Thursday, August 19, 2010

Tropical Plant Fats: Coconut Oil, Part II

Heart Disease: Animal Studies

Although humans aren't rats, animal studies are useful because they can be tightly controlled and experiments can last for a significant portion of an animal's lifespan. It's essentially impossible to do a tightly controlled 20-year feeding study in humans.

The first paper I'd like to discuss come from the lab of Dr. Thankappan Rajamohan at the university of Kerala (1). Investigators fed three groups of rats different diets:
  1. Sunflower oil plus added cholesterol
  2. Copra oil, a coconut oil pressed from dried coconuts, plus added cholesterol
  3. Freshly pressed virgin coconut oil, plus added cholesterol
Diets 1 and 2 resulted in similar lipids, while diet 3 resulted in lower LDL and higher HDL. A second study also showed that diet 3 resulted in lower oxidized LDL, a dominant heart disease risk factor (2). Overall, these papers showed that freshly pressed virgin coconut oil, with its full complement of "minor constituents"*, partially protects rats against the harmful effects of cholesterol overfeeding. These are the only papers I could find on the cardiovascular effects of unrefined coconut oil in animals!

Although unrefined coconut oil appears to be superior, even refined coconut oil isn't as bad as it's made out to be. For example, compared to refined olive oil, refined coconut oil protects against atherosclerosis (hardening and thickening of the arteries) in a mouse model of coronary heart disease (LDL receptor knockout). In the same paper, coconut oil caused more atherosclerosis in a different mouse model (ApoE knockout) (3). So the vascular effects of coconut oil depend in part on the animals' genetic background.

In general, I've found that the data are extremely variable from one study to the next, with no consistent trend showing refined coconut oil to be protective or harmful relative to refined monounsaturated fats (like olive oil) (4). In some cases, polyunsaturated oils cause less atherosclerosis than coconut oil in the context of an extreme high-cholesterol diet because they sometimes lead to blood lipid levels that are up to 50% lower. However, even this isn't consistent across experiments. Keep in mind that atherosclerosis is only one factor in heart attack risk.

What happens if you feed coconut oil to animals without adding cholesterol, and without giving them genetic mutations that promote atherosclerosis? Again, the data are contradictory. In rabbits, one investigator showed that serum cholesterol increases transiently, returning to baseline after about 6 months, and atherosclerosis does not ensue (5). A different investigator showed that coconut oil feeding results in lower blood lipid oxidation than sunflower oil (6). Yet a study from the 1980s showed that in the context of a terrible diet composition (40% sugar, isolated casein, fat, vitamins and minerals), refined coconut oil causes elevated blood lipids and atherosclerosis (7). This is almost certainly because overall diet quality influences the response to dietary fats in rabbits, as it does in other mammals.

Heart Disease: Human Studies

It's one of the great tragedies of modern biomedical research that most studies focus on nutrients rather than foods. This phenomenon is called "nutritionism". Consequently, most of the studies on coconut oil used a refined version, because the investigators were most interested in the effect of specific fatty acids. The vitamins, polyphenols and other minor constituents of unrefined oils are eliminated because they are known to alter the biological effects of the fats themselves. Unfortunately, any findings that result from these experiments apply only to refined fats. This is the fallacy of the "X fatty acid does this and that" type statements-- they ignore the biological complexity of whole foods. They would probably be correct if you were drinking purified fatty acids from a beaker.

Generally, the short-term feeding studies using refined coconut oil show that it increases both LDL ("bad cholesterol") and HDL ("good cholesterol"), although there is so much variability between studies that it makes firm conclusions difficult to draw (8, 9). As I've written in the past, the ability of saturated fats to elevate LDL appears to be temporary; both human and certain animal studies show that it disappears on timescales of one year or longer (10, 11). That hasn't been shown specifically for coconut oil that I'm aware of, but it could be one of the reasons why traditional cultures eating high-coconut diets don't have elevated serum cholesterol.

Another marker of cardiovascular disease risk is lipoprotein (a), abbreviated Lp(a). This lipoprotein is a carrier for oxidized lipids in the blood, and it correlates with a higher risk of heart attack. Refined coconut oil appears to lower Lp(a), while refined sunflower oil increases it (12).

Unfortunately, I haven't been able to find any particularly informative studies on unrefined coconut oil in humans. The closest I found was a study from Brazil showing that coconut oil reduced abdominal obesity better than soybean oil in conjunction with a low-calorie diet, without increasing LDL (13). It would be nice to have more evidence in humans confirming what has been shown in rats that there's a big difference between unrefined and refined coconut oil.

Coconut Oil and Body Fat

In addition to the study mentioned above, a number of experiments in animals have shown that "medium-chain triglycerides", the predominant type of fat in coconut oil, lead to a lower body fat percentage than most other fats (14). These findings have been replicated numerous times in humans, although the results have not always been consistent (15). It's interesting to me that these very same medium-chain saturated fats that are being researched as a fat loss tool are also considered by mainstream diet-heart researchers to be among the most deadly fatty acids.

Coconut Oil and Cancer

Refined coconut oil produces less cancer than seed oils in experimental animals, probably because it's much lower in omega-6 polyunsaturated fat (16, 17). I haven't seen any data in humans.

The Bottom Line

There's very little known about the effect of unrefined coconut oil on animal and human health, however what is published appears to be positive, and is broadly consistent with the health of traditional cultures eating unrefined coconut foods. The data on refined coconut oil are conflicting and frustrating to sort through. The effects of refined coconut oil seem to depend highly on dietary context and genetic background. In my opinion, virgin coconut oil can be part of a healthy diet, and may even have health benefits in some contexts.

* Substances other than the fat itself, e.g. vitamin E and polyphenols. These are removed during oil refining.


JLL said...

Heh, I'm just writing a post on palm oil and cholesterol for my blog (should be up later today). Like with coconut oil, the results from animal studies are very mixed. Sometimes LDL goes up, sometimes it goes down, sometimes nothing happens. Like you said, the rest of the diet matters as well. Good post!


David Moss said...

Interesting summary. So in terms of the most common western sources of coconut (at least in the UK) creamed coconut, dessicated coconut etc, how would these fit into the refined/unrefined categories? Although not 'virgin' I'd have hoped that these sources of coconut meat and oil would've retained a reasonable number nutrients.

New York said...
This comment has been removed by the author.
New York said...

Overall, these papers showed that freshly pressed virgin coconut oil, with its full complement of "minor constituents"*, partially protects rats against the harmful effects of cholesterol overfeeding.

I was under the impression that the ostensible link between cholesterol consumption and heart disease had been disproven (except in rabbits). So that there would be no "harmful effects" of cholesterol to protect against.

trix said...

We've been buying Tropical Traditions Gold Label Virgin Coconut Oil, and their Expeller-Pressed version. The expeller-pressed coconut oil is naturally refined without solvents. Is that considered the same as the 'refined' you mentioned? (It no longer has the 'coconut' flavor.) We've also used MCT oil which doesn't solidify.

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

One issue people need to be more aware of.

Whatever the benefits of coconut oil, adding it to your existing diet rather than using it to replace other fats will stop weight loss and/or lead to weight gain.

This outcome has been reported by dozens (if not hundreds) of people posting to low carb diet boards.

John said...

Hi Stephan,

This is a great overview. Although there are some contradictory conclusions, I think the strong majority of studies show coconut oil as being positive rather than negative or neutral. It's nice that there are so many across several different types of animals as well. A good source for MCT/coconut oil studies is

This is interesting:

By the way, references 14 and 15 were the same web site for me.

John said...

Here is the full text of the link I posted. There are good follow-ups references too.

David Csonka said...

You don't drink purified fatty acids from a beaker?

Get with the program man! :D

Anna said...

Favorite dessert this summer:

Small bowl of fruit (cantaloupe melon, banana, orange slices, blueberries, or plums, etc.) drizzled with coconut milk and grated fresh coconut - mmmm!

I've using my rotary coconut grater a lot this summer - freshly grated coconut is a really nice addition to lots of foods, including savory Thai-inspired curries and homemade coconut milk ice cream. After I grate up a whole coconut, I freeze the grated coconut "meat" in small quantities for convenience (covered ice cube trays are perfect).

Richard A. said...

Too many coconut oil studies are omega 3 deficient -- and if the the coconut oil has been hydrogenated omega 6 deficient as well.

benn686 said...

Maybe I just missed it, but is refined supposed to be the Copra oil?

And Extra virgin coconut oil is preferable to Copra?

mike250 said...

I have yet to find a way to incorporate coconut oil in my diet. Cooking with it is piss-poor as it ruins the taste of the food its cooked with. I am thinking of putting it in my oatmeal or in a smoothie. What do you guys think of that?

P.S coconut oil isn't very good for fat loss. your body ends up using it instead of burning your own fat.

Jee said...

Has anyone read Mary Enig's, Eat Fat, Lose Fat? Is it a worthwhile read?

Hi Moataz,
Thai food works well with coconut oil. Also, if you eat tubers, try baking sweet potato slathered in coconut oil with a little sea salt and pepper -- add some chopped pumpkin seeds if you like. Tasty!

John said...


There are several studies showing fat loss benefits with coconut oil consumption--just do a Pubmed search.


It's okay, but if you're familiar with WAPF articles, I wouldn't recommend it.

Anna said...


I have Fallon & Enig's Eat Fat, Lose Fat book. I don't recall that much of the non-recipe text is very different from other WAPF info or other coconut-boosting books (Bruce Fife, etc.), but I have made a number of the recipes and they are generally good. There are a couple of chicken recipes I particularly liked and have adapted, as well as an easy Thai-inspired coconut milk soup.

I think you are right about suggesting Thai and other Southeast Asian-inspired recipes for coconut oil use. Coconut oil often just isn't going to taste right in many American and European origin recipes. For instance, I'm not wild about cooking my over-easy eggs in coconut oil, yet I love coconut milk baked egg custard. I also cook non-grain pancakes for my son in coconut oil or a blend of coconut oil and ghee/butter (mashed half banana, generous tablespoon almond butter, beaten egg, optional tablespoon Dutch process cocoa powder). With meat and poultry recipes, "Eastern" spices such as ginger, cardamon, lemongrass, tumeric/curry blends and so on are natural partners to the taste of coconut oil and coconut milk.

Anonymous said...

coconut oil reminds me of acai berries as the next big thing...everyone promotes their weightloss abilities and oh so healthfulness... kind of like the vitamin D which i only think people should be supplementing if they dont go outside

i dont use coconut oil for consumption just on my skin periodically, or vita D personally. i read excess vita D is stored as fat(maybe an evolutionary adaption??)

Jee said...

Thanks for the info John and Anna!

I use coconut oil, along with ghee and lard, to replace vegetable oil for cooking. If it contributes to fat loss, then that's ok too. But from what I've been learning on this awesome blog, there's compelling evidence to suggest our choice of cooking oils/fats have a great impact on our whole health, whereas our choice of antioxidants... not so much. I think half the battle is knowing which ones to fight:)

bee said...

I hail from Kerala in southern India where coconut forms a huge part of the calorific intake. However, not much coconut oil is consumed. Most of the coconut used is of the fresh, grated variety. Some is in the form of fresh, green coconut (which is low in fat), and some in the form of coconut milk.

All that fiber probably has something to with the fact that people from this part of India have traditionally been lean and healthy.

JLL said...

Here's that post on red vs refined palm oil and cholesterol I mentioned earlier. Also a couple of comparison studies between palm oil and coconut oil referenced in there.

At least where I live, red palm oil is a lot cheaper than virgin coconut oil (although I use both).


Brandon said...

Thank you so much Stephan for providing these sources for us! It really helps out those who need to see the actual literature, rather than go on what other people tell us. You're awesome in my book.

Anonymous said...

very good study about Heart Disease.

David said...

"Keep in mind that atherosclerosis is only one factor in heart attack risk."

wow, you are really stretching Stephan! Atherosclerosis is more than "only one factor" in heart attack risk - in all but rare cases it is the sine qua non of heart attack risk.

Unknown said...

Good study! Though there is lots of benefits from coconut oil but its taste is not good at all. I’ve tasted a south Indian food once made by coconut oil. It was awful.

Healthcare Recruitment

Stephan Guyenet said...

Hi David,

I beg to differ. Other important factors include plaque stability, hemostasis, susceptibility to fibrillation, myocardial susceptibility to hypoxic damage, and certainly others. In aggregate, these factors are probably more influential of heart attack risk than atherosclerosis itself. Atherosclerosis has been emphasized at the expense of these other factors, probably because it's a visible process that is also related to blood lipids.

The studies of non-industrial cultures that don't suffer from heart attacks make it clear that atherosclerosis per se is not sufficient for increased heart attack risk (to a point). Several autopsy studies in Africa indicated that individuals with atherosclerosis similar to Western adults (and age-matched) weren't having infarctions. These were very large studies and about as close to definitive as we're going to get. The authors speculated that Africans have a much lower clotting tendency and this underlied their lack of heart attacks. In support of that, they cited studies showing much lower rates of embolism in Africa.

Unknown said...

Stephan, this is a great and informative post. I have a question going back to you interview with Jimmy Moore. You cited the benefits of oligofructose and chicory root. I've been looking at this subject since but I can't figure out which products/supplements derived from chicory root have sufficient oligofructose content. Can you please advise? Would dried chicory root (brewed as coffee) sufficient? What specific products/sources/amounts would you recommend?

Best regards,

Chris M

radconsultancy said...

This is an excellent post on health.
It is very informative and interesting. I really appreciate your blog.

IF someone want to know more about Reactive attachment disorder. You can contact us.
And Visit our blog :

Roger L. Cauvin said...

Stefan, I was wondering if you had analyzed this meta-study that concluded switching from saturated fats to polyunsaturated fats from vegetable oils lowered the incidence of heart disease:

They conclude:

"These findings provide evidence that consuming PUFA in place of SFA reduces CHD events in RCTs. This suggests that rather than trying to lower PUFA consumption, a shift toward greater population PUFA consumption in place of SFA would significantly reduce rates of CHD."

Stephan Guyenet said...

Hi Roger,

Yes, I reviewed it here:

I'm dismayed that a paper of such poor quality made it through the peer review process. They blatantly ignored their own selection criteria.

pooti said...

Chris M - I'm thinking that maybe he's referring to inulin which can be made from chicory?

Roger L. Cauvin said...

Thanks, Stephan, for pointing me to your review of the meta-study. I was hoping you had written one!

John said...


Comparing coconut oil to beef tallow, plasma insulin and glucose are lower in the CO group (calves) but the CO group had an 18-fold increase in hepatic triglycerides. Is this necessarily a negative?

In rats, when comparing fish oil to coconut oil, the hepatic trigs go down w/ more coconut oil

Now this is what I would expect. Care to offer your idea? Is it simply a matter of different animals?

Stephan Guyenet said...

Hi John,

I'm really not sure. It could be the different species, or the dietary context, or both.

Stephan Guyenet said...

Thanks JBG.

شركة نماء لالحاق العمالة بالخارج said...

nice work
for more details about me visit the following blog and web site

bovinedefenestration said...

I figured that I'd chime in, what with all the people claiming that coconut oil impedes fat loss. Personally, I find I can successfully lose weight only when I include a good amount of coconut oil. I have no idea why, but I feel much better, more energetic, and the weight comes off faster with fewer side-effects.

I just eat plain spoonfuls of the stuff. Virgin coconut oil isn't so bad that way.

Steve_Roberts said...

"humans aren't rats"

Exactly. And rats aren't humans - specifically they are adapted to eat grains, and we are not. We need to very cautious in drawing conclusions about human nutrition from studies of rodent nutrition.

benn686 said...

Is there anything particularly healthful or nutritious about Coconut sugar?

I'm looking to use Cerylon Cinnamon, or Coconut sugar as an alternative to the more mainstream alternatives... stevia (Sweet Leaf, Puritan's Pride), sugar alcohols (Xylitol/Erythritol), luo han guo extracts (Sweet Fiber), Orange Peel / Chicory Root (JustLikeSugarInc) or sucrose/fructose/ (whey low)...

Does coconut sugar have any of the good properties of coconut oil?

David Pier said...

For anyone who thinks it is hard to use coconut oil, use coconut milk! Coconut milk should be at least as good as coconut oil in terms of "minor constituents". You can make a delicious smoothie with equal parts of frozen fruit and refrigerated coconut milk. A little extra sweetener might be your liking, but it isn't necessary. Strawberry/pineapple, peach/cherry/pineapple, and blueberry/pomegranate are our favorites.


Nice post.

Have you found any new studies on unrefined coconut oil and humans?