Wednesday, April 16, 2008
Don't Believe Every Medical Study that Comes Down the Pike
You can't believe a lot that you read these days. It turns out that Merck wrote many of the Vioxx drug studies and that the published authors didn't write up or conduct most of the research. The former might be forgiven, but the latter, I don't think so. If you are going to be an author of a drug study, you or your lab had better have done the research. What those people did seems a bit unethical to me.
This study's underlying assumptions seem flawed. I could understand if someone died from an overdose of a fat soluble vitamin, but here's what one expert said about the study:
Blumberg raised concerns about the use of "all-cause mortality" as a yardstick for antioxidants' influence on health and life. "All-cause mortality" includes deaths resulting from everything from cancer to a train wreck.
Blumberg said: "There is no basis in biology to presume that one or more of these nutrients can kill through any and all possible mechanisms of action."
Medical statistics looks for trends, but if the underlying assumptions are garbage - garbage in equals garbage out.
The same day, the same web site has a mouse study using a synthetic form of Vitamin D that prevents certain breast cancers from forming. Hopefully, the researchers used sound statistical methods.
This study's underlying assumptions seem flawed. I could understand if someone died from an overdose of a fat soluble vitamin, but here's what one expert said about the study:
Blumberg raised concerns about the use of "all-cause mortality" as a yardstick for antioxidants' influence on health and life. "All-cause mortality" includes deaths resulting from everything from cancer to a train wreck.
Blumberg said: "There is no basis in biology to presume that one or more of these nutrients can kill through any and all possible mechanisms of action."
Medical statistics looks for trends, but if the underlying assumptions are garbage - garbage in equals garbage out.
The same day, the same web site has a mouse study using a synthetic form of Vitamin D that prevents certain breast cancers from forming. Hopefully, the researchers used sound statistical methods.
Labels: Real science versus PR science
