st: Performing mediation analysis while using the clogit command in Stata

["Pillas, Demetris" <d.pillas06@imperial.ac.uk>]

Dear all,

For my analysis, I am currently examining the association between a disorder and mortality. Since the data I'm using follows a matched case-control group design, I am using the clogit command [since the clogit command fits conditional (fixed-effects) logistic regression models to matched case-control groups]. I ran the analysis and I identified a significant association.

However, I would also like to examine whether the significant association identified through running the above clogit analysis is mediated by other variables (e.g., behavioural / psychological variables) (it does not have to be that many variables mediate the association, it could be just one).

Is there any way to test for this 'mediation effect' in the stata software while using the 'clogit' command?

Thanks in advance,

Demetris

*
*   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/