Tupelo3,

Thanks for finding this.

You write:
"This is probably too complicated for most of us (certainly above my pay grade)"
You do yourself a disservice: ask for a pay rise!

I played with it briefly. It was interesting following the links in the network. It's still early days, but it could grow into something important.

I've proposed in the past producing a belief network showing the linkage between any two factors involved in PD

Correlations could be used. They're not perfect: correlation does not imply causality and correlation is not transitive (A is positively correlated to B, and B is positively correlated to C, does not imply that A is positively correlated to C). Nevertheless, it would be interesting.

Can someone add to this?

did smoke ----> get PD ++++> constipation

A bonus mark to someone who can find a contradiction.

John