Interaction and confounding in logistic regression

#1
Hope someone here can help me out with the following problem!

I’m looking for explanatory factors for a certain dichotomous health outcome, the patient either has the disease or not. I have a set of about 12 possible factors. Now I want to select the factors that play role in the occurrence of this disease outcome. However, several of the factors are correlated (underlying disease and treatment are two of the factors, they don’t have a one-on-one relation but are correlated).
My question is: how do I take into account possible interactions and confounders when selecting factors with logistic regression?
So, I don't have a preselected independent factor and other factors which I want to correct for. Every factor can be an explanatory factor as well as a confounder etc...
Also, some factors need to be added as dummies, since they have multiple non-ordered levels.

Thanks for any suggestions!
 
Last edited: