02 September 2011

Necessary or Sufficient?

I’d never heard of the ‘necessary or sufficient’ argument before I studied for a BA with the Open University. It appeared during a discussion about skyscrapers.
The standard received story is that the first skyscrapers were in Chicago, and not all that tall — about 16 storeys. At the time they were built it was known that people wouldn’t climb more than six flights of stairs to an office. So, lifts were ‘necessary’ for the development of skyscrapers, yet were not a ‘sufficient’ reason for their development. We had to think of other factors.
I’ve never yet a medical professional who has heard of this type of argument, though from the way they frame things, they are probably unconsciously aware of it.
Consider this: cigarette packets are covered in warnings about how bad it is to smoke, and how smoking gives you cancer. This isn’t quite correct. Most cases of lung cancer are associated — note the careful use of words — with smoking, but around 5% are not. And yes, I do know that there are several varieties of lung cancer.
But, only about 20% of cigarette smokers get lung cancer — yes, perhaps the others die of other smoking related causes. I’m not trying to say that smoking is good for you, or even safe. I’m just trying to get behind some of the facts.
And the facts simply say that for most cases of lung cancer, smoking is necessary, but it is not sufficient. If it was a sufficient cause, then (almost) all smokers would get lung cancer. They don’t. Therefore, there are other factors to be considered — these days we would consider genetic susceptibility, and perhaps other environmental nasties.
The same is true for alcoholic cirrhosis of the liver. This is now a major problem in younger people, and has been linked to binge drinking. It’s clear that alcoholic cirrhosis is caused by ‘excessive’ alcohol consumption, and there are guidelines about units per day. In passing, I note that these recommendations were figures plucked out of the ether; it was seen to be important to tell people how much they could ’safely’ drink. Yet there are plenty of people who drink far more than the recommendations, and who don’t get cirrhosis. Alcohol is necessary for alcoholic cirrhosis, yet not ‘sufficient’. Again, there seems to be evidence of a ‘genetic weakness’.
And none of the above is to be interpreted as encouragement to smoke at all, or to drink to ‘excess’. It isn’t.
On the other hand, mesothelioma — a malignancy of the pleural lining of the chest — is directly related to exposure to asbestos fibres, and is (almost) unknown if there has been no exposure. Asbestos is not only necessary, it is sufficient.
Next time you hear an expert talking about associations and causation, just ask how much of it is necessary and how much is sufficient.

No comments:

Post a Comment