I am aware of observational studies. But if you want to quote some, please find ones which examine total mortality and at least discuss the confounding factors.

E.g. this is a better example, but note that in this case they lump fish and chicken together under "white meat".