How Placebo's Evil Twin Makes You Sicker

Whenever a pharmaceutical company tests a new migraine prevention drug, nearly 1 in 20 subjects will drop out because they can't stand the drug's side effects. They'd rather deal with the headaches than keep receiving treatment. But those suffering patients might be surprised to learn that the drug they've quit is only a sugar pill: the 5 percent dropout rate is from the placebo side.

Lurking in the shadows around any discussion of the placebo effect is its nefarious and lesser-known twin, the nocebo effect. Placebo is Latin for "I will please"; nocebo means "I will do harm." Just as the expectation of feeling better can make our symptoms ease, the expectation of feeling worse can make it a reality.

In a review paper published last week in the German journal Deutsches Ärzteblatt International, researchers say doctors and drug companies are unwittingly introducing patients to the demon of nocebo. Led by Winfried Häuser of the Technical University of Munich, the authors say that nocebo in the doctor's office can add unnecessary pain and distress to ordinary procedures. In clinical drug trials, it can create side effects that shouldn't be there—and perpetuate them in the patients who will take that drug in the future.

Chemically, nocebo seems to use the same toolkit that placebo does. Say you have a headache and treat it however you normally like to—maybe with an ibuprofen, or a few drops of homeopathic whatever under your tongue. If you expect to start feeling better soon, your body will use internal molecules such as dopamine and opioids to start creating its own pain relief. (Depending on what treatment you've used, you may or may not get some chemical backup once it kicks in.) It's good old-fashioned conditioning, just like Pavlov's hungry dogs salivating before food was anywhere in sight. But in nocebo, when you expect your headache to get worse, your body turns the pain-relief machinery down instead of up.

Nocebo doesn't need a doctor's help to find you. But a doctor can harness it too. The standard assumption in medicine, Häuser and his coauthors write, is that patients should be warned ahead of time about anything painful ("You're going to feel a little pinch!"). But telling a patient to expect discomfort might actually make it worse. In one study, patients getting an injection felt more anxiety and pain when their doctors used words such as "sting," "burn," or "bad," even if the doctor was only trying to express sympathy.

In another study, women receiving epidural injections felt more pain when they were warned that the "big bee sting" would be the worst part of the procedure. When women were instead reminded that the injection would numb them and make them more comfortable, they experienced less pain. The authors point out that patients in emergency situations or facing major surgery are often in a "trance state" that makes them even more suggestible than usual.

Nocebo can really throw a wrench into clinical drug trials. Placebo is well accounted for; these trials always include a large placebo group in which patients are given a sugar pill or other fake treatment. To minimize the effect of suggestion, neither doctors nor patients know which group they're in. After the trial, researchers can subtract the positive effect seen in the placebo group from that in the patients taking a real drug, and see how much good their treatment really did.

In both the real and placebo groups, subjects report any side effects they experienced. When drug companies report the results of their trials, the Food and Drug Administration asks that they only report side effects (or "adverse events") that they have some reason to believe were caused by the drug. But the FDA acknowledges that this is "a matter of judgment."

As we saw with the migraine patients, side effects can be common even with a sugar pill. In one study, 44 percent of lactose-intolerant people reported gastrointestinal problems after taking a fake lactose tablet. (Impressively, a quarter of people without lactose intolerance also reported digestive troubles after taking the tablet.) And in a somewhat cruel prostate drug study, one group of subjects was told that sexual dysfunction was a possible side effect, while the other group wasn't. The better-informed group reported sexual side effects at a rate of 44 percent, compared to only 15 percent in the blissfully ignorant group.

Whatever side effects are attributed to a new drug, doctors may increase patients' odds of feeling those effects just by mentioning them. In Germany, Häuser says, "most of the product inserts contain very many potential non-specific adverse events, raising the risks of nocebo effects." So how can doctors avoid making their patients sicker?

Häuser and his coauthors have a couple of suggestions. Patients could consent to not be informed about mild side effects, knowing that just hearing about these effects makes them more likely. And doctors can phrase their warnings more positively, emphasizing that most patients respond well to a treatment rather than focusing on potential negatives.

"Doctors can and should be trained to positively use the power of their words," Häuser says. If we know where nocebo is lurking, we may be able to keep it far away.


Winfried Häuser, Ernil Hansen, & Paul Enck (2012). Nocebo phenomena in medicine: Their relevance in everyday clinical practice. Deutsches Ärzteblatt International : 10.3238/arztebl.2012.0459

Image: takgoti/Flickr