Wednesday, February 27, 2013

Guest Post: More Facts on Fats

Last Saturday I linked to a story from Consumer Reports that highlighted the state of the evidence on fat. That story seemed to upset some folks and so I reached out to the article's authors, Kevin Lomangino, the editor of Clinical Nutrition Insight and asked him if he would be willing to expand on his piece and also to cover what the term, "state of the evidence" means as it would seem that some of the comments left that Saturday were left by folks who aren't familiar with the concept that rather than meaning "fact", it means this is what the data looks like today.

More Facts On Fats
Kevin Lomangino

Last week, Yoni linked to a Consumer Reports story called “Fat Facts and Fat Fiction” that some readers of this blog weren’t very happy with (scroll down to the comments for Yoni's post). This was no surprise, since dietary advice about fat is always controversial, and even experts disagree about what recommendations are supported by current evidence. As a contributor to the Consumer Reports piece (which was also extensively edited and reviewed by nutrition scientists), I thought I’d address some of the concerns raised in the comments and try to provide more detail and links to the evidence that we considered. Although I don’t harbor any illusions about changing people’s strongly held beliefs about food, I do believe that our recommendations are sound and were based on the best scientific evidence available. The following comments are my own opinion and haven’t been endorsed or reviewed by Consumer Reports.

Revisiting Saturated Fat

Is saturated fat bad for you? A growing number of people think that the answer to this question is no. But to my mind, this question can’t really be answered without first considering a second question: What are you comparing it with?

If you reduce your saturated fat intake, chances are very good that you’ll be replacing those calories with something else. And if that “something else” is refined carbohydrate, then evidence increasingly shows that you won’t improve your health and might be worse off. On the other hand, if you replace those saturated fat calories with unsaturated fat instead, then there’s consistent evidence that you may well reap some cardiovascular benefits from the change.

As noted in the Consumer Reports piece, a recent Cochrane Group meta-analysis, widely regarded as the “gold standard” in critical evaluation of medical evidence, found a 14% reduction in cardiovascular events across randomized controlled trials that involved reduction and/or replacement of saturated fat with unsaturated fat (including both mono and polyunsaturates). Another meta-analysis that looked specifically at studies replacing saturated fat with polyunsaturated fat (trials involving monounsaturated fat were excluded) found a 19% reduction in coronary heart disease events.

Admittedly, the reduction is not that big, which is something we tried to communicate by providing the absolute number of events that occurred in these studies (77 per 1000 people on a regular diet vs. 66 per 1000 on a reduced saturated fat diet in the Cochrane review). Still, the difference is substantial enough to provide meaningful benefits if large numbers of people carried out the recommendations.

These studies have limitations, of course, and some researchers say there’s not enough proof to conclude that replacing saturated fat with polyunsaturated fat is causally related to heart benefits. Then again, diet studies are hard to do and almost always yield messy evidence. It’s unlikely that better studies will be available any time soon to guide our thinking. Meanwhile, a variety of prestigious groups have looked at the evidence (see this and this), including epidemiologic and biomarker studies, and concluded that it all points to a benefit for replacing saturated fat with unsaturated fat. That’s a well-supported stance based on the current state of the evidence.

Benefits of Omega-6

Commenters also raised concerns about our endorsement of vegetable oils that are high in linoleic acid, an omega-6 polyunsaturated fatty acid. This reflects longstanding worries that high intakes of omega-6 are pro-inflammatory, and that increased intake may well lead to higher risk of cardiovascular disease and other adverse health outcomes.

As the article pointed out, this is an area where research is rapidly evolving and could quickly swing one way or the other based on new evidence. Nevertheless, studies conducted to date don’t show that we have much to worry about with linoleic acid. Despite theoretical concerns about inflammation, recent systematic reviews found no evidence that higher linoleic acid intake increases inflammatory biomarkers or their precursors. And as noted above, a meta-analysis of clinical trials showed cardiovascular benefits from replacing saturated fat with polyunsaturated oils. In these studies, participants replaced saturated fat primarily with oils rich in linoleic acid such as soybean oil, corn oil, and safflower oil.

As some researchers have pointed out, there may be a difference between studies that increased linoleic acid exclusively and those that also simultaneously increased intakes of omega-3 fatty acids such as alpha-linolenic acid, EPA, and DHA. They say that the increased intake of omega-3s in these studies may have negated the harmful effects of linoleic acid and obscured its risks. This view is supported by a recent analysis that unearthed old data from one of these studies – the Sydney Diet Heart Study, originally conducted in the 1960s – that replaced saturated fat exclusively with safflower oil, which is 75% linoleic acid and has no omega-3s. The authors reported a higher rate of death from cardiovascular disease in the safflower oil group compared with controls who consumed their usual diet (17.2% vs. 11%).

This study hadn’t yet been published at the time our Consumer Reports piece was released, and I think it will take some time before these results are fully digested. However, early reaction to the study has raised concerns about its validity – particularly the fact that the margarine consumed by participants in the safflower oil group apparently had considerable amounts of trans fat, which might have contributed to the cardiovascular harm seen in that group.

Personally, I wouldn’t fault anyone for being a little more cautious about safflower oil in the wake of these results. But it’s hard to see how the results implicate linoleic acid in general, as similar studies involving soybean oil, which is about 50% linoleic acid and has only a small amount of omega-3s (about 7%), have shown cardiovascular benefit. Also, you’d have to work very hard to obtain 15% of your daily energy from safflower oil as they did in this study. (The average intake of linoleic acid in the U.S. is currently 6.7%, less than half that level.) If you “choose a variety of plant-based oils, plus low-mercury fish such as salmon twice a week” as recommended by Consumer Reports, you can be confident of getting a good mix of unsaturated oils that are likely to be protective.

What About Hexane?

Lastly, some commenters expressed concern about hexane, an industrial solvent that is used to process many cooking oils, including soybean oil (as well as the defatted soy protein flakes that are used in veggie burgers and other soy foods). As we noted in the article, hexane is a neurotoxin, and testing has reported trace amounts residual hexane in soybean oil (about 10 parts per million). While that may sound disturbing, there’s really no evidence that such minute exposures pose any kind of health risk. Hexane has been shown to cause toxicity in places like shoe factories and furniture finishing shops, where workers were breathing in much higher amounts of hexane and absorbing it through their skin from rags. But these types of exposure are likely more hazardous than oral consumption, because they bypass the liver.

According to an EPA toxicological assessment, there has never been a study of oral exposure to hexane in humans, and very few studies in animals. John L. O’Donoghue, VMD, PhD, a toxicologist who was a coauthor on one of these rodent studies, told me that they had to give “very large quantities” of hexane before they observed any neurotoxic effects. How much are we talking about? This Slate article estimates you’d have to eat the equivalent of more than 1 million veggie burgers a day to approximate this level of exposure! O’Donoghue told me, “I don’t know of any reason to consider [hexane in cooking oil] a problem.” With that being said, I realize that we can’t definitively rule out adverse effects from chronic long-term exposure to small amounts of hexane, because the studies simply haven’t been done. And you may also wish to avoid hexane-processed oils for other reasons, such as worker safety and air pollution concerns. Still, in a world full of risks, hexane in cooking oil doesn’t rank as priority for personal health. You can avoid hexane by choosing expeller-pressed oils, but realize that these products will cost quite a bit more than hexane-processed oil, because the extraction process is much less efficient.

Take-Home Message

Why is fat so controversial? In my mind, it’s an example what Marion Nestle called the conflict between science- and belief-based perspectives. Many people believe that vegetable oils simply can’t be good for you, because they haven’t been a significant part of our diet for more than a few decades. They are much more comfortable eating the foods that we’ve always eaten, which more often than not will be animal-based and higher in saturated fat. It’s a valid view, but the science—at least for now—does not support this approach. The verdict could change if and when we get more high-quality research on this topic.

Meanwhile, the overall science-based message is not to worry too much about one specific food or nutrient. Where reasonable and feasible, replace butter and solid fats in your diet with a variety of plant oils such as olive, canola, soybean, and peanut. Eat fatty fish twice a week. Limit processed junk food that is usually rich in saturated fat and refined carbohydrate, and fill up most of your plate with fruits and vegetables instead.

Kevin Lomangino is editor of Clinical Nutrition Insight, a monthly evidence-based newsletter for physicians and dietitians, and a reviewer for He tweets as @Klomangino.

Bookmark and Share


  1. Interesting article, however I find these studies rather myopic.

    It's all very well switching from saturated to unsaturated fats to reduce your risk of cardiovascular diseases, but it could be rather futile if all it means is you die earlier of something else!

    From the cochrane review: "There were no clear effects of dietary fat changes on total mortality (RR 0.98, 95% CI 0.93 to 1.04, 71,790 participants) or cardiovascular mortality (RR 0.94, 95% CI 0.85 to 1.04, 65,978 participants)"

    I for one am not going to swap butter and bacon for vegetable oil and tofu on the basis of the data available at the moment.

    TBH, I think I'd need some pretty irrefutable evidence that I'd be guaranteed an extra decade of quality life to make those kinds of changes!

    1. Kevin Lomangino9:22 am

      I think that's a reasonable and rational approach. Some people want to do whatever it takes to increase their chances of staying healthy; others need a big incentive to change a way of life that they enjoy. Thanks for reading and commenting.

  2. Even if the overall conclusions of many observational studies suggest that canola oil is slightly better for me, I will continue to use olive oil and organic butter, even though I know these foods are not perfect choices.
    I am not convinced that using a vegetable oil that has been refined using n-Hexane, a solvent made from crude oil is a better choice than one that is not.
    Canola oil is made at a processing facility by slightly heating and then crushing the seed. Almost all commercial grade canola oil is then refined using hexane. Finally, the crude oil is refined using water precipitation and organic acid, "bleaching" with clay, and deodorizing using steam distillation.
    n-Hexane is a chemical made from crude oil. Pure n-hexane is a colorless liquid with a slightly disagreeable odor. It evaporates very easily into the air and dissolves only slightly in water. n-Hexane is highly flammable, and its vapors can be explosive….When rats are exposed to n-hexane in the air, they show signs of damage to their nervous systems very similar to those seen in people who became ill after workplace exposure. As in people, these effects in rats depend on the concentrations of n-hexane in air and how long exposure lasts.
    Even if n-Hexane does not cause any harm as a result of residues in the food that I eat, a much larger amount of this toxin enters the air, soil and water during the manufacturing process. Since n-Hexane is petroleum product, the production of canola and soybean oils is dependent on a non-renewable fuel.
    However, it is also true olive oil has to be shipped from Italy to Canada in order for me to buy it at the store, so petroleum is used in the transportation process (and probably in the manufacturing process). No doubt, some aspects of raising local, grass-fed organic cows is also energy-consuming.
    I think that olive oil and butter are the lesser evils in this case. Of course, I have no idea if it would ever be possible to supply enough organic butter to feed the world! It seems unlikely.
    Health is more than simply the effect of a particular food on your body when you consume it. Health is the result of the whole environment including the water, air and soil. If the processes involved in creating our foods contaminate the environment, then the foods aren’t really healthy.
    I guess that’s why there is a movement to produce and consume local foods. I know that’s not always feasible, especially in a really cold climate like ours. There are no perfect solutions, but I don’t think that canola and soybean oils as they are now produced should even make the long list of acceptable solutions to feeding the world.

    1. Kevin Lomangino9:31 am

      We did not factor in the larger environmental perspective, as the evidence was daunting enough when we restricted it to personal health! I appreciate your comments.

    2. Anonymous1:21 am

      Hexane oil extraction plants don't belch out hexane, it's too expensive and explosive to do so.

  3. The Sydney diet heart study was from 1978 but in the original study CVD deaths were not reported. The data may have been "lost" so that the study could get published. I talked to the researcher from this study:

    5 Journals refused to publish his work because it did not agree with current advise that saturated fat is bad and polyunsaturated fat is good.

    How many other studies in the meta analysis "lost " data so the studies could get published, we will never know.

    You mentioned trans fats in the Sydney study. This brings up an interesting point. In almost every study I have looked at concerning saturated fat that was done before 2000 grouped saturated fats and trans fats together. The studies in the meta analysis mention were all from before 1993.

    As an example of the problems with meta analysis is that one analysis you linked to included the Dayton 1968 – Los Angeles Veterans study. This study failed to control for smoking as there were twice as many smokers in the control group. Its hard to compare CVD risk when you don't control for smoking. As well cancer rates increased greatly in the polyunsaturated group in the last couple years of the study and were higher than in the control group, even though the control group had twice as many smokers.

    1. Was that the same study where they re-used the butter over and over again but the corn oil (or whatever they used) was new every time? I think Chris Masterjohn did an excellent overview of that awhile back. And I believe that, toward the end of the study (~8 years) the poly group's rate of death was starting to shoot through the roof!


    3. Kevin Lomangino9:43 am

      Publication bias -- studies getting "lost" or not published because of negative findings, is a problem in nutrition and many other areas. And I agree that there are many limitations and problems with the existing studies. As I recall, Gary Taubes had a good discussion of these problems in his book, Good Calories, Bad Calories. Nevertheless, most diet advice is based on observational studies. When you have trials with real outcomes that are mostly pointing in the same direction, that's a strong signal. I look forward to the day when we have better evidence to guide recommendations.

    4. Yes Andrew that was the study Chris was talking about.

      Kevin, The Sydney Heart Study was a trial, but when originally reported on in 1978 they did not include the CVD deaths. They did report a drop in cholesterol so this study was used to support the idea that polyunsaturated fats reduce the risk of heart disease since cholesterol dropped, even though more people died of heart disease. This is an example of how even clinical trials can be manipulated to provide the outcome that the sponsors of trials want. In some cases they are no better than observational studies.

      As well even the best controlled trials before 2000 did not differentiate between sat fat and trans fat. Most people in the control group eating sat fat were also eating more trans fat so we can't tell if it was the sat fat or trans fat that resulted in poor outcomes.

  4. Kevin, You wrote, "Commenters also raised concerns about our endorsement of vegetable oils that are high in linoleic acid, an omega-6 polyunsaturated fatty acid. This reflects longstanding worries that high intakes of omega-6 are pro-inflammatory, and that increased intake may well lead to higher risk of cardiovascular disease and other adverse health outcomes. As the article pointed out, this is an area where research is rapidly evolving and could quickly swing one way or the other based on new evidence. Nevertheless, studies conducted to date don’t show that we have much to worry about with linoleic acid."

    You're right. Human studies don't suggest that lenoleic acid is harmful. That's because there aren't any studies in which lenoleic acid intake was deliberately restricted. However, there was a study in which lenoleic acid was inadvertently restricted. Here's what happened:

    "The only long-term trial that reduced n-6 LA intake to resemble a traditional Mediterranean diet (but still higher than preindustrial LA intake) reduced CHD events and mortality by 70%. Although this does not prove that LA intake has adverse consequences, it clearly indicates that high LA intake is not necessary for profound CHD risk reduction."

    To access the article, Google "PDF Dietary Fat Quality and Coronary Heart Disease"

    We all know that studies have limitations. That's why scientists do experiments. So what does the experimental evidence suggest? Excerpts:

    "Fish oil supplements fed to mice already on a diet rich in vegetable oil interfered with the ability of tissues in the gastrointestinal system to repair themselves, according to recent research by Sanjoy Ghosh published in the British Journal of Nutrition. Additional unpublished results point to a similar effect on cardiac tissues. The cellular disruption that led to tissue injury — called oxidative stress — appears to be caused by the combination of omega-6 fatty acids in vegetable-based oils and the omega-3 fatty acids in fish oil, according to Ghosh, an assistant professor of biology."

    Here's more:

    Southampton researchers have demonstrated that mothers who have higher levels of n-6 polyunsaturated fatty acids (PUFAs), which are found in cooking oils and nuts, during pregnancy have fatter children.

    While a diet rich in saturated fat and polyunsaturated fats from olives, nuts and fish is still thought to benefit overall health, excessive amounts of omega-3 fats, especially with omega-6 fats already concentrated in the body, could trigger inflammation responses in tissues.

    A recent study suggests a close association between dietary omega-6 and the development of overweight and obesity. Omega-6 is a type of fat found in certain vegetable oils which is present in large amount in processed and junk food. New results from experiments using animal models show that a high intake of omega-6 led to overproduction of signalling compounds that stimulate the appetite, with the result that the animals ate more and developed obesity.

    The question is, how much evidence do we need to raise suspicions that lenoleic acid may exhibit toxic effects when consumed in amounts exceeding one to two percent of total caloric intake?

  5. Anonymous10:41 am

    I'd like to "follow the money" on this one...there is a lot of incentive to keep us all eating stuff we shouldn't. Telling people to eat canola and soy in particular is terrible advice. Wow.
    I'm surprised you subscribe to this conventional wisdom, Dr. Freedhoff.

    1. Kevin Lomangino11:02 am

      I should have disclosed that I have been a consultant for the Informed Medical Decisions Foundation and the American Association of Diabetes Educators. Not that either of these relationships has anything to do with vegetable oil, but in the interest of full disclosure...

    2. Anonymous2:32 pm

      Well, there is a link between Canola and the AADE, as well as big Pharma, and no offense, but this is the kind of nonsense that is in keeping with those alliances...

    3. Kevin Lomangino4:19 pm

      I don't see the connection, but you are correct to ask for full disclosure and can draw your own conclusions.

    4. Cleethorpes8:53 am

      You don't see a conflict of interest defending your organisations' mainstream advice based on similar flawed evidence?

  6. "According to an EPA toxicological assessment, there has never been a study of oral exposure to hexane in humans, and very few studies in animals. "

    I would prefer to base my personal health decisions on the precautionary principle. A novel addition to the human diet should need to be proved safe rather than haphazardly added and then maybe removed if it is proven unsafe.

  7. People get altogether too worked up about diet, and dietary fats in particular. The three subjects people have the strongest opinions about (and often, the least objective information about) are religion, politics, and food. In addition, blog comments sections always represent the most vocal (frequently, most angry) readers and do not necessarily represent the "silent majority" that may have derived value from your writing.

    I think Lomangino did a good job of representing the majority position of the diet-health research field, and not getting overconfident about the state of the evidence.

    That said, I have reservations about the idea that dietary saturated fat is harmful. Nearly every prospective observational study has found that SFA consumption is not associated with heart attack risk, as reflected by the Siri-Tarino et al. meta-analysis. I understand the argument that we have to consider what is replacing the SFA, but frankly I find that to be somewhat of a cop-out when it comes to the observational evidence. That's not how epidemiology works-- either people who eat more SFA have more heart attacks (after "controlling" for other diet and lifestyle factors), or they don't. If they don't, then the evidence doesn't suggest it's harmful, regardless of how protective it is to eat nuts instead. The point is, even if increasing nut consumption at the expense of butter reduces heart attack risk, that doesn't mean butter is harmful, it simply means nuts are protective. I think this is more than an academic point. Someone could read these papers, get the impression that butter is going to hurt them, and eat less butter without increasing consumption of nuts (instead increasing other components of the typical diet). They would likely derive no benefit from that change, and it may even be harmful.

    As mentioned by a previous commenter, the controlled diet trials also must be interpreted with a great deal of caution. The Finnish mental hospital trial is frequently included in these meta-analyses and it often accounts for a large proportion of the "effect", but it suffered from multiple fatal methodological flaws, including a completely unblinded design (i.e. the physicians who were diagnosing disease and doing autopsies knew which diet group patients were in), no randomization, much greater use of cardiotoxic drugs in the control groups, and other problems.

    Another problem is that the control groups in these studies typically consumed much more trans fat than the intervention groups. These interventions did not really test SFA replacement with PUFA-- they tested SFA and trans fat replacement with PUFA.

    Here's my opinion. The evidence from controlled trials suggests overall that replacing trans and saturated fats with omega-3 containing PUFAs is likely to reduce heart attack risk, at least among people who already suffer from cardiovascular disease. Whether we would see a similar effect after controlling for TFA is another question, and one that has not been answered definitively to date.

    1. I imagine Kevin will weigh in. The only line in the original article I took issue with was the first line on saturated fat stating it was bad. The rest of the article, and this guest post, clearly indicate that it's not that saturated fat is bad, but rather than replacing it with unsaturated fat has fairly conclusively been shown to be good. Full stop. Wish that singular line wasn't in the Consumer Reports piece, but certainly that sentiment isn't echoed in Kevin's excellent post here.

    2. Fair enough. I also want to acknowledge the uncertainty of my own position, partially because this isn't my research specialty. It still seems possible to me that SFA will be proven harmful in the end, I just don't feel the evidence is there yet. But I doubt there's any harm in eating a diet low in SFA-- it's certainly not an essential nutrient.

    3. Kevin Lomangino4:10 pm

      Thanks for these thoughtful comments. In retrospect, I agree that the saturated fat answer in the Consumer Reports piece could have been more nuanced. The point is not that saturated fat is a directly harmful substance that should be avoided at all costs, but rather that it would be healthier for most people to replace it where possible with unsaturated fat. I hope that's the main point that people took away from that part of the discussion.

      I think we were sensitive to Stephan's concern that people might cut out the butter and start eating something less healthful, which is why we were pretty careful to specify that it had to be replaced with a healthier alternative (not refined carbs).

      We also didn't want to give people the idea that there was nothing at all to worry about when it comes to saturated fat. When you look at where most saturated fat in the diet comes from these days, after full fat dairy, you have pizza, grain-, and dairy-based desserts. You can argue about how much blame saturated fat should take for the "badness" of these foods, but the bottom line is these are not healthy foods, and people would be better off if they cut down on them and replaced them with something better.

      Lastly, I agree that there are many limitations in these studies, although our conclusion was that the evidence still pointed in the direction of benefit. I am curious Stephan about your opinion on replacing trans and sat fat with omega-3 containing PUFA. Do you include soybean oil in the category of omega-3 containing PUFA due to its ALA content? I think a lot of people have an aversion to soybean oil due to its link to industrial agriculture and its high linoleic acid content, but many of the controlled trials that showed benefit replaced saturated/trans fat with soybean oil.

    4. Anonymous10:27 am

      No you are right, SFA is not an essential "Nutrient" as we are capable of assimilating our own from glucose. In regards to PUFAs from nuts etc I cant see how these can be healthful. It doesn't matter whether it is from vegetable oil, nuts, seeds or fish it is still an unstable fat with dangerous breakdown products.

    5. Pizza, grain and dairy-based desserts are full of white flour and/or sugar. Many desserts (such as store-made cookies) are made with vegetable oils and are not high in saturated fat. The problem isn't fats (except perhaps trans-fat). It's what we do with them that gets us into trouble!

  8. Wow, my favourite nutrition blog just got some link love from my favourite film blog...Worlds are colliding, Dr. Freedhoff, and I thought you'd like to know!

    Link is near the bottom of the page...congratulations!! :o)

  9. There are many examples of groups of people who have traditionally eaten 20-50% of their calories from saturated fat and were very healthy. The Masai, Inuit, Pacific Islanders like the Tokelau and even the starch loving kitavans ate close to 20% of their calories from saturated fat if my research is correct. How many traditional cultures get 20-50% of their calories from polyunsaturated fat? If you want to be part of this experiment to see how refined omega 6 oils that have only been consumed in the last 100 years affect our health go ahead. I will stick to the butter, coconut and olive oil that have been consumed for thousands of years.

    What is the biochemical mechanism that makes saturated fat more unhealthy than polyunsaturated fats. I'm not a biochemist but it seems to me that a stable fat that the body likes to make when it has excess calories would be a better choice than a fat that oxidizes easily and is not made by the body.

    1. Cleethorpes9:04 am

      Yes, but you wouldn't make as big a profit as you can from selling soy bean oil, and imagine having to admit that the theory that's been advocated for the last thirty plus years as healthy eating advice was wrong. Not going to happen until somebody gets sued.

  10. Very informative blog! I think I can understand(or atleast claim to) why these studies on replacement of saturated fats with other nutrients yielded messy and inconclusive results. One of the reasons could be the amount of physical work or workouts that each person, who was the subject of study, did.

    A person who worked out regularly and yet consumed saturated fats on regular basis might be less prone to heart diseases than the one who consumed same saturated fats but did not do any workouts.

    1. I have enjoyed this vigorous dialogue so far, but I still have as many questions as many of you, presumably because, as was pointed out, that so much nutritional science is simply observational, and what isn't has been hard to control.
      Despite all of what has been said about the supposed evils of saturated fat, it appears that an actual mechanism of how dietary saturated fat ratchets up total and/or LDL has yet to be elucidated, let alone the atherogenesis of SFAs, given the subfractions of LDL now being discussed. Further, there has been described differences in the atherogenicity of individual SFAs, with Lauric, Myristic, and Palmitic leading the way, and Stearic being equivocal, yet it is not clear what the significance of these observations are, given that we do not eat these fats individually. And given that we instead eat foods that usually contain a mixture of several fats--those that swear by olive oil are still getting 15% of their fat as SFAs, so they can't be that bad for you..

      In the meantime, I like Melissa McEwan's sober perspective (that a novel item in the food chain should have the burden of proof of safety before it is haphazardly added, then hastily removed if unsafe) very effectively argues against Marion Nestle's point against "believed-based" vs science-based decisions. A healthy scepticism of novel foods (and medicines for that matter) is a much safer position (some might even call it common sense) than hiding behind the veil of scientific dogma that always starts with, "There is no evidence that ..." Haven't we learned anything from limited science? Thalidomide? Smoking? There is probably "no evidence" that eating specific vegetable like cauliflower is good for you in some way, but most scientists are happy to extrapolate its value as part of the large group of vegetables it is a part of. But the minute someone purifies a component of it, whether as an oil or anything else, I, and most people would prefer to say that there is no evidence that this IS safe to eat in quantities never taken in before, than to state that there is no evidence that it is unsafe, so go ahead and be a guinea pig. We all have one life to live, so why not be cautious about novel foods?

  11. Kevin Lomangino11:09 am

    I appreciate this perspective M doc, but respectfully disagree that unsaturated vegetable oils (extracted with hexane or otherwise) can be considered novel additions to the food chain at this point. Our body requires a certain amount of these oils or we develop a fatty acid deficiency, so the presumption should be that we are biologically hardwired to consume them. Clearly we are consuming more than we have historically. But at this point, many thousands of people have been followed in epidemiologic studies and clinical trials. When they consume more unsaturated oil (most of it no doubt extracted with hexane), the evidence suggests that they are more healthy, not less. There is certainly much wisdom in the precautionary principle, but I just don't think comparisons to smoking and thalidomide are apt in the case.

    1. "Our body requires a certain amount of these (industrially processed seed) oils or we develop a fatty acid deficiency..."

      Until it is demonstrated otherwise, my position is that omega-6s are toxic at levels exceeding one to two percent of calories. To consume higher levels than that is to risk exceeding the body's ability to control their action. In addition, the research shows that the requirement for omega-6s (if there actually is one) is extremely low.[1] Excerpt:

      "These experiments were done in rats consuming an extremely purified sucrose-casein chow. Refined sucrose has been shown to increase the EFA intake requirement. Facilitating the conversion of LA to ARA are several vitamins and trace elements that modulate the EFA requirement. Vitamin B6 is so influential (desaturation of D6D) that later it would be shown to cure the deficiency when provided alone."

      Another point. It is the Amount of omega-6 in the diet that is the novel addition.[2] Excerpt:

      "One of the major differences between our post-industrial diets and the evolutionary and traditional foods of our past is in the kinds of fat we eat. One huge change has to do with the polyunsaturated fatty acids (or PUFAs), which come in several varieties, but most commonly omega 6 and omega 3. PUFAs are "essential fats," meaning we can't make them from other types of food, and we must eat them. However, never in the history of humankind have we eaten novel omega 6 fatty acids in such massive quantities."


  12. Is there any evidence which shows that the consumption of saturated or animal fats in the absence of refined carbohydrates (i.e. a paleo-style diet) is atherogenic?

  13. Anonymous10:11 am

    Thanks for a thoughtful discussion.

    But doesn't the fact that the Hooper study is the only one you can cite give you pause? After all, it showed only a 1% decrease in events (not deaths) in men (not women), but "there were no clear effects of dietary fat changes on total mortality . . . or cardiovascular mortality". Is that the most damning evidence we have against SFA, after decades of looking? Isn't it a rather pathetically weak foundation on which to base such an enormous edifice of dietary advice? And that's before you consider all the other meta-analyses showing no relationship between SFA and disease, eg:

    Siri-Tarino PW, et al. Meta-analysis of prospective cohort studies evaluating the association of saturated fat with cardiovascular disease. Am J Clin Nutr. 2010 Mar;91(3):535-46.
    "…no significant evidence for concluding that dietary saturated fat is associated with an increased risk of CHD or CVD."

    Mente A, et al. A systematic review of the evidence supporting a causal link between dietary factors and coronary heart disease. Arch Intern Med. 2009 Apr 13;169(7):659-69.
    "Insufficient evidence of association is present for intake of … saturated or polyunsaturated fatty acids; total fat … meat, eggs and milk."

    Skeaff, C. M., & Miller, J. (2009). Dietary Fat and Coronary Heart Disease: Summary of Evidence from Prospective Cohort and Randomised Controlled Trials. Annals of Nutrition and Metabolism, 55(1-3), 173-U287 (://000269790400010)
    "Intake of SFA was not significantly associated with CHD mortality... Similarly SFA intake was not significantly associated CHD events"

    Meanwhile, it is not clear that PUFAs (n-6 at least) are all that good for you. Even before the latest negative results from the Sydney study, Ramsden's previous meta-analysis concluded that: "RCT that substituted n-6 PUFA for TFA and SFA without simultaneously increasing n-3 PUFA produced an increase in risk of death that approached statistical significance (RR 1.16; 95% CI 0.95, 1.42). Advice to specifically increase n-6 PUFA intake, based on mixed n-3/n-6 RCT data, is unlikely to provide the intended benefits, and may actually increase the risks of CHD and death."

    Ramsden, C. E., Hibbeln, J. R., Majchrzak, S. F., & Davis, J. M. (2010). n-6 Fatty acid-specific and mixed polyunsaturate dietary interventions have different effects on CHD risk: a meta-analysis of randomised controlled trials. British Journal of Nutrition, 104(11), 1586-1600. (://000285463900003)

    The Astrup study you cite concurs, saying only that fish oil is likely to be better than SFA.