Saturday, November 18, 2006

Cholesterol Bogeyman

I was going to suggest it is currently accepted wisdom that an elevated blood cholesterol level is a "risk factor" for heart disease. That is not strictly true as there is a small group of medical practitioners who object to the idea that elevated cholesterol is a cause of heart disease. This even applies to those doctors who are so enthusiastic about cholesterol lowering drugs (statins) that they preach that everyone should take a statin IRRESPECTIVE of their cholesterol level. I find it hard to find a more convincing argument that cholesterol is irrelevant.

Anyway, let's look at one of the few reasonable cholesterol lowering trials ever completed. It was done in Japan. It simply involved taking 47,294 men, all of whom had a total cholesterol level above 240mg/dl (that's 6.15mmol/l in new money, but the paper is written in Noddy units). Everyone got either 5mg or 10mg of simvastatin per day (I presume based on body weight, the authors forgot to say how they decided!). They followed them for six years, then looked at death rates.

Now one strange thing about humans is that we are all different. If you give a big group of people the same dose of a drug most people will respond to it. Some by a lot, some by a little, a few not at all. That's exactly what happened. So now we can split those 47,294 men up in to those who lived with high, low or medium cholesterol levels for six years, all of whom had the same dose of simvastatin kicking around in their blood stream.

What do you get? I like death rates as a measure of outcome. There is no arguing with an outcome of being dead or a live. It's pretty clear cut. Even to a cardiologist. So what happened to death rates?

Those men who's cholesterol level ended up between 200 and 219mg/dl had the lowest risk of dying. In fact if the value ended up anywhere between 180mg/dl and 259mg/dl the risk of dying was pretty much the same as in the lowest risk group. Anything above 260mg/dl was associated with increased mortality. Above 280mg/dl the effect was most marked. Mostly heart disease. I'll write about inherited familial hypercholesterolaemia another day.

Wow, cholesterol must be really bad for you! Except there were a number of men who developed cholesterol levels below 160mg/dl on this dose of drug. Now this is a cholesterol level which would might once have made a cardiologist very happy. I believe they are harder to satisfy nowadays. How good for your overall health is a cholesterol level below 160mg/dl? Well, in this study, by six years later you are considerably more likely to be dead than if your level had only dropped to 210mg/dl. In fact you are 2.76 times more likely to be dead. This is actually a marginally higher death rate than if you had failed to respond to the drug at all. But your cardiologist would still be happy because the excess death rate is not due to heart disease.

The men who dropped their cholesterol below 160mg/dl tended to died of cancer.

Imagine going to your doctor and being offered a pill which would switch your future life from one ending in heart disease to one ending in cancer. Well, we've all got to die some time. Which disease would you prefer? Go on, really. A quickie heart attack or the big C?

By the way the Japanese appear somewhat more clued up about heart disease that the West. A nice commentary here. Pity the free full text is in Japanese!

No comments: