Here’s a study: http://www.reuters.com/article/idUSTRE66C66D20100713 that shows some of the common errors researchers make when studying acupuncture. The first problem is their “sham” protocol was not sham, the second problem was their “real” acupuncture was not real.
Regarding the latter point first. When a skilled acupuncturist gives a treatment, he or she diagnoses the patient’s condition according to Chinese medicine’s theory and provides an individualized treatment for the person’s specific condition. With respect to infertility, there are at least half a dozen main possible causes according to Chinese medicine, which would each call for a different treatment. One does not just put needles into points “that are related to fertility” and do the same thing for everybody.
With respect to the “sham” acupuncture, it’s been well established that needling into almost any point has some beneficial, non-placebo effect. Secondly, it is not clear that the authors of the study understood acupuncture theory well enough to select points that would have no effect at all on fertility (and depending on the patient’s condition, almost any point would be likely to have some effect).
The authors acknowledge a possible issue with their sham protocol, but conclude that the way to remedy this is to use a sham protocol where needles aren’t inserted into the skin. However, some styles of acupuncture, including the one I specialize in, include non-inserted needle techniques; in addition, some studies have shown non-inserted techniques to have therapeutic efficacy.
Unfortunately, these types of problems occur over and over in the field of acupuncture research.
Manu's Alternative Medicine Blog 