The cholesterol-lowering drugs called statins can reduce the risk of subsequent cardiovascular events among those who have had one already, and possibly prevent an adverse event from ever occurring. Can they also prevent you from dying of cancer? That s what a number of headlines are suggesting today, based on a poorly conceived study published Thursday in the New England Journal of Medicine.
Unfortunately, it s what we at the American Council on Science and Health politely describe as bupkis. (I prefer the more hifalutin term data dredging, interrupted ACSH s Dr. Ross). Study leader Dr. Sune Nielsen, a University of Copenhagen biochemist, used the Danish Cancer Registry to analyze the medical records of 5.5 million Danes, including 295,925 cancer patients. Dr. Nielsen compared the cancer death rate, and the overall death rate, of the 18,721 patients with cancer who were also taking a statin drug, with the rest of the group who were not on statins. The results: the statin-cancer group had a 15 percent lower risk of dying.
ACSH s Dr. Gilbert Ross wonders why the respected New England Journal of Medicine would publish such a study and calls this pure data dredging of the grossest kind. This is a perfect example of how not to do epidemiology. (Data dredging is the mining of statistics so that searchers, after scouring a vast amount of data from various inputs vs. various outcomes, finally find some statistically significant linkage in the test sample. However, that shotgun method has a high probability of finding, out of thousands or millions of random combinations, some association, albeit purely by chance.)
