Sunday, October 21, 2007

Teetotaling Vindicated

Recently, a professional acquaintance was surprised that I choose not to drink alcohol. I have always said that I am willing to trade the health benefits from a class of red wine a day for full control over my cognition. If it's the health benefit you seek in consuming ethanol (or is that just an added benefit), it turns out that epidemiological evidence is tipping the balance in favor of my decision.

Check out this article summarizing an unnamed, unreferenced study (I hate it when the lay media does this!) about the risks of drinking as little as one drink a day. Here are some cherry-picked quotes for you:
According to data compiled by the Centers for Disease Control and Prevention (CDC), alcohol consumption is the third-biggest cause of preventable death in the United States, after smoking and obesity. The CDC estimates that drinking caused nearly 93,000 deaths in 2001.
Alcohol reduces the risk of heart attacks and strokes caused by blocked arteries by 10 to 15 percent. That's probably because alcohol increases good cholesterol and prevents blood platelets from clumping together. "On the other hand, alcohol is detrimental for more than 60 diagnoses," said Juergen Rehm, head of public health and regulatory policies at the Ontario Center for Addiction and Mental Health.
If after I track down these articles there are elements that are misrepresented here, I'll be sure to update you. In the interim, I'll raise a glass of Diet Coke to your health. It will probably be a little while longer before word comes my way of that being bad for me.


golob said...

So far as I can guess, it was this circ research review that prompted the news article.

The conflating of absolute and relative risk makes this article silly. Yes, cancer risk might go up. The absolute risk of even common cancers (like colon and breast carcinoma) are grossly outweighed by the absolute risk from cardiovascular disease. Hence: the overall mortality data says a drink or so a day is a net positive. Check out figure 2 in the review--a 'U' shaped delight.

thomas said...

Yes, but what of this basic error in the controlling for confounding variables (from the article):

The same data show that about 30,000 fatal heart attacks were prevented by moderate alcohol consumption, defined as fewer than two drinks a day for men and less than one drink a day for women. But Dr. Robert Brewer, head of the CDC alcohol team, said the evidence for alcohol preventing heart attacks is less reliable.

"We need to be cautious about interpreting the studies that suggest a benefit," Brewer said. "People who drink moderately are different from people who don't drink — they exercise more, they have better medical insurance, their body-mass index is lower."

In other words, they might have had fewer heart attacks because of factors other than alcohol intake.

I think, however that it was some other papers that prompted the lay press article, but I am too lazy/busy to find them right now.

PS: I am disappointed that you decided to run the "closing I-5 changed Seattle's temperature" shtick over at Dear Science. You definitely made that question up.

golob said...

"shtick!" What a harsh PS!

Honestly, I didn't write the question.

And I think my article is defensible. Hell, this line alone is key: "The changes were modest, the data sets tiny, and the data subject to the variation of the climate. (In other words, do I believe the change? Maybe. Maybe not.)" It wasn't exactly unequivocating.

Still, this is off-topic.

thomas said...

fair enough...

but it's hard to balance an entire article against a complex disclaimer. am i allowed to bring up off-topic ideas on my own blog? ; )