Saturated fat and heart disease: studies old and new

A study appeared this week sure to drive members of the low-fat and vegan tribes sprinting for their Protexid.

Ron Krauss and his group published a paper in the Articles in Press section of the American Journal of Clinical Nutrition (AJCN) stating there is no evidence that saturated fat intake increases the risk for heart disease.  The paper, titled Meta-analysis of prospective cohort studies evaluating the association of saturated fat with cardiovascular disease, is not a study per se, but is a meta-analysis, a compilation of numerous studies looking at the relationship between saturated fat intake and the risk for developing heart disease.

As I’ve discussed before on these pages, meta-analyses are not my favorite types of studies.  I’ve attacked them when they’ve been used to ‘prove’ the low-fat diets are better, so I can’t very well embrace meta-analyses when they present a conclusion I agree with.  And I really can’t embrace meta-analyses when they are compilations of observational studies, which are themselves next to worthless.

For those who don’t know, meta-analyses are compilation studies in which researchers comb the medical literature for papers on a particular subject and then combine all the data  from the individual studies together into one large study.  This combining is often done to bring together a collection of studies, none of which contain data that has reached statistical significance, to see if the aggregate of all the data in the studies reaches statistical significance.  I think these types of meta-analyses are highly suspect, because they can lead to conclusions not warranted by the actual data.

To give you an example of what I mean, let’s assume that we have a study looking at a flipped coin.  If a researcher flips a coin 10 times and comes up with 6 heads and 4 tails, runs this through a program checking for statistical significance, he/she will discover that the 6-4 ratio isn’t a statistically-significant difference because of the low number of overall flips (10).  Now, let’s say that 50 researchers did the same kinds of study and some found that their coins came up heads 6 times out of 10 or 4 times out of 10, etc.  If a researcher then wants to ‘prove’ that heads comes up more times than tails on a coin flip, he/she can gather all the studies showing heads come up more times than tails, add them together in a meta-analysis and come up with 25 studies, each with 10 flips, showing that heads came up 63 percent of the time.  Now we’re talking 250 flips and we would probably reach statistical significance.  We know that over the long run a flipped coin is going to come up heads about 50 percent of the time and that the more the times it is flipped the more likely the number of heads will close in on the 50 percent figure.  But, the meta-analysis that selected the studies showing the 63 percent heads is statistically significant because the studies were cherry picked.

Researchers using meta-analyses set up selection criteria to pick which studies will be included in their final product, which leaves the door open for all kinds of mischief.  For example, let’s say a researcher wants to make the case that low-fat diets reduce cancer. He/she would create a set of criteria, do a literature search for all the studies that meet those criteria, then do a statistical analysis of all the data.  If the data demonstrate that low-fat diets are linked to lower rates of cancer to a statistically significant degree, the researchers submit their paper for publication.  But let’s say that when the data is crunched, it doesn’t show any such relationship?  It’s easy to go through all the studies and find which ones strongly show the opposite of what the researchers want to show and then figure out how to change the study-selection criteria in such a way as to keep those studies from being selected, run the whole process again, and repeat until enough studies are found to make the meta-analysis show the link between low-fat diets and lower rates of cancer.

Sad to say, this is often how it is done.  Which is why I don’t give a lot of credence to meta-analyses.

But having said all this, I’m still happy to see a researcher with the academic credentials of Ron Krauss coming out with a meta-analysis showing no correlation between saturated fat intake and cardiovascular disease risk.  And getting it published in the AJCN, probably the world’s most prestigious nutritional journal, no less.  It’s called putting your money where your mouth is.  Many academics whom I’ve spoken with admit that there is no correlation, but wouldn’t risk their academic reputations doing a meta-analysis to ‘prove’ it.

I’ve had many people tell me that it’s really nice to finally see some studies coming out vindicating saturated fats.  Or at least not attacking them.

I have to tell them that pro-saturated fat studies have been around for years.  Not just observational studies or meta-analyses, but real controlled studies looking at death rates from heart disease as a function of fat intake.

Let’s look at a couple.

Over 40 years ago, way back in 1965, there were two studies published showing that heart patients – the kind of people who today assiduously avoid saturated fat – who ate saturated fat were more likely to survive than those who didn’t.

One paper titled Low-Fat Diet in Myocardial Infarction, published in The Lancet, looked at the survival of subjects who had suffered heart attacks who went on either low-fat diets or their regular high-saturated-fat diets.

Here’s what they did:

264 men under the age of sixty-five, who had recently recovered from a first myocardial infarction and who had been in the Central Middlesex, Edgeware General, or West Middlesex hospitals took part in the trial.  On leaving hospital they were allocated at random to one of two groups at each hospital.  One group was placed on a low-fat diet, which the other group continued with their normal diet.

The trial, which ran from 1957 to 1963, was managed by four research medical registrars working at the three different hospitals.
What was the low-fat diet?

Patients in the diet group were allowed to take 40 g fat daily [under 20 % fat].  The daily allowance included 14 g (1/2 oz) butter, 84 g (3 oz) of meat, 1 egg, 56 g (2 oz) cottage cheese, and skimmed milk.  The nature of the fat consumed was not altered, nor were any additional unsaturated fats given.  The diet was often unpleasant, [my italics] and where possible, it was modified to suit individual tastes.

The body of the article states that the control subjects on their regular diet consumed about 2.5 times the fat eaten by those on the low-fat diet. (106-125 g for the former; 44-45 g for the latter.)  I ran the saturated fat calculations on the low-fat study diet and found that it contained about 30 g saturated fat, which is about 13.5 percent of total calories.  Most ‘experts’ today recommend keeping saturated fat under 10 percent of total calories.  Given how the data was presented in this paper, there was no way to tell how much saturated fat the control group got, but we can estimate their total fat intake to be about 46 percent, which was the average fat content of the typical American diet when I first got into this biz way back in the early 1980s just as the low-fat jihad was kicking off.  I would guess that the control diet contained 60-70 g of sat fat or about 25 percent of calories.  You can see the difference in fat intake in the graph above on the left.

The patients on the low-fat diet had pretty close counseling during the course of the multi-year study, and, consequently, they hewed fairly closely to their diet.  The researchers knew this because the study group consumed about 400 fewer calories per day as compared to those subjects on their regular diet and lost weight.  The researchers also used serum cholesterol levels as a measure of compliance to the diet.  In 1965 it was well known that reducing fat in the diet, especially saturated fat, made cholesterol levels go down.  As you can see from the chart on the right, cholesterol levels went down on the low-fat diet and stayed there.

What did the researchers find after observing these subjects for years?  They found that putting people on unpleasant low-fat diets didn’t help them live any longer nor avoid another heart attack.  Over the course of the study, the same number of subjects died in both groups.

What were the recommendations of the authors of the study?

It is concluded that in men under the age of sixty-five who have survived a first myocardial infarction, a low-fat diet does not improve their prognosis.

Summary

A controlled diet of a 40 g low-fat diet was carried out on 264 men who had survived a first infarction.  Despite a lowering of the blood-cholesterol and a greater fall in body-weight in the treated group, the relapse rate was not significantly different in the two groups.

A low-fat diet has no place in the treatment of myocardial infarction.

Ah, how things have changed since 1965.  And not for the better.

Here is another.

A paper published in the British Medical Journal (BMJ) in 1965 titled Corn Oil in Treatment of Ischemic Heart Disease looks at the differences in the rates of death or a second heart attack in patients following one of three diets: Their regular diet (control diet), a high-olive-oil diet, or a high-corn-oil diet.  After determining the caloric intake of the control group, the researchers had subjects in the other two groups restrict their intake of fat from foods as much as possible and replace it with supplements of either olive or corn oil in amounts calculated to match the calories they reduced by getting rid of animal fat.  The subjects getting one of the two oils ended up getting about 80 g per day.

The aims of the study were as follows:

Our purpose was to study the effects of prescribing a vegetable oil and a restricted fat diet to patients with ischaemic heart disease.  The primary interest was in an unsaturated oil with a cholesterol-lowering effect.  But large doses of any oil may have secondary effects on diet and nutrition, so that differences between an unsaturated-oil group and a control group might be due to these secondary effects rather than to unsaturated fatty acids as such.  It could, for example, be relevant that mortality from heart disease is low in Italy and Greece, whose inhabitants consume much olive oil; this oil has no major effect on serum cholesterol level, its main fatty acid (oleic acid) being only mono-unsaturated.  The trial was therefore designed to study the effects not only of a more highly unsaturated oil (corn oil) but also of olive oil.  It seemed likely that if any differences emerged between the olive-oil and corn-oil groups these would reflect the specific effects of polyunsaturated fatty acids.

After starting the diets to which they were randomized, the subjects were followed closely for two years.  As with the last paper, the researchers used serum cholesterol levels to monitor compliance with the diet.  You can see the differences in serum cholesterol in the three groups in the chart below.  Note that the cholesterol levels in the control group did not change a significant amount, which would be expected.  The same held true for the olive oil group: no significant change.  But those subjects in the corn-oil group dropped their cholesterol levels significantly.

Over the course of the study a number of patients died or had a second heart attack.  The researchers knew which subjects were on the control diets but were blinded (as were the subjects) and so didn’t know which were consuming the olive oil or the corn oil.

When the codes were broken and the data analyzed, it turned out that 75 percent of subjects following their standard high-fat, high-saturated-fat diets were remaining alive and free from a second heart attack whereas only 57 percent of subjects on the olive oil had done so.  The group with the worst outcome was the corn-oil group.  Only 52 percent of those subjects remained alive and heart-attack free.

The authors’ summary:

Eighty patients with ischaemic heart disease were allocated randomly to three treatment groups.  The first was a control group.  The second received a supplement of olive oil with restriction of animal fat.  The third received corn oil with restriction of animal fat.  The serum-cholesterol levels fell in the corn-oil group, but by the end of two years the proportions of patients remaining alive and free of reinfarction (fatal or non-fatal) were 75%, 57%, and 52% in the three groups respectively.

It was concluded that under the circumstances of this trial, corn oil cannot be recommended in the treatment of ischaemic heart disease.

In this same issue of the BMJ appeared an editorial about this study.  The author of this editorial points out that

the patients treated with corn oil had the worst experience, though initially their outlook was apparently similar to that of the other groups.  There is a 1-in10 to 1-in-20 chance that corn oil had a deleterious effect; the probability of its having any beneficial effect is remote.

This came at a time when corn oil was being touted on advertisements everywhere as the best oil to prevent heart disease because it is polyunsaturated.

The editorial goes on to grumble about the outcome and discusses a few other studies with conflicting outcomes.  The writer finally declares that maybe the problem is that this and other studies have been done on subjects who already have heart disease.  Maybe that’s too late in the game to make a difference.  (The outcome of this study wouldn’t indicate that, but the writer didn’t let that fact get in the way of his opining.)

Maybe it doesn’t help to lower cholesterol or increase polyunsaturated fats in those already afflicted; maybe what really needs to be done is to increase polyunsaturated fats and lower cholesterol levels in healthy people with no sign of heart disease.

A different approach, and a formidable one, is the prevention of ischaemic heart disease by altering the diet of healthy people.  A study of the organization of such a scheme in the U.S.A showed that it was practicable, and an anti-coronary club for men has been in existence in New York since 1957.  Its 814 members take a “prudent diet” in which fat is moderately reduced and equal proportions of saturated, monounsaturated, and polyunsaturated fats are eaten.  Already there is evidence  that the development of “coronary events” is being prevented.  Again, we await confirmatory evidence.

What the editorialist is waiting for is evidence to confirm his bias that reducing fat generally and saturated fat specifically (while increasing polyunsaturated fat) and the lowered cholesterol levels arising from such changes will prevent the development of heart disease.  Unfortunately, for him, this confirmatory evidence was not forthcoming.

From Gary Taubes’ Good Calories, Bad Calories (pg 36 hardcover):

Overweight Anti-Coronary Club members were prescribed a sixteen-hundred-calorie diet that consisted of less than 20 percent fat.

[It was reported] in February 1966 that the diet protected against heart disease.  Anti-Coronary Club members who remained on the prudent diet had only one-third the heart disease of controls.  The longer you stayed on the diet, the more you benefited, it was said.  But in November 1966, just nine months later, the Anti-Coronary Club investigators published a second article, revealing that twenty-six members of the club had died during the trial, compared with only six of the men whose diet had not been prudent.  Eight members of the club died from heart attacks, but none of the controls.

Like the maze shown at the top of this post, the people who have a bias against fat are trying to make things more complex than they are.  The simple solution is to look at the mortality, which no one wants to look at because it doesn’t confirm their bias.  They all want to look at more complex issues that have little bearing on the most important issue – whether one lives or dies.

Even the authors of the study showing the members of the Anti-Coronary Club members dying at enormously higher rates than non-members and dying with heart attacks want to look at other more complex information.

Gary Taubes continues

This [the deaths by heart attack of the club members] appeared “somewhat unusual,” Christake [the author of the paper] and his colleagues acknowledged.  They discussed the improvements in heart-disease risk factors (cholesterol, weight, and blood pressure decreased) and the significant reduction in debilitating illness “from new coronary heart disease,” but omitted further discussion of mortality.

Classic behavior from someone whose mind is made up.  Ignore the evidence denying your hypothesis and focus on that confirming it.  Instead of focusing on which people actually die of heart disease, let’s spend our time running through the maze looking at how our beloved low-fat diet reduces supposed risk factors. Which brings to mind a wonderful Winston Churchill quote:

However beautiful the strategy, you should occasionally look at the results.

How many people have died or been incapacitated with heart disease since 1965 when the evidence above was presented?  How many fathers, mothers, aunts, uncles, grandfathers and grandmothers could have had more years of productive lives if only the people who do these studies had looked at just the two mentioned above and taken the tack that maybe they had been going down the wrong path?  Had they done that instead of ignoring these results and continuing to try to prove an hypothesis that can’t be proven, how many lives might have been saved?  I’m glad it’s not on my conscience.

For maze at top
hat tip to
FAILblog.org

Please note: I reserve the right to delete comments that are offensive or off-topic.

Leave a Reply

Your email address will not be published. Required fields are marked *

188 thoughts on “Saturated fat and heart disease: studies old and new

  1. I was wondering if you were going to report on this study. Glad to see you have. I doubt seriously that the media will even mention it. The findings are not PC.

    • dr how much do i take of krill oil, I need to lower my “bad” cholestrol but my triglycerides are acceptable range and my total cholestrol is a little high just over 200. I have strong maternal and paternal history of family diabetes and heart disease. I am slightly overweight(working on it though) and I am female help?

  2. Your comments on meta-analyses reflect why you’re one of three people (Taubes and Robert Lustig are the other two) I trust to present evidence-based dietary advice.

  3. @Peter Silverman

    The only studies showing a relationship one way or another are observational studies, which don’t prove causality.

    @Amy Alkon

    Should my love life ever go South – God forbid! – you would be the ONLY one I would rely on for sound advice.

  4. ALERT! If you are looking for The Slow Cook from the ping back above, you will hit a dead end. That’s because the blog post that links to this piece won’t appear until tomorrow, Wednesday, Jan. 27. There was computer hiccup when I was scheduling the post, which sent out the inadvertant ping. My apologies. Check back tomorrow.

    Ed Bruske
    The Slow Cook

  5. There needs to be a day of reckoning for anyone in in any sort of position of Nutritional Authority-dom who is still recommending low fat/high carb diets in light of all the hard science available in this day and age. Heads of bodies like the AHA, ADA, need to hauled before a judge to explain why they continue to make recommendations that kill thousands.

    I call to order the Trial of Jane Brody!

    • Dr Eades,

      what proportion of the benefit from saturated fats do you suppose is due to the fact that, for many people, animal fat is their main, sometimes only, source of vitamin D and pre-formed retinol?
      Everybody ought to know how important vitamin D is these days; the amount in sat fats is low compared to sunlight and supplements, but compare it to vegetable fats, which lack any … plus, you need cholesterol to make vitamin D from sunlight, and low cholesterol seems to reduce vitamin D conversion (increasing cancer risk, etc).
      As for vitamin A, we’ve been told (by the usual suspects) that all we need is the carotene from half a carrot. This is wrong, and based on a mis-interpretation of the facts; beta-carotene from supplements can have a high rate of conversion, because it comes disolved in fats; beta-carotene from veges has a conversion rate of 12:1 – and increasingly reasearch is finding that many, perhaps most people don’t convert carotene to retinol at the advertised rate regardless of where it comes from (we don’t all have the genetic programming to do it, and may lack essential co-factors, such as zinc).
      Carotenoids are beneficial in many ways, but do not necesarily provide enough retinol for good health. The RDA is probably far too low, and we may not be getting it from veges in any case.
      Which may be why so many teenagers end up dependent on synthetic retinol analogues (toxic in ways retinyl palmitate isn’t – no-one has ever OD’ed on non-fortified cod liver oil). Pretend the vitamin is toxic and scare people away from it, then sell a drug based on the vitamin when the results of deficiency become obvious…
      I predict that retinol (as retinyl palmitate) will be the next Vitamin D.
      In any case, here are two good reasons to eat saturated fat, aside from the lipid – insulin science.

      • Vitamin D production is positively correlated to cholesterol levels (1). So, low cholesterol = low vitamin D. Low vitamin D is linked to increased cardiovascular disease.

        ——-
        1. Bogh, M. K. B., Schmedes, A. V., Philipsen, P. A., Thieden, E. & Wulf, H. C. Vitamin d production after uvb exposure depends on baseline vitamin d and total cholesterol but not on skin pigmentation. Journal of Investigative Dermatology 130, 546-553 (2009).
        2. Parker, J. et al. Levels of vitamin d and cardiometabolic disorders: Systematic review and meta-analysis. Maturitas (2009). URL http://dx.doi.org/10.1016/j.maturitas.2009.12.013.

  6. Did you mean to say low-carb in the first sentence? Or low-fat?

    Low-fat, of course. Thanks for the heads up. It’s now changed.

  7. I’m almost positive that the modern day low-fat group will conclude that the low fat group in the 1965 paper didn’t go low enough. The total cholesterol was still above 200. This would “prove” that everyone needs to be on a statin too. Argh!

  8. What amazes me (guess I’m too easily amazed) about the articles about the low carb / blood pressure reduction is that the writers and quoted doctors seem mystified by how a low carb diet could reduce BP.

  9. Doc, I just read a new book which backs up what you say in your books. “Fat and Cholesterol are Good for You” by Uffe Ravnskow, MD, PhD, published n 2009. Starting the 6-Week Cure today and am looking forward to a new belly. Thanks for being a beacon of light and hope in the still dark world of nutrition science. I’m lucky I have a doc who knows nutrition.

  10. @Michael Byrnes

    Yes, it is simple. Reduce carbs, reduce insulin, reduce sodium resorption in the kidneys, reduce blood volume, reduce blood pressure. Or reduce carbs, reduce insulin, reduce stiffness of the arterial wall, reduce blood pressure. Or both. It’s as simple as that.

  11. This is less a comment than a question.

    My wife, of 50+ years, has Type 2 diabetes, mostly under control – over the past 11 years she has generally kept her HbA1c results between 6.5 and 7.0. For the most part we follow the same diet and exercise regularly; we have a comfortable relationship together – as you might have assumed.

    However, she and I occasionally argue about what I’ve read in your books, and your blog, (and Gary Taubes book). Her source of information is varied but often includes the magazine her MD recommended, “Diabetes Self Management”(DS-M). The information in this and other publications she reads have convinced her that what is good for me, (the general public), does not necessarily apply to her, as a diabetic. This month, DS-M published an article by Shauna S. Roberts, PhD, supported the notion that the Mediterraenan diet, which Ms. Roberts characterizes as including eating meat and meat products sparingly, (emphasizing whole grains and monounsaturated versus saturated fats), “was associated with a lower HbA1c level and lower after-meal blood glucose levels in people with Type 2 diabetes.” (Diabetic Medicine, Vol 26, #9, Sept 2009, pgs 900-907)
    The second study Ms.Roberts cited involved 23,000 Greeks who did not have diabetes. ((BMJ online Vol 338, June 23, 2009 b2337) That study was not consistent with the other she cited in her article because it concluded that “the least important component (of the Mediterranean diet) (was eating few dairy products and eating lots of grains”. There were other caveats in the studies which Ms.Roberts acknowledged, and excused. Her conclusion:”Together these two studies confirm the beneficial effects of the Mediterranean diet. These studies show that following a Mediterranean diet prolongs life and improves diabetes control…”
    Here’s my question: Do you believe that the diet you advise is NOT suitable for Type 2 diabetics? or if suitable, still needs to be modified significantly? (Beyond the more-obvious modifications of alcohol and sugars…)

  12. @Andy DeGiralamo

    I believe the diet I recommend works spledidly to treat those with type II diabetes. It’s the diet I used successfully in my practice on numerous diabetic patients. As to the Mediterranean diet, a recent study shows that a low-carb Mediterranean diet is works better than either the ADA diet or the traditional Mediterranean diet to raise HDL levels and control blood sugar.

  13. How come nobody ever mentions the huge vegetable oils contain of omega-6 essential fatty acids (EFAs)?
    High consumption of vegetable oils disrupts the omega-3 to omega-6 balance in our diet. From a healthy 2 to 1 ratio the balance today is about 25 to 1. See: http://www.longevinst.org/data1.htm for the details

    We use EFAs to make prostaglandins. With a 25 to 1 EFAs balance, we make much more prostaglandins from omega-6 than from omega-3 EFAs.
    Omega-6 prostaglandins increase blood coagulation, inflammation and stimulate immune system reaction, while omega-3 prostaglandins maintain blood fluidity, reduce imflammation and immune reaction.
    The actual omega-6 to omega-3 ratio in our diet is responsible for undue coagulation, widespread inflammation, increase in allergic reaction and auto-immune diseases.
    Doc EDv

  14. When it’s all said and done, I’m certain that the ‘anti-fat’ crusade will be another example of ‘science gone awry’.

    I’ve watched for over a decade as people change lifestyles and take costly medications to chase lab numbers, convinced they will prevent heart attacks — all based on a stack of assumptions and inferences.

    I can only hope that this will all unravel while I’m still around to enjoy the spectacle.

  15. As a nurse that works for a cardiac-specialty home care company (doing mostly computer and regulatory work), it disgusts me that there is so much more that we can do for cardiac patients to make thier lives better, and so manythings we do that make thier lives worse. All we do is put them on a low fat, low cholesterol diet and give them statins (sometimes two at a time) and diuretics. Nobody gives a thought to the cognitive issues that the statins can cause, or that those with the higher (200-250) cholesterol live longer, or that we should be dosing these people with COQ-10 or ubiquinol, getting them off grains etc. Not to mention what we could be doing for thier diabetes… the list goes on.

    There’s no joy for me in nursing anymore. I hate belonging to a broken system that’s never going to get better.

    • I am also a Registered Nurse also and share your views. I worked in both cardiac and stroke areas and 2 years ago left the broken system and work as a wellness coach now promoting the lifestyle recommendations outlined on this blog. I couldn’t keep watching people getting worse either. Unfortunately, people seem to trust the USDA and ADA…which is interesting as our government has proven itself to be money driven again and again versus the best interest of the people.

  16. Hi Dr. Eades,

    I think the authors of the Rose et al corn oil study must have meant “There is a 9-in10 to 19-in-20 chance that corn oil had a deleterious effect; the probability of its having any beneficial effect is remote,” rather than “There is a 1-in10 to 1-in-20 chance that corn oil had a deleterious effect; the probability of its having any beneficial effect is remote.” I think the result was on the cusp of p < 0.05 or maybe even below it.

    I don't know why they didn't just crunch all the p values and publish them in the paper. It makes me wonder if the corn oil result was significant but they didn't want to openly acknowledge it.

  17. Actually the results from this study correlate quite well with the data so far seen in clinical trials of ezetimibe. If we think of ezetimibe as a medically enforced low fat diet the results thus far have trended in a negative direction. Enhance study no significant difference in plaque build up with ezetimibe added to the statin than the statin alone, that despite significanly lower LDL numbers; in fact the trend was in the wrong direction more plaque with ezetimibe.

  18. Notice the trend. Cholesterol goes down, more people die. Maybe we should try to increase cholesterol instead. I’m just saying.

    Thanks Dr Eades.

  19. I heard something very sad on the radio yesterday. A trial in Boston had begun over the death of a 4-year old. This child had been diagnosed bipolar at 2! Huh, Wuh?. She was prescribed POWERFUL drugs. I in no way espouse the views of Scientologists (they eschew all psychiatry), but what is going on? My first thought, after just feeling horrible for this poor child, was maybe she didn’t get enough FAT, FATTY FAT FAT FAT to eat. If she was fed a SAD and soy-based formula diet and cheerios, and juice, bread, skim milk, margarine blah, blah, blah, she may have had no chance from the start. But, a 2 year-old diagnosed with bipolar!!! When is this madness going to stop?

    • My parents did this to me as a third grader. I had a number of behavioral problems, and because of a family history of bipolar, they hauled me into a psych office and essentially forced them into the diagnosis. I was lucky enough that the psychiatrist was against medicating children. It’s bad enough when your family tells you you’re crazy, and you believe it. Being forced onto medication that dulls your entire existence would be even worse.

      I started getting really weird in high school and ended up on a ton of drugs that only kind of helped. I’m off those now, but some of the side effects are lasting. It looks like one of the real problems was shoddy parenting. The whole “low-fat crap” diet they fed me probably didn’t help, either (so starved for fat I ate margarine out of the tub, regularly). My docs think the weirdness that started in high school might be Kleine-Levin Syndrome (WTF, seriously), so we’re in the process of ruling everything else out.

      Unfortunately there are parents out there that see behavior they don’t like or expect and assume mental illness. As the kids don’t know how to speak for themselves when it comes to mental health, the parents have far too much pull. Every behavior of the child’s will be put through the filter of their choice, and that’s what the physician sees.

      Don’t even get me started on the kind of things that happen in child “therapy”. The treatment and understanding of abnormal psychology in adults is sketchy at best, and I don’t think children should be subjected to it at this point. Certainly not with medications.

      The madness will stop when the medical community takes off its blinders and starts treating their profession as a science instead of a pseudo-religion with public-health moral imperatives.

  20. To Andy DeGiralamo: I’m a female type II diabetic (diagnosed at 55) who was lucky enough to find Dr Richard Berntein’s Diabetic Solution and follow his low carb diet, along with Protein Power and Atkins. I take no diabetic medications and walk about an hour a day (no other exercise). I lost the 35 sudden pounds and was told I am no longer diabetic, am now pre-diabetic at just under 6 (although Dr Bernstein thinks under 5 is better). The only way to tell your personal glucose reaction to any particular food is to test with your glucose meter. (Now I’m hoping to get my husband lower carb to get off blood pressure medication.)
    Also Gary Taubes wrote a masterpiece.

  21. p.s. When I had “Diabetic Counselling” with nutritionists provided by unversal Canadian health care, we were offered a diabetic-suggested breakfast of toast and jam, cereal and milk, extremely strong coffee, and cheese. I had the cheese and was wired on the coffee for the rest of the session.

  22. My father is a pediatrician and is the smartest man I know (only thing he doesn’t know is how to get back the money he spent on my education). He takes it with darn near hostility when I suggest his cardiologist – who’s suggesting statins for my 72 year old dad with an LDL level of 130 – might be missing the more obvious step of reducing his carb and milk intake instead. Truthfully, I don’t know that the statins are a bad cost benefit analysis for him, but I can’t imagine a 130 LDL even begins to approach criteria for taking such a potent med. I’m surprised that my father, who’s as open minded about knowledge as anyone I know, would be so uncurious about a non-status quo perspective of the science of diet/health. Like Taubes said, the science has to be as simple as possible but not more. The LDL/low fat drums have been beating so loud and long that even my dad is under the spell. The greater irony is that in the 70s he told me “it’s the triglycerides that really matter.” My brother, an ER doc, seems as incurious, for example, and won’t read Good Calories Bad Calories. Frustrating.

  23. Dr. Mike,

    I believe you have recommended supplementing with vitamin b6 while on low-carb. Why is it that meat and other low carb fare won’t provide enough b6?

    I came across the suggestion that the biochemical process of converting protein into glucose uses draws on b6, so that a person who relies on protein for fuel could become deficient, get kidney stones, and so forth. Is there some other reason that you could get a b6 deficiency?

    Thanks

  24. I don’t understand. As Appolloswabbie points out in specific, drs. seem not to be curious about anything other than what they would like to believe. I thought one of the requirements for being a doctor was to be curious. Did I miss something here?

  25. To Andy,

    I am a 58 yo woman with T2 diabetes, diagnosed 5 yrs ago. Went LC (probably 50 – 60 gms/day) and got my glucose down quickly but no lower than the pre-diabetic range (6 – 7 A1c). Just couldn’t get it into the normal healthy range of less than 6 no matter what I did — exercise, eat less, lose weight, etc. Then a year ago, I learned that pre-diabetes is not low enough for truly good health as it is also associated with some complications and most crucial for us diabetics, REMAINING AT THE PRE-DIABETIC LEVEL DOES NOT PREVENT THE PROGRESSION OF THE DISEASE. This is due to continuing higher than normal glucose and insulin levels which continue to damage beta cells and will eventually wear them out… with or without meds. That’s why most docs believe that there is nothing a diabetic can do except slow the damage, not PREVENT it. But then they rarely see a diabetic not on meds get his A1c in the truly normal range so how would they know? Anyway, that’s when I went on a VERY LC diet — 20 – 30 grams carbs per day; almost exclusively green vegetables and some dairy (whole fat only). I increased my fats, especially saturated, to replace the carbs. My A1c began to drop down into the 5s right away. I do have some seasonal fluctuations, especially late autumn but it has not gone up over 5.6 for over a year now and lately has been 5.3 I test monthly with a very accurate home test (is spot on with my lab results). I hope your wife reads this and other similar testimonies everywhere online for LC. It’s the only way for a diabetic to overcome the effects of the disease. Keeping the A1c in the pre-diabetes range is not enough to escape it. She has to go on a VERY LC diet to stay healthy otherwise, as all docs say, it will catch up with her.

  26. “I really can’t embrace meta-analyses when they are compilations of observational studies, which are themselves next to worthless.”

    First, I have learned so much from you. Thanks.
    Second, having learned so much, you’ve got me thinking. That might be dangerous.

    Let’s do a thought experiment, where we are gods and we watch the little, lowly humans struggle. There is an imaginary human disease: let’s say blisters on the left elbow. Let’s say that this disease has a number of causes, one of them being the regular dietary consumption of rutabaga.

    Some human scientists adhere to the “rutabaga hypothesis.” Others use all their rhetorical skills to propagate the competing “turnip hypothesis,” arguing that turnips are a primary cause of left elbow blistering.

    Someone does an observational study and finds that there is no “association” between turnips and left elbow blistering. If there are a number of fairly selected additional observational studies showing no association, I think that that would qualify as reasonable evidence that turnips do not “cause” (in any simple sense of the word) blisters.

    On the other hand, if someone does an observational study and does find an association between rutabagas and blisters, that alone is not sufficient to conclude that rutabagas are a cause of blisters. This is because (1) the observations are neutral with respect to the direction of the causal arrow, it might be that blisters cause rutabaga cravings; and (2) there might be some third factor that is arbitrarily connected with rutabagas that is really causing the blisters, and the rutabagas have nothing to do with blisters. For example, it might be that Finnish immigrants to the US eat rutabagas in higher proportions than any other US community, and Finnish immigrants to the US also play a lot of left-elbow-rubbing sports. So it is their sports playing that causes the blisters, and not their rutabaga consumption.

    So, what can we conclude? It seems to me that we observational studies are relatively effective for disconfirming causal hypotheses, though they are not very helpful for confirming hypotheses. That is, if observational studies show no association between A and B, it’s hard to imagine (with the important caveat “other things being equal”) how either A could cause B, or B cause A. However, if observational studies do show an association, it tells us nothing about whether A is the cause or not. Yet again, if someone believes that A and B are causally related, there should be an observable association between A and B. If no association is observed, that is good evidence against a causal relation.

    Back to Krauss, cardiovascular disease, and saturated fat.

    The lipid hypothesis is that fat causes disease. Observational studies that find an association between elevated fat consumption and elevated disease can never legitimately confer a causal role to fat consumption. On the other hand, observational studies that fail to find an association between elevated fat consumption and elevated disease do legitimately serve to undermine the strength of any lipid hypothesis.

    1. If lipids cause disease, then lipids will be associated with disease.
    2. Lipids are not associated with disease.
    3. Therefore, lipids do not cause disease.

    The logical form of the above argument cannot be questioned. So, only the assumptions (1 and 2, above) are open to debate. Point 1 seems self-evident to me. Am I missing something here? Point 2 depends on empirical data. Krauss has just provided such data. So, it seems to me, Krauss has performed a useful service to us all.

  27. Another good post by Dr Eades. I’ve been spending some time in the Amazon health discussion and its really disturbing how anti-fat some folks are. Calling people tubs of lard and the reason they are fat is because they are lazy and eat gross amounts of food.

    • Hahahahaha. That’s just the tip of the iceberg. Go anywhere online where people can vent their true feelings, and you’ll get a bunch advocating death for the obese (because the gluttony and sloth of the obese drive up their health premiums and offend their precious, saintly eyes!) among other patently stupid things. The fat are some of the last targets that are socially acceptable to discriminate against without reservation.

      People are stupid, hateful creatures.

  28. The case you make against corn oil would seem to support the notion that meat animals fed a corn-based diet are more deleterious to health than animals fed a grass based diet. Can you shed any light on the anecdotal evidence that grass derived animals are better for human health, with respect to Omega 3 to 6 ratio, CLA content, etc, than corn fed? It would seem, based on this discussion, that both groups are superior to plant fats.

    Thank you for a fascinating post.

    • Grass-fed animals have a much lower fat content overall, with less percentage saturated fat, more CLA, and much more omega-3.

      http://www.eatwild.com/healthbenefits.htm goes over many of the differences.

      Pulled from somewhere down in the page:

      1. Lower in total fat
      2. Higher in beta-carotene
      3. Higher in vitamin E (alpha-tocopherol)
      4. Higher in the B-vitamins thiamin and riboflavin
      5. Higher in the minerals calcium, magnesium, and potassium
      6. Higher in total omega-3s
      7. A healthier ratio of omega-6 to omega-3 fatty acids (1.65 vs 4.84)
      8. Higher in CLA (cis-9 trans-11), a potential cancer fighter
      9. Higher in vaccenic acid (which can be transformed into CLA)
      10. Lower in the saturated fats linked with heart disease [we know this is BS but it’s a selling point)

      Also, if you get a good producer, it tastes MUCH better than grain-fed, in my opinion.

      • It’s also generally tougher than grain fed as well. Which makes it nice to have a Sous Vide Supreme so you can cook it for several hours to tenderize while keeping the steak medium rare. A shameless plug, I know. :-)

  29. I would not assign much weight to these studies, based on the information that you show.
    In the first paper, why did the people on the control diet have a 20% reduction in cholesterol, almost as much as the low-fat? This would cause me to question whether the control group did something that changed their risk as much as the low-fat diet group.
    With regard to the ingestion of 80g of corn vs. olive oil – there seem to be a lot of reports of a link between diets high in omega 6 fatty acids and ill health. However, what was the quality of the olive oil? Rancid or refined olive oil could have caused the olive oil to be nearly as bad as the corn oil.
    In the anti-coronary club, the group self-selected, i.e. the two groups were not assigned at random. Were the members of the club more likely to join because they had relatives, co-workers, or friends who had died early,and so were at a higher risk of mortality?
    I like Moises examples, but the studies have to be properly constructed before anything can be inferred. Also, the meta-analysis only makes a conclusion regarding levels of saturated fats, not total fat. I’m kind of glad that I was trying to eat a low-fat diet when I was also consuming high-trans-fat margarines and cooking with corn oil, back in the 80’s, so I had less of that stuff. And I regularly splurged on ice cream, chocolate and cheese when I was climbing mountains and could take in the extra calories – my fellow climbers observed that the marathon runners who ate high-carb diets did not adapt/perform as well.

    To Kab – thanks for sharing your success.

  30. thank God for that butter facebook link: I joined it immediately!

    slightly off topic I know: my latest dessert – usually right after eggs and beef, 1 can of organic coconut milk with 1 capful of vanilla extract and 1.5 tsp of ZSweet, fantastic – that keeps my carb count for the day at 12 g (according to the back of the Whole Foods organic coconut milk can)

  31. My wife and I are in week two of the six-week cure. We both lost 6-8 pounds in the first week, but have been putting weight back on (2-3 pounds) during the second week. We are folowing the plan religiously. My wife is getting a little doscouraged, so can you pass along any knowledge you have of instances where this has happened to others?

    We are sticking to the two shakes and one solid meal a day, and have not varied from the receipes and suggestions you’ve made in the book.

    Would increasing exercise levels help?
    Many thanks for your reply!

    • All I can think of is that you need to make sure that you are adhering to the recipe for the shakes to the letter. You can take a look here at how a couple of other people have made mistakes in their reading of the instructions. Increasing exercise is always good for a host of reasons, but it doesn’t really help with weight loss.

  32. Interesting dissection of the meta analysis methodology. I had to intervene with my Mom’s doctor trying to push more treatment on her for “high cholesterol” I reminded him she was 93, and just what was he trying to prevent?

  33. Wouldn’t it be interesting to find that the lower total cholesterol level observed in the old studies with low-fat diets or diets promoting polyunsaturates was reflective of a lowered HDL? We now know a higher HDL is healthful and protective, and aren’t we finding out that eating saturated fat actually supports a higher HDL?

  34. Dr Eades, would you care to comment on this post?

    http://adipo-insights.blogspot.com/2009/…rning.html

    It explains to me why I have to limit my calories to less than 1500 a day for low carb/no carb maintenance or I gain weight.

    I have not found a description of this physiological process in your books. Is it there? Is it valid?

    Thanks.

    • I couldn’t get the link to work, so I can’t comment. I’m traveling in Mexico right now and have spotty and extremely slow internet service so I don’t know if it’s the link or the internet.

      I’ve put up numerous posts on this subject on this blog over the past few years. You should be able to find them using the search function.

    • For the most part I would agree with Kendrick’s statement. Gary Taubes in GCBC listed an ancient study done on inpatients in a mental institution that showed a slight decrease in heart disease from reducing saturated fat. But as far as I know that is the only study. Others have shown just the opposite.

      Glad to see another Spiked reader. I put up this same Kendrick article on Twitter a day or two ago.

  35. I think these studies are fine. Why would they use rancid olive oil? The goal is to prove that olive oil saves lives (which it doesn’t, and this isn’t the only study saying so).

    As for a reduction in cholesterol for the control group, total cholesterol varies quite a bit, even day by day. Uffe Ravnskov has a table in his book “Saturated fat and cholesterol are good for you” (and also in The Cholesterol Myths, if you can find that), where he started eating a certain number of eggs then added one egg a day until he was up to 8 eggs a day, I believe. He analyzed his total cholesterol each day, The range was quite varied (and no, his cholesterol did not increase). He stated that was because the error in the test is high.

    Also, these people were followed closely for two years, which means that their fat intake would also be followed.

    • If the olive oil was purified as to be indistinguishable from the corn oil, there’s a chance that some of the unsaturated fats were damaged in both extraction (if high temperature) and the bleaching/deodorizing process. If that was the case, they wouldn’t have known, and it wouldn’t have been intentional.

      Oils purchased at the supermarket can (and tend to be) be partially rancid without color or odor effects thanks to processing. People shopping don’t give it a second thought, and most researchers don’t seem to, either.

      Either way, it can’t be known, and is just speculation.

  36. Dear Good Dr. Eades, A book recommendation for you. “The Other Brain” by R. Douglas Fields, PhD. It’s about the 85% of the cells in our brains that aren’t neurons, but glial cells. Our brains are fatty. The glia and axons (axons are in the body also, all 100,000 miles of them if stretched out and laid end to end) are fatty fat fat fat fat fat fatty fat. I just got the book. Can’t put it down.

    • @ Laurie
      not only are glial cells in the brain, relatives of these caretaker cells line all nerves and blood vessels – and the Hepatic Stellate Cells that become activated myofibroblasts in liver fibrosis, causing cirrhosis, are the liver’s glial cells (there are also stellate cells in the Pancreas; my bet is they will be found wherever there is circulation, that is, everywhere). The HSC glial cells, like the neural ones, respond to, produce, and metabolise neurotransmitters; this provides a mechanism whereby mood can impact on liver fibrosis.
      Insulin can activate hepatic stellate cells, to bring this discussion back on theme.

  37. @Andy DeGiralamo and Mediterranean diet in general

    All right, I’m down there, and my question has always been… whose smart idea was it that Mediterranean diet was all about whole grains, low fat and a few olives? Have people SEEN what guys over here eat? Has anyone really bothered? I’ve lived here all my life, and all I can say is that, for the first time I heard of fat free dairy was over the internet. The same for the mighty worshipped pork NOT being a staple in the rest of the world. 😛

    • People haven’t seen, or don’t want to see. It’s like the view that Americans are all loud, rude, obnoxious, and anti-intellectual, eat nothing but obvious junk food, and drive nothing but giant behemoths of vehicles.

      There’s a grain of truth in such things, but that grain of truth sprouts into a giant bush of silly, colored by the preconceptions of those who spread information.

      I live in the pork capital of the US, and while the meat is cheap, you really don’t see a lot of it, and most of what’s around is bland factory meat. It’s also depressingly lean. There are pigs everywhere, and no pork belly to be found, and no good prosciutto!

      Lack of good, properly cured Serrano and prosciutto hams is a crime against humanity. (FDA, I’m lookin’ at you)

  38. I’ve been low carbing for a few years now, and doing so really seriously for the past 6 months or so. I just got back from discussing screening labs with the doc. Basically there was no change over last year’s tests. Blood pressure 108/69 (normal for me and it causes no problems), fasting blood glucose 85, total cholesterol 169, HDL 62. My BMI is 29 and the doc had the good sense not even to mention it. I’ll be 58 next month, and I come from a family with a strong history of diabetes (types I and II) and a great deal of obesity including a couple of cases of clinically morbid obesity.

    Thanks, Dr. Eades.

  39. Hi Dr. Mike,

    My husband is concerned that the levels of phosphorus consumed from an adequate protein, moderate fat and restricted carbohydrate diet is going to short change him on the life of his kidneys. He read some article somewhere and now is trying to prove it isn’t so…

    I did mention that there is really no possible way for a study to be designed to watch a bunch of people for their whole lives and see if their kidney function is reduced by x-percentage because of all the variables, et cetera.

    He asked me to ask you of your knowledge on the subject of excess phosphorus and renal damage, if there is such a thing, and more importantly if following PP and PPLP will end his years early (as if all the stuff he did as a teenager won’t already :) ) Please let me know if there’s anything in PP or PPLP that will help, too.

    Cheers!
    Wanda

    • I would tell him that he doesn’t need to worry about it. He’s right in that there is no possible way for a study to be carried out in which people are randomized and followed for their entire lives. But, there are no such studies for whatever diet he is on now. There is no evidence that a PP type diet harms the kidneys, but if he doesn’t want to follow such a diet, far be it from me to force it on him.

      • Hi again, Dr. Mike,
        Just wanted you to know that Hubby actually does follow the plan, and was just trying to find an answer to allay his fears. He does like the plan, and does trust your advice, as do I. Nonetheless, I did a little more digging and did find out that the only way he would have kidney damage due to high phosphorus intake would be calcification of non-skeletal tissues ONLY if his renal function was already severely compromised, most likely with end-stage renal failure. I’m sure you wrote about it in PP, but I don’t have my copy yet (still shipping!).

        Thanks for all that you do!
        Wanda

  40. Two weeks ago I read Dr. Eades’s book “Protein Power.” For the last eleven days I have been as close to carb-free as it is possible to be.

    Eleven days ago my blood pressure was 131/91. This morning it was 109/69. The high-carb/vegan crowd has it all wrong.

    • Nice article. Thanks for sending. A few years ago I spoke with the author, Hannah Sutter, at length at a conference at which I gave a talk. We had an email correspondence for a while after. I’m glad to see her book is finally coming out.

  41. http://www.reachmd.com/xmsegment.aspx?sid=4806
    (about 10 min. audio targeting physicians; may require simple registration)

    I listened to this today and couldn’t believe what I was hearing. The expert, Lawrence Rudel, PhD, is listed as a professor and section head of pathology and lipid sciences at Wake Forest’s med school. He astonishingly recommends that our main source of dietary fats should come from vegetable oil!! He recommends avoiding olive oil and other monounsats, along with sat fats. Am I missing something here?

  42. I was always a bit dubious about saturated fats. It seemed that half the people are fine with it but half are cautious. I always tended to think that if it was bad it certainly was not as bad as other dietary factors such as omega 6, carbs etc. It would seem though that my concerns about saturated fat were unfounded.

  43. Off topic, sorry
    Hi, I’m reading ‘The Other Brain’ by Dr. Fields, it’s about the 85% of our FATTY brain cells that are not neurons, but glia. Fascinating.
    I started ruminating on the fact that there is a blood-gut barrier (tight junction control and regulation between the intestine and blood) that is made leaky by gluten. This is problematic to say the least. So I wondered about gluten and the blood-brain barrier and sure enough there is some speculation about this by a Dr.Kaslow. My interest is peaked for several reasons. I recently heard about non-celiac gluten sensitivity and so you don’t have to be diagnosed with celiac to have problems. Combine that with Lierre Keith of ‘The Vegetarian Myth’ alerting me to the fact that grains are addicting. AND we have zero requirement in human nutrition for grain and wheat….zero, zip, nada. Humans are adapted to eat grains in one and only one dramatic way…..it helps us (and has helped us for a very long time) reproduce like mad. The more we’ve reproduced the more people there have been; the more chance for genius. Up until about 100 years ago, natural forces kept our population in check.
    Anyway, I’m starting to see that the idea of an inviolable (to mischievous gluten) blood-brain barrier is a dangerous assumption.

  44. Dr Eades:

    As requested, here are the results of my recent VAP test completed after being on PP for about 90 days. First, here are baseline results from last August just before I started low-carbing from mid-October to mid-January: H: 5′-8″ W: 203 lbs; BP 130/80 HDL:47; LDL:150; TG:191; TC:235.

    The new me and new results are now: W:178 lbs; BP 120/75; HDL:68; LDL:118; TG:59; TC:202; LDL Size Pattern: A; ApoB100: 87; Lp(a): 10.

    Now, I knew there would be improvement because I feel better than ever – but I had no idea the improvement would be so dramatic!

    I’ve already seen the ‘answer’ to the moderately high Lp(a) in PPLP (eat more saturated fat!) so there’s some grass fed full cream in my future!

    Once again, I cant thank you and MD enough!!

    Paul

  45. Squire Hola…Does anyone know the definitive caloric ‘region’ whereby the body goes into starvation mode please ?

    • Depends upon whether one is on a low-fat diet or a low-carb diet. As I showed in this post on calories, subjects on a 1550 kcal low-fat diet starved about starved while those on a 1550 kcal low-carb diet didn’t. I would say, though, that in general, a diet of any length providing under 1000 kcal per day would put one in starvation mode.

  46. Pamala Brink posted a link to an article on fat storage can happen even on a very low carb diet. She gave the wrong link. Here is the correct link – http://adipo-insights.blogspot.com

    I myself an a low carber and have lost about 15 IBs in the last few weeks but this article does raise some very interesting points about how fat storage can occur through other mechanisms. I know it’s kind of unrelated to this blog post, but I thought I’d post the correct link since Pamala brought it up.
    Steve

  47. I’m not interested in losing weight, per se, but I’d love to see a post on what simple changes people can make to their diet and lifestyle to reduce CVD. All the searches lead to recommendations to reduce saturated fat intake, or even more outdated, reduce cholesterol consumed. A post from you would slide up the google search charts pretty quickly.

  48. I read the Perfect 10 diet by Dr. Michael Aziz,
    and it appears that it is the only book on the market that has the right information.

  49. Thank you so much for that. There’s definitely a book in it. Reducing CVD risk (and overall mortality) is a much bigger incentive to change than losing weight for most men, and the idea of reducing carbs rather than all the delicious fat and protein is very attractive. I would be sad to cut down on all the fruit I eat, though.

  50. Dr. Eades,

    Your “coin-flip” analogy is flawed.

    ****************
    If a researcher then wants to ‘prove’ that heads comes up more times than tails on a coin flip, he/she can gather all the studies showing heads come up more times than tails, add them together in a meta-analysis and come up with 25 studies, each with 10 flips, showing that heads came up 63 percent of the time.
    *********************

    This is not how meta-analysis works. A good meta-analysis does not select only the positive trials….it selects all trials that meet a particular selection criteria (i.e., RCT, certain methodology requirements, etc.).

    Studies in good meta-analyses are not chosen by their outcomes….they are chosen by their methodology. However, in your analogy, the coin-flip trials are being chosen by their outcomes.

    Second, most good meta-analyses do a formal analysis for publication bias, which helps indicate whether there is a bias in the literature for positive outcomes. If there is significant evidence of publication bias, then certainly, the results should be viewed with skepticism.

    ***************
    This combining is often done to bring together a collection of studies, none of which contain data that has reached statistical significance, to see if the aggregate of all the data in the studies reaches statistical significance.
    ******************

    This is also not true. There are many instances where studies included in a meta-analysis do have statistical significance. Sometimes meta-analyses are used to obtain an estimate of the magnitude of a treatment effect, particularly if the effect size seems to vary across studies. It can also be useful in situations where studies have conflicting results….i.e., some studies show statistical significance while others do not.

    *************
    But let’s say that when the data is crunched, it doesn’t show any such relationship? It’s easy to go through all the studies and find which ones strongly show the opposite of what the researchers want to show and then figure out how to change the study-selection criteria in such a way as to keep those studies from being selected, run the whole process again, and repeat until enough studies are found to make the meta-analysis show the link between low-fat diets and lower rates of cancer.

    Sad to say, this is often how it is done.
    *******************

    Dr. Eades, now you are making a blanket accusation statement of dishonesty among a lot of researchers who do meta-analyses. It’s also a statement for which you have little to no evidence to support.

    You could easily make the same blanket accusation for ANY scientific research, as data can always be manipulated in any scientific study. However, something tells me you would never make such accusations against researchers whose studies support your viewpoint. Your logic here implies that no scientific research should ever be trusted.

    The point is moot, anyway, because in a meta-analysis, someone who is familiar with the literature in a particular area will know what studies were not included in the analysis. If questions arise, that individual can easily write a letter to the publishing journal to bring this up. And if the publishing journal doesn’t publish the comment, the individual is always free to try to replicate the analysis.

    • My analogy is flawed? Maybe in a perfect world, but not necessarily in the real one.

      I suspect most authors of meta-analyses go into the project with an end in mind. I doubt they sit around and say, Gosh, I wonder if statins really work; let’s do a meta-analysis of all the studies out there to see. The academic world (as is the non-academic world) rife with bias and intellectual dishonesty. Countless papers are published that have been ghost authored by writers from the pharmaceutical and other industries. I don’t want for a minute to imply that all research funded by industry is bad, because it isn’t. Some is very good, but we see a skewed perspective because research supported by industry that shows results unfavorable to the supporting industry never sees the light of day.

      Observational studies are of value only in establishing hypotheses that can then be tested in RCTs. Observational studies are of no value in establishing causality. Meta-analyses of observational studies are, again, of value only in that they lay the foundation for establishing an hypothesis. Having said that, I do believe that observational studies showing no association between between two variables are a little stronger in showing no link than those observational studies that do show an association. The same no doubt holds for meta-analyses, but with the same caveat, i.e., that someone with an end in mind probably sets the parameters for which studies are included.

      • By the way, Dr.Eades, Kreiger is the guy who recently blogged about Gary Taubes, showed himself to be totally confused about what Taubes was saying, then kicked hell out of a straw man and declared victory. I’m sure he’s still congratulating himself………..

      • So I take you don’t accept that smoking is bad for your health? After all, we have few RCTs, if any, that show smoking increases your risk of cancer, heart disease, etc.

        • This is almost a reductio ad absurdum. Yes, the notion that smoking was problematic came from observational studies. And, no one wanted to do RCTs with people by randomizing them to either smoke or not smoke. But many studies were done in which animals were randomized into ‘smokers’ or ‘non-smokers.’ When ‘smoking’ animals manifested the same changes as seen in human smokers, it was a pretty easy case to build.

          But you raise a good point. Some observational studies are so strong that their findings have to at least strongly imply causality. For example, if I ran a camp in the middle of nowhere, and I noticed that everyone who drank from a certain well developed vomiting and diarrhea while those drinking from a different well didn’t, I would have done an observational study. But the correlation is so strong, that I could reasonably conclude that the water in the first well was the cause of the vomiting and diarrhea. If, however, three people (out of several hundred) got sick after drinking from the first well while only one got sick after drinking from the second well, then it’s time for an RCT.

      • In other words, Gary had no cogent rebuttal? Because so far, all I’ve seen from the low carb absolutists on James’ site are attempts to defend Taubes with straw men arguments and other fallacies. Taubes’ silence speaks volumes here.

          • Interesting comment, doc. Would you care to elaborate on why you think James is an “idiot”? I’m coming from a place of trying to get it right – not “be” right. I respect your intelligence – you are clearly a very intellectual and lucid fellow. For this reason, I am a little surprised that you would launch an ad-hominem rather than making an attempt to debunk something that Mr. Krieger said about Taubes’ research.

            Is it possible that you may have a wee bit of a confirmation bias when examining data and critiques?

        • No, you misread me. If someone – idiot or otherwise – decides to spend half a day coming up with a critique of one of my books or a blog post, why should I immediately leap into the fray and spend a half day or more laying out my own analysis of the critique of me? I have no obligation to do that. And just because I don’t doesn’t mean I can’t or that I’m afraid to. It just doesn’t make sense from a time-wasting perspective. I just spent a half a day writing up a critique of Anthony Colpo’s nonsense. Why? Because back when I was younger and stupider I promised to do it after he had attacked me in much the same way this guy has attacked Gary. Will it make any difference? Of course not. Colpo will blow it off with some unintelligible gibberish and all his sycophants will believe what he says because they can’t think for themselves. All the people who are my readers already probably believe I’m right and Colpo’s wrong, so what have I accomplished? I’m sure there are a few people out there on the fence who might come over to my perspective, but was it worth spending half a day – more like 3/4 of a day, actually – pulling studies and writing this up? I seriously doubt it.

          Gary is working on a book right now and always has a couple of articles in progress. He is a meticulous researcher who leaves no stone unturned when he reports on the issues he reports on. Plus he has a heavy lecture schedule – he’s lecturing at the NIH in a couple of days, in fact – along with his writing schedule, and gets paid for all of it. Why should he take productive time away from paying work to debate for free with anyone who should decide to put up a post critiquing him? He has these debates with academicians all the time in the Q & A after his talks. Why should he do it with every Tom, Dick and Harry who has a blog. For free.

          I don’t know what you do for a living, but I seriously doubt that you would provide a half or 3/4 of a day of it away just because someone you don’t know demanded it of you.

  51. Krieger sets up a straw man argument with the claim regarding
    Prediction 3. Fructose, a carbohydrate that generates very little change in blood sugar and very little insulin secretion, should result in less fat gain than glucose

    Listen to what Taubes says about fructose here, at slide 48 about 47 minutes into the talk.
    (Use Thumbs, Slides, tabs at top of page if you haven’t an hour for all the lecture, however the slides immediately preceding 48 set the context).
    Taubes Dartmouth lecture
    “Fructose (from sucrose and HFCS) is the carbohydrate converted most efficiently into glycerol-3-phosphate.
    One of several reasons why it is considered the most lipogenic carbohydrate”

    Before moving onto the next slide you hear Taubes say “It might be 90% of the problem”
    Krieger clearly hasn’t read or understood Good Calories Bad Calories pages 198~203

  52. *******************
    Maybe in a perfect world, but not necessarily in the real one.
    ******************

    But again, you have absolutely zero evidence that most people who do meta-analyses are doing what you claim they are doing, so you cannot state that your analogy works “in the real world.”

    **********************
    I suspect most authors of meta-analyses go into the project with an end in mind. I doubt they sit around and say, Gosh, I wonder if statins really work; let’s do a meta-analysis of all the studies out there to see.
    ************************

    Read what I wrote. A meta-analysis isn’t just performed to see if there’s a treatment effect. It may also be performed to estimate the magnitude of that treatment effect, even if an effect is known.

    And your criticism is absurd anyway. There are numerous meta-analyses that are done to see if something actually does work. Your statin example is a poor example….the fact that statins work was never in doubt (meta analyses on statins are done to estimate the magnitude of an effect as I mentioned before). However, for other areas where the evidence is not as consistent, then certainly researchers will do a meta analysis to get a better idea of what may be going on….particularly when you’re dealing with issues of statistical power and type II errors.

    *****************
    The academic world (as is the non-academic world) rife with bias and intellectual dishonesty.
    *********************

    Care to give evidence for that statement? How many papers would you consider as dishonest? Give me a percentage. Of course, I doubt you have any data to back such a number up….you will probably just make something up.

    Second, if academia is so rife with dishonesty, then why do you use academic papers to support your own viewpoints? You constantly cite research in your blog to support your views. I don’t see you accusing the researchers you cite as being dishonest. Or is the only “honest” research the one you happen to agree with?

    It is true that dishonesty is present in the world of academia, just as any profession. But that is not an excuse to write off an entire collection of studies (meta-analyses in this case). To do so is to be intellectually lazy. Every scientific paper should be dealt with on its own merits.

    ***************
    Some is very good, but we see a skewed perspective because research supported by industry that shows results unfavorable to the supporting industry never sees the light of day.
    *****************

    Doc, I could make the same argument about you. You have a vested financial interest in what you write, as you sell books and products related to a certain dietary philosophy. That means the perspective we get from you can be just as skewed as the perspective that one gets from research supported by industry.

    **********************
    Meta-analyses of observational studies are, again, of value only in that they lay the foundation for establishing an hypothesis.
    **********************

    I agree with this. However, you talk as if all meta-analyses are on observational studies, which isn’t even close to the truth. A large portion of meta-analyses are on RCT’s.

    I will respond to other people’s comments here at a later date when I have the opportunity.

    • Okay, I’m going to spend the time going through this exercise one more time. From here on, though, why don’t you just savage me on your BS Detective site and be done with it. If I feel the need to spend the time to respond, I will. This is one of these arguments that is never going anywhere. I certainly don’t mind having back and forth with readers who have legitimate differences of opinion with me, but I think there is a little more going on here than just a difference in opinion. These kinds of things have a way of turning into real time wasters, and I’m not interested in wasting a whole lot more time.

      First, you wrote:

      But again, you have absolutely zero evidence that most people who do meta-analyses are doing what you claim they are doing, so you cannot state that your analogy works “in the real world.”

      I would posit that you have as much factual support that they don’t work this way in the real world as I do that they do.

      Your statin example is a poor example….the fact that statins work was never in doubt …

      I would love to see the evidence of this. Not evidence that statins lower LDL, we all know they do that, but evidence that they decrease all-cause mortality in anyone other than men under 60 with diagnosed heart disease. No observational studies allowed.

      Would I care to give evidence that the academic world is rife with intellectual dishonesty and bias?!?! Are you kidding me? That’s what I spend most of my time on this site doing, displaying it for all to see. I don’t think every academician is dishonest and I do think there are some very good researchers and scientific papers out there. I have critiqued a number of papers that present conclusions I do agree with. The saturated fat vs CVD paper is a case in point. I wrote that I was glad someone had the balls to write a paper like this in today’s climate, but that it didn’t really mean a whole lot because it’s a meta-analysis of a bunch of observational studies. Here is one that I remembered off the top of my head, and here is another. I’m not going to take the time to go back through my countless posts to dredge up the ones I critiqued that had conclusions I totally agreed with, but you are more than welcome to.

      Doc, I could make the same argument about you. You have a vested financial interest in what you write, as you sell books and products related to a certain dietary philosophy. That means the perspective we get from you can be just as skewed as the perspective that one gets from research supported by industry.

      I agree. My perspective could indeed be just as skewed as any research supported by industry. But I came to my opinions based on many years of medical practice taking care of thousands of patients with obesity and other metabolic disorders. Through a lot of reading and a lot of trial and error, I learned what worked and what didn’t. I tried various therapies and evaluated countless lab results and had follow up visits with God only knows how many patients. Based on what I’ve seen and experienced first hand, I’ve got a pretty good idea whether or not what I read in a scientific study about subjects that I have great familiarity with makes sense or not. I came to most of these conclusions long before I ever started writing this blog, so if my thinking is skewed, it is skewed because of the 25 years of medical experience I’ve had. And I leave it up to my readers to decide.

      I agree that many meta-analyses are of RCTs. Often these meta-analyses are used to group studies that haven’t reached statistical significance together so that the whole of the data does. I’m somewhat skeptical of this because it leaves the door open for a lot of chicanery, but I’m not totally opposed to it if the selection criteria are valid. I’m a big fan of the Cochrane studies for this very reason. I have confidence that the people doing these meta-analyses are unbiased and are very, very careful in their selection criteria.

      Feel free to post as many comments as you want to post on the other people’s comments and I’ll be happy to put them up. But I’m out of this debate. I’m sure by now everyone pretty much knows where we both stand.

      • **************
        I would posit that you have as much factual support that they don’t work this way in the real world as I do that they do.
        *****************

        Shifting the burden of proof fallacy

        http://www.nizkor.org/features/fallacies/burden-of-proof.html

        The burden of proof lies on the individual making the positive claim (which is you in this case).

        There is a reason why we have “innocent until proven guilty” in a court of law. Apparently you prefer the “guilty until proven innocent” approach when it comes to your claim on meta-analyses.

        *******************
        but evidence that they decrease all-cause mortality in anyone other than men under 60 with diagnosed heart disease.
        *********************

        This is a red herring. I don’t really care about what statins do or don’t do. My point was that that there’s many other reasons for doing meta-analyses other than just finding out if something works.

        **********************
        But I came to my opinions based on many years of medical practice taking care of thousands of patients with obesity and other metabolic disorders.
        *************************

        I have no problem with that. In fact, I have been a supporter of low-carb dieting as an alternative to traditional approaches for a long time, including a scientific publication in AJCN that supports low-carb approaches.

        But it should be noted that my clinical experience, coming from a weight loss program that has been treating over 400 patients per year since the early 90’s, that we had outstanding results using an approach different from yours, with 40 lb weight losses in 3 months on average. Probably the only major thing in common with your approach is the high protein intake. But our carbohydrate intake was quite moderate and not even close to being a low-carbohydrate diet.

        I have no doubt low-carb dieting works. My problem is that the low-carb camp has moved from science to dogma, and really have become no better than the people on the low-fat end from the 1980’s – 1990’s. In the “fight against the establishment”, it seems that the people fighting the establishment have become no better than the establishment. This includes selective quotation of research and confirmation bias, demonization of a single macronutrient, creation of false dichomoties surrounding many issues, along with the many other fallacies that the establishment was guilty of during the low-fat, high-carb craze.

        The fact is, as long as protein intake is high, there are many dietary approaches that can work well

        • Mr Krieger.

          “Innocent until proven guilty” is a concept of criminal law, and and the “burden (weight) of evidence” for the claim (of committing a crime) should be “beyond all reasonable doubt” in eyes of (at least) 9 of a 10 person jury.

          In civil jurisprudence, it is the weight of evidence, which, “on balance of probabilities” favours the claim in the eyes of a judge, who decides the proposition that claim made should be upheld.

          In both trial situations the claimant makes a sanction – jail time or financial restitution – on the defendant.

          MREades does not make a “claim” to which a defendant raises a “defence” and which he must prove before some tribunal.

          He has merely postulated something about numbers, which action is a statement of his belief, and neither needs to be supported by third parties, nor sanctioned by them.

          Talk about a straw man!

          • Desmondo,

            When someone asserts that “most” people who engage in meta-analysis are engaging in data manipulation, then this is an assertion that requires evidence to support.

            The term “most” is the key term here, as it certainly implies that 70-90% of individuals who publish meta-analyses are manipulating their data. Logic dictates that extraordinary claims require extraordinary evidence….something that Eades does not have for this particular assertion.

            A postulate is something that is considered self-evident. However, there is absolutely no basis to consider such an assertion as self-evident. Thus, my statement about the burden of proof being on Eades still stands.

        • @ Krieger

          you say “during the low-fat high carb craze” as if it is all over now and we can all go home safely to our beds.
          But every time I read the paper or go shopping I see the same “saturated fat is fattening” dogma being milked for all it is worth (a lot, apparently) and spouted as proven fact by supposed experts.
          Whenever a food is marketed as “low fat” it has more carbs (usually from sugar) to compensate; and most “cholesterol free” products are overloaded with PUFAs.
          Whatever the slimming industry might be up to now, the situation out on the streets is worse, not better.

        • JK said:

          “There is a reason why we have “innocent until proven guilty” in a court of law. Apparently you prefer the “guilty until proven innocent” approach when it comes to your claim on meta-analyses.”

          Yes or no James – “All researchers are honest.” Ever heard of Ancel Keys?

          The burden of proof does not lie with anyone when we already know that the proof exists. If I say that getting hit by a car at 80 MPH will injure you, I don’t need to prove it.

          JK said:

          “This is a red herring. I don’t really care about what statins do or don’t do. My point was that that there’s many other reasons for doing meta-analyses other than just finding out if something works.”

          You’re only calling it a red herring because he nailed you. You did indeed imply that statins work to reduce all-cause mortality in those who take them. Take it on the chin like a man James.

          JK said:

          “But our carbohydrate intake was quite moderate and not even close to being a low-carbohydrate diet.”

          So what? I’ll bet you in every case you saw in your clinic where fat loss occurred, the clients/subjects were eating far less carbs than they were when they were eating ad libitum.
          You think it was the calories and not the decrease in carbohydrates. But you can’t prove that can you since total carbohydrates were also reduced.

          JK said:

          “The fact is, as long as protein intake is high, there are many dietary approaches that can work well.”

          Shallow thinking. Not stupid, just not deep. What does this mean really? Sounds like your done thinking. Case closed.

          And what is ‘high protein’ exactly?

          You suggest here that if two groups of people ate the same amount of protein with one group eating the rest of their intake as fat and the other pure starch with total calories being the same – there would be no difference in body composition after say a 12 month trial.

          You are what your body does with what you eat James.

          This will help you get it.

          http://www.youtube.com/watch?v=mNYlIcXynwE

          • *******************************
            Yes or no James – “All researchers are honest.”
            ***********************************

            Just because some researchers may be dishonest, doesn’t mean most of them are.

            Fred, you constantly demonstrate how you are unable to engage in any type of thinking other than dichotomies.

            *********************************
            The burden of proof does not lie with anyone when we already know that the proof exists. If I say that getting hit by a car at 80 MPH will injure you, I don’t need to prove it.
            *************************************

            That is an idiotic analogy. There is dramatic empirical evidence that getting hit by a car at 80 mph will injure you. Not only that, if I asked for the evidence, you could easily provide it. You could easily obtain data on car accidents of this type and look at injury rates. You could also provide data from physics, the structure of the human body, and other facts to show how the body could not withstand such forces without experiencing some type of injury.

            However, claiming that 70-90% of people who perform meta-analyses are manipulating their data requires empirical evidence….evidence that I know Eades does not have nor can he obtain. Thus, it is an unsupportable claim.

            *********************
            You’re only calling it a red herring because he nailed you
            ************************

            No, he did not “nail me.” I caught him in an assertion that he could not support with empirical data. I asked him to provide such data and he talked about how research is rife with dishonesty. But that still does not support his assertion.

            Washington D.C. is rife with crime. So does that mean that most people who live in Washington D.C. are criminals? Of course not.

            **********************************
            . You did indeed imply that statins work to reduce all-cause mortality in those who take them.
            ***************************************

            You obviously need to read posts closer. Eades made a comment how a meta-analyst might say, “Hey, I wonder if statins work….let’s do a meta-analysis to see.” I replied that they “work” was never in doubt (and when I say “work”, I meant reduce LDL)….I NEVER made any claims about mortality…..I simply stated that a meta-analysis might be done to estimate the magnitude of a treatment effect.

            It was Eades who made the statement about all-cause mortality….I never made any comments about all-cause mortality in any of my posts before that, nor did I even imply anything about it.

            ************************
            So what? I’ll bet you in every case you saw in your clinic where fat loss occurred, the clients/subjects were eating far less carbs than they were when they were eating ad libitum.
            You think it was the calories and not the decrease in carbohydrates.
            *****************************

            I think it was a combination of things, which includes protein, carbohydrate, fat, calories, education, stress reduction, sleep improvement, accountability, increased physical activity (not just exercise), etc. I’m not stuck in the “it’s all about the carbs” mindset like you are.

            *********************
            Shallow thinking. Not stupid, just not deep. What does this mean really? Sounds like your done thinking. Case closed.
            *************************

            The only shallow thinking is your “it’s all about the carbs” mentality.

            *************************
            And what is ‘high protein’ exactly?
            ****************************

            It depends on the individual, but typically we had our clients ingesting probably around 1.6 – 1.8 grams per kilogram, although I don’t recall exactly….we were in line with Dr. Donald Layman’s recommendations.

            *************************
            You suggest here that if two groups of people ate the same amount of protein with one group eating the rest of their intake as fat and the other pure starch with total calories being the same – there would be no difference in body composition after say a 12 month trial.
            ***************************

            No, I said there are “many” approaches that will work, not “any” approach that will work. Maybe you need to stop reading at such a shallow level.

  53. Research can be undermined by ”questionable research practices” that go “beyond falsification, fabrication and plagiarism” (Martinson, Anderson & De Vries, 2005). A survey that collected empirical evidence on questionable research practices found the top questionable behaviours by research scientists were ‘overlooking others’ use of flawed data or questionable interpretation of data, changing the design, methodology or results of a study in response to pressure from a funding source, dropping observations or data points from analyses based on a gut feeling that they were inaccurate and inadequate record keeping related to research projects”. These behaviours do not fall within the “narrow definition of misconduct” (Martinson et al., 2005) but should be addressed.

    • What bugs me about meta-studies is that apples and pears are added together, and conclusions about plums are published.
      It is very rare to find two studies that use the same treatment in the same dose with the same type of population under the same conditions.
      Usually this only happens when a) the same researchers try to repeat a study to confirm the result, or
      b) an extremely conscientious second research group tries to repeat the first experiment.
      You are really unlikely to find more than 3 or 4 studies that can truly be said to be testing the same hypothesis.
      I can’t help thinking it a sign of intellectual disrespect to combine other people’s studies for a purpose which was not intended by the designer of any of them.
      I’ve read a couple of complaints along the lines of “my study didn’t say that at all” online, and that was from contributors to studies used as raw material in two meta-studies (warfarin and antioxidants) by Cochrane, which dr Eades considers the most reliable provider; if that’s really so, then the allegation of “most” meta-studies being manipulated doesn’t seem such an exaggeration.
      (there was also a discussion of what looked like a rather serious undisclosed conflict of interests in a Cochrane study; it seems that Cochrane uses an honour system, and doesn’t actually check for them. If I was in pharmaceuticals, I’d want to get my boy onto a Cochrane job).
      Whatever the credibility of a meta-study, it can’t tell you what you’d learn from reading the original papers and making your own assessment. The devil is in the details.

        • I can’t help feeling that, where the point of combining data into a meta-study is to push insignificant results into statistical significance, then that is manipulation of data prima facie, and it is only excusable where the experiments used are truly identical in all meaningful ways.

          • George,

            The combination of non-significant studies to show significant results is not data manipulation. It is to address the problem of statistical power.

            So many people think that, if I have 10 studies that don’t show a significant difference, then that must mean there really is no difference. But if all 10 studies are underpowered to detect the difference in the first place, then a meta-analysis is useful to investigate whether there’s a trend or not. Barring the issue of publication bias, then if there really is no difference, then we would expect that there would be no trend among all the studies. But if there is, it at least suggests that any future research in the area needs to dramatically increase subject numbers.

          • I specifically wrote that “If we did a meta-analysis of these studies, we might find that adding the subjects together would end up showing a significantly difference in weight loss,” which is a far cry from actually combining the studies and trying to assert that they had reached statistical significance. I made it clear that these studies had NOT reached statistical significance as they were. But I did make the case that had these studies contained a larger number of subjects and still shown the same weight loss difference, these wt loss differentials may well have reached statistical significance. And I was also making the point that true metabolic ward studies are extremely expensive to do, and, consequently, usually involve few subjects making statistical significance difficult to reach.

          • Doc,

            In that post, you’re only 1 statistical analysis away from a rudimentary meta analysis….I wouldn’t call that a “far cry.” And essentially you weren’t doing anything that much different from what is done in some meta-analyses….looking at a body of studies and seeing if there is a trend among them. The only difference is that a meta-analysis uses formal statistical techniques to see if that trend is statistically significant. Otherwise, why would you have even bothered posting that information in a blog where you’re refuting Anthony Colpo’s assertions on the lack of metabolic advantage? You obviously think there is a trend there….you even state it as such in the post.

          • Were I to do a meta-analysis of this data, I would be doing precisely what I accuse others of doing: Coming to the literature with an hypothesis in mind in advance and picking the studies that confirm it. I used these specific studies because they were the ones Colpo proffered to bolster his hypothesis. I was merely showing how these same studies could argue for the opposite conclusion than the one he was using them to ‘prove.’

          • Definition of “most”:

            a great majority of; nearly all

            That’s means a helluva lot more than 51%

            Even if it was 51%, that’s still a number that requires evidence to support, as it’s at least 1 in 2.

            And yes, I’ve read it cover to cover. I still don’t see it like you do.

          • If you own 51% of a sports team do you own most of the shares? You certainly do.

            You don’t see that Gary’s hypothesis is not a valid explanation for how and why we store excess body fat?

            And I don’t “see it” in any particular way other than how it really is. You are still mired in conventional thinking.

      • ***************
        It is very rare to find two studies that use the same treatment in the same dose with the same type of population under the same conditions.
        *******************

        This is actually not a problem when it comes to meta-analysis. With advances in computing power and modern statistics, there are sophisticated statistical models that can account for the fact that no two studies are exactly alike. These statistical models incorporate the variance between studies into the levels of uncertainty in the final estimates.

        ******************
        You are really unlikely to find more than 3 or 4 studies that can truly be said to be testing the same hypothesis.
        ********************

        It’s not as hard as you think. I recently published a paper that included 14 studies that all essentially tested the same hypothesis.

        On top of that, the papers don’t need to test the same hypothesis (although they should be close). If I’m trying to estimate the magnitude of a treatment effect, not every study needs to compare the treatment to a control. I can actually gain more precision in the estimate with studies that compare the treatment to another treatment, for example.

        ***************
        I’ve read a couple of complaints along the lines of “my study didn’t say that at all” online, and that was from contributors to studies used as raw material in two meta-studies (warfarin and antioxidants) by Cochrane, which dr Eades considers the most reliable provider; if that’s really so, then the allegation of “most” meta-studies being manipulated doesn’t seem such an exaggeration.
        ******************

        This is a hasty generalization. I really don’t see how a “couple” complaints translates to “most” meta-analyses being manipulated.

        • The latter point is fair comment so far as that goes; however, I wasn’t searching for either posting when I stumbled on it, and suspect they are the tip of an iceberg. That meta-studies are considered controversial (to say the least) by the very people whose work is being relied on ought to give pause for thought. This is different from people outside a study disputing its findings.
          I imagine that where a drug is being tested in a certain dose for a certain diagnosed condition, it is possible to combine many studies. And I agree that there is no need for a control if you have good statistics about the performance of other medicines, or the normal course of an untreated disease (the “placebo effect” is just the tribute science pays to magic; show me any serious metabolic condition that can be significantly improved, long-term, by placebo pills). Any error that might creep in is probably no greater than that between disparate controlled studies.

          I really wonder where the need for meta-studies is supposed to exist. Anyone interested in the subject can read the real studies and make their own synthesis. If they exist to convince the great uneducated masses that a consensus exists, then they are surely prime targets for manipulation.

  54. If you wanted a rough estimate of what fats to include in your diet, you could do worse than look at the proportion of fat already in the human body. This includes many fatty acids the body has made itself (at a cost in nutrients and energy) to get the balance right.
    There is about 40-50% monounsaturated (mostly oleic), abbout 30% saturated (mostly palmitic, some stearic), the rest (about 10%) is omega 6 and 3 PUFAs (and a trace amount of short-chain fats). The turn-over of the PUFAs used in ecosanoid synthesis is probably greater than that of the more stable fats, and this does represent depot fat (but it is not that different from milk fat, for example).
    the point is, that if you do not consume palmitic and stearic saturated fatty acids in your diet, you have to synthesise them, and you could be using the enzymes instead to synthesise rarer FAs, for example, making EPA and DHA from ALA.

    • (I’m not saying that synthesis of palmitic acid uses the same elongation and desaturation enzymes as that of EPA and DHA, but that both elongation processes use the 2-carbon product of the citric acid cycle, and that different elongation enzymes may use the same co-enzymes and co-factors)

        • You are quite right. Really I am only sure of the involvement of the citric acid cycle, and of the biological need for palmitic and stearic acid.
          Interestingly some of the vegetable oils that are thought to be most healthy – the olive oils and rice bran oils – are relatively high in saturated fats; virgin olive oil about 14%, rice bran oil 21%, about the same as milk chocolate.

  55. I have before me a Cochrane meta-study from 2005 about increased mortality from high-dose vitamin E supplementation http://www.annals.org/content/142/1/37.full
    This is a subject I’ve a personal interest in, supplementing about 1,200iu daily; as you can see from my comments on the vitamin D debate page, I’m open minded that some antioxidants may have adverse interactions.
    So I really wanted to understand what this study was saying. Are the negative results due to the (inexplicable) use of dl-tocopherol by some researchers? Unfortunately the meta-study doesn’t specify which studies used racemic tocopherol, in fact states their equivalence for purposes of the study. Perhaps positive studies were excluded? There is a clue in the exclusion criterea; studies with fewer than 10 deaths were excluded. Obviously, the better the result from vit E, the fewer deaths there are going to be in a study, all else being equal, so positive studies were more likely to be excluded. (The larger all-antioxidant Cochrane meta study excluded studies with no deaths, which is better, but both criterea will inevitably push non-significant results into significance, making a nonsense of the published “percentage” results).
    One problem is that the Cochrane governing body cannot agree on exclusion criterea a priori, so leaves them up to the editor(s) of each study, just as they leave any disclosure of financial ties to the pharmaceutical industry up to the personal conscience of contributors. Both of these are matters of oversight that I believe Cochrane has to address to deserve the reputation they seem to have acquired.
    The Vitamin E meta-study didn’t give me the information that would have helped narrow down the likely reasons for the cited effect. The effect itself was small (0.35% increased mortality from all causes). None of the studies was life-long, and you can be sure of an effect on mortality when everyone is dead; some people may live shorter lives, others may well live longer lives, and the sum total may be towards longer lives with vitamin E (I actually suspect this may be the case, and it is interesting that the longest running population studies, the Nurses and Doctors studies, do show marked health benefits from vitamin E).
    My last complaint had to do with the conclusions;
    Conclusion: High-dosage (≥400 IU/d) vitamin E supplements may increase all-cause mortality and should be avoided.
    You see, none of the contributing studies could have drawn this conclusion. Commonsense would say that Vitamin E supplementation should be used with due caution; that other factors should be considered; that treatment and dose should be adjusted depending on early results, and that the best (most natural) form of the vitamin should be used. These are problems that relate to the nature of cotrolled clinical trials, not just meta-studies; if people use a vitamin as a drug monotherapy, and lock patients into a fixed dose long-term (sick patients, who really deserve better care), is it any wonder the results are not so great?

  56. Doc,

    I don’t know whats more impressive, that you wrote a book as universally praised by experts and laymen as PPLP, or the restraint and grace you showed in handling the hooligans here.

    I tried to read the your friend’ study on the thermodynamic case for a metabolic advantage but my brain locked up. Thankfully, I cling to an evolutionary proof of a metabolic advantage that suits me well enough. Basically, my thinking is that even if a calorie equaled a calorie equaled a calorie 40,000 years ago, adaption would have created an advantage since then. We would have, at a minimum, created a metabolic disadvantage to sugar, which would only be around in large quantities around the time of year when extra fat stores would ensure our survival. (“Lights outs: sleep sugar and survival” is a jaunty look at at that that and many other topics).

    And if sugar is disadvantaged, that which isn’t sugar is, um, advantaged. I know, I know, I too am a towering colporeigarian intellect and thus my logic is unassailable, but I’m glad to see that the math backs me up. Or rather, I’m glad to hear that the math back me up.

    Keep up the good work.

  57. Dr Mike

    I can’t pretend that I can follow the argument that Krieger has set up, and I really don’t care since low carb is working for me. In the very first sentence he claims that Taubes is only a journalist and not a scientist. I believe your friend and colleague Taubes has studied applied physics at Harvard and aerospace engineering at Stanford and has a masters in journalism from Columbia, and that his entire career has been in science journalism. He is the only journalist to be awarded the Science in Society Journalism Award three times. And was also awarded an MIT Knight Science Journalism Fellowship in for 96-97.

    I will take Taubes’ word and work over Krieger’s anytime. I find it amusing that Krieger criticizes Taubes’ credentials for writing, when his own are somewhat lacking. They simply are not in the same league as Taubes.

    To reiterate, I just don’t care. I have become more and more sensitive to carbohydrates with each passing year, and the calories in-calories out, energy thingy, fructose vs glucose, insulin-or-not stuff doesn’t matter. I have to avoid carbs as much as possible.

    I know that you have had twenty-five years of success with patients. I don’t know that four hundred patients in an unnamed obesity program losing 40 lbs in three months means anything. I too, have lost 40 lbs in three months in years past, and put it all back with additional pounds. When any given 400 patients in any given obesity program can keep off the weight they’ve lost for five years, then I’ll consider taking the advice from that obesity program.

    Until then, I am losing weight very, very slowly right now, I know that for me, a very low carb, super slow weight training program is what gives me the best chance to achieve and maintain a normal weight for the rest of my life.

    By the way, Dr Mike, I’m 5′ 2″ and leg pressing 552 lbs.–not bad, don’t you think?

    • LCforevah,

      The number of awards someone has been given in regards to journalism is irrelevant to whether a book on nutrition is scientifically complete or accurate.

      And whether Taubes has studied physics or aerospace engineering is also irrelevant. That’s like saying someone with a PhD in computer science has the background and knowledge to write a scientifically complete book on nutrition.

      Now, I would agree with you that I was not being completely fair in criticizing Taubes’s background in journalism. To do so is to simply engage in an ad hominem of my own, which is a fallacy. Taubes’s background is irrelevant to whether his thesis is correct, and I should’ve just stuck with arguing the evidence (or lack thereof).

      On the same token, Taubes could have three PhD’s in nutrition and biochemistry-related fields and it would be irrelevant to whether his thesis is correct. To argue based on someone’s credentials would be to commit the appeal to authority fallacy.

      If you’re going to debate on the grounds of how many awards someone has received, then that would imply that George Bray (who has criticized Taubes) should be given more credence than Taubes, as Bray has more awards and more publications.

      Now, I’m not saying that I agree with Bray….all I’m pointing out is that someone’s credentials are irrelevant to the validity of their argument.

  58. Krieger, you implied in the first sentence of your post that Taubes was “only” a journalist, skipping his science background. That was sloppy on your part; if that happened in the first sentence, there are probably other things that you were sloppy about in the post. Given that I have read the book from cover to cover, I, and others commenting here, can tell that you have not.

    Agreed about appealing to authority. I was pointing out in a nice way that his credentials are better than yours. It’s more about having the kind of critical thinking that you intimated Taubes hasn’t. Taubes did his research concerning the causes of obesity across many disciplines, not just nutrition.

    Readers on this site followed for years Dr Eades’ posts on Taubes work. We have a pretty good idea of what it took. I think your dismissive attitude about both his research and his conclusions isn’t based on facts.

    • ********************
      Given that I have read the book from cover to cover, I, and others commenting here, can tell that you have not.
      **************************

      Why? You think that just because I don’t automatically fall for Taubes’s arguments that it suddenly means I haven’t read the book? You guys make strawmen out of the arguments I’m making, so how can you claim that I haven’t read the book.

      I think it’s hilarious how anybody who happens to think Taubes is grossly incomplete in his presentation of the data must “not have read the book.”

      • James you said:

        “Why? You think that just because I don’t automatically fall for Taubes’s arguments that it suddenly means I haven’t read the book? You guys make strawmen out of the arguments I’m making, so how can you claim that I haven’t read the book. I think it’s hilarious how anybody who happens to think Taubes is grossly incomplete in his presentation of the data must “not have read the book.”

        James do you realize that you accuse virtually everyone who nails you of creating a ‘strawman’ argument? Why don’t you say something a tad more productive for a change and write for us here the three main errors Taubes makes in GCBC. Please cite the chapters.

        And DID you read GCBC in its entirety? A yes or no will suffice. FWIW, I’ve read it twice and am currently on a third turn. The first 200-300 pages are a masterpiece. To belittle the book as you do indicates quite strongly that you have not read it. And if you have, you need to read it again like so many of us have.

        And where did Dr. Eades say that 70-80% of all researchers lie about their research?

        • Fred, it doesn’t surprise me that you think GCBC is a masterpiece. And you’re only proving my point in my previous statement…you just cannot believe that someone could read the book and not see it as you see it. Maybe it’s because I’m a helluva lot more versed in the obesity-related literature than you are, so I’m not nearly as easily convinced by well-told stories like you.

          And if you think the term “most” means anything less than 70% or so, then you must be speaking a different version of the English language from everyone else

          • Wow, at least 3 basic logical fallacies in 1 paragraph. Continue this way and people will start to take you seriously, not!

          • “Fred, it doesn’t surprise me that you think GCBC is a masterpiece.”

            ****Strawman. I didn’t say GCBC was a masterpiece. I said the first 200-300 pages were.

            “And you’re only proving my point in my previous statement…you just cannot believe that someone could read the book and not see it as you see it. Maybe it’s because I’m a helluva lot more versed in the obesity-related literature than you are, so I’m not nearly as easily convinced by well-told stories like you.”

            *****Where is your evidence that you are a lot more versed in obesity-related research than I am? And did you read GCBC cover to cover, yes or no, James? You have evaded this question several times. I KNOW you did not. And I know you won’t lie about it.

            “And if you think the term “most” means anything less than 70% or so, then you must be speaking a different version of the English language from everyone else.”

            ****What English dictionary do you use? I’d give it the boot if I were you. 51% of anything can be considered most of something or the majority. 70% would be considered ‘the vast majority.’

    • *****************
      Taubes did his research concerning the causes of obesity across many disciplines, not just nutrition.
      *******************

      I wish I could believe that. But when I know of so many papers that he doesn’t cite that don’t fit with his story, it makes me wonder.

      • James said: “I wish I could believe that. But when I know of so many papers that he doesn’t cite that don’t fit with his story, it makes me wonder.”

        Dear Lord, James please spare us the self-righteous know-it-all ‘tude and cite the studies here. Give us 5. Since it’s his site, Dr. Eades can then judge if you’re right or not. I’ll even send them to Gary and ask him for ya how’s that?! :)

  59. M.S,M.S. and no B.S….You did all of that writing just to point out the validity of Taubes journaling skills contribute nothing to the facts. Why didn’t you just state facts…if Taubes was incorrect in anything that he said, you could have easily pointed that out. Since you do not have your B.S. degree don’t spend so much time shoveling it.

    Bloviating is so aggrivating. We, as lay people need accurate answers and solutions. We don’t have time for speculation and puffery. It is time to express facts that are not colored by opinons.

    No, I am not a doctor of medicine nor do i play one on TV. But I am an observer and reader of nutrition. I am a participant and I am learning what is true and what is false.I eat completely opposite from the way mainstream “experts” claim to be ideal. What I have experienced from eating the “wrong” way had been ideal blood lipid control, healthier weight, glucose/insulin/glucogon levels are as ideal as they can possibly be ( considering the fact that my system has suffered damage due to eating the “right” way for 45 years). I don’t getit. I don’t understand why it is soooo hard to accept a hgih fat diet for what it is. Why do I have to experiment on myself, discover what is true, then told by experts that it is wrong? Dr. Eades has come out with more than enough information that supports my dietary success.Now I have to watch my family suffer because they continue to believe “mainstream” experts.

    I am sorry but i just had to get that out!

    • ********************
      Why didn’t you just state facts…if Taubes was incorrect in anything that he said, you could have easily pointed that out.
      ***********************

      That’s not my point. My point is that Taubes leaves out information that doesn’t fit with his “gripping story.” Leaving out important information is just as bad as stating incorrect things, because it then leads people to incorrect conclusions.

      ***********************
      . Since you do not have your B.S. degree don’t spend so much time shoveling it.
      ***************************

      Ummm, I do have a B.S. degree. The “No BS” is stuck on there as a sort of joke, reflecting my “No BS” attitude.

      *******************
      What I have experienced from eating the “wrong” way had been ideal blood lipid control, healthier weight, glucose/insulin/glucogon levels are as ideal as they can possibly be ( considering the fact that my system has suffered damage due to eating the “right” way for 45 years). I don’t getit. I don’t understand why it is soooo hard to accept a hgih fat diet for what it is.
      ***********************

      I’m not saying what you’re doing is wrong. In fact, I am a supporter of low-carb diets for people that find them effective. I even have a research publication that supports low carb diets.

      My problem is with low-carb *dogma*, which is NOT supported by the science, but which books like Taubes’s have led to. My problem is with the overly simplistic (and wrong) concept that obesity can be pinpointed to a single cause (too many carbohydrates). Funny that Taubes talks about Occam’s Razor in his book (yes, I have read it for the people who think that I haven’t), but doesn’t realize that Occam’s Razor doesn’t apply to the problem of obesity, because you’re trying to apply a simple solution to a highly complicated biological organism that has many interacting and redundant biochemical pathways. Taubes says that Occam’s Razor must apply to science, but there is no rule that it does have to apply to science (at least not in the way that he’s applying it).

      • James said: “That’s not my point. My point is that Taubes leaves out information that doesn’t fit with his “gripping story.” Leaving out important information is just as bad as stating incorrect things, because it then leads people to incorrect conclusions.”

        Spit them out James. C’mon – you can do it. And why not save us the bravado and write for us here on Dr. Eades blog what Taubes’ basic error is. Should be easy for you since you claim to know.

        Hey you have two masters degrees – it should be a breeze for you.

  60. ***************
    Coming to the literature with an hypothesis in mind in advance
    *****************

    Ummm, yeah, doc, that’s the way science is supposed to be carried out…..you started with a hypothesis and then you develop an experiment to test it

    *****************
    and picking the studies that confirm it.
    *****************

    Please. Like I stated earlier, if somebody does that, they can easily be called out on it if you are familiar with the literature in a particular area. If certain studies seem to be left out for no good reason, anybody can write to the journal and point out the glaring omission. Or better yet, they can replicate the analysis if they think the methodology was suspect.

    • Somehow we’re engaged in a circular argument. You just repeated in different words what I just wrote. I’m off the merry go round on this one. I’ll give you the last word.

  61. I always get a kick out of you, Fred. Your arguments and your methods have become so completely predictable that it’s laughable.

    You act very tough when your ensconced into a low carb-zealot-friendly comfort zone, but anyone who isn’t off-the-charts low carb and/or super slow training-biased thinks you’re a complete joke.

    James has been picking apart your sacred cows for years now. While I admire your dogged pursuit of your dietary and exercise ideals, you’ve been given every opportunity to provide evidence of your claims and you have failed. Miserably. You are clearly frustrated that James has science slapped you all over the internet and has repeatedly called you out on your dichotomous generalizations, mistruths, half truths and flat out lies.

    Your method of arguing is hilarious and anyone who disagrees with you is aware of your tactics and logical fallacies. Your bias is clear Fred and to claim that you, Eades or Taubes has the same objectivity as James is quite funny – especially since James has repeatedly stated he is in no way against low carb.

    If you actually had any evidence against James’ assertions against Taubes’ conclusions, you would have had something substantive to say about it on his site. Instead, you whined about how “scientist don’t talk that way” and proceeded on your usual red herring diatribe about Bray’s critique – which had little to nothing to do with James’ critique.

    Tangentially, I wonder why you are even engaging in a debate with James. According to your logic (from one of your posts on the Colpo thread), “those lacking formal nutritional education can only argue from a layman’s perspective”. Does that mean we should all take your opinion as simply a “layman’s perspective”? From that logic alone, why should anyone listen to you over James? Or is parroting spoon-fed information from Fein, Eades, Taubes etc considered ok?

    I guess it would be easier just to buy into every aspect of GCBC without questioning anything – you know, ‘cause the first 200 pages are a “masterpiece” (therefore the entire book is infallible.)

    I understand your tendency to lash out here, Fred. You and others wish that people like James didn’t exist so you could go merrily along your way spewing whatever nonsense you want without accountability. The problem is that anyone with an ounce of intelligence, logic and objectivity will call you on your BS pretty quickly. This is evident from your own blog, the JP fitness site, Alan Aragon’s blog and Baye’s site. I’ll provide some links later.

    The bottom line is that you’ve clung to your absolutist beliefs and nobody is going to change your mind about it. You will continue to argue in circles, moving goalposts, throwing out red herrings and relying on personal anecdote, other irrelevant examples (like your 13 yr. old 7 footer) and appeals to authority to prove your case. Eades got called out for making a spurious and unsubstantiated claim and instead of retracting, he danced in circles and launched an ad hominem. This is the original contention which has been lost amongst the LCT who prefer to just whine about James’ presence and Geiger-count for any smidge of misstep on his part. James addressed his comment on Taubes’ background as a journalist. Perhaps Dr. Eades and Fred can retract on their follies, too?

  62. Pingback: Pro!
  63. There are example of foods which contains a high proportion of saturated fat include animal fats such as cream, cheese, butter, and ghee; suet, tallow, lard, and fatty meats; as well as certain vegetable products such as coconut oil, cottonseed oil, palm kernel oil, chocolate, and many prepared foods.