We just read a new paper, available here, which seems very well done. Its message is its title: LDL-C[holesterol] Does Not Cause Heart Disease

That LDL cholesterol causes heart disease is very widespread common medical wisdom. (Except, that is, with Dr. Mike, who has been saying not to worry about it for at least 20 years.)

However, this belief, also known as the lipid hypothesis, is so widespread and ingrained, that if the aforementioned paper were tattooed on the arms of the entire medical profession, we would still be getting statins shoved at us and saturated fat warnings for the foreseeable future. Medical Sacred Cows are very resilient.

In the paper, chief author Uffe Ravnskov of Sweden, attacks several sacred cows and their offspring, by developing evidence for the false hypothesis. Simply stated, if a benefit is claimed (i.e. lower LDL cholesterol=less heart attacks), then the various studies should support this. If they do not, then the hypothesis, sacred cow or otherwise, is disproven.

Sacred Cows on Parade

This leads to several very interesting assertions, all of which seem to hold up. Here is a summary.

• Excluding people with familial hypercholesterolemia, a rather special case, people with high cholesterol do not have more atherosclerosis than people with low cholesterol.
• Cholesterol lowering treatments, with a sole exception, do not improve arterial health. The sole exception: exercise. (And there is a good reason, see this.)
• Straight out fraud? The 1987 Framingham study had this: For each 1 mg/dl drop in TC [total cholesterol] per year, there was an eleven percent increase in coronary and total mortality’. Read it twice. It’s a total shocker. Less cholesterol, more death. There it is, 30 years ago. HOWEVER, this is not how the American Heart Association reported it three years later. Supposedly quoting Framingham, they stated ‘a one percent reduction in an individual’s TC results in an approximate two percent reduction in CHD risk.’ They deliberately flipped the result to jibe with the conventional wisdom. You can’t make this stuff up.
• The same American Heart Association report allegedly misquoted 10 other articles, and apparently ignored contradictory ones.
• There is only weak association between heart disease and total cholesterol, and no association for those over 50. For women over 60, it’s inverse: more cholesterol, less heart attacks.
• If the lipid hypothesis is correct, then heart attack victims should have higher than average cholesterol. However, they, in fact, have lower than normal.
• Elderly people with high total cholesterol live longer. How can this be if the lipid hypothesis is true?

Statins

Then the author turns his cannons to our ever-popular cholesterol-lowering drug: statins. He examines several highly published trials and studies, and finds a fairly consistent pattern of ignoring contrary data. I.e. a study looks at results from 25 trials, but uses only 12 for assorted “reasons.”

Most studies are sponsored by drug companies and show weak benefit from statins. So Ravnskov, unsurprisingly, decided to look at the excluded trials. The results from excluded trials were inverse: more death for statin users.

Were the drug companies cooking the books? The author sates this: For some years, many researchers have questioned the results from statin trials because they have been denied access to the primary data. In 2004–2005, health authorities in Europe and the United States introduced New Clinical Trial Regulations, which specified that all trial data had to be made public. Since 2005, claims of benefit from statin trials have virtually disappeared.

Since 2005? Benefits have disappeared? You would never know it. Statins are still pushed as a miracle drug.

Ravnskov also makes a strong case that statin side-effects are heavily underreported. Why? Because in the drug trial “run in” period, participants are excluded if they have side effects. Hence, no side effects are ever reported.

If This Proof?

The author concludes that he has disproven the lipid hypothesis and the efficacy of statins. Accept this as you will. We do.

The principle is this. You cannot prove causality from association, but you can disprove it. If, for instance, lowering cholesterol was associated with decreased mortality, then the hypothesis is supported, but not confirmed, however, if the opposite result occurs, the hypothesis is disproven. While we may live, to some extent, in a post-fact-based world, most scientists would subscribe to this.

If associations, tests, trials, etc. support a hypothesis for a long time, the hypothesis may become accepted as proven, but this is not the case for the area covered in Rasnskov’s exposé. Far from it. The medical treatments hundred of millions of people have been subjected to for decades are all based on very selected data, and often on no data at all, or even in the face of contradicting data.