Couldn't there be countless other differences due to environment, food, culture, genetics, etc that would be hard to separate? I'm no scientist but I was under the impression that the control group should be as close as possible a cohort other than what is being tested.