st: generalised estimating equations with unequally spaced data

[NL Turner <Nicholas.Turner@bristol.ac.uk>]

Hi all,

I have a question about interpreting results from generalised estimating 
equations with unequally spaced data

My understanding is that normally with equally spaced time intervals e.g. 
on an annual basis, then a repeated exposure-outcome association can be 
modeled using in stata with the xtgee command to give an average effect 
over time e.g. the average effect of the exposure on outcome 1-year later. 
However, with a data set that I would like to use I have unequally spaced 
data, there is data at age 12-13 then 14-16 i.e. a 1-year interval for the 
first two measures and then a 2-year interval for the second two measures. 
I know that with certain correlation structures the xtgee command can cope 
with unequally spaced data but I'm not sure what the 
interpretation/time-frame of the output would be? Would the output be 
giving the average effect of the exposure on outcome 1.5-years later or 
something else?

Thanks
Nick


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