The paper is organized into three sections:
- A comprehensive review of the observational studies that have examined the relationship between high-fat dairy and/or dairy fat consumption and obesity, metabolic health, diabetes, and cardiovascular disease.
- A discussion of the possible mechanisms that could underlie the observational findings.
- Differences between pasture-fed and conventional dairy, and the potential health implications of these differences.
We wrote this paper because after reviewing the evidence, we found it to be surprising and fairly contradictory to conventional ideas on nutrition and health. I wrote the sections on obesity, metabolic health and diabetes, Mario wrote the sections on cardiovascular disease and fatty acids, and Ton wrote the section on husbandry practices and dairy fat composition. Mario was the lead author and did most of the editing/formatting, submitted the manuscript, etc. Our paper went through a rigorous peer review process.
Here are our basic findings:
- High-fat dairy consumption is not associated with obesity, in fact, 11 out of 16 studies found that higher dairy fat intake is associated with lower body fat and/or less fat gain over time. None identified an association between high-fat dairy consumption and fat gain, although some did find an association between low-fat dairy consumption and fat gain.
- High-fat dairy consumption is not associated with poorer metabolic health. Six of 11 studies found that higher high-fat dairy consumption is associated with better metabolic health, while only one found that it was associated with one marker of poorer metabolic health (and this study used an odd design).
- The association between high-fat dairy intake and diabetes risk is inconsistent. Zero of eight studies found that high-fat dairy consumption is associated with diabetes risk, and three found that it was protective. However, three studies also found that low-fat dairy intake was inversely associated with diabetes risk, compared to no association with high-fat dairy, suggesting by inference that the fat content of the dairy could be harmful. These studies all adjusted for body fatness. Since body fatness is a key risk factor for diabetes, and dairy fat intake is inversely associated with body fatness, this is obviously a major confound. We discussed this and other potential confounds in the paper.
- The evidence on cardiovascular disease is inconsistent, with a number of studies suggesting a protective association, a few suggesting a harmful one, and several suggesting no association.
- Dairy fat is a complex substance. There are major differences in the fatty acid composition of dairy from pasture-raised vs. conventionally raised cows, and many of these fatty acids are bioactive and could influence human health.
- We also discuss the limitations of observational studies in some detail, and many other issues that I won't touch on here.
This is the first comprehensive review of studies on the association between high-fat dairy intake and obesity, metabolic, and cardiovascular health. Typical dietary advice includes the recommendation to eat low-fat or skim dairy products. This is based on the hypothesis that avoiding the (mostly saturated) fat in dairy will reduce the risk of obesity, metabolic problems, and cardiovascular disease. This idea is logical, but not every idea that is logical is correct when tested scientifically, particularly when it pertains to a complex natural food. We asked the question "what does the evidence say about this hypothesis?"
The research to date suggests that high-fat dairy overall does not have a negative impact on obesity risk, metabolic problems, diabetes risk, or cardiovascular disease. In fact, these studies offer fairly strong support to the hypothesis that high-fat dairy may protect against obesity. However, there was variability between studies and this may be explained by factors such as a) differences in the quality of dairy products between countries/regions, b) the form in which dairy is consumed (e.g., traditional cheeses vs. ice cream and pizza), and c) other confounding factors discussed in the paper.
Please keep in mind that these studies are observational and therefore can not establish cause and effect. They're best viewed as a springboard for future research.
What This Paper Doesn't Mean
I want to be very clear about this. This paper does not mean that adding butter to all your food will make you lose fat or become healthier. In fact, if you do that you will most likely gain fat and become less healthy. Say what?? The studies we reviewed examined the role of high-fat dairy in the context of normal varied diet patterns. They did not compare people eating normally to people who put extra butter on everything, which is an excellent way to increase your calorie intake. Essentially they compared people eating high-fat dairy to people eating other types of fats as part of a mixed diet. The difference is subtle but critical to understand: addition vs. replacement.
So does this mean that replacing other types of fats with dairy fat (pasture-raised in particular), in the context of a normal varied diet, could lead to less fat gain and perhaps even better health over time? Perhaps. That is what the studies suggest overall. But again, these are observational studies with major limitations, so we'll have to wait for more evidence before we can hang our hats on the idea. In the meantime, it's clear that typical dietary recommendations to favor low-fat dairy over high-fat dairy are on thin ice.
Stephan & Co.
I was in the course of eating 250 gms of Natural Full-Fat Irish Pro-Biotic Yoghurt, when I came across your latest blog entry.
I was curious about what constituted terciles, quintiles or even deciles of consumption in the studies - but, tant pis, I'm not prepared to pay Springer for the privilege.
Congrats on another publication. I think it's time we stopped worshiping low-fat dairy.
Note: I *do* have a connection with the dairy industry. I recently borrowed two Jersey cows to help eat down my pastures. Here in Vermont, pastured dairy cows are the norm.
Of course, if you put butter on everything you might actually end up eating less of it.
Especially if you put butter on your butter.
Good one, this stuff was worth reviewing.
They put free milk back into schools in N.Z. - but it's low fat milk.
They reintroduced it because poor kids were turning up with no breakfasts as their parents were hopeless.
So - reduced fat milk. Because, presumably the reason these kids are starving is, they are getting too many calories from fat!
Some people really are idiots.
Check out the difference in dairy from cows fed with palm kernel expeller
Now, I think that study is actually far from showing CVD causation. The claims made are quite ridiculous given the data. But it does show different composition of the milk and perhaps a loss of CLA.
If you put butter on everything, it doesn't make sense that this will result in a total increase in caloric intake. The implication is that somehow butter causes the hunger/satiety biological mechanisms to break, and thus allow a frivolous increase in caloric intake.
The hunger/satiety mechanisms will make sure to make you eat when you need to, and make you stop eating when you've had enough. This includes calories. However, I have yet to find any evidence for a calorie receptor so I'm not sure how a calorie could be a biological signal for anything. But I digress.
The point is that if adding butter to everything does increase total caloric intake, then the hunger/satiety mechanisms would make sure to compensate in several ways. For example, increase the time before the next hunger signal, decrease hunger signal amplitude, make food taste not as good, etc.
Your own paper seems to indicate that. You noted an association between all kinds of benefits and butter intake. Thus, those who ate more butter probably were more healthy because of that, therefore had better working hunger/satiety mechanisms as a result of that better health.
Nevertheless, your paper is not enough to tell people to do anything about butter. For that, we'd need to test whatever hypothesis your paper generated. I don't know if there's already studies done with butter, but low-carb studies will do just as well to demonstrate the concept. It makes sense. In low-carb studies, when subjects cut out carbs, they compensated by eating more fat. It follows that if subjects were instructed to eat more fat outright, they'd probably cut out carbs to compensate. Low-carb studies also show low-carb subjects spontaneously reduced total caloric intake. Therefore, it's reason to conclude that either carbs make us eat more, of fat is more satiating. Either way, it further opposes your suggestion that adding butter to everything would lead to an increase in total caloric intake, which would then lead to weight gain.
funny. I've been naturally cutting beef and chicken for more cheese and yogurt.
Butter definitely makes me eat more popcorn and drink more beer.
Our pastured Jersey cow and two milk goats will be most happy to hear this in the morning. Also our 6 children. Will be interesting to read the original article; and pass that or your summary above on to those who advocate low-fat dairy.
Perhaps such info, in time, will help small-scale, grass-based dairies survive.
The following posts suggest that this finding should not be surprising to you:
Consequently, I would speculate that other factors besides your knowledge and understanding of the literature influence the content of what you write. This seems concerning.
That's good to know. I myself have always thought high fat dairy was a healthy thing, and now to know that spending those extra dollars for organic grass fed products is worth it makes me feel even better. I just hope that your article will get some good pub in the mainstream media. Prestige is one thing...what the news media says and what Dr. Oz and Dr. Ornish say is of course (and sadly) quite another. I still don't get the insulin thing, though...that's the one thing that is just so hard for me to believe, since I have a diabetic child. I see what insulin does everyday, so to not consider it a major/important pathway for obesity is hard for me, despite studies saying otherwise...
For science to be self-correcting, we need scientists willing to wonder (even in print) if reigning paradigms are accurate.
Philosophers of science have long ago discarded the notion of a purely objective scientist, but not the ability to objectively test a well-formed hypothesis.
The real question is the quality of the science we choose to practice, support, report and rely on to set dietary guidelines. What concerns me is reliance on informational cascades, without a critical examination of the peer-reviewed research supporting such views.
Being in a rural, hilly area, the family cow is a mainstay. Converting oft rocky land into high quality foods. As I teach foods and nutrition, I regularly have students who wonder aloud: If high-fat dairy is such a danger, why are there so many elderly in my community who've daily consumed it? With what Stephen reports, I can more easily say their observations may be more than anecdotal.
Any funnel plots in the paper? I love funnel plots :-)
Hi, Stephan. :)
It's nice to see that you commented about what the study is not saying.
Very few Internet websites are run by scientifically literate people. That is what separates yours.
Urgelt ( whom I owe a lot to) taught me back in 2007 that it is very important to recognize know what a study is NOT saying- among other things.
The low-carbers all think that butter is filling.
It's filling for some of you, but some of us find bland starches or protein to be far more satisfying than fat.
Adding extra butter to my food *would* just make me consume more calories.
@Stephan : Congratulations on your new article
"The implication is that somehow butter causes the hunger/satiety biological mechanisms to break"
J Biol Chem. 2012 Jul 18.
Short-term voluntary overfeeding disrupts brain insulin control of adipose tissue lipolysis.
"However, I have yet to find any evidence for a calorie receptor so I'm not sure how a calorie could be a biological signal for anything"
Cell Metab. 2011 Jan 5;13(1):35-43.
Cellular energy depletion resets whole-body energy by promoting coactivator-mediated dietary fuel absorption.
Interesting references Gys. I would add that there are many "calorie receptors" in the human body of one form or another. These are involved in a variety of negative feedback loops at the cellular and whole-body level. AMPK and mTOR are cellular energy sensors. Leptin, insulin, and other circulating peptides are whole-body energy sensors. There are various gut peptides and nerve signals that operate on a meal-to-meal basis in proportion to energy intake, and one could also call those "calorie receptors" roughly speaking.
Also, satiety mechanisms do not have to break to cause overeating and fat gain. You can make lean people overweight by asking them to eat more, and there are many other factors that can make people eat more independent of hunger and true energy need. These include things like palatability, social environment, and ease of procurement/consumption.
Stephan. I just read this important and well written paper addressing those hugely important questions. Nutrition has for long been undervalued as a scientific subject, not least by doctors. Physicians commonly rely on recommendations given by public health authorities and respective medical associations, usually without analyzing them thoroughly. Most doctors are not very interested in discussing nutritional aspects with their patients. I know this from my own experience. It may be due to lack of knowledge, lack of interest or both. Maybe they think their limited time with their patients will be better spent discussing other issues.
Back to your fantastic paper. It answers a lot of questions, but an analysis like yours indeed raises more questions than answers. These questions become more important due to the fact that your conclusions contradict general public health recommendations on diet and nutrition. So, I have no doubt that you will get enormous feedback.
I am not going to bombard you with questions although I surely would like to. I wonder however what your results are really telling us about the popular low fat dairy products generally promoted by public health authorities. Would it be fair to conclude from your analysis, that selecting low fat dairy products rather than high fat ones, does not reduce the risk of obesity, cardiovascular disease or diabetes. Is there ever a reason, from a health perspective to choose a low fat rather than a high fat dairy product?
What's interesting is that cheese intake does not raise LDL:
"This effect may be due to the high calcium content of cheese, which results in a higher excretion of fecal fat."
Perhaps this could help explain the inconsistent relationship between dairy and CVD? Yogurt may exhibit this property as well, though the effect seems most reliable in cheese.
Also, I wonder if pastured dairy has much lower levels of certain LDL-raising saturated fats like palmitic acid, lauric acid, and myristic acid compared to grain-fed dairy.
I would argue that there is no such thing as a high-fat dairy food. There is normal fat, low fat and no fat. Dairy foods lacking fat are completely abnormal, so I think it’s absurd to think of them as “healthier.”
Since butter is lower in fat than any of the vegetable oils massively consumed by virtually all North Americans, even it cannot be considered unusually or unnaturally high in fat.
I supposed eating lots of butter could make one fat, but what if no other fats are consumed -- or very few? I’m willing to try it.
The paper I quoted above actually throws the calcium/fat excretion hypothesis into question, since fat excretion was the same in both cheese and butter groups. I think it may be something about the fermentation process that is causing the effect.
when you enroll in PreviLEAN Natural Health programs in nutritional consulting, you learn the fundamentals of natural health including essentials to a healthy diet,supplements, herbal medicine, homeopathy, exercise etc.,
I congratulate you on your great success. Nice post. Diet Program.
Stephan, I love whole milk, unsalted butter, and probiotic plain yogurt, and I have been amazed by the wide variety of ailments that are improved by yogurt consumption.
With one exception: I am concerned about research that indicates that lab rats on a yogurt diet became very healthy -- but they also went blind with cataracts.
If I consume lots of dairy, will I become a b;ond healthy person?
I meant "blind healthy person."
I find that I have to eat fiber, protein, "adequate carbs" and fat mixed together to feel truly satiated.
When I used to pour massive amounts of heavy cream onto my coffee that did not turn out to well. It was adding useless, excessive amounts of energy. I am not a low fat guy by any means, but things are less now.
Drinking caloric rich berevages is not a good thing, especially if yu struggle with weight. I have more room than most. But, for a victim of morbid obesity, it would do no favors for you ( or most people).
Overeating disregulates our natural system of weight regulation. So does dieting.
If possible, try to pay close attention to satiation and hunger and eat slowly. Linda Bacons' HAES is pretty good. It has been scientifically tested at least ( and done well for health) - which is something Internet gurus cannot claim for their books.
I'm so relieved after reading this post! I'm vegetarian and I live in Italy where there are many faboulos fat cheese (like the "burrata" for example, which is a kind of shredded mozzarella very, very fat and delicious...). I know, I don't have to exagerate, but that is obvious, isn't it?
please check out my blog http://essere-bene-essere-nepi.blogspot.it/
The practical advice you offer at the end of your post is a bit ambiguous.
At points it sounds like you're merely asserting the obvious “addition vs replacement” claim:
'If you add more calories to your diet while ensuring that your caloric intake and output otherwise remains completely equal, then you will gain fat.'
Elsewhere it sounds like you're making the stronger claim that:
'If you add butter to your food you will (most likely) gain fat.'
The first claim is so obvious no-one could possibly be confused about it. This claim is true whether the food is brocolli or butter. Even the most wild straw man low carb position doesn't claim that adding more fat, independent of anything else happening, makes you thinner.
The second claim is a much more controversial empirical claim, based on whether making foods with a higher proportion of the calories coming from dairy causes you to eat more or less, which isn't at all obvious.
The contrast you draw between people eating a “normal diet” and people adding butter to a normal diet doesn't seem to mean anything, unless you mean to compare a set amount of food to that very same amount of food plus butter. Otherwise, how could there have been people who were eating a normal diet and eating more, rather than less, dairy fat? In one obvious sense, these could be described as people eating a normal diet, with added dairy fat- they are, at least, eating something equivalent to some-one who did just that.
You also describe “replacement” in terms of replacing other fats with dairy fat (rather than simply replacing other calories with dairy fat calories). If this is the limits of your claim though, then your conclusions don't support high fat dairy versus low fat dairy, because there you would not be swapping fats, but adding fats and replacing some other food (perhaps replacing larger quantities of low fat dairy). If you are supporting high fat dairy over low fat dairy, then it seems obvious that this is, in one sense, taking low fat dairy and adding pure dairy fat to it.
Bravo! It's nice to see someone else talk about how science is done. I get sooo tired of people saying that "good science" is when you falsify your hypothesis. It's so silly. This notion that anyone can make up a hypothesis and declare it the "null" is one of the more ridiculous notions out there. I blogged on this topic as well: Testing Scientific Hypotheses v. Statistical Hypothesis Testing
Nice post as always. And as per usual comment thread is solid and a good read as well.
I wonder if it's worth thinking about and/or researching purposeful over-feeding? Individuals in both the powerlifting and bodybuilding communities do this.
I mention this because I'm currently doing it. I'm 5'11", 170 lbs, male, fit, very active. I'm looking to add lean muscle mass.
I had been eating 2300 calories daily.
I have upped it recently to 2700-3100 daily.
At first I thought this would be fun. I love eating and while my goal is not simply eating to eat but to add good lean mass, I thought it'd be fun because more tasty stuff sounded great!
In all honesty after a certain point it is kind of a pain. Not literal pain but it ceases to be as enjoyable as just a little bit of extra cake, etc seems.
For me, up until about the 2700 calorie mark it was still all good and fun. Upwards of that amount and, while everything still literally tastes as good and pleasurable as it did before, the sheer bulk of getting into your stomach results in mild discomfort. Well, just a very full feeling of being full I suppose.
My digestion and elimination is still fine. No gas, no bloating, etc. I don't even get sluggish, which many mention. I'm just talking about feeling literally full of food.
I read a comment that mentioned a "calorie monitor" or something in the body that might sense some amount of calories. While I'm not too sure I do feel that the amounts I'm eating do indeed create quite strong signals to me that I have clearly eaten enough. Again, it still tastes good, the food still seems rewarding I suppose, but I could just as easily not eat the last "meal" as eat it.
This is different than when merely eating a surplus of, say, 300-500 calories. In that smaller range of surplus I still very much have a drive and total pleasure in eating the extra bits. So, perhaps the amount of reward in the surplus calories is a diminishing return?
I do add butter and cream to everything but have lost 12 kgs doing so over the last few months. High fat low carb is the only way I can lose weight and keep it off apparently and all my biometrics are great too...
I see we have another Jane. Jane, please could you use a different name?
There is another angle to the "low fat" dairy that needs to be considered.
And that is, that if you look at the label of most "low fat" or "fat free" dairy products, you will likely see an ingredient that says "modified milk protein/powder/solids" or some such.
This could be a variety of things, but is often spray dried skim milk powder, and this presents a few problems.
Firstly, the milk proteins themselves can be denatured by the drying or "modifying" process, which can make them indigestible to some people.
Secondly, the milk solids contain lactose, which makes the low fat product actually a "high sugar" product (as are many low fat products), and, subsequently, evokes a greater insulin response.
There is nothing wrong with true skim milk (i.e. no added solids) other than the fact it tastes very watery. But any dairy products with modified milk ingredients are just that - modified. Do your body a favour and stay away from them.
Hmm, don't quite know why my previous comment went up as "unknown", but anyway...
I can't leave the "don;t put butter on everything" comment alone.
I think if you put butter on a lot of things, it is likely to make you healthier, as you will eat certain things, or more of them, than might otherwise be the case.
Consider almost any green, cooked veggie (asparagus, kale, spinach, peas, broccoli etc) ALL of them taste better with butter (and salt). Try getting any kid to eat steamed veggies, and then try the same veggies with butter and salt - they are much more palatable.
That is the beauty of butter - it makes a lot bland - but healthy - things taste good, so you will eat them. And you can better absorb the carotenes and other fat soluble nutrients in them.
If the teenager in question actually eats their broccoli with butter, they are more likely to feel satisfied, and less likely to be eating a burger/fries/soft drink later. And will be healthier for it.
Also, of course, eating butter can improve your brain function too, as demonstrated by Seth Roberts
Great post & well done on the publication. This is certainly an interesting area of nutrition. As a dietitian I've always thought there's room in a healthy diet for regular dairy so long as eaten in moderation. I wonder if a similar hypothesis can be made with full sugar products eg jams or even sweetened yoghurts, compared to artificially sweetened product?
Stephan, You said
"Typical dietary advice includes the recommendation to eat low-fat or skim dairy products. This is based on the hypothesis that avoiding the (mostly saturated) fat in dairy will reduce the risk of obesity, metabolic problems, and cardiovascular disease. This idea is logical, but not every idea that is logical is correct when tested scientifically, particularly when it pertains to a complex natural food."
I would be interested to know why you consider the idea that avoiding the (mostly saturated) fat in dairy will reduce the risk of obesity, metabolic problems and cardiovascular disease is logical. The notion seems totally illogical to me for many reasons.
For one thing, in the western diet, dairy fat provides more than 70% of vitamin K2 with most of the remaining K2 coming from other animal fats. The Rotterdam study found that those in the lowest tertile of vitamin K2 intake more than double the risk of CHD mortality and severe aortic calcification. The Heidelberg study found that those in the lowest quartile of K2 intake had a 250% increased risk of advanced prostate cancer and of lung cancer mortality. Followers of USDA dietary recommendations to avoid animal fats will put people in the lowest quartile of K2 intake for sure.
Sardinia has the longest life expectancy on the planet for men.
I have seen what the Sardinians eat and it is full fat dairy and fatty pork among many other things. They had a feast that was not low on fat by any standard.
There is a video of it on YouTube as well.
Now, of course it could be great genes are by far the largest reason they live the longest of all men. But, apparently, full fat dairy has not harmed them in any noticeable way to my knowledge.
That's why I buy Cabot butter instead of more expensive grass fed. It's a nice yellow color so I know it's from mostly grass fed cows in VT. And their full fat Greek yogurt is so good.
He's saying the obvious one, because to some people that's NOT obvious and he wants to be thorough. The "replacement" idea doesn't necessarily invalidate the fact that high fat dairy > low fat dairy because at the very least the high fat dairy eaters did better than the low fat dairy eaters because of a decrease in consumption of these "other fats". Or to go further, if the caloric balancing happens only with milk (and rest of diet stays the same), then this paper would show that one change itself was beneficial.
@Axel F Sigurdsson If you're doing carb refeeds (maybe on an intermittent fasting protocol), or want to make a fast-absorption protein shake with milk (though I think the "anabolic recovery window" is actually quite large): http://suppversity.blogspot.com/2011/10/intermittent-thoughts-on-intermittent_23.html
@Paul N - I agree with you, and I think Stephan does as well. He can correct me if I'm wrong, but the point was more "if you're eating your normal diet + butter (which isn't very filling), you will increase calories and will gain, not lose, weight. And then you'll complain to me that my paper lied to you". Butter addition is great if you're eating low carb and thus highly satiating foods, so the addition of butterfat will simply augment that process. On a standard american diet or high-carb meal, where one often eats a large volume of food, butter on everything will turn your food into butter with a side of food.
@Jack C - I think he does not believe that. Maybe he's being nice to them for thinking up a good idea (less satfat in dairy = less disease), but then he states that it doesn't necessarily pan out like that when you actually.. you know.. do the science. Maybe he feels sorry for them (I know I do).
@Kate - the one that's like all fat and protein? Mix it up with blueberries and sliced bananas in layered fashion, and you've got yourself a parfait. In the words of Donkey from Shrek, "Parfait's gotta be the most delicious thing on the whole damn planet!"
OK - science discussed in detail is why i come - however, for what is worth - N=1
we eat exclusively organic and grass fed dairy - a good amount of it. and we DO add lots of butter to our servings but the servings are modest. no grains, very little sugar. my partner is 120 lbs after 4 years of this eating style and before that - always 140# to 150# al her life. we 3 (4.5 year old) are very healthy and lean, muscles seem to be toned with very little specific exercise.
i love stephen's dedication to science - but at some point you have to take all the oft-conflicting science blabber and make a diet for oneself and high fat dairy (given that we are of northern euro decent) is hugely successful for us.
Congratulations! Stephan, why do you write: "This is based on the hypothesis that avoiding the (mostly saturated) fat in dairy will reduce the risk of obesity, metabolic problems, and cardiovascular disease. This idea is logical, ..."
I find it quite illogical that eating a a few dozen grams of animal fat, will cause any problems in a body that contains many kilograms of animal fat. The only reason it would be logical would be if it would cause digestive issues. In all other cases, I would reverse your statement and say that one should only assume fat to be bad if the is scientific evidence.
Funnily, I just started reducing my pot belly by stopping eating full-fat yoghurt every breakfast, which I found surprising as I never noticed any digestive problems. However, that may also be due to eating more eggs as replacement.
Stephan, I just want you to know that you have made the prime-time news in New Zealand with this study.
Given that we are a largely dairy economy and stuck in a high-carb health cycle phase, this was a big deal.
You paper - with your names clearly legible - was shown on TV1's Close Up (investigative journalism and human interest after the 6pm news) and previewed in the main bulletin.
Not only congratulations, but the thanks of all New Zealanders are in order.
Here is the link:
It is the second item after alcohol on 13th August.
So you also get to see what pissheads Kiwis are, and that our government is owned by the breweries. Shame!
May I ask, what do you think of this study: http://www.youtube.com/watch?v=-DQ2itumKu8
It found that inheriting polymorphisms which cause lower LDL dramatically lowered the incidence of heart disease.
I don't know if my original question posted due to Blogger's identity process but did this review take into account what type of casein--A1 vs A2--those studied were eating?
Post a Comment