Scientists say they have found that the "placebo effect" of dummy drugs can relieve anxiety as well as pain.
Simply taking a tablet may make people feel better
The effect is when a person is successfully treated by a dummy drug, just because they believe it works.
Swedish volunteers were shown a series of unpleasant pictures and then given an anti-anxiety drug. The test was later repeated, but with a fake drug.
The effects on calming the people's nerves were fairly similar, the scientists told the journal Neuron.
However, there is still controversy over whether the placebo effect actually exists.
In the Swedish experiment, Dr Predrag Petrovic and colleagues at the Karolinska Institute showed their volunteers images of mutilated bodies, in order to measure anxiety, which they asked the participants to rate on a scale of 0 to 100, 100 being the most unpleasant.
They then gave the subjects a genuine anti-anxiety drug - a benzodiazepine - and told them that it should reduce any unpleasant emotions.
They then administered an antidote to the benzodiazepine, telling the volunteers that this would restore the unpleasant emotions.
They repeated the test the next day with the same subjects in the same way, telling them that they were receiving the same drugs.
However, instead they were given dummy drugs.
Benzodiazapine had reduced the average unpleasantness rating from 51 down to 29. This reverted back to almost 61 after the antidote was given.
The placebo had a similar effect - unpleasantness rating dropped to 36 after the placebo and rose back up to 51 after receiving the fake antidote.
During all of the experiments the subjects' brains were scanned using functional MRI, which shows blood flow.
The scans revealed that the placebo reduced activity in the brain's emotion centres and this reduction correlated with the unpleasantness rating, meaning subjects who reported the largest placebo response also showed the largest decrease in activity in the emotional centres.
Dr Petrovic said: "The placebo changes what we expect. When we expect that something unpleasant should become less unpleasant, it really does."
He said it was unlikely that placebos could be used to treat anxiety, because of the ethics involved of telling patients they were being given a treatment but not revealing that it is a sham.
However, he said the findings might help with finding better drugs and better cognitive treatments that activate similar centres in the brain as the placebos.
Dylan Evans, senior lecturer in Intelligent Autonomous Systems at the University of the West of England, who has written books about the placebo effect, said: "It does provide much better proof that placebo can relieve some certain forms of anxiety.
"There has been some evidence of this before going way back to the 1970s.
"Researchers at Newcastle found patients with anxiety responded better when given green tablets than red or yellow, even though they were the same in every other way."
He agreed that there were ethical problems with giving placebos as therapies.
"Doctors have a duty to care for their patients but they also have a duty to tell them the truth.
"Placebo seems to pull those two duties in opposite directions."
He said it was unclear why the placebo effect worked for some conditions and not for others.
He suggested that it might be that the effect is active for conditions that involve activation of something called the acute phase response - the inborn immune system that reacts crudely through things like swelling and pain.
However, he stressed that his was only one theory and that it needed testing.