Reply #101: Oh go back to school...

[skids]

We aren't idiots here. A proper review of anecdotal evidence would only be used to find clues as to where to look with a proper follow-up clinical study, or what data to start collecting for a proper statistical treatment. More routine data collection by the health care establishment would help tie those anecdotes to co-variables and multiply the value of the data, not to mention be useful in its own right for proper statistical modeling.

No correlation does not imply causation, however it is a good indicator of where to look for either causation or cofactors. For example, take vitamin D. They looked at the breast cancer correlation to latitude, tried testing suppliments in older women and found that was a no go. There's something there but it's not adult D bioavailability.

Then they did the same for multiple sclerosis, also correlated negatively with sun exposure, and lo and behold we find that prenatal vitamin D helps correctly express genes that prevent MS. Science wins because someone didn't treat a correlation as meaningless noise.

Perhaps someone will put two and two together and look at breast cancer gene expression vs prenatal vitamin D exposure.

"Correlation is not causation" is an important thing to know. As a mantra, however, it's destructive.