I don't think there is a "general" answer to how you calculate something as complicated as the dielectric constant, the latter is -in general- frequency and temperature dependent and what model you use will depends on which processes are involved for that particular frequency/temperature.
Rememeber that all absorbtion processes will also give rise to a change in the dielectric constant, so it is probably a good idea to first think about absorbtion and then worry about the dielectric constant.
It is fairly straightforward to e.g. calculate the change (and in some cases even the absolute value) if you can model your system as an uncoupled ensemble to two-level systems. The way to do this is to first calculate the absorbtion via the Bloch equations (textbook stuff if you are dealing with an ensemble of spin 1/2 systems) and then use the Kramer-Kronig relations to get the permittivity.
I don't know much about cold atoms, but presumably it should be possible to use a relatively simple model (maybe even just the Bloch equations) for the absorbtion and still get a reasonable answer, right?