start to have this combination of changes ultimately leading to reduced beta cell mass. The beta cells are getting more and more abundant and now they're starting to contract a bit in number. So in some people, you'll see that insulin levels had gone from normal to really really high to now kind of high — which really amplifies the situation because now you're combining both the insulin-resistant state with less insulin overall. Because the hyperinsulinemia — whether it was causing it or a consequence of it — is a compensation. Even in the midst of all of it, if a little bit of insulin wasn't working well, then more maybe will work well enough to control blood sugar.

In some people, insulin stays really high. In other people, it will come down to high — but it will never go to low. If it's true type 2 diabetes — and not a person who just coincidentally develops type 1, which can happen — then it will always still be high, multiples higher than where it should be. That's really important for people to understand, because if you can appreciate that type 2 diabetes is always a state of hyperinsulinemia, then the term "insulin is insufficient to control blood glucose" becomes obviously more confusing. Insufficient — well, sure. But what's the alternative? Pushing it right back up to here.

And this manifests obviously clinically when you give a type 2 diabetic insulin. There have been meta-analyses after meta-analyses published documenting the health outcomes of giving a type 2 diabetic insulin therapy to control their glucose. The more aggressively you're giving them insulin to control their blood glucose — even in the midst of normal glucose levels — you're killing them faster. Their risk of dying from heart disease triples. And that matters because heart disease is the leading cause of death for people with type 2 diabetes. And type 2 diabetes is the leading risk factor for heart disease, the leading killer in the United States right now.

And then again, tragically, if you give that type 2 diabetic insulin to try to control their glucose — it's not the glucose that's killing them. It's the hyperinsulinemia. But that's not the common view. The conventional clinical care only has that glucose-centric paradigm. And that leads us not only to miss the marker that matters more — which is the fasting insulin as opposed to the fasting glucose — but compounding the problem, it leads the clinician to prescribe the wrong approach. "Well, if your glucose levels are high, I'm just going to give you insulin therapy." They not only have not measured the insulin, but now they're giving them more of what's killing them. It really is like giving an alcoholic another glass of wine, hoping that the extra alcohol will cure them of their alcoholism. You're giving them more of the very thing that's making them fat and killing them faster. And they do die faster — heart disease triples, cancer mortality doubles, Alzheimer's risk doubles. You're making them fatter and killing them faster by giving them more of what's actually harming them. The glucose is the side effect. It's the symptom of the actual underlying problem.

"Giving a type 2 diabetic more insulin to control their glucose — it's not the glucose that's killing them. It's the hyperinsulinemia. The glucose is the side effect. It's the symptom of the actual underlying problem." — Dr. Ben Bikman

Why fasting insulin isn't on your standard panel — and why it should be

Mike Haney: Right. That's maybe a good lead-in into markers then. The first question we get all the time is: we've just talked about how critical insulin is — why don't we measure it? Why is it not part of a standard panel?

Ben Bikman: I'm glad to hear that gets Rob Lustig riled up — then I feel better getting riled up too. I actually believe, if I could articulate it in what I hope is a clear and even somewhat empathetic way — because I sound very critical of clinicians, and I could imagine a clinician saying, "Well, it's easy for you as a PhD to say because you're in the lab, you don't have to interact with patients, and you don't have the constraints that modern medicine has" — like with insurance providers, where I may be a clinician who wants to measure insulin but if the insurance won't cover it, am I going to put that burden on the patient? I really do appreciate those limitations.

As much as it seems that I'm being a bit harsh, I think there is a very relevant sort of historical and scientific precedent here. So two things.

Starting with the historical: the main manifestation of what we call diabetes — in fact the very word diabetes comes from the manifestation of polyuria, or a lot of urine being produced. Why is that urine produced? Because of the hyperglycemia. Why were flies attracted to the urine as if they were flies coming to honey? That's where the mellitus term comes from — the honeyed urine. All of that is the diabetes mellitus term. It's because of the hyperglycemia, where when glucose levels reach around 200 milligrams per deciliter, you are overwhelming the kidney's ability to reabsorb that glucose, and so the glucose is spilling into the urine and pulling with it a lot of water. And in that urine is a lot of glucose, attracting flies like flies to honey. So historically, whether it was India or China or the UK hundreds of years ago, the earliest manifestations of the disease historically was the hyperglycemia. That gives some empathy in having a glucose-centric paradigm.

And then scientifically: we've been able to measure glucose for well over a hundred years in urine and plasma. It is so simple you can miniaturize the whole process and strap it on your body — that's how simple it is to measure a nutrient like glucose in the blood. Insulin is not so tidy. We've only been able to measure insulin for a handful of decades, and even then the earliest way of measuring it was a radioactive test. Even now the cheapest way for my lab to measure insulin is a radiolabeled enzymatic assay. That's complicated to use — radioactive chemicals in a lab, or heaven forbid a clinic. The science of measuring a peptide in the blood at such fantastically picomolar levels is brutally challenging. And so modern healthcare was influenced by the ease with which we could measure glucose as opposed to the challenge we face when we want to measure insulin.

A real-time insulin monitor — I think optimistically it's 10 years out, and I said that probably 10 years ago and we're no closer now. There are incredible hurdles to measuring insulin in real time. However, we can measure it very easily in a lab nowadays. And so to me, both of those explanations — while I state them with some compassion and empathy to my clinician brothers and sisters — it is getting less and less forgivable to continue to overlook it. More and more insurance providers are covering it. The cost is coming down. Even if a patient does have to pay, it's in the order of like $10 or something to just check that little box. More than anything, we just need the clinician to have that paradigm shift — to go from the narrowed glucose-centric paradigm to one that encompasses insulin and indeed other markers too.

But at a minimum, if I could have an executive order that influenced the modern shift happening in healthcare — which thrills me on one hand — I would have it be that fasting insulin is measured on every test. Because that is the metabolic canary in the coal mine. That is going to be, in most people, the earliest signal that we can look at and then project them down the metabolic road into metabolic decay and say, "Oh, hey, we got this early signal. Your glucose is still normal, but thank goodness — at a fasted state, your glucose is still coming in normal, but we looked at your insulin and it's high. Thank goodness we detected it earlier rather than waiting 10 years for the fasted hyperglycemia to start to manifest."

So to me, overlooking fasting insulin once upon a time was somewhat forgivable. It's less forgivable now, and I'd like to think that it's mostly a matter of education. To my great delight, when I'm invited to give seminar talks to clinicians at continuing medical education events, they are extraordinarily receptive. When they see the data and hear the logic of the argument, they are overwhelmingly convinced and then leave with an absolute conviction to start measuring insulin. I do think it's happening. I just wish we could ramp it up and work it into every medical school curriculum.

Reading the markers: fasting insulin, triglyceride-to-HDL ratio, and why they go together

Mike Haney: If a person manages to convince their doctor to give them the insulin test, how should they think about the result? There's a simple scale of optimal to unhealthy, but beyond that, how should I think about changes over time and that result in relationship to other relevant markers?

Ben Bikman: I'll mention my sort of three categories of good, maybe good, and then warning. Six microunits per milliliter and less is generally going to be a really good sign that you are insulin sensitive. From around seven to the mid-teens is going to be a bit of a warning — I'll come back to that in a moment. And then if it's anything in the high teens and beyond, that's very likely a warning sign that you are insulin resistant.

Now, why do I have that wiggle room in the middle? Because insulin wiggles — like most hormones, it has a circadian rhythm to it. And it's possible that when that person went in for their fasted test, they simply got measured at the peak. They go in another time or their sibling goes in at the same time and got measured a few hours differently — measured at their low point. So it's important to realize that like many hormones, insulin has a wiggle to it. Now it's not as big as some, like cortisol, which has a much bigger rhythm, or growth hormone, which has a much bigger rhythm. But insulin still has one, and that's all the more relevant because it exists in such low levels. Insulin is so powerful that it is able to do what it wants to do at these picomolar concentrations. So it's really, really strong.

So if someone looks at their insulin measurement and thinks, "Well, I'm in that kind of middle area that Ben mentioned" — how can you get a little more clarity? I actually think one of the most overlooked markers is a simple ratio that you can get from your normal lipid panel. The lipid panel is always going to have total cholesterol, LDL cholesterol, HDL cholesterol, and triglycerides. Take those last two. The triglyceride-to-HDL ratio is a really good surrogate for insulin resistance — and actually also a good surrogate for cardiovascular risk. Take your triglycerides number, divide it by your HDL cholesterol number, and if that answer is 1.5 or less, that's a good sign that you're insulin sensitive. That surrogate tracks really well with insulin levels, and it's going to be a little more stable. If you've gone in 8 to 12 hours fasted, it's going to be running at a flat line and will for the next almost 24 hours, depending on your exercise and such. But insulin will continue to be a little noisy. Those two are going to be rock solid.

So look at the triglyceride-to-HDL ratio. If it's above 1.5, that's a warning problem area. Below, that's a good sign. And the lower, generally the better.

Mike Haney: And why is that a useful proxy? What's actually going on in the body that makes those two things related?

Ben Bikman: That's just further reflection of how powerful insulin is with regards to controlling all metabolism. When I'm asked to give insulin a singular definition to explain what it does everywhere: it controls metabolism. It tells cells in the body what to do with the energy they have access to. Triglycerides and even HDL are not exempt from this.

On the HDL side, it's a little more modest, but hyperinsulinemia will accelerate HDL influx into the liver. So it'll clear more HDL from the plasma more rapidly, thereby giving a lower HDL. At the same time, there is no signal in the body that is going to stimulate the synthesis of new fats from the liver more than insulin does. So when insulin goes up, it's telling the liver that there's an abundance of energy and that we want to store fat in fat tissue. While insulin can tell fat tissue itself to store fat, it also recruits the liver to help with that process. And so it will activate lipogenesis — the process of synthesizing new fats. Then the liver will take free fatty acids, bind them into a triglyceride, and dump them into the blood, carried on a triglyceride-rich liver protein like VLDL and LDL. And we would detect that as higher triglycerides. So the hyperinsulinemia — not the insulin resistance — will both reduce HDL and at the same time increase triglycerides.

"The triglyceride-to-HDL ratio is a really good surrogate for insulin resistance. Take your triglycerides number, divide it by your HDL cholesterol number, and if that answer is 1.5 or less, that's a good sign that you're insulin sensitive." — Dr. Ben Bikman

Fasting glucose, A1C, and their limitations

Mike Haney: How about the glucose markers? We talked about how insulin can be rising for 10-plus years before the glucose markers move. But how should I think about my glucose markers — both A1C and fasting — in terms of the static ones?

Ben Bikman: My view on fasting insulin and A1C is a little nuanced, because I have a little bit of frustration that we've become a little overly reliant on them — which is one reason why I'm so enthusiastic about the democratization of CGMs. I really do believe fasting glucose levels are fantastically limited in their takeaways.

Now, having said this, fasting glucose levels — conventional care will say we want that to be around below about 110, and then above 110 up to the mid-120s is that warning range. And then mid-120s and beyond — boy, you're type 2 diabetic, time to drug you up in conventional clinical care.

Unfortunately, before I leave fasting glucose behind, I think it's important to note that if you look backwards over the past 60 years or so, you would find that what was considered the threshold for diabetes and problem glucose levels has been becoming increasingly more strict. Once upon a time, type 2 diabetes was defined at around 150 milligrams per deciliter, and then it went down into the 130s, now it's down to the mid-120s. One view — a generous view — could be that we've learned more about what happens with hyperglycemia. I'm not a very generous-minded individual in that regard — at home but not at work, and not with these sorts of things. To me I actually have a bit more of a cynical view, which is: the more you lower the bar, the more people are going to trip over it. Which is to say, the more people are going to get diagnosed with a problem sooner and then you can drug them sooner. That's a cynical view and I acknowledge it as such, but I don't think I'm wrong.

With A1C — I actually find that perhaps even more problematic, because of what it has enabled the clinician to do, which is to not even tell the patient to fast. I'm collaborating right now with a clinic who is looking at low-carb diets, and we're analyzing all of their clinical data only to find out that the first hundreds or so of patients weren't even getting their blood tested in a fasted state. So everything we just got done talking about — not only fasting glucose, but forget about it if you don't know what they just ate. Insulin, forget about it. Triglycerides, forget about it. All of those are going to be influenced by what they just ate and so they're useless markers. But if you just look at A1C, A1C is not influenced by what they just ate. And so to a clinician who has a glucose-centric view, and many insurances — including Medicare and Medicaid — enable that A1C-centric view where all of the outcomes are pegged to A1C, the clinician feels justified in telling them that they don't even have to fast. Which again negates any other marker that would need to be done in a fasted state.

But it also overlooks the other part of the A1C variable, which is the lifespan of the red blood cell. A1C is the degree to which — the percent of red blood cells that have been glycated, where the glucose molecule has come and irreversibly bound to the hemoglobin, which does have real consequences. If you've glycated the hemoglobin, its ability to carry oxygen is reduced. So that does have a real consequence beyond just being a clinical marker of glucose levels.

But we too often look at A1C as only an indicator of glucose. What if it is also an indicator of how long-lived the red blood cells are? If you have an individual who has long-lived red blood cells, it's just more likely that a red blood cell will have undergone a degree of glycation. So long red blood cells that are living longer in one person — they may have normal blood glucose levels, it's just because they're so long-lived, it's just more likely that a few more are going to get glycated. That ends up being a false positive, if you will, where it looks like they have hyperglycemia but their actual fasted glucose levels are totally normal.

In contrast, you could have an individual who has very fragile red blood cells due to iron deficiency or some other bone marrow problem or some hemolytic anemia. They may be hyperglycemic, but the red blood cells die and are turned over so quickly that they don't have time to undergo glycation. Thus the hyperglycemia is actually missed and they have a false negative — they pass the test, their A1C looks great, but they actually do have a glucose problem.

So I don't like A1C. While I appreciate its utility, it has enabled us to have bad blood tests — done in a fed state — causing us to discount or not even be able to look at any other fasted metabolic markers. And it's also ignoring the complicating factor of the red blood cell.

As an aside, this is one reason why I believe people who go on very nutrient-dense carnivore diets may notice that their A1C is going up all in the midst of having ideal blood glucose levels. I think it's because their red blood cells are living long — they have such an abundance of iron in the diet and such an abundance of all the necessary lipids for the membranes, omega-3s included, that they just have really robust red blood cells.

Mike Haney: So the story that emerges when we think about these markers together — if my insulin is low, that would suggest I have good metabolic health, and therefore my glucose markers should, setting aside food and things that could be influencing it, be healthy. And yet the thing we see a lot from folks who get these things tested is they come to us with confusion because they've read our blog and go, "How come I'm having borderline pre-diabetic A1C but my insulin is still under 10, still kind of healthy? Why do these discordant results happen?"

Ben Bikman: Well, I can only speculate now — I'm leaving the realm of solid peer-reviewed studies behind, unfortunately, only because I'm unaware of evidence that has looked at that particular situation. So I have to speculate a little bit, but it'll all be based in sound science.

One, the easiest answer would be that they're measuring their insulin at the low point — their insulin is actually a little high and it's in that somewhat troublesome range. Now someone would say, "Yeah, but Ben, I'm measuring it multiple times and it's always good." All right. And moreover, the triglyceride-to-HDL ratio may be supporting and giving a supporting number that suggests good insulin sensitivity.

Then I think we get into the realm of speculation. The first point of speculation for me would be what I don't want to give birth to as a new term, but might call glucagon dominance. Throughout this conversation I've actually tried and been successful at not bringing in the alpha cell and glucagon. You'd mentioned the pancreas earlier and I actually wondered whether I ought to bring it up then. Right next to every beta cell is an alpha cell. And the beta cell is making its insulin, trying to reduce blood glucose. The alpha cell is the yang to the yin here — it's creating glucagon, whose singular job is to increase blood glucose. So it puts them at odds. Some people have alpha cell insulin resistance. Dr. Roger Unger identified this and he is really the face of it — a legendary scientist. Anyone who wants to learn more about the glucagon-centric ideology of type 1 diabetes, I implore you to look up the work of Roger Unger.

In fact, we have a paper that we've not published yet — we're working through a massive amount of data that Levels actually funded — and it was to answer the question why is it that some people on a low-carb diet have either a sticky high glucose, or the glucose paradoxically goes up? What we found, by giving people isolated macronutrient meals — a singular load of carbohydrate, a singular load of fat in the form of olive oil, a singular load of protein in the form of whey — is that when people ate whey, some people had a dramatic glucagon response, some people didn't. The people who had a more dramatic glucagon response were the same people who tended to have underlying, constantly elevated glucose levels in the low hundreds even in the midst of a very low carbohydrate diet. So we found that some people were responding to dietary protein with a much higher glucagon response than other people, and those were the same people who tended to have this paradoxical increase in blood glucose levels while adhering to a low carbohydrate diet. It could be this individual component of alpha cells just responding a little too much to the amino acid load — certain amino acids are going to stimulate glucagon release.

So I would say some people appear to be having, for lack of a better term, a glucagon-dominant phenotype where they may be responding a little too much to the protein that they're eating. And I think this is why we partly see some people on the so-called carnivore diet seeing blood glucose levels that actually look surprisingly good — I think you'd see the same thing if they went on like a pure fat diet. Some people just have an enhanced response to protein than other people. Hopefully we'll publish those results soon.

Dynamic testing: the oral glucose tolerance test, CGMs, and glucose curve shapes

Mike Haney: Let's talk about the dynamic test then. So we've got something like an oral glucose tolerance test, which people who are pregnant would undergo — otherwise we don't tend to get those. And then you've got something like a CGM, a continuous glucose monitor, which will tell you your glucose over time. How do you think about using those tools to understand insulin resistance? And are they in some ways more instructive than even fasting insulin because you're essentially watching your insulin response in real time?

Ben Bikman: In fact, I like the way you've phrased that. Could the dynamic glucose be more reflective of a metabolic problem than any of the fasted levels? Yeah, I actually think it could be. As painful as it is for me to put insulin in a corner here, I think any dynamic test with glucose — and of course if you add insulin to that, now you're really cooking with gas, but we can't do that really at the moment, certainly not in someone's own home like the CGM can do — but yeah, I think these same people who may say, "Well, my insulin levels are coming in normal all the time and yet my glucose isn't" — all the more reason to look at the dynamic response. If you are a person who goes and eats 50 to 60 or so grams of carbohydrate like a couple pieces of bread, and you're watching that glucose, and if it hasn't come back down to below 100 in two hours — that's a pretty good sign that there's a problem. You should have seen that drop, especially if your insulin levels are good in the single digits. That should be a reflection of insulin sensitivity, and yet if your glucose tolerance is such that it's not dropped back to below 100 at two hours, you're not responding well to glucose.

And it's important to note that glucose tolerance is not the same as insulin sensitivity. This is why people will say, ignorantly, that a ketogenic or low-carb diet causes physiological insulin resistance. That is laughably wrong, and we ought to laugh at those who say this so that they stop saying it and feel humiliated, because it's just a wrong view. What it is, however, is a reflection of a temporary metabolic inflexibility — which has more to do with insulin production than insulin sensitivity. So that's on the production end of the beta cell.

But all of that to the side: if you don't see that rise and drop within that two-hour span, that's an absolute warning sign. Similarly, if you find you eat a carbohydrate-heavy load and you not only go high but then you go quite low — that's also a sign of an insulin dynamic that is aberrant, where the insulin has gone too high. That's reflecting a hyperinsulinemia in a pre-diabetic state. And there can be some compounding factors, like did you get up and start moving around in the midst of this, which could force you to go a little lower than you would have otherwise. But if you just ate that carbohydrate load and you're sitting down quietly working or reading, you go high and you don't get back down within two hours — but then when you do go down, you go really low — that's another warning. That rebound hypoglycemia is another thing you would only identify in a dynamic case.

Now, Mike, you'd actually mentioned pregnancy — just to honor my darling wife and our three pregnancies. She is a petite gal, so I want to have an emotional connection with any of the moms that may be listening and share my frustration with the way that test is done. It is a standard load of glucose regardless of body type, and that should not be the case. It leads to false positives in petite women like my wife. She is a small body, and for her to clear the same amount of glucose as, say, a 6'2" woman who has 100 more pounds of body mass than her — she's going to struggle. The bigger gal, even if she's big and healthy, just has so much more body with which to clear the glucose. So I think another waving of my magic scientific wand — if I could change healthcare — would be to make that be body-weight-based, a certain number of grams of glucose per kilogram body weight rather than a standard 75 grams regardless of body size. I think that causes a lot of petite healthy women, even low-carb women, to fail that test — not because they can't handle the glucose well, but either because it's a lot for their small body, or they don't have a lot of pre-made insulin on hand to clear the glucose as rapidly, like in the case of a low-carb diet. But then on the other hand, you could have a woman who's a lot bigger who may actually have a problem, but there's so much body to her that she can clear that amount of glucose without it triggering the warning.

Mike Haney: You mentioned something I want to just hit on again because it perked my ears: insulin sensitivity is not the same as glucose tolerance. Can you unpack that a little bit?

Ben Bikman: Yes, absolutely. I believe it's this mistaken view that has led to people saying a ketogenic or low-carb diet causes physiological insulin resistance. I remember the first time I ever saw someone state this on social media and they were preening like a peacock — so proud of invoking this multisyllabic word and sounding so clever. And I couldn't help but chuckle thinking they got it totally wrong. There is such a thing as physiological insulin resistance. But just as a reminder, insulin resistance only happens when there's also hyperinsulinemia. And you put someone on a low-carb or ketogenic diet, and fasting insulin levels plummet to the point that they get off medications — even if they're on an insulin therapy as a type 2 diabetic, their insulin sensitivity gets to such a good range that they get off their insulin and it goes to a normal even single-digit realm. That immediately throws out the idea that they become insulin resistant. Even if you want to call it physiological and try to sound clever, that is not right. There is no insulin resistance without hyperinsulinemia. There is not hyperinsulinemia on a low-carb diet. What there is, is a temporary or transient glucose intolerance.

So you could have an individual eating a mixed macronutrient meal, eating carbs all the time, and then they go take an oral glucose tolerance test and assuming they're still healthy, they pass it just fine. But then they go on a low-carb diet. Six months later they take an oral glucose tolerance test and they don't pass it — it takes them a lot longer to clear that glucose. They may say, "Well, it's because I've become insulin resistant." No. If you give them a load of insulin, they will respond to that insulin exquisitely well. But it's because they don't make as much insulin as they did before. And that's a transient problem.

So insulin will be released in two phases. There's the first phase and then a second phase. The first phase constitutes the insulin that the beta cells have made previously and have in storage — holding on to it ready for the moment the hyperglycemia starts, and then flooding the system with pre-made insulin, then seeing what's happening and making new insulin from scratch to fine-tune the final phase of the response, all in an effort to bring the glucose back down to normal.

When you have restricted carbohydrates for as short as 20 hours, the beta cells are so efficient — and as a person who abhors clutter in the home, I'm sympathetic to this view — the beta cell will look at all of this pre-made insulin cluttering up the space and say, "Well, clearly I didn't need this." And that can happen in as little as one day of carbohydrate restriction. What does the beta cell do? It gets rid of it — it breaks down all of the pre-made insulin back into its component amino acids and does something else with them. Now all of a sudden, if you flood the system with 75 grams of glucose in an oral glucose tolerance test, the beta cells panic, if you will — and I'm being a little silly, you can tell I'm a professor who teaches 18 and 19-year-olds and has kids at home. The beta cell will say, "Well, shoot. I would normally have had all this insulin on hand, I would have been ready, but I didn't expect this glucose load because I hadn't seen a glucose load in the last day." And again, that's all it takes. Now what happens is the glucose level is allowed to get much higher before the beta cell is finally online and now pumping out the insulin that it's making from scratch.

So to say all of this in a simpler way: any carbohydrate restriction, even in as little as a 20-hour fast, is sufficient to have the beta cells get rid of their pre-made insulin. And now you've lost that first-phase insulin response. Thus the person would fail that oral glucose tolerance test — and in that state, I would say they're very insulin sensitive but they are temporarily glucose intolerant. It's almost like a reverse of the classic concept of metabolic inflexibility. The classic view of metabolic inflexibility — as identified by David Kelly and Brett Goodpaster at the University of Pittsburgh when they first talked about this — described people who, when you look at whether they're sugar burning or fat burning, would go into a fasted state and yet still be sugar burning. The conventional view of metabolic inflexibility is being stuck sugar burning. But with a low-carb or ketogenic diet, you actually get into a state where you're at least temporarily stuck in fat burning, because your chronically low insulin levels have facilitated such a state of fat burning that you're actually making ketones. But when you then flood the system with glucose, no surprise that it may take one or two stimuli for the body to say, "Oh, okay, we're back to burning glucose. All right, well then I'm going to get all this insulin on hand again."

And in fact, therein lies the solution. If someone listening adheres to a low-carb diet and they know they have an upcoming oral glucose tolerance test, eat some carbs the day before. Then you will go in and you will pass that test with flying colors. Even someone who's not adhering to a low-carb diet — they may actually work against themselves by thinking, "I'm going to crush this test and fast for 36 hours before I go in." No — you'll fail or you'll be close to failing, because you will have told your beta cells to get rid of all of the pre-made insulin and you'll have lost your first-phase insulin. But that's something that can come back with the stimulus of one time. Eat some carbs, remind the beta cell of what hyperglycemia requires of them. They will put their insulin back in place on the shelves, and so when you go in for that test, they'll have both phases ready to go and you'll pass it really well.

Mike Haney: One more question on the dynamic test because we help folks get a CGM — we help them watch their glucose curve for the first time. And one of the things we're always sort of combating is people freaking out about a spike. How do you think about the shape of a glucose curve? If you're wanting to help somebody understand their metabolic health — which is to say probably their insulin resistance — what shapes are you looking for and what shapes concern you?

Ben Bikman: Well, one of the things I like about the Levels app is that it can appreciate the area under the curve. I do think it's important to point out that, as much as I have a view that carbohydrates are the main dietary problem and I'm very confident in that view, I don't like to imply that any moment of hyperglycemia is problematic — I think that's overly simplistic. To me, it would be how long does it take to come back rather than the peak that would be more reflective of a problem. And then second, what happens when it comes down? Does it stay down or does it now start swinging everywhere up and down for the next four hours?

So that variability is also problematic and it is pathogenic — it's harmful. A paper was published within the past six months finding that retrospectively, when they looked at a group of patients in their cognitive decline, it wasn't underlying hyperglycemia that was the most predictive metabolic variable of early stage Alzheimer's. It was the glycemic variability — high then low, high then low. That itself is where the dynamic monitoring of the CGM becomes tremendously helpful and insightful.

So to me, an individual spike I find quite benign. Now, the more often it's happening, we would have to appreciate that behind every spike would be an insulin response that was a little longer. And the more frequently you're stacking those together, the more you are living a life of constant hyperinsulinemia — and insulin, I believe, is the main driver of insulin resistance. So it comes back to that behind-the-scenes character of insulin. But one individual spike, particularly, of course, if it's food-induced rather than exercise-induced or the noise of a sauna — if it's a food-induced spike, I would say rather than the spike, appreciate the frequency with which it's happening. Are you spiking often? How long does it take to come back down? And then what happens when it comes back down — does it start to lead to a lot of static, or does it just quiet back out again?

Mike Haney: And is that pattern of glucose variability post-spike — as part of the sort of recovery — what is that reflecting within my physiology?

Ben Bikman: I think it's probably reflective of some erratic insulin production — a burst of insulin that went too far, went into a rebound hypoglycemia, which would then lead to a burst of epinephrine or glucagon, which then leads to another shot of insulin because those two overshot. So I think it would be some interplay between insulin at a minimum and then maybe some of these other counter-insulin hormones playing off each other.

Insulin is so good at lowering glucose levels that it takes a small army of other hormones to offset it — glucagon and growth hormone and cortisol and epinephrine. All of those hormones have in common a desire to increase blood glucose. And just that one little insulin at its picomolar concentrations is so powerful that it alone will work to challenge that little mob of hormones. So looking at the dynamic changes, I think it would be some interplay between the two, like bouncing a ping pong ball back and forth. Metabolic dysfunction can manifest not only in that too-high area under the curve — that prolonged spike, meaning you're not getting the rapid or appropriate response that you want — but also a sort of erratic response can also be representative of some level of metabolic dysfunction.

What to actually do about it: managing macros, supplements, and the case for CGMs

Mike Haney: In the last few minutes we have left, I want to get to probably the meat of this, which is what do we do about it? You mentioned carbs before. The really simple story is if you've got insulin resistance, if you've got hyperinsulinemia going on, if you certainly have high glucose — cut the carbs out. One of the things we try to be careful to tell people is we're not saying everybody has to go keto. There are very healthy ways to eat carbs — we're concerned about sugar and refined carbs. So maybe talk a little bit about the carb story, but I also want to talk about what else we can do beyond just cutting carbs.

Ben Bikman: My view is that there are three macronutrients and I think macros matter most, so we've got to manage our macros. I start with carbs because I do believe it's the main culprit. My maxim is: control carbs. Now that can have varying levels of rigor to it. But at a minimum, for the average person who's worried about metabolic health — stop getting your carbs from bags and boxes with barcodes. Whole fruits and vegetables: enjoy them liberally. Eat them — don't drink them — but enjoy them liberally. But then the more processed, the more you avoid them. That's it.

Now, if someone is a raging type 2 diabetic — I once heard an endocrinologist say, and it's something I appreciated, that fruit is a candy to a type 2 diabetic. So there are some fruits they would need to be careful with, like the more tropical fruits — bananas, pineapples, mangoes. But for the most part, whole fruits and vegetables, enjoy them.

So that's control carbs. The next two kind of come together because in nature they do, which is: prioritize protein, and don't fear fat. And by that I mean don't fear the fat that comes with the protein. And more and more I even have a view that revisits the old butter-in-the-coffee approach. I think some people can really thrive by doing what I call fat fasts — they're not eating anything, but they're just drinking in their chocolate electrolyte drink a little butter. I have found I can fast effortlessly if I add a little butter to a hot drink throughout the day.

Now, someone would say, well, you're not technically fasting. I have two versions of a fast as a metabolic scientist: a caloric fast, where you're not taking any calories in; or a metabolic fast, where you are setting up the metabolic milieu in the body to behave as if it's fasting — which is to say, insulin is low. Insulin is the hormone of the fed state, as Dr. George Cahill described it — a legend in the realm of fasting science. It was insulin being high that defined the fed state. It's insulin being low that defined the fasted state. And as far as the cells are concerned, if insulin is low, autophagy is happening and the body is in a fasted state. So I think there's a good justification for — if someone wants to go on a fast and wants it to be somewhat easier — adding a little butter to that warm drink.

So those are the three rules for me pertaining to each of the macros. Now, there are some hacks you can add on. Things like berberine — I'm a huge advocate of apple cider vinegar, it works surprisingly well. And then I'm an advocate of a small brief moment of physical activity after your most carb-heavy meal — just getting up and walking for 10 or 15 minutes. The moment the muscle starts moving, it will start taking in glucose. In the absence of insulin, the contracting muscle has a back door, if you will — insulin, I can't wait on you. I'm hungry because I'm up and moving. I'm just going to pull the glucose in. And you will correct that glycemic excursion much more quickly — even to the point of it being 50% of what it would be otherwise.

I would also say, in the interest of overall health, I actually credit the Levels app as helping me identify one of the most problematic habits I had with sleep. I have always been — well, the moment our kids were born, I became a bad sleeper. And that was born from my desire to just be the best husband in the world and give my darling sleep-deprived wife a break. And so if the kids woke up and they'd already been nursed, I was on night duty. And that has just put me in a position where I'm still a terrible sleeper even though my kids are all well beyond that stage now. But I found that I would go to bed and I'd be wondering: why am I so hot? Why am I so anxious? My heart is pounding. And the easy answer was, well, it's anxiety. And yet I would think, I'm not anxious about anything. Everything's fine. I didn't realize what I was doing every evening when I would indulge in some little treats and I'd go to bed with a hypoglycemic rebound. I credit the CGM for helping me identify this.

Most people don't appreciate that when you are hyperglycemic, you activate your sympathetic nervous system, and the sympathetic nervous system is the fight-or-flight response. That's a terrible time to activate the fight-or-flight response when you're trying to go parasympathetic and go to bed. And so I would say: if you want to indulge in a carbohydrate-heavy meal, put it earlier in the day. For me, the best time to eat if you know you're going to eat a carb-heavy meal — let it be lunch. Or at a minimum, let it just be your early dinner. At worst, it's at bedtime, which unfortunately is when most people find they want carbs the most. No one is sitting around on an evening craving protein and fat — we're craving carbs, and that's the worst time to eat it.

How reversible is insulin resistance? The 90-day case study

Mike Haney: The last thing I want to leave us with, because this is the message that I try to get out to folks when I'm talking to them about this, is how reversible all of this is. There are a lot of things that can go wrong in the body that are awfully hard to move. But the thing I find fascinating about insulin resistance is that — again setting aside type 1 and autoimmune conditions — type 2 and that level of insulin resistance can be changed. We can make the cells pay attention again. We have some evidence that the beta cells can come back from being sort of quiet or offline. How reversible is insulin resistance?

Ben Bikman: Mike, you framed that so well. I even like how careful you were with the word easy, because I'm quick to note that while the principles are simple, the implementation of the principles is not easy — because we start to talk about deep-seated habits, dare I say addictions, when it comes to eating. So I appreciate that, and I want anyone listening to not think that I'm painting this in an overly simplistic way. The concept is simple and remarkably effective, but that doesn't mean it's easy to implement. That's why I'm such an advocate of these external technologies — where we can monitor what's happening and have an objective, quantifiable thing telling us what's actually happening. The CGM becomes its own coach in a way, which is the reason I'm such an advocate of it.

But the concept is simple. The reversibility is the good news — it's sort of the gospel of metabolic health. The good news is that these are very reversible problems because they are lifestyle. So it is lifestyle that caused it. Whether it's the culprit or the cure, it is generally what we're eating, how much we're eating of it, and when and how much we're eating. Back to those manage-the-macros principles.

And then the timeline of it — we actually published a report working with a local clinician who's become a good friend, someone I sort of convinced of this metabolic theory of chronic disease. We did a case study of 11 people in his clinic who were just diagnosed with type 2 diabetes. They were left with a decision: do you want to leave the office with a prescription for a medication, or do you leave with a prescription for changing your lifestyle? According to the three principles I just outlined, these 11 people chose the lifestyle. Within 90 days, not a single one of them had a diabetic marker. So you can take someone who is in the throes of type 2 diabetes, and in as little as 90 days — without a pill popped or a syringe injected — they can have reversed the problem entirely, to the point that there is not a single marker that reflects the disease state.

"The reversibility is the good news — it's sort of the gospel of metabolic health. You can take someone who is in the throes of type 2 diabetes, and in as little as 90 days — without a pill popped or a syringe injected — they can have reversed the problem entirely." — Dr. Ben Bikman

So I would say in most people 90 days is going to be a pretty realistic realm to expect a return. Now, these people had coaching, they had monitoring of the diet — and it depends on how well you're able to fine-tune the procedure and stick to it. But to me, the good news is you can take someone who has been on this pathway to metabolic decline for years and they can get to metabolic health in a substantially shorter time than it took them to get to where they are. I often, when it comes to weight loss, try to encourage people to be sympathetic and kind to themselves and say, "How long did it take you to gain the weight? Give yourself that long to take it off." Although it can happen much faster than that. Even more so when it comes to reversing insulin resistance and this metabolic decline — you don't have to expect that if it took you 10 years to get there, you can't reverse it. It could be well below 10 months to reverse all of it.