QUICKLINKS AND VIEW OPITONS
Cholesterol Myths by Uffe Ravnskov MD: Myth 6: Lowering Cholesterol Will Lengthen Your Life (Part 2)
Wednesday, November 12, 2014 10:16 am Email this article
An expedient byproduct
Parallel with the mentioned study of healthy men the Finnish researchers performed another experiment on men who already had had a heart attack. About 600 such individuals participated, all of them worked at the same companies as those in the original Helsinki study.
The result after five year was disheartening. Seventeen of those who took gemfibrozil had died from a heart attack; compared to only eight in the placebo group.
Dr. Frick and his coauthors were eager to stress, that this difference was most probably a product of chance. In the summary of the paper they wrote: the number of fatal and non-fatal heart attacks did not differ significantly between the two groups.
They were right, because in contrast to their fellow-directors of the other trials they used the correct formula for determining the effect of a treatment, the two-sided t-test. If they had used the one-sided test as diet-heart supporters usually do when the allegedly positive effects are measured, significantly more had died in the treatment group.
But they had modified the result in another way. In the group “cardiac deaths” they had included a small group called “unwitnessed death.” That death is unwitnessed means that we do not know the cause of the death. It is not self-evident that an unwitnessed death is due to a heart attack and such deaths should of course have been classified otherwise.
If they had excluded the unwitnessed deaths there were more than three times more fatal heart attacks in the treatment group; sixteen against five. And this difference was indeed statistically significant.
The directors of the study admitted that the result was not “in accord with previous experience,” but they had a number of explanations.
As the trial was only “an expedient byproduct” of the original trial the number of individuals had been too small to give reliable results, they said. They were especially concerned about the low number of heart attacks in the control group. It was unlikely that it reflected the incidence in the general population. Most probably the individuals in the control group by chance had been less affected by coronary atherosclerosis than those in the treatment group.
Short guts and long lives?
At the University of Minnesota Medical School the surgeon Dr. Henry Buchwald had a bright idea. He had noted that when the last part of the small intestine is taken away from a patient (because of cancer or another disease) his blood cholesterol level decreased sharply. The explanation is that much cholesterol and bile acid is taken up in this part of the intestine, and as cholesterol is used for the production of bile acid in the liver considerable amounts of cholesterol are lost in the stools after the operation. Could this same operation be used to treat patients with too much cholesterol in their blood?
In 1963 Dr. Buchwald and his team performed the first ileal bypass to lower cholesterol. At this operation the last third of the small intestine, the ileum, is cut and closed, and the open end of the upper two-thirds is connected with the large intestine. Many such operations have been performed since then, mainly on patients with familial hypercholesterolemia. Unfortunately, only few researchers have studied the effect of the operation in a controlled study.
Two of them were Dr. Pekka Koivisto and Dr. Tatu Miettinen from Helsinki, Finland. Twenty-seven patients with familial hypercholesterolemia had this operation performed and after ten years their course was compared with twenty-seven control patients matched for a large number of the usual risk factors for coronary disease and treated with cholesterol lowering drugs only.
The ileal bypass was indeed effective, more effective than the drugs. The final level of cholesterol was 23 percent lower in the operated patients and the LDL-cholesterol was even “better.” But there was no difference as regards the outcome. After ten years five of the operated patients and four of the controls had died from a coronary, and three patients in each group had had a non-fatal coronary.
In spite of this disappointing result Dr. Miettinen and his colleagues recommended ileal bypass as a treatment against high cholesterol. They saw the operation as a partial success because those who had had a coronary in the bypass group had only lowered their cholesterol by only 25 percent while those who hadn’t had a coronary had lowered their cholesterol by 33 percent. Obviously they meant that the lowering should have prevented these heart attacks had it been more pronounced.
But it is difficult to see how the cholesterol level had any importance at all, because in the control group the lowering was about ten percent both in patients who had had a coronary, and in those who had not, and cholesterol in controls who hadn’t had a coronary was almost 20 percent higher than in the operated patients who had suffered one. Thus, if anything, the bypass operation had induced heart attacks instead of preventing them.
Dr. Buchwald himself, the inventor of the bypass operation, has conducted the largest trial of ileal bypass. In cooperation with 23 colleagues and 51 advisers he studied more than 800 middle-aged patients, mostly men, who had had at least one coronary. Half of them were randomized to ileal bypass, the other half were controls.
After ten years 32 in the surgery group had died from coronary heart disease against 44 of the controls. In all, 49 had died in the surgery group, 62 in the control group. These differences were far from statistical significance; they could have been due to chance. But in a subgroup analysis Dr. Buchwald and his co-authors found that if only those who had suffered a less serious coronary initially were considered, the difference was almost statistically significant. (But among those who had had a more serious coronary initially, more had died in the surgery group.)
There were other bright spots. In the control group there were more non-fatal coronaries, more attacks of severe angina, and many more patients underwent an operation to get a new coronary, a so-called coronary-artery bypass grafting. If all these events were taken together the difference between the two groups was highly significant.
Apparently a success. However, a study like this is of course neither single nor double blind. It is necessary to remind you of Professor Cornfield’s and Dr. Mitchell’s conclusion from their early overview of cholesterol lowering trials: open trials are successful, blind trials are not. To decide whether a patient has had a coronary or something else, or whether a patient should have a coronary graft or not is of course a highly subjective matter. You must be divine to avoid irrational motives from influencing your judgment in a million-dollars trial with so much glory and prestige at stake.
The authors argued that the higher rate of coronary grafting in the control group had improved their survival and blunted the trend toward a reduction in mortality in the surgery group. However, new coronaries may eliminate your angina, but most studies have shown that they do not prolong your life. On the contrary, a net excess of two deaths in the control group could be ascribed to complications of the coronary grafting, thus a further reduction of the difference.
Other complications were produced by the ileal bypass itself. Each year four percent had a kidney stone, a total of about 135 attacks at all, 14 had their gall bladder removed, and 57 had symptoms of bowel obstruction, 15 of whom required an operation. And there was more.
Lack of the ileum means not only loss of bile acid. When bile acid is lost the fats transported with the bile are lost also and make the stools frequent and loose. Loss of fats means loss of calories. On average, ileum bypass patients had a weight loss of 5.3 kg (11.7 lbs).
You may probably recall Dr. Kannel’s words about the “formidable risk” of coronary heart disease which is added by obesity. An ileal bypass is an effective treatment against obesity, and obese patients must therefore have been in great excess in the control group. It is difficult to know how many extra heart attacks among controls were due to obesity, but at least this bias should have been mentioned in the discussion and in the summary of the report.
Can atherosclerosis disappear?
Trials including thousands of individuals are laborious and of course utterly expensive. In recent years scientists have taken a shortcut. Instead of coronary deaths they have used regression of coronary artery disease as a measure of treatment effect. By regression they mean a widening or at least a less rapid narrowing of the coronary arteries as seen on coronary angiography. An increase of the mean diameter of the coronary vessels during treatment is said to be due to disappearances of atheromas, the scientific name of the vascular lipid deposits seen in atherosclerosis. Angiographic trials are much cheaper because much fewer test individuals are necessary and the result is possible to evaluate after a much shorter time.
Laboratory changes instead of number of deaths as a measure of treatment effect is called surrogate outcome. The term surrogate is used because it is not self-evident that laboratory changes can be translated to clinical effects such as lowering of mortality. It can be questioned also, if a widening of a coronary vessel seen on angiography means disappearance of atheromas and nothing else, but let me come back to this question a little later. Let us first have a look at the angiographic trials.
The National Heart, Lung and Blood Institute supports the diet-heart idea to one hundred percent. Together with the American Heart Association they administer more than 90 percent of all grants for cardiovascular research. Also on the American Heart Association they are convinced about the danger of cholesterol. In fact, most of those who have introduced the cholesterol campaign or have advocated it most vigorously are members, or have previously been members, of these institutions.
On the National Heart, Lung, and Blood Institute they decided to study the effect of cholesterol lowering directly on x-ray angiograms. To lower cholesterol they had chosen cholestyramine, the same drug that was used in the ongoing LRC trial. In five years 116 male patients with coronary heart disease and high blood cholesterol were treated; one half were given cholestyramine, the other half a placebo drug.
The result was again disappointing. In the treatment group the coronary arteries widened a fraction of a millimeter in four patients, but they widened also in four of the untreated patients.
Before the trial had started the investigators had decided to analyze their results using the one-tailed t-test that is scientifically unacceptable if the outcome can be both positive and negative. On the National Heart, Lung, and Blood Institute however, they said, that it was OK to use it because the weight of laboratory and epidemiological evidence suggested that reduction of blood cholesterol would retard coronary artery disease. The result could only go in one direction.
If you haven’t skipped too many chapters you will probably agree with me that the weight of laboratory and epidemiological evidence suggest nothing of the kind. Let me only mention that when the study they initiated was published in 1984 not fewer than seven controlled cholesterol lowering trials had resulted in an increase of coronary mortality in the treatment groups.
By using the one-tailed t-test and by putting the figures together in different ways the fifteen directors of the study headed by Dr. John Brensike found a combination that gave statistical support for the benefit of the treatment. They admitted that the result was not exactly what they had expected, but they returned in a further paper stating that the improvement was proportional with the changes of blood cholesterol they had seen in the patients independent of whether they had been treated or not. With other words, the coronary vessels had less often worsened if the cholesterol was low than if it was high.
It is easy for cholesterol researchers to get caught up in circular reasoning. We do not yet know the cause of atherosclerosis. What we do know is that high cholesterol is a risk factor. Theoretically high blood cholesterol may be the cause, but as I have said before high cholesterol may be secondary to the real cause; the causative factors may have induced atherosclerosis and at the same time it may have raised the level of cholesterol in the blood. The aim of the study was to see which of these alternatives were true by lowering blood cholesterol. Thus, the only valid finding is a possible effect of cholesterol lowering on atherosclerosis in the treated patients. If cholesterol is the bad guy a reduction in its concentration should be followed by a decrease in atherosclerosis, or at least by a halting of its progress.
If cholesterol is only an innocent bystander witnessing the crime and being influenced by it, then a reduction in its concentration would not have any effect because the unknown villain continues his activity. Crime is not prevented by killing the witnesses.
Supported by the National Heart, Lung, and Blood Institute and the drug company Upjohn Dr. David Blankenhorn and his group started a new angiographic trial, called CLAS, the Cholesterol-Lowering Atherosclerosis Study.
This study included 162 patients who had undergone coronary bypass operation. After routine laboratory tests all major arteries in the bodies were examined by angio?graphy. The patients were then randomly assigned to two groups of equal size. One group took cholesterol-lowering drugs, the other took ineffective placebo tablets. Neither the patients nor the doctors knew which group got treatment and which did not. To be sure that cholesterol was lowered properly, two drugs were given at the same time, colestipol and nicotinic acid.
After two years of treatment an angiography was performed again. In 16 percent of those who had been given the drugs but only in two percent of the control patients the coronary arteries had widened. In 38 percent of the treated patients the diameter of the vessels had decreased, but even this finding was seen as a success as the vessels had narrowed in still more of the controls, about 56 percent.
And the differences were statistically significant according to Dr. Blankenhorn and his colleagues, but only with the one-tailed test.
It may seem petty to take exception to the details of statistical formulas, and if the authors had underlined the weakness of their study and made reservations for the questionable results themselves no further discussion had been necessary. But the authors did not hint at any weakness. Their remarkable deviation from standard statistical practice was not even mentioned in the summary of the paper, nor when Blankenhorn’s study was cited in other publications. Furthermore, there was another problem.
The side effects of nicotinic acid, one of the drugs used in the study, are so obvious that no one can have any doubt of who is taking the drug, least of all the patient. Shortly after having swallowed the tablet the patient feels his skin is burning and itching as if stung by nettles. No doubt, the patient will tell about the side effects to his surroundings, including his doctor.
If the doctors still wanted to know for certain whether the patient had treatment or not they could look into the laboratory records. To be sure that the cholesterol really went down all patients had eaten the drugs during a period of three months before the start of the trial, and those whose cholesterol did not go down as much as was anticipated were excluded.
And cholesterol went down. On average blood cholesterol decreased by 26 percent in those who took the drugs; the “bad” cholesterol decreased by as much as 43 percent. On their regular visits at their doctors the patients might as well have had a sign around their neck telling to which group they belonged.
Thus, the trial was neither single nor double blind, which the authors also admitted. They called it “selective blind”; a new, but striking description of the condition on the branch of the research tree where the cholesterol hawks are breeding.
A researcher may be utterly impartial and dedicated to the truth, but he is probably not a saint. All experience tells that if the doctors knew to which group the patients belonged, their judgement must have been influenced in at least some of the cases no matter how much they tried to avoid this bias.
But Dr. Blankenhorn’s group was so certain of their results that they immediately called a press conference at the National Heart, Lung, and Blood Institute and with much fanfare announced their sensational findings. Now, at last, “for the first time” they had shown “a strong and consistent therapy effect from cholesterol lowering at the level of coronary arteries.”
Obviously, Dr. Blankenhorn had forgotten, that Dr. Brensike and his coworkers, claimed to have proven this several times. As you probably recall, it was Dr. Brensike’s “strong laboratory and epidemiological evidence that allowed them to use the one-tailed t-test.”
But Blankenhorn went further. He also claimed that his study demonstrated, that blood cholesterol should be lowered to a level of 185 mg/dl and this was also the limit that was set at the National Heart, Lung, and Blood Institute and the American Heart Association after Blanken?horn’s study. What an amazing perspective! Remember that even the strongest supporters of the diet-heart idea do not believe that diet is sufficient to lower cholesterol as much as in Blankenhorn’s trial; instead drugs must be used. How many adults have blood cholesterol above 185 mg/dl (4.85 mmol/l)? According to Dr. Basil Rifkind, one of the strongest advocates for the diet-heart idea, it probably amounts to 40 million healthy Americans. Said Robert Levy, former head of the American Heart Association, the results were “exciting” even if they were not quite unexpected. And the directors of the drug producer, Upjohn Company must have been delighted.
In a scientific report, it is customary to include discussion of the results of other investigators, especially when they deviate completely from one’s own results. How did Dr. Blankenhorn and his coworkers and the brain trust at the National Heart, Lung and Blood Institute comment on the disheartening results of Drs. Bemis, Kimbiris, Shuh, Kramer and their groups, who didn’t find the slightest connection between changes in cholesterol deposits in the coronary vessels and changes in cholesterol levels in the blood? How did they explain the fact that the coronary vessels improved in their own experiments, but not in the many previous studies where cholesterol went down just as much or more? Why did they place more importance on their own dubiously positive results than on the many indisputable negative ones? I cannot give you an answer because they did not comment on them at all.
Three years later, in 1990, the results from a new angiographic trial was published by Dr. Greg Brown and his team in Washington, again supported by the National Heart, Lung, and Blood Institute. Obviously they were not impressed by the “strong and consistent” effect of Blankenhorn’s treatment. In their paper they wrote that the lipid changes in the previous trials were small and the clinical benefits limited. Therefore Dr. Brown and his colleagues had used two drugs at the same time. Thirty-eight men received lovastatin and colestipol, 36 received nicotinic acid and colestipol and 46 received placebo, and the trial was designed as a double-blind study. Most of the men participating had familial hyper?cholesterolemia.
And indeed cholesterol was lowered; by 34 percent in the lovastatin-colestipol group and by 23 percent in the niacin-colestipol group. The devil himself, LDL-cholesterol was lowered even more, by 45 percent and 32 percent, respectively.
Before and after treatment the width of the coronary arteries were measured at various points, using a fivefold magnification of the x-ray pictures. Magnification was necessary, because on average, the vessel diameters in the niacin-colesti?pol group increased by a mere 0.04 mm only, whereas the diameters in the lovastatin-colestipol decreased by 0.002 mm and in the control group by 0.05 mm. Indeed small differences, but they were statistically significant, and Dr. Brown and his colleagues saw them as a proof of therapeutic success.
I am sure you noticed that the vessel diameters decreased in those treated with lova?statin-colesti?pol, those who had their cholesterol lowered the most. In some places of the arteries the diameter had increased, but on average, taking all measurements together, the diameters had decreased. A decrease of the diameters means of course that the coronary vessels had become narrower, which is certainly not an improvement, although the diameter in the control group decreased even more. The authors had no comments about this striking finding, neither about the fact that the only two deaths in the study, and the only heart attack, were seen in the lova?statin-colestipol group. In fact, the title of their paper said the opposite: “Regression (improvement) of coronary artery disease as a result of intensive lipid-lowering therapy…”
The anguish of angiography
Let us still have in mind, that a change of the coronary diameter is nothing but a surrogate outcome. It is assumed that a widening of a coronary vessel on an X-ray means less atherosclerosis and thus a better chance to avoid a heart attack, but this is only an assumption.
Artery walls are surrounded by smooth muscle cells. When such cells contract, the artery narrows. When they relax, it widens. Various factors may stimulate the smooth muscle cells of the coronary arteries. Most important, mental stress, anxiety, exposure to cold, and even a sustained handgrip may lead to contraction. The latter effect was studied six years earlier by Dr. Greg Brown, the same Dr. Brown who led the angiogra?phic trial mentioned above. He found that a handgrip sustained for a few minutes was followed by a 35 percent decrease of the vessel diameter.
Consider that the changes seen in the trials were only a few percent on average. What do you think you would do yourself if somebody were to put a long catheter all the way from your groin up to your heart and into your coronary vessels? If you are not a stuntman or an astronaut I think that you probably would have gripped the nurse’s hand or something else very tightly, at least during the first examination. How, then, did the researchers know whether an increase in blood vessel diameter at the second examination was due to the patient being more relaxed or to the vessels being less atherosclerotic?
Also, drugs which relax the coronary vessels, and which are used by almost all coronary patients, may have disturbed the study. In the trial Dr. Brown and his coworkers were aware of that problem. The use of such drugs was “duplicated as exactly as possible.” This can’t have been too easy because the level of any drug in the blood depends on a large number of factors which are difficult to standardize. And Dr. Brown and his colleagues didn’t write anything about duplicating possible handgrips or anxious feelings because such duplication is, of course, impossible. So, any factor which may influence the state of the smooth muscle cells in the coronary vessels may have influenced the vessel diameter much more than the possible appearance or disappearance of a tiny amount of cholesterol.
There are more uncertainties. Dr. Seymor Glagov and his colleagues from University of Chicago studied the hearts of 136 deceased individuals and found that when vessels become sclerotic, they widen to compensate for the narrowing brought about by the deposition of cholesterol in their walls. In fact, this widening overcompensates for the deposition until the cholesterol deposits occupy about 40 percent of the area beneath the muscle wall. Only thereafter does the vessel become steadily narrower. In other words, an increase of vessel diameter may be due to disappearance of cholesterol in a highly sclerotic vessel, but also to a compensatory widening during the first stages of cholesterol deposition. How could the trial directors know whether the increase of vessel diameter was due to a disappearance of deposited cholesterol, or to a compensatory widening due to an appearance of deposited cholesterol?
Risk factors and coronary vessel
The fact, that a lowering of cholesterol may reverse sclerotic lesions also conflicts with the results from a number of previous, angiogra?phic long-term studies of the coronary arteries. The aim of those studies was to explore which factors dictated the development of atherosclerosis. Were the risk factors primary or secondary? Should they be sought among the mercenary troops or were they something which followed the vestige of war as hunger and cholera? Was smoking of any importance? Did the blood pressure influence the development of atherosclerosis? Did a high blood cholesterol?
According to the conventional wisdom, atherosclerosis should increase if the cholesterol over a longer period of time is high or if it goes up, no matter why. Likewise, atherosclerosis should decrease, or at least it should not increase, if cholesterol is low or if it goes down.
One of the first who studied the inside of the coronary vessels with such questions in mind was Dr. Charles Bemis. The year was 1973. Together with his team at the Peter Bent Brigham Hospital in Boston he studied about seventy patients and found that the only factor which could be connected with the degree of atherosclerosis was the level of the blood lipids at the start of the investigation. In patients with high blood cholesterol at the start, coronary atherosclerosis had increased at the following angiographic examination a couple of years later.
So far, the results were as anticipated. Once again, it was demonstrated that high blood cholesterol is a risk factor for coronary heart disease. Now to the interesting finding.
In twenty-four patients, cholesterol had decreased by more than 25 percent between the two angiographies, a lowering which was considerably greater than in most cholesterol lowering trials. Among these twenty-four patients, atherosclerosis had increased in sixteen, while it was unchanged in just eight. In twelve patients cholesterol had increased, but only in four of them had atherosclerosis increased.
Said in another way: two out of three whose cholesterol went down had become more sclerotic, while this was the case in only one out of three whose cholesterol went up. It should, of course, have been the other way around.
Dr. Bemis’s result was confirmed the following year by Dr. Demetrios Kimbiris and his group in Philadelphia. These investigators also found that cholesterol was unimportant. The coronary arteries of seventeen out of twenty-five patients with high cholesterol had worsened, but they had also worsened in seven out of ten patients with a low cholesterol.
Similar results were achieved at the famous Mayo Clinic by Dr. Clarence Shub and his colleagues. They found that coronary atherosclerosis had increased in all patients whose cholesterol had decreased by more than 60 mg/dl, a lowering which should have been considered more than acceptable in any cholesterol-lowering trial.
In study after study the startling finding of Bemis and his colleagues was confirmed. In their paper, Dr. John Kramer and his colleagues at the Departments of Cardiology and Biostatistics, The Cleveland Clinic Foundation, concluded: “… medical treatment directed toward ‘secondary prevention’ may be unsuccessful in retarding or reversing the development of progressive arterial lesions and their clinical consequences.”
But nobody listened. More prudent scientists should have questioned the diet-heart idea, facing the fact that coronary atherosclerosis is worsened just as fast or faster when cholesterol goes down as when it goes up.
In a scientific report it is a rule to discuss also the results of other investigators, especially when they deviate completely from one’s own results. How did Blankenhorn and his coworkers and the brain trust at the National Heart, Lung, and Blood Institute comment on these disheartening ?results? How did they explain that the coronary vessels improved in their own experiments, but not in the many previous studies where cholesterol went down just as much or more? Why did they place more importance on their own dubiously positive results than to the many indisputable negative ones?
I cannot give you an answer because nothing was said about them.
Recently, a team led by Canadian Dr. David Waters published yet another study including 335 patients. They found that when the coronary vessels after a two year interval had narrowed by more than 15 percent, the risk for cardiac death or a non-fatal coronary increased considerably. Certainly not an unexpected finding. Their conclusion was that the changes seen on coronary angiography was a good substitute for cardiac events in clinical trials.
But they didn’t comment that on average the cholesterol of those whose atherosclerosis had progressed did not differ significantly from the cholesterol of those whose atherosclerosis had not progressed. Remember that the cholesterol in the angiographic trials went down by 30-49 percent, but the change of the vascular diameter was much less than one percent. In Waters’s study the diameter change was more than 15 percent, but the decrease of cholesterol an insignificant two percent. Such results do not suggest that the diameter changes have anything to do with the cholesterol changes.
Certain treatments are easy to assess. The right antibiotic, for instance, will cure nine out of ten women with an uncomplicated urinary infection, which means that after having treated fewer than a dozen patients and controls for a few days you already know for certain that the drug is effective. But after the many cholesterol-lowering trials, scientists still don’t know whether the treatment could change mortality. Statisticians say that to prove a beneficial effect on mortality, many more test individuals are necessary, probably more than 100,000. If the beneficial effect of treatment is so difficult to prove, aren’t we justified in concluding that high cholesterol cannot be that dangerous for our health?
But the problem may be solved in another way. The solution is called meta-analysis.
In a meta-analysis, data from all studies that satisfy certain standards of quality are put together in the hope that they will provide a large enough sample for statistical reliability. For medical trials, at least, three standards are mandatory. Trials should be double-blind, they should be controlled, meaning that on average, all risk factors are similar in the two groups, and the test individuals and the controls should be chosen randomly. Also, in order to use the accumulated results from many trials, it is necessary for the same kind of treatment to have been used in each trial and, of course, the result of the treatment—the outcome or the end point—should also be the same.
By now you may have realized that if all standards are to be satisfied, very few trials can qualify. Very few trials have been performed in a true double-blind fashion, and besides cholesterol-lowering, some trials have also used other kinds of intervention. Nevertheless, let us have a look at the entire body of trials that have been published before 1992, all of them performed before the introduction of the statins.
Several meta-analyses on cholesterol-lowering trials have been published. When I prepared the first edition of this book in Swedish it annoyed me that most of these analyses had excluded a number of trials, preferably the unsupportive ones. So I decided to perform a meta-analysis myself that included all randomized and controlled trials where the aim had been to lower cholesterol, whether by diet, or by drugs, or whether they had used other kinds of intervention also. I accepted open trials, since it was not possible to do a fair selection of double-blind studies because even trials that were designated double-blind were in fact more or less open for the reasons I have given above. In table 6B the raw figures from this analysis are given.
Table 6B. Overall result of 26 controlled cholesterol-lowering trials. The number of individuals in the three calculations are not identical because a few trials did not give the number for all end points.
|Number of individuals
|Non-fatal heart attacks; percent
|Number of individuals
|fatal heart attacks; percent||2.9
|Number of individuals
|Total number of deaths; percent
As you see, the number of deaths from a heart attack was equal in the treatment and in the control groups, and the total number of deaths was greater in the treatment groups. In one study total mortality had decreased significantly, in two others it had increased significantly, and in no trial was coronary mortality changed more than could be attributed to chance. More sophisticated calculations did not change the picture.
There was a small reduction in the number of non-fatal heart attacks. Calculated in the way diet-heart supporters usually do the difference was 10.4 percent; calculated the simple way, the difference was 0.3 percent. Due to the large number of individuals studied this small difference was statistically significant, but most probably it was a result of bias. Not only were the trials open and partly multifactorial; there was another finding that definitely proved that cholesterol lowering does not make any benefit.
If cholesterol lowering could reduce the risk of coronary heart disease, a pronounced and prolonged lowering should of course lower the risk more than a slight and short one. But there was no relationship between the degree of cholesterol lowering and any of the end points, not between individuals in each trial and not between trials. And on average, total mortality was equal in short and long trials, and coronary mortality was higher in long trials than in short ones.
So, although some of the trials also included physical exercise, weight loss, reduction of blood pressure, and smoking advice, and although most trials were open, the number of non-fatal heart attacks was not reduced by more than 0.3 percent. And even if the doctors had been totally uninfluenced by their knowledge about the patients’ group affiliation, remember that what doctors—even experienced ones—call a heart attack very often is something else.
After its publication in the British Medical Journal, my meta-analysis provoked harsh comments from diet-heart supporters. According to my critics, the most serious mistake was to include trials using hormones, since such drugs are now considered toxic to the heart. But in one of the first controlled trials published, conducted by Professor Jeremiah Stamler, the researchers used low doses of the female sex hormone estrogen and that trial had the best result of all. High doses of estrogen are possibly harmful for men, but whether low doses are harmful is an open question. In women, at least, low doses, such as those used in post-menopausal hormone replacement therapy, seem to protect against heart attacks. The results were also just as unsupportive in the subgroup calculations, even in subgroups that did not include the hormone trials.
Another objection was that I had ignored the angiogra?phic trials, because they “can lead to regression of atheroma,” as one of the critics noted. Let us therefore look at a more recent meta-analysis which included the angiographic trials.
In this analysis, Dr. George Davey Smith at the Department of Public Health, University of Glasgow, Scotland, and his coworkers excluded the multifactorial trials to study the effect of cholesterol lowering only, and they limited their analysis to total mortality. They ranged the trials in order of risk according to the coronary mortality in the control groups. In high-risk trials, many control individuals had died from a heart attack, in low-risk trials relatively few.
In the high-risk trial group (including 5,115 individuals in ten trials on the uppermost part of the risk list) mortality had decreased. In the median-risk group (including 24 090 subjects in 15 trials), mortality was unchanged. In the low-risk group, the largest one (including 27,918 subjects in ten trials on the lower-most part of the list), mortality had increased. Both the decrease and the increase of total mortality was statistically significant. Overall, in trials where drugs had been used to lower cholesterol, mortality from non-coronary causes had increased significantly. The authors’ conclusion was that benefits from cholesterol lowering drugs seem to be produced in only a small proportion of patients at very high risk of death from coronary heart disease. Thus, in the future cholesterol lowering should include only individuals at very high risk. But here a problem appears.
Individuals in the so-called low-risk group were only at low risk compared with individuals in the other two groups. In comparison with normal individuals, they were at high risk also. For instance, their cholesterol was 278 mg/dl (7.13 mmol/l) on average, higher than in the other two groups! Remember that at that time the lower limit for drug treatment according to the early recommendations of the National Heart, Lung, and Blood Institute was 185 mg/dl (4.75 mmol/l); according to the cholesterol campaign it was 240 mg/dl (6.15 mmol/l). Furthermore, half of the trials in the so-called low-risk group were of the secondary preventive type, which means that they included patients who already had suffered a coronary. Such patients always have been considered as being at high risk. Many of the individuals in the low risk group were exposed to other risk factors also, so indeed this group was a sample of high-risk individuals. How should we discriminate between these high-risk individuals and the high-risk individuals who are said to prosper from cholesterol lowering? The simple fact is that we can’t. Even doctors who treat high-risk individuals only may shorten the lives of their patients instead of prolonging them.
It is also questionable if mortality really was lowered in the high-risk trial group. One of the trials for instance, was in fact a multifactorial trial that had been included by mistake. The good result in that trial could therefore have been caused by something other than cholesterol lowering. The other trials were very small, and in only one of them was mortality lowered significantly. There was only one reasonably large trial in the high-risk group, but in that trial mortality had increased.
Anyone who has read scientific papers or official recommendations about cholesterol and the heart know that we are told another story. Listen, for instance, to the most recent recommendations of the European Atherosclerosis Society: “Clinical trials of secondary prevention by lowering plasma cholesterol, when studied together by meta-analysis, show that morbidity and mortality from coronary disease are reduced; there is also a trend to lower total mortality.”
This misleading statement is not unique; in fact, it is typical of diet-heart writings. Similar statements are found in numerous scientific papers from the supporters, and from the mass media.
There is more than one explanation for the inappropriately optimistic messages from doctors and scientists. Most important, scientists prefer to cite only the supportive trials. I have already told about the few citations from the unsupportive trial by Miettinen and the many citations from the allegedly supportive LRC trial. On average, I found that trials considered supportive by their directors have been cited almost six times more often than unsupportive trials. The fact that a trial was cited frequently had little to do with its quality or whether it had been published in a famous or a less well-known journal. The trial directors themselves were especially unwilling to cite an unsupportive trial; since 1970 up to 1992, no trial considered unsupportive by its directors had been cited in another trial report. Even authors of meta-analyses had selected their trials according to their outcome.
A successful dietary trial
The idea that a Mediterranean diet—whatever that is—would be beneficial for cardiovascular disease inspired the French researcher Dr. Michel de Lorgeril and his team from Lyon, France, to start a new dietary trial, the Lyon Diet-Heart study. About 600 patients who had survived a first heart attack were included. Half of the patients were instructed to adopt the so-called Mediterranean diet by including more bread, more root and green vegetables and more fish, and by reducing consumption of animal fat and red meat. They were also instructed to eat fruit every day, to replace pork with poultry and to replace butter and cream with margarine, which was supplied free for the whole family. In contrast to the previous trials, where the dietary fat was dominated by vegetable oils with a high content of omega-6 polyunsaturated fatty acids, these researchers used a margarine made with rapeseed oil, which has a high content of ?-linolenic acid, a polyunsaturated fatty acid of the omega-3 class. This special type of margarin was supplied free for the whole family. Control group individuals were also given dietary advice, but were recommended the usual “prudent” diet.
This design was chosen because in a previous study people from Crete, the Greek island where heart attacks were rare according to the Seven Countries study, had three times more of this fatty acid in their blood than the people from Zutphen, the Netherlands. The French researchers therefore thought that ?-linolenic acid might be protective because the rate of heart attacks was much higher in Zutphen than on Crete although their cholesterol levels were almost identical.
After forty months, a significant difference was found; in the control group 20 individuals had died, compared to only eight in the treatment group. Most surprising was, that in the control group eight of the patients who died from a heart attack had died suddenly, which almost always means that they have died from a disturbance of the heart rhythm. This was not seen in any of those who died in the treatment group. After four years the trial was ended because the improvement in the treatment group had continued; 24 had died in the control group but only 14 in the treatment group. And the difference between the number of nonfatal heart attacks was even larger—25 in the control group, but only eight in the treatment group—and it was therefore considered unethical to continue the trial.
For the first time a dietary trial had succeeded in lowering the risk of dying from a heart attack. Evidently it was a good idea to lower cholesterol by dietary means.
But blood cholesterol was practically identical in both groups after the trial had ended. In fact, it was a little higher in the treatment group.
If the difference couldn’t be explained by the participants’ cholesterol, what else could? Was it the additional ?-linolenic acid or was it the extra fruit or vegetables? Or the extra bread? Or the extra fish or chicken? Or, as both fish and ?-linolenic acid is rich in omega-3 polyunsaturated fatty acids, perhaps a better balance between omega-3 and omega-6 polyunsaturated fatty acids? According to the food frequency questionnaire at the end of the study the intake ratio of omega-6/omega-3 in the control group was about 20/1, in the intervention group about 4.5/1.
There is much evidence, both from animal experiments and epidemiological studies that a high ratio between omega-6 and omega-3 polyunsaturated fatty acids may predispose to heart arrhythmia, the main cause of sudden death in patients with heart disease. A few years later a new trial gave further support to that idea.
The GISSI-Prevenzione trial
One of the largest controlled, dietary trial was started in Italy, named Gruppo Italiano per lo Studio della Sopravvivenza nell’Infarto miocardico Prevenzione trial, or GISSI. More than 11,000 patients who had survived a first myocardial infarction were enrolled and divided into four equally large groups. One group were treated with a capsule containing a mixture of three different omega-3 polyunsaturated fatty acids, one group a capsule with vitamin E, one group were given both capsules, and the fourth group was used as control. All of them received standard treatment for coronary patients.
On average the patients were followed for 40 months. At follow-up 3.5% in the control group had died from sudden death, 2.3% in the vitamin E group, 2.4% in the combined group, and 1.9% in the omega-3 group. Except for the result in the vitamin E group these differences were statistically different; indeed, the difference between the control and the omega-3 group was greater than in any of the following statin trials. Relatively seen, mortality due to sudden death was 46% lower than in the control group. Also the total number of deaths was significantly lower, and again, no differences were seen between the groups as regards blood cholesterol.
Interestingly, no differences were seen between the number of non-fatal heart attacks, a further argument for the idea that the omega-3 polyunsaturated fatty acids primarily protect the nerve conduction system of the heart.
Those who still believe that high cholesterol is the main villain perhaps may argue that the disappointing results from the trials I have discussed in this chapter may be because we haven’t lowered cholesterol sufficiently. But to-day we have got a pharmaceutical method to lower cholesterol much more than by using any of the previous drugs; the statins. And according to the directors of the statin trial intensive cholesterol lowering by statin treatment is harmless and also an effective means to prevent cardiovascular disease.
Here are links to the other chapters in the book.
- Cholesterol Myths by Uffe Ravnskov, MD, PhD: Forward to Book by Michael Gurr, PhD
- Cholesterol Myths by Uffe Ravnskov, MD, PhD: Author’s Foreword
- Cholesterol Myths by Uffe Ravnskov MD, PhD: Introduction: The Diet-Heart Idea: A Die-Hard Hypothesis
- Cholesterol Myths by Uffe Ravnskov MD, PhD: Myth 1: High-Fat Foods Cause Heart Disease
- Cholesterol Myths by Uffe Ravnskov MD, PhD: Triglycerides
- Cholesterol Myths by Uffe Ravnskov MD, PhD: Myth 2: High Cholesterol Causes Heart Disease
- Cholesterol Myths by Uffe Ravnskov MD: Familial hypercholesterolemia—not as risky as you may think
- Cholesterol Myths by Uffe Ravnskov MD, PhD: Myth 3: High-Fat Foods Raise Blood Cholesterol
- Cholesterol Myths by Uffe Ravnskov MD, PhD: Myth 4: High Cholesterol Blocks Arteries
- Cholesterol Myths by Uffe Ravnskov MD, PhD: Myth 5: Animal Studies Prove the Diet-Heart Idea
- Cholesterol Myths by Uffe Ravnskov MD, PhD: Cholesterol lowering in children
- Cholesterol Myths by Uffe Ravnskov MD: Myth 6: Lowering Cholesterol Will Lengthen Your Life (Part 1)
- Cholesterol Myths by Uffe Ravnskov MD: Myth 6: Lowering Cholesterol Will Lengthen Your Life (Part 2)
- Cholesterol Myths by Uffe Ravnskov MD PhD: Myth 7: The Statins — Gift to Mankind (Part 1)
- Cholesterol Myths by Uffe Ravnskov MD PhD: Myth 7: The Statins — Gift to Mankind (Part 2)
- Cholesterol Myths by Uffe Ravnskov MD PhD: “The most exact data base”—the screenee
- Cholesterol Myths by Uffe Ravnskov MD PhD: Myth 8: Polyunsaturated Oils are Good for You
- Cholesterol Myths by Uffe Ravnskov MD PhD: Dr. Ornish and The Lifestyle Heart trial
- Cholesterol Myths by Uffe Ravnskov MD PhD: Myth 9: The Cholesterol Campaign is Based on Good Science
- Cholesterol Myths by Uffe Ravnskov MD PhD: Insider Insight
- Cholesterol Myths by Uffe Ravnskov MD PhD: Myth 10: All Scientists Support the Diet-Heart Idea
- Cholesterol Myths by Uffe Ravnskov MD PhD: Epilogue
- Cholesterol Myths by Uffe Ravnskov MD PhD: References
This chapter is from the book
The Cholesterol Myths: Exposing the Fallacy that Saturated Fat and Cholesterol Cause Heart Disease
by Uffe Ravnskov, MD, PhD.
Dr. Ravnskov has given me the permission to share this version of his book to help educate the world about the cholesterol campaign.
Information about Uffe Ravnskov, MD, PhD is posted here.
More information about Cholesterol Myths is posted on his website here.
Dr. Ravnskov posted his book for free here.
Uffe Ravnskov, MD, PhD is the founder of The International Network of Cholesterol Skeptics (THINCS.org) which can be found here.
Articles on the same subject can be found here:
Please feel free to share your comments about this article.
© Copyright 2003-2017 - Larry Hobbs - All Rights Reserved.