Fake acupuncture works just as well as the real thing in relieving migraines, scientists have found.
Needles are inserted at specific sites on the body
In a study of more than 300 patients, both genuine and sham acupuncture reduced the intensity of headache compared with no treatment at all.
But real acupuncture was no better than needles placed at non-acupuncture points on the body, the Journal of the American Medical Association reports.
It goes against recent research showing acupuncture works in its own right.
It has long been debated whether acupuncture works in a unique way or whether any benefits gained are merely down to a person's expectation that the treatment will work.
The latter is called the placebo effect.
To investigate this, Dr Klaus Linde and colleagues randomly assigned 302 people with migraines to one of three groups.
One group received 12 sessions of genuine acupuncture over eight weeks.
Another group received 12 similar sessions of acupuncture, except the needles administered were not placed in parts of the body thought to relieve migraine pain, hence any benefit would likely be placebo rather than real, according to the researchers.
The third group received no treatment but were placed on a waiting list to see a migraine doctor.
All of the patients kept diaries about their migraine symptoms.
While the patients on the waiting list continued to have headaches just as often, the ones who received acupuncture - sham or real - had fewer headaches.
The average number of days blighted by a headache went down from about five to two.
This may be due to "non-specific physiological effects of needling, to a powerful placebo effect or a combination of both", said the researchers.
But Dr George Lewith, who recently published work suggesting acupuncture has an effect above and beyond placebo, said although the present study was well conducted, it did not truly test the placebo effect.
"We do not know whether this sham acupuncture is active or not. To test for placebo effect you have to use an intervention that only raises a patient's expectations.
"The authors note themselves that something else could have been going on as well."
The British Acupuncture Council said that using pre-prescribed acupuncture points for all patients might have skewed the results.
"Acupuncture treatment is different for each person. The formulaic treatment part of the study would be inappropriate for some patients. This would reduce the apparent effectiveness in the acupuncture group."
The council said there was good evidence to suggest acupuncture was helpful for treating migraine.
Ann Turner of the Migraine Action Association said: "Acupuncture may be a good treatment option for migraine sufferers to explore."
More than one-in-10 people in the UK experience migraines, two-thirds of whom are women.