That's true though it's still interesting the three patients cited in the paper they reference improved. Here are some better links perhaps

http://alzheimer.neurology.ucla.edu/Curcumin.html

http://www.ncbi.nlm.nih.gov/pmc/articles/PMC2781139/