Call it the "negative placebo effect." Patients who don't believe their medicine will work may find they're actually right.

The placebo effect has been well documented. It's the effect seen when patients given dummy sugar pills actually report relief from symptoms, purely because they thought the pills would work.

But this study find the opposite can be true too: Patients who make up their minds that a highly effective painkiller won't relieve their pain will report that in fact, it didn't.

And brain imaging techniques used in the study back that up, showing that the brain can agree with a mind.

To test the "nocebo effect," Irene Tracey of the Centre for Functional Magnetic Resonance Imaging of the Brain at Oxford University gathered a team of British and German researchers to conduct some experiments.

They first recruited 22 volunteers who were willing to put up with a little pain in the interests of science.

They hooked the volunteers up to an IV line and functional MRI brain scanners to monitor their brain activity. They then used a laboratory heat device to induce pain in the lower legs of the volunteers.

When they asked the volunteers to rate their pain a scale of 1 to 100, on average, they rated it a 66.

For the first part of the study, the scientists began infusing the volunteers' IVs with a powerful, fast-acting synthetic opioid painkiller called remifentanil -- without telling them.

The participants reported feeling slightly less pain, rating their pain at about 55.

Next, the volunteers were given remifentanil and told they were being given a painkiller. That dropped their reported pain to 39 — even though the volunteers were receiving the same painkiller in the same way as the first experiment.

Finally, they were again given the painkiller in the same way, but told they were receiving no drug and might even experience more pain.

The volunteers reported their pain intensity increased to 64 -- almost as bad as it had been immediately after the heat device was first used – even though the drug was still being administered at the same dose.

The brain scans revealed that the volunteers' pain networks matched the volunteers' reports of pain. Activity in the brain's pain networks went down when the volunteers thought they were getting pain relief, and went back up again when they were expecting and reporting pain.

The results appear in the journal Science Translational Medicine.

The study authors say their experiments show the power of the nocebo effect and say that doctors need to be aware of the power of expectations on their patients.

"Doctors shouldn't underestimate the significant influence that patients' negative expectations can have on outcome,' Tracey said in a statement.

"For example, people with chronic pain will often have seen many doctors and tried many drugs that haven't worked for them. They come to see the clinician with all this negative experience, not expecting to receive anything that will work for them. Doctors have almost got to work on that first before any drug will have an effect on their pain."

Tracey says there may also be lessons for the design of clinical trials.

Such studies are often carried out comparing an experimental drug against a placebo to see if there is any effect of a drug above and beyond that of the placebo. This study suggests that volunteers' expectation could skew the results.

"We should control for the effect of people's expectations on the results of any clinical trial. At the very least we should make sure we minimize any negative expectations to make sure we're not masking true efficacy in a trial drug," she said.