Key points are not available for this paper at this time.
Using genetic variants as a proxy for modifiable environmental factors that are associated with disease can circumvent some of the problems of observational studies
Smith et al. (Thu,) studied this question.