Search: GO!

Placebo Effect

A placebo can be used in two ways. One way is as a therapeutic procedure devoid of all specific activity that is given deliberately to have an effect on a patient, a symptom, a syndrome, or a disease. The second way is as a substance that is inert in relationship to a desired effect, but otherwise indistinguishable in appearance from a substance to be tested.

Placebos were a common part of clinical medicine in the days before the reductionist Cartesian model of modern medicine split the mind from the brain and body and focused on detecting and eliminating physiologically provable pathology. Since around the end of the Second World War and the development of modern drugs, medical and surgical techniques, the placebo effect has languished in disuse except as an important control in randomized double-blind placebo-controlled clinical trials.

Placebos certainly are critical elements in developing evidence-based therapies, however, research into placebos themselves and their effects seldom follow evidence-based procedures. Relatively few controlled trials include a no treatment group, yet often any effect encountered in the placebo group is considered to be due to the placebo. This is actually a bit of a stretch as it is not possible to attribute or quantify effects without the group that received no treatment, because among other factors, symptom variation or better adherence to health-improving behavior may improve the condition without having anything to do with the placebo effect. Measurement and methodology effects as well as treatment effects need to be distinguished from placebo effects. These effects include investigator and subject biases, natural evolution of the disease, measurement reliability, measurement reactivity, regression to the mean, confounded therapeutic procedures, and spontaneous remissions. Adding to the confusion, relationships between the placebo effects and treatment effects may be dynamically interactive.

Nevertheless, placebos have been reported to create beneficial clinical results at rates that vary widely and are often reported to be much higher than the frequently quoted 35% (). Similar to active medications, they seem to have time-effect curves, peak, cumulative, and carryover effects. In a review of clinical studies, Benson and Friedman () have found beneficial results attributed to placebos in 60 to 90% of diseases.

Placebo effects on pain are reported to work in both directions: i.e., analgesia or algesia/hyperalgesia can be induced by the proper instructions. It is said that if someone is told that a placebo is a pain provoking substance, pain may occur or be increased. This negative effect of a placebo is sometimes called a nocebo, and in extreme cases, it has been reported to result in severe pathology ().

A study of double-blind placebo-controlled trials of rheumatology inpatients found that 75% of the placebo arm (control) subjects who experienced either a nocebo effect or significant improvement may believe that they have received that active drug (). If true, the impact on differences between the 2 arms of a study could be profoundly affected.

Researchers have identified three common factors that contribute to the placebo effect ():

positive expectations by the patient;
positive expectations by the health care provider; and
a good trusting relationship between the patient and the health care provider

Does it exist?

One of the most quoted articles is the classic 1955 article by Beecher, “The Powerful Placebo” (1). He claimed that in 15 trials including 1,082 subjects, 35% of the patients adequately found relief by placebo alone. He states that, “It is evident that placebos have a high degree of therapeutic effectiveness in treating subjective responses, decided improvement, interpreted under the unknowns technique as a real therapeutic effect, being produced in 35.2±2.2% of cases.”

Used as a therapeutic intervention, placebo proponents claim that they may pacify the raging hormonal and other chemical onslaughts that are caused by stress and contribute to poor health. They stimulate the release of calming chemicals that improve health. In general, they claim as a therapy they may balance or enhance the immune response by helping to coordinate immunity in a way not to damage our selves, or by stimulating killer functions of immune warriors and their weapons (2). But, what is the evidence?

Is there really a placebo effect? In the mid 1990s, Kienle and Kiene and others began to investigate the evidence (, ). They began by going back to the sources. In 1996, Kienle and Kiene first investigated the published original article source material that creates the scientific basis for claims such as we reported in the opening paragraphs of this article (e.g., Benson and Friedman () have found beneficial results attributed to placebos in 60 to 90% of diseases). They found that the studies on which such ideas are based, except possibly bronchial asthma, “do not in any way justify the conclusions drawn from them” (). They concluded that the literature relating to the “magnitude and frequency of the placebo effect is unfounded and grossly overrated, if not entirely false.” They wonder whether the “so-called placebo effect is itself not largely or indeed totally illusory.” They felt that the placebo effect could be misperceived for several reasons, including:

lack of clarity of the placebo concept itself resulting in misjudgment
insufficient conceptual differentiation resulting in confusion with conditioning and experimental subordination
the fact that sometimes the placebo is not inert at all and that it has a specific action on the condition under investigation
natural history of the disease
regression to the mean
concomitant treatments
obliging reports
severe methodological shortcomings in the studies

To determine if there is any credible evidence that placebos improve patient-reported or observer-reported outcomes, Hrobjartsson and Gotzsche conducted a Cochrane review on RCTs through December 1998 (, ). In the overwhelming majority of the reports, the placebo effect is defined as the difference from baseline in the condition of patients in the placebo group in an RCT after treatment. In these reports, it would be impossible to distinguish the effects of a placebo from the natural course of the disease, regression to the mean and the other factors mentioned in Kienle and Kiene’s reviews. But, for purposes of the meta-analysis the reviewers defined placebo as any intervention labeled as such in the report of a clinical trial. A typical physical placebo was a procedure performed with an instrument that was not functioning. The typical pharmaceutical placebo was a lactose pill. The typical psychological placebo was a non-directional neutral discussion.

The reviewers’ conclusions were:

There is no evidence that placebos in general have clinically important effects.
Although there were moderate effects on subjective continuous outcomes especially on pain, these effects could not be clearly differentiated from bias.
So, there is no justification for the use of placebos outside of the setting of clinical trials.

The question we need to ask is how do we reconcile apparently well-documented placebo effects with the results from the Cochrane review? The answer has been hotly debated. For a more detailed discussion, please refer to Psychoneuroimmunology for the Health Sciences. But, here let us just present one of the viewpoints that diverge from the Cochrane conclusions.

One reviewer, Bailar, while comparing the fall of the placebo effect to Toto pulling the curtain away and revealing that the Wizard of Oz was really a very ordinary man, believes that the sweeping condemnation may be a bit hasty (). He makes the following 6 points:

They did find some evidence for pain.
In spite of the large sample size, they may not have had enough power to detect possible small but clinically effective benefits in subgroups other than pain.
Heterogeneity of results such as they found in binary studies cannot occur unless there is a real although unidentified effect in at least one trial.
The RCTs that were included dealt with severe illnesses. There might be a significant effect in less severe illnesses.
The lack of methodological quality of some trials may have obscured beneficial results.
Research settings may obscure placebo effects. So, how can one study the non-research effects of the placebo without conducting a research study: a mutual contradiction?

Believe it or not

In light of the above discussion, we cannot vouch for the validity of placebo claims in these studies. But, they are intriguing.

Not all placebos are created equal. They need to be tailored to the individual. Nevertheless, some generalities can be made. In the realm of pills: white, angular are better than white round; colored are better than white; gel caps are better than colored pills, and bigger are better. An injection is better than a pill, and an i.v. is better than an injection. A placebo given in a hospital is better than one given at home. Ethnicity also appears to play a role in perception/expectations, which may predict compliance and efficacy (, , , ).

Surgery apparently can also produce substantial placebo effects (). Here are some surgery classics. A Seattle cardiologist subdivided a group of patients experiencing angina pectoris (chest pains) into two groups: one that received the standard operation and the other that had sham surgery. Ninety percent of individuals in both groups reported less chest pains. In a group of 10 orthopedic knee surgery candidates, 5 received the actual operation and 5 a sham operation. Six months later, all 10 felt substantial reduction in pain. The ethics however of sham surgery is a topic for considerable debate (, ).

In addition to differences in types of pain, the method of analgesia delivery, pharmacokinetics and pharmacodynamics, and non-specific activation of endogenous opioid systems can induce variability in response to analgesia (). An RCT reported in 2001, showed that different verbal instructions about certain and uncertain expectations of analgesia result in different placebo analgesic affects which in turn can trigger a dramatic change in behavior leading to significant reductions in opioid intake ().

When antidepressants were compared to active placebos (a placebo that mimics some of the side effects of the antidepressants) little difference was seen (). This suggests that trials with inert placebos may have unblinded the groups and effects of antidepressants may be inflated as a result.

Prevalence of whiplash symptoms following exposure to a placebo rear-end collision has been reported to be 20% in one study (). They found that certain psychological profiles predispose certain people to this occurrence.

Laboratory Studies

Research has progressed in animals in neurobiological models to explain how psychological phenomena interact with the central nervous system, but it is difficult to extrapolate from the cognitive processes of an animal to a human mind. There are relatively few human studies at this level.

There is evidence that cognitive factors and conditioning create and balance in different ways opioid and nonopioid-induced placebo analgesia (). Conditioning activates specific subsystems. That is, if conditioning is performed with opioids, the placebo analgesia is mediated by opioid receptors. If the conditioning is trained by non-opioids, the placebo analgesia is produced via non-opioid pathways. Cognitive expectations trigger endogenous opioids.

A controlled clinical trial published in Science in 2001, demonstrated that the placebo effect in Parkinson’s disease is powerful and is mediated through the activation of the damaged nigrastriatal dopamine system (). Positron emission tomography showed in vivo evidence of substantial release of endogenous dopamine in the striatum of Parkinson’s patients in response to placebo.

Summary

Two recent National Institutes of Health (NIH) meetings highlighted current research into the placebo effect. They focused on aspects of a model in which placebo effects function through psychosocial mechanisms, such as, belief, conditioning, expectancy, and meaning response. These effects would then elicit physiological responses that may modify pathways in neurological, endocrine, immune, gastrointestinal, cardiovascular and/or other organ systems to relieve disease symptoms. A meeting assembled by the John D. and Catherine T. MacArthur Foundation, Network on Mind-Body Interactions, and the NIH focused on the neurobiology of emotions, the biology of social interactions, and neural factors in inflammatory and infectious disease.

As part of the Public Health Service led “Healthy People 2010” initiative, elucidating underlying mechanisms of the placebo effect is the goal of an NIH Request for Applications (RFA). In 2002, this RFA began to fund this research to the tune of several million dollars. Their goal is to integrate research from fields such as sociology, psychology, cell biology, physiology, genetics, and/or molecular biology to probe neurological, endocrinological, immunological and other systems to define mechanisms underlying the placebo effect (RFA-AT-02-002).

Perhaps results from these research efforts will redefine the placebo effect and clarify its existence.

References (Not PubMed Indexed)

Beecher HK. The powerful placebo. JAMA 1955; 159:1602-1606.
Charnetsky CJ, and Brennan FX. Feeling good is good for you; how pleasure can boost your immune system and lengthen your life. Rodale Press. 2001


Our Workshops Workshop Overview Our Staff Our Location About Heather Hills Our Endorsements Free Stress Test Endorsements: "I will always appreciate the support Elizabeth provided during my time of distress." Read More		Heather Hills Evidence Base: Health Consumer PNI Evidence Articles Health Care Professional PNI Evidence Articles Psychoneuroimmunology for the Health Sciences