How much omega-3 is too much?
A. There is no UL (tolerable upper limit) for omega-3 or omega-6 fatty acids (PUFAs). The USDA feels it has "insufficient evidence" to set one. Instead, you might want to consider how much total fat you're going to eat and how you want to divide that up. For example, consuming large amounts of PUFAs might either cause you to miss out on the benefits of other fats, such as monounsaturated fats, or lead you to eat too much total fat.
Many experts recommend that PUFAs should make up 5-10% of your dietary calories and that seems reasonable to me. The rest of your fat calories would come from monounsaturated fats like those found in olives and avocados and/or saturated fats from meat, dairy, and tropical oils. Trans fats from hydrogenated oils and fried foods should ideally be kept to an absolute minimum.
In your example, you'd be eating 50g of PUFAs, for a total of 450 calories. Unless you're eating 4500 calories a day, that might be a little out of balance. Without knowing the particulars of your situation, sounds as if you might want to think about ratcheting your intake of BOTH kinds of PUFAs back a bit.
Building your dietary fat "profile"
Here's how a healthy distribution of fats might look for a typical person eating 2500 calories a day:
30% of calories as fat = 83 g
Of that:
15% of calories from monounsaturated fat = 40g
7% of calories from saturated fat = 20g
7% of calories from polyunsaturated fats = 20g
To maintain a 4:1 ratio of omega-6 to omega-3, that would be 16g omega of omega 6 and 4g of omega 3.
Fish or flax?
As far as the dangers of taking too much omega-3, it depends whether you're talking about 10g of flax oil, which provides alpha-linolenic acid (ALA), or 10g of fish oil, which contains the much more biologically active forms of omega 3 (EPA and DHA).
Ten grams of fish oil could potentially create some issues, such as interfering with blood-coagulation and immune function (not to mention digestive distress). Whether or not these might be concerns for you is a question for a licensed nutrition professional who knows the details of your situation.
 older posts |
home | newer posts  |
| Do overweight people eat less than thin people (or do they just think they do)? | | Nutrition Data Blog | | Purslane as a source of omega-3: a reality check |
Posted by: greg | Oct 19, 2009 2:32:45 AM
@O
OK, we can drop it here as I don't have the energy to read another amateur site. Lets just agree to disagree.
As for the study and paper, I didn't "derive any conclusions," I just read them. To avoid beating a dead horse, I'll drop it here but I would encourage you to read the paper carefully for yourself and you will see the authors draw the exact opposite conclusion than you think.
I'd urge you to keep reading but to find a few more sources to get additional perspectives. The WHO report I linked to below is not very user-friendly but it is packed with information if you are willing to wade through it all.
I wish you good luck and good health :)
Posted by: O | Oct 19, 2009 12:19:03 AM
The corrected link to "Fat and cholesterol are good for you" by Uffe Ravnskov:
Posted by: O | Oct 19, 2009 12:16:42 AM
@Greg
I simply repeated and summarized the discussion of the cited study from the two additional links. They explained their analysis quite well, and it was logical. That you would derive a different conclusion from the result does not surprise me.
Uffe Ravnskov has spent the last 20 years debunking the lipid hypothesis by reading all the papers that supposedly support it, and debunk them in his own papers. In a lot of these papers the authors simply derive the wrong conclusions! Here is a nice new article by him:
http://www.spacedoc.net/saturated_fat_is_good_for_you_1
He also has a great new book - "Fat and cholesterol are good for you:"
Posted by: greg | Oct 18, 2009 6:08:35 PM
@O
First, I think you go too far trying to put the feed-the-animal link on the same level as peer-reviewed epidemiological studies and then brushing it all off as “correlation is not causation.” The amateur analysis at feed-the-animal has not even identified a real correlation because they used a statistically meaningless technique (regressing a variable against a rank) and have not attempted to adjust for even the most obvious confounding variables (health care, smoking etc.) I doubt it is even possible to get a statistically significant result analyzing data aggregated at the level of national consumption as they are trying to do.
On the other hand, epidemiological studies done correctly are not at the mercy of “the bias of the authors.” With a large enough data set and a properly controlling for confounding variables they can be quite powerful. That said, they should be used primarily to form hypothesis for further study and not taken alone. Unfortunately, in many cases, big clinical trials would be impractical so we have to live with the best studies we can get.
The Mozaffarian study and “American Paradox” article you linked to are more interesting and properly raise some of the nuance involved. But these articles DO NOT support your summary and your conclusion, “in lean, insulin-sensitive subjects SFA [do] even more good,” is the OPPOSITE of the author’s of both papers.
The study found SFAs were associated with relatively slower progression IN A LOW FAT DIET and may (P=0.09) be associated with LOW TOTAL FAT consumption. That is an interesting, confusing result but does not support the blanket suggestion that more SFAs are good for you. In particular, extrapolation of the results is difficult since this is a very, small observational study of ill, overweight, post-menopausal women with metabolic syndrome – certainly not representative of the population at large!
The article suggests that overweight, insulin-insensitive seem to be protected from the harmful effects of SFAs relative to lean, insulin-sensitive subjects – i.e. SFAs are harmful to the lean but not this narrow group. The article doesn’t draw any conclusions – they describe the Mozaffarian study as hypothesis-generating only – but draw attention that there is more to cardiovascular health than SFA intake alone.
To bring this discussion full circle, Monica’s advice/method is right in the center of current scientific understanding and it makes sense for most people. To get a broad sense of where the science currently is, Monica recently posted a big review by the World Health Organization on dietary fats and health and the article “Fats and Fatty Acid Requirements for Adults” has recommendations in line with her method.
http://blog.nutritiondata.com/ndblog/2009/09/how-much-and-what-kind-of-fat-should-you-eat.html
In addition, a recent meta-analysis finding the healthful effects of PUFAs relative to SFAs was impactful to the participants in WHO review.
Reading the comments on this post, it is apparently a point of conviction for some that PUFAs are dangerous and not SFAs. I wish such adherents well, but today there is not anywhere near enough evidence to support such a claim. And in fact, what evidence there is exactly the opposite: replacing SFAs with PUFAs (up to 10% of calories) improves your cardiovascular health. We won’t know the whole truth (for example, the biological factors or dietary interactions that impact outcomes) for some time (if ever) and in the meantime we have to eat. We can’t do better than basing our choices on the best science we have today.
Posted by: O | Oct 18, 2009 1:58:30 AM
@Greg
A discussion of the cited paper is also here:
http://www.ajcn.org/cgi/content/full/80/5/1102#FN2
The authors cite evidence that in insulin-resistant subjects, like the women in the study, the effects of SFA were blunted, which would mean that in lean, insulin-sensitive subjects SFA does even more good.
Posted by: O | Oct 18, 2009 1:49:16 AM
@Greg
Fair enough, the data does not prove anything. Correlation is not causation. But then, a lot of "serious science" supposedly validating the lipid hypothesis are correlation studies, and thus do not prove anything beyond the bias of the authors.
Let's discuss peer reviewed papers then, starting with this one, from the American Journal of Clinical Nutrition:
A quick write-up on the conclusions from the paper is given in this blog post:
http://high-fat-nutrition.blogspot.com/search/label/Arteriosclerosis%20and%20saturated%20fat
In summary:
1. more saturated fat leads to less progression of heart disease
2. more carbs leads to more progression of heart disease
3. PUFAs lead to more progression when replacing other fats (!), except if they replace carbs
4. mono fat had no association
Posted by: greg | Oct 17, 2009 9:34:16 PM
@O
We seem to be changing the topic from “what is current scientific understanding” to “how did that web site get the answer it does?” That is fine but let’s be clear that a private website is not indicative of the state of our understanding.
As for the link, these amateur web sites are not serious science. With time on your hands and a spreadsheet, you can make a graph to show anything. A quick look at the link you posted reveals some major problems:
1) It is mathematically non-sense. You can’t regress a variable against a rank order as they have in this website; you regress a variable against a variable. The graph might look pretty but it is meaningless. (Though I can’t be sure what they did since they don’t actually disclose the formula they used or any significance tests… I’m just surmising from the text and the graph.)
2) Second, a fundamental challenge in regression analysis is under-specification; if there is a variable that influences the outcome but you don’t include it in your regression, your results are worthless. In the free-the-animal regression, notice how it is all poor countries on the left hand side of the graph (Bosnia, Georgia, Tajikstan) and all the rich countries on the right hand side (Netherlands, Switzerland, France.) To try to get a meaningful result, one would have to do a lot more work, including variables like per capita income and health care expenditure. At a minimum, one would also need to adjust for other known risk factors (smoking comes to mind) or else the exercise is a waste of time. This "analysis" doesn't do any of the basics required to get a legitimate result.
To get a feel for the state of real science, stick to peer-reviewed studies (or else be lazy like me and let Monica do it for you!) There are a lot of hobbyists out there that don’t know what they are doing putting nonsense on the web. But trust me, your friend at free-the-animal hasn’t discovered something that the EU, OECD and WHO have missed.
Posted by: O | Oct 17, 2009 6:33:42 PM
@Greg: A CT 64-slice calcium heart scan is a "personal anecdote" ???
It is the only diagnostic test that can conclusively determine whether a person has or doesn't have heart disease.
Get this book:
http://www.amazon.com/reader/0595316646?_encoding=UTF8&ref_=sib_dp_pt#reader-link
Also, how would you explain that the countries with the least amount of heart disease have the highest intake of saturated fat?
http://freetheanimal.com/2009/09/saturated-fat-intake-vs-heart-disease-stroke.html
Posted by: greg | Oct 17, 2009 1:40:48 AM
@PJ, haha... I think we have misunderstanding. My "PS" was really a post script (refering to O's question to me) and not directed at you.
Welcome to the board. I hope you find this an excellent resource.
Posted by: PJ | Oct 17, 2009 12:28:31 AM
Hmmn, meant that to be a comment on the article, not on the comment above it, sorry. ;-)
Posted by: greg | Oct 17, 2009 12:27:24 AM
@O
I think you have a half truth there. Certainly scientific understanding has evolved from the overly simplistic conception that "healthy eating" means "avoiding dietary fat."
That said, it is certainly not the case that there is "massive scientific evidence that SFA do not clog your arteries." On the contrary, even Taubes (to make him the champion of your position) accepts that each individual step in the causal chain of the diet-heart hypothesis has been convincingly demonstrated since the 1970s.
What we understand now, however, is that this is not the whole story. FAs mediate serum cholesterol levels and this influences CHD risks but only weakly and there are clearly more factors at play. Among them, oxidative stress from PUFAs is one of many hypothesized additional risk factors.
So, in the end, the best evidence-based advice we can give today is:
1) limit SFAs - they increase CHD risk;
2) replace SFAs (or carbs for that matter) with PUFAs - they decrease CHD risk; but
3) keep PUFAs below 10% of calories - to constrain the risk of LDL oxidation and coincident increased CHD risk.
PS: I don't mean to be unresponsive to your "heart scan" query, but I'd just assume avoid devolving to personal anecdotes since I think that trivializes the discussion and they are neither interesting nor informative.
Posted by: PJ | Oct 17, 2009 12:27:11 AM
When I radically upped my fat intake (animal fats) and dropped my carb intake, I ate more-to-fullness than I ever had in my life and have lost about 140 pounds. That 30% fat content in a diet recommended by the author here is crap. That's going to force the majority of the diet to be carbohydrates -- hey, have a little food with your sugar why don't you. Grow up already, if you're going to be preaching to the world via internet, start reading real science with half a brain, and not the bogus translations that abstracts and 'media' spoonfeed to the ignorant public, mostly based on money from companies with something to sell by convincing us fiber is good because cheerios aren't as bad as froot loops, or that fat is bad because someone's ability to kill a rat with soybean oil won't be noticed if they leave that detail out and pretend that it was all those steaks that killed off the rats. All the same public advice for the last 30 years and the problems just get worse; there's an obvious answer here somewhere.
Posted by: O | Oct 16, 2009 9:33:08 PM
@Greg: What you are saying is not based on the current scientific understanding, but the dated lipid hypothesis which has never been proven.
Today, there is a massive scientific evidence that SFA do not clog your arteries. What causes heart disease is inflammation caused by overload of PUFA fats and carbohydrates.
"Extra sugar molecules attach to a variety of proteins that in turn injure the blood
vessel wall. This repeated injury to the blood vessel wall sets off inflammation. When you
spike your blood sugar level several times a day, every day, it is exactly like taking
sandpaper to the inside of your delicate blood vessels.
by consuming excessive Omega-6, the cell membrane produces
chemicals called cytokines that directly cause inflammation.
" - Dwight Lundell, MD
I eat 70g of saturated fat each day on my primal (low carb) diet. Like many others that eat similarly, I have a CT heart score of 0 (no plaque i.e. no heart disease).
What is your heart scan score Greg?
Posted by: greg | Oct 16, 2009 7:55:07 PM
Monica's advice/method seems to be right inline with current scientific understanding (as limited as that may be): keep SFAs as low as possible, substituting with PUFA when you can, and keeping both below 10% of calories for optimal heart health.
@Dave
I'm not sure we understand the efficiency of conversion of ALA to EPA and DHA.
Some studies do show low levels of converstion of ALA to EPA (even lower than 10%.) However, more studies seem to show increases (though not linear) of EPA with dietary ALA but not DHA. Still others find the background diet to strongly influences the synthesis of DHA: long term (several months) of high ALA diet has been shown to increase DHA formation. Also women are thought to be more efficient at the formation of DHA than men. I think all of this needs to be confirmed with more study...
Regardless, to answer your question, I understand that ingested ALA that is not metabolized is largely thought to be catabolized into acetyl CoA which is reused or else further catabolized into CO2.
Posted by: O | Oct 16, 2009 7:33:58 PM
I also disagree with Monica, and agree with the comments by David Moss and Pam Schoenfeld. There is no need to eat so many PUFAs, especially given the danger of them being oxidized. Also, PUFA vegetable oils like corn and soybean oil are produced by chemical extraction using solvents. Why would you ever want to put anything like that in your body?
In practice, it is easy to minimize omega-6, and get enough omega-3. Simply avoid all vegetable oils, which means avoiding processed foods. Use coconut oil and butter for frying, sauteing, and cooking in general, and olive oil for salads. Eat free range meats, especially if they are fatty. Eat fatty fish like salmon a few times a week.
It is time to retire the tiresome fat-phobia. There is no need to set an upper number on total fat. When I moved from a zone like diet to my current primal diet, my fat intake went from 30% to about 55% of calories. About a third of that daily fat is saturated. Thanks to the increase in saturated fat, my HDL went from 49 to 88, and my testosterone doubled from 294 to 606. My bodyfat percentage went down a few percentage points as well.
Posted by: Pam Schoenfeld, RD | Oct 16, 2009 5:07:14 PM
I concur with most of these comments, not those of the poorly informed author. There is no benefit in eating such high levels of polyunsaturates, and indeed there is great potential for harm from lipid peroxidation.
This article is misleading on a number of levels, in particular that ALA has biological activity in of itself before it is converted to DHA and EPA (it does not). Plant sources of omega-3s can only increase oxidative stress in the body when consumed in large amounts, especially when you add in the excessive amounts of omega-6s that most people already eat. Humans need DHA in very small amounts, balanced with ARA. (see www.cholesterol-and-health.com for an excellent overview of the real need for omega-3 and omega-6 fats in the human diet by Christopher Masterjohn).
Monounsaturated and saturated fats, are a good source of calories and nutrients when they come food sources like cold-pressed olive oil and pastured animals, as well as coconuts, etc. They help keep insulin levels in check as compared to comprable calories consumed from carbohydrates. Plus cold-pressed olive oil unfiltered is full of polyphenols, and pastured animal fat is a good source of CLA, among other nutrients.
Posted by: Sean Hannigan | Oct 15, 2009 10:26:34 PM
It seems that the same question continues to arise in this debate....flax seed or fish oil.
There are other sources of Omega 3's. ...instead of increasing the flax seed or fish and/or fish oil intake for the increase in Omega3's, we should look at safer plant based sources of Omega 3.
Mila by Lifemax (a blend of Chia) contains the highest and safest levels of Omega-3s on the planet. ... Mila provides the greatest known delivery system of Omega-3s to your body.
Read more about Mila at www.mykindofgrain.com
Posted by: Peter Sturm | Oct 15, 2009 5:50:09 PM
I am constantly reading research which shows that the conversion from the fats in the flax oil to the DHA is pretty much nonexistent. I guess that if this is true, then the flax oils should be used only as a addition to the fish oils (or the eating of fatty fish). What are your thoughts about this, Monica?
Posted by: susan allport | Oct 15, 2009 2:27:24 PM
I thought you would be interested in my article on omega-3s in Prevention Magazine: http://health.msn.com/nutrition/articlepage.aspx?cp-documentid=100245164
Posted by: David Moss | Oct 15, 2009 10:49:04 AM
@ Monica: Why would getting monounsaturated fats be necessary? I'm aware of them providing benefits relative to nutrients they're replacing, but not as having any vital function themselves. Also any-one eating meat and dairy is likely to be getting pretty substantial quantities of monounsaturated fat anyway. Most nut/seed/vegetable oils also contain substantial quantities.
Conversely I am aware of harms caused by having more PUFA to *saturated* fats, such as reduced testosterone, which being a hormone one can infer to be rather integral!
Also, why/how would you set an upper limit for total fat? Studies linking various harms to high fat consumption aren't that suggestive if the fats consumed are pointedly unhealthy fats (like masses of omega 6), which clearly isn't an issue in this case. If the issue at hand is really fat containing lots of calories, then saying you should limit fat arbitrarily is rather counter-productive (as per your immediately previous post).
I would certainly exhort the limitation of the total amount of PUFA (to far under 10% of calories) in fact. PUFA are highly reactive, oxidisable and whatever positive effects they produce on total blood cholesterol they've been shown to result in more oxidised LDL cholesterol (which is a far better predicter of negative cardiac outcome than any other cholesterol measures).
Also I'd essentially reverse the advice to hold back on fish oil but be more easy-going about ALA. You would expect the negative effects of high omega 3 you cite to result even from either form of omega 3. Conversely various of the benefits of omega 3 can only be gained from the long chain form.
Consequently the ideal, in my opinion, would be just enough long chain omega 3 to meet your specific needs (brain function etc), the RDA for which may or may not be accurate. Then essentially as little omega 6 as possible (yes minimal amounts are essential but in any western diet virtually impossible to miss). Then the rest of one's fat intake (if you don't just consume pure carbage) should be mono-unsaturated (or saturated fat in my view, but that's a contentious further debate).
Posted by: Dave | Oct 15, 2009 10:12:03 AM
I'd be more concerned about PUFA increasing oxidative stress, particularly when divorced from a whole food. Something like a flax seed or salmon has mechanisms in place to protect the PUFA from oxidation.
Humans convert ALA to EPA inefficiently, around 10%. When you take 10 grams of flax oil, what happens to the other 90%?
