ASA Connect

If you cannot establish that the exposures temporally preceded the development of the disease, it is a cross-sectional study.

Thank you!

Yours would be a cross-sectional study.  It may be helpful to view a case-control study as a retrospective, sub-study of a cohort study, where you oversample the people from the cohort who had the event of interest, and also take an appropriate sub-sample of those without the event. In some situations, the base cohort from which the cases and controls are sampled is obvious, but in some other situations it is not so explicit.  In the latter case, it would be difficult to generalize the inference from the case-control study to the target population, since the target population is not well characterized.  You should certainly consult a good Epidemiology text such as Rothman, Greenland and Lash's Modern Epidemiology for more in-depth understanding.

Hope this is helpful,

Ravi   

Hi.

This is the comment from Dr Moyses Szklo:

It is a case control study if cases are incident, with cross sectional ascertainment of the exposure - thus making it impossible to establish temporality. If cases are prevalent, it becomes a cross sectional case control study. 

I'll add to what others have said -- just in case -- that it's still a case-control study in the sense that the sample has pre-determined the proportion of outcome vs non-outcome datapoints; so you can't draw any conclusions about risk ratios for any covariate versus breast cancer (although you still have odds ratios).