This article describe why many of the "benefits" of alternative medicine cannot be distinguished from the placebo effect.
Swear-by-it stories and anecdotal reports of benefit are one thing. Proving a treatment helps is quite another. Many alternative medicine studies have not included a placebo group — people who unknowingly get a dummy treatment so its effect can be compared.
Acupuncture is especially hard to research. Positive studies tend to lack comparison groups that have been given a sham treatment. Or they are often done in China, where the treatment is an established part of health care.
One U.S. study found that true acupuncture relieved knee arthritis pain better than fake acupuncture, in which guide tubes were placed but no needles were inserted. But a European study involving twice as many patients and using a more realistic sham procedure found the fake treatment to be just as good. The conclusion: Pain relief was due to the placebo effect.
The one bad thing about reports in mass media news like this is that they never give exact citation for all of these research and results that they are quoting. It's as if we are expected to just accept these things without having to know the source of what they used to draw up their conclusions.