Cholesterol And Heart Disease — Just a Myth?
A Review of Uffe Ravnskov's The Cholesterol Myths: Exposing the Fallacy that Saturated Fat and Cholesterol Cause Heart Disease
August 21, 2008
Uffe Ravnskov's 2000 book The Cholesterol Myths was a blockbuster among skeptics of mainstream health and nutrition dogma. With the heavy credentials of an MD, PhD author, the book formed the impetus for a resurrection of once-popular cholesterol skepticism that eventually led to the formation of The International Network of Cholesterol Skeptics in 2003, of which Ravnskov is the spokesman.
There is no doubt that the book is a good read. After all, it was set on fire on Finnish television in 1992, before the English translation came out. Any book that raises the ire of nutrition experts to such a degree that they would burn it on national television certainly carries enough controversy to be worth your while!
Ravnksov's book is an important contribution to the debate about the role of cholesterol in heart disease, but it is not without its flaws. The idea that eating saturated fat or having a high cholesterol level inexorably leads to a heart attack is so pervasive and until recently had been so unquestioned that we were in dire need of someone like Dr. Ravnskov to organize the contrary arguments and make them accessible to the public. Ravnskov convincingly shows that this simplisitc idea is full of gaping holes.
Ultimately, however, Ravnskov leaves the reader with the idea that blood lipids have virtually nothing to do with heart disease, that existing animal models can teach us almost nothing about human heart disease, and that good scientists should never have given the cholesterol theory much weight because it was built entirely on a foundation of sand. In reality, blood lipids do matter, animal models are relevant, and though the cholesterol theory is a lopsided house, it is in fact built on a solid foundation of evidence that needs to be carefully looked at.
These are the simple facts: on the one hand, if there was never any good evidence for a role of blood lipids in heart disease, such a massive "cholesterol education" and cholesterol-lowering drug empire could never have been built; on the other hand, were the mainstream theory not seriously flawed, Ravnskov could never have written his book. Before we review some of its flaws, however, let's look at some of its remarkable strengths.
In This Review:
Familial Hypercholesterolemia - Is It Relevant to the Rest of Us?
Atherosclerosis Is Good for You? Certainly Another Cholesterol Myth!
The Diet-Heart Hypothesis: The Cholesterol Myth Crumbles
The diet-heart hypothesis is the idea that dietary saturated fat and cholesterol cause heart disease. Ravnskov not only destroys this myth with the hard data but shows it was never based on honest science in the first place.
Ancel Keys first argued this theory by charting heart disease mortality against fat availability for six countries, showing the more dietary fat available, the higher the rate of mortality. There was just one pesky problem: data was available for 22 countries at the time. Include the other 16, and the association falls apart.
Ravnskov also reviews numerous examples of populations consuming diets rich in saturated fat who are virtually immune to heart disease, such as the Masai of Africa, who eat a diet almost exclusively composed of meat, blood, and milk.
Are such native populations merely genetic anomalies? If we ask this question, we implicitly assume that the research conducted in modernized Western populations shows something different. But Ravnskov compiled the 27 studies following 34 cohorts or groups composed of more than 150,000 individuals that researchers had published up until 1998. In three of the 34 groups, those who ate more saturated fat were more likely to have heart disease, in one they were less likely, and in the other thirty the amount of saturated fat they ate had nothing to do with how likely they were to develop heart disease.
What about dietary cholesterol? Some studies have showed small differences favoring the diet-heart hypothesis and others have shown the opposite. Pool the results together, and healthy groups have eaten, on average, just slightly more cholesterol than groups with heart disease!
Should You Lower Your Cholesterol or is This Just Another Myth?
Ravnskov's extensive analysis of cholesterol-lowering treatments spans over 80 pages. He shows conclusively that cholesterol-lowering drugs are expensive, a waste of time, and even dangerous.
The earliest drug trials were a disaster. In the famous World Health Organization (WHO) trial, a cholesterol-lowering drug called clofibrate did indeed reduce the incidence of heart disease, but the heart disease mortality rate stayed the same, and the total mortality rate increased! In other words, people who took the drug were more likely to die earlier than people who did not.
The Lipid Research Clinic (LRC) Coronary Primary Prevention Trial (CPPT) used a drug called cholestyramine. This trial, said to have "proved" the cholesterol theory, reduced the risk of heart disease but left the total mortality rate unchanged. The investigators pointed out that many of the non-heart disease deaths in the treatment group were accidents, suicides, and in one case a homicide. Obviously, they said, the drug could not have caused these deaths. According to Ravnskov, this is not so obvious. Not only has later research shown that cholesterol-lowering drugs hurt mental performance, but many of the side effects recorded in the CPPT were neurological problems.
Since cholesterol is essential to the brain and nervous system, then, it is quite possible that cholesterol-lowering drugs increase the risk of accidents and suicide by decreasing reaction time, increasing aggressiveness, and ruining one's mood. Without doubt, they can certainly destroy your memory.
Some of the early trials even increased the risk of heart disease, and others were almost certainly manipulated to produce favorable results.
The new generation of statin drugs, however, have produced a never-ending euphoria in which physicians have been enamored with their power to consistently reduce heart disease and total mortality and investors have been enamored with the tens of billions of dollars of revenue the drugs take in every year.
Are these drugs really all they are cracked up to be?
Even if we are generous enough to assume that they benefit women and elderly as much as they benefit middle-aged men, something that has yet to be sufficiently studied, statins may well hurt more than they help. The benefits are greatly exaggerated by using "relative risk" statistics, when the truth is that dozens or even hundreds of people need to take the drugs in order for one person to benefit.
They cause cancer in laboratory animals, and the trials looking at quickly developing cancers or examining the effect in elderly subjects have shown substantial increases in the cancer rate. Investigators don't even look for the side effects most likely to result from treatment, like poor mental performance, memory loss, and poorer social and family life. They deliberately design trials to underestimate side effects by using younger people with lower risks of cancer and only looking for cancers that take much longer to develop than the length of the trial, and by eliminating half the people who sign up during the screening process, allowing them to weed out the people most likely to exhibit side effects.
No, they are not all they are cracked up to be, and Ravnskov masterfully pops the bubble of statin euphoria with his sharp and skeptical analysis.
Polyunsaturated Fatty Acids (PUFA) Are Good for You — Another Cholesterol Myth
In 1981 all the big names in the cholesterol business met to discuss the ominous finding that low cholesterol levels are associated with the risk of cancer. The panel, of course, concluded that the low cholesterol levels were not causing the cancers (even though they are quick to conclude that high levels of cholesterol cause heart disease) and that the risk was modest.
Ravnskov agrees. But if the low cholesterol levels are not causing the cancer, why would they be associated with it?
It certainly isn't that the cancers cause the low cholesterol levels, because the low cholesterol levels are often found many years before the cancers could consume meaningful amounts of choelsterol. Ravnskov has another idea - a high intake of PUFA can cause both low cholesterol and cancer.
The idea is quite plausible. As discussed in my PUFA Report, research has long shown that even small amounts of PUFA in the diet cause substantial increases in the risk of cancer. And the medical establishment has been recommending PUFA for decades based precisely on the fact that consuming PUFA generally lowers cholesterol levels.
Ravnskov also shows that dietary PUFA accelerate aging and cause brain damage. He comes short of arguing that PUFA cause heart disease, mostly because he rejects the animal models of experimental atherosclerosis - a mistake we will discuss below.
The bottom line? The medical establishment recommends increasing PUFA intake to prevent heart disease, but doing so increases the risk of cancer and brain damage, makes you age faster, and may even increase your risk of heart disease.
It is the nail in the coffin of the diet-heart hypothesis. Not only has Ravnskov destroyed the myth that saturated fat is the "bad fat," but now he has destroyed the myth that PUFA is the "good fat." In fact, if PUFA is indeed the "bad fat," it only logically follows that the fats rich in saturated and monounsaturated fatty acids such as coconut oil, butter, and other animal fats, are in fact good fats!
Flaws in The Cholesterol Myths — A Critical Review
While The Cholesterol Myths is a monumental contribution to the debate about cholesterol's role in heart disease, it is not without its flaws. The principal flaws result from lumping things together when they should be analyzed separately, and failing to connect other things that should be connected. He posits missing links that do not exist, dismisses animal evidence that is very relevant, and throws out entire hypotheses when they should be modified instead. The result is that we learn nothing from the animal studies and conclude that blood lipids have almost nothing to do with heart disease, come to the absurd conclusion that atherosclerosis might actually be good for you (!), and, most importantly, fail to realize the full extent of the dangers of PUFA.
A Note on the Scientific Method
Ravnskov introduces an approach to the scientific method at the beginning of the book that could best be described as the "throw-the-baby-out-with-the-bathwater" approach. On page 12, he writes the following:
[I]f a scientific hypothesis is sound, it must agree with all observations. A hypothesis is not like a sports event, where the team with the greatest number of points wins the game. Even one observation that does not support a hypothesis is enough to disprove it. The proponents of a scientific idea have the burden of proof on their shoulders. The opponent does not have to present an alternative idea; his task is only to find the weakness in the hypothesis. If there is only one proof against it, one proof that cannot be denied and that is based on reliable scientific observations, the hypothesis must be rejected.
Not exactly. The hypothesis could be rejected, or, it could be modified. If there are many facts that seem to support a hypothesis and many that seem to contradict it, that is a sign that we might need to modify the hypothesis instead of rejecting it outright.
A fact rarely if ever refutes a hypothesis per se. Instead, it refutes a particular prediction of the hypothesis. The prediction rarely represents an entire hypothesis but instead represents a particular part of it. It could also represent a particular interpretation of its implications, so it might be our understanding of the implications of the hypothesis that needs to be modified rather than the hypothesis itself.
As we will see, this is especially important when considering the "lipid hypothesis."
The Lipid Hypothesis — Fact and Fiction
Ravnskov uses the terms "diet-heart idea" and "lipid hypothesis" interchangeably (see page 9). These two terms, however, are properly understood as two separate hypotheses, a point that lipid hypothesis proponent Daniel Steinberg emphasizes in his book The Cholesterol Wars.
The diet-heart hypothesis states that saturated fat and cholesterol in the diet (or just one or the other) cause heart disease. The lipid hypothesis states that an elevated level of cholesterol in the blood causes heart disease.
The two are not one and the same. Proponents of the diet-heart idea usually rely on the lipid hypothesis to make their argument - the mechanism by which saturated fat and cholesterol in the diet cause heart disease is by raising the level of cholesterol in the blood, they say - but the lipid hypothesis does not in any way rely on the diet-heart hypothesis.
For example, when Nikolai Anitchkov described the cholesterol-fed rabbit model and other experimental models of atherosclerosis, he concluded that blood levels of cholesterol were the principal (but not exclusive) factor in the disease, but he never concluded that human atherosclerosis was caused by eating cholesterol. In a 1933 English-language review entitled "Experimental Arteriosclerosis in Animals" published in Cowdry's Arteriosclerosis: A Survey of the Problem (New York: Macmillan), Anitchkov 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.
So the lipid hypothesis in its foundation never had anything to do with the diet-heart hypothesis. As Steinberg points out, the two hypotheses were "short-circuited" in the 1950s and 1960s because diet was the only means available of lowering cholesterol. It was the Ancel Keys crowd, the data manipulators and interventionists, who made that ridiculous connection, not good scientists like Anitchkov.
To Ravnskov, the fact that cholesterol-feeding produces atherosclerosis in some animal models and not others casts serious doubt on the diet-heart hypothesis, aka the lipid hypothesis. "How do we know whether man reacts like a rhesus monkey or a baboon or in some other way?" he asks.
Steinberg analyzes the problem in a totally different way. By separating the diet-heart hypothesis from the lipid hypothesis, Steinberg looks at whether raising cholesterol levels in the blood produces atherosclerosis. This cannot always be achieved by cholesterol feeding alone. But when it has been achieved, it has produced atherosclerosis in baboons, cats, chickens, chimpanzees, dogs, goats, guinea pigs, hamsters, monkeys, mice, parrots, pigs, pigeons, rabbits and rats.
To throw out the lipid hypothesis, which concerns blood levels of lipids, with the diet-heart hypothesis, which concerns dietary lipids, is to throw the baby out with the bathwater. But is the lipid hypothesis really true? Or does it need to be rejected - or modified? In order to address this question, we need to reexamine the value of the cholesterol-fed rabbit model.
The Cholesterol-Fed Rabbit — Relevant or Irrelevant?
Ravnskov describes the cholesterol-fed rabbit model as if it produced an effect that looked nothing like human atherosclerosis whatsoever. Unlike the atherosclerotic human, the cholesterol-fed rabbit has its eyes and internal organs stuffed with cholesterol. "It is true that cholesterol is also deposited in the arteries of the rabbit," he writes, "but these deposits do not even remotely resemble those found in human atherosclerosis."
It is true that rabbits fed large doses of pure cholesterol dissolved in sunflower oil develop cholesterol deposits in all their internal organs, but Anitschkov produced atherosclerosis without major deposits in internal organs simply by feeding lower amounts of cholesterol over longer periods of time. For example, feeding rabbits milk produced a much more mild result that primarily affected the arteries.
If one reads Anitchkov's English-language review in Cowdry's 1933 collection, Arteriosclerosis: A Survey of the Problem, one is struck by how precisely the nature of the rabbit lesion resembles that of the human lesion. First fatty streaks develop beneath the endothelial layer (the layer that has contact with the blood); then circulating white blood cells called monocytes embed themselves beneath the endothelium, start to engulf the lipid, and transform into large, lipid-filled phagocytic cells; smooth muscle cells migrate from the media to form a fibrous cap, resulting in "elevated thickenings or plaques in the arterial wall." The core of the lesion contains lipid-rich white blood cells and crystals of cholesterol that "often contain large quantities of calcium."
The lesions do not appear in random positions but localize in a predictable pattern. It is not exactly the same pattern as in the human, but it is largely the same, and as in the human, the type of blood flow that the arterial wall experiences in a given area seems to determine whether it develops such a lesion. Since the rabbit's anatomy is not exactly the same as the human's, it would be astounding and completely unexpected if the dictates of blood flow caused the lesions to develop in exactly the same places as in humans. What we should expect is that the same underlying physiological principles would dictate where they develop.
The largest difference that Ravnskov notes is that the lesions in the rabbit never ulcerate and never rupture - most strikingly, they never lead to a heart attack. This, however, is another example of lumping together things that should be analyzed separately. The factors that produce the lesion are not necessarily the same as the factors that cause it to rupture.
According to a recent review by Brigham and Women's Hospital's Peter Libby, the principal factor determining plaque rupture is the balance between collagen synthesis and collagen degradation. T cells, foam cells, smooth muscle cells, and endothelial cells within the plaque can all secrete inflammatory compounds that enhance the degradation of collagen or decrease its synthesis. While oxidized LDL can enhance this process, it is far from the only factor.
The importance of collagen raises a further point. The production of collagen is dependent on the availability of vitamin C. Most animals, including rabbits, make their own vitamin C; humans do not. It is quite possible that atherosclerosis never leads to plaque rupture in the rabbit because the rabbit has a much higher vitamin C status and thus is capable of producing enough collagen to protect the lesion from rupture.
The cholesterol-fed rabbit actually demonstrates a striking similarity to human familial hypercholesterolemia (FH). In severe cases of homozygous FH, meaning FH caused by two defective genes, one from each parent, cholesterol is deposited in eyelids, joints, internal organs, and other areas of the body in addition to the arteries. The mild cholesterol-fed rabbit model, where atherosclerosis is the main effect, corresponds to heterozygous FH, meaning FH where there is only one defective gene, where atherosclerosis is more severe and appears earlier in life but is otherwise like the atherosclerosis everyone else gets.
Ravnskov makes this connection but claims that FH, like the cholesterol-fed rabbit model, is irrelevant to everyone else.
Familial Hypercholesterolemia — Is It Relevant to the Rest of Us?
Familial hypercholesterolemia (FH) is a condition where the LDL receptor is defective and cells cannot efficiently take in LDL particles from the blood. Steinberg argues that, since cells jealously maintain their own cholesterol concentrations by adjusting the rate of cholesterol synthesis, a defective LDL receptor has only one effect in the body: to raise the level of cholesterol in the blood. Thus, FH offers a substantial piece of evidence in favor of the lipid hypothesis. Ravnksov, by contrast, argues that this is a genetic defect that most people do not have, so we should consider it as an entirely seprate phenomenon.
Actually, we don't know how similar or different it is to people without the defect. Other people might have fully functional LDL receptors, but they might degrade them at a higher rate or deliver them to the surface of cells at a lower rate. The LDL receptor requires thyroid hormone to function, and many people have suboptimal thyroid status. So it may well be the case that high cholesterol is often caused by decreased functioning of the LDL receptor.
At the same time, it simply is not true that a defective LDL receptor changes nothing but the concentration of LDL in the blood. It also causes LDL to be in the blood for a longer amount of time. I'll discuss this more in the next section.
Ravnskov dismisses the relevance of FH to the rest of us because in its extreme form, it looks so different:
The peculiarities of individuals with familial hypercholesterolemia is best seen in the rare homozygous form, the form that appears when both parents have the deficient gene for the LDL-receptor. Autopsy studies of such individuals show that cholesterol deposition is increased, not only in their vessels, but generally throughout their bodies. Many other organs are impregnated with cholesterol, just as we find in the cholesterol-fed rabbits.
More precisely, this is just what we see in the severe model of the cholesterol-fed rabbit, where pure cholesterol dissolved in vegetable oil is used rather than cholesterol-rich foods such as milk. What about the more mild heterozygous form of FH?
The vascular changes seen in people with the more common heterozygous form of familial hypercholesterolemia are more difficult to analyze because these changes must partially be due to the metabolic error and partly to common atherosclerosis. And how do we know whether the possible effects of treatment stem from reduction of the changes caused by the inborn error or from reduction of atherosclerosis? Thus, any conclusion that may be true for individuals with familial hypercholesterolemia cannot possibly be valid for the rest of mankind.
Thus we are to assume that the changes due to FH and those due to "common atherosclerosis" are two entirely separate things. The differences between the two are so "difficult to analyze" that Ravnskov cannot name any; we simply must believe that these two phenomenon look exactly the same and occur simultaneously but nevertheless are two completely unrelated things.
The fact that FH produces the same type of atherosclerosis as the rest of us develop suggests that we should study it to try to learn things that are relevant to the rest of us from it. The fact that severe FH looks just like the severe cholesterol-fed rabbit model and mild FH looks just like the mild cholesterol-fed rabbit model suggests that we should study those models and try to learn things from them that are relevant to us. Ravnskov does hint at an important lesson from the rabbit model, but his excessive partitioning of all these phenomena into different boxes stops him from concluding that the lesson has any relevance to common heart disease.
Oxidized LDL - Making the PUFA Connection
Ravnskov points out that the antioxidant probucol prevents the development of the fatty streak in the Watanabe rabbit, a rabbit model of familial hypercholesterolemia. He reviews the science showing that oxidized LDL, but not regular LDL, accumulates in foam cells, a critical part of forming the "fatty streak." Ravnskov, however, believes that the fatty streak is not a precursor to the atherosclerotic lesion and that the Watanabe rabbit model is relevant only to people with FH and not to the rest of us. He concludes, then, that oxidized LDL may contribute to the development of fatty streaks in people with FH, but not to atherosclerotic lesions or heart attacks in these people, and not to fatty streaks in the rest of us.
Since The Cholesterol Myths was published, researchers have produced a similar result with the antioxidant resveratrol in the cholesterol-fed rabbit model (see here.) The Watanabe rabbit develops raised atherosclerotic lesions just like the cholesterol-fed rabbit and humans with FH. The cholesterol-fed rabbit model established clearly that fatty streaks progress to complex lesions, and as Steinberg points out, human studies show that complex lesions develop in the same places that fatty streaks developed years earlier.
Thus, there is no rational reason to compartmentalize all these different findings into separate boxes as if they had nothing to do with one another. The animal models are relevant, familial hypercholesterolemia is relevant, the fatty streaks are relevant, and the antioxidant studies are relevant. The question is, if we do not simply dismiss them all and throw them out with the bathwater, exactly what is it we can learn from them?
Steinberg, as mentioned before, is wrong that decreased LDL receptor function only affects the concentration of LDL in the blood. By blocking its entrance to the cell, it also increases the time LDL spends in the blood.
LDL is a complex particle that is surrounded in a membrane of phospholipids, which in turn contain polyunsaturated fatty acids or PUFA, which are highly vulnerable to oxidation. In the cell, PUFA are protected by antioxidants and antioxidant enzymes that the cell continuously produces. If PUFA begin to oxidize, the cell increases its production of these protective compounds and enzymes even further. When the liver stuffs cholesterol into an LDL particle, these PUFA leave the comparative safety of the liver and are packaged with only a limited supply of antioxidants. Ideally, the particle reaches its destination before the antioxidants run out.
Sending LDL from the liver to another cell in the body is somewhat like taking a bottle of oil out of the refrigerator and using it. If you open it up, it is exposed to air. It contains vitamin E and other antioxidants that protect it from going rancid so long as it is only exposed to air intermittently and for a short duration. If you use the oil quickly and return it to the fridge, the oil does not go rancid. But if you open it and leave it on the counter, it goes rancid.
Keeping LDL sitting in the blood for an extended period of time is like opening a bottle of oil and letting it sit on the counter. Over time, it goes rancid.
As described by Peter Libby here and here, oxidized LDL contributes to the deposition of the fatty streak, the adhesion of white blood cells called monocytes, their transformation to macrophages and then foam cells, and the secretion of inflammatory compounds that lead to plaque rupture. Many studies, including this one and this one suggest that the principal components of oxidized LDL that transform monocytes into macrophages and foam cells are oxidized derivatives of linoleic acid, a PUFA found in vegetable oil.
So, we can ignore all the evidence as irrelevant, or we can pay attention to it and learn the following strategies for preventing heart disease:
Eat a diet rich in antioxidants, especially coenzyme Q10, the master antioxidant of LDL.
Eat a diet low in PUFA, especially from vegetable oils (for more on PUFA, see my PUFA Report, "How Essential Are the Essential Fatty Acids?").
Maintain a robust rate of cholesterol utilization (the nutritional implications of which direly need more study).
These are important lessons to learn. Ravnskov is correct that the focus on cholesterol levels has led to an enormous waste of talent, effort, and money that could have been used to research other aspects of heart disease. At the same time, those of us in the natural health movement are wasting an enormous amount of time and effort arguing whether cholesterol has anything to do with heart disease when we need to be studying natural ways of increasing LDL utilization rather than leaving the unconvinced to choose between statins and the equally dangerous red yeast rice.
Atherosclerosis Is Good for You? Certainly Another Cholesterol Myth!
There is one claim made in the book that just seems totally bizarre to me - the idea that atherosclerosis is good for you. Ravnskov suggests this on page seven:
When we grow old our arteries become stiff. The smooth muscle cells and the elastic fibers that surround our blood vessels when we are young are gradually replaced by more or less fibrous and rigid tissue. At the same time, or later on, cholesterol, various fats and even calcium become embedded in the blood vessel wall. Arteries probably become stiff as a protective measure, to prevent the pressure of the blood inside them from causing them to widen too much. Thus, the remodeling of the arteries does not occur evenly. It is most pronounced where the strain to the artery wall is highest, for instance, where the blood vessels branch. Such localized thickening is called an atheroma or plaque. . . . For unknown reasons, in some people the embedding of cholesterol in the arterial wall becomes irregular and protrudes into the interior of the artery. Sometimes these localized protrusions, called raised lesions, even change into a material similar to limestone. The embedding of cholesterol and lime may also progress until the vessel becomes so narrow that the heart gets too little blood and thus too little oxygen. It is these constrictions that are considered to be the cause of heart attacks, either directly, or by starting the formation of a clot.
So the embedding of cholesterol and calcium is bad when it gets to the point where it produces a raised lesion, but when cholesterol and calcium embed themselves in the arterial wall just enough to stiffen it, this is a good thing!
Ravnskov did not develop this idea. He does supply two citations for it, so we can put most of the blame on those authors instead. Also, we can blame the passage of time because when these hypotheses were suggested back in the 1990s, we may not have had a good explanation for why hemodynamic factors cause lesions to develop in certain places and not others. Nevertheless, it still seems bizarre.
To begin with, Ravnskov points out that most animals do not naturally develop atherosclerosis, and he never suggests that lifespan or blood pressure can account for why some do and some don't. Do their blood vessels grow too wide over time with no plaque to keep them in their place? Apparently not.
Second, the cholesterol-fed rabbit model shows clearly that fatty streaks progress to plaques, which progress to raised lesions. If the cholesterol feeding stops, the plaques regress. The rabbit naturally gets fatty streaks during infancy when suckling milk from its mother, and these disappear later on. The rabbit does not develop fatty streaks or plaques as an adult unless it is fed cholesterol or cholesterol-rich foods, in which case they appear in the same locations that the fatty streaks appeared in during lactation. These experiments show clearly that the raised lesions are a progression of the same phenomenon as the fatty streak. They further call into question why, if this is a good thing, they only occur in the rabbit on an unnatural diet.
Third, Ravnskov states quite clearly later in the book that the accumulation of LDL into the fatty streaks only occurs if the LDL is oxidized. LDL that is heavily oxidized kills most healthy cells; it is toxic. It damages the production of nitric oxide, necessary for blood vessel dilation. It causes foam cells, smooth muscle cells and endothelial cells to produce inflammatory compounds that accelerate the progression of the obviously harmful stages of plaque development. If the accumulation of foam cells behind the endothelial layer of the blood vessel wall is health-promoting, why on earth would the body choose something so toxic as oxidized LDL to accomplish it?
What we now know about nitric oxide allows us to form a more reasonable explanation of why lesions form in some places and not others. As Peter Libby explains, the places where lesions develop experience disturbed blood flow. Laminar, or parallel, flow causes shear stress, which causes the blood vessel lining to produce nitric oxide. Disturbed blood flow reduces laminar flow and counteracts its effect. Nitric oxide prevents the oxidation of LDL, dilates blood vessels, and prevents the adhesion of white blood cells to the blood vessel lining. One of the ways exercise might protect against cardiovascular disease, in fact, is by increasing laminar blood flow and thus boosting nitric oxide production.
So yes, blood flow determines the localization of atherosclerotic lesions, just like Nikolai Anitchkov described almost 100 years ago in the cholesterol-fed rabbit. But no, that does not mean atherosclerosis is good for you.
Conclusions and Who Should Read This Book
Despite its flaws, The Cholesterol Myths is a very important book. It convincingly demonstrates that there is something seriously wrong with the mainstream theory of heart disease. It destroys the diet-heart hypothesis, and makes a compelling case against the use of cholesterol-lowering drugs.
An excellent companion to this book would be Steinberg's The Cholesterol Wars. Anyone interested in the controversy over the role of cholesterol in heart disease should read at least one book from each side. So far, Steinberg's book is the only one I know of on the "pro" side, and it is certainly the most recent one.
Steinberg's case for the relevancy of the cholesterol-fed rabbit is much stronger than Ravnskov's case against it. For anyone interested in a deeper look at this question, I would recommend Anitchkov's (often spelled "Anitchkow") article in Cowdry's 1933 collection, Arteriosclerosis: A Survey of the Problem, which one would most easily obtain through interlibrary loan.
On the other hand, Ravnksov's demonstration that PUFAs are guilty and saturated fat is innocent is much more convincing than Steinberg's attempt to do the opposite.
The major competition for Ravskov's skeptic-side argument is Anthony Colpo's The Great Cholesterol Con. The differences between the two books can be summarized as follows:
Colpo's book contains many more references, whereas Ravnskov's book takes a much deeper look at each individual reference.
Colpo spends half the book answering the question "what does cause heart disease?" whereas Ravnskov's book focuses much more tightly on the cholesterol question.
Colpo focuses more on politics whereas Ravnskov, while addressing politics in several places, focuses much more tightly on the science.
Ravnskov has more credentials (MD, PhD), which is useful for convincing certain types of people.
Ravnskov's book is more professional in tone, contains funny comics, and is easier to read because of its layout (Colpo's book has very small print).
Uffe Ravnskov's The Cholesterol Myths provides a rich and thought-provoking analysis that, like any other argument, should be evaluated critically, but is an important addition to the library of anyone interested in the cholesterol controversy.