In 2001, the Expert Panel on Detection, Evaluation, and Treatment of High Blood Cholesterol in Adults issued perhaps the most influential document in the history of modern American medicine. Written as part of the National Cholesterol Education Program, the updated guidelines incorporated the findings of the most recent clinical trials into concise recommendations designed to assist doctors in reducing their patients’ risk of developing coronary heart disease (CHD). The recommendations are bold and offer the tantalizing hope that coronary heart disease in all but the very old will become a far less common occurrence. This goal can be reached, according to the guidelines, by increasing the number of Americans taking statin drugs, from 13 million to 36 million.
The new recommendations call for doctors to measure adult patients’ cholesterol and triglyceride levels every five years. A two-step assessment of risk is then suggested: First, “major risk factors” are identified, including cigarette smoking, high blood pressure, low HDL (good) cholesterol (less than 40 mg/dL), a strong family history of coronary heart disease, and older age—i.e., men who have reached the age of 45 and women who have reached the age of 55. For people with two or more major risk factors, the probability of developing coronary heart disease over the next 10 years is then calculated based upon a “risk score” developed from the findings of the Framingham Heart Study. If the risk of developing CHD over the next ten years is 10 percent or more, and the LDL cholesterol level remains 130 mg/dL or higher* after a trial of diet and exercise, the new guidelines call for treatment with a statin drug to head off coronary heart disease at the pass.
The excitement generated by these new guidelines was unprecedented. Dr. Claude Lenfant, the director of the National Heart, Lung, and Blood Institute, under whose auspices the NCEP does its work, told the New York Times that if the new guidelines were followed, coronary heart disease “would no longer be the number one killer [in the United States].” With even greater enthusiasm, the lead author of the guidelines, Dr. Scott M. Grundy, said, “These statins are amazing drugs.. . . When you say you can’t put that many people on drugs, you’ve got to balance that against the tremendous devastation of coronary heart disease.” This certainly seems like due cause for excitement.
The new guidelines were formulated by a panel of 14 experts and approved by representatives from 22 prestigious nonprofit medical societies, including the American College of Cardiology and the American Medical Association. The guidelines make specific recommendations for men, women, and people 65 and older who do not have coronary heart disease (“primary prevention”), and for those who already do (“secondary prevention”). An 11-page executive summary of the full report was published in the May 16, 2001, issue of JAMA. Virtually all cardiologists and primary care doctors who treat adults are familiar with these recommendations. But few have read the uninviting 284-page full-length version of the NCEP expert panel’s report, available on the Internet. Most doctors probably figured it wasn’t necessary. The summary assures its readers that the full document is “an evidence-based and extensively referenced report that provides the scientific rationale for the recommendations contained in the executive summary.”
The updated guidelines rely heavily on the findings of five large clinical trials of CHD prevention with statins that had become available since the previous version of the guidelines were issued in 1993. The executive summary published in JAMA is clear about its primary approach to CHD prevention: after a brief discussion of the changes introduced in the new guidelines and the new method of individual risk assessment, the bulk of the recommendations address LDL cholesterol as “the primary target of therapy.” Largely as a result of these guidelines, cholesterol control has become the main focus of preventive health care in the United States.
Doctors aren’t required to follow these (or any) guidelines to the letter, but most do for several reasons: they want to practice the best medicine possible; they want their medical decisions to be consistent with community standards; and they know that if they do not follow current guidelines, they have a greater risk of getting sued should something go wrong. For these reasons, plus the enormous publicity that continues to surround the issue of cholesterol, the updated guidelines are playing a larger role than any guidelines that have come before in shaping both doctors’ and patients’ health care priorities.
But careful scrutiny of the full report of the NCEP—and of the key studies supporting its recommendations—reveals a picture strikingly different from what is presented in the executive summary. Rather than presenting a balanced interpretation of the scientific findings, the report seems intent upon building the case for greater use of statins, preferentially presenting data that support greater statin use and even misrepresenting findings reported in the original articles. And rather than promoting a balanced approach to coronary heart disease prevention and overall health promotion, the guidelines seem more intent upon getting doctors to focus on lowering LDL cholesterol.
The link between elevated cholesterol and increased risk of heart disease was initially identified by researchers from the Framingham Heart Study. Beginning in 1948, this study enrolled 5000 residents of Framingham, Massachusetts, with the goal of identifying the factors that contribute to coronary heart disease. In 1957 the study first reported that high cholesterol levels increased the risk of heart disease. The different effects of HDL and LDL cholesterol were described in 1977. (Cholesterol is a waxy fat that is attached to a protein for transport in the blood. The difference between “good” cholesterol, HDL, and “bad” cholesterol, LDL, is the protein to which the cholesterol is attached.)
Coronary heart disease develops when one or more of the arteries that supply blood to the heart muscle become blocked, depriving the muscle cells of the oxygen and nutrients needed to function properly. The process starts when LDL cholesterol particles circulating in the blood pass into the walls of the coronary arteries. This sets off an inflammatory reaction that leads to the buildup of white blood cells and other material, collectively called plaque, within the walls of the arteries. HDL cholesterol, on the other hand, seems to pull cholesterol out of the arteries, like a scavenger, and transport it back to the liver.
The buildup of plaque on the inside of artery walls is a slow process, taking many years. Usually there are no symptoms until the internal diameter of a coronary artery becomes narrowed by at least 60 percent. Then, like a garden hose that is being stepped on, the capacity of the partially blocked coronary artery to carry blood to the heart muscle becomes compromised. When the blood supply to an area of the heart is inadequate to meet increased metabolic demand (during exercise or emotional stress, for example), patients often feel crampy, pressure-like pain in the left side of the chest, known as angina. When a coronary artery becomes completely blocked by plaque and there is no other blood supply to the downstream heart tissue, muscle cells die. This is known as a heart attack.
Most heart attacks are not, however, caused by a gradual buildup of plaque. The more frequent scenario is that a smaller area of plaque, for unknown reasons, breaks open (“fractures”) or becomes eroded on its surface. This causes the tiny platelets circulating in the blood to become sticky and form a small blood clot, or thrombus, on top of the plaque. Without any warning, thrombus formation can quickly and completely obstruct the flow of blood through a coronary artery, causing a heart attack. (Aspirin decreases the risk of heart attack by decreasing the stickiness of platelets, thereby making thrombus formation less likely.)
Lowering total and LDL cholesterol with medication, the theory goes, lowers the risk of coronary heart disease by decreasing plaque formation. From the 1960s to the 1980s, drugs of the fibrate class were used to lower cholesterol (though exactly how they worked was not known) and decrease the risk of heart attack. This they did quite well for people who did not yet have heart disease. But after many years of use, as a study done by the World Health Organization found, clofibrate (brand name Atromid-S) increased the overall risk of death by 47 percent. (About half the excess deaths were due to cancer.) Similarly, a study done by the National Public Health Institute at the University of Helsinki, Finland, showed that the death rate among people taking the other popular fibrate, gemfibrozil (brand name Lopid), for 8.5 years was 21 percent higher than that for the people taking placebos.
The current era of treatment to lower total and LDL cholesterol began in 1987, with the introduction of the first statin drug. Most of the cholesterol in our bodies does not come from our diet but is manufactured in the liver. Statins were developed specifically to mimic one of the intermediate molecules produced in the process of cholesterol synthesis, and thereby trick the liver into slowing down cholesterol production. This decreases the amount of total and LDL cholesterol circulating in the bloodstream. The first statin, Mevacor, is now available as a generic called lovastatin, which costs less than half as much as the brand-name statins. The best-selling statins as of 2003 were Lipitor, Pravachol, and Zocor. The newest entry into the U.S. statin market, Crestor, was approved by the FDA in August 2003.
It is important to keep in mind that cholesterol is not a health risk in and of itself. In fact, cholesterol is vital to many of the body’s essential functions. For example, cholesterol is the most common organic molecule in the brain (this could explain why statins have a small but statistically significant negative effect on cognitive function). It is also an essential building block of many of the body’s most important hormones, such as stress hormones, blood sugar–regulating hormones, and sex hormones. (One study that specifically looked for sexual problems associated with cholesterol-lowering therapy found that statins increase the frequency of sexual dysfunction by about 50 percent in men.) Cholesterol is also necessary for the transmission of signals from one nerve cell to the next, and is an integral component of the membrane that surrounds each cell.
The essential role that cholesterol plays in many of the body’s biological functions is easily forgotten when one is reading and following the updated guidelines for cholesterol management, and getting caught up in the growing cholesterol-lowering frenzy. The real goal of medical care is, after all, to improve overall health—in this case to decrease the risk of coronary heart disease, serious illness of all kinds, and premature death from all causes—and not simply to lower blood levels of LDL cholesterol.
Even though much of what we know about the relationship between cholesterol and coronary heart disease comes from the Framingham Heart Study, the mother of all cholesterol studies, some of its most important findings will come as a surprise—especially to doctors. An article published in the Archives of Internal Medicine in 1993 analyzing data from the Framingham study showed that higher total cholesterol levels significantly correlate with an increased risk of death from coronary heart disease only through the age of 60. This correlation does not extend to age 70 or beyond. More important, the article showed that elevated total cholesterol levels correlate with an increased overall risk of death only through the age of 40, and not once the age of 50 is reached. An even more alarming finding from this study, given the current cholesterol-lowering craze, is that the risk of death from causes other than coronary heart disease increases significantly with lower total cholesterol levels for men and women after they reach the age of 50. (The authors specifically considered but rejected the hypothesis that people with lower cholesterol levels had higher noncardiac death rates because of undiagnosed underlying illness.) Other data from the Framingham Heart Study published in 1999 show that physical activity, unlike total cholesterol levels, is highly correlated with overall mortality rate: The most active third of the original 5000 men and women in the study had a 40 percent lower death rate than the least active third.
So why has our collective national attention become so narrowly focused on lowering LDL cholesterol as the single most important preventive health strategy when the evidence shows that it plays a relatively limited role in our overall health? Part of the answer has to do with the experts’ intellectual commitment to the role of cholesterol in heart disease, having dedicated their careers to furthering the scientific understanding of this relationship. Another part of the answer may have to do with some of the authors’ potential conflicts of interest: Five of the 14 experts who participated in writing the guidelines, including the chair of the panel, disclosed financial relationships with manufacturers of statin drugs. Four of these five, including the chair of the panel, had relationships with all three manufacturers of the best-selling statins.
And curiously, although the guidelines recommend reduced intake of saturated fat and cholesterol, the words “egg,” “beef,” and “dairy” do not appear anywhere in the executive summary. (Animal products such as egg yolks, red meat, and dairy fat are the primary dietary sources of saturated fat and cholesterol, and therefore reducing intake of these foods is an integral part of the “therapeutic lifestyle interventions” suggested in the guidelines.) Perhaps the omission has something to do with the fact that, according to the Center for Science in the Public Interest, several of the authors and expert reviewers of the guidelines have, or have had, financial ties to one of the following organizations: the American Egg Board, the National Cattlemen’s Association, and the National Dairy Promotion and Research Board. In contrast, increasing fiber intake is mentioned five times in the executive summary. (One of the authors and two of the reviewers have done research on fiber funded by Procter and Gamble, the manufacturer of Metamucil.)
The drug companies have plenty to gain from the pro-drug orientation of the updated guidelines. If the guidelines are followed, sales of cholesterol-lowering statin drugs will increase by at least $20 billion to $30 billion per year. (The total cost will be even more because of the extra doctor visits and blood tests necessary to make sure that the drugs are not causing side effects such as inflammation of the liver or breakdown of muscle tissue.) Of course, experts’ financial ties to corporations do not necessarily mean that the report itself is biased by corporate influence. A close comparison of the data and recommendations presented in the guidelines to the available scientific evidence speaks for itself.
Two of the five major studies incorporated into the new guidelines tested the effectiveness of statin therapy in the primary prevention of coronary heart disease. One was the West of Scotland Coronary Prevention Study (WOSCOPS), published in the New England Journal of Medicine in 1995. Western Scotland has one of the highest rates of heart disease in the world, and the 6600 men included in the WOSCOPS study were at particularly high risk: their average LDL cholesterol level was 192 mg/dL; 44 percent of them smoked; and, though men with a history of heart attack were excluded from the study, one out of five had symptoms of blocked arteries, such as angina or leg pain with exertion. The men were randomly assigned to receive either a statin drug, Pravachol 40 mg per day, or a placebo.
After almost five years, the men who took Pravachol had 31 percent fewer heart attacks (statistically significant) and 22 percent fewer deaths (not quite statistically significant) than the men in the control group. More relevant than these large reductions in relative risk in determining the real benefit of treating these very high risk men with Pravachol is the actual number of heart attacks and deaths prevented. This turns out to be a much less impressive number: 100 men in the study had to take Pravachol for two full years in order to prevent a single heart attack.* The cost of treating 100 people for two years with Pravachol 40 mg per day was $336,000 (for the drugs alone). In order to prevent a single death, 100 men in the WOSCOPS study would have had to take Pravachol for five and a half years.
Of course, when I was in practice, I wanted to protect my very high risk patients from suffering a heart attack, and recommended they take a statin. At the same time, it is important to remember that even among the highest-risk men without coronary heart disease, 99 out of every 100 would take Pravachol for two years without any benefit. The problem is that we can’t know in advance who the hundredth man is going to be, and for that individual the protection conferred by the statin drug is very important.
In contrast to the very high risk men in the WOSCOPS study, the other major study of primary prevention, the Air Force/Texas Coronary Atherosclerosis Prevention Study (AFCAPS/TexCAPS), included people with only a moderately elevated risk of developing heart disease. Researchers randomly assigned 6600 healthy middle-aged and older people with mildly elevated LDL cholesterol and below-normal HDL cholesterol levels* to be treated with lovastatin (brand name Mevacor) or a placebo for five years. The study then compared the frequency of several health outcomes in the two groups: occurrence of coronary heart disease, any serious disease, death from coronary heart disease, and death from any cause. The guidelines summarize the results: “AFCAPS/TexCAPS is important because it showed that LDL-lowering therapy in persons with only borderline-high LDL-cholesterol levels produces a large reduction in relative risk.”
Well, sort of. The report neglected to specify the relative risk of what, exactly, the statin reduced. The relative risk of developing coronary heart disease was indeed significantly lower (37 percent) in the people who took the statin compared with those who took placebos. However, the guidelines failed to disclose an even more important measure of the impact of treatment with a statin, a finding that was reported in the original article published in JAMA in 1998. The risk of developing any serious disease (the kind of illness that requires hospitalization or that causes death) was identical in the people who took the statin and those who took the placebo. The new guidelines report that the effect of statin treatment on the overall risk of death in the AFCAPS/TexCAPS study was inconclusive. Not so. A definitive conclusion can be drawn: In a five-year study involving 6600 people with moderately elevated LDL cholesterol levels, treatment with a statin did not decrease overall mortality. In fact, a few more of the people who took the statin died (80) than those who took the placebo (77).
In other words, the net result of treating people with moderate risk of developing coronary artery disease with a statin was simply to trade coronary heart disease for other serious diseases, with no overall improvement in health.
Disregarding, for the moment, the fact that statins had no effect on the overall risk of serious illness or death, even the “large reduction in relative risk” of developing coronary heart disease touted in the guidelines translates into a much less impressive reduction in absolute risk. One hundred people in this study would have to be treated with a statin drug for two and a half years to prevent a single episode of heart disease—and the other 99 people would not have derived any benefit. In order to prevent one death from cardiovascular disease, 100 people in this study would have to have been treated with a statin drug for 25 years.* The updated guidelines relied heavily on AFCAPS/TexCAPS to develop the two-step risk-assessment process that doctors now use to determine which patients should take statins. A look at how those criteria actually apply to the men in this study, however, reveals a paradoxical result. At least 85 percent of the men are in the risk category for which the executive summary of the guidelines states, “Use of LDL-lowering drugs at this risk level reduces CHD risk and is cost-effective.”
Notwithstanding the lack of overall health benefits in the people who received statin therapy in AFCAPS/TexCAPS, the guidelines later contradict their own recommendation, concluding that the “incremental cost per additional year of life gained would be >$100,000 for the whole cohort of AFCAPS/TexCAPS,” and that this is too expensive to justify statin treatment. In other words, if you do the arithmetic, the guidelines recommend statin therapy for everyone who has a health profile similar to 85 percent of the men in AFCAPS/TexCAPS. But the report then goes on to say that the cost would be prohibitive. And notwithstanding this internal contradiction, it is simply nonsensical (or perhaps a diversion) to even contemplate how much each year of life gained in AFCAPS/TexCAPS would cost, because the overall death rate was higher, not lower, in the people who took a statin drug.
The bottom line is that these two studies show that men who do not have CHD but do have very high cholesterol levels might benefit from taking a statin drug, though not nearly as much as all the talk of “miracle drugs” leads us to believe. The case for prescribing statins for men with only moderately elevated cholesterol levels is far less compelling.
The updated guidelines are definitive: “In recent trials, statin therapy reduced risk for CHD in . . . women, in those with or without heart disease.” The NCEP’s full report convincingly cites six references to back up this statement.
None of the six references cited, however, provides significant evidence to back up the claim that statin therapy reduces the risk of CHD in women without heart disease: three apply to people who already have heart disease; one is the WOSCOPS study, which included no women; one is AFCAPS/TexCAPS, in which there were only a total of 20 episodes of heart disease among the women in the study—not nearly enough to reach statistical significance; and one was a compilation of five other studies. A review article published in JAMA in 1995 concluded: “There is no evidence from primary prevention trials that cholesterol lowering affects total mortality in healthy women.”
It is hard to believe that the only evidence from randomized controlled (gold standard) studies that addresses the primary prevention of heart disease in women with statins consists of these statistically inconclusive 20 episodes of coronary heart disease, which are being used to justify putting millions of healthy women on statins. This fact is not disputed in the guidelines, which later state under the heading “Special Considerations for Cholesterol Management in Women (Ages 45–75)”: “Clinical trials of LDL lowering generally are lacking for this risk category; rationale for therapy is based on extrapolation of benefit from men of similar risk” (italics mine).
The cholesterol guidelines’ cavalier extrapolation to women of the benefit of lowering cholesterol in men, especially after the recent surprise provided by the results of the “Heart and Estrogen/Progesterone Replacement Study” (HERS study),* demonstrates the medical industry’s unrepentant opportunism in its attempt once again to impose scientifically unsubstantiated medical treatment on women. Such a sweeping recommendation, without unimpeachable evidence from the gold standard of at least one large randomized clinical trial, and optimally several such trials, makes a travesty of the claim that American medicine upholds its standards of excellence by adhering strictly to scientific evidence.
The guidelines are very enthusiastic about the prospects for reducing heart disease in people age 65 and older who have an increased risk of, but have not yet developed, coronary heart disease. Specifically for this group, the report states that recent trials have shown “aggressive LDL-lowering therapy [meaning with a statin drug] is effective in reducing CHD (see Table II.2-3).” The table cites nine references to support this strong claim.
Checking each reference produced the same results as checking the references that supposedly showed that women without heart disease benefit from statin therapy: only one of the nine references (AFCAPS/TexCAPS) is at all relevant to primary prevention of CHD in the elderly, and even in this study only one-fifth of the people involved had reached the age of 65, and the reduction in their risk of heart disease was not statistically significant.
What about the other studies? Six were of secondary prevention (people who are at much higher risk because they already have CHD); two were of primary prevention, but one of these did not include patients over the age of 64, and the other was published in 1978, long before statins were available, and the average age was 51. So, the nine references notwithstanding, no significant evidence from randomized controlled trials has been presented to support this recommendation.
The guidelines also refer to population-based (epidemiological) data to justify the increased use of statins: “The relationship between serum cholesterol levels and lifetime risk for CHD has been evaluated in the Framingham Heart Study.. . . Even at age 70 the lifetime risk for CHD remains high.” Though both halves of this statement are true independently, their juxtaposition could easily lead the reader to draw the wrong conclusion.
Of course the lifetime risk for CHD is high once the age of 70 is reached; eventually hearts give out in people lucky enough to reach a ripe old age and otherwise in good health. And it’s true that the Framingham study examined the relationship between increased cholesterol levels and the risk of heart disease at different ages. But, as mentioned before, the data showed exactly the opposite of what the guidelines imply: total cholesterol is not significantly related to mortality from coronary heart disease beyond the age of 60. Indeed, the authors of the paper, based on data from the Framingham Heart Study, warned: “Physicians should be cautious about initiating cholesterol-lowering treatment in men and women above 65 to 70 years of age. Only randomized clinical trials in older people can settle the debate over the efficacy and cost-effectiveness of lipid-lowering interventions for reducing mortality and morbidity in this population.” At the time the updated guidelines were published, no such studies had been reported. Furthermore, a study published in the Archives of Internal Medicine in 1999 and referenced by the guidelines showed that “None of the lipid measures (total, high-density lipoprotein, and low-density lipoprotein cholesterol or triglycerides) was associated with the risk of MI [myocardial infarction] in this population [of people aged 65 and older].” In other words, there is not even an increase in the risk of heart attack associated with higher cholesterol levels once the age of 65 is reached.
The guidelines then rely on two assumptions that make statins appear very effective in the primary prevention of CHD in people who have reached the age of 65. The first is that statin therapy reduces the risk of coronary heart disease in people between the ages of 65 and 80 “by approximately one-third.” The only problem is that none of the nine references cited to justify this assumption provides any significant evidence of benefit from statin therapy after the age of 65 is reached.
The second assumption is that in the elderly, the risk of coronary heart disease increases with increasing cholesterol levels. The article cited to justify this assumption does indeed show that the risk of developing coronary heart disease remains high in the elderly, but the article does not examine the relationship between cholesterol levels and the risk of developing CHD.
As we will see at the end of the chapter, when a randomized controlled study of the effect of statins in elderly patients without heart disease was published in 2002, the overreaching estimates of the benefits of statin therapy for people in this age group were not borne out.
Most should. Clearly, the people with the highest risk of heart attack are those who already have CHD. This is the situation in which statins are most helpful (called secondary prevention). The results of three large randomized controlled studies of statin drugs in secondary prevention were incorporated into the 2001 guidelines: the 4S study of Zocor, and the CARE and LIPID studies of Pravachol. The primary difference among them is the average LDL cholesterol levels of the people in the studies.
In the 4S and LIPID studies, LDL cholesterol levels averaged 188 and 150 mg/dL, respectively. In both of these studies, treatment with a statin significantly reduced the risk of another heart attack, the risk of dying of coronary heart disease, and the overall mortality rate. In the CARE study, LDL cholesterol levels averaged 139 mg/dL—very close to the average for people who develop coronary heart disease (140 mg/dL)—and the results were not quite as impressive: there was a significant decrease in the risk of heart disease but no significant reduction in the risk of death from heart disease or the overall risk of death.
Just how many heart attacks are prevented by treating people who already have heart disease with a statin drug? In the CARE and LIPID studies, the reduction in fatal and nonfatal heart attacks in the people treated with Pravachol was 0.6 percent each year. This means that 166 people need to be treated for a full year to prevent one heart attack. If my patient is that 1 out of 166, I certainly want him or her to be taking a statin; yet given all the media hype about statin drugs, it is easy to be misled into believing that statins will help everybody or almost everybody who has already had a heart attack.
The guidelines refer to the three major clinical trials of secondary prevention to support the statement that statin therapy effectively reduces the risk of recurrent CHD in women. The 4S study shows that treatment with a statin does decrease the risk of recurrent CHD for women but does not decrease the risk of death from CHD. Furthermore, the overall death rate was actually 12 percent higher in the women who took the statins compared with the women who took the placebo (not statistically significant). The CARE study similarly showed that statins significantly reduce the risk of recurrent heart attack, but not the risk of death from coronary heart disease in women, and not the risk of death from all causes in the study population as a whole. The LIPID study failed to show a significant reduction in the risk of recurrent CHD in women and did not present mortality data for women. At best, these studies show that statins probably lower the risk of recurrent CHD in women with coronary heart disease, but do not appear to lower their overall mortality rate.
Wouldn’t it be interesting to be able to test the recommendations of the 2001 cholesterol guidelines against the older standard of care to see how much benefit would actually result from tripling the number of Americans taking statins? Of course this would be completely impractical, requiring another large, costly, many-year clinical trial. Even more of an obstacle is the fact that such a study would be highly unethical—volunteers assigned to receive treatment according to the old standards would receive therapy that was less than the best medical care currently being recommended.
In a remarkable coincidence, the findings of almost exactly such a study were published in JAMA about a year and a half after the 2001 guidelines were released. The ALLHAT study had been designed in the early 1990s in a way that fortuitously turned out to be a good test of the expanded use of statin drugs recommended in the 2001 guidelines.
Starting in 1994, the ALLHAT study enrolled more than 10,000 patients at high risk of coronary heart disease—equal numbers of men and women age 55 and older whose risk factors for heart disease would have qualified about 90 percent of the men and 75 percent of the women for statin therapy according to the new guidelines. The patients were randomly assigned to take a statin drug (Pravachol) or simply to receive the usual care from their own doctors and let the chips fall where they may with regard to drug therapy. By the end of the study, 83 percent of the group assigned to take Pravachol were still taking a statin and 26 percent of the “usual care” group had been started on cholesterol-lowering therapy by their own doctors—a perfect test to see how much more heart disease could be prevented if the number of Americans being treated with statins were tripled.
The study found that tripling the number of people on statins neither prevented heart disease nor decreased the overall risk of death. There was no benefit to increasing the number of patients taking statins beyond the community norm of the mid-1990s: not for people age 55 to 64 or 65 and older, not for men or women, not for those with or without diabetes, not for those with or without heart disease, and not for those with LDL cholesterol higher or lower than 130 mg/dL. The only group that derived any significant benefit from more statins were African Americans who had fewer episodes of heart disease but no fewer deaths.
These findings should have been major news. Yet the Wall Street Journal was the only major newspaper that I could find that carried the story; otherwise there seemed to be a virtual press blackout. Within the medical journals, the study findings were largely rejected by experts on the basis that so many of the people in the usual-care group had been put on statins that the difference in cholesterol levels between the Pravachol and usual-care groups was not enough to show the benefit of statins. But that was exactly the point of the study. High-risk patients treated by their own doctors according to the prevailing standards of the mid-1990s were already getting the maximum benefit of statin therapy. Tripling the number of people taking statins—coincidentally, almost exactly in line with the recommendations of the new guidelines—provided no further benefit.
Dr. Richard C. Pasternak, the cardiologist who wrote the editorial that accompanied the publication of the ALLHAT study results in JAMA, concluded, “Physicians might be tempted to conclude that this large study demonstrates that statins do not work; however, it is well known that they do.” So much for evidence-based medicine. From my perspective, Pasternak was wrong. The study did not tempt me to think that statins don’t work—it just made me think that tripling the number of people on statins doesn’t provide any additional benefit. Dr. Pasternak was one of the original 14 authors of the 2001 cholesterol guidelines. He declared financial relationships with nine drug companies in the “financial disclosures” that accompanied his editorial in JAMA.
Like the ALLHAT study, the PROSPER study (Pravastatin in Elderly Individuals at Risk of Vascular Disease), published in The Lancet in 2002 also got very little press coverage—no news of breakthroughs or paradigm shifts, though in a negative way it, too, should have contributed to both. The study tested the effect of statin therapy in high-risk elderly patients between the ages of 70 and 82. For those who did not already have heart disease, statin therapy did not reduce their risk of developing heart disease or stroke. It did, however, significantly increase their risk of developing cancer (p = .02). This risk increased each year these elderly patients took the statin, so that by the fourth year of the study there was more than 1 extra case of cancer for every 100 people taking a statin each year.
The 2001 cholesterol guidelines were reassuring about statins’ not causing cancer: “There is no evidence that currently used cholesterol-lowering drugs promote development of cancer.. . .” However, a paper titled “Carcinogenicity of Lipid-Lowering Drugs,” published in JAMA in 1996, was apparently overlooked. This paper pointed out that statins cause cancer in laboratory animals at blood levels only three to four times those typically achieved in cholesterol-lowering therapy for people. The authors raised the possibility that increases in the risk of cancer caused by statins could take many years to be detected, and therefore would not be evident in the large studies that had been done to date. (The average duration of the studies is about five years, and the average age of patients is less than 60.) We cannot know whether statins will turn out to increase the risk of cancer when used long term, nor can we know from one study whether older people might be particularly susceptible. The one thing we can know is that the denials of the evidence of such a relationship in the 2001 cholesterol guidelines and by the authors of the PROSPER article (sponsored by Bristol-Myers Squibb) suggest that the principle “First do no harm,” which should be fresh in our minds from the increased rate of breast cancer caused by HRT, seems, once again, to have been forgotten.
Also published subsequent to the 2001 guidelines were two randomized controlled clinical trials that supported the guidelines’ target of reducing LDL cholesterol below 100 mg/dL in people who already have heart disease (the Heart Protection Study and the Prove It study). This leaves unanswered the question why some studies show no benefit to cholesterol-lowering in heart patients whose LDL cholesterol level is less than 125 mg/dL, while other studies support the lower target. The only conclusion we can draw about the optimal level of LDL cholesterol in people who already have heart disease is that it is still to be determined.
The disparity between the recommendations of the 2001 guidelines and the subsequent findings of the PROSPER and ALLHAT studies cannot be dismissed as simply due to the normal zigs and zags of science. The findings of these two studies directly contradict the guidelines’ recommendations, but they don’t contradict the best scientific evidence that was available at the time the guidelines were formulated. The guidelines’ interpretation of the scientific evidence stretches credulity beyond reasonable limits to recommend statin therapy for primary prevention of heart disease in women, people over 65, and men with only moderately elevated cholesterol levels. The experts must have believed that future studies would validate their assertions regarding statin therapy, but the results of the studies that came out the following year did just the opposite. Even when the contradictory results were published, the response of many recognized experts in the field seemed to focus more on damage control to protect statin sales than on unbiased evaluation of the scientific evidence.
Why do doctors accept such obvious distortions and commercial intrusions into what ought to be the inviolable standards of medical science and medical care? There are several reasons: Practicing doctors are far too busy to do this kind of research for themselves and, as we have seen, the drug companies play a large role in determining how doctors are “educated” about the new developments in medicine (including all the “good news” about the benefits of statins). Another reason is that significant conflicts of interest have become a normal part of American medicine today. Dr. Scott Grundy, the chair of the panel that created the latest cholesterol guidelines, told the Wall Street Journal, “You can have the experts involved, or you could have people who are purists and impartial judges, but you don’t have the expertise.” Unfortunately, in American medicine that expertise is now virtually inseparable from financial ties to industry. When asked why a more balanced approach to heart disease prevention gets pushed aside by these guidelines, Dr. Walter Willett, professor of epidemiology and nutrition at the Harvard School of Public Health, said, “Drug companies are extremely powerful. They put huge efforts into promoting the benefits of these drugs. It’s easier for everyone to go in this direction. There’s no huge industry promoting smoking cessation or healthy food.”
The ultimate impact of the 2001 cholesterol guidelines is this: competent and caring physicians trying to provide the best possible care for their patients are being misled—and are misleading their patients. As shown in Chapter 13, inexpensive, easily accessible, and often more effective interventions to prevent heart disease and improve overall health are being abandoned in favor of expensive drugs. This issue highlights the question of whether our society ought to or needs to tolerate medical care that serves the interests of the drug companies and medical industry before it serves the health of the American people. The 2001 cholesterol guidelines provide one important example of just how far the pendulum has swung toward the interests of the drug companies.
The obvious question is this: Who will benefit from expanding the number of Americans on statins from 13 million to 36 million? The most honest answer (though admittedly taken out of context) probably comes from the Morgan Stanley Dean Witter newsletter: “Who will benefit most from an expanding [statin] market? We have identified three likely incremental winners in the 2006 statin market—AstraZeneca, Schering-Plough, and an undisclosed marketing partner for Crestor.” The newsletter continues: “there are not likely to be any outright losers.” No mention is made of the patients and the doctors who are more concerned about their own and others’ health and well-being than about pharmaceutical company profits. We are the losers.