Same answer as in your question concerning PD. The difference is that since most of the new drugs (except acetyl choline esterase class) fail in AD, these publications are probably considered authentic? It is only when the trial results are positive, we start looking for flaws and bias in the study. I have covered some of the failed trials in AD in my review of a new drug Bapineuzumab knol and Solanezumab (Medpedia article).

Bapineuzumab (Pfizer, J&J, Elan) Review
http://knol.google.com/k/krishan-maggon/bapineuzumab-pfizer-j-j-elan-review/3fy5eowy8suq3/119#

http://goo.gl/ZM8v.qr

http://goo.gl/ZM8v

If you wish to find an overview of effectiveness existing or new class of drugs is best to start with review articles which can be found on NIH Pubmed, and through sources such as Cochrane reviews. All clinical trials have some degree of bias in what is, or is not, disclosed.. There are certain number of trials on Alzheimer's disease conducted by academic groups where data is disclosed without much bias that can be found through Pubmed searches; see http://www.ncbi.nlm.nih.gov/pubmed/19328435 as an example. Clinical data from trials of proprietary compounds are typically only selectively disclosed by companies, making accurate evaluation of the clinical trials somewhat difficult.