Sunday, December 02, 2007


The other day I was chatting to a work colleague about a lab result showing hypercholesterolaemia. She'd ruled out any medical problems likely to elevate blood cholesterol in a cat, so we decided it was clinically insignificant. If you are really dumb you'll add, "if you believe it's significant in human heart disease anyway".

I'm really dumb some times.

When your colleague comes back with "my father in law is on a statin and I've been meaning to find out more about them" you some times say stupid things like "does he have muscle pain?". "Yes". "Oh".

*******HEALTH WARNING*******

Next link has serious factual errors about statins. Click at your own risk.

*******HEALTH WARNING*******

It's called myalgia and even the most mis-informed hardcore true statin believers realise it's real and it's Q10 responsive. The antistatin folks have been on the ball a bit longer, they withdraw the statin as well as giving the Q10. They are well aware that the muscle pumping your blood is struggling as much as the muscles which can't get you up the stairs any more.

If you mutter something about coenzyme Q10 to most clinicians they will just look blank. Don't you get that at health food shops? Or in face cream, the anti wrinkle type?

For those with elderly relatives on a statin, especially pravastatin, the following paper deserves careful reading:

First, as I'm afraid it has to, comes the body count. This is a European study and so they give you the mortality figures. There were 2913 patients in the placebo group. A total of 306 died during their 3 years of not taking a statin. That is 10.5% died. In the treatment group there were 2891 patients and 298 died, that's 10.3%. Bear in mind that these were high risk cardiovascular patients, the sort for whom statin therapy is supposed to be effective in saving lives.

If you consider these percentage figures expressed as mortality per 1000 patients it means that 105/1000 died in placebo group and 103/1000 in the pravastatin group. That looks very much as if you had to treat 500 people for 3 years to save one life. Wow! Statistically this level of benefit is pretty certainly due to chance. That's a lot of pills for no benefit. Actually it's roughly 3 X 365 X 2891 pills, which is 3,165,645 pills. At 40mg/pill that's about 128kg of pravastatin.

They forgot to include the above information in the abstract.

They did mention the increased frequency of new cancer diagnoses. They could hardly avoid it, given the p value, which was p=0.02, well below the 0.05 needed for statistical significance. The actual cancer death rate never made statistical significance, a mere p=0.082, but in general it takes longer to die from most cancers that the three years the trial lasted, so no surprise there. Thank goodness the trial stopped when it did.

An elderly person developing cancer while on pravastatin will be pleased to hear that if you mix this study with loads of other statin studies, especially those using younger patients, this cancer increase can be made to go away. Phew. It's nice to see what meta-analysis can do. But you still have cancer.

The bottom line is that pravastatin saved some deaths due to heart disease and replaced them with deaths from other causes, the commonest of which was cancer.

Clearly no self respecting cardiologist would ever want you to die of a heart attack. Cancer, obviously, is much better and has the advantage of being an SEP (Somebody Else's Problem). Personally I'm not too certain that cancer is a good trade for a heart attack. You decide.

If you have access to The Lancet full text you want Table 2: Primary End Points of PROSPER, at the bottom of page 1625.

Best quote from the discussion:

"the most likely explanation is that the imbalance in cancer rates in PROSPER was a chance finding, which could in part have been driven by the recruitment of individuals with occult disease".

My comment?

But they mostly were recruited in to the pravastatin group, with a value of p=0.02.


Duh, we wuz ded unlucky Guv. Ded unlucky.


No comments: