Skip to Content

From the Far Corner of the Basement

To go along with that recent CETP trial news, here’s another one for the “We don’t know much about human lipid handing” file. A dietary study originally done back in the 1960s and 1970s has been (almost literally) resurrected, with data pulled out of yellowing stacks of paper, old cardboard boxes, and ancient-format computer tapes.

What it shows is that, under about the most controlled conditions possible in a large human trial (institutionalized patients being served standard meals), that replacing saturated/animal fat in the diet with vegetable-derived fats and oils provided. . .no cardiovascular benefit whatsoever. In fact, the lower the cholesterol levels of the patients, the higher their death rates. This was in over 9,000 subjects over five years, probably the largest study of its kind ever conducted, and it had only produced one (not very thorough) paper in 1989 that didn’t make much of an impression.

After all, Everyone Knew by that point that saturated fat was bad for you – higher cholesterol, atherosclerosis, cardiovascular mortality, and the case was closed. But the evidence for this has never been as strong as you’d imagine. Most of the studies that have backed it up are observational (with all the problems that entails), and some of them have never been fully published themselves. That Stat link in the first paragraph has more on this, and Gary Taubes has a great deal more in this book.  (Whatever you think about his own dietary recommendations, it’s hard to refute his evidence that the entire official-dietary-recommendations experience has been a shambles). And there have been meta-analyses of the published data suggesting that even it doesn’t support the unsaturated-fat-good/saturated-fat-bad view. It’s interesting, in light of current evidence, to go back and read some popular cookbooks and dietary plans from the 1980s, which basically tell you to (1) cut out all saturated fat, (2) cut out the rest of the fat as much as possible while you’re at it, and (3) eat as many carbohydrates as you can hold. Well, that’s an exaggeration, but the introduction to the original edition of, say, Jane Brody’s Good Food Book starts to sound like that.

Beyond the dietary issues themselves, there’s an interesting psychological component to this story. Why was there only one paper from this huge study, and why did it take sixteen years after its completion to reach print? Its lead author was a hard-working, dedicated medical researcher:

The Frantz children always felt fortunate that their father brought his work home, his beliefs about the dangers of saturated fat shaping what the family ate. “Other kids would have ice cream; we had ice milk,” recalled Ivan Frantz. Bob said they were “reared on margarine,” foreswearing butter’s saturated fat.

It’s possible, Bob Frantz said, that his father’s team was discouraged by the failure to find a heart benefit from replacing saturated fats with vegetable oils. “My feeling is, when the overall objective of decreasing deaths by decreasing cholesterol wasn’t met, everything else became less compelling,” he said. “I suspect there was a lot of consternation about why” they couldn’t find a benefit.

The coleader of the project was Dr. Ancel Keys, author of the Seven Countries Study, Time cover subject, and the most prominent advocate of replacing saturated fat with vegetable fat. “The idea that there might be something adverse about lowering cholesterol [via vegetable oils] was really antithetical to the dogma of the day,” Bob Frantz said.

His father, he said, “was always committed to discovering the truth. He would be pleased this is finally coming out.”

It’s impossible to know for sure, but it seems likely that Franz and Keys may have ended up regarding this as a failed study, a great deal of time and effort more or less wasted. After all, the results it produced were so screwy: inverse correlation with low cholesterol and mortality? No benefit with vegetable oils? No, there must have been something wrong. The boxes full of data sat for decades, unlabled, in a far corner of a basement, and you wonder if Dr. Franz thought about them. Did he regret all the time spent on the study? Did he regret that more wasn’t done with the data it generated? We’ll never know – but what we do know is that he never threw those boxes away.

25 comments on “From the Far Corner of the Basement”

  1. roger says:

    The Guardian had a story recently entitled “The Sugar Conspiracy,” (www.theguardian.com /society/2016/apr/07/the-sugar-conspiracy-robert-lustig-john-yudkin) which details the story of John Yudkin. Upon noticing that sugar consumed was turned into fat by the liver, Yudkin pronounced sugar the cause of obesity and other maladies, Pure, White, and Deadly, 1972. Since this contradicted the fat-cholesterol theory he was professionally destroyed. It is a fascinating read.

    1. Over and Out says:

      Really enjoyed this piece when it came out, especially the paragraph about the 2015 paper “Does Science Advance One Funeral at a Time?” and the kind of ingrained academic nepotism that bully’s disfavoured view points into submission. Fat vs Sugar is certainly a hot topic at the moment!

  2. Random Scientist says:

    Derek – I saw this article the other day and I’m glad you brought it up. In addition to providing the best evidence that we really don’t know much about human lipid handling and cholesterol, in a broader sense it also provides a good teaching example of the way in which personal ego and selection bias influences what is published / not published.

  3. c says:

    The constant upheaval of nutritional advice from “experts” discredits science generally in the public’s eyes. If these people can’t be trusted to figure things out year-to-year, why should we trust the conclusions of experts in climate change, vaccination, or evolution?

    Of course, it’s the typical problem that the loudest most confident voice gets heard. Nobody wants to hear that biology is complicated and that your health is (mostly?) unpredictable. Having a boogeyman feels better than being helpless.

    1. NJBiologist says:

      c: It’s worth keeping in mind that the reproducibility of results–and stability of findings–varies by study type. Generally, larger, better-controlled experimental studies are more reproducible than smaller, less-controlled experimental studies and larger, better-controlled observational studies; those tend to be more reproducible than smaller, less-controlled observational studies. So, for the fields you mention:

      Nutrition–generally observational, sometimes badly riddled with control issues
      Climate change–only observational data; controls hard to assess
      Vaccination–generally large and well-controlled experimental studies (clinical trials)
      Evolution–much observational data, but also some unintentional experiments (moths in the English countryside during the industrial revolution; bacteria during antibiotic treatment)

  4. ScientistSailor says:

    The human body is such a complicated system, that it’s silly to think that health outcomes can be linked to one variable. Genetics, diet, activity level, microbiome, state of mind, are all interdependent in such a way as to prevent clear deconvolution.

    Chalk this up to another fail for super-reductionist approach.

    1. Hap says:

      It’d just be another thing we didn’t know if people hadn’t enforced it as holy writ. It’s probably also another data point for “Just because you’re really certain something is true doesn’t mean it is true.” and “A little humility is probably always in order.”

  5. watcher says:

    Let’s not forget the work by Brown and Goldstein on the roles of “good” and “bad” cholesterol which begot statins (and a lot of heavy advertising, particularly by Merck, to doctors and the public alike) and won the two a Nobel Prize. Which begot statin absorption inhibitors that ultimately did not have much if any effect on outcome, yet made a lot of money for their sponsoring companies. And seeking the holy grail of a magic pill to alter the HDL/LDL ratios for the better.
    Ironically, statins do appear to help in long term survival, but likely not for the reason they were initially devised. And for this, the studies required huge patient studies run over many years. And then there has been the use of Fish Oil to lower lipids. “Lies, Big Lies, and Statistics”? (Mark Twain).

    Now, how do we “uneducate” thousands of doctors from insisting the goal is to lower cholesterol (but not too low) and to clarify the roles of LDL and HDL.

    1. John Wayne says:

      It is true that our understanding of biology is constantly evolving. This often leads to theories of health being changed; usually a result of the clearing away of ignorance. Lots of folks think we know everything, and portion out knowledge in dribs and drabs. In reality, we are wrong a lot. In the research lab there is a lot more misleading data than misleading people. In advertising, it is the reverse.

      How do you educate doctors? They are just like every other profession; some are good, some are coasting and not paying attention, and some are terrible. The best cause change is with cold, hard clinical data that makes ignoring it unethical. Designing those studies is difficult, but it is the best thing we can do to start to shake off our crushing ignorance of biology.

  6. Levorotatory says:

    Cognitive dissonance is the hobgoblin of the scientific method.

  7. steve says:

    Mark Twain summed it up a long time ago: “”It ain’t what you don’t know that gets you into trouble. It’s what you know for sure that just ain’t so.”

  8. Ms. Zxy Atiywariii says:

    “Education consists mainly of what we have unlearned.” (Mark Twain)
    It’s sad how many people’s health is harmed in this ever-evolving process.

  9. Nick K says:

    This is a stunning result, completely at odds with conventional wisdom, but how does it square with the well-established health benefits of the Mediterranean diet (lots of olive oil)?

    1. David Morgenstern says:

      Also lots of meat and dairy fat.
      The bottom line is that the main factor deciding the influence of nutrition on health is caloric intake. if you eat as much as you spend, you’ll be healthier than if you’re intake is higher, simply as that – regardless of which diet you’re using – japanese, chinese, indian or Med.

    2. HFM says:

      All else being equal, if you swap out corn oil (mostly omega-6 polyunsaturated fat) with olive oil (high in monounsaturated fat), you are almost certainly doing your health a favor. It’s pretty well established that corn oil (and others like it) are pro-inflammatory.

      Also, the Mediterranean diet does some other things right. If you’re going to eat carbs, better to have slow-digesting starches cut with some fat, so that they won’t give you insulin spikes. A chunk of bread dipped in olive oil or a handful of nuts is going to be much more satisfying than the fat-free cookies I grew up snacking on. (Yes, I had health-conscious parents…and I was straight-up obese, despite their efforts to keep me on a “healthy” diet.)

      It seems to be the case that a typical Western diet (high sugar, high protein) causes pathological increases in cholesterol, and that by trimming the excess sugar and protein with a “Mediterranean-style” diet, you can decrease total cholesterol in a way that genuinely reflects improved health (mostly via lowering triglycerides). It also seems to be the case – and this is what the field has resisted, kicking and screaming – that by limiting non-polyunsaturated fats, you will produce a pathological *decrease* in total cholesterol (largely via lowering HDL) that increases the risk of certain diseases, particularly stroke.

  10. cthulhu says:

    The human nutrition researchers need to be hit upside the head with Diax’s Rake: http://anathem.wikia.com/wiki/Diax's_Rake

  11. Quercus says:

    I’m reminded of my own experience as a young grad student in the early 90s: I kept generating data that conflicted with well known dogma in a closely related field. Despite using a range of experimental techniques and conditions, my results very consistently pointed to the “wrong” result.

    The consensus among my advisors was that I must have been doing something wrong and that it would be foolish of me to pursue the results, especially if I wanted to finish my Ph.D. in a reasonable amount of time. I had several very prominent scientists on my thesis committee, including someone who was seen as a likely Nobel candidate. Since I was still a pretty new student, I grudgingly accepted their advice, and I dropped the work to start on a “safer” project. I figured that they must “know better.”

    Three years later, another lab overturned the dogma with a paper that made the cover of Nature, presenting experiments not very different from the ones I performed.

    Key lesson: if you design and perform your experiments well, including your controls, then trust your data, not your dogma!

  12. johnnyboy says:

    A few years back in a CRO I supervised a study in ApoE ko mice where not only fish oil extract but also the pravachol control actually increased the size of atherotic plaques, pretty clearly and with a nice dose-response. Didn’t hear from that sponsor again…

  13. Renee says:

    I certainly remember growing up in the 1960s when butter was constantly blasted as leading to heart disease and various other maladies, and how we should all give it up and only use margarine. There were constant ads on TV for this and that brand of margarine. But nowadays, margarine is rarely advertised, and there’s a lot less of it on store shelves. And it seems butter is back, though now there are lower fat versions of it, as well.

    Back to the issue of doctors and cholesterol – I’m a woman in her 50s, who recently was sent to a urologist for a possible kidney stone. While taking my medical history, he asked me which cholesterol-lowering drug did I take. I told him I don’t take any, since I don’t have high cholesterol. He glumly said “Well, you’re one of the few.” Is it de rigueur for people to be pressured into taking these drugs?

    1. Carya says:

      Interesting question. I have mildly elevated total cholesterol (200-230 mg/dL, with HDL, HDL/LDL, and triglycerides generally OK). About three months ago my primary care physician recommended that I initiate statin therapy. When I indicated a desire to first do some cost/benefit and side effects research, he unilaterally terminated me as a patient in his practice. And here I thought doctors wanted their patients informed and involved in their own health care. Silly me.

      1. Anon says:

        Carya, your physician is subject to PQRS for part of his reimbursement. PQRS requires treatment in this case, or he gets dinged. You’re well shut of him, he doesn’t treat patients- he treats lab results and uncritically follows the current fads.

        If you take a look in sources like Documenta Geigy (old enough to be unbiased on the subject of normal values, but good luck finding a copy) you’ll see that 200-230 mg/dl in an adult was WNL 50 years ago.

        A good knock-out question for interviewing a physician is when to treat with statins.

  14. Eat yr beans says:

    This study used hydrogenated oils fed to the control and the intervention diet with more vegetable oil. From the article:
    “As common margarines and shortenings of this period were rich sources of industrially produced trans fatty acids, the control diet contained substantial quantities of trans fat.”

    Further in the article:
    “Because the trans fatty acid contents of MCE study diets are not available, one could speculate that the lack of benefit in the intervention group was because of increased consumption of trans fat”

    But who knows? They said later in the discussion they then say the original authors (Frantz and Keys) likely tried to decrease the trans fat content…but I don’t buy it personally. They keep saying margarine throughout the article and I only know I can buy corn oil as a liquid. I’m glad Walter Willett of the Harvard public health says the same in his response
    http://www.bmj.com/content/353/bmj.i1246/rr-10

    The conclusion: saturated fat is bad but trans fat is worse for you. Nothing new here. Low LDL throughout your lifetime is important and the best approach is not to wait until you have symptoms of heart disease because it might be too late. That’s why the American Heart association recommends less than 7% of calories (16 g/day) for 2000 calorie diet for those over 2 years old. Prevention is the cheapest option for our system.

    I’m not surprised this is getting this much press as people do love to hear that they aren’t putting their long term health at risk eating animal fats (also coconut, palm oil and of course trans fats) on a daily basis. That is not what the best available evidence shows. A large meta-analysis showing the benefit of replacing saturated fats with complex carbs and/or polyunsaturated fats. It also shows the benefits of eliminating cholesterol:
    http://www.bmj.com/content/314/7074/112.long

    1. HFM says:

      The linked article does show that substitution of saturated fats for either polyunsaturated fats or carbohydrates reduces blood cholesterol. This is well-known (and consistent with the Minnesota trial results). It doesn’t show a health benefit of this reduction – given that the studies were on average 1 month long, there’s no real way to do so.

      As for the trans fat issue, I agree that it’s a potential confounder – but given the “control” group was using regular margarine, I don’t see how cutting said margarine with corn oil increases trans fat intake. The “intervention” group was almost certainly eating less trans fat, which makes the lack of benefit even more remarkable.

  15. Dr. Lloyd T J Evans says:

    Interesting research indeed, but it could be potentially dangerous. It does look like saturated fats have been absolved of responsibility for causing heath problems. But if this is more widely publicized, it could perversely lead to even greater problems. How so? Because if people think that eating saturated fats is ok, they will probably feel far less guilty about stuffing their faces with crap from McDonalds or KFC.

    We can quibble all we like about saturated fats, monounsaturated fats or polyunsaturated fats and what they might do to our HDL or LDL cholesterol levels. Likewise, we can debate until the end of time about which type of cholesterol is good or bad. But there is one aspect which cuts through all of the babble, which is the calorific value of the fats we consume.

    It is undeniably true that saturated fats have a higher calorific value than unsaturated oils. Likewise, any sort of fat or oil has a far higher calorific value than proteins, starches or sugars. So if your goal is to lose weight and stay healthy, it is hard to argue that eating less fat (of any sort) is somehow a bad idea. A diet of fish, rice and vegetables is always going to be better than one of cheeseburgers, fried chicken and chips. I can’t see any study disproving that any time soon.

Comments are closed.