Monday, May 25, 2015

An Ambiguous "Scientific" Breakfast

Steak with scrambled eggs (made with milk and diced tomatoes), a corn muffin with butter, and coffee and/or tea sounds like a great breakfast. But as an article from Vox reports, all of those ingredients have published studies finding that they both cause and prevent cancer. Of course, I left out the wine. Perhaps it could counteract the rest of the breakfast!

Plotted above is the relative risk of cancer, that is, the probability of cancer by consuming one of the ingredients divided by the probability of cancer by not consuming it. A relative risk greater than one indicates that the cancer risk is greater with the ingredient. A relative risk less than one indicates that the cancer risk is less with the ingredient. As it is clear, in a collection of such studies, variability is large. Poor design, small sample size, small effect sizes, etc. can deceive, making us too quick to label, and publish, a cause or celebrate a prevention. Be careful following any new finding.

No comments: