Please remember that statistical analysis is not intended to prove anything, just to provide evidence for something. Of course double-blind testing doesn't establish whether or not a difference exists, especially when the situation in point is only a case study, and a very VERY small sampling in the study as well.