Thursday, September 18, 2008

A New Toy

I bought a new toy the other day: a blood glucose meter. I was curious about my post-meal blood glucose after my HbA1c reading came back higher than I was expecting. A blood glucose meter is the only way to know what your blood sugar is doing in your normal setting.

"Glucose intolerance" is the inability to effectively control blood glucose as it enters the bloodstream from the digestive system. It results in elevated blood sugar after eating carbohydrate, which is not a good thing. In someone with normal glucose tolerance, insulin is secreted in sufficient amounts, and the tissues are sufficiently sensitive to it, that blood glucose is kept within a fairly tight range of concentrations.

Glucose tolerance is typically the first thing to deteriorate in the process leading to type II diabetes. By the time fasting glucose is elevated, glucose intolerance is usually well established. Jenny Ruhl talks about this in her wonderful book Blood Sugar 101. Unfortunately, fasting glucose is the most commonly administered glucose test. That's because the more telling one, the oral glucose tolerance test (OGTT), is more involved and more expensive.

An OGTT involves drinking a concentrated solution of glucose and monitoring blood glucose at one and two hours. Values of >140 mg/dL at one hour and >120 mg/dL at two hours are considered "normal". If you have access to a blood glucose meter, you can give yourself a makeshift OGTT. You eat 60-70 grams of quickly-digesting carbohydrate with no fat to slow down absorption and monitor your glucose.

I gave myself an OGTT tonight. I ate a medium-sized boiled potato and a large slice of white bread, totaling about 60g of carbohydrate. Potatoes and bread digest very quickly, resulting in a blood glucose spike similar to drinking concentrated glucose! You can see that in the graph below. I ate at time zero. By 15 minutes, my blood glucose had reached its peak at 106 mg/dL.

My numbers were 97 mg/dL at one hour, and 80 mg/dL at two hours; far below the cutoff for impaired glucose tolerance. I completely cleared the glucose by an hour and 45 minutes. My maximum value was 106 mg/dL, also quite good. That's despite the fact that I used more carbohydrate for the OGTT than I would typically eat in a sitting. I hope you like the graph; I had to prick my fingers 10 times to make it! I thought it would look good with a lot of data points.

I'm going to have fun with this glucose meter. I've already gotten some valuable information. For example, just as I suspected, fast-digesting carbohydrate is not a problem for someone with a well-functioning pancreas and insulin-sensitive tissues. This is consistent with what we see in the Kitavans, who eat a high-carbohydrate, high glycemic load diet, yet are extremely healthy. Of course, for someone with impaired glucose tolerance (very common in industrial societies), fast-digesting carbohydrates could be the kiss of death. The big question is, what causes the pancreas to deteriorate and the tissues to become insulin resistant? Considering certain non-industrial societies were eating plenty of carbohydrate with no problems, it must be something about the modern lifestyle: industrially processed grains (particularly wheat), industrial vegetable oils, refined sugar, lack of fat-soluble vitamins, toxic pollutants and inactivity come to mind. One could make a case for any of those factors contributing to the problem.


.^ said...

Well at least you won't die from diabetes. Maybe from blood loss ...

Anonymous said...

At least those cultures in the tropics were consuming a lot of coconut. As you make reference in your post, would this not have slowed the absorption rate and therefore isn't really comparable to the carbohydrate laden, Western diet?

Thackray said...


Steven said: “The big question is, what causes the pancreas to deteriorate and the tissues to become insulin resistant?”

The dangerous combination seems to be saturated fats that induce insulin resistance and simultaneous high glycemic load carbohydrates. See Peter’s blog here:

Here is a test tube study that shows the same thing.

This researchers note the change in terminology from “lipid” hypothesis to “glucolipotoxicity” hypothesis to describe dietary influences on the development of diabetes (and all the down stream problems associated with diabetes).

Philip Thackray

Stan Bleszynski said...

You may find this graph useful:

Stan (Heretic)

Stephan Guyenet said...


Yes, I'm sure it would have. But keep in mind they weren't eating very much fat- 21% of calories.

Anna said...


I'm so jealous of your BG numbers :-). I can get similar or only slightly higher numbers, but with a strict low sugar/low starch diet only. And my A1c has remained in the 5.5-5.9% range.

I did my DIY test with Cheerios. It's like a bowl of predigested sugar, IMO. But I measured for 75 gms of CHO, slightly higher than your amount. Usually, for GTTs during pregnancy, the lab administers 100 gms of glucose syrup. Not sure why they challenge with such a high amount, but it feels just awful to put that much glucose syrup in one's body that fast. Burning stomach, sickening dizziness, "wired" feeling, trembling, ugh, I hate to event think about it.

BTW, your A1c could be a bit higher due to fructose consumption from produce (lots of nice vine-ripened heirloom tomatoes this summer?), or as you mention, your own rate of RBC turnover, etc. If you read Jenny's blog, A1c is a rather crude measurement in some ways, and is often hard to make an exact correlation between BG levels and A1c.

Fats and proteins do slow down the rate of glucose absorption, dampening BG spikes, but the body still has to crank out the insulin for every gram of glucose consumed (that's what many high-carbing normoglycemic people don't realize).

A glucose intolerant person eating a high carb food combined with protein and fat might have only moderately high BG (still damaging) for a longer time or later after eating. I've experienced that with my meter and various food tests and still don't like the numbers I get. Though, if I do indulge in something usually off-limits on my diet, like commercial ice cream, I do make sure it has ample fat or protein. MY BG numbers are not as high with ice cream as with things like non-fat frozen yogurt (sky high!). But of course, except when I was on vacation in Italy this summer and indulged in a small gelato fairly often, most of the time I make my own ice cream and it is very low in sugar (& high in butterfat/coconut fat and eggs) instead of commercial ice cream.

So just slowing a lot of CHO with food combinations isn't necessarily a practical option, at least not without monitoring BG over time. For those of us with IGT, keeping CHO consumption to a low enough level to maintain tight, normal BG control is a better option.

You said, "The big question is, what causes the pancreas to deteriorate and the tissues to become insulin resistant?"

That's my question, too, and I can't get my endocrinologist to help me investigate it, either. It probably isn't entirely possible to determine the cause in each person (there are so many possible causes or combinations of causes), but still, it frustrates me to not have answers. "Inquiring minds want to know", right?

And I want to know for my son, too, because between whatever genetics we have passed to him, and epigenetic influences from high BG during my pregnancy (the first two trimesters, before I knew about my high BG and normalized it with diet) might be expressed or not expressed by factors well within our control (if we knew what they were).

Also, as you point out, using FBG as a screening tool has disastrous results on health, because it catches BG problems too late. Most people have had years or decades of damaging high post meal BG before their doctor ever even mentions. By that time, so much damage is already done. And worse, the treatment is usually drugs instead of good advice to limit the foods that raise BG - concentrated sugars and starches.

lizzicat said...

I wonder what your reticulocyte count is. Probably unbelievably low. Perhaps you turn over your red blood cells every 4 months.

Stephan Guyenet said...

Hi Thackray,

I can't say I'm convinced by the theory that sat fat causes insulin resistance. From reading the paper Peter linked to in his post, I can see that what they're referring to is long-term insulin resistance, rather than meal-by-meal. That clearly didn't happen to me, despite my high intake of saturated fat plus carbs, nor did it happen to the Kitavans at 17% energy from sat fat plus a high GL high-carb diet.

Stephan Guyenet said...


Another point. In the US, saturated fat consumption has decreased in the past 30 years, n-6 consumption has increased, and diabetes has increased dramatically.

Stephan Guyenet said...


Thanks for the link.


I'll be monitoring my BG during my normal meals, so we'll see how that looks. I've noticed I get a lower reading after I wash my hands-- is that due to something on my hands increasing the reading?


That would be nice to know. Apparently erythrocytes get phagocytosed in response to oxidative changes in their membranes. Maybe my low-PUFA diet plus high HDL makes my erythrocytes resistant to oxidation, so they turn over more slowly. But I'll continue monitoring my BG after meals just to be sure it isn't elevated.

Unknown said...

As Stephan mentioned, toxic pollutants are one of the factors involved in the increase of insulin resistance. Recently there's been more articles in the popular press referencing studies such as this:

Anna said...


Testing tips:

You definitely should wash your hands before testing if you can. There can be traces of glucose on your skin from that salt water taffy someone left at the coffee station after their vacation :-). I'm not aware of glucose in skin secretions, but it's good practice to at least rinse to avoid potential glucose contamination. The skin doesn't have to be sterile, though, just generally clean. Alcohol pads and those anti-bacterial gels/lotions are really drying and can make a testing area too dry and painful over time.

Dr Bernstein says that if you can't wash your hands, at least suck your finger clean a bit. That sounds gross, perhaps, but it is better than getting a false reading from "donut" hands (not you, of course). And it isn't good to have hands so dirty all the time anyway, right?

BTW, some tips if you test frequently: it also isn't necessary to change the lancet every time you use the device. If you are the only one using it, you can't contaminate yourself. The lancets actually get less painful after some use. When they start to hurt again (takes about a month's use for me), then I change the lancet tip.

Of course, the above doesn't apply if you are using the lancet device to test someone else - then a new lancet is necessary for each person. I'm sure you know that, but one never knows who reads this and doesn't think of that.

Used lancets and test strips are considered biohazards and are not supposed to be disposed of in common garbage disposal receptacles, though I'm sure most people don't follow that (though increasingly there are laws and regulations about this). An old film canister or similar sealed container works fine, but of course, drug stores sell special "sharps" containers for those who use needles, lancets, test strips, etc. I think it makes sense to take care to dispose carefully to eliminate others risk of coming into contact with the lancets or the test strips, but a biohazard box? Common sense will do.

Also, many people find it more comfortable to test on the pinky or ring fingers, as one uses those digits less often. Also, testing on the side of the finger tip generally is more comfortable than on the pad of the fingertip. Some prefer the top of the fingertip, though I don't. Forearm testing can be done, but the lag time is greater than for fingertips, according to Jenny/Diabetes 101.

Additionally, adjust the lancet device for a good blood droplet, but at the lest discomfort level. It can take some practice to find the right setting. Hand temp makes a difference, too. 10 years ago I didn't know I was hypothyroid, and my hands were always so cold and numb, I had trouble getting a good droplet without really sticking my finger hard. Now that I take thyroid hormone, my temp is nearly normal and I don't have cold hands so often. I used to swing my arm in a circle first, to get the blood down into my hand. I was told a small amount of "milking" the finger from the base to the tip is ok, but not to overdo it.


Stephan Guyenet said...


That study is really disturbing. I'm going to post about BPA soon.


Thanks for the tips.

Robert Andrew Brown said...

Thanks Stephan and everybody for more thought provoking and informative material.

I go with Stephans simplistic questions as to what has changed to drive insulin resistance. Clearly it is multi factoral.

I believe that Omega 6 has a big part to play in insulin resistance by a number of mechanisms.

It all comes back to Omega 6 linoleic LA acid being a true external agent of enormous influence in body function that is scarce in the environment and I suggest provides a functional link to the fecundity of the environment.

LA is key to breeding at many levels. LA arguably aggressively drives the body to store LA and other fats. This drive overrides other insulin instructions. At the same time excess LA in the form of free fatty acids drives oxidative stress. Add to this increased sugar availability. High level of Omega 6 in seeds pips nuts etc came with carbohydrates and sugars, all conveniently prepacked for storage as fat. Operating parameters arguably tell us to store and not burn these excess carbohydrates, and Omega 6 is the conductor.

Excess LA is outside operating parameters and leads to malfunction through a number of mechanisms - over activation of the eicosaniod pathway - oxidative stress - high storage of Omega 6 providing permanent excess of raw materials etc.

I cannot see why saturate or monosaturates would be causal. The body makes these fats. They may be predictive.

It all interlinks and flows from the way the plant 18 carbon EFAs are utilised in the body.

Insulin increase desaturase function. Archidonic acid and or its derivatives increase insulin production. This creates potential for a self blocking loop.

Free fatty acids drive peroxisomal oxidation. LA in particular will drive downstream oxidative stress.

A significant portion of excess LA is made back into saturates and monosaturates. LA has mechanisms to drive these fats into storage...

I am suggesting that no matter what other signals are being produced in the insulin glucose fructose energy pathways, those produced directly and indirectly by EXCESS Omega 6 have the whip hand.

Robert Brown


Omega Six The Devils Fat


Anna said...

Stephan said, "I've noticed I get a lower reading after I wash my hands-- is that due to something on my hands increasing the reading? "

I can see it now, some internet search will find your comment, leading to headlines reading, "Recent Research Study Shows Washing Hands Lowers Blood Glucose Readings" or "Dirty Hands Increases Risk of Higher Blood Glucose Readings". Sales of antibacterial soap skyrocket.


Drs. Cynthia and David said...

Hi Stephan,

Cool study! My sympathies for your fingers. I really hate sticking myself, and I think you're very brave (as are all those diabetics out there who stoically go through these motions regularly!)

Here is a recent paper that redefines the equation to calculate estamated average glucose from HbA1c values. Seems to indicate you may have had lower ave BG than using the older equation.

The authors said:
“Of note, the regression equation in this study provides lower eAG values, compared with the widely used equation derived from the DCCT, and the scatter around the regressionline is less wide (18). The most obvious explanation for the difference between AG calculated from the DCCT and that calculated in the current study is the difference in the frequency of glucose measurements used to calculate AG (a single seven-point profile with no overnight measurements during 3 months in the DCCT compared with numerous CGM [continuous glucose monitoring] and seven-point profile measurements that captured a median of 52 days in ADAG), providing a more complete and representative measure of average glucose in ADAG.”

“Linear regression of A1C at the end of month 3 and calculated AG during the preceding 3 months. Calculated AGmg/dl = 28.7 x A1C -46.7 (AGmmol = 1.59 x A1C - 2.59) (R2 = 0.84, P < 0.0001).”

There's a nice graph too showing their linear regression.

Stephan Guyenet said...


Why do you think n-6 pathways are dominant over other pathways in health?


Haha, here's one- "Eating donuts and not washing hands causes a 537% increase in blood sugar"

Cynthia and David,

Thanks for the link. I used the formula and it says my average BG is 118 mg/dL. Considering I haven't gotten a reading over 116 mg/dL yet (and my fasting is in the low 80s), I have a hard time believing my average is that high. I'm still testing just to be sure, but I think there's something else going on.

Robert Andrew Brown said...


"Why do you think n-6 pathways are dominant over other pathways in health?"

Three years reading and being a total Omega obsessive nerd.

1. Because it has more influence in more functions in the body than any other factor I have seen. Because it is a key part of structure.

2. Because it is a true external agent.

2. Because it reflects the fertility of the world in the most accurate and direct way. There is no other agent I can think of that that as accurately reflects the fecundity of the land environment AND is capable of the necessary control of body function.

3.Because it regulates body function behaviour and breeding in response to the fecundity of the environment.

And generally just because it makes a very great deal of sense.

Robert Brown

Omega Six The Devils Fat

Stuart Buck said...

Have you seen this article? Title: "Cancer patterns in Inuit populations."
It looks like it would be interesting.

Stephan Guyenet said...

Hi Stuart,

Thanks, I did see that article. I may post on it at some point.

Drs. Cynthia and David said...

Hi Stephan,

Interesting that it still doesn't calculate out to a reasonable approximation of your average BG. The formula was developed using types I and II patients as well as normal controls (presumably eating a standard higher carb diet (SAD as Peter calls it!)). But maybe this formula doesn't fit with low carbers? Or is it as Lizzicat says that your RBC turnover is slower making your av BG anomalously high? Maybe there is even causation there- keeping carb intake and BG low leads to slower turnover, less wear and tear, oxidative damage etc.? This could be fodder for a whole other study and maybe a new discovery: low carbing leads to a slowed rate of aging. It's not so far-fetched really. Barzilai found that centenarians tend to have mutations in their IGF-1 receptors. I wonder if RBC turnover could be used as a measure of aging rate...

Keep up the good work!


Stephan Guyenet said...

Hi Cynthia,

That thought has crossed my mind. I know dietary fat affects erythrocyte membrane composition, and apparently oxidative damage to the membrane is a signal for erythrocyte elimination. Perhaps a diet high in SFA and low in PUFA slows erythrocyte turnover.

I still haven't fully ruled out that it's my blood sugar though. I'm going to keep testing it. If it stays normal after several more meals, I'll feel comfortable saying the HbA1c was an anomaly.

Scott W said...

Thanks for posting on this. I am working through a personal paradigm shift. I had been a very dedicated low carber for years, even down to zero carbs. But I never did obtain the loss of body fat that I wanted and at the same time I eventually had to give up working out at almost any would just destroy me. I couldn't figure it out.

My problem was that the low-carb books are really written for the general population, not those who actively pursue fitness. Eventually, I began to learn about muscle and liver glycogen, etc.

Adding carbs back into my diet, particularly during/post-workout, made an amazing difference. Suddenly, I could lift weights again.

I followed your articles on the Kitavans with great interest. It seems that many of the paleo gurus hold an image of a temperate forest dweller as their diet model. What of the tropical cultures, where people thrived as well? Even for Savannah dwellers, meat was a fairly rare event, shared widely and individually limited. I read a very interesting book about pre-contact bushmen, and was interested to find that they relied extensively on carb sources, including many roots and tubers.

I do agree on the problems with grains, but that is not really an issues as there are many other starch sources.

Now I am working through finding proper carb levels to support the amount of training that I want to do be be functionally fit. But I must admit that I have lingering fears, driven by years of believing that carbs are pure evil, of ruining my pancreas and metabolism in general. So, to bring this long ramble to an end, I was quite interested to hear of your experiment in measuring response to glucose, which I plan to replicate for myself. question after all of did you choose the blood glucose meter you used? There seems to be a ton of them. What brand are you using?

Thanks so much for your work in this blog. I check it every day.

Stephan Guyenet said...

Hi Brian,

I went through something similar this year. I discovered that I'm stronger, I put on muscle better, and I feel better overall if I'm eating carbohydrate.

Reading more about human archaeology and the Kitavans turned me around. We've probably been eating starchy tubers/roots for 1-2 million years.

I bought the TrueTrack meter from Bartell's (available at other common drug stores as well). It was the cheapest available. $20 for the meter, $55 for 100 test strips. It's been giving me some strange readings these past few days though, so I'm not sure I can recommend it. It may just be a bad batch of strips.

Your best resource in my opinion is Jenny Ruhl's "Blood Sugar 101" webpage. Go to the "diagnosing diabetes" link at the top of the page and click on the "Am I diabetic?" link. She gives instructions for buying a meter and doing a makeshift OGTT.

Keep in mind you have to get your body used to carb again before you do the OGTT. Good luck and keep us filled in on your progress!

Anna said...


One of the main differences in meters is the amount of data they hold and ability to analyze the data. I have two One Touch meters, complex and one small and very simple; both are made by the same manufacturer and use the same test strips.

The "big" meter that I use most often is a One Touch Ultra Smart, which will allow for a fair amount of additional data and customization to be added, such as comments for each reading: Food (I think you add gm of CHO for this one); Health 1 (stress, feel hypo,illness, menses); Health 2 (same options as Health 1); Exercise (I think this one you can put in minutes). Medication data can also be entered; I don't use any diabetes meds so I haven't explored that one. The data entry isn't perfect, and not as useful as some of the PDA programs out there, but it is in one device, so that is useful. More details would have to be hand entered into another log, either written or electronic.

The UltraSmart also allows you to customize your target BG range and will prompt you if you are lower (do you need a snack?) or higher (add a comment?) than your range.

The data analysis is the feature I like most. I know there are non-meter ways to do this, but keeping the data in one place is convenient. I like scrolling back in time and viewing in graph form to see how my BG readings fall in relation to my target BG range. You can also compare by meals or exercise, as well as time of day.

There is a way to use a PC to back up data or print reports or for data transfer at the doc's office, but I use a Mac. As far as I can tell there are few meter/Mac options available currently, or for the foreseeable future (arrrgh!).

The case for the UltraSmart is made of zipped semi-rigid neoprene and holds the meter, a vial of strips (or two), a lancet device, and a zippered mesh pocket for control solution, extra lancet tips, Quick Reference guide card, tissues, etc. It's about the size of an medium sized aim-n-shoot camera case or a woman's large wallet.

I bought this meter kit at Costco for perhaps about $65 (memory is murky), which at the time, was the best out of pocket price I could find for a meter with this data capability. I think manufacturer's mail-in rebates covering much or all of the cost of a new meter is common, so be sure to check if you buy one.

I have another compact meter, the One Touch Ultra Mini, which retails for about $20. I think I bought it on sale for about $15. The meter itself is small, the size of a very thick pen or marker, and the lancet device is half the size of the pen-sized regular version (but the mini one makes a more painful jab than the standard size - perhaps the spring load is different). There is also room for one vial of test strips, and a small zippered mesh pocket. This case is soft fabric, slightly longer, but narrower than the other one, with a belt loop and gusseted pocket on the outside. This case really is only about 25% smaller in bulk than the other one, but the smaller contents (without the case) would fit easier into a small purse or pocket than the standard sized kit. The big advertising feature seems to be the range of colors this model comes in. Big woo! But that might appeal to teenagers.

The UltraMini only has a memory for BG readings and date/time, with scroll buttons to scroll through. More data or data analysis would have to be done some other way.

I had a free Bayer Contour that I received for signing up for an online diabetes newsletter (I never used it and gave it to a friend so I can't comment on its use), but I think that offer is expired now. Many diabetes supply companies offer free meters, but they usually require a doctor's Rx. And you don't get to choose the meter features yourself. When I was pregnant, I was given a meter kit by the Diabetes Nurse Educator. As an aside, 10 years ago TODAY was one of the best days of my life - I gave birth to my son - at this point in the day I was still knitting between contractions. :-) Seems like yesterday.

Until I got a Rx for test strips, I paid out of pocket for my strips for about a year. I bought boxes of 100 (4 vials of 25) at Costco, which was the best price I could find. Now my copay is the same no matter where I buy them so I get them from a local independent pharmacy (the line is shorter, too). The strips dont' have an indefinite shelf life, so check the expiration dates. And use up an open vial within 3 months or you may not get accurate results. Keep test strips tightly capped in the vial at times. If your climate is humid, I would estimate test strip life span might be less than three months after opening.

I used my meter while I was having a lab 3 hr GTT, so I have a good sense that my meter is very accurate and closely calibrated to the lab tests (samples taken within 2-5 minutes of each other). I've also tested with both meters at the same time and with two test strips from different vials to check accuracy. So I don't really worry about meter inaccuracy, as long as the test strips are pretty fresh and stored appropriately and I don't' drop the meter.

Jonty said...
This comment has been removed by the author.
Jonty said...

According to Lutz' Life Without Bread (highly recommended), you are hypersecreting insulin like fat kids do. If you weren't eating too much sugar-equivalent food, you'd have a peak. I have a graphic from his book showing a curve just like yours which I'm just itching to post, but I can't figure out how to do it!

LOG ME IN said...

(I thought I posted this comment last week, but then I couldn't find it. I hope it didn't get deleted for any reason, as I'm nothing but a fan of this blog.)

Thanks much for this post! I've been looking for some decent recommendation of what my blood sugar ought to be after meals at various intervals. I bought a glucose meter a few weeks ago, but I had trouble digging up that information. The web site for Blood Sugar 101 seems helpful though. From what it says, I gather that I never want to go over 140, as that's the point of damage. I should stay under 120. And, by two hours after a meal, I should have returned to under 100, at least.

Does that sound right? Or should I shoot for less than 120 (or even less) as my peak?

I've been wanting to run a series of tests with my glucose meter, to find out what foods raise my blood glucose significantly. So I recently tested my standard breakfast of:

1/2 c homemade raw milk greek yogurt
1 medium peach
2 tbsp raw walnuts

Before eating (fasting): 82
30 minutes after eating: 94
60 minutes after eating: 95
90 minutes after eating: 85
120 minutes after eating: 75

That seems pretty damn good, I think. (I'm definitely going to have some fun with these tests. I love playing guinea pig with myself.)

On a more general note, thanks for all the great blogging. I'm yet another person who radically changed her diet in recent months -- in part guided by your blogging -- and I'm feeling way, way better than I ever have.

Also, I've added you to my blogroll. I've been blogging for over six years, but mostly cultural and political commentary. I've just started blogging on nutrition and exercise, albeit only on occasion; I'm going to make it a regular feature on Saturdays. My first two posts are here:

The New Diet -- on the changes that I've made to my diet over the past few months.

Experiments in Eating -- on the process of finding out what foods work well for one's body.

Based on the volume of comments on those posts -- over 150 so far -- my readers do seem interested to hear more. Of course, I'll be liberally linking to all the good stuff on your blog, but please do feel free to chime in if you like. I've been reading voraciously, but I'm still very much a novice.

-- Diana Hsieh
NoodleFood (it's a metaphor!)

Anna said...


In my not-so-humble-opinion, you've got great numbers, though the carb load in your test meal wasn't terrible high by most standards, and the fat, protein, and fiber of those foods will slow the glucose absorption quite a bit. Try testing/challenging again a few times with a more easily absorbed, high carb food to see how your system handles a big dumping of glucose into the bloodstream. Think of it like a glucose stress test.

Data from your BG during/after normal eating patterns is useful, too.

Jenny's Diabetes 101 numbers are pretty good numbers to go by for your testing. She provides a lot of references to demonstrate that 140 mg is on the upper end of the BG limit (or try to reserve foods that get you in that range for rare occasions, like monumental celebrations).

Trying to eat so that your BG remains below 120 mg at all times seems more prudent than 140 in teh long run, though, so if you can, do that. If your glucose regulation is still normal, I'll bet you can.

Some people with robust glucose metabolisms can still stay below those numbers for a long time with a pretty big amount of dose of glucose, but even then, there's likely a long-term price to pay for the constant insulin production generated by lots of glucose.

LOG ME IN said...


Thanks for the info and advice! I've continued to test various meals since running my initial test. It's slow going, however, as the five measurements interrupt my work, so I can only manage to do one or two tests per week. (I'm writing my dissertation for my philosophy Ph.D, so it's just too easy to slip into procrastination mode when interrupted!) I'm sure that I will do the kind of improved oral glucose tolerance test that Stephan did; that would be interesting.

Most of my results have been very good, but I did get a pretty serious spike after a lunch of a small chicken breast, a cup of raw milk, and a large ear of corn:

Just before eating: 72
30 minutes after eating: 120
1 hour after eating: 133
1 hour 30 minutes after eating: 101
2 hours after eating: 85

The combination of the corn and milk was clearly more than I want to subject my body to on a regular basis. So that's really good to know. Notably, with a glass of raw milk alone, my peak was 106.

In contrast to that lunch, I had bacon, eggs, and some leftover sauteed vegetables (zucchini, onions, tomatoes) for lunch today, and my blood sugar never got above 90. Yeah!

Stephan Guyenet said...

Hi Jonty,

Why do you think I'm hypersecreting insulin?

Hi Diana,

Sorry your comment didn't show up. I'm not sure why that would have happened. I have only deleted comments that contain spam or are inappropriate so far. What you said sounds right. You definitely want to stay below 140, and below 120 is better.

Judging by your numbers, you are clearly not diabetic, so that's good. 133 is on the high side though, so it might be good to keep an eye on that. Make sure to wash your hands before testing. I often get high readings if my hands aren't perfectly clean.

I read your posts; it sounds like you are making some great progress! It's amazing what real food will do for a person. Keep us up to date.

Jonty said...

I think you're hypersecreting insulin because Wolfgzng Lutz found the same truncated pyramid response peaking under 100mg% in overweight teenagers. After some time on less carbs, their responses changed to a pyramid peaking at about 130mg%, same as his normal-weight pts - they secreted less insulin when challenged.

He says you hypersecrete in response to excess carbs, his book Life Without Bread is about his 40 years of positive experiences prescribing <72g carbs per day for CHD, IBS, diabetes, metabolic syndrome, cancer, hypertension, hypoglycemia, overweight etc.

BTW I'm an insulin-dependent diabetic on a very low-carb diet for 10 years whose HbA1c is 5.5% - normal. My story's at

Tell me how to post a jpg file and I'll put Lutz's graph up along with his commentary. Best, J

Stephan Guyenet said...

Hi Jonty,

I'm not sure I understand the argument. People who low-carb undersecrete insulin in response to a carb challenge because the body adapts to having to deal with less glucose. That doesn't mean peaking at 130 is normal. If those same people ate more carbohydrate, their pancreas would adjust over several days and the peak would be lower after a challenge.

The other thing to keep in mind is my meter is not super accurate. It's within about 10% of my actual glucose level, so any single point could be off by that much. I'm not sure you can read much into the shape of the curve. If I did it again and got the same result, then I might believe it.

I don't understand the argument of "hypersecretion in response to excess carbs". You secrete more insulin because you have more glucose to deal with, how is that a problem? I suspect there is more to the argument that I haven't captured.

I don't know how to post jpg files here, sorry. I'm a bit computer-challenged.

Jonty said...

The kids who had curves like yours had been overindulging refined carbs and had gained weight: see graph from Life without Bread at , the curve is strikingly similar to yours (you're right, could be an artifact).
His remedy for “carbohydrate disease” was to restrict carbs to 72g per day which changed the kids curve to the one labeled Normal in the figure and they lost weight. Of course you’re right, restricting carbs elevates insulin resistance, but who cares, my HbA1c stays below 5.5% so long as carb intake stay below 15% of calories. My doc's Rx is go for the insulin sensitivity and die of complications, he's a regular Albert Schweitzter. For all I know you eat nothing but lard but the similarity of the curves is striking. I think it likely that insulin hypersecretion is widespread, the precursor of insulin resistance and metabolic syndrome - the problem is one eats refined carbs, then one secretes more insulin than is necessary (reason unknown, it's an observation). There can be overshoot and hypoglycemia, which is why the mainstream Rx for hypoglycemia is to cut out sugar (refs at

Stephan Guyenet said...


I see the resemblance between my glucose curve and the overweight adolescents' curve. What I'd like to see is the curve of a non-overweight, carbohydrate-adapted person. The person whose glucose shot up to 130 could be considered glucose intolerant. Of course I wouldn't call them that if they were eating low carb prior.

I agree with you that insulin oversecretion can be an early sign of insulin dysfunction, but I don't see how it counts as oversecretion if there's no hypoglycemia going on.

On a related note, I've been sitting on a study that showed wheat bran to induce insulin resistance and insulin oversecretion in a controlled trial. I'd be inclined to think it's an effect of specific foods rather than carbohydrate in general.

Jonty said...

A "normal" curve in a carbohydrate-adapted person is like the peaky curve - see - search for "uncommon" - for normal curves from Hawkins and Pauling's Orthomolecular Psychiatry. The blacked-out portion is the normal area.

But let's say for the purposes of argument that the fat kids are insulin-resistant; this causes the hypersecretion of insulin and overshoot of bg correction. Let's call this subclinical hypoglycemia, since hypoglycemia is so narrowly defined as simultaneous low bg & symptoms. Let's further say that fructose feeding causes insulin resistance. Kids eat 25%+ of cals as sugar or HFCS; and one in every 3 of these kids is overweight or obese, destined to become an obese adult - an epidemic according to the WHO. If these things are causally connected, it's grounds to think as I do that there's oversecretion of insulin going on even tho there's no overt hypoglycemia, and that this is the precursor of metabolic syndrome, the greatest man-made nutritional catastrophe in the history of the galaxy.

I think you're spot on with the food sensitivity thing, but I think it's additive and not the whole story. Oatmeal is the most inflammatory food for my bg I've ever tested in spite of its mild GI, and it must work thru provoking insulin resistance since I don't make insulin.

Anonymous said...

Stephan, omega-3 supplements raise hba1c, according to many studies. I read about this from Brian Peskin's EFA Analysis and Dr. Guy Schenker's Nutri-Spec Newsletter. Ray Peat has also blamed diabetes on eating high PUFA oils of nuts, seeds, and fish.

Here's an excerpt of Brian Peskin's article. He discourages fish oil in even small amounts (1-2 g/d) and he says the problem is that people are eating too much damaged omega-6 fat and not that they don't have enough omega-3 in their diets. He believes that a high omega-6 ratio is better for health and has many testimonies from athletes and other people.

"The glycemic [blood sugar] control of the four insulin dependent diabetic patients worsened during the fish oil administration.
"[T]he insulin dose of the subjects had to be increased throughout the six-month period of fish oil administration to maintain constant
blood glucose and glycosylated hemoglobin concentrations (HbA1c ~ average blood sugar level).
"Despite the stable body weight by patients on the basal diet, glycosylated hemoglobin levels after six months of fish oil administration increased 16% from 4.9% to 5.7%. Note: This is an awful effect for a diabetic.
"Another important finding of our investigation was that consumption of a fish oil-enriched diet worsens glycemic tolerance."

Here's the Nutri-Spec newsletter on the "Omega-3 Propaganda Machine." I also recommend the six newsletters prior to this (2005-11 to 2006-04). They form a series that completely demolishes the idea of eating high PUFA oils and foods for health.

Of course, Ray Peat has written a lot about these topics. Here's his article on diabetes and the truth about what causes it.

Stephan Guyenet said...

Hi Bruce,

Thanks for the links. I am highly skeptical of the Brian Peskin article because it sounds like he's trying to sell something, and his advice does not seem to square with the diets of healthy non-industrial people.

The Nutri-Spec newsletters make some good points. The biggest challenge to the idea that PUFAs need to be kept low is the traditional Inuit culture. I'm glad to see he addressed that.

He says that the average Inuit diet was about 10% n-3, which I can believe. Seal oil, one of their big staples, is between 15 and 30% n-3. Fish fat has a similar amount of n-3, and land mammal fat (which they also ate) has much less. None of those sources ate much n-6. He says that the Inuit were healthy despite their 10% n-3 intake, rather than because of it, which may be true. That kind of n-3 intake is certainly not required for good health.

I skimmed over Ray Peat's article as well. As usual, he is so far away from the mainstream that it's difficult for me to even evaluate his points. Sometimes I feel he gets a bit too disconnected from reality by reading cell culture experiments and other in vitro data. But I can't really evaluate the article at this point. I'd have to spend a while thinking about it.

Anonymous said...

The Inuits also ate raw fish often. Brian Peskin's point is that people are eating damaged, cooked, refined omega-6 and that's the problem, not the lack of omega-3 or a high ratio of omega-6. Just because primitives ate certain amounts of omega-3 does not prove that this was optimal. It is only anecdotal. Also, they lived in a different world.

IIRC, the Nutri-Spec articles said that Eskimos ate 10% PUFAs, not 10% omega-3 by calories. They got some omega-6 from land animals and fowl, they didn't just eat fish. They had known problems resulting from their diet, like pathological bleeding & hemorrhage. Stefansson believed the Inuits aged faster than Westerners of his time, who were eating a low PUFA diet relative to now.

I like the fact that Ray Peat is so far away from the mainstream. It is like he's talking another language, his ideas are so revolutionary. But he's not the only one making a case against dietary PUFAs. There's lots of research showing that omega-3 is pathological. The studies saying it is beneficial are short-term (weeks or months), and they look at people eating the typical diet filled with PUFAs, trans fat, wheat, and sugar. They need to try removing variables instead of adding them.

Stephan Guyenet said...


Very interesting observation about the oatmeal. I wonder if the same would happen if you bought raw whole oat groats (they're typically toasted) and soaked them 12 hrs before cooking. That would get rid of many of the lectins and anti-nutrients. If that improved your response to it, that would implicate them in the problem.

I've been reading about the exorphins in gluten lately, and their ability to increase insulin secretion. I wonder if that plays a role. I did eat a large slice of white bread with my makeshift OGTT, which is rare for me. I should repeat the test with just potatoes.

Anna said...

I saw Diana's first posting, so it was up for at least a while. I chuckled over Stephan and Diana describing 133 as a big spike. I guess it's all relative.

Just curious, Diana, how come you are still using a fairly moderate (relatively) amount of carbs in your meal to test? Why not really give yourself a challenge and see how high? You'll only go really high (or stay high) if you have impaired glucose tolerance; wouldn't that be a good thing to know?

I've had BG readings all the up to the 250s when I've consumed *really* challenging high carb foods (I tried out my BG meter right after buying it at Costco with one of their food court blended iced mocha drinks - ok, food-like substance, and saw a reading in the 250s - yikes). At that point, I'm not so concerned about a potential 10 mg meter error margin, for obvious reasons.

A large piece of Costco pizza will push me up to around 200, give or take 10-15 mg. I don't eat those things typically, of course, just when I was really challenging my BG to see how bad it was and gathering data before I brought this up with my doctor and asked for a GTT. But others eat these things or worse on a regular basis without a thought.

Jonty said...

Why not test your fasting insulin and find out if you're insulin-resistant: 2-3, not to worry, 10+, you're "carbohydrate intolerant." Easy to fix at your age.

I'll leave it to others to experimenting with oats - hyperglycemia happens too fast to control with insulin, causes insulin resistance, takes me 4 days to get back on track. Nothing could say "carbohydrate intolerant" more clearly. The medical surprise is that I'm insulin-dependent and my HbA1c is 5.5%: diabetic complications are iatrogenic!

Stephan Guyenet said...


My fasting insulin is 2.3.

I take it you are a type 1 diabetic? 5.5% is a fabulous HbA1c. So you get that reaction with oats specifically, and not other carbohydrates?

Jonty said...

Yes I'm Type I, I think normal bg's are within the reach of any diabetic who avoids starches.

I react to carbs more or less according to their GI but porridge oats were out of control, grape nuts were #2.

2.3 is good, if you are hypersecreting it'll likely rise eventually, I started putting on weight at 30. Easy to fix by cutting carbs. Good luck!

Jonty said...

Wow, I just found your Sept 11 post with your tests - they say: re-test! So many SDs away from the mean just screams "Error."

(i) I've had back-to-back chol tests which differed by nearly 100 in TC in my doc's lab - which is unregulated(!) - and commercial labs tested with errors of + or - 18%. (ii) My ex-wife always tested 100 points higher TC when under stress. (ii) Lindeberg emailed me that HbA1c tests aren't standardized - all the diabetics in his 2007 paper had HbA1c under 5%(!) - so it might be worth looking at averages for the lab to find out where you stand.

It'd also be really interesting to retest after 90 days of lowcarb ...

Stephan Guyenet said...

Hi Jonty,

Thanks for the information. I probably won't get re-tested because I just did it out of curiosity. Plus I haven't gotten the bill yet! But you're right, it would be interesting to try all sorts of things and re-test.

Regarding the high HDL, the RN told me it's the highest he's seen in 22 years. So it's not just that the lab consistently gives high readings.

That's amazing that the diabetics in Lindeberg's study were all under 5%! I hope he had a control group for that one!!

Anonymous said...

"I used the formula and it says my average BG is 118 mg/dL. Considering I haven't gotten a reading over 116 mg/dL yet (and my fasting is in the low 80s), I have a hard time believing my average is that high. I'm still testing just to be sure, but I think there's something else going on."

Not uncommon.

Over time I've met several people who can predict their A1c quite accurately from spot BG readings. Others are consistently and reproducibly high, or low.

My BG used to bang about several times a day mostly between 180 and 70 (seldom "genuine" diabetic numbers but well high enough to do the same damage): my A1c was 5.3

After nailing my BG so it seldom exceeds 120 and seldom drops to 60, my A1c went UP to 5.6. So yeah, other factors probably are involved. Theoretically blood cells are only reversibly glycated initially so don't react to brief BG spikes (other tissues aren't so fortunate) but other factors make some people high or low glycators. What these factors are and whether they can be changed I'm not sure.