Defining Placebo

Defining Placebo



The term “placebo is a Latin word meaning “I shall please.” In medicine, it implies “I shall please the one who asks me to prescribe a treatment.”



A placebo is a therapeutic measure of zero or low intrinsic efficacy with no logical connection with the disease. It is designed to let the patient believe that it is an active treatment. “A placebo is not only an inert substance, but also a mode of intervention with a variety of sensory and social stimuli which inform the patient that the treatment will be beneficial” (Benedetti et al., 2011 p.339).

There are two types of placebo. A pure placebo is a treatment that has the appearance and taste of the treatment it replaces, but is devoid of any active substance (e.g., saline, starch, sugar, lactose, acupuncture in non-active zones). An impure placebo is an active treatment which theoretically proves ineffective for a given pathology (e.g., vitamin C prescribed for chronic migraines; vitamin C is known instead for its efficacy in the treatment of scurvy).

 The placebo effect

The actual placebo effect corresponds to “the psychobiological phenomenon occurring in the patient’s brain after the administration of an inert substance, or sham physical therapy, such as sham surgery, along with verbal suggestions of clinical benefits” (Benedetti al., 2011, p.339). It results in an improvement of the patient’s condition. It corresponds to the positive variance between the observed therapeutic result and the expected therapeutic effect according to biological models (Lemoine, 1996). A negative variance is the consequence of the nocebo effect.

Avoid confusion

Placebo products cannot be effective without the contribution of the placebo effect. By contrast, an active product may be effective because of the placebo effect. If the administration of aspirin should calm a patient within 10 minutes, that would be due to the placebo effect – and not to the active ingredient of aspirin. Indeed, the physiological action delay of the active substance in aspirin is approximately one hour.

 One should not confuse placebo and placebo effect.

Placebos are also in non-pharmacological interventions

Placebos are also found in non-pharmacological interventional research. They are useful to assess the actual efficacy of non-drug interventions. For example, in randomized controlled trials focusing on acupuncture, experimenters may be asked to follow the traditional ritual and use traditional needles but to place them instead in areas of the body which will yield no effect according to meridian theory.

Size of the placebo effect

The placebo effect usually amounts to about 30% of the therapeutic effect of any treatment. It can reach 70% in the treatment of migraines and depression. A review of questions on the efficacy of antidepressants in patients with major depressive disorder showed that the placebo effect accounted for 51% of the therapeutic effect, while the effects of the active substance and of spontaneous remissions only accounted for 25% and 24% respectively (Kirsch and Sapirstein, 1998). The placebo effect may be even more important in preventive actions.

Mechanisms involved

The contribution of the brain’s dopaminergic system to the placebo effect is now well-known (Benedetti et al., 2011). The placebo response consists in the stimulation of the production of endorphins in the body and the generation of an analgesic effect in patients (Levine et al., 1978). As a result, patients experience less pain. But limiting the placebo effect to a simple physiological process does not make sense (Benedetti et al., 2011). It really is a cultural, social, relational, emotional, cognitive and physiological process. Beliefs and education play an amplifying role as well. The medical context and the patient-doctor relationship are also critical for fostering mutual confidence and efficacy expectations. The placebo effect is therefore the result of combined psychobiological and psychosocial mechanisms.

In fact, to understand the mechanisms of the placebo effect, researchers must first eliminate confusion factors (Colloca et al., 2008). These factors are:

– The natural history of the disease (also known as natural history or spontaneous remission),

– Co-interventions – identified or not (e.g., a change in patient diet, whether intentional or not, such as a lack of appetite),

– False positives (misdiagnosis),

– The Hawthorne effect (feeling of being selected for a successful effect),

– Regression toward the mean (statistical phenomenon due to patient selection bias),

– Other methodology flaws, such as the lack of data for the subjects lost in the trials and which may skew the results; intended (sponsoring) or unintended (the evaluator is also the designer of the study) conflicts of interest; the absence of double-blind assessment; the measurement tool learning effect; and, finally, the assessment of efficacy criteria in non-standard conditions.

After ruling these biases out, one may highlight the real psychobiological mechanisms of the placebo effect (Benedetti et al, 2011; Colloca and Benedetti, 2005). Advances in brain imaging have contributed key knowledge to the subject. Research on pain is the most frequent, and placebos appear to be adapted in this domain. According to specialists, the placebo effect works regardless of age, socio-cultural level, personality and lifestyle (Benedetti et al, 2011; Kaptchuk et al., 2008)

The actual placebo effect involves several mechanisms acting in parallel, but not always with the same intensity:

– Expectation: patients prepare their body and mind to anticipate events and address them. When in pain, patients activate the primary sensorimotor cortex, the insula and anterior cingulate cortex.

– Anxiety: a placebo can modulate emotions (Petrovic et al., 2005). The endogenous opioid system is activated. Attention, produced by the prefrontal cortex, plays an important role in allowing – or not – a focus on pain stimuli. The placebo lowers patient anxiety.

– Reward: the dopaminergic mesolimbic system is activated when patients expect the placebo effect to be felt.

In the placebo effect, endogenous opioid and dopaminergic mechanisms combine… 

– Individual learning: for example, patients who have had to cope regularly with pain (like headaches) are more likely to associate a placebo characteristic with the phenomenon of pain reduction. Patients may have been conditioned through regular treatment use (e.g., the color of a pill). Repeated experiential learning stimulates the immune system, the endocrine system and the muscular system.

– Social learning: patients can learn the placebo effect by imitation and observation. A placebo has a positive effect on patients who are informed that they take a one, compared to a control group who does not take one and knows so.

…but the placebo effect is the result of learning and previous experience…

– A relationship based on trust: strong convictions about healing outcomes, the strength of the doctor-patient relationship and the optimism of health professionals are all essential factors which may potentiate a placebo effect. When physician treat patients with conviction and optimism, vs. skepticism, patients become convinced and optimistic themselves. Their brains produce more ‘endomedicine,’ giving rise to the placebo effect. A study compared pain intensity when morphine was delivered by a machine only to when it was delivered and explained by a doctor. Pain decreased more rapidly when the morphine injection took place in the presence of the doctor (Colloca et al., 2004).

…and finally, a doctor-patient relationship based on trust, conviction and optimism.

An exciting research area for both mechanistic and methodology researchers

The placebo effect is a fascinating, supernatural and inexplicable process which has bewildered clinicians, researchers, methodologists and experimentalists until today.

“From this initial part of our experiments, we may conclude that the most inert substances, such as starch, when administered homeopathically, that is to say by acting on the imagination of patients, produce effects which are as strong as those produced by the most powerful homeopathic medicines.”

Armand Trousseau (1834), Hotel Dieu Hospital, Paris

Several mechanisms are involved in the placebo effect, sometimes simultaneously, sometimes successively (Benedetti et al., 2011). Researchers are trying to understand how these mechanisms interact and how to potentiate them.

The advent of meta-analysis provides reconciliations and data compilations with significant additional statistical analysis power. For example, in their systematic review, Hrobjartsson and Gotzsche (2001) questioned the alleged placebo effect of 35% popularized by Beecher in 1955 (see below). To do so, the authors identified 114 randomized controlled trials comparing groups of patients who received a placebo treatment (a pharmacological pill, a manual gesture such as a therapeutic manipulation, or a psychological intervention) with groups of patients receiving no treatment. The analysis included data for 8,525 patients with 40 health disorders (including asthma, anemia, obesity, smoking, epilepsy, Parkinson’s disease, depression, schizophrenia). The authors analyzed the data according to performance criteria which were objective (e.g. blood sugar level measured with a blood test) or subjective (e.g. a questionnaire completed by patients), and binary (e.g. presence or absence of symptoms) or continuous (e.g. body temperature). The authors found no statistically significant difference on continuous objective variables between the groups receiving placebo treatments and those receiving no treatment. On the other hand, they did find significant differences in the continuous measurements of subjective variables. While this study does not question the foundation of the placebo effect, it warns researchers about the difficulty of measuring its amplitude.

A controversial subject

The placebo effect is a controversial subject: it has supporters and deniers both among clinicians, researchers and patients. Skeptics want to understand its mechanisms before using it and highlight the various methodological issues that have led to infamous frauds in the history of medicine (see below). Others, who are more pragmatic, were able to experience it personally as patients or prescribers, and are interested in the fact it does “work”.

A long story

The first documented experimental use of a placebo was in 1800 (Pignarre, 2010). Between 1795 and 1796. The American physician Elisha Perkins invented and patented metal rods he modestly called “Perkins Tractors”. He alleged they were made in an original alloy which endowed them with healing powers. By passing them over aching body parts, the tractors were supposed to relieve various diseases such as rheumatism and headaches. The English physician and epidemiologist John Haygarth repeated Perkins’ experiments on patients with metal rods and wooden sticks. He obtained similar results with both devices. Four patients in five reported they felt much better. Haygarth revealed the deception and described the effect in a book published in 1800 entitled “On the imagination as a Cause & as a Cure of the Disorders of the Body.”

The placebo effect is born with randomized controlled trials

The concept of placebo effect was scientifically confirmed with randomized controlled trials in the 1950s (Sullivan, 1993; Kaptchuk, 1998; Harrington, 2002). The historical turning point was the publication of an article in the Journal of American Medical Association in 1955 by Henry K. Beecher. Based on scientific and medical evidence, the author concluded that the placebo effect, which he defined as the difference between the patient’s condition before and after taking a placebo, occurred, on average, in 35.2% of therapeutic cases. This significant percentage could no longer be ignored, or even negated by practitioners. Ultimately, from that perspective, the explanatory mechanism mattered less than the fact that a placebo treatment could reduce pain, relieve symptoms and contribute to the well-being of patients.

Thus was the placebo effect demonstrated: starting with evaluation criteria for groups treated with placebo in randomized controlled trials, researchers managed to prove that it consists in a change in clinical condition following the administration of a placebo.

What it means for Patients

The placebo effect works with any therapeutic, whether pharmacological or non-pharmacological. It is always puzzling to behold the mysterious influence of mind over body.

What it means for Health Professionals

Strong convictions about healing outcomes, a doctor-patient relationship based on trust and the optimism of health professionals are all critical factors which help potentiate a placebo effect. When physicians treat their patients with conviction and optimism – versus skepticism – patients are more likely to become convinced and optimistic themselves. Their brains produce more ‘endomedicine’ to foster the placebo effect.

What it means for Researchers

Shedding light on the mechanistic underpinnings of the placebo effect and their combination has proven challenging and humbling. The placebo effect offers a fascinating field of research for biologists and human science specialists alike. However, because of its many confusing factors, it requires tremendous theoretical and methodological discipline. The therapeutic potential of placebo has been explored more often than its prevention. To date, pain has been the most widely studied marker.

What it means for Policymakers

A placebo is an effective treatment against pain. It could prove quite useful in prevention programs, but additional research remains necessary.


Beecher HK (1955). The powerful placebo. Journal of American Medical Association, 159, 1602-1606.

Benedetti F, Carlino E, Pollo A (2011). How placebos change the patient’s brain. Neuropsychopharmacology Reviews, 36, 339-354.

Colloca L, Benedetti F (2005). Placebos and painkillers: is mind as real as matter? Nature Review Neuroscience, 6, 545-552.

Colloca L, Benedetti F, Porro CA (2008). Experimental designs and brain mapping approaches for studying the placebo analgesic effect. European Journal of Applied Physiology, 102, 371-380.

Colloca L, Lopiano L, Lanotte M, Benedetti F (2004). Overt versus covert treatment for pain, anxiety, and Parkinson’s disease. Lancet Neurology, 3, 679-684.

Hrobjartsson A, Gotzsche PC (2001). Is the placebo powerless? An analysis of clinical trials comparing placebo with no treatment. New England Journal of Medicine, 344, 1594-1602.

Kaptchuk TJ. (1998). Powerful placebo: the dark side of the randomised controlled trial. Lancet, 351, 1722-1725.

Kirsch I, Sapirstein G (1998). Listening to Prozac but hearing placebo: a meta–analysis of antidepressant medication. Prevention and Treatment, 1, 1-16.

Lemoine P. (1996). Les mystères du placebo. Paris: Odile Jacob.

Levine JD, Gordon NC, Fields HL (1978). The mechanism of placebo analgesia. Lancet, 2, 654-657.

Petrovic P, Dietrich T, Fransson P, Andersson J, Carlsson K, Ingvar M. (2005). Placebo in emotional processing-induced expectations of anxiety relief activate a generalized modulatory network. Neuron, 46, 957-969.

Pignarre P (1995). Les deux médecines: Médicaments, psychotropes et suggestion thérapeutique. Paris: Editions de La Découverte.

Sullivan MD (1993). Placebo controls and epistemic control in orthodox medicine. Journal of Medicine and Philosophy, 18, 213-231.

To reference this Blog en Sante © article.

Ninot G (2014). Definition of placebo. Blog en Sante, L29.

© Copyright 2014 Grégory Ninot. All rights reserved.

Thanks to Syl Billere for the English Revision.