Jessica
I like the idea of using a case-control study to evaluate effects, though some aspects of your description are unclear. In a case-control, you'd enroll cases (people with a specific outcome such as psoriasis), and then select control people who would have been eligible to be cases in your study if they had developed psoriasis - and the way to identify controls is usually the most challenging aspect of a case-control study. Then you compare the exposure status, e.g., methylation patterns of cases and controls. The idea of embedding this in a long term study where they already collected samples at birth (and there are many such studies) would make it more feasible, or using tissue that is easily accessible in adults also makes sense. I am not sure to what extent the tissue that the disease affects corresponds with the tissue where the methylation pattern is altered. But immune function seems an appealing category - perhaps white blood cells would show different methylation patterns.
Is this related to your research?
Maria