Well yeah, I have been at many meeting where correlations are used and some correlations are really strong when data is mapped out and statisticians come in. What do you want somebody to hit you on the head, just kidding of course but I don't think what we are looking at is cherry picking either. Obviously we should do whatever we want, and that is the way medicine should go, patient decides path, but sometimes correlations help (and not just fire engines at fires). I chose Tulsa Pro 1.5 years ago, but no docs are advising Tulsa just yet till something hits them in the head and gives them the machines and training. Things take time on all these things, but we make our own paths. I have really bad inherited cholesterol, so maybe I inherit one source of the problems that contribute, yet eating plant food since start of the year and I have the best cholesterol in 30 years+. So maybe things help and correlations are pretty good too. Maybe you inherit a whole bunch of other things and it isn't related to fats or foods for you but some other gene, or radon gas you were exposed to as a kid, or who knows? I don't, we just work with data as we see for ourselves.