This is basically just facilitating people data dredging their own health. If you measure one outcome variable and enough "candidate" inputs you'll eventually f

This is basically just facilitating people data dredging their own health. If you measure one outcome variable and enough "candidate" inputs you'll eventually find something that looks like a pattern. As a doctor I measure CRP when I'm specifically looking for something - but that's in the context of having formed a differential diagnosis, then having considered the pre-test probably of each differential, and forming a view that the test result with alter the overall likely outcome. The last thing I would even want to do (for me or my patients) is to start randomly measuring CRP when generally well. Why not just eat the ginger?