[rightwingcrazy]

I'd guess that sick people are more likely to take supplements, so, naturally, people that take supplements are at increased risk. What is the evidence of causality?

[barkeep]

Good point rwc, very possible skew on the results. Still we have this lovely chance to do a long-term experiment on natural selection, let's not blow it by informing the cohort.

FDA cannot regulate anything classed as a "supplement". Lots of boomers are becoming junior Ponce DeLeons looking for the fountain of youth. Too few of them took enough sciemce courses to apprehend that in the dose lies the poison (you'd really think as many of them as are buying into botox this lesson would be second nature, but go figure).

By all means, let's let a few million of them kill themselves using the "if a little is good, more must be better" theory. Maybe we should offer a mega-dose selenium supplement just to speed the process.

If this works out, it could be better than tobacco for relieving stress on Soc Sec and Medicaid, dead folk don't use benefits.

[darth]

ABSOLUTELY!!! The error I see in many medical papers is CAUSALITY. Sick people often try megadoses as medication. Just because there is correlation does not mean that supplements cause the increased mortality. The major drug companies have been on a campaign for decades against supplements because supplements supplant expensive prescription drugs. For example, I tore my achilles tendon while packing a friend's deer out of the mountains. My orthopods were unable to do anything for me. After 4 years of being incapacitated I heard about glucosamine-chondroitin; my torn tendon grew back in 3 months.

[Hawthorn]

> I'd guess that sick people are more likely to take supplements, so, naturally, people that take supplements are at increased risk. What is the evidence of causality? <


Excellent point! And that's why one should always be skeptical of studies that aren't based on large, randomized samples and even better, a "double-blind" methodology.

[Stentor]

I'd guess that sick people are more likely to take supplements, so, naturally, people that take supplements are at increased risk. What is the evidence of causality?

I'd like some evidence that the people who did this study ever took an undergraduate statistics course.