power-of-placebo-effectTruly understanding placebo effects (note the plural) is critical to science-based medicine. Misconceptions about placebo effects are perhaps the common problem I encounter among otherwise-scientific professionals and science communicators.

The persistence of these misconceptions is due partly to the fact that false beliefs about placebos, namely that “the” placebo effect is mainly an expectation mind-over-matter effect, is deeply embedded in the culture. It is further exacerbated by recent attempts by CAM proponents to promote placebo-medicine, as their preferred treatments are increasingly being demonstrated to be nothing but placebos.

One idea that proponents of placebo medicine have tried to put forth is that you can have a placebo effect without deception. The study most often pointed to in order to support this claim is Ted Kaptchuk’s irritable bowel syndrome study. However, this study was flawed in that it told participants that placebos can heal, so it wasn’t exactly without deception.

The role of conditioning

A recent study published in the Journal of Pain focuses on conditioning as a component of placebo effects. The study design is interesting, but I have doubts about its applicability to the clinical setting.

The researchers used a heat model of pain applied to the forearm. They did various runs where participants reported their pain level in response to different temperatures. Their forearms were treated first either with a control cream (petroleum jelly) or a placebo cream (petroleum jelly plus blue food coloring). They were told the placebo cream contained a powerful analgesic that would block pain. The cream was then removed, and a hot stimulus applied.

When subjects were treated with the placebo cream they believed to be active, they were then tested, without their knowledge, with a warm but not painfully hot stimulus. This was meant to convince them that the placebo cream worked. (So again, the protocol used active deception to achieve its ends.)

One group of subjects experienced a single set of runs in this fashion. Another group experienced four sets of runs spread out over different days. Then finally came the test – all test subjects were told about the deception, that the placebo cream was inactive and in fact identical to the petroleum jelly except for the dye. They were then given another run with the control and placebo cream.

The researchers found that in the group that had the long run (four-day series) subjects still experienced reduced pain sensations with the placebo cream, but not the control cream. There was no difference for the short run (single day, single series) subjects. There were controls in place to rule out simple habituation to the stimulus as a factor.

What did we learn about managing pain clinically?

If we take the results of the study at face value, what do we learn? The study does not establish that you can have placebos without deception. The method used in this study depends explicitly on deception.

What the study does potentially show is that conditioning can play a role in placebo effects. This idea is nothing new, as conditioning has been on the list of placebo effects for years (certainly since I have been writing about placebo effects). The study does demonstrate that conditioning alone, without expectation of benefit, is sufficient to produce at least a temporary effect for a subjective symptom (pain, in this case). This is entirely unsurprising, but it’s good to have an experimental verification.

Conditioning probably plays a significant role in many placebo rituals, such as acupuncture, or even non-placebo treatments such as taking medication or getting a valid medical procedure. The process of the treatment becomes associated in our minds with feeling better, and so experiencing the process makes us feel better. Perhaps the conditioned stimulus triggers the release of endorphins, for example.

The main limitation is that the conditioning in this case required a contrived situation, in which patients were deceived by receiving a non-painful stimulus they were told would be painful. They therefore attributed the lack of pain to the placebo analgesic. How would we apply this in the real world?

I suppose one way to accomplish a similar effect is to give the patient a real analgesic along with an associated placebo treatment. Then after sufficient time for conditioning to take place, give the placebo treatment without the analgesic. According to this study, however, you will have to secretly give them a real analgesic, an ethically dubious practice.

There are other ways to shift from active treatment to placebo, but they were not the subject of this study. Follow up studies that attempt to remove all deception would be interesting.

Conclusion – What do we know about placebos?

Taken together, the scientific literature on placebos indicates that it is a complex assortment of various effects. These include conditioning, as well as reporting bias, statistical effects such as regression to the mean, confounding factors, observation bias, and other effects. Expectation of benefit is only one element, and is not necessary by itself.

Often the absence of expectation is used by the naïve public to argue that placebo effects are ruled out, but this is not true. For example, it is frequently argued that babies and animals cannot have placebo effects because they cannot have expectation, but there are many other sources of apparent placebo effects, as this study partly demonstrates.

We further know that placebo effects are only measurable for subjective outcomes. Placebo effects won’t cure cancer or make you live longer. They may cause you to report less pain or nausea, however. Whether you are actually experiencing less pain or just reporting less pain is unclear. Placebo effects are also short lived.

The ultimate question is whether or not placebo effects are clinically valuable and whether attempts to provoke them are worthwhile. My position is that they are of severely limited value, and are not worth compromising the relationship with the patient by incorporating deception into the treatment. It is certainly not worth instilling in the patient false and pseudoscientific ideas about health and medicine.

Any placebo effects worth having can be achieved with legitimate treatments given without deception to fully informed patients.



Posted by Steven Novella

Founder and currently Executive Editor of Science-Based Medicine Steven Novella, MD is an academic clinical neurologist at the Yale University School of Medicine. He is also the president and co-founder of the New England Skeptical Society, the host and producer of the popular weekly science podcast, The Skeptics’ Guide to the Universe, and the author of the NeuroLogicaBlog, a daily blog that covers news and issues in neuroscience, but also general science, scientific skepticism, philosophy of science, critical thinking, and the intersection of science with the media and society. Dr. Novella also contributes every Sunday to The Rogues Gallery, the official blog of the SGU.