Thursday, December 2, 2010

Living an Extra Seven Years - Medical Implications

Part 1  Part 2  Part 3  Part 4  Part 5
Church abstinence is today what smoking was in 1954. Everyone who’s looked at this knows it’s bad for you. There is debate on the mechanism of action, but not on the effect itself. What do we do? Or perhaps a better question is: did the Surgeon General do the right thing in condemning smoking before it was fully understood?

Because this is a behavior and not an opinion or belief, we can, in good conscience, recommend it to everyone. An atheist can go to church without changing a single belief. He can attend church potlucks and volunteer with church members without agreeing to a single point of theology. It wouldn’t be easy; an atheist might get withdrawal symptoms like a smoker, but if this can be pushed through, it’s probably a good thing in the long run. Pushing belief on an unwilling person had ethical implications, but a doctor can push a behavior like attendance with no ethical misgivings.

Even if we don’t care about atheists (or are afraid of their anger at such a reasonable suggestion), doctors should have absolutely no misgivings with asking Christians about church attendance and encouraging them to increase it. We have very solid evidence to support such a recommendation.

In the Hippocratic Oath, Doctors swear to ‘do no harm.’ If you went to your doctor, it would be unethical for the doctor to ignore your smoking habit; he has information that could prevent you harm; withholding it is wrong. How is this information about church attendance different? Can we doctors, in good conscience, withhold this information from our patients? Would this be anything other than harm?

Unfortunately doctors are terrified of religion. An enlightening essay was published in the New England Journal of Medicine on this topic (5). The author (a non-physician) told physicians that they should not talk about religion. Even though 77% of hospitalized patients want physician to consider their spiritual needs and 48% of patients want their physician to pray with them, the authors explain, “Patients often ask for things that are unrealistic or that may not be in their best interests.” After all, “physicians are not trained to engage in in-depth conversations with their patients about their spiritual concerns.” In other words, doctors are so bad at spiritual conversations, they shouldn’t even try. This idea comes straight from the Hippocratic Oath; all doctors swear to, “Never attempt a thing I’m not already good at.”  Actually that’s a lie. We don’t swear silly things like that. And neither should we live by them.

It is clear to me that doctors ought to recommend church attendance to every one of his patients. It is likely that patients will ignore him as they do with smoking. But if he truly cares about their wellbeing, he must at least mention the risk his patient is taking by staying at home on Sundays.

Consider the magnitude of this public health problem. Every year in the US, 2.5 million people die; if we consider the frequency of church attendance and adjust robustly, the number of deaths that can be attributed to missing church is 568,000 (9, 11, 12, 13). Every year, half a million people die because they missed church. This is one-fifth of the total mortality rate. In 2004, a paper was published to try to calculate the actual causes of death in the US (14). They tabulated that smoking causes 435,000 deaths per year. Church attendance is more important than smoking. Inactivity and poor diet cause 400,000 deaths per year. Church attendance is more important than diet and exercise. Deaths attributable to car crashes, alcohol, illicit drugs, sexual behavior, microbes, and incidents involving firearms together add up to 324,000 deaths per year. Church attendance is more important than all of these combined. Church abstinence is the single strongest association with mortality in the United States, but it doesn’t even get mentioned in a paper on the subject.

[All references are listed under the first post]


  1. Great article, David. It really made me think! But, I'm having trouble with the statement, "Church abstinence is the single most important cause of mortality in the United States". I hope I can explain it correctly . . . the part 'cause of mortality' is what's getting me. Smoking causes the lungs to fail and people die. Drinking causes the liver to fail and they die. Not going to church causes what part of the body to fail? The other items cause a body part to fail and there's a direct correlation to why that part failed. When people who abstain from church die, wouldn't you see them die from a certain body part failure(the same body part failure - like lungs, heart, etc.) in order to say that church abstinance was the cause? Of those 568,000 people, did they die of the same body part failure? I think you could say that those 568,000 could have had a longer life if they'd gone to church, but can we say that not going to church actually caused their body part, lung, heart, etc., to fail? I think you could say that it contributed to their death, but I don't know if you could say it "caused" it.

  2. You've caught an error! I was inconsistent. I have corrected it and the sentence now reads: "Church abstinence is the single strongest association with mortality in the United States." As Dan pointed out in a previous comment, I have not proven causality.

    But causality is a very high bar, particularly for negative effects. My argument throughout the paper is that we acted against tobacco before we had conclusive causal linkages, and I think we should do the same with church abstinence.

    There are (at least) two things that can be happening here. It may be that 1. church abstinence *causes* death, or 2. it may be that something else causes death that simply correlates with church abstinence.

    If it is case 1, what are they dying of? One of the papers tried to look at that. With minimal adjustment, non-churchgoers are more likely to die of respiratory disease (e.g. emphysema), diabetes, infectious disease, and circulatory disease (e.g. heart attack). But the sample size wasn't big enough to give clean results. The only effect that remains statistically significant after adjustment is respiratory disease (and that's not enough to explain the high-level effects). The never-attenders were 299% (p<0.01) more likely to die of respiratory disease than the average person (with simple adjustment); if you adjust for smoking and social ties, it becomes 111% more likely (p<0.05). Most of the effect was from smoking, but not all of it. So, at least, church attendance correlates with healthy lungs. Why? I have no clue. But that's what the data seems to show.

    Case 2 (only association) is hard, too, because most of the stuff we know is good or bad for you is already accounted for. The best theory here is that there is some social health which church people have that non-church people don't, but that's very squishy. I don't know what that means. I'd love to hear your speculations on this one.