This article could easily be subtitled “The Study That Tells You Everything You Need To Know About Insulin, Blood Sugar, Carbohydrates, Satiety, And Obesity”. Yes, I admit to a degree of hyperbole—but this study is so well instrumented and controlled, and its results so informative, that I believe it’s important for everyone to read it.
PEDIATRICS Vol. 103 No. 3 March 1999, p. e26
High Glycemic Index Foods, Overeating, and Obesity
David S. Ludwig*, Joseph A. Majzoub*, Ahmad Al-Zahrani*, Gerard E. Dallal, Isaac Blanco, and Susan B. RobertsThanks to the American Academy of Pediatrics for making the full text of their archives freely available.
You should really click on the fulltext link above and read the study yourself, because it’s very clearly written…but as not everyone has that kind of time, I’ll cover the important parts.
The design of the study was simple.
- Take twelve obese teenage boys.
- Admit them to the research center the evening before. Feed them dinner and a bedtime snack (the same each time).
- In the morning, feed them one of three different breakfasts, each with equal caloric value but dramatically varying composition.
- Measure blood samples and subjective perception of hunger every 30 minutes.
- Feed them the same meal for lunch.
- Repeat blood and hunger measurements for the next 5 hours.
- Allow them to request food at any time after lunch. Measure when and how much they ate.
- Repeat after 1-2 weeks, until everyone’s been measured for all three meals.
The three breakfasts and lunches in question:
- “High-GI”: Instant oatmeal with 2% milk, a tablespoon of cream, and glucose plus an unspecified “artificial sweetener”. Milk was treated with lactase in order to increase GI. 64% of calories from carbohydrate, 16% from protein, 20% from fat.
- “Medium-GI”: Steel-cut oatmeal, prepared as above—but without the lactase, and with fructose instead of glucose and sweetener. Same macronutrient composition.
- “Low-GI”: Vegetable omelet made of appx. 1 whole egg and one egg white, low-fat cheese, spinach, and tomato, plus some grapefruit and apple slices. 40% carbohydrate, 30% protein, 30% fat.
All three meals contained the same number of total calories, and weighed approximately the same. (And, if anything, caloric availability would have been greater with the instant oatmeal than with the steel-cut oats.)
How “Heart-Healthy Whole Grains” Make Us Fat—In Pictures!
Now let’s look at the graphs!
Square = "high-GI meal" (instant oatmeal). Circle = "medium-GI meal" (steel-cut oatmeal). Triangle = "low-GI meal" (omelet + fruit).
“The mean area under the glycemic response curve for the high-GI meal (284 mmoles-min/L) was twice that of the medium-GI meal (141 mmoles-min/L; P < .001) and nearly fourfold that of the low-GI meal (76.6 mmoles-min/L; P < .001)."
Is anyone surprised that a big pile of high-carb oatmeal spikes blood glucose and insulin, and hammers glucagon?
Do you see the huge epinephrine (= adrenaline) spike four hours after the instant oatmeal, when the sugar hit wears off? (And the start of one an hour later, with the steel-cut oatmeal?) How do you think that makes you feel? Nervous, irritable, and desperately in need of another sugar hit?
Moving on: here’s the subjective hunger level, charted over time, after the three test breakfasts. Again, these results shouldn’t be a surprise to anyone:
Again: square = instant oatmeal, circle = steel-cut oatmeal, triangle = omelet + fruit.
Note that there is no time at which the hungriest omelet-eater was more hungry than any oatmeal-eater…and that after five hours (approximately the time between breakfast and lunch), both the instant and steel-cut oatmeal-eaters were approximately 65% hungrier than the omelet-eaters!
What do you think that means at lunchtime? It means the “heart-healthy” oatmeal breakfast will leave you ravenously hungry at lunch (if you even make it there without snacking), whereupon you’ll gorge and suffer an hour or two of “food coma”. Goodbye, afternoon!
Now here’s the punchline: the test subjects were fed the same meal for lunch as they had for breakfast. Over the next five hours, they were allowed to request a snack platter and eat all they wanted, as often as they wanted.
How about that?
These results speak for themselves:
“Voluntary energy intake after the high-GI meal (5.8 megajoule [mJ]) was 53% greater than after the medium-GI meal (3.8 mJ), and 81% greater than after the low-GI meal (3.2 mJ).”
[...]
“In addition, mean time to the first meal request after lunch (2.6, 3.2, and 3.9 hours for the high-, medium-, and low-GI meals, respectively) differed between test meal groups (high GI vs low GI; P = .01; high GI vs medium GI, not significant).”
That’s not a misprint. People consumed 81% more calories during the five hours after eating instant oatmeal than after eating the same number of calories as an omelet and fruit—and 19% more calories after eating steel-cut oatmeal than after eating an omelet and fruit. (Note that the hunger curve for both kinds of oatmeal was rising precipitously at 5 hours, whereas the omelet + fruit curve flattened out. Do you ever have to work late? Is dinner always five hours after lunch?) Furthermore, the omelet-eaters took 50% longer to request any food at all.
That’s not all! A modern “heart-healthy” oatmeal breakfast, as mandated for schoolchildren by the new US government guidelines, would use skim milk and no cream (instead of 2% milk plus a tablespoon of half-and-half), driving the GI even higher and fat content even lower. A traditional* breakfast would use whole eggs, butter or coconut oil, and full-fat cheese (or ham, or bacon, or just an extra egg), driving the GI even lower and the fat content much higher! (Recall that between the apple and grapefruit, the low-fat cheese, and the egg whites, the “low-GI” omelet breakfast was still 40% carbohydrate and only 30% fat.) So real-world differences will be even greater than this experiment demonstrates.
* These days “traditional” is confined to a few ghettos called “paleo”, “primal”, or “Weston A. Price”. But less than a hundred years ago, nearly everyone who could afford real food was smart enough to eat it.
Conclusion: Eat Real Food, Not Birdseed
Do you want to be fat, constantly hungry, or both? Keep eating that birdseed. (Known to our overlords as “heart-healthy whole grains”.) The US government pays Big Agribusiness billions of dollars of our tax money each year to overproduce grain—to the point that we’re forced to put corn ethanol in our cars at a net energy loss just to get rid of the excess. Your poor health and shortened life help keep their profits high.
Do you want to be healthy, happy, and strong? Eat real food.
Live in freedom, live in beauty.
JS
Interested in learning more? You might also enjoy my three-part series on carbohydrate addiction: Why You’re Addicted To Bread, Kicking Your Cereal Addiction, and The Myth Of "Complex Carbohydrates". Or, start with the index.
Meanwhile, you can help rescue others from being crushed under the “food pyramid” by using the buttons below.
That’s some nice graph work there. Great post. I remember my rice milk, cereal and honey days…with great contempt. That was not fun! I didn’t even need to lose weight but I thought it was healthy, durr. I had lamb balls, sweet potato and creamed coconut for breakfast today and didn’t need to eat for 7 hours.
My one criticism would be that some of it will have to do with leptin and insulin sensitivity. The insulin response won’t be as potent and hunger won’t be quite as bad in ideal metabolic situations, but the secretion of peptide YY, CCK and all that is obviously better with fat and protein than carbohydrate so all of these dieticians try to coerce people into starving themselves should logically recommend the eggs, bacon and butter. Fat chance.
I also think that the inflammatory nature of grains ties into things, but that’s a discussion in itself.
Great post – having dabbled in the “body building” scene and always being an avid workout enthusiast (?) I can say I paid attention to the “GI” of things. That made me spiral WAY downwards, especially all the small meals. After a while it became insane the amount of work I had to put in. The irony? After I dropped the oatmeal (which I wasn’t eating much of anyhow), skim milk (I preferred full-fat raw from Jersey cows anyhow. Tastes like icecream), and low-fat nonsense; I started feeling better. By that I mean, even though I got down to ~3-4% bodyfat it was rough on me, my body, and my mind.
Low-fat (and especially with moderate-carb to low-carb) REALLY does affect you. I never felt full, I lost all my libido [who thought a twenty-something could lose THAT?], and I always felt tired and hungry. I won’t be making that mistake again. I’m pretty sure that I’m still recovering from it, but lesson learned. Plus, I forgot how good animal fat tastes. Try eating broccoli, steamed, without butter or coconut oil for a while. You’ll learn quickly how DAMN GOOD it tastes.
Thanks again JS!
Hopefully some of the people that do eat bird seed will stumble upon this thoughtfuly analyzed study. But unfortunately, us paleoers already comfortably skip snacks and don’t rip peoples heads off before dinner.
Anyway, thanks for taking the time! It’s nice to see those numbers. I’m sure I’ll probably reference this from time to time.
About this study. It tested meals with different Glycemic loads, and different ratios of protein, carb and fat.
It shows that a high carbohydrate meal made of oats is high GL, and has a bad response. A meal of starchy tubers may also show the same result, if the same GL was picked. That does not prove that grains are a problem, it proves that a high GL, high carb meal is a problem. To show grains are the culprit, you would have to compare a similar high starch paleo meal to the grain meal.
You would also have to compare a balanced 40:30:30 protein carb fat meal – one using grains as carbs, the other paleo carbs to compare paleo and grain carbs to see if there is a different impact. This would be another study.
What we can say from this study is that a high carb, high GL meal is much worse than a higher protein, higher fat,low GL meal.
Stabby:
No, insulin isn't the whole story, and leptin and PYY are other parts of an interdependent system I don't claim to completely understand. I also agree that wheat has special 'appetitogenic' properties. What I take from this study is that a quickly absorbed meal of whole grains, e.g. sugar (glucose) with impurities, leaves us all more hungry, and makes us eat more, than a slowly absorbed meal of real food, such as meat and eggs.
Aren't lamb balls kind of small? How many of them do you need to make an entire breakfast?
Chris:
Getting yourself down under 9-10% bodyfat is rough anyway…I can't imagine doing it on a high-carb diet. Ouch! I'm glad you're back in a happier and healthier state.
And no, I don't like steamed vegetables either. I'll have mine sauteed, please, in butter or coconut oil. Grilled is acceptable too if I can butter them afterward.
Peggy:
I do my best to write articles that you can forward to people who aren't already paleo, so they'll understand why you won't eat table bread or their delicious homemade muffins. It's tough to overcome decades of dogma…I have the perhaps naive belief that reason and logic will help.
Julianne:
I agree with you that the results aren't as simple as “GI drives satiety”, which is why I continually refer to the meals by content (instant oatmeal, omelet, etc.) instead of the “high-GI meal” and “low-GI meal” terms used by the study authors. There are a lot more differences between the meals than just their GI, or even their macronutrient composition.
I partially disagree with you, however: the study does indeed prove that “heart-healthy whole grains” are a problem. I agree with you that they're not the entire problem: other high-GI starches served with very little protein and fat are also very likely to be fattening…but as the study didn't measure that, I felt I should limit my conclusions to what was explicitly tested. (I also find it interesting that the omelet + fruit was jiggered precisely to hit the Zone 40-30-30 proportions…I bet at least one of the authors was a Zone dieter at the time.)
I agree that it would be interesting to test some of those other comparisons to see if there are uniquely fattening characteristics for different starches, e.g. wheat vs. oats vs. potatoes.
JS
I wish I’d found this study a few weeks ago when I was writing a similar post! I’m emailing this to family members though. It seems like common sense to those of us who already eat this way, but hopefully as more research like this comes out, this info will be more mainstream.
Katie:
Isn't it a great study? I found a link to it in the comments section of some other paleo blog…it wasn't featured or anything, just sort of left there as a “by the way”.
It's so hard to find anything that compares (mostly) REAL FOOD to hearthealthywholegrains: most studies just compare whole grains to junk food and conclude that whole grains are better…which is much like comparing “low-tar” cigarettes to regular cigarettes and concluding that the “low-tar” versions are healthy.
JS
New to your site, here via TWIP. Nice post. As you say, “eat like a predator, not like prey”. I’m afraid this study, though, will not be front page news. Now, are all yor posts this interesting?
Darn – just did big post and your site didn’t like it.
I agree with what you say. Re the author was zoning – likely it was done in Boston children’s hospital and they designed a weight loss programme OWL as a result of this study. Sears lives in Boston too.
This study and the graphs are great – I’m glad you highlighted it.
I had an argument with Sears on this topic – as he only seems to have one paradigm 40:30:30. I argue that food quality matters more, he argues that ratio is paramount.
http://blog.zonediet.com/2011/01/24/increased-sati…..ight-loss/
Julianne:
That's a great dialogue, and I commend you for introducing paleo principles to the Zone people. From what I know of the Zone diet, I agree with your take on it.
I'm sorry your post got eaten! That's the first I've heard of that…I'll try a few tests myself and see if I can reproduce it.
Meanwhile, I make a practice of using ctrl-A and ctrl-C to copy my comments before submitting them to other people's websites, just in case that happens to me (which it has often enough to make the extra keystrokes worth it.) You can paste it into Notepad or TextEdit if you're not sure it copied correctly.
JS
I notice that posts with several links get eaten – they go into the spam filter I think. I did save and re-post, but I got the message back that it was a duplicate! So I started again.
I got to your blog from MDA and I gotta say, you are great. Witty and to the point. That was one heckuva interesting study – but not at all surprising.
Kem:
Thank you for the implied compliment!
My strategy is to post a limited number of articles: it allows me to give each topic the time and energy it deserves, and it takes the pressure off to come up with something new every few days. I'm happy with the results, and I think you'll find my other articles are of a similar depth.
Julianne:
OK, that makes sense: I'm running the Akismet spam filter. Contact me next time a comment is eaten and I'll see if I can fish it out. Also note that if you actually create a forum account (links in the right sidebar) and log in to it, your posts should bypass the spam filtering process, as well as automatically linking your website.
JS
I found it interesting that the high GI meal caused the highest spike in serum growth hormone. I am implying that that a low GI meal might be better for fat loss, while a high GI meal might be better for muscle growth. Then again, muscle growth is probably best for fat loss…
All other graphs except growth hormone and epinephrine is about the same after the 4 – 5 hour mark. I wonder if it would be beneficial to eat a low fat, low protein, high GI meal 4 – 5 hours before a workout and then eat a protein rich and medium GI meal after lunch to allow an insulin spike to shuffle nutrients into the cells, and utilize the high GH levels.
It certainly rhymes with the leangains approach ( 16 hour fast / 8 hour fed state with fasted workouts ), which I am loosely following.
I’m only a n=1 experiment, but after eating primal for about 600+ days and working out I added about 150 grams of carbs per day into my diet, and dialed back about the same amount of calories in fat and I am now another notch in on the belt, while the weight stays the same.
Thoughts?
I think your approach – fewer posts in greater depth is the way to go. I bookmarked your site and I’ve been following paleo since Art De Vany and partake of all the paleo/low carb usual suspects. If you didn’t add value I wouldn’t be here.
Sungrazer:
Yes, that is indeed an interesting piece of data. I'll have to do some research, though, to determine whether the spike is significant relative to that produced by, say, exercise. It might also be an artifact of random noise, as GH secretion is very spiky and occurs periodically throughout the day. I've filed the diet/GH connection under “something to look into further.”
As far as adding carbs, I think that the more active you are, the more carbs you can/should ingest in order to keep your glycogen stores reasonably full. I'm not sure how liver and/or muscle glycogen levels affect metabolism and, particularly, energy storage: that's another connection to file under “look into this further”.
150g is about 600 kcal, which is about 20% of 3000 kcal, which is a reasonable daily intake for an active person. I believe that <20% of calories from carbohydrate is more of a weight-loss tool than a state to permanently aspire to, and I personally don't do as well under that threshold, so I think I agree with you. (Although some people do have metabolic issues that require them to stay VLC or else gain weight dramatically.)
Walter:
I appreciate your support! It's good to know that people who have been around much longer than I have still find my articles valuable. Attention is the scarce resource on the Internet, and I try to only demand it if I have something to say.
It's not the quick way to get readers — but I'm not doing this for the money. I'd rather have the respect of others in the field, and my small crew of perceptive commenters, than a million page views per month and pointless flamewars that need constant moderation.
JS
[...] “Heart-Healthy Whole Grains” Make Us Fat How “Heart-Healthy Whole Grains” Make Us Fat - GNOLLS.ORG People consumed 81% more calories during the five hours after eating instant oatmeal than after [...]
[...] articles: how heart healthy whole-grains make us fat recipe: paleo french salad dressing [...]
[...] How Heart Healthy Grains are Making Us Fat [...]
Foods that spike insulin the most in humans lead to the greatest satiety and lowest food intake at subsequent meal. http://www.ncbi.nlm.nih.gov/pubmed/20456814
@Sven-Are:
That study was done on “twenty-two lean, healthy men”. And “participants consumed four liquid test meals”.
Has anyone dissected that study to see why the results contradict the one mentioned in the OP?
Sven-Are, LCC:
That study compares the effects of different types of pure protein, fed as chocolate-flavored liquid shakes:
“Different dietary proteins vary in their ability to influence satiety and reduce food intake. The present study compared the effects of four protein meals, whey, tuna, turkey and egg albumin, on postprandial glucose and insulin concentrations as well as on appetite measures and energy intake in twenty-two lean, healthy men.”
I don't see what that says about the relative satiation caused by actual food – or even that of carbohydrate vs. fat and protein, as there was nothing but protein in this study.
I've shown previously that fat is the major driver of glycemic index (and a major potentiator of satiety)…and as the study in this article shows, carb-driven satiety is inversely related to glycemic index. What's going to get you through to lunch: a shake of whey protein isolate in water, or a three-egg omelet?
It's an interesting study, though, and I thank you for bringing it to my attention.
JS
(Also keep in mind that satiety was measured just four hours later…is dinner four hours after lunch? No.)
[...] sugar and hormonal consequences from a study that compares a carbohydrate meal to a balanced meal. How heart healthy whole grains make us fat. And stop [...]
[...] I’m reading this week: Some more info on grains, and it appears grains can make us fat! http://www.gnolls.org/2052/how-heart-healthy-whole…..ke-us-fat/ Workout of the [...]
[...] I posted this study on my facebook page. The study had three groups of kids being fed breakfasts of instant oatmeal, [...]
[...] How “Heart-Healthy Whole Grains” Make Us Fat by J. Stanton of Gnolls.org [...]
[...] also see similar results in the exhaustively instrumented study I reference in my previous articleHow “Heart-Healthy Whole Grains” Make Us Fat. Yes, it’s theoretically possible that the results in Bray et. al. won’t apply to humans…but [...]
Thanks for reinforcing this …I am new. After years and years of low fat, low cholesterol, I am still throwing it all out and repacking my brain. Lost 12 pounds in the last month and the crazy hungry feeling is almost gone. It feels like I lived with stress and am relaxed now. Any studies on that?
lynn:
One of the main functions of cortisol is to raise your blood sugar by causing your liver to break down glycogen and release glucose into the bloodstream. High cortisol also makes you feel stressed out.
It could be that regaining metabolic flexibility by eating high-fat and lowish-carb, and getting periodic intense exercise, allows you to get by with less cortisol. I'll have to look into that.
JS
I love this study.
Of the six graphs reproduced above, I find the ones on the left to be the most scary.
Check out that epinephrine spike!
AHH:
It's nice to see the instrumentation, isn't it? Especially with so many people wanting to blame everything on “food reward”. Food reward is real, and it's one part of hunger…but it only makes sense as part of a larger picture, and ignoring or deprecating the basic physiology of nutrient ingestion, storage, and retrieval will get us in trouble.
JS
[...] Brain Health with the Paleo Diet The Fear of Saturated Fat and Cholesterol Why Grains Are Unhealthy How “Heart-Healthy Whole Grains” Make Us Fat Evolutionary Genetics and the Paleo Diet What Are The Benefits of the Paleo Diet The Weston A. [...]
I eat grains every day and maintain 7% body fat all year round.
Grains don’t make you fat, you make yourself fat.
Just eat them in the right amount, like 1 whole pitta a day MAX
and you will unfat yourself without any exercise
Aaron:
My grandfather smoked two packs of cigarettes a day for 60 years, and he lived to a ripe old age — well above the American average. That doesn't mean smoking is healthy.
It seems we're in agreement that most people eat too many grain products. As the metabolic disruption caused by gluten grains is well-established, and the nutritional value of all grain products is extremely low compared to animal and vegetable foods, I feel the optimal amount of them to eat is zero.
Some people can tolerate grain better than others, myself included — but I'm not interested in what we can tolerate. I'm interested in what's best.
JS
[...] different meals. The study compares a high carbohydrate grain meal to a balanced meal with protein. How heart healthy whole grains make us fat. And I’ll stop [...]
You mention the epiniphrine spike, but epinephrine triggers lipolysis. Could there be a benefit to spiking epinephrine but then NOT responding to it by eating? So the hardest time not to eat is the most productive for fat loss?
Domenic:
It's a truism that you don't lose weight by eating: you lose weight by not eating. And yes, the reason for the epinephrine spike is to get some energy into the bloodstream RIGHT NOW.
To that end, epinephrine causes both fat and glucose (via glucagon) to be released into the bloodstream. However, this won't do you any good if you're not burning the energy that has just been released – which is why exercise is generally a better way to raise epinephrine than hypoglycemia. Otherwise some quantity of the fat and glucose just gets reabsorbed.
The problem with starving yourself is that you tend to shut down your HPTA, which makes you tired and cranky and weight loss even more difficult. I feel weight loss is best achieved as a side effect of a healthy diet: a healthy body, replete with nutrients, will have a healthy metabolism, which will naturally tend to converge to a healthy weight.
JS
J,
Thanks for the response, im just playing devils advocate, I guess what im wondering is whether there is any benefit to a hypoglycemic spike. I agree with the exercise.
Love the site!
First, let me admit that you guys are way smarter than I am on grains, GI of foods, etc! The following will all be in layman’s terms.
Having read all this great research, I’m very interested about the role of exercise in GI and insulin spikes. I regularly run 10-12 miles about 3-4 times a week. I start the day with toast and butter, head out for my run an hour later, and am oddly not hungry, fine not eating until lunch, which is usually volumetric-ish (1 cup cooked regular oatmeal or two cups cereal, 1 cup fruit, two eggs or a cup of yogurt).
I’ll pay more attention to spikes now because I have two pieces of fruit between then and dinner; how much of that is driven by a high-ish GI lunch? For dinner, I confess to eating any ol’ thing: pizza, stir-fry, etc, but always with salad, broccoli or that kind of thing.
The interesting for me is that on days when I don’t run (after my usual toast and butter) I crave a muffin or something like that mid-morning. So exercise seems to me to be in the equation somewhere.
Again, you all impress me with scientific understanding of the research. Apologies for my layman’s curiosity.