Validity of the scientific category of evidence

As outlined in one of my previous blogs, Roy Walford in his textbook entitled "The 120-year Diet " states evidence falls into 4 categories namely,"... testimonial, argumentative, correlational and experimental..."-the latter being the gold standard normally required before the medical profession will consider recommending the use of a new drug and/or medical procedure. Let us analyze the experimental design. It includes what is normally called "the double blind procedure ", where neither the investigator nor the subject(s) know who is receiving the drug and/or who is just getting the "placebo ". I should mention that when it comes to a surgical procedure , this anonymity is achieved for example in the case of cardiac procedure like coronary artery bypass by just making a surgical incision of the skin and not doing the coronary artery bypass. Life Extension is one of the many newsletters that I receive or a regular basis. These newsletters- which I highly recommend to all investigators present fascinating evidence of medical benefits. For example the importance of the level of Vit D in Alzheimer's and the value of exercise in delaying its onset, etc. Recently one of their weekly newsletter discussed the importance of knowledge of your blood groups which suggested that those with type 0 had a lower risk of heart disease compared to type A,B or AB with those with the rarest blood type namely AB having the highest increase of heart disease risk at 23%! My question is: Do the investigators know the frequency of these and other factors in their 2 groups of subjects? There are literally thousands of physiological processes going on in each one of us which may be slightly different, similarly the home environments may be different as well. In short , in my view there is no such thing as the experimental category because the investigators cannot hope to know all the details of their patients.The only closest approach to such experiments would be conducting these experiments with identical twins, unless of course we had the luxury of using inbred strains