The term placebo effect describes an improvement in the condition of a patient after being given a placebo--an inert substance (e.g. sugar pill) the patient expects may hold some benefit for him. The placebo effect has long been recognized as an unavoidable aspect of medical treatment. Physicians before the 1950s often took advantage of this knowledge by giving patients treatments like bread pills or injections of water with the understanding that patients had a tendency to feel better when they were given something--even if it was inactive--than when they were given nothing at all. In the years following World War II, it became recognized that the placebo effect is more than just a medical curiosity--it is an extremely potent influence on patient psychology and physiology. With this realization came the determination that a condition where participants are given a placebo is a necessary component of an experiment designed to assess the effectiveness of a drug; for, if just the act of receiving treatment makes patients feel better, then that improvement must be subtracted from the overall strength of a drug's action to determine the true efficacy of the substance. This awareness led to the use of placebo conditions in clinical trials of pharmaceutical drugs being commonplace, and to the modern conception of the placebo effect as an important component of drug effects.
While many of us are aware of the use of placebos to test the effectiveness of drugs, we may be less likely to realize that some fraction of the benefit of any drug we take is likely due to the placebo effect. Because we expect the medications we take to help us feel better, they generally do to some extent; this influence is added to the efficacy of the mechanistic action of the drug to produce its overall effect. The magnitude of the contribution of the placebo effect can range from minor to the majority of the drug effect, depending on the medicine in question. Thus, the placebo effect in medicine is something that influences many of us on a daily basis, and all of us at some point or another.
The potency of the placebo effect
The magnitude of the placebo effect is often under-appreciated. Although placebos have no active ingredients, they have been shown to influence both psychology and physiology, and in some cases the effects of a placebo have been found to be stronger than the effects of the medication being compared against it. Placebos can improve quality of life, mitigate the burden of a disability, and--amazingly--have even been associated with decreased mortality. For example, in studies of patients with congestive heart failure, those who adhered to taking placebos regularly were 50% less likely to die than those who were in the placebo group but didn't adhere to taking their "medication." Those who faithfully took placebos were also less likely to experience cardiovascular events like stroke or heart attack.
The effects of placebos on a number of physiological systems have now been well documented. Placebos have been found to influence the activity of the autonomic nervous system, such as heart rate, gastrointestinal activity, and respiration. Placebos can elicit changes in hormone levels across various functional systems; effects have ranged from reducing stress hormone levels based on expectation alone to decreasing levels of appetite-stimulating hormones by convincing participants they had just eaten a very calorie-rich food (even though they hadn't). Researchers have even found that placebos can affect the activity of the immune system. In one study, investigators elicited immunosuppression as a placebo-induced response, and in another study it was found that watching advertisements for the antihistamine drug Claritin led to the drug being more effective than it was in participants who didn't receive any pro-drug messages.
Despite all of the experiments that have documented placebo effects, there is still a great deal to be learned about which neurobiological systems are necessary for creating the placebo effect. It may be that the neurobiology of the placebo effect is different depending on the type of stimulus or the expectancies involved. In other words, it is not clear if there is a group of brain regions and/or pathways that are activated whenever the placebo effect occurs--regardless of the circumstances--or if there are different regions activated depending on the context of the placebo administration. Recent research has used neuroimaging to attempt to unravel the mechanism underlying the placebo effect and, while the effect is complex and still poorly understood, these studies have provided some insight into which areas of the nervous system may be important to mediating it.
Neuroimaging of the placebo effect
Much of the experimental evidence regarding the placebo effect comes from studies of the impact of placebos on pain. This is due in part to the early recognition that the experience of pain is amenable to manipulation by the use of placebos. Pain is also useful to study because it is a ubiquitous problem that has relevance for clinical practice; additionally, we have a fairly good understanding of the components of the nervous system that are involved in pain sensations.
There are several brain regions that receive direct innervation from pathways that carry nociceptive (i.e. pain-related) information from the body to the brain. These include: the thalamus, through which pain signals must pass as they travel to the cortex; the somatosensory cortex, which is the cortical area where sensory signals from the body are initially processed; the insula, which is thought to be involved in mediating the intensity and emotional response to pain; and the anterior cingulate cortex, which is also believed to be involved in emotional responses to pain. Treatment with a placebo has been found to decrease activity in all of the above areas, and several studies have found that larger placebo responses were associated with a greater reduction of activity in these regions.
In addition to affecting these pain "centers" in the brain, placebos have also been found to activate pathways that travel down from the brainstem to the spinal cord to inhibit pain responses. The best known of these pathways runs from an area in the midbrain called the periaqueductal grey, down to the spinal cord. Activation of the periaqueductal grey can be initiated by a variety of cortical areas, and leads to increases in levels of natural painkillers known as endogenous opioids, which act to suppress pain. Endogenous opioids are part of an adaptive mechanism that allows us to tolerate pain, presumably to ensure we can extricate ourselves from an acutely dangerous situation before we become preoccupied with pain sensations.
Placebos can activate these descending pain modulatory pathways involving the periaqueductal grey to cause increases in levels of endogenous opioids. Some studies have found that increased activity in the periaqueductal grey is associated with the degree of placebo analgesia experienced. Additionally, administering a drug called naloxone that blocks the receptors where endogenous opioids normally exert their effect causes a decrease in placebo-induced analgesia. Thus, it seems that activity in the periaqueductal grey is an important component of placebo-induced pain relief.
Placebos also affect activity in higher brain regions like the prefrontal cortex, amygdala, and striatum. Changes in activity in these areas may cause alterations in levels of endogenous opioids and/or may involve changes in affective and anticipatory states, which may influence the perception of pain. Connections between the prefrontal cortex and periaqueductal grey seem to be important for placebo analgesia, as placebos can cause increased activity in areas of the prefrontal cortex; this activity is associated with increased periaqueductal grey stimulation and endogenous opioid release. Placebo treatments also elevate levels of endogenous opioids in the amygdala, and reduce activity there. The role most commonly attributed to the amygdala involves the detection of threats in the environment and the generation of anxiety about those threats, and thus reduced activity in the amygdala may mitigate the anxiety-producing impact of pain. Placebo treatments also cause increases in both dopaminergic and endogenous opioid activity in the striatum. Dopamine activity in the striatum is generally associated with learning, motivation, and emotion; it has been hypothesized that the striatum may be involved in encoding information about the rewarding nature of pain relief and the aversive aspects of pain itself, and thus in the learning and behavior associated with pain avoidance.
Although the placebo effect has been explored most comprehensively in regards to pain, it is not confined to mitigating painful sensations; placebos have been found to affect experiences ranging from emotion to movement in Parkinson's disease. In many cases, the same systems discussed above in the context of pain are thought to be involved. For example, being given a placebo anti-anxiety medication led to decreased activity in the amygdala in response to a series of negative images; participants also rated the images as less unpleasant after taking the placebo. Studies in Parkinson's disease patients have found that taking a placebo that is expected to facilitate movement can cause increased dopamine levels in the striatum, which is associated with improvements in mobility.
Thus, it seems that the brain areas mentioned above may not be specific to the type of placebo effect explored, and may be part of some underlying neural circuitry that mediates the placebo effect in general. However, it is also likely true that we are just scratching the surface with the identification of these common areas. The full neural circuitry of the placebo effect is probably more complex than the collection of regions outlined above, and presumably includes a more intricate neurochemical basis than just endogenous opioids and dopamine. For example, recent research has identified roles for hormones like cholecystokinin and oxytocin in the placebo response as well.
Additionally, it is unclear if brain regions like the prefrontal cortex, which are activated in different types of placebo responses, are activated to serve the same purpose in each context. For example, in a pain context the prefrontal cortex may be involved in activating the periaqueductal grey; in a situation where someone is given a placebo anti-anxiety drug, however, the prefrontal cortex may be involved with regulation of areas like the amygdala. Thus, it is uncertain if this shared neural circuitry is actually working in the same manner in different placebo situations.
Research will therefore continue into the phenomenon of the placebo effect, for more than just the sake of curiosity. For, if we can learn more about the placebo effect and how it is mediated by the brain, we can use that knowledge to better predict which patients might be likely to experience a large placebo effect, and which would not. An ability to predict placebo response in patients could be immensely valuable, and could turn the placebo effect from a quirky aspect of medical care to something that can be directly manipulated in order to improve the effectiveness of treatment. And, while we may not return to the days of giving bread pills without consent, we may be able to better evaluate the efficacy of medications if we are able to better understand the contribution the placebo effect is having.