Close Window


Mechanisms of Placebo Effects

There has been a recent upsurge of interest in placebo effects (e.g., Blakeslee, 1998 a, b; Brown, 1998; Harrington, 1997; Holden, 2002; Moerman, 2001; Petrovic, Kalso, Petersson, and Ingvar, 2002). The word placebo (from the Latin, I shall please) has two related meanings: (1) a substance containing no medication but prescribed or given to reinforce a patient's expectation to get well; (2) an inactive substance or treatment used as a control in an experiment to test the effectiveness of a drug or medical treatment. Many reports show that placebos have significantly greater effects than giving no treatment, so rather than employing placebos simply as controls in experiments or clinical trials, there is increasing study of the mechanism(s) of placebo effects. As in many cases in biological psychology and biomedical sciences, no single mechanism accounts for all the effects.

Some of the proposed mechanisms are these: (1) As the text notes (pp. 241-242), it was found in the 1970s that placebo reduction of pain is mediated in part by endogenous opioids, demonstrated by the fact the effect could be reduced by the opioid-blocker naloxone. This suggested to some that the placebo effect is "real" because it is mediated by a known physiological mechanism. This could not be a complete explanation for placebo effects, because they occur not only for relief of pain, but also for other situations such as asthma attacks, where opioids would not normally relieve the symptoms. (2) It was also found in the 1970s that immunosuppressive effects can be conditioned to originally ineffective stimuli (p. 500). (3) The placebo might reduce stress, allowing the body to regain a natural, optimal level of functioning. Each of these mechanisms would be expected to be systemic, affecting the whole body, but cases were found in which the placebo effect held for only a specific part of the body. Thus, in an experiment in which a placebo in the form of a topical anesthetic was administered to one index finger and then shock was applied to both index fingers, most subjects reported less pain in the finger with the sham anesthetic (Montgomery and Kirsch, 1996). Such results support the interpretation that the expectancies of people cause placebo effects. The role of the basal amygdala in interpretation of stimuli and expectancies is discussed on pp. 483-485, and further work is to be expected on how expectations translate into bodily responses.

The variability of effects among placebo studies causes difficulties in accepting these effects. Moerman (2001) compared 117 double-blind placebo-controlled ulcer studies from all over the world. Doctors used the same drugs, the same placebo pills, and studied images of the stomach lining before and after treatment to rate effects. Among the studies, drugs caused improvement in 38-100%, whereas placebos worked from 0 to 100% of the time. The placebo effects were more variable than the drug effects, but among the studies the rates of healing for placebos and drugs correlated significantly (r=.40). The world average for placebo healing effects in these ulcer studies was 36%, and results for the United States were close to this value, yet placebos were effective in 59% of the patients in Germany but 22% in the neighboring counties of Denmark and the Netherlands, and only 7% in Brazil. To try to find whether national variation in placebo effectiveness is general across diseases, Moerman (2001) examined the results of controlled studies of treatments for moderate hypertension and for generalized anxiety disorder. Here the placebo effects for Germany were among the lowest for treatment of hypertension and middling for anxiety disorder, and no generality was found across nations. Differences in effectiveness of placebos have been attributed in part to cultural factors, including the differences in the enthusiasm and conviction with which physicians present the treatments, but the variability within nations indicates that this question is far from being solved.

In spite of the variability of results, Benson and Friedman (1996) urge physicians to harness the power of the placebo effect, since it is effective in so many situations, is safe and inexpensive and has withstood the test of time. Noting the negative connotations of the term "placebo effect," they suggest renaming it "remembered wellness."

Further understanding of the placebo effect is provided by a recent study in which PET brain imaging was used to compare the neural mechanisms involved in both opioid analgesia and placebo analgesia (Holden, 2002; Petrovic, Kalso, Petersson, and Ingvar, 2002). Six experimental conditions were included: heat pain and treatment with a rapidly acting opioid injected 40 sec. before the start of heat stimulation (pain-opioid [POP]), non-painful warm stimulation and opioid treatment (WOP) heat pain and placebo treatment with a saline injection (PPL), non-painful warm stimulation and placebo (WPL), heat pain only (P), and non-painful warm stimulation only (W). Heat pain was produced by pressing a metal stimulator at 48 degrees C on the back of the left hand for 70 sec. 48 degrees may not seem very hot, but the pressure on the back of the hand prevented reduction of heat by blood circulation, and subjects rated the pain an average of 65 on a scale where 100 meant unbearable pain (personal communication from Martin Ingvar, Feb. 18, 2002). The control stimulation was 38 degrees C. Although there was high inter-individual variation in placebo effectiveness among the 9 subjects, most showed lower pain ratings during the PPL condition than in the P condition.

Brain imaging showed that both opioid analgesia and placebo analgesia activated the rostral anterior cingulate cortex and a region in the pons. The authors suggest that in some cases cortical regions may exert control over analgesic systems of the brainstem, not only during opioid analgesia but also during placebo analgesia. Those subjects who showed a greater placebo effect also showed greater activation of the anterior cingulate cortex. Thus this study helps to understand a mechanism involved in the placebo analgesic effect.

The concept of effects of expectations on bodily processes has been extended to include the "nocebo effect" (from the Latin, I shall harm). This is the concept that an inactive substance or treatment can cause distress, illness, or even death (such as have been asserted to occur as a result of “voodoo” curses). Understandably, there has been much less research on the nocebo phenomenon than on the placebo effect. In one study, 48% of 40 asthmatics who were exposed to water vapor and told they were inhaling irritants or allergens experienced substantially increased airway resistance (Luparello et al., 1968). The twelve subjects who developed full-blown attacks were relieved by the same saline solution when it was presented as a therapeutic treatment. Thus, the same substance, presented differently, was either a nocebo or a placebo. Further experiments extended these findings (Luparello et al., 1970). Recently physicians have become concerned that nonspecific nocebo side effects of medications may distress patients, lead them to refuse to use prescribed drugs, and may cause physicians to discontinue what would otherwise be an appropriate therapy (Barsky, et al., 2002). They suggest that health care personnel can attempt to ameliorate nonspecific side effects to medications by identifying patients most likely to develop them and establishing a collaborative relationship with the patient to help understand and tolerate these bothersome but nonharmful symptoms.


Barsky, A.J., Saintfort, R., Rogers, M.P., Borus, J.F. (2002). Nonspecific medication side effects and the nocebo phenomenon. JAMA, 287, 622-627.

Benson, H. and Friedman, R. (1996). Harnessing the power of the placebo effect and renaming it "remembered wellness." Annual Review of Medicine, 47, 193-199.

Blakeslee, S. (1998 a). Enthusiasm of doctor can give pill extra kick. New York Times, Science Times, October 13, p. D4.

Blakeslee, S. (1998 b). Placebos prove so powerful even experts are surprised. New York Times, Science Times, October 13, pp. D1, D4.

Brown, W.A. (1998). The placebo effect. Scientific American, 278:1, 90-95.

Harrington, A. (Ed.)(1997). The placebo effect: An interdisciplinary exploration. Cambridge, MA: Harvard University Press.
Holden, C. (2002). Drugs and placebos look alike in the brain. Science, 295, 947-948.

Luparello, T.J., Lyons, H.A., Bleeker, E.R. & McFadden, E.R. (1968). Influences of suggestion on airway reactivity in asthmatic subjects. Psychosomatic Medicine, 30, 819-825.

Luparello, T.J., Leist, N., Lourie, C.H. & Sweet, P. (1970). The interaction of psychologic stimuli and pharmacologic agents on airway reactivity in asthmatic subjects. Psychosomatic Medicine, 32, 509-513.

Moerman, D.E. (2001). Cultural variation in the placebo effect: Ulcers, anxiety, and blood pressure. Medical Anthropology Quarterly, 14:1, 51-72.

Montgomery, G. and Kirsch, I. (1996). Mechanisms of placebo pain reduction: An empirical investigation. Psychological Science, 7, 174-176.

Petrovic, P., Kalso, E., Petersson, K.M. and Ingvar, M. (2002). Placebo and opioid analgesia -- imaging a shared neural network. Sciencexpress. Available on Internet at