st: how can I compute a risk difference with a confidence interval from a logistic model with a categorical*continous interaction?l

[Eva Katz <ekatz@unc.edu>]

I've been using prvalue command in stata to compute risk difference (RD) 
estimates with confidence intervals from a logistic model (citation: 
Confidence Intervals for Predicted Outcomes in Regression Models for 
Categorical Outcomes Jun Xu and J. Scott Long Indiana University August 
23, 2005). I have a model with an interaction term (dichotomous x 
continuous variable) and I would like to compute the RD for each level 
of the continuous variable - compared to a referent value - within each 
level of the dichotomous variable. For example, my continuous variable 
is BMI and my categorical variable is a genetic factor (present/absent) 
I want to know the RD associated with each level of BMI (compared to the 
referent BMI of 21) among individuals with the genetic factor and among 
individuals without the genetic factor. However, I do not know how to 
get a CI when the genetic factor is present.
How do I do this?

thanks!
*
*   For searches and help try:
*   http://www.stata.com/help.cgi?search
*   http://www.stata.com/support/statalist/faq
*   http://www.ats.ucla.edu/stat/stata/