November 23, 2020

Why is Blood Sugar Sometimes Higher Than Expected on Keto?

We have a numbers problem in the keto community. 

We place too much stock in numbers without a good understanding of what they mean. 

We measure the level of every little molecule without an appreciation for the myriad feedback mechanisms and overlapping biochemical pathways that govern the systemic regulation that determines those levels. We assign way too much meaning to single measurements taken at single points in time rather than looking at longer term trends and patterns. It’s like seeing one frame of a movie—one isolated, frozen frame—and making wild guesses as to what the movie is about. It’s ridiculous. 

Allow me to tell you a story. Sometime in the early 1980s, when I was a little girl, my family took a trip to Disney World in Florida one winter. We lived in New York, so Florida was a way for me and my sister to see Disney World and for us to spend a few days away from the freezing cold and bask in the warmth of the Sunshine State. Well, a freak cold front happened while we were there. Yes, a freak, unexpected cold snap in Florida. Sea World was closed because some of the water was frozen. We had to buy winter coats down there, because who packs winter coats to go on vacation in Florida? 

Imagine if we knew nothing about Florida and based our assessment of Florida’s climate on the short-term weather of those few days. We would have assumed Florida was a bitterly cold place, rather than the hot, steamy swamp it usually is. The lesson here is, don’t mistake an acute occurrence for a general pattern, trend, or tendency. 

What does this have to do with blood glucose? 

(If you’re pressed for time and you want an abbreviated version of my take on this topic, watch this video. If you have a minute or two to read, the first part of this post will give you a brief synopsis of the most salient points. But if you have some extra time to spare and you want the juicy details, cozy up, settle in, and happy reading.) 

Don’t freak out over seeing a blood glucose level you think is “high” on a ketogenic diet. There are numerous reasons why blood glucose (BG) is occasionally higher than you would expect when your carbohydrate intake is very low. Don’t mistake these one-off measurements for the metabolic status of your body most of the time. 

I’ve said before and I’ll say again: I wish people were required to sit through a 5-hour biochemistry lecture before being allowed to purchase a glucometer or ketone meter. It’s fine to want to track some of these data but you have to understand how to interpret the numbers you see. Otherwise, you risk making yourself neurotic about your blood sugar level because you don’t have a basic understanding of the various mechanisms and feedback loops responsible for controlling it. (It’s even worse with ketones.) 

The wide availability of handheld glucometers and meters for measuring blood and breath ketones is a blessing and a curse. The positive aspect is that this technology can provide people with valuable feedback about which foods or combinations of foods work best for their body. The negative aspect is that with such easy access to these tools, people can get carried away. They might measure too often and jump to unscientific conclusions because they don’t appreciate how wonderfully complex the human body is. Worse, they become stressed out, scared, and worried, when they’re actually in perfect health and there’s nothing wrong with their body. 

Not everyone tracks blood glucose on a low carb or ketogenic diet, but people who do are sometimes surprised by “high” blood sugar. I have “high” in quotation marks because the blood sugar level they find alarming isn’t high-high; it’s just higher than they expect to see based on their very low carbohydrate intake. What gives? Are these occasional blips on the radar increasing risk for type 2 diabetes and other metabolic illnesses? Is a slightly higher glucose now and then going to glycate every single protein in your body and put you on the fast track to neuropathy, nephropathy, blindness, amputations, and early death? Is an occasional slightly elevated blood glucose something to be worried about? In most cases, probably not. 

Before I go on, let me say right here that measuring your glucose or ketones is absolutely not required to be successful with a low carb/ketogenic way of eating. People were using carbohydrate restricted diets and doing great long before these technologies existed. If you like tracking the numbers and compiling data, then cool. Do it. But if you’re like me, and you’re not into putting your blood glucose, blood, breath and urine ketones, heart rate, REM sleep, and daily steps taken into a spreadsheet, please know that not doing any of this is totally cool, too. (Note: testing blood glucose and possibly ketones is important if you have type 1 diabetes or are taking insulin and/or oral medications for type 2 diabetes. For fat loss and overall general health, though, it’s not necessary.) 

I’ve had the privilege of getting to know some of the low-carb and keto-friendly physicians in our community and I’ve been surprised—pleasantly surprised—at some of their perspectives on blood glucose testing. Many of them just aren’t phased by a glucose that some people might consider high—or, at least, high for someone following a ketogenic diet. There’s no definitive cutoff point for what they would consider problematic; it’s more of a spectrum. If your fasting BG is 200 mg/dL (11.1 mmol/L), obviously that’s a problem. But 110 mg/dL (6.1 mmol/L)? Even 120 (6.7 mmol/L)? Without knowing anything else about a person’s individual situation, these numbers, by themselves, are not necessarily concerning. 

If someone in your favorite keto Facebook group, Reddit forum, Twitter feed, or wherever else, throws out a range of blood glucose that you must stay within, or makes a claim that a certain hemoglobin A1c is “too high,” please take all that with a grain of salt. Your blood glucose level at any point in time is the product of numerous processes and crosstalk between several organs and tissues, such as your liver, pancreas, skeletal muscle and adipose tissue (body fat). Sometimes it’s elevated because it’s totally normal for it to be elevated under that particular circumstance. Your blood sugar isn’t just some random number completely unrelated to anything else in your body. If anything, it’s the exact opposite: barring the presence of type 1 or type 2 diabetes, your body regulates blood sugar very well and if you are metabolically healthy, if your BG is a little higher than you expect, it’s probably at the level it’s at because that’s where your body needs it to be at that particular moment in time. 

This is what’s missing from discussions of “optimal” BG: respect for normal, necessary fluctuations. When you measure BG, you see what’s in your blood at that one point in time. You don’t know how much is being used by various cells types. Maybe there’s some extra glucose flowing through your blood because certain cells need it at that particular time, and it needs to travel through your bloodstream to get there. 

If you’re on a very low-carb or ketogenic diet and your BG is occasionally higher than you expect it to be, what, exactly, are you worried about? If your carb intake is very low, then you know your food isn’t what’s raising your BG. So if it’s not your food, then what is it? Where is that glucose coming from? From your own body! I’ve heard Dr. Jason Fung and Megan Ramos talk about some of their fasting patients, and how people doing extended fasts can occasionally see an unexpectedly high blood sugar. They’re fasting. No food! So if that glucose isn’t coming from food, it’s coming from inside their bodies—either being released from a place it was stored in, or their bodies are generating it anew for some reason. Either way, whatever that glucose is doing there, the body probably has a reason for releasing or generating it, and it’s probably better than if it had come from a blueberry muffin or glass of apple juice. 

The point is, blood glucose is very much context dependent. And unless you have an understanding of and an appreciation for these contexts, you will drive yourself crazy and terrify yourself into catatonia every time you prick your finger and see a number you think is “high.” 

The unfortunate funny thing is, from what I see, for the most part, the people measuring and tracking obsessively are the people who have the least need to do so. They’re metabolically healthy, athletic, lean, active, and are the least likely candidates for metabolic illness driven by chronic hyperinsulinemia or hyperglycemia. They’re not sick and they’re not overweight. But boy, do they loooove to measure stuff. And good for them. I have nothing against measuring. What I want, as I said, is for people to understand how to interpret the numbers they see. Otherwise, they spin themselves into a tizzy and send me terrified emails asking why their blood sugar “skyrocketed” to 108 mg/dL after a fasted workout. *Sigh.* 

Please stop being terrified of having glucose in your bloodstream. It’s normal to have glucose in your bloodstream. And depending on the context, it’s normal to have a “high” amount of glucose in your bloodstream. What you don’t want is high glucose all the time. If your glucose is high once in a while—and you’re on a very-low-carb or ketogenic diet—there’s probably a perfectly logical explanation for why it’s high, and instead of freaking out, you can be grateful that your body is working exactly as it’s supposed to. 

With that enormous preamble out of the way, let’s get into things.  

You’ll see and hear all sorts of wacky things on the low carb and keto interwebs these days. Try to keep in mind that the more adamant and dogmatic someone is about a particular claim, the less likely they are to have an appreciation for context, nuance, and the complexities of human biochemistry and physiology. These three things—context, nuance, and the complexity of the human body—will be a recurring theme throughout this post. I’ve tried to make it a recurring theme on this entire blog, in fact. Not everyone needs a strict ketogenic diet to lose weight and be healthy. Not everyone needs to fast. Some people do just fine on higher carb diets. 

The vast majority of people who adopt a low-carb or ketogenic diet experience dramatic improvements in blood glucose control. In fact, carbohydrate restriction is so effective for type 2 diabetes that researchers have said it should be “the first approach in diabetes management and “the default treatment” for type 2 diabetes (T2D) and metabolic syndrome. Keto is so effective for keeping blood glucose in a healthy range that many people with T2D who adopt this way of eating are able to reduce or eliminate many of their medications, including insulin. 

However, after some length of time on a low-carb or keto diet, a small number of people see a rise in fasting blood glucose (FBG). And healthy people and people with diabetes alike might measure glucose at random points throughout the day and find it unexpectedly elevated—again, not overtly high, but higher than they expect based on their diet and lifestyle. Considering that elevated FBG is part of the diagnostic criteria for both type 2 diabetes and metabolic syndrome, should they be worried? 

As is true for so much in health and nutrition, it’s all about context. A fasting blood glucose that’s higher than one would typically expect in someone adhering long-term to a very low-carb diet doesn’t automatically indicate anything pathological. There are many reasons why FBG might be elevated, and many ways to assess metabolic health and gluco-regulation beyond this one measurement. I cannot stress this point enough: your health is not reflected by a single number from a single blood draw at one particular moment on one particular day. Rather than these momentary snapshots, look at trends over time, and look at several different measurements together to get a better picture of things. Health is a mosaic. Mosaics are made up of many individual tiles. If you look at only one tile, you’ll have no idea what the whole mosaic looks like. Blood glucose—whether it’s your fasting level or the level measured at a random point during the day—is just one tile. It tells you very little about what the whole picture looks like. 

Do not—repeat, do not—become alarmed over one measurement in isolation.  

With this in mind, let’s look at four reasons why blood glucose is sometimes “high” on a ketogenic diet. And we’ll end with exploring ways to fill in what the rest of the mosaic looks like—that is, other things to look at besides blood glucose, to get a better sense of overall metabolic health. 


Reasons for Elevated Blood Glucose on Keto


Dawn Phenomenon


The dawn phenomenon is a situation in which the first morning blood sugar measurement is high, possibly even being in the range of pre-diabetes or diabetes. Dawn phenomenon is the body’s response to a hormonal surge that occurs in the very early morning hours. A couple of hours before we wake up, our bodies release cortisol, epinephrine (adrenaline), norepinephrine, and other hormones that prompt us to wake up and get moving. (Coffee, of course, does the rest, at least for me.) These hormones are responsible for numerous effects, one of which is raising blood glucose.  

In a non-diabetic person, this slightly elevated BG comes back down quickly—quickly enough that by the time they check it, it’s already back to normal. In someone with T2D, however, the glucose stays elevated for longer. In this situation, when the person checks their morning BG, it’s still elevated from that normal early morning hormonal release. To emphasize, this early morning rise in BG happens in many people, including healthy people with good insulin sensitivity. Dawn phenomenon can happen even when these individuals follow a low-carb diet, and even when their overall blood glucose management is much better than it was before.


Should You Worry About Dawn Phenomenon? 

When someone’s first morning glucose reading is higher than normal even after being on a ketogenic diet for a while, gluco-regulation as a whole is most likely vastly improved. If fasting glucose remains stubbornly high, it’s not automatically a sign that that someone’s diabetes is not improving. (And if someone doesn’t have diabetes but has an elevated FBG anyway, it’s still not necessarily reason for concern. More on this toward the end of the post, when I talk about HOMA-IR.) 

Diabetic or not, for people concerned about a higher-than-expected fasted morning glucose, the best way to assess glucose control is to test BG at various times throughout the day to get a better picture for what the levels are most of the time. Testing before and after meals can reveal how different foods and food combinations affect glucose levels. Peri-mealtime and random testing doesn’t have to be done forever; a few days or weeks of measuring will give a more accurate picture than the fasting level alone. People with a diabetes diagnosis can get a continuous glucose monitor (CGM) prescribed for them. These devices report BG throughout the day and night in near-real-time. (Non-diabetics might also be able to test out a CGM, but a prescription or doctor’s requisition is usually required to get one.) 

Your blood glucose level throughout the day is more important than the level first thing in the morning. A few years back, I explained that exercise is a pretty ineffective tool for fat loss, because what matters more: the “calories you burn” during one, two, or even three hours at the gym, or what your body is doing metabolically the other 21, 22, or 23 hours of the day? So with regard to BG, which should you be more concerned about: your BG at just one moment first thing in the morning, or the range it’s staying in throughout the whole rest of the day?   

Bottom line: assessing BG control is best done by looking at a few different measurements together, rather than judging solely by the fasting level. 

(Click here to read a nice explanation of the dawn phenomenon from Dr. Jason Fung.)


Ways to Improve Dawn Phenomenon 

If your hemoglobin A1c falls within the healthy range and your postprandial (after meal) glucose levels are normal, you don’t really need to do anything at all. (More on HbA1c in a bit.)  But if you’re concerned about a high fasting BG and you’d feel better—and be less stressed out—seeing a lower number first thing in the morning, there are ways to reduce the severity of the dawn phenomenon. One that might surprise you is having a bedtime snack. Contrary to what you might assume, having a small, protein-rich snack before bed tends to result in lower FBG the next morning. Good options include a hard-boiled egg or two, plain, full-fat Greek yogurt, a small piece of leftover chicken or steak, or any other small amount of high-quality protein-rich food. 

Something surprising that also does the trick for many people is taking in a little bit of vinegar at bedtime. When taken with a meal, vinegar has some very interesting properties with regard to lowering BG and insulin levels, but it also appears to have these effects even when taken with just a very small amount of food. One study showed that type 2 diabetics who consumed 2 tablespoons of apple cider vinegar along with 1 ounce of cheese at bedtime had greater reductions in fasting glucose than subjects who consumed the cheese with water. The reduction wasn’t large (only about 4-6%), but if you’re especially worried about dawn phenomenon (and the stress this causes you could be making things worse), this little trick might be worth trying.

One of the keto doctors I know advises his patients with dawn phenomenon to do a few jumping jacks or jog in place for a couple minutes, rest a bit and then test BG. Give that bit of glucose “somewhere to go,” so to speak. Let the working muscles take it up, and then when you test, your BG will likely be within the normal range. 

All that being said, I want to emphasize that a slightly elevated FBG when other glucose measurements are within a healthy range—especially the postprandial levels—is not something to become alarmed about. If it’s especially high, this could be a sign that someone should take additional steps to control their diabetes a bit better. But when it’s only slightly above the normal range, a keto-savvy physician familiar with the person’s medical history and current diet and lifestyle can weigh in as to whether it’s a threat to long term health. Because of the unfortunate side-effects many diabetes medications have, some physicians prefer their patients to have slightly higher BG off of medication (or on lower doses of medication) than to have lower BG achieved through heavy medication.  

According to Jenny Ruhl, author of Blood Sugar 101: What They Don't Tell You About Diabetes, slightly higher FBG doesn’t necessarily mean someone’s at risk for the long-term complications of poorly managed diabetes. From her article on dawn phenomenon, which is highly recommended reading: 

“The reason that fasting blood sugars over 100 mg/dl (5.6 mmol/L) are associated with poor outcomes in the research is that most people who have those kinds of fasting blood sugars also have post-meal blood sugars surging up to 180 mg/dl (10 mmol/L) or higher after every meal, which we know does cause complications over time. But if your elevated fasting numbers are not associated with big post meal numbers, and your readings throughout the rest of the day are mostly under 110 mg/dl (6.1 mmol/L) a few hours after eating, you should be fine.” 

As always, work with your physician to determine the course of action that’s right for you.


Physiological Insulin Resistance, a.k.a. “Adaptive Glucose Sparing” 

 Another reason for higher than expected fasting glucose is a phenomenon called physiological insulin resistance. This sounds like something bad, so before you get scared, let me explain. This is a phenomenon that occurs in some people who’ve followed a very low-carb or ketogenic diet for a significant length of time. In people who are keto-adapted or fat-adapted—that is, they’ve adhered to a very low carbohydrate intake for a long time—most of the body’s cells run happily on fatty acids and ketones, with a much lower requirement for glucose than in people on a higher carb diet. Since these cells are fueled mostly by fats and ketones, they need only minimal glucose.  

The majority of glucose is spared for tissues with a requirement for it, such as the brain and neuronal tissue. The biggest sink for glucose is your skeletal muscle tissue. (We use the term “skeletal” muscle for things like your glutes, hamstrings, biceps, quadriceps, biceps, etc. This distinguishes these muscles from cardiac muscle [your heart] and smooth muscle [muscle cells that line your GI tract and blood vessels].) With skeletal muscle tissue “refusing” the glucose in order to keep it available for the brain and other glucose-dependent tissues, the early morning hormone surge that releases glucose (even in someone on a keto diet) can lead to the first morning BG being a little higher than you would expect. This is called “physiological insulin resistance” in order to differentiate it from the harmful or pathological insulin resistance we associate with metabolic syndrome and T2D. But to avoid confusion, a better name for it is “adaptive glucose sparing.” It’s an adaptation some people’s bodies make as a healthy, normal, and logical response to a very low carbohydrate intake, to spare glucose for cells that must have it. (I don’t think this is an accepted medical term. Dave Feldman is the first person I ever heard say it. He may have even coined the term; I don’t know for sure.)   

In fact, this adaptation is well-recognized enough that if someone who’s been on a low-carb or ketogenic diet for a substantial period of time must undergo an oral glucose tolerance test (OGTT) for some reason, keto-savvy physicians recommend they “carb up” for several days beforehand, to get their bodies readjusted to handling large amounts of glucose. (I’ve heard recommendations for as much as 150 grams of carbs per day for 5-7 days before the test.) Without this re-acclimation period, it’s possible that the keto dieter will get results indicative of diabetes when the truth is, they’re completely healthy. (Again, most of your tissues are humming away handily on fats and ketones, so they don’t need a whole lot of glucose. So if you all of a sudden drink 75-100 grams of liquid glucose after being on a ketogenic diet for a few months or years, of course your BG is going to be sky-high. You will appear to be diabetic when your body is simply accustomed to predominantly using fuels other than glucose.)  

If you’d like to learn more about this idea, I recommend this article on physiological insulin resistance by veterinary doctor Petro “Peter” Dobromylskyj, author of the popular, science-heavy Hyperlipid blog. (Hyperlipid is very technical and scientific. I often comprehend only about 70% of what Peter writes. I wish there were a “Hyperlipid for Dummies” blog where someone would translate his work into plain English. Until someone creates that, I try to glean whatever nuggets of wisdom I can from the parts I do comprehend.) He mentions typically measuring his own FBG around 100-120 mg/dL (5.6-6.7 mmol/L), yet his HbA1c is very low, at 4.4%. His usual diet is less than 50g total carbohydrate and high in saturated fat, and he’s in ketosis more often than not. With such a low HbA1c, his fasting glucose is not indicative of constantly high BG throughout the rest of the day, nor over the course of weeks or months. If he only ever looked at his FBG and nothing else, he would have a wildly false sense of his metabolic status—including possibly thinking he was pre-diabetic! (Remember the Florida thing: short-term weather versus long-term climate.) I say again for emphasis: your fasting blood glucose level doesn’t necessarily indicate anything about the typical range of your BG throughout the rest of the day, and the latter is far more important. 

Don’t be alarmed by one measurement in isolation. If you’ve been keto for a while and your FBG is higher than you’d expect, look at the total picture of your metabolic health, including HbA1c, postprandial glucose, INSULIN, and other measurements your physician might suggest. 



You can’t read a health-oriented magazine or listen to a podcast in the same genre without hearing the message that stress negatively impacts health. If you’re still a doubter, checking your blood sugar after a stressful commute will make a believer of you.  

If your morning commute is typically a major source of stress for you, measure your blood sugar before you leave home and measure again when you get to your destination—before getting out of your car, if possible. You might see a substantial rise. Clearly this isn’t from food, because you didn’t eat anything in the car. It’s solely the stress response. (Easier to see this if you don’t eat in the morning, but it would likely also happen even you ate a morning meal, as long as it was very low carb.) I mentioned in this video I did on higher BG on keto that my friend Colin Champ, MD, has seen his BG rise as much as 30 points just from a stressful commute. 

This is the “fight or flight” mechanism—your sympathetic nervous system kicking into gear to help you survive what the ancient, hard-wired parts of your brain perceive as an emergency. You might not have noticed, but your heart rate was probably up a bit at the height of the stressful incident, and your blood pressure probably was, too. The stress hormones—adrenaline, cortisol, and others—signal your body to release glucose in order to give you quick-acting energy to help you “fight” for your life or “flee.” This is an evolutionarily hard-wired mechanism that served us well in prehistoric times. In fact, if you’re reading this right now, it’s in part because your hominid ancestors survived life-threatening situations thanks to this neat trick of the body flooding itself with glucose at a moment’s notice. The fact that you’re alive right now means your ancestors survived every accident and wild animal encounter they ever experienced. 

Unfortunately, this powerful survival mechanism doesn’t serve us so well in the 21st Century. You’ve probably never had to run from a wild animal with its fangs bared and its claws reaching out for you, but you have probably gotten caught in an aggravating traffic jam or had a heated argument with a family member or coworker. These are our modern-day stressors, and our bodies think they’re doing us a favor by flooding our blood with glucose. A few hundred thousand years ago, they were. The problem today is that we don’t need a flood of glucose to give us an energy surge when we’re stuck behind the wheel of a car or sitting at our desk or kitchen table, stewing in anger.   

The influence of stress on blood glucose levels goes to show that many things affect BG besides the foods we eat. Another non-food factor that influences BG is exercise.



Many people are surprised to see their BG elevated after an intense exercise session. At first this seems paradoxical—after all, generally speaking, exercise is supposed to improve insulin sensitivity and lower blood sugar, right? Overall, this is true, but in the acute sense—during and just after your workout—it’s normal for your BG to be on the higher side. NORMAL. Exercise is great for managing blood sugar. When you exercise regularly, your blood sugar will generally be lower. But during exercise, it’ll be higher. 

The first thing to know is, glucose isn’t poison. You do need some of it in your blood. Like I said earlier, it’s all about context. Too much of anything can be deadly, including water and oxygen. No matter how keto- or fat-adapted someone is, the body still uses some glucose during exercise. And the higher the intensity of exercise, the higher the demand for glucose, even in fat-adapted athletes. (Fat-and keto-adapted athletes use more fat during exercise than carb-adapted athletes do, but they still use some glucose. I don’t think anyone has ever shown that human metabolism, whether during a workout or when someone’s just sitting around—can be fueled 100% by fat. You’re always oxidizing at least a small amount of glucose – and that’s okay!) 

The second thing to know is, exercise increases the body’s demand for glucose. Some of the glucose used to fuel the muscles will come from glycogen, the carbohydrate stored in muscle tissue. If the muscles burn through their glycogen and still need more glucose, this is where the liver takes charge. The liver can break down its own stored glycogen into glucose and release that glucose into the bloodstream. It can also make more glucose through gluconeogenesis. (Gluconeogenesis isn’t always a bad thing! Stop being afraid of it.) Because your intensely exercising muscles need more glucose during exercise, there’ll be more of it in your blood. Instead of feeling worried by this, be grateful that your body is doing exactly what it’s supposed to do in this situation. Again: it’s all about context. This rise in BG is temporary and totally normal, and it’s helping you power through the activity you’re engaged in. This kind of transient rise is a world apart from chronically high blood sugar (hyperglycemia) all day, every day, in someone who consumes a high-carb diet and is sedentary. Please stop being afraid of normal human physiology. 

Nutritionist and athletic coach Alessandro “Alex” Ferretti reports seeing elevated fasting glucose in many of the athletes he trains, especially the ones on low-carb diets. These are high-level performers, yet they have FBG that would put them in the pre-diabetic category. This is likely an example of the adaptive glucose sparing or physiological insulin resistance discussed earlier, because by all other accounts, these individuals appear to be in excellent metabolic and cardiovascular health. 

In a study looking at small cohorts of elite, highly-trained ultra-endurance runners who were adapted to either a high-carb diet or a ketogenic diet, fasting glucose levels were 95 mg/dL (5.3 mmol/L) for the high-carb group and 96 mg/dL for the keto group. These levels aren’t “high,” but they’re higher than we might expect for elite athletes, especially the ones in the keto group. It goes to show that the human body is highly adaptive and that we can’t take numbers at face value without understanding why they are what they are. (Interestingly, even with nearly identical FBG to the high-carb athletes, the keto athletes had slightly lower fasting insulin levels, substantially higher HDL, and lower ratios of triglycerides-to-HDL, a key marker for heart disease risk. So even though their fasting glucose was basically the same, the keto athletes appeared to be in better metabolic health. Look at the overall pattern, not the one number in isolation. The whole mosaic, not one tile.) 

I’m a big fan of Dr. Rob Cywes, who’s one of the few people in our community who talks about carb addiction. (Check out his YouTube channel, and I recommend listening to pretty much any podcast interview he’s ever done.) If you want to hear from a true expert about the effects of exercise on BG in someone who is highly keto-adapted, I cannot recommend this interview highly enough. He described how his BG almost doubles during a run, going from 60s-70s fasted to as high as 120. During a swim, it rises even higher—to as high as 220 mg/dL. Should he be worried about this? Is someone whose BG is in the double digits the vast majority of the time be worried about it being high during intense activity in which his body needs more glucose? My guess is no. 

And here are two tweets from people in similar situations: 

A keto-adapted woman whose BG is normally within a tight (low) range, and her BG practically doubles during a cardio boxing session. 

A keto-adapted man whose BG doubles during a hard 5km run. He’s a physician, and he called this a “predictable (normal) brief spike in blood sugar during a hard 5km run.” YES! Predictable, normal, and brief. Thank you, Dr. Tobin, for sanity and trying to convey to the public that this is basic human physiology. 

Okay, so if fasting glucose or a one-time glucose snapshot at any other part of the day, for that matter, is not all that reliable an indicator of anything, what should you look at?  


HbA1c – Better, But…

If fasting glucose isn’t a very good indicator of your blood sugar control, you might be wondering if HbA1c is any better. (HbA1c is used as an approximation of your average BG over the previous 3 months or so.) In some ways, yes, it is better. As one example, let’s say your fasting glucose is always normal, but your HbA1c is in the pre-diabetic or diabetic range. That would suggest that your BG is elevated most of the day, but it just so happens to come back down to normal overnight so that when you test in the morning, it always looks normal. If you only ever tested FBG, you would miss this fairly major issue. 

As a second example, let’s go back to Peter D, whom I mentioned earlier. As he tells it, his FBG is typically around 100-120 mg/dL (5.6-6.7 mmol/L), yet his HbA1c is only 4.4%, which works out to an estimated average glucose of 80 mg/dL (4.4 mmol/L). He would have the opposite issue from that above if he only ever tested FBG: he would think his blood sugar is typically much higher than it really is. 

However, HbA1c is not without flaws and shortcomings. There are many reasons why HbA1c might be higher or lower than expected, regardless of whether someone’s following a ketogenic diet. (Check out this article I wrote on this topic for details.) The much more serious issue, however, is that in quite literally millions of people, HbA1c may be normal, but insulin is through the roof. (The chronically high insulin is keeping the blood glucose—and therefore HbA1c—in check.) Chronically elevated insulin (a.k.a. hyperinsulinemia) even when blood sugar is normal is a major factor in cardiometabolic illnesses, such as hypertension, gout, PCOS, skin problems, erectile dysfunction, enlarged prostate, and more. (I wrote a multi-part blog series on this critical issue a while back. If this is all new to you and you want to get up to speed, start here.) Looking at HbA1c tells you nothing about insulin.

So instead of looking at FBG or HbA1c, both of which indicate only glucose, why not look at insulin, too? This is what HOMA-IR does.



HOMA-IR stands for homeostatic model assessment of insulin resistance. Where HOMA-IR shines over fasting glucose and HbA1c is that it takes insulin into account: How much insulin is required to keep your blood glucose in a normal range? That is, how hard does your body have to work in order to maintain homeostasis? (Hence, the “homeostatic” model assessment.) 

It’s not an exaggeration to say that in a staggering number of people, FBG and HbA1c are normal because dangerously high insulin levels are keeping them in check. Through testing only FBG and HbA1c, potentially millions of people with impaired insulin sensitivity are lulled into a false sense of security with regard to their metabolic health. HOMA-IR gives a clearer picture and it can help identify patients at risk for myriad conditions stemming from chronically elevated insulin, even when blood glucose is normal. 

Here’s how HOMA-IR is calculated:

  • Excellent insulin sensitivity:  ≤ 1
  • Average insulin sensitivity:  1.75
  • Insulin resistant:  ≥ 2.75 

Here’s an example of HOMA-IR in action: 

Patient A’s fasting glucose is higher than Patient B’s, but Patient A’s insulin is much lower. By taking both glucose and insulin into account, the HOMA-IR scores show that even with a lower fasting glucose, Patient B is at greater risk for metabolic complications down the road. Their body has to work harder and they require much more insulin in order to maintain a normal glucose level.   

Some of the keto-oriented MDs I know are more concerned with patients’ insulin levels than with their glucose. Chronically high glucose is a problem, don’t get me wrong. But even when glucose is normal, chronically high insulin is probably an even bigger problem. And glucose is so much more fickle than most people realize. Putting too much stock in it can cause unnecessary alarm for a patient, for all the reasons I’ve mentioned in this post. 

Here’s a great graphical representation of this, courtesy of Ted Naiman, MD:

What Does It All Mean?


Understand that fluctuations in blood glucose are normal, even on a ketogenic diet. Your diet obviously has an enormous impact on your BG, but it’s not the only thing that affects it. If you see a measurement that surprises you, think about the context. Did you eat something that had more carbs than you realized? Were you under substantial emotional stress? Did you just finish exercising? Had you been skimping on sleep? Any of these can make your BG a little higher than you might’ve been expecting.  

As for fasting glucose, a slight elevation in long-time low-carbers or keto dieters is typically not cause for concern. The full picture of cardiometabolic health should be considered so that each measurement—be it glucose, HbA1c, insulin, triglycerides, LDL, or something else—can be interpreted as part of a dynamic system and within the proper context. Each of these—plus about a hundred measurable things I haven’t even mentioned—is just one tile in the mosaic of your health. You can’t know what the full image looks like by studying only one tile. 

If you’re concerned about your blood glucose management, work with a doctor to see if there are diet and lifestyle changes that can get you on a better path.  

I’m constantly saying that low carb/keto is supposed to make your life easierSimpler. More enjoyable. I see the constant data mining doing the opposite for many people: they’re nervous, worried, anxious, and they narrow their entire existence to keeping certain numbers within a certain range. Why? Keto is supposed to make your life better; it’s not supposed to be your life.

P.S. If you’d like to read a bit more about this, I recommend this article from, the most popular LCHF/keto site in the world: Is your fasting blood glucose higher on low carb or keto? Five things to know.  



Disclaimer: Amy Berger, MS, CNS, is not a physician and Tuit Nutrition, LLC, is not a medical practice. The information contained on this site is not intended to diagnose, treat, cure, or prevent any medical condition and is not to be used as a substitute for the care and guidance of a physician. Links in this post and all others may direct you to, where I will receive a small amount of the purchase price of any items you buy through my affiliate links.


  1. Thank you Amy,
    This is a wonderful blog post. All of what you describe I am experiencing. Stress is the biggest one.

  2. very good summary... thank you. one comment though: it is not just "rotten alcohol" but the real thing that lowers blood sugar for the morning. I guess it is better be sugar free like dry wines and certain boozes, in moderation. port wine with 20% alcohol and 10% sugar {5% fructose} must be best gout generators there is... medieval vodka coke.

  3. πŸ‘πŸ‘πŸ‘πŸ‘πŸ‘πŸ‘πŸ‘πŸ‘πŸ‘πŸ‘πŸ‘ŒπŸ‘ŒπŸ‘ŒπŸ‘πŸ‘πŸ‘

  4. Many Thanks Amy. With this amazing post you've really answered all my questions. πŸ™πŸ™πŸ™

  5. Thank you, I panicked this morning due to my keto fast blood glucose reading of 7.1. I shall continue to test throughout the day for a more accurate picture.

  6. perfect analysis Amy, the best out there

  7. I think if you'd warned us how putdowny you were going to be for the first three pages I wouldn't have bothered. And in the end, you don't deal with the issue. If I am not consuming glucose or high protein, how is my body manufacturing that glucose? How. And if I'm supposed to be able to run my brain on ketones, why are my muscles refusing glucose. YOu are highly interested in morning glucose, which is often created by overeating carbs the night before - but don't address anything else. I am diabetic, I have been on under 30 carbs a day for 3 weeks, when are my fasting bloods going to come under 6.5 and how is that going to take place. That is the question you need to answer. Not just a long speech about people being fixated and neurotic about tracking sugars. I am never going to get this 20 minutes back.

  8. Great explanations on FBS and A1c. i agree with you the high risk of a CGM creating fear and anxiety when a person is not familiar with interpreting the numbers correctly. Thank you for this informative document. It was easy to read and understand. I am so grateful to you for sharing your wisdom. Thank you Amy BergerπŸ˜€

    1. Thank you for such a nice comment! I wish you weren't "anonymous" so I could thank you personally. But since I can't, please know I appreciate it. :-)

  9. Hi! Thank you so much for your work! I would love for you to address measuring insulin by way of the glucose ketone index/GKI. I use the KetoMojo meter and want to interpret the test results correctly. Any help is appreciated!

    1. Before I comment, I would want to know *why* you are measuring your GKI. What is your goal in following a ketogenic diet? My answer for you would be different if you're using keto mainly for weight loss or overall metabolic health versus using it as an adjunct therapy in cancer treatment or for a psychiatric issue. Most people don't need to ever even think about the GKI, but again, it depends on why you're following a ketogenic diet. Feel free to reply here or contact me here: