Wednesday, December 1, 2010

Living an Extra Seven Years - Clarifications

Part 1  Part 2  Part 3  Part 4  Part 5
This effect is very interesting because it is so tightly linked to behavior, not belief. It doesn’t matter how ‘spiritual’ you feel or how religious you say you are (6). You only live longer when you actually go to church (though there is some suggestion that private religious practice counts). It’s very much like the behavior of smoking: people who do X, die. People who smoke, die. People who stay home on Sunday, die. It’s as simple as that.

Because the models have been adjusted, we can’t recommend anything other than the thing itself. For example, if it were the case that church only worked by reducing smoking, then we could say: either go to church OR quit smoking (if you can...). When they found out smoking was bad, they didn’t know entirely what parts of the cigarette caused the cancer; they could only recommend quitting altogether (today they understand the mechanisms better and sell ‘smokeless’ cigarettes). One day, that might be possible for church; but for today, we can recommend nothing but simply attending church.

It is important to note that we don’t have final ‘proof’ that church causes lower mortality. ‘Proof’ in a modern medical sense demands a particular kind of study called a Randomized Control Trial. Correlation is not causation (4). We cannot be absolutely certain that this effect is caused by church. That is to say, if a doctor recommends a non-churchgoer to go to church, we are not certain that the effect would remain.

The skeptic might say that there might be some other effect (maybe a yet-unobserved healthiness factor) which makes people go to church and also makes them live a long time. But it’s also possible that smoking doesn’t cause lung cancer, but that there is some yet-undiscovered sickness factor which makes people who smoke get lung cancer (we could certainly have argued this in 1950). To silence the skeptic, we’d have to get a big group of people and have them each flip a coin. Heads gets treatment 1 (smoking/no church); tails gets treatment 2 (no smoking/church). The problem is that we have good reason to suspect that treatment group 1 is going to die, and for better or worse, we decided killing people for science advancement was generally a bad idea. And even if we weren’t sure, nobody objecting to recommendations on church attendance would support an experiment randomly assigning people to church or no church. The experiment is ethically impossible. We package cigarettes with a warning label and doctors push patients into smoking cessation programs without ever having done a perfect study on their deadliness. How is church abstinence different?

[All references are listed under the first post]


  1. This is fascinating - looking forward to parts 4 and 5!

  2. I'm glad!

    Also, nice new blog! I hope your philosophical ruminations continue :)