Some of the benefits of medicine undoubtedly come from a patient’s belief that a particular drug will work as intended—easing arthritis or relieving wheezing, for example—but there’s a flip side to this placebo effect. It turns out that our belief in a drug’s side effects may actually cause us to experience them.

Recent research has demonstrated that when doctors and nurses inform patients about a laundry list of symptoms that a drug can cause—such as headaches, anxiety, dizziness, and nausea—they may unintentionally trigger these symptoms via the power of suggestion, according to a review paper published in the July issue of the journal of the German Medical Association.

The phenomenon, called the nocebo effect, draws on the “imagination and fears of patients,” study author Dr. Winfried Häuser, an associate professor of psychosomatic medicine at the Technical University of Munich, said via e-mail. But, he added, there’s no evidence that the nocebo effect produces quantitative physiological events associated with some drugs such as kidney stones or high blood pressure.

Advertisement - Continue Reading Below

“We see it as the evil twin of the placebo effect, but it’s much harder to study because you don’t want to put people in situations where you’re intentionally causing them harm,” said Dr. Ted Kaptchuk, a placebo researcher at Harvard Medical School.

A 2006 study he conducted to examine the benefits of sham versus real acupuncture treatments to relieve chronic arm pain found that one-quarter of those who were given the sham treatment experienced skin irritation, a side effect that can occur when acupuncture needles actually pierce the skin. The study also found that some of the participants who had their arm pain treated with placebo pills instead of the drug amitriptyline experienced severe fatigue that’s associated with the real drug.

“We had people who called us up while taking a placebo pill complaining that they were too tired to get out of bed,” said Kaptchuk. “We cut their placebos in half, and they complained that the pills were still too strong.”

The only way to completely eliminate the nocebo effect would be to keep patients in the dark about a drug’s potential side effects—which is clearly not ethical. In a March paper published in the American Journal of Bioethics, Kaptchuk and his colleague Dr. Rebecca Wells proposed that doctors practice what they call “contextualized informed consent,” which takes into account the possible side effects, the patient being treated, and the disease involved; doctors can then tailor the information they convey about side effects to provide the most complete picture of the treatment with the least potential to cause harm.

That, of course, can be tricky. Some patients known to develop every side effect they’re told about might benefit from being told less, but how much less?

Wells said patients don’t necessarily need to be informed about vague, non-specific symptoms such as difficulty concentrating, dizziness, and headaches because those symptoms can occur with any drug and aren’t life-threatening. On the other hand, she said, those considering statins to lower their cholesterol should be warned about muscle pain because that can be associated with a rare and dangerous condition. They should also be told about other dangerous side effects associated with certain drugs such as an increased risk of kidney stones if they take the anti-seizure drug topiramate or the possibility of stomach ulcers if they take ibuprofen long-term.

National Institutes of Health researcher Dr. Luana Colloca said doctors could use framing to help reduce the nocebo effect. “It’s different to say 5 percent of patients will have an adverse event than to say 95 percent of patients won’t experience any problems,” she said. Her research found that patients given an infusion of morphine for postoperative pain experienced far more discomfort after they were told specifically when the computerized pump was being turned off compared with those who were told the pump would be stopped at some point but not informed when.

Another study found that women in labor whose anesthesiologists warned them that their spinal epidural injection would feel like “a big bee sting” reported feeling more pain from the needle compared to those who were simply told that they were getting an injection that would numb the area without any pain warning.

Perhaps, added Colloca, simply telling patients about the nocebo effect could help mitigate it. “They need to know it’s a phenomenon in their brain and not necessarily due to the treatment they’re receiving.”