I wander if the effect is due to the milk solids or the type of fatty acids. If it is the milk solids making ghee will solve the problem.

So the low fat control meal of pasta and bread gives the LEAST amount of insulin response!

So, to avoid insulin spikes you must minimize all types of fats.

Fat+Pasta seem to trigger the insulin.  The glucose response does not seem to matter all that much.

Thanks for the great info as always.

Is it correct to say, in view of the graphs, that this type of dairy-induced insulin elevation is not mirrored by blood glucose readings, so we couldn't check for it at home with a glucose meter?

butyric acid causes an insulin response, as does casein...interestingly, although i can't find it, i saw a paper that showed the insulin response to white bread was actually less when combined with butter...

What if you eat it without any carbs (e.g., like I do with my eggs in the morning)? If there's nothing to trigger the insulin in the first place, butter can't make it worse, right? Or am I wrong?

That's an interesting study indeed.  It seems to indicate that eating no fat at all (the control meal) provided better results in every category except glucose rise, and then it was only with the rest of a fairly tight pack.  Adding any fat at all spiked the insulin significantly higher.   This is consistent with the charts in your post "The Timing of Blood Sugars".  Yet you concluded there that carbohydrates should never be consumed without fat.  Based on these studies, it sure looks like a person is trading a little glucose rise for a larger insulin spike when consuming fat.  Which begs the question, which is worse?  Does insulin cause more problems than glucose?   You did say it was the insulin that triggered weight gain...

I remain unconvinced that I should rush out and follow Ornish.

Interesting post -- and my experience confirms it.  Here's something that I recently posted to the OEvolve e-mail list:

As folks might know, I've been struggling with my weight due to my hypothyroidism.  Even though many of my symptoms have improved, I continued to gain weight at the very consistent rate of 3 pounds per month.  Sometimes, mostly due to a bit of fasting, I'd be down a pound or two, but then I'd suddenly gain back all that weight plus more within a day or two, so that I'd be on that steady upward slope again.  

In late February, I was up to just over 150 pounds; that was my pre-paleo weight.  That's not so terrible in and of itself -- although I hated it -- but the continued upward progression was really alarming.  Plus, I'd outgrown almost all my pants!

After listening to Robb Wolf talk about the growth-promoting, insulin-spiking effects of dairy, I realized that I'd been eating a huge amount of dairy lately -- far more than ever before.  I've been buying those huge blocks of cheese from Costco, not to mention those delectable half gallons of cream.  So instead of eating meat-and-veggie leftovers for lunch, I was usually eating some high-fat dairy.  And for breakfast, I would eat cheese rather than meat with my eggs.  

A few weeks ago, I decided to cut my dairy down dramatically -- to basically just one serving per day at most.  So I'll have a cup of raw milk kefir or cheese on meatza, but not much more than that.  That was really hard for me to do for the first week, as I really was used to eating it as a staple.  I'm also not eating nuts, absent some "must have food as I run out the door" emergency, as I know they can hamper weight loss for me.

The results have been good so far.  I've not gained any more weight in the last two weeks.  Instead, I've lost four pounds.  Given what I had been experiencing, that's pretty remarkable.

I'm not sure how much more weight I'll lose just due to cutting back the dairy.  I definitely think my metabolism isn't quite up to speed yet due to the hypothyroidism. However, my experience suggests that eating boatloads of dairy -- even good-quality, high-fat dairy -- is an excellent way to gain weight.

That being said, I don't think that it's necessary to cut out all dairy to lose weight.  I'm still eating some dairy, and I'm losing weight.  Plus, I lost my original 18 pounds with my 1.5 shares of raw milk -- that's 1.5 gallons per week.  (I'm now down to just 1/2 shares, meaning 1/2 gallon per week.)  I just think dairy is something to consider reducing if you're not meeting your weight loss goals.

***

I've not lost much more weight since writing that, but I'm definitely not gaining any weight.  That's huge: I was petrified that I'd be 200 pounds by next summer.  

I've probably stalled in that weight loss because my metabolism is still screwy due to my not-yet-fully-managed hypothyroidism.  Plus, I'm not seriously trying to lose weight right now: I don't want to stress my already over-stressed body.  But perhaps cutting out dairy entirely would make a difference.

Anyway, thanks for putting some science behind my experience, Dr. Davis!  I'll definitely add a link to it in the discussion of dairy on the principles page of Modern Paleo.

-- Diana Hsieh (Modern Paleo)

My working assumption has been that I don't need to worry about any increased insulin resistance with high saturated fat intake since I don't eat foods which spike my glucose. Am I correct?

Yikes! Never would have thought butter could do this. Are there any other non-carbohydrate foods that are known to cause insulin secretion?

If anything, it seems like the study supports the idea that all fats are bad, doesn't it?

The zero fat control meal does better than all the fatty meals. Look at insulin for instance.

I notice that you conveniently ignored the pasta (sugar/gluten) in the study, and attributed ALL of the negative effects to butter. Try again, this one failed.

Hmmmm, Wheat Pasta and Wheat Bread?
All bets are off. Would be curious to have conducted tests without either of those included.

Could butter still make you fat if you're eating below BMR?  It seems to me that if you were eating 1000 calories per day consisting of only butter, and your BMR was 1800, you would still lose weight.  Is there anything that scientifically proves this to be false?

Any ideas why ?

I believe insulin drives the rate of ageing, and this would certainly make me scale back my butter consumption.  

But what part of butter causes this ?

Dear Dr Williams, I find this topic very interesting, but being a lay person would like to know a bit more. What would the "meal" they talk about be? Would that be grains? Is the resuls they got from combining grains with cream or butter? Would butter or dairy products have the same effect in the absence of grains?

Doesn't this study mainly show that butter is a bad idea when eaten together with lots of carbs?

Is it really likely you would get the same insulin response if the butter was eaten without all that carbs?

After removing dairy, grains, and anything else processed. I'd imagine it would be hard to eat too many calories when all you're left is meat, veggies, and fruit but not too much. Just curious, is that all we have to choose from?

Interesting.

But the meals were carb-heavy, too.

"The subjects then ingested, within 15 min, a fat-rich meal consisting of dietary fat [50 g/m2 body surface area of butter, refined olive oil (ROO), high-palmitic sunflower oil (HPSO) or a mixture of vegetable and fish oils (VEFO) along with a portion of plain pasta (30 g/m2 body surface area), one slice of brown bread, and one container of skim yogurt.

I wonder if the same effect would be noted if were JUST butter, or oil.

Is it the butter? Or the wheat pasta, brown wheat bread and lactosey yogurt?

Thanks for taking the time to post a response to inquiring commenters.

Interesting. Here are the numbers from the study for insulin AUC:

19960 +- 2766 control
27970 +- 2107 VEFO
29619 +- 4975 MUFA
34749 +- 1167 HPSO
37582 +- 4364 SFA (butter!)

Unless I've misread the study, the "control" meal has no fat at all, and is not isocaloric with the test meals. "the macronutrient profile was as follows: 72% fat, 22% carbohydrate, and 6% protein (see Table S1 under "Supplemental data" in the online issue). The subjects also consumed the same test meal containing no fat as a control meal". So the butter test meal has nearly 4x as many calories as the control meal. The AUC for insulin between the two is about double for the butter, but that seems like a useless measure.

Comparing like-for-like, the spread between average insulin AUC for butter meal and the isocaloric alternate fat meals  shows the butter to be about 34% higher than VEFO and only 8% higher than HPSO. (I have to be pedantic here and point out the spread between averages is not as good as would be the average spread, but I can't get that unless I have the raw data from the study.)

If you want to claim that butter has some unique ability to raise insulin vs. other fats, this is not strong support.

You say "butter's insulin-triggering effect, doubling or tripling insulin responses", and "butter makes you fat". I don't see support for either claim when I read that study.

Why ?

How about cream and fatty milk, or other dairy products ?

I guess coconut is again the safest bet here...

I'm not an MD nor a research scientist. But decades ago, I was a math major. If I read the study and do my math right, the subjects ate an 800 kcal meal of which 72% was fat, for 576 kcals worth of fat (the other control meal components were pasta, bread, and skim milk yogurt).

576 kcals worth is nearly 6T of butter -- or 3/4 of a stick! To me, that is important additional context for the increased AUC for both the insulin and the triglycerides.

Considering my typical use (~1T/meal) and considering the potential downside of replacing carb calories with more PUFAs (and omega 6s -- even 1T of EVOO has over 1g of omega 6), I'm not ready to throw out my ghee just yet.

Is it the milk solids that do it?  if so, then ghee may be an alternative.

as a  personal anecdote - eating butter (even in amounts of 200 g. a day) speeds up my fat loss - i feel as if i'm burning inside (elevated metabolism?)...

I see that I touched a nerve.

This study is not meant to stand on its own, but taken in the context of other studies.

I think that you should eat butter . . . in small quantities and occasionally. I do not believe that the data argue for liberal use every day, or else you chance triggering insulin. Remember: It's the interpretation of the data in the context of the broader experience that leads you to practical conclusions.

What about coconut oil as a butter substitute? There is a lot of hype and hoopla about it being a "safe" medium-chain saturated fat to use as a butter substitute.  Anyone tested if it spikes glucose?

Pastured butter rocks. My belief with butter is "a stick a day keeps the doctor away."

It's as though the more butter I eat the leaner I get!

The anthropological evidence really contradicts the data you and loren cordain cite. The pancreases of many people in traditional cultures carried them toward 100 years old, reacting to dairy every day, multiple times per day.

Here is a 112 year old woman from a very long lived people who use dairy as a staple.

http://news.bbc.co.uk/2/hi/europe/8550374.stm

"Fat, in general, does not make you fat. But butter makes you fat."

Why make such a comment in regards to butter?

Dr Davis.

First of all, thank you for your wonderful site and your willingness to impart life changing and saving advice for anyone to access.

Peter at hyperlipid has written a counter on this particular blog entry of your.

Would you care to comment?

http://high-fat-nutrition.blogspot.com/2010/03/butter-insulin-and-dr-davis.html

So let me get this straight: Ingesting carbohydrates including the bad guy wheat does not do much to blood glucose levels and insulin response. But adding fat to the mix at least doubles, or, in the case of butter, quadruples the peak insulin response? That seems like the best refutation of all the low carb/paleo/primal/ef approaches I've ever seen. So where is the catch?

The last sentence made me cry.

It seems like butter (or cream) without carbs does not create insulin spikes according to the study that Peter at Hyperlipid discusses here:

http://high-fat-nutrition.blogspot.com/2010/03/butter-insulin-and-dr-davis.html

If you eat fat, ASP levels will rise, which will make you gain fat. ASP levels increase in response to an oral fat load, and ASP is one of the most potent stimulant of triglyceride synthesis.

So, if you eat carbs, insulin secretes and you store it as fat, if you eat fat, ASP levels rise and you store it as fat. Why is it only the insulin's fault?

http://www.jlr.org/cgi/content/abstract/30/11/1727

The oracle's take on your post doc:

http://high-fat-nutrition.blogspot.com/2010/03/butter-insulin-and-dr-davis.html

I don't know about the effect of butter, but when I eat a fatty meal (like a cheesy omelet with an avacado), my BG goes down several points.

Here's a study that isolates cream and shows a very small insulin response: http://drbganimalpharm.blogspot.com/2009/12/insulin-and-aging-how-paleo-works.html

This is becoming more confusing by the minute. First, no carbs,only fats and protein.

Now, no butter, no dairy, no, carbs, just a few drops of fat and protein.

I am going to cry, like the previous poster.

"Fat, in general, does not make you fat. But butter makes you fat."

This seems wrong. When I eat more butter, I lose fat/weight. Eat more carbs, gain fat/weight. Been noticing this for years. Sure, put a load of butter on noodles or rice, and you'll gain plenty, but just keep the carbs low and there's no problem. This seems like a case of being blinded by science. Maybe the science isn't exactly wrong, but just incomplete or misapplied.

I think I fugured it out...

The AMA or AHA has kidnapped Dr Davis and they are forcing him to post these confusing refutations.

What will we see next?  Studies that show how Cocoa Puffs actually will help reach our 60-60-60 goal after all?

After reading your post and the hyperlipid post I'm left with a few questions.

1.The big one, why is butter raising insulin so much? I had previously thought it was only high GI carbs that did that. Hyperlipid seems to think it was the small amounts of casein.

2.Is it right to say that since the control carb meal only contained 174 calories compared to the butter meal which contained 800 this study is fairly useless? It raises some interesting questions but you can't conclude anything until you have a proper control containing the same amount of calories.

3. Are there any studies which measure insulin secretion after a pure butter meal with no carbs?

Even though the study is inconclusive, I'm lowering my butter consumption until I'm sure it isn't causing an agrovated insulin response. Even though butter causes much higher levels of FFA's which mean more of the fat is being burned, insulin levels that high can't be healthy

All this information hurts my brain...ugh. All I know is when I eat low carb... I lose weight... feel better and more energetic. Sometimes there's just too much analysis and information about what constitutes a healthy diet. It's all opinion and based on research. Then other research refutes that idea.

My Grandma used to say, "Eat when you're hungry and drink when you're dry."

BTW... Grandma lived to be 95.

Dr. Davis,

I'm not too sure this is right.  I don't have the time to hunt down the links right now, but there was a study on Swedish children recently that showed those eating full-fat dairy (butter fat) were leaner than those who ate low-fat dairy.  Also, butter contains a lot of CLA, which is now being sold as a fat-loss supplement.  I can't parse this study on its own merits like Peter or Pyker (though I thank them for doing so), but I remain unconvinced.  

I also agree with other posters about the in-vivo, long-term evidence of dairy-reliant cultures and their longevity.  

Again, with so many cautions of what not to eat, I'd love to see a Dr. Davis-approved diet plan.  If I were just following all the Don'ts, I'd go crazy (and hungry).

The palmitic acid content of the butter would probably trigger  insulin when there is an abundant source of glucose in a meal.  But, if you look at this post by Stephan you will see that the body stores dietary fats when there is a glucose source of energy available.  

http://wholehealthsource.blogspot.com/search/label/fats

As the glucose is depleted the fat is released to be used as energy.  The graph in Stephan's post clearly shows the cycling that occurs in normal metabolism.


The information Dr. Davis presented showed an insulin spike, not continual high levels of insulin that would inhibit the release of FA as an energy source.
I believe it was Dr. Eades who recently posted about the constant exchange of fats in and out of storage, and the implication is that storage is only a problem when cells remain saturated with glucose.  The fat never comes out because it isn't needed.  In this scenario insulin resistance and fat storage is protective.

http://high-fat-nutrition.blogspot.com/2010/03/getting-fat-is-good-official.html

My understanding of this subject is, by limiting carbs, especially the sugar surrogates that are refined grain products, you can keep the normal cycling of energy sources going.

When everything is working right, your body signals hunger and satiety appropriately, and the body runs at an equilibrium level of energy intake and energy expenditure.

I suspect that the FA composition of butter makes it available as an energy source sooner than the other FA's tested in the study.  Hence the higher insulin level.


I don't see where a moderate amount of butter, when there is a limited or low availability of glucose sources will make you fat, expecially if you are VLC or on a ketogenic diet.  I can see where it could be a problem if you were eating 150 or more grams of carbs per day and have metabolic syndrome.  Then high levels of insulin would spike even higher prompting the liver to pump out more glucose, and the fat just has no where to go except into storage.

Take your vitamin D3 to get your adiponectin levels up.  Take fish oil while limiting Omega-6's to normalize inflammatory responses. Limit carbs, and get the glucose-FA cycle working for you.

Eating carbs with butter produces a big insulin response because butter induces temporary insulin resistance (IR) which means that more insulin has to be secreted to get the same net effect. Looking at the blood glucose response, the net effect is the same. So stop panicking. The temporary IR lasts for only a few hours.

"Fat, in general, does not make you fat. But butter makes you fat."

If you based your conclusion on the graph in this study, all fats make you fat.

Protein also stimulates insulin, so protein makes you fat.

Carbs stimulate insulin, so carbs make you e biggest ingredients that women find that helps with their skin health is mineral products. They are natural and very breathable on a woman's skin.

DD: I think your velvet lip pencils are by far the most extraordinary lip pencil on the market. What are some of your other standout products your customers love?

JM: Pure Press Mineral Foundation is still our #1product. But the fastest growing product is Magic Marker Eyeliner. It is easy to use, doesn't smudge and lasts all day.

DD: Anything new on the horizon for Mirabella that you can share with us?

JM: In August we launch CC crème. This product has all the good for you ingredients to help with Anti-Aging like avocado oil, argan oil and Acai (Assai) berry. Plus it is a mineral formula, gluten-free, and paraben-free. And it has an SPF of 20. One of the biggest issues that women have with aging is lips. That is why we put Litchi Chinesis Fruit Extract in our Colour Vinyl lipstick. Then in your favorite Velvet Lip Pencil, we put Pomegranate Extract, Vitamin C and E in to assist with in Anti-Aging.

DD: Is Mirabella only sold in the US or do you have international distribution as well.

JM: We are sold in Canada, Australia, Finland and Russia.

DD: Where can we purchase your cosmetics?

JM: Our products are available at www.mirabellabeauty.com and at over 1,500 of the finest salons and spas. Go to our salon locator to find a retailer near you.