Home • AltMed Home • Search

Scientific Studies

    There is not a single physiologic explanation that completely describes the phenomenon of the placebo effect.  There also is a minority opinion in the scientific literature that argues against the existence of a placebo effect and suggests that the response is due to statistical regression towards the mean, improper study design or a expectancy reaction to positive interactions with the physician.

    The original article on the placebo effect in 1955 examined 26 studies and determined that 35% of the patients treated with placebo responded.  In Beecher's study he identified placebo effects ranging from 21% to 58%.  Beecher's article has been criticized for not considering the natural course of disease and poor statistical quality of some of the studies he used.  A recent study by Kienle and Kiene reanalyzed the studies from Beecher's paper and concluded that "In contrast to his claim, no evidence was found of any placebo effect in any of the studies cited by him".  The authors of this paper further conclude that many of the mistakes made by Beecher and the studies he examined that lead to the appearance of a placebo effect can still be seen in much of the contemporary literature on the placebo effect.

    Benedetti has published several studies that support the opioid explanation of the placebo effect.  He found that respiratory depression following administration of a narcotic could be elicited with a placebo if the patient was conditioned with the opioid agonist buprenorphine. The respiratory depression caused by the placebo could then be completely blocked by naloxone suggesting that this response is mediated by endogenous opioids.  Another study by Benedetti examined the existence of a spatial placebo effect by applying a placebo cream and a analgesic cream to treat pain induced by intradermal capsaicin injections.  Placebo or analgesic cream was applied to one or two of the four capsaicin injected limbs in a double-blind fashion. Subjects consistently reported reduced pain in limbs treated with the placebo cream.  It was further found that these "spatial" placebo effects were alleviated with the administration of naloxone.

    Moerman's article "Deconstructing the Placebo Effect and Finding the Meaning Response" proposes what he refers to as a new perspective on understanding the placebo effect and uses published studies by other researchers to illustrate his points.  He argues that the placebo effect has less to do with the use of a placebo treatment and is instead related to "meaning".  Moerman defines the meaning response as "the physiologic or psychological effects of meaning in the origins or treatment of illness; meaning responses elicited after the use of inert or sham treatment can be called the "placebo effect" when they are desirable and the "nocebo effect" when they are undesirable".

    An article by Hrobjartsson and Gotzsche is a systematic review of clinical trials comparing placebo treatment to no treatment.  The trial identified 727 potentially eligible trials and finally included 114 trials in the review following the exclusion of trials that were unacceptable due to improper blinding, dropout rate >50% and other study design flaws.  This review examined trials that compared placebo pills, sham electrical nerve stimulation or psychological placebo (which was a nondirectional, neutral discussion between the patient and the treatment provider).  The authors conclusions were "We did not detect a significant effect of placebo as compared with no treatment in pooled data from trials with subjective or objective binary or continuous objective outcomes. We did, however, find a significant difference between placebo and no treatment in trials with continuous subjective outcomes and in trials involving the treatment of pain".  They further concluded that the effects attributed to placebo were larger in small trials and in trials that were not double blinded and attributed this to bias.

    A recent article published in BMJ by Kaptchuk compared the placebo effect of a sham acupuncture device to an inert pill.  This study examined 270 adult patients with persistent arm pain in a single blinded fashion.  All patients began the study with a two week placebo run in with either sham acupuncture (using a validated sham acupuncture needle that has a blunt tip and does not peirce the skin and has been clinically shown to be indistinguishable from a traditional acupuncture needle) and then wither continued placebo treatment or were switched to traditional acupuncture or amitriptyline.  The authors conclusions were that "The sham device had greater effects than the placebo pill on self reported pain and severity of symptoms over the entire course of treatment but not during the two week placebo run in".  Objective measures using a function scale and grip strength showed no statistical difference between the two groups.  Interestingly the study also found that the side effects (or nocebo effect) differed between the two groups and closely mimicked the differences in the informed consent forms of the respective placebo therapy.

    A 1983 article by Mcdonald makes the argument that the placebo effect is nothing more than statistical regression to the mean (i.e. the natural tendency for "patients selected for abnormalcy to improve").  The authors examined studies that used statistical or design methods designed to correct for regression and compared them to older clinical trials that did not take these precautions.  The study concludes that in the older studies a high percentage of patients seem to respond to placebo and that in the studies that corrected for regression there was no statistically significant placebo effect.

    The 2004 Cochrane review titled "Placebo interventions for all clinical conditions" concludes that "There was no evidence that placebo interventions in general have clinically important effects. A possible small effect on continuous patient-reported outcomes, especially pain, could not be clearly distinguished from bias".

    Several recent studies have used brain imaging to attempt to elucidate a physiologic mechanism that explains the placebo effect.  A study by Wager concluded that there is imaging evidence that the endogenous opioid system plays an integral role in the mechanism of the placebo effect and that there are changes in opiate release that occur directly in the regions of the brain most closely related to feelings, reward and pain.  Zubetia, using MRI and PET scans, has come to the conclusion that there is direct evidence that the endogenous opioid system, specifically opioid receptors acticity, mediates the placebo effect.  One image from his research can be seen below:

    Images of the brain responding to pain (left) and to pain plus a placebo (right) show activation of receptors that are part of the brain’s endogenous opioid system (red). Placebo-activated regions are associated with cognitive factors, such as expectation of pain relief. (Credit: Jon-Kar Zubieta/University of Michigan)