Even when you know it’s fake: The strength of the placebo effect

Almost everyone has heard about the placebo effect – the finding that treatment that have no particularly relevant effect (like a sugar or vitamin pill, or a behavioral equivalent) can make patients feel better. The placebo effect is actually the reason that all FDA approved drugs have to go through a double-blind placebo-controlled clinical trial before being approved for use – It has to be shown that using a specific medication is more beneficial than a non-active placebo even when the experimenter (or doctors) and the patient have no idea which treatment the patients are receiving. Otherwise, companies could simply continuously create placebos, show that they produce improvements in patients, and bring in the dollars.
Until now, it’s been assumed that in order for placebos to work, the patients have to be told that they are effective medications, amounting to an unethical lie by the doctors that prescribe them. This is the reason that very few doctors use pure placebos, though in a recent survey more than 50% of doctors reported using mild prescriptions (like over-the-counter pain medications) that they don’t believe are actually relevant to the condition as “impure placebos.” Given the large placebo effect, we can expect that these treatment, even if unethical, resulted in significant improvements in conditions that those same doctors were unable to treat using conventional methods. Still, it doesn’t feel good to know your doctors have to lie to you to make you feel better, right? Well they might not have to.

This fake placebo pill might make you feel better

A very recent study conducted at Harvard and published in the journal PLoS One examined whether placebos would still work for irritable bowel syndrome (IBS) patients even if they were explicitly told that the pills were inactive. The 80 IBS patients were randomly selected to receive either no treatment (the control condition) or a pill they were told was inactive “like a sugar pill” without any medication in it. Patients in the placebo condition were also told about the placebo effect and that such inactive sugar pills have been shown to produce significant mind-body self-healing processes. The placebo pills were marked with clear labels that read “placebo pills” so that there would be no confusion and so that patients would be constantly reminded that they pills they were taking were placebo pills 2 times a day. Amazingly, the placebo effect was still found to be present.

At the end of 3 weeks of treatment, participants in the placebo group reported significantly greater improvement in their symptoms on a number of different scales used for IBS as well as an overall Quality of Life measure. Even more amazingly, nearly twice as many patients in the placebo condition (59% versus 35%) reported that the treatment gave them “adequate relief” from their IBS symptoms!!! Patients got better taking pills that did nothing even though they knew that the pills were completely inactive. In fact, the effect was so great it was similar to the effects commonly seen with actual approved medications for IBS.

Conclusion and implications

Overall, these findings are encouraging on a number of fronts: If confirmed by future studies and using other conditions, these results suggest that doctors don’t have to lie to their patients since the placebo effect is strong enough to work even when patients are aware of it as long as the person giving the treatment is trusted. In fact, even this last piece should probably be put to the test since as far as I know, it has never been put to the test itself. Still, there is a whole body of literature in psychology telling us that when people of authority, people we respect, tell us to do something we react favorably.

Unfortunately, this has resulted in some pretty disturbing results including Stanley Milgram’s 1960s studies that revealed the power of authority in directing people to harm others (if you’ve never looked into these experiments you really should – here’s a video link to get you started). However, this time the results suggest that doctors may be able to help patients for whom there is no specific approved treatment. Obviously this is of interest to me because of the relatively low success in clinical addiction treatment. It may actually also suggest that a number of the treatments we’re using now – especially those that have never been subjected to a Randomized Placebo-Controlled Clinical Trial – may in fact simply be placebos. I have my own sneaking suspicions about some specific treatments…

Like every other study, this one is not without limitations. First of all, these results have to be replicated with other conditions and in other hospitals with other patients before doctors should feel comfortable simply prescribing placebos. There could be specific aspects of this sample that made the results so convincing. Indeed, even the no treatment condition got some benefit from their interaction with the medical and research staff involved in the study (or maybe just due to the passage of time). It’s as if we are now comparing the placebo effect to the time/human-interaction effect, gradually making our way towards some “no-effect” condition.  We also want to see these sort of results with a larger sample, and although conducting a double-blind trial isn’t possible (one group takes pills and the other doesn’t so everyone knows who is in what condition), I’m certain that future experiment will figure out more subtle controls.

Still, this study definitely indicates that giving placebos openly can work for patients who are open to it. It should also make everyone aware that just because someone tells you a treatment is objectively helpful doesn’t mean it is. In the end though, if the psychological placebo effect brings about actual improvement in symptoms, shouldn’t we consider it effective?


T. J. Kaptchuk, E. Friedlander, J. M. Kelley, M. N. Sanchez, E. Kokkotou, J. P. Singer, M. Kowalczykowski, F. G. Miller, I. Kirsch, A. J. Lembo (2010) Placebos without deception: A randomized controlled trial in irritable bowel syndrome. PLoS One, 5.


© 2010 Adi Jaffe, All Rights Reserved

Adi’s Mailing List |  Adi’s eMail |  Follow Adi on Twitter

Become a Fan on Facebook |  Connect with Adi at LinkedIn