There was a little girl,
Who had a little curl,
Right in the middle of her forehead
When she was good
She was very good indeed
But when she was bad she was horrid.
So goes the familiar poem by Henry Wadsworth Longfellow, and so goes most of our low-carb diets I suspect. When we are in diet mode and are deeply committed, we are very good indeed, but when we break and hit the carbs (and this includes yours truly) we are horrid. A paper in this week’s JAMA presents data confirming what I’ve long suspected: carb bingeing now and then could actually cause worse free radical damage than regularly eating more carbs on an ongoing basis.
The paper titled Activation of Oxidative Stress by Acute Glucose Fluctuations Compared with Sustained Chronic Hyperglycemia in Patients With Type II Diabetes shows that patients with diabetes who have fluctuations in their blood sugars incur more free radical damage than those with high but not fluctuating blood sugar levels. Before we get into the nitty gritty of the paper, I will step back a little and go over some of the biochemistry involved so that the data will be more understandable to those who might not have a technical background.
First, let’s look at how blood sugar is measured. The standard way is to draw blood (usually after the patient has fasted for 12-14 hours) and test it for a blood sugar level. Although this is the way that diabetics were monitored for years, it’s not a particularly accurate means of managing patients with diabetes. The fasting blood sugar test itself is reliable, but it only tells what the blood sugar is at the precise moment the blood was drawn–not what it was the week before or the day before or even the hour before. The typical drill was that diabetic patients would come to the office, have their blood sugar levles checked, and based on the results, leave with their medications or insulin adjusted as necessary. Many diabetic patients who didn’t like getting lectured by their doctors on following the proper diet, exercising, etc., learned that if they were very good indeed for the few days before their office visit, their blood sugar levels would be close to normal. They would leave with the congratulations of their doctors ringing in their ears and head back to the trough until a few days before their next visit.
A while back a clever person figured a way to detect blood sugar levels not just at the precise moment the blood was drawn but in a way that measures the average blood sugar level for the previous month or two. The test, called a hemoglobin A1C (pronounced, logically enough, hemoglobin A one see), tells the doctor what the patient’s average blood sugars have been over the past couple of months. No more can patients be very good indeed for a day or two before their blood draw because a day or two of lower sugar levels won’t change the test values. As you might imagine, the advent of the Hgb A1C test allowed doctors to manage their diabetic patients much better.
Just for grins I’ll explain how the test works. If you’re not interested, skip on down to the next paragraph. If you take a glass of water and stir some protein into it along with some sugar, over time the sugar will react with the protein in an irreversible way. The sugar will actually attach itself to some to some of the protein chains. The process takes a fair amount of time, but the longer you leave the sugar in the water with the protein, the more of it will attach. If you measure the amount of the protein-sugar complex that forms, you will have a rough measure of how long the sugar and the protein were together. If you don’t find much of the protein-sugar compound, then you can figure that the two haven’t been together for long; or if you find a lot, then you know they’ve been mixing it up for a while. The same goes for the amount. If you stir the protein into the water and add a tiny bit of sugar, after a time you will find a small amount of the protein-sugar compound; if you add a lot of sugar, then after the same amount of time you will find more of the protein-sugar compound. By taking multiple samples over time and tabulating the results you will be able to come up with a scale that says if you find this much of the protein-sugar compound after a certain number of days, then you can determine how much sugar was added. This is precisely the way that the Hgb A1C measurement works. Sugar in the blood attaches to protein in the blood. By measuring one of those proteins-sugar compounds (Hgb A1C) you can tell what the average amount of sugar in the blood has been over the past couple of months.
The second test I want to discuss that is critical to an understanding of this paper is the urinary measurement of 8-iso prostaglandin F2 alpha (8-iso PGF). When free radicals attack fatty acids that are a part of the cell membrane the damaged products produced are called isoprostanes. Isoprostanes in general are indicators of free radical damage and oxidative stress. Researchers can perform various experimental procedures and measure the release of isoprostanes in the urine and determine how much oxidative stress and free radical damage their procedure caused. 8-iso PGF is simply a specific isoprostane commonly used experimentally to determine the degree of oxidative stress/free radical damage.
Now that you know what all the tests are and what they mean, the paper is a pretty straightforward affair. It is well known that elevated blood sugar levels cause oxidative stress and free radical damage, which is probably the primary reason diabetic patients have increased risk for atherosclerosis and accelerated aging. The authors of this paper wanted to see if a steadily elevated blood sugar caused the production of more 8-iso PGF, i.e., more free radical damage, than wildly erratic blood sugars that averaged out to about the same as the steady state ones. In other words would a blood sugar level that stayed at around 180 (definitely diabetic) most of the time cause the same, more, or less free radical damage than a blood sugar that fluctuated between 100 (upper end of normal) and 250 (pretty high), but averaged by Hgb AiC measurement at about 180?
The researchers recruited 21 subjects with diabetes and 21 age and sex matched non-diabetic controls. They monitored the diabetic patients on a round the clock basis to determine not just their average sugar levels but their fluctuations as well.
The data showed that the diabetic subjects had much higher urinary levels of 8-iso PGF than did the non-diabetic patients, which would be expected. But the data also showed that diabetic subjects who had large fluctuations in blood sugar levels had even greater 8-iso PGF formation than diabetic subjects with more stable, yet still high, blood sugar levels.
These data were in diabetic patients but I suspect the situation holds true in non-diabetic patients as well. And, unfortunately, especially in followers of the low-carb diet. Why so?
Following a low-carb diet makes one a little glucose intolerant, which is the reason that the instructions for a glucose tolerance test always include the admonition to eat plenty of carbs in the week before the test. Why? Because all the macronutrients–glucose, fat and protein–are broken down by enzymes during the metabolic process. And all the enzymes necessary for the metabolism of the various macronutrients are made on demand but not immediately. If you are on a high carbohydrate diet, then you will have plenty of enzymes on hand to deal with the carbohydrates you consume. If you switch to a low-carbohydrate diet, it takes a while to manufacture the enzymes in the quantities needed to deal with the extra fat and protein that your metabolic system hadn’t been exposed to. This deficiency of protein/fat metabolizing enzymes is the reason people starting a low-carb diet become so easily fatigued–they’ve got plenty of enzymes on hand to break down carbs, they just don’t have the carbs to metabolize. Once they produce the enzymes necessary to deal with the load of protein and fat, which takes a few days, they become low-carb adapted and no longer feel fatigued.
Once people become low-carb adapted–as I hope we all are–then the same thing happens if they go face down in the donuts. They don’t have the enzymes on board to deal with the sudden influx of glucose, and, as a consequence, their blood sugar spikes higher than it would on a person eating the same amount of carbohydrate who is already carb adapted.
This paper shows that these carb spikes are not benign. As the paper points out

Risk factors of atherosclerosis such as hypertension, regular smoking, hyperlipidemia, and obesity have been described as being associated with elevated urinary excretion rates of isoprostanes.

Since the best thing we can do for ourselves is limit free radical damage as much as possible, the obvious way to do so is to maintain a constant low level of blood sugar, for which the low-card diet is just the ticket. In view of these recent findings when we’re good, we should be very good indeed, but when we’re bad maybe we shouldn’t be quite so horrid.


  1. Dr. Eades, I want to thank you very much for this post, for today I was considering binging on a stack of pancakes from Ihop. (Oh god, I love pancakes.) Earlier today I had rationalized in my head: “Hey! I’ve been pretty good this month! Why not treat yourself?”

    Then I read this. Stay the course, shall I.

  2. Interesting article, Dr. Eades, thank you. I just wanted to mention that an old doctor of mine in the States (I now live in Europe) told me years ago (eight or ten) that dramatic BG fluctuations were worse than high but stable BG levels in terms of risk for and development of diabetic complications.

  3. I’ve been guilty of going face first into the donuts. For one thing, I think the medication I take (an SSRI for depression) makes me carb intolerant. It only takes about 90g net carbs one day to put my face first in donuts the next. I just have no willpower at that level. When I keep the carbs under that I’m okay, and I seem to do best at about 60g as far as maintaining my weight, etc.
    This is certainly an article to remember since I seem to fall off the wagon about once a month.
    This probably also explains why my diabetic father seems to be okay even though his fasting glucose runs about 150+ (this is *with* insulin injections).

  4. Thank you so much for this article.
    I now understand why a diabetic friend died of a heart attack. He would go off on heavy carb binges and not pay attention to his blood sugar and then resume his low carb diet to lose weight.

  5. but if I dont have diabetes, and if I exercise a lot every day, and if I only carb-binge when I go on vacation or on major holidays, I’ll be ok, right? right? puhleeez???

  6. I usually binge on carbs once a week (saturday night). Do you think that this is regular enough to keep the carb eating enzymes active without damaging the low carb diet too much?

  7. In your book the Protein Power Life Plan, you advise against eating grains, especially wheat and corn, because of the Leaky Gut Syndrome which can lead to a variety of autoimmune diseases. However, I have reading in various articles and magazines like AARP and the Harvard Health Newsletters touting the benefits of whole grain bread and cereals. Can you clarify this situation for me?

  8. In your book The Protein Power Life Plan, on page 409, you recommend a visit to the “Honey tree”. How do you reconcile this article (the damage done to our bodies) with the statement in your book?
    Was 320lb, currently 240lb, target 190lb

  9. I don’t have the book in front of me as I write this, but I recall that the recommendation was to eat from the Honey Tree once in a while, not every other day or even every other week.
    I don’t know anyone–and I know a bunch of extremely diligent low-carbers, low-carbers who put my diligence to shame–who doesn’t succumb from time to time. I think the point is to control when you succumb and make amends for it ASAP by hopping back on the low-carb wagon.
    As a friend of mine is fond of saying, “Pleasure is a nutrient.” I agree with him entirely. A little ‘forbidden’ pleasure once in a while is probably good for us despite the temporary spike in free radicals.

  10. –Following a low-carb diet makes one a little glucose intolerant, which is the reason that the instructions for a glucose tolerance test always include the admonition to eat plenty of carbs in the week before the test.–
    I’m about to have a glucose tolerance test (2 hr.). How many carbs should I be eating leading up to the test and for how many days before?
    You need to consume 150-200 grams of carbs for at least 4 to 5 days prior to the test. That’s why MD and I almost never did glucose tolerance tests on our patients.

  11. Dr. Eades, In preparation for a 3 hr glucose tolerance test–150-200 gm of carbs- I can’t imagine that many grams! Can you please give me an example of what to eat in one day to equal that amount? Will a week of the high carb diet be adequate for the test? Thank you.
    Hi Margie–
    Follow the standard American diet and you’ll double that carb intake.
    What to eat? I would probably eat a potato and some bread and maybe a bunch of rice. You want to eat starch so that you will get glucose. Don’t eat a bunch of stuff with sugar or HFCS because you’ll get a lot of fructose.
    You need at least three days of the higher carb fare for the glucose tolerance test to be accurate. I usually have patients follow such a diet for a week just to be sure. But, I don’t usually do a glucose tolerance test, I do an insulin challenge instead, which is much more revealing.

  12. Dr Eades, After trying to up my carbs for about a week or so, I finally had my 3 hr glucose tolerance test done. Literally a pain. My results at fasting= 107, 1/2 hr= 177, 1 hr=177, 2 hr=138 and 3hr=114. It is great to be on a low carb diet again! Never realized how sluggish you feel with all of those carbs in your system. Do you think that my results are high due to my carb intolerance?– as I had been on low carb for almost a year. Thanks, Margie
    Hi Margie–
    Your results are within the normal range, at least at the lab I use. You can indeed get higher readings if you’ve been on low-carb for a long period before the test. That’s why it’s always recommended that those scheduled for a GTT up their carbs for a few days before.
    Carbs do indeed make one sluggish after a long period on low-carb. Makes you want to get back to the proper diet in hurry.

  13. Currently with low carb and exercise alone my glucose levels range from 115 to 155. I have a morning spike pretty much every day as high as the 155. I am VERY consistent in my low carb lifestyle always remaining in ketosis. I do 30-60 minutes exercise a day focusing on weight training with walking to fill out the time. I feel i should have lower BG numbers being so consistant in the last 3 months. I am gaining strength but not losing much weight (nor inches.. conundrum).
    Is there some factor I am missing? help! anyone!
    actorkent at
    Hi Kent–
    I can’t really make a call without a whole lot more information.  And even if I had the information, I couldn’t diagnose and/or treat over the internet.
    If your sugar levels are consistently as high as you report and haven’t responded to a low-carb diet and exercise, I would recommend that you find a good physician experienced in the use of low-carb diets and work with him or her to get things under control.
    Good luck–

  14. I am a 62 year old former hobbyist pastry chief now diabetic who has his sugar well controlled by a low carb diet. I want to know if I should stay away from a product called Diabeti Sweet, a sugar substitute that contains isomalt and acesulfame-K? The reason for its use is that it comes in the form of a brown sugar substitute. Thanks for your help.
    Hi Jay–
    DiabetiSweet contains a sugar substitute called isomalt that is really a sugar, but one that has a pretty low glycemic index.  I’ve not had any direct experience with isomalt, but it supposedly doesn’t break down and absorb as well as does sugar, leading to much lower blood sugars after its consumption.
    I would say that your best bet would be to try a couple of teaspoons of it, then check your blood sugar with a glucometer in 20 minutes or so to see what happens.  If your blood sugar goes up significantly, then I would avoid it.
    If you do the test, let me know the outcome.

  15. Hi Dr Mike,
    I find this really interesting.
    I low-carbed through all 3 of my pregnancies. (Advanced maternal age all 3, at 35, 37, 39 y.o., no complications, normal unmedicated vaginal deliveries without complications)
    The first PG, the GD screening was required to be fasting, and I got very sick and even nearly fainted when I drank the liquid and my sugar was 114 at one hour.
    The second PG, I could eat before (different OB’s) and remembering how sick I had gotten, I had a big protein breakfast of bacon and eggs. I felt a little yucky when I drank the liquid, but nothing terrible. My result at one hour was in the 90’s, I forget the actual number.
    The 3rd PG, same as the 2nd, and my result after one hour was 89.
    I tell you all this to say that I thought my low-carbing before the test (I refused to eat any differently the week before the test b/c I wanted to the test to reflect results based on the way I usually eat) would somehow show that eating a LC diet would protect one from high blood sugar even after a load of glucose? My hypothesis (eat low carb, have normal screening) held true 3 times, or so I thought.
    Now I read that the result was likely HIGHER that it would have been had I just been eating a high-carb diet??? Good Lord, if a HIGHER result was 89, then what would it have been??? 60?? 40??
    I’m very confused now.
    What, if anything do you make of the fact that my nunmber were great, even though I was low-carbing for months (years even) before the tests? I’m very curious. I thought I was doing the right thing at the time…now, not so sure?
    Kids are just fine, BTW, at 6, 4, and 2 years old!
    Hi Elle–
    I think you are fine.  You obviously did not develop insulin resistance during your pregnancy, which is a good thing.  Had you done so, your blood sugar levels would have been much higher.  I think some people adapt nicely to the low-carb diet (especially those who have followed it for a long time), increase their insulin sensitivity, and have no problem with a glucose tolerance test.  You are one of those people.

  16. I wish I had read this a year ago. I had a borderline diabetic result on a 2 hour GTT last year, but I’d been eating a low carb diet (15 or less grams per day) for about 3 months prior to the test. I asked my doctor at the time if I should eat more carbs the week before the test, and he told me no, to eat as I had been. Because of the results of this test, I’ve been labeled a diabetic, even though my A1C has never been higher than 5.3, and my fasting glucose level is 85-93 daily.
    I honestly think I’ve been misdiagnosed, and should be retested, but my doctor disagreed and put me on several meds that caused my glucose levels to crash several times per day. I have moved on to a different doctor (and gotten off all meds), and plan to revisit this again. I’ll bring a copy of this article with me when I do. Thank you.
    Hi Robin–
    Sounds like you’ve made the right call.  Let me know the outcome.

  17. Hi Dr Mike,
    I posted above in #15 and I wanted to give you an update.
    This month I was off-plan for a week b/c I had family visiting. I decided to test my fasting blood sugar (diabetic Daddy was visiting) and I was astounded to see that the result was 135!!
    It has always, always been 89-90ish when I’m adapted to whatever style I’m eating, whether it be LC or see-food diet (I see food, I eat it!).
    You could have knocked me over with a feather. I was incredulous of the increased sugar response on LC b/c of my experience above, but I’m a believer now! I’ve seen it with my own eyes (Thomas that I am)!
    Hope that it doesn’t mean I’m IR now? Oh well. I go for another physical in the fall. My weight is good now, so I figure I’m at holding it at bay with being normally weighted.
    Thanks for all your great info. I have just recently discovered your blog…don’t think I can ever get caught up…
    Hi Elle–
    I doubt that you’re IR now.  Just check again after you’ve been back on the plan for a while.  I’m pretty sure it will be okay.  Keep me posted.

  18. Hey, Dr Mike,
    Love your blog! Thanks so much for having it on the net.
    I was wondering what you thought of CKD. It seems that most athletes who practice this are consuming VLC for 6 days, then for one day they consume TONS of carbs. Many of them have PWO carbs, but usually restrict that to about 20g or so. In relation to this article on carb binging, I was wondering what your take on this regular carb binging was? How much harm are these athletes doing? Or is there some trick to CKD that makes it safe?
    Also, if one is on a near zero carb diet, can they still build muscle effectively? And, if so, what is the best way? More protein or what? What would be a good PWO meal, and why?
    BTW… I’ve always been curious as to why glycogen stores go up and down depending on carb consumption. If some glycogen is needed, and will be produced via gluconeogenesis, why is excess glycogen shoved into liver and muscle when carb consumption is high? How did this come to be during our evolution, where carb consumption must have been extremely miniscule if not totally absent? What is the physiological advantage (or disadvantage) to putting so much sugar into this type of storage?
    Thanks again for such a great blog!
    Hi Dean–
    The fact that these folks eat a ton of carb once a week is mitigated by the fact that they are at zero carbs for the rest of the week.  It’s kind of like intermittent fasting.
    People can build muscle effectively if they eat plenty of leucine, one of the branched chain amino acids.  It requires about 8-12 gram per day to drive muscle synthesis.
    Glycogen is a storage reservoir for carb to basically get it (carb) out of the blood quickly where it can cause real damage and into the storage space in the cells.  Although primitive man didn’t eat a lot of carbs, I’m sure when he figured out about honey be binged from time to time, and ate a fair amount of sweet fruit when he could get it during the short period that it was ripe.  The glycogen storage system is designed for just such occasions.

  19. I have so much more energy when I eat low carb. I’m bloated and tired when I eat carbohydrates, short term or over a period of months. Eating low carb, I can bike 25+ miles. Please don’t make blanket statements.

Leave a Reply

Your email address will not be published. Required fields are marked *