MALCOLM KENDRICK, MbChB, MRCGP (email - malcolm@llp.org.uk )

November 28, 2002

Part 1, Part 2, Part 3, Part 4, Part 5

Statins Do Not Prevent Heart Disease - At Least Not By Lowering LDL/Cholesterol Levels


Statins reduce the risk of dying of coronary heart disease (CHD). There, I said it. You probably thought I didnt believe this, but you cant argue with the results from the clinical trials. Big, long, well-controlled studies that have all shown pretty much the same thing - stains provide protection against CHD.

Should this be a surprise? Statins were, after all, specifically designed to block the synthesis of cholesterol in the liver, and thus reduce LDL/cholesterol levels in the blood - and they do this very well. Furthermore, by reducing the level of LDL, statins were then supposed to reduce the risk of CHD - on the basis that a raised LDL level was the primary risk factor for the disease - and they do this too.

On the face of it, a glorious vindication of the cholesterol hypothesis of CHD, and it looks pretty damned inarguable doesnt it? Raised LDL levels cause CHD, and when they are lowered, the risk of CHD drops. Cause and effect flushed out into the open. Experimental proof. Touchéé.

So, Batman, argue your way out of that!

Let us start the argument with a little diversion into the world of risk. You may have seen figures stated, such as, statins reduce the risk of CHD by 40%, even 50%. You may not know what such figures mean. They certainly sound super-impressive, and suggest that a statin will save one persons life for every two people taking the drug. Right?

Wrong.

For that risk reduction is a relative risk reduction - not an absolute risk reduction. To explain.
The risk of being struck by lightening (Im guessing here), may be one in five million over five years. Were I to develop a hat with a copper wire reaching from it down to the ground, the amazing copper-o-matic, I may reduce your risk of a lightening strike over five years to one in ten million, an amazing 50% reduction in risk.

The other way to look at this is that your absolute risk of being struck by lightening has changed from 0.000.02% to 0.000.01%. Or an absolute risk reduction of 0.000.01%. Take you pick, a massive 50%, yes 50% risk reduction. Or a measly 0.000.01% risk reduction.

They both mean exactly the same thing, but one sounds a bit more impressive than the other, and were I to try to sell you my copper-o-matic - $19.95 from all major stores - I know which type of risk reduction I would be promoting (And no, it wouldn't be absolute risk reduction)

Thus, if we look at the recent Heart Protection Study, hailed as the most amazing trial ever, at least by the chief investigator anyway, this showed that tens of thousands of lives could be saved each year by the use of statins, with a risk reduction of nearly 50%.

True, all true, but how many people would you need to treat to save fifty thousand lives?

The figures from the HPS were that you could save fifty thousand lives, over a five year period, if you treated…….. Have a guess.

Ten million people.

This works out at one life saved for every two hundred people treated, or an absolute risk reduction of 0.5%. Maybe not quite as awe inspiring as you may have thought. About the same risk reduction, in fact, as is achieved by aspirin.

And, if we take these figures a little further, it is possible to work out something else quite interesting, which is that a lifetime reduction of LDL using statins will result in a, maximum, 3% risk reduction of dying from CHD.

To explain. If a statin reduces the risk of dying of CHD by 0.5% over five years, and you take a statin for thirty years, basically a lifetime of drug taking, then you would have a 0.5% x 6 reduction in risk. Or 3%. Which opens up an interesting thought. Interesting to me, at least.

In general, statins bring LDL levels down to the normal level, but by doing so they decrease your risk of dying of CHD by a somewhat titchy 3%. Which, even if you do believe that statins work by lowering LDL levels, leaves a rather large 97% of CHD caused by something else, other than a raised LDL level - and what might that be?

Anyway, to return to the main point, which is that, whilst statins do normalise LDL levels, they dont actually provide that much protection against CHD. Which means that it could well be that their protective effects may be due to something else, other than LDL lowering. (On the other hand, if the absolute risk reduction was 50%, then I think this would be case proven for the cholesterol hypothesis).
Even so, I recognise that to suggest that the cardioprotective effects of statins are, in effect, a coincidence, seems a bit of a stretch. But bear with me, for I think that case for coincidence is overpoweringly strong.

There are four main strands to the coincidence argume

• Statins act far too quickly for it to be through any LDL lowering effect
• They work independently of the LDL level - or by how much they lower the LDL
• Statins protect against CHD in the elderly, in whom a raised LDL level is not a risk factor

They have many other effects, other than lowering LDL levels, a great number of which have been clearly demonstrated to have an beneficial impact on both blood clotting and endothelial damage - and thus CHD

Speed of action
It is generally accepted that a raised LDL level takes many years to cause CHD. Exactly how many is unclear, but the major clinical trials on statins lasted five years, so the assumption here was that it would take at least five years on a statin to show any significant effect on the rate of CHD.

If, however, statins reduce CHD risk within a much shorter time period, this makes it almost certain that they are not working through LDL lowering.

So what of the: Myocardial Ischaemia Reduction with Aggressive Cholesterol Lowering (MIRACL) trial on the short term use of statins?

MIRACL demonstrated that intensive treatment with atorvastatin, begun immediately after an acute coronary event, produces beneficial effects that are apparent within several weeks. This provides evidence that the addition of intensive lipid-lowering therapy to the standard of care may help improve the outcomes of these patients. - Dr Gregory Schwartz.

Statins work within weeks, not months or years. Actually, to have a measurable effect within weeks, they must be working immediately - instantly. Because If they didnt start working for weeks, they wouldnt show any effect for months. And this is not a one off result; it has been shown in many trials.

Lipid-lowering therapy after acute coronary syndromes (ACS) reduces risk of 6-month mortality by one third after adjustment for confounding factors and should be prescribed to patients in order to reduce short-term mortality. - Dr Herbert Aronow Lancet Apr 2001

On thing is absolutely certain, these short term protective effects cannot be due to LDL lowering.

Dose response
Moving on to the next point. If statins do work by lowering LDL, then the more that LDL is lowered, the more CHD protection you should see. In fact, a number of trials have shown the exact opposite:

……(in the CARE trial) In addition, there was no linear relationship between the extent of LDL reduction and percent reduction of events. Patients whose LDL levels on treatment were between 101mg/dl and 125mg/dl had a 46% event reduction, while those treated below 100mg/dl only enjoyed a 32% event rate reduction. - Thomas Bersot MD. AHA 71st Scientific Sessions Nov 8 - 11 1998

And if you look at other trials on statins, you cannot see any dose response, just a general reduction in CV events with statins, no matter what the starting level of LDL or how much it is lowered by.

Effect in the elderly
As demonstrated in earlier articles, there is no doubt that a low cholesterol/LDL level is associated with higher rates of CHD in the elderly. Yet, if you give statins to elderly patients they are protected against CHD. In addition, in the recent Heart Protection Study, statins also protected against Ischaemic stroke, and a raised LDL level is not a risk factor for this condition.

So statins work immediately, they work independently of their LDL lowering effect, and they work in populations, and conditions, where a raised LDL is not a risk factor. All of which makes it inarguable that their CHD protecting effects have nothing to do with LDL lowering. They must operate in another way.

Other effects
When statins first came out and started to show protection against CHD, I must admit that my conviction that LDL levels have nothing to do with CHD was severely shaken. Indeed, for a few years I gave up on my alternative hypothesis.
However, it didnt take too long for data to start emerging that cast serious doubts over the statins protect against CHD by lowering LDL levels concept. I wasnt the only one to notice the difficulties with the data:

The statins correct plasma lipid levels optimally, yet the real magnitude of their benefits is marginal and certainly not better than attained with agents that do not affect plasma lipid levels. It is suggested that some of our recommendations and actions relating to plasma cholesterol levels and to atherosclerosis are based on concepts that are fundamentally flawed and need to be revised. - Krut LH Am J Cardiol 1998

So what else were statins doing?

The beneficial effects of statins on clinical events may involve nonlipid mechanisms that modify endothelial function, inflammatory responses, plaque stability, and thrombus formation……These nonlipid properties of statins may help to explain the early and significant cardiovascular event reduction reported in several clinical trials of statin therapy. - Rosenson RS JAMA 1998

In reality, statins do all sorts of things that could easily provide protection against CHD, dividing into three basic areas:

• Plaque stabilisation
• Endothelial protection
• Anti-coagulation

As you may be aware, the latest hot thing in CHD research is to measure C - reactive protein levels (CRP). The CRP level provides a reasonable indication of endothelial damage, with higher levels suggesting active plaque formation and growth. So if statins work by reducing endothelial damage, rather than by lowering LDL, we should see statins lowering CRP levels. And guess what…….statins reduce CRP levels.

It will not be too long (in fact it is happening now) before mainstream researchers start to actively promote the fact that statins protect against CHD by protecting the endothelium, and thus lowering CRP. In time the LDL lowering effects will, like the Cheshire cat, gradually disappear until you will never know they were there at all.

At which point the entire diet-heart/cholesterol LDL hypothesis should just roll over quietly and die, killed, ironically, by the statin data. But this hypothesis has proven before that the small matter of several mortal wounds has no effect upon it. The inescapable fact that statins do not, and cannot, work by lowering LDL levels will, I predict, prove merely an inconvenience.

For the cholesterol hypothesis managed to survive the fact that cholesterol in the diet has no effect on cholesterol levels in the blood - by claiming that it was saturated fat that mattered all along. It easily shrugged off the data showing that many people with normal cholesterol levels died of CHD - mainly by lowering the definition of normal. I think it has now reached 4.5mmol/l, but it has a few more mmol/l to fall yet.

So the statin data will be blithely ignored. The diet-heart hypothesis will simply change its shape and grow again, stronger than before. It is the original 1950s B-movie monster hypothesis.