Kendrick's The Great Cholesterol Con
A Review of Malcolm Kendrick's The Great Cholesterol Con: The Truth About What Really Causes Heart Disease and How to Avoid It
August 22, 2008
Reviewed by Chris Masterjohn
If you want a few good laughs, read this book.
Malcolm Kendrick's 2007 book, The Great Cholesterol Con, is full of sarcastic humor. It parodies some of the most outrageous scientific absurdities ever to find their way into print, and it is impossible not to laugh out loud while reading it. Although it makes a number of excellent serious points, readers with a background in the relevant science might also laugh at some of the egregious scientific errors in the book and some of Kendrick's poorly conceived speculations - or at least find themselves scratching their heads.
The Tall Tale of Teleoanalysis —
The Study That Never Was
One of the most outrageous absurdities Kendrick reveals to us is a technique called teleoanalysis. In short, this is how you perform a study to determine what would have been the result of another study you never carried out.
Really.
In the words of the researchers who developed the technique, Law and Wild, it "provides the answer to studies that would be obtained from studies that have not been done and often, for ethical and financial reasons, could never be done." The problem they faced was that "a meta-analysis of randomized trials suggested that a low dietary fat intake had little effect on the risk of ischaemic heart disease," despite the fact that, as we all already know, "saturated fat intake increases the risk of ischaemic heart disease."
Solution? Simple. If you calculate the effect of saturated fat on cholesterol levels, and then calculate the effect of cholesterol levels on heart disease risk, you can "teleoanalyze" the true effect of saturated fat on heart disease risk without ever studying it. As Law and Wild explained, quoted by Kendrick:
It may also be necessary to quantify the individual effects that relate to separate steps in a causal pathway - that is, the effect of factor A on disease C is determined from the estimate of the effect of A on an intermediate factor B and the estimate of the effect of B on C, rather than by directly measuring the effect of A on C. This exercise is like putting together the pieces of a jigsaw puzzle.
Thus, they conclude that "the effect of a significant reduction in dietary fat can easily be underestimated, even when it is based on the results of randomized trials."
Are you laughing yet?
The Statin Empire and The Cholesterol-
Lowering Money-Making Machine
Kendrick offers a scathing critique of the use of cholesterol-lowering statin drugs, concluding that if ten million people (males with established heart disease, that is) took these drugs for a year they would experience, on average, a life lengthened by a whopping two days at a price to the British government (he hails from Scotland) that could otherwise pay for an extra 70,000 nurses or the construction of two new university-sized hospitals every year.
He offers disturbing testimony from Dr. Marcia Angell, former editor of the New England Journal of Medicine, on the corporate takeover of the research process, a phenomenon that has progressed from mere funding to complete control of the data and relegation of the researcher to a cog in a money-making machine.
If you have a cynical sense of humor, you might laugh when he quotes Harvard Medical School Associate Professor Daniel Simon as saying that most researchers without conflicts of interest "are not truly expert."
No, That is Not How It Actually Works
Kendrick mocks virtually everything in his path. When the subject of his mocking is truly absurd, this is funny, but sometimes he takes the sarcasm too far and it backfires.
The good doctor finds it absurd that saturated fats could raise cholesterol levels. Though admitting that certain parts of some fats might here and there wind up as part of a cholesterol molecule at one point or another, he presents us with acetyl CoA, the basic building block of cholesterol (see my cholesterol synthesis flow chart), and points out that it is full of phosphorus, sulphur, and nitrogen, and contains several ring structures, none of which are found in fats. "Perhaps I should start a new competition," he writes. "In Fig. 10 of Acetyl CoA, can you 'spot the fat'?"
There is just one problem. Acetyl CoA is not a building block of cholesterol - acetate is. The CoA molecule is just a carrier for the acetate. When the two are joined together, it is called "acetyl CoA," but the CoA is just a bus on which the acetate is a passenger. Acetate is made of carbon, hydrogen, and oxygen, all of which are found in fats.
The idea that saturated fat raises cholesterol levels was based mostly on highly controlled metabolic ward studies showing that "milkshakes" made with animal fat produced higher cholesterol levels than the same "milkshakes" made with vegetable fat. Modern textbooks generally attribute this effect to polyunsaturated fatty acids (PUFA) binding up cholesterol and keeping it stuffed in the liver, so it would seem that it is more of an effect of PUFA reducing cholesterol levels than saturated fats raising them. Nevertheless, the idea that cholesterol could be made from saturated fats is not nearly as absurd as Kendrick makes it out to be. And while the evidence that saturated fats do independently raise cholesterol levels is less than stellar, Kendrick's assertion that the phenomenon is not only "biologically implausible" but "biologically impossible" is a gross exaggeration.
This is not the only such mistake in the book. Consider a few others:
- He states that HDL has more cholesterol than any other lipoprotein, when this distinction belongs to LDL (which according to Mary Enig's Know Your Fats has nearly four times the cholesterol content as HDL).
- His figure of a lipoprotein shows triglycerides covering the outside of the particle, whereas every other source I have seen has put these components in the core of the particle, since they are fat-soluble.
- He states that the reason trans fats are unnatural and dangerous is because our enzymes are designed to work with cis bonds, but every time we burn an unsaturated fatty acid for energy, we convert it into a trans fatty acid with an enzyme called 3,2-enoyl CoA isomerase so that the oxidation enzymes, which usually work with saturated fatty acids, can use it. Thus it appears that at least one enzyme designed for utilizing saturated fatty acids, if not all such enzymes, utilizes trans fatty acids perfectly fine.
The Lipid Hypothesis — Dead Wrong or Half Right?
Where Kendrick really goes wrong, however, is his treatment of the role of blood lipids in heart disease.
To be clear at the outset, I do not believe that "high cholesterol" levels cause heart disease. I do, however, believe that oxidized LDL plays an important role in heart disease, and I believe this indicts dietary PUFAs from vegetable oils rather than dietary saturated fats or dietary cholesterol. For a more extensive discussion of this theory, see my review of Uffe Ravnskov's The Cholesterol Myths.
The idea that high blood levels of cholesterol cause heart disease is called the "lipid hypothesis." Kendrick's presentation of the history of the lipid hypothesis bears little resemblance to what actually happened. According to Kendrick, Rudolf von Virchow found that atherosclerotic plaque was loaded with cholesterol in the mid-1800s. The next fifty years were filled with silence until Nikolai Anitschkov produced atherosclerosis in rabbits by feeding them cholesterol. "So, cholesterol in the diet is deadly, and causes heart disease in humans - case proven? Well, Anitschkov certainly thought so."
One gets a remarkably different sense of the history by reading Anitschkov's 1933 review in Cowdry's Arteriosclerosis: A Survey of the Problem (New York: Macmillan). Before the cholesterol-fed rabbit, researchers tried producing atherosclerosis in animals by the following means: ligating, pulling, pinching, wounding or cauterizing blood vessels; increasing blood pressure; inducing kidney lesions; directly injuring the nervous system; injecting adrenalin; injecting a wide variety of toxins; and injecting bacteria or bacterial byproducts.
Anitschkov never concluded from his experiments that dietary cholesterol caused heart disease in humans. Quite the contrary, he wrote the following:
[I]n human atherosclerosis the conditions are different. It is quite certain that such large quantities of cholesterin are not ingested with the ordinary food. In human patients we have probably to deal with a primary disturbance of the cholesterin metabolism, which may lead to atherosclerosis even if the hypercholesterinemia is less pronounced, provided only that it is of long duration and associated with other injurious factors.
Later Kendrick states that "you should always bear in mind that the cholesterol hypothesis started life as the diet-heart hypothesis, with cholesterol in the diet as the major culprit substance." Thus, the idea that LDL in the blood is the culprit is just backpedaling. The idea that oxidized LDL is the culprit - or a culprit - is backpedaling on top of backpedaling.
Atherosclerotic plaques occur behind the innermost layer of the artery called the endothelium. Kendrick spends three entire pages telling us that no one has ever come up with a molecular mechanism by which LDL or oxidized LDL can find its way behind this supposedly impermeable barrier into an atherosclerotic plaque.
Really?
This is contradicted by the following 2.5 pages, which Kendrick spends describing a molecular mechanism (full of long words and intimidating acronyms) by which oxidized LDL can be taken into endothelial cells as if this were the main hypothesis of how oxidized LDL gets behind endothelial cells (it isn't), which he finally concludes is mind-baffling "bull--."
Kendrick posits a hypothetical five-year-old capable of taking a step back from the scientific mumbo-jumbo and pondering why, if LDL causes atherosclerotic plaques, these plaques only occur in some places but not others when the concentration of LDL everywhere is the same.
The explanation of how LDL or oxidized LDL can get behind the endothelium is much simpler and much more established than Kendrick suggests, and it even answers his question of why plaques occur in specific places!
Anitschkov showed that molecules stained with dye permeated the vascular endothelium - the innermost layer of the blood vessel - in all the same spots where plaques develop, and that inflammation increased the permeability.
Places that experience disturbed blood flow, such as arterial branch points, experience reduced shear stress, which is the force caused by blood flowing parallel to the lining of the vessel. Shear stress increases the production of nitric oxide, which inhibits the oxidation of LDL, inhibits the ability of white blood cells to invade the endothelium and embed themselves behind it, and inhibits blood clotting. A recent study showed that the low level of shear stress experienced by the blood vessels in locations where plaque develops leads to increased vascular permeability. Easily visualizable gold particles the size of LDL particles slip right in between the endothelial cells under these conditions, because cells make fewer of the proteins that keep them locked together.
Once again, the theory makes quite a bit more sense than Kendrick would lead one to believe.
Lipoprotein (a) — An Alternative Hypothesis?
Although Kendrick finds it absurd that either LDL or oxidized LDL have any role in the development of heart disease, he is perfectly happy to admit such a role for lipoprotein (a), or Lp(a). Lp(a) is a subfraction of LDL that associates with a protein called apolipoprotein (a) or apo(a). So "apo(a)" refers to the protein, and "Lp(a)" refers to the protein with LDL stuck to it.
Atherosclerotic plaques are loaded with Lp(a) and patients with familial hypercholesterolemia (who have a dramatically elevated relative risk of heart disease in their younger years) have high blood levels of it.
Kendrick favors a hypothesis that eliminates any role for the lipid fraction of this particle: the protein involved, apo(a), inhibits an important enzyme that breaks down factors involved in blood clotting. This is particularly useful, because Kendrick's own theory of heart disease, discussed further below, centers on blood clotting.
To date, this hypothesis seems to have more evidence against it than in favor of it. Consider the following abstract from a 2007 review:
Lipoprotein (a) (Lp (a)) may be pro-thrombotic in humans due to its apolipoprotein (a) (apo(a))-mediated decreases in fibrinolysis. Such decreased fibrinolysis arises putatively from interference with plasminogen conversion to plasmin due to the considerable homology between apolipoprotein (a) and plasminogen. However, in vitro, most studies have shown that human Lp (a) decreases agonist-stimulated platelet aggregation while in vivo it appears to decrease aggregation as implied by increased bleeding times with higher blood serum concentrations of Lp(a). . . . However, it is concluded more in vivo work needs to be done to fully understand whether, if at all, Lp(a) in varying concentrations and isoforms, favours reduced platelet aggregation or fibrinolysis.
In other words, test tube studies show that Lp(a) can inhibit the breakdown of clotting factors, but they also show that it exerts separate anti-clotting effects. Which effects play out in the live human? Clotting or anti-clotting? People with higher blood levels of Lp(a) have increased bleeding times, suggesting the anti-clotting effects win out.
There is another recent hypothesis about the role of Lp(a) in heart disease presented here. Blood levels of Lp(a) are strongly correlated to blood levels of oxidized phospholipids. Oxidized LDL, whose membrane phospholipids are the first targets of oxidation, is found almost exclusively associated with Lp(a), and "donates" its oxidized phospholipids directly to apo(a). These authors believe that at low concentrations Lp(a) may function as a type of "mop" to mop up the oxidized phospholipids and either stop them from coming into contact with delicate tissues or maybe even detoxify them, but at higher concentrations its capacity to manage them may be overwhelmed.
Kendrick offers Lp(a) as an alternative to LDL when attempting to explain why familial hypercholesterolemic patients have a high relative rate of heart disease in youth and middle age. Familial hypercholesterolemia (FH), however, is caused by a defective LDL receptor, meaning that cells cannot effectively take in LDL from the blood. Kendrick never explains how FH can lead to increased Lp(a) levels independently of LDL levels. Thus, Kendrick never explains how this hypothesis is an alternative to any of the LDL-related hypotheses.
The conventional understanding of FH, based on Nobel Prize-winning science, is that only the LDL receptor is involved. Thus, the only thing going on is that LDL is not making it into the cell. Cells are also believed to jealously guard their own cholesterol concentrations by increasing or decreasing cholesterol synthesis as needed.
If these understandings are correct, then there are only two reasonable explanations for how FH can raise Lp(a) levels: either a higher concentration of LDL in the blood raises Lp(a), or LDL that has spent a longer amount of time in the blood raises Lp(a). If Lp(a) is indeed a reaction to LDL oxidation, then FH probably causes increased Lp(a) levels by allowing LDL to sit around in the blood for an extended period of time, allowing it to oxidize.
Thus, even if Lp(a) does promote blood clotting, it is not an alternative explanation to LDL- or oxidized LDL-related theories - it is just an adjunct to them.
Kendrick's Theory of Heart Disease —
Stress and Clotting
Kendrick spends the last couple chapters of his book discussing the role of stress in heart disease. I have little doubt that he is correct that stress plays an important role in this disease, but I doubt it plays out in precisely the way Kendrick suggests it does.
Kendrick's evidence is primarily ecological, meaning he looks at different countries or ethnic groups over time, rather than at individuals. He suggests that heart disease rates have always been highest during periods of "social dislocation" and makes an interesting case, even though it is primarily circumstantial in nature.
He also discusses Michael Marmot's very interesting study showing that the risk of heart disease for Japenese immigrants was determined more by whether they kept their cultural traditions than by whether they kept their traditional diet, and he presents interesting research showing that Chinese and Japanese Americans die more often on the fourth day of the month than on any other day and that Westerners die more often on Mondays.
Apparently "the number 4 evokes discomfort and apprehension in many Chinese and Japanese people" while the rest of us just hate getting up for work after a relaxing weekend.
He cites a few other studies showing that people with heart disease are more likely to have certain pyschosocial risk factors or have other signs of stress. Interesting studies, but mostly the stuff of generating hypotheses rather than confirming them. Kendrick stretches the conclusiveness of his data somewhat:
Right is that enough for you? If you don't believe that social dislocation causes heart disease by now, I'll never be able to convince you.
So how does stress cause heart disease, according to Kendrick? Stress causes the release of certain hormones, especially cortisol. These somehow damage the endothelium. A blood clot comes to the rescue to patch up the endothelium, but since the arterial wall cannot risk letting a clot go loose in the blood, it sucks it in to metabolize it. Cells from the bone marrow become new endothelial cells and layer themselves on top of the clot.
As evidence for this, Kendrick cites a study showing that some plaques contain bits of fibrin-related proteins. Fibrin is involved in blood clotting.
That might be evidence that many plaques have experienced minor ruptures here and there along the road, but it leaves a lot left unexplained.
Why is it that plaques have a fatty core if they are made up of old blood clots? Kendrick says it is because blood clots contain Lp(a), which in turn contains LDL, but this seems to be a gross misinterpretation of the hypothesis he had previously cited which, to date, appears unlikely to be true.
If Lp(a) inhibits the breakdown of a clot by binding to the enzyme that would otherwise break down the clot, this would not make the Lp(a) part of the clot. As far as I know, blood clots are not generally filled with fat and cholesterol.
Then we must ask, of course, what happens to all these layers of endothelial cells? Shouldn't we see something that looks like a layered cake, alternating old blood clots with old endothelial layers? On the other hand, if the majority of the clots get metabolized and all the endothelial cells get metabolized, why can't the fatty core get metabolized?
Heart Disease — It's More Than Atherosclerosis
To my knowledge, no one has yet tried to produce atherosclerosis in animals by injecting cortisol, but early experimenters did inject adrenalin into rabbits. It caused arterial lesions, but they didn't look anything like human lesions and they didn't occur in the same places. If hypercholesterolemia, which we now know is hyper-oxidized-lipoproteinemia, was combined with adrenalin injection, the adrenalin accelerated the development of atherosclerotic lesions and modified their location somewhat.
Of course, there are many other degenerative changes that happen to blood vessels besdies atherosclerosis. In diabetes and aging, the media calcifies and becomes stiff. This occurs in the middle of the artery instead of in the intima, the part near the blood; it occurs evenly rather than in isolated lesions; and it does not involve fatty degeneration. Numerous toxins can cause local damage to blood vessels and elicit inflammation. Electrical impulses in the heart can misfire, and energy insufficiency can lead to congestive heart failure.
Ceratinly stress plays an important role in cardiovascular disease. But does it lead to the specific fatty plaques that characterize atherosclerosis? If it contributes to them, is it sufficient in and of itself, is it the primary cause, or does it merely aggravate the phenomenon? Kendrick singles stress out as "the primary cause" of heart disease and posits atherosclerosis as the main mechanism by which it causes heart disease, but his case just ultimately fails to convince me on this point.
Conclusions and Who Should Read This Book
Kendrick's The Great Cholesterol Con is easy to read, set in a conversational tone, and full of good laughs. For someone who plans on reading multiple books on the cholesterol controversy, it would not be a bad read. However, the reader must be discerning and intent on either fact-checking much of the basic science or taking it with a grain of salt. It makes many good points, but few points not made by other cholesterol skeptics.
Anthony Colpo's book of the same title covers the role of stress in heart disease without entangling it in a mess of highly speculative and almost certainly erroneous science, and covers many other contributors to heart disease that Kendrick does not cover to boot. Of course, the format of Kendrick's book makes it easier to read (Colpo's book is in ridiculously small print), but the quality of Colpo's content is much better.
That said, if you are into sarcastic humor and want to spend a few days reading Kendrick make fun of other researchers, then this is the book for you.