You generally can't really directly associate specific cases of disease with specific lifestyle choices.

EX: Some non smokers get lung cancer, so a percentage of smokers would have gotten cancer either way.