Why Nobody Talks About the Carb Side of Metabolic Flexibility
Got a great batch of questions from a reader working through the course material on metabolic flexibility. There were six of them, all related, so instead of splitting these into six separate posts, I’m answering them together here. Buckle up.
1. Why does everyone ignore the carb side of the equation?
The question: most people talk about metabolic flexibility as “can you burn fat.” Almost nobody talks about the flip side—can you actually burn carbs well? We assume that because there’s a lot of carbohydrate in the average diet, we must be good at burning it. The Pop-Tart example says otherwise.
Honestly, I don’t know why other professionals gloss over it. Metabolic flexibility was coined by Dr. David Kelley back in the early 2000s, and I started digging into it around 2006. Every piece of primary literature defines it as both sides—carb use and fat use. My best guess is most people just don’t read primary literature anymore, especially now.
Here’s the mechanism, using Type 2 diabetes as the clearest example (it’s a spectrum—you don’t have to be diagnosed to be somewhere on it). Blood sugar creeps up, A1C creeps up, and the body has a hard time clearing glucose out of the bloodstream. High blood glucose is toxic long-term, but the body is more scared of the acute problem (going hypoglycemic) than the chronic one, so it biases toward running a little high rather than risking a crash.
To deal with that, the body pumps out more insulin, which shoves glucose into tissue—muscle and liver glycogen if you’re lucky, but also into a process called de novo lipogenesis, converting some of that glucose into fat that gets stored as intramuscular triglyceride or contributes to fatty liver. Those local fat deposits then throw off compounds like ceramide that make insulin resistance worse, at both the muscle and liver level. Insulin goes up more to compensate, and now you’re getting squeezed from both ends: the rising insulin that’s trying to fix the glucose problem is simultaneously shutting down your ability to down-regulate and burn fat.
So yes—when you retrain the body to use fat well, you’re very often supporting healthy glucose handling at the same time. They’re not separate systems the way older biochemistry textbooks taught it. They’re intricately related. If someone can’t down-regulate fat oxidation well and then tries fasting or stretching out time between meals, the body pushes appetite up to chase glucose, because it still can’t tolerate going hypoglycemic. That appetite dysregulation shows up constantly in practice—high blood glucose, high A1C, high fasting insulin, high C-peptide, and a genuinely hard time fasting. We’ve put a few of these people on a metabolic cart and watched them simply fail to use fat as fuel. You could target either end of that spectrum: exercise is probably the single biggest lever, along with temporarily reducing dietary carbohydrate, sleep, and the rest of what’s in the course.
2. Does low carb cause a transient insulin resistance that reverses with a short carb refeed?
The question: does low carb create a temporary insulin resistance that clears up with something like 150 grams of carbohydrate a day for 3–7 days—especially in people who’ve restricted carbs for a long time?
You’re correct. Put someone on a low carb or keto diet and you will see what’s called non-pathological insulin resistance at the muscle level. Confusing name, simple idea: it’s not a disease state, it’s a temporary adaptation. The brain is greedy for glucose and can’t run on much else—yes, it can use some ketones, yes, some lactate, but it still needs a glucose floor. If dietary carbohydrate is low, the body pulls from liver glycogen first, and if that runs low, it can make some glucose from protein (gluconeogenesis), but that’s slow—we’re talking hours, not minutes.
So the body’s workaround is to make muscle more insulin resistant, sparing that glucose for the brain instead of letting muscle soak it up. The mechanism is a change in an enzyme called PDH (pyruvate dehydrogenase)—think of it as the gatekeeper to glycolysis, your ability to actually use carbohydrate once it’s available. Go keto (say, 50 grams of carbs a day or less) and you will up-regulate fat burning and produce more ketones. The tradeoff is that PDH enzyme changes, and even if you throw a pile of carbs at the body afterward, it won’t use them well.
This is exactly why the old “keto-adapt then carb-load before the race” strategy for endurance athletes never produced the record-smashing results people expected. It doesn’t work because resetting that PDH enzyme takes a lot longer than a day or two of carb-loading. Muscle biopsies on these athletes show glycogen stores are actually full—it’s not a fuel shortage; it’s an access problem. The carbohydrate side was never retrained to use what’s already there.
The enzyme does reset over time, but nobody has a precise timeline. In my experience, it’s taken anywhere from a few weeks to a few months for people who’ve been strict keto for a while. What I actually do is slowly raise carbohydrate intake and pair it with as much movement as possible—even just walking, because walking drives non-insulin-mediated glucose uptake (contracting muscle can pull glucose out of the blood without needing insulin to do it). Over time, most people get back to handling blood glucose well again—in my experience, that’s taken anywhere from two weeks to two months. If there’s solid data on a specific 3–7 day / 150g protocol reversing this, I’d genuinely like to see it; short-term refeeds like that can work, but I’d want the evidence.
One more note: I like to see carbohydrate intake pushed above 100 grams a day fairly quickly when reintroducing carbs. Somewhere between 50 and 100 total grams a day is what I call the metabolic no-man’s-land—not low enough to be truly keto or get meaningful ketone production, but not high enough to support good carbohydrate handling either.
3. Is there a carbohydrate intake that’s “too low,” and does it differ for women?
The question: people with insulin resistance often get told to go low carb, but low carb itself can dampen glucose handling—which sounds like exactly the trap this approach is designed to avoid by retraining both systems instead of just one. Is there a “too low” threshold, particularly for women of reproductive age who often do better with more carbohydrate for hormonal and cycle reasons?
I’ve tried for decades to give a clean number here and unfortunately it varies. What I typically do is pull someone out of keto and put them at 100–110 grams of total carbohydrate almost overnight. That’s enough to get them out of ketosis and start retraining carbohydrate handling, without being so much that it causes real problems. People might feel a little off for a couple of days—keep fluids and electrolytes up—and I’ll simultaneously try to increase some form of movement, whether that’s NEAT, walking, or resistance training. Carbohydrate then increases gradually from there. Body weight may tick up slightly from glycogen and water, but people should start to feel better, and I’m watching symptoms and exercise performance closely, especially response to a small dose of higher-intensity interval work.
In practice, I see far too many women, especially of reproductive age, who’ve cut carbohydrate too low and are stuck in that no-man’s-land: exhausted, underperforming, disappointed with results, and stuck in a cycle that just repeats. There’s a real range here rather than one number, but chronically low energy and tanked performance are the signals to watch.
4. Is metabolic inflexibility a spectrum rather than a yes/no diagnosis?
The question: inflexibility used to get lumped in with severe insulin resistance as a diagnosis, but it’s increasingly understood as tissue-specific, showing up even in people who aren’t systemically insulin resistant. Is there a range of severity, and can someone be good at one piece (burning glucose, burning fat, or switching between them) while being poor at the other two?
You’re right that it’s tissue-specific. We usually talk about metabolic flexibility at the whole-organism level, but the tissues doing the actual work—liver and muscle—can vary independently. Some glucose does convert to fat via de novo lipogenesis, but in humans that process is relatively minor unless you’re eating several hundred grams of carbohydrate day after day; mostly the body just upregulates carbohydrate oxidation and burns off the excess instead. Since it is tissue-specific, the fix is also tissue-specific: movement is the main lever for muscle, along with sleep and nutrition. For the liver, something like fasted high-intensity interval training can help—deliberately doing glucose-demanding work after an overnight fast when liver glycogen is already low, sometimes getting muscle glycogen down to as low as 40%, which is close to the floor you’d see on a biopsy.
As for whether someone can be good at one piece and bad at the other two—yes, and I’ve seen it in the data. I worked with an elite half-marathon competitor who was excellent at burning carbohydrate but, on a metabolic cart ramp test, never crossed over to using even 50% fat as fuel. On the flip side, the classic keto example shows high fat oxidation but impaired carbohydrate use because of that PDH enzyme issue. These extreme splits do show up in the literature (Jeff Volek’s work is a good reference point here), but they’re not common. The good news is that if you fix both endpoints—carbohydrate metabolism and fat metabolism—the switching between them tends to take care of itself.
5. If insulin resistance means trouble burning carbs, is there a name for trouble burning fat?
The question: insulin resistance describes trouble handling carbohydrate at the cellular level, and since an insulin-resistant cell essentially becomes a “sick cell,” it also has trouble using fat. But is there a recognized name or diagnosis for someone who’s specifically poor at using fat, independent of carbohydrate handling?
It’s a fair and confusing question. Insulin resistance is the dominant framework because it disrupts the whole system, but there isn’t a clean, separate diagnostic label for “poor fat adaptation” the way there is for carbohydrate handling. The closest thing is looking directly at someone’s fat max—either a dedicated metabolic cart protocol or referencing the fat-max research. You can likely train this: fasted training, low-to-moderate intensity exercise, and longer fasting windows (in some cases we’ve gone 18–24 hours) followed by moderate exercise to really push insulin down.
One piece that doesn’t get enough attention: if someone’s VO2 max is genuinely low—say, in the neighborhood of 25 ml/kg/min—trying to optimize their fat-burning is a losing battle. That’s trying to run a big fuel load through a tiny three-cylinder engine. The fix there isn’t fat-max optimization; it’s building a bigger aerobic engine first. Research backs this up: even when the percentage of fat used doesn’t change dramatically, the absolute amount of fat burned per minute goes up substantially once the aerobic engine improves. Practically, that means getting any reasonable estimate of someone’s aerobic capacity—a 12-minute Cooper run test, a 2,000-meter row, a Rockport walk test, whatever’s accessible—and building that up first. In practice, this is the single most common thing I see: someone’s aerobic system, as measured by VO2 max, is just genuinely small, and no amount of fat-max tinkering fixes a tiny engine.
6. What do I think of the Lumen metabolic tracker?
I tested an early Lumen device and met with the founders—genuinely good people. In my case, it told me I don’t use carbohydrate well, which didn’t match my own metabolic cart data at all. Lumen is built on a repurposed CO2 sensor originally from the automotive industry, and the CO2 sensor itself is quite accurate. The catch is that measuring CO2 alone, without oxygen, doesn’t let you calculate a true RER—so you only have half the equation for how much fat versus carbohydrate someone is actually burning. The rest is filled in with algorithms and assumptions.
I want the device to work, because if it did, it would be dramatically cheaper and easier than a metabolic cart that costs thousands of dollars, and the software itself isn’t bad. I just haven’t found the accuracy holds up. People who do report improvement using it are, more often than not, simply in a caloric deficit for the first time, sleeping better, and doing the fundamentals—exercise, sleep, micronutrition—that were covered in the course anyway. They were asked to be an affiliate partner given how much metabolic flexibility content they were pushing, and passed.
Applied knowledge with love,
Dr. Mike
