What is the evidence that “elevated cholesterol” causes atherosclerosis? There are 4 supporting arguments in my view. (1) Atherosclerotic plaques are easily produced experimentally in herbivores (e.g., rabbits, monkeys) simply by feeding these animals cholesterol (e.g., egg yolks) or saturated fats. Indeed, atherosclerosis is probably the second easiest disease to produce experimentally. (The first is an endocrine deficiency—simply excise an endocrine gland.) (2) Cholesterol is present in atherosclerotic plaques in experimentally produced atherosclerosis and in plaques in human beings. (3) Societies and subjects with high serum cholesterol levels (total and low-density lipoprotein [LDL] cholesterol) compared to populations and subjects with low levels have a high frequency of atherosclerotic events, a high frequency of dying from these events, and a large quantity (burden) of plaque in their arteries. (The best study in my view supporting this thesis is the Seven-Country study. ) (4) Lowering total and LDL cholesterol levels decrease the frequency of atherosclerotic events, the chances of dying from these events, and the quantity of plaques in the arteries. No one has produced atherosclerosis experimentally by increasing arterial blood pressure or glucose levels or by blowing smoke in the faces of rabbits their entire lifetime or by stressing these animals. The only way to produce atherosclerosis experimentally is by feeding high-cholesterol and/or high–saturated fat diets to herbivores. (Atherosclerosis is not a disease of carnivores and it is not possible to produce atherosclerosis in carnivores [dogs, cats, tigers, lions, etc.] unless the thyroid gland is removed or made dysfunctional before a high-cholesterol or high–saturated fat diet is administered. )

Why has the proved causal relation between abnormal serum LDL cholesterol and atherosclerosis been so difficult to accept by so many extremely intelligent physicians? One factor, in my view, is that this cholesterol–atherosclerosis causal relation has been diluted by the concept of multiple atherosclerotic risk factors and the idea that atherosclerosis is a multifactorial disease. The Framingham study, which has taught us all so much, introduced the concept of “risk factors” and fostered the view that the larger the number of risk factors present, the greater the chance of atherosclerotic events. As a consequence, increased cholesterol became just 1 of several risk factors and was perceived as essentially having no more influence than increased systolic blood pressure, diabetes mellitus (“glucose intolerance”), cigarette smoking, abdominal obesity or lack of regular physical activity, family history, or left ventricular hypertrophy except in younger patients. The view that atherosclerosis is a multifactorial disease has muddled the waters in my view. This is not to say that cigarette smoking, increased blood pressure, diabetes mellitus or obesity, and inactivity are not harmful—of course they are—but if serum LDL cholesterol is <60 mg/dl or serum total cholesterol is <150 mg/dl, there is no evidence (with extremely rare exceptions ) in my view that these other risk factors cause atherosclerosis.

A second factor is the introduction and propagation of the thesis that atherosclerosis is an inflammatory disease. Yes, a few mononuclear cells are regularly seen in experimentally produced atherosclerotic plaques but not commonly in plaques of patients with fatal coronary disease or in plaques excised by endarterectomy. Yes, some blood inflammatory markers are commonly increased in patients with atherosclerotic events. However, many patients have atherosclerotic events when high-sensitivity C-reactive protein (hs-CRP) is normal (<1 mg/dl) and patients with the highest levels of hs-CRP (e.g., those with rheumatoid arthritis or systemic lupus erythematosus) have only a slightly higher frequency of atherosclerotic events than do others of similar age and gender with normal or near normal hs-CRP levels. The same principle, however, does not apply to cholesterol. Patients with the highest serum levels of total and LDL cholesterol, namely those patients with homozygous familial hypercholesterolemia, have an incredibly high frequency of atherosclerotic events and they have these events at very young ages—teenage years. In addition, patients with the next highest serum LDL cholesterol levels, namely those with heterozygous familial hypercholesterolemia, have atherosclerotic events often in their 30s and 40s.

A third factor preventing acceptance of the causal relation between abnormal serum LDL cholesterol and atherosclerosis has been the observation that of adults with nonfamilial hypercholesterolemia but similar levels of serum LDL cholesterol, some develop atherosclerotic events and others do not. It is in this group particularly in my view that the other risk factors and high-density lipoprotein cholesterol levels come into play. Of 2 patients of similar age and gender and similar serum LDL cholesterol levels, say 130 mg/dl, the patient whose systolic systemic blood pressure is 170 mm Hg versus the other patient with a systolic pressure of 115 mm Hg is at much greater risk of an atherosclerotic event. Cigarette smoking may work in similar fashion. Nevertheless, if serum LDL cholesterol is <60 mg/dl, maybe <50 mg/dl, irrespective of degree of blood pressure increase or number of cigarettes smoked daily, atherosclerotic plaques do not develop.

Another factor may be the use of multiple atherosclerotic risk factors in guidelines for who to treat and who not to treat with lipid-lowering drugs. Although guidelines do focus on serum LDL cholesterol level, the number of other risk factors present play a prominent role in this therapeutic decision. If no other nonlipid risk factors are present or only if 1 non-LDL cholesterol risk factor is present and there have been no previous atherosclerotic events and diabetes mellitus is not present, the magical drug treatment number is an LDL cholesterol ≥190 mg/dl. Refraining from drug intervention until this very high LDL cholesterol level is reached plays down or even nullifies the importance of cholesterol in preventing events. (It is important to realize that the lipid-lowering drug guidelines [1988, 1993, 2001, and 2004] have to do only with decreasing atherosclerotic events. They do not concern themselves with preventing atherosclerotic plaques in the first place. Of course, if atherosclerotic plaques are prevented, atherosclerotic events do not occur!)

Such high guideline drug treatment levels keep, in my view, many patients deserving of lipid-lowering drug therapy from receiving these magical agents. The danger of high cholesterol levels to longevity were recognized by life insurance companies in the 1930s but not by physicians. The normal range of serum total cholesterol in laboratory reports for decades was listed as 150 to 300 mg/dl. In 1972, 1 of the world’s most prominent lipidologists reported that his total cholesterol “worry level” for patients was a value >300 mg/dl. If the expert uses such high levels, what importance can be placed on cholesterol by the nonexpert community? Incidentally, for the first several decades of the Framingham study, increased cholesterol was defined as a serum total cholesterol level >250 mg/dl. At this level, it is easy to understand how this risk factor did not separate itself from the others.

It is time to move on from a goal “to decrease risk” to a goal “to prevent plaques.” To do so requires much lower levels of LDL cholesterol than advocated by guideline publications. My goal for all subjects worldwide is a serum LDL cholesterol ≤100 mg/dl and ideally <60 mg/dl. The beauty of the Justification for Use of Statins in Prevention: an Intervention Trial Evaluating Rosuvastatin (JUPITER) is that it dramatically demonstrates what incredible decreases in events can be produced in a short period (<2 years) by decreasing LDL cholesterol by 50% even when starting from a level considered by many to be normal (<130 mg). The mean level (108 mg/dl) might be considered “good” or even “great” by many physicians but its lowering to 55 mg/dl (by rosuvastatin 20 mg/dl) decreased all events by >40%, indeed nearly 50%, including a decrease in stroke by 48%! This trial beautifully shows that we can drastically decrease or even prevent atherosclerotic events and expensive procedures by taking a single pill every day and do it safely. Most Americans will not reach JUPITER treatment levels (LDL cholesterol 55 mg/dl) by diet alone. Statin drugs have been ingested by humans for nearly 30 years and their safety and thus benefit/risk ratio may be the best of any proved-useful medication. Toxicity resides mainly in atherosclerosis, not in the drug.

I consider it unfortunate that there continues to be so much criticism of statin drugs, which I consider to be the best cardiovascular drugs ever created. ^⁎

⁎ I have no investments in pharmaceutical or device companies, I receive no grants from them, and I am on no advisory boards of industry. I do, however, give talks periodically sponsored by pharmaceutical companies.

The lower the LDL cholesterol, the better, and this principle has been established repeatedly despite voices of the anticholesterol, antistatin fallacy mongers! It’s the cholesterol, stupid !