The Placebo Response ‒ A Powerful Phenomenon
Complete the form below to unlock access to ALL audio articles.
In 1807, in a letter to Dr. Caspar Wistar, Thomas Jefferson penned, “one of the most successful physicians I have ever known has assured me that he used more bread pills, drops of colored water and powders of hickory ashes, than of all other medicines put together.”
Despite early observations of placebo being used in medicine and the benefits documented by patients receiving placebo treatments, the precise mechanisms driving these physiological responses ‒ or placebo effects ‒ have remained elusive.
In a recent interview, Technology Networks had the pleasure of speaking with Professor Ted J. Kaptchuk, leading figure in placebo studies and a scholar of East Asian medicine, about the importance of humanistic dimensions of care, what is currently known about the neurobiology and genetics of placebo effects, open-label placebo studies and ethical use of placebos in clinical practice.
An extraordinary path to alternative medicine
“I was very involved in the antiwar movement in the 1960s, which was impacted by the Civil Rights Movement and Vietnam War. But at some point, I realized I had to earn a living. And I thought, why don't I study this weird Chinese stuff?” says Kaptchuk.
He provides me with some background on his life during this period and explains how a specific event led him to pursue alternative medicine, in particular traditional Chinese medicine, as a career: “There was a grand jury hearing in New York City. I was asked to testify about some of my friends who had gone off the deep end and really became terrible.”
However, he realized he didn’t want to testify, and to avoid doing so, Kaptchuk moved into the San Francisco Red Guard House. The Red Guard Party was an Asian American organization founded in San Francisco's Chinatown in 1969, affiliated with the Black Panther Party, according to Kaptchuk.
“I was there hiding for three months and during that time I read China's national news magazine Peking Review, which is now called Beijing Review. On the back page of every other issue, featured a story that said something along the lines of ‘Chinese medicine is a great treasure’.”
He knew he didn't want to pursue a “conventional” profession and decided to study Asian medicine. Kaptchuk traveled to China in 1971 and studied to become a fully trained Chinese doctor at the Macao Institute of Chinese Medicine.
When he returned to the United States in 1976, Kaptchuk opened a small private practice in Cambridge, Massachusetts. “I practiced for a long time both in a private clinic and also ran a pain clinic in a public hospital for 10 years,” he explains.
It was during this time this Kaptchuk began to notice that some of his patients were getting better very quickly. “They [patients] would walk out of my office with the prescription that I'd written. But they already looked like they were getting better as they walked out, their gait was upbeat and they looked more spirited,” notes Kaptchuk.
This puzzled him, what was going on? “I just held on to that question. I never really pursued it until I was recruited to Harvard Medical School to do research on alternative medicine.”
Making the switch to placebo studies
In preparation for the move to Harvard, Kaptchuk began studying with statisticians and epidemiologists more closely. “I had the opportunity to ask one of the great 20th century statisticians, Fredrick Mosteller, who was one of my great mentors and heroes, a key question – what is my job going to be exactly?”
It turns out that Kaptchuk’s job was to find out whether alternative therapies were more than just placebo effects. But first, he himself needed to understand exactly what the term “placebo effect” meant, so Kaptchuk asked his mentor: “He said, that it [the placebo effect] was the effect of an inert substance ‒ and I thought to myself ‘don’t say anything Ted, but that’s an oxymoron ‒ the effect of a substance that has no effect?’. And so, I began to wonder whether I should study the placebo effect instead of alternative medicine, since no one knows much about it.”
Kaptchuk switched focus and submitted some grant applications on placebo effects to the National Institutes of Health, which were approved, and his work was funded heavily.
“It was at that moment I started my career as a placebo researcher,” says Kaptchuk.
Clearly unsatisfied by his colleague’s definition of the term, I asked Kaptchuk how he would describe the placebo effect:
“I would say that the placebo effect is the positive benefits that patients sometimes receive from being immersed in a clinical interaction. It has to do with rituals, symbols, words, silences and the behaviors that go on there.”
He adds, “I don't want to say what the exact mechanisms are, because I think there's much more work to be done. What I’ve provided is a descriptive explanation, which I think is accurate. Although my colleagues don't like it, because they think they already know how placebo effects work.”
The value of placebo medicine and humanistic dimensions of care
In 2008 Kaptchuk et al. published a study in BMJ exploring whether placebo effects could be unraveled into three components. And whether these components could be recombined to produce an incremental improvement in symptoms, in adult patients with irritable bowel syndrome (IBS) (N=262). The trial participants were randomized to three groups:
- Group one: Assessment and observation only
- Group two: A therapeutic ritual (Assessment, plus placebo treatment with a truncated patient‒practitioner interaction)
- Group three: A supportive patient‒practitioner relationship. ritual (Assessment, plus placebo treatment with an engaged patient‒practitioner interaction)
“Adequate relief” was lowest in group one, higher in group two and highest in group three. The six-week trial demonstrated that placebo could offer patients more therapeutic benefit than just time itself (no treatment), but when placebo was coupled with engaged patient‒practitioner interaction, there was a significant boost in relief.
Mistaking a stick for a snake
The brain can function as a prediction machine, which is something Kaptchuk discussed in detail during his TEDMED talk in 2014. The example he provides “sticks” with me: “You’re in a forest, you’re walking along, and you know that the forest is full of snakes. You see a long slender object ‒ your brain’s visual processing will see that object as a snake, even though it’s a stick.”
Anticipatory mechanisms are critical for human survival, explains Kaptchuk, and while his snake/stick analogy is easy to digest, he shares a more relevant example in the context of placebo: “When a person feels sick and down and they go into an environment that is designed to help them and the people in that environment want to help you and have access to the kinds of technology or treatments that society has to make them feel better, the brain’s processing of sensations, of self-awareness and symptom changes – sometimes in a very positive way.”
Kaptchuk highlights what he describes as an “iconic experiment”, conducted by researchers based in Vancouver and published in Science that demonstrates the point above. The team sought to determine the mechanism underlying placebo response in patients with Parkinson’s disease. The study involved six participants that were all initially on a stable dose of levodopa ‒ a precursor to dopamine used as a dopamine replacement agent. All six were then taken off this medication, resulting in a reduction in the levels of dopamine in their brains. Their symptoms worsened.
“Dopamine plays a part in the misfiring of neurons that are responsible for the shakiness and rigidity in Parkinson’s,” notes Kaptchuk. Using positron emission tomography, the team was able to measure and visualize diminishing dopamine levels in the brain.
The participants were then given either placebo or their original dose of levodopa. The researchers noted that for those given placebo, there was a “substantial release of endogenous dopamine in the striatum.” This increase in dopamine was equivalent to the levels of dopamine that would be observed if they had been given the stable dose of levodopa.
“There’s a veritable pharmacy in each and every one of us,” notes Kaptchuk.
The power of words
Information that is communicated to patients when physicians provide prescriptions is thought to play a part in the way that patients respond to therapies. To investigate this hypothesis, Kaptchuk and colleagues conducted a study that sought to establish whether it’s possible to change how migraine patients (N=66) respond to an active drug (Maxalt (rizatriptan)) or placebo by changing a single word. Their findings were published in Science Translational Medicine.
The team studied 459 migraine attacks. After an initial session in which patients documented their degree of pain and any associated symptoms (30 minutes, and 2.5 hours post migraine onset), they were given six envelopes containing pills to be taken for each of their next six migraine attacks.
Of these envelopes, two were labeled “Maxalt” (the name of a common migraine drug), two envelopes were labeled “placebo”, and two envelopes labeled “Maxalt or placebo”. The pills supplied to participants were indistinguishable.
When patients were given placebo labeled as Maxalt, Kaptchuk explains that it “probably gave them the anticipation of relief” as they believed they were receiving the active drug, and on average individuals reported a 30% reduction in pain in 2.5 hours.
When patients were given Maxalt labeled as placebo, on average patients reported a 38% reduction in pain in 2.5 hours. “When patients received Maxalt labeled as placebo, they were being treated by the medication ‒ but without any positive expectation,” noted Rami Burstein, one of the study’s senior authors, in a press release about the study. “This was an attempt to isolate the pharmaceutical effect of Maxalt from any placebo effect.” Conversely, the inert placebo labeled as Maxalt was an attempt to isolate the impact of the placebo effects from pharmaceutical effect.
When patients were given Maxalt labeled as Maxalt, on average they reported a reduction in pain of 62%.
“Even though Maxalt was superior to the placebo in terms of alleviating pain, we found that under each of the three messages or labels, the placebo effect accounted for at least 50% of the subjects’ overall pain relief,” says Kaptchuk.
Placebos provide relief, but they rarely cure
The positive therapeutic effects associated with placebos are not attributed to altered pathophysiology of diseases, rather they are limited to the symptomatic manifestations of the disease. Kaptchuk explains that while a placebo can’t shrink a tumor, it could help to improve symptom management, for example, cancer-related fatigue.
In a perspective article, published in the New England Journal of Medicine, Kaptchuk and American bioethicist Franklin G. Miller, noted that “while placebos may provide relief, they rarely cure”.
“Placebo effects predominantly occur in illnesses that are described by self-appraisal, what doctors call subjective outcomes, for example, conditions like pain, fatigue, nausea, migraines, irritable bowel syndrome,” Kaptchuk notes.
“There’s not one placebo effect. I always try to specify placebo effects, plural. They're different in different diseases and illnesses. Placebos aren’t only about the imagination and the mind, neurotransmitters are involved. For example, endorphins as well as endogenous opioids, cannabinoids, dopamine and maybe serotonin. We know that neurobiologic mechanisms and brain circuitry are involved in eliciting placebo responses, but obviously, there is more complexity there.”
The mechanism of action for many drugs work via this same circuitry, Kaptchuk adds: “in fact, drugs for symptom management often mimic what the body can already do, not the other way round.”
While we don’t currently have a clear picture of the entire process behind how a patient experiences a placebo effect, Kaptchuk explains that “we do have a good understanding of little pieces of the puzzle.”
He continues, "Certain quantitative, relevant areas of the brain are engaged when people respond to placebo. Neuroimaging can be used to highlight these specific areas.” Studies have shown that regions including the rostral anterior cingulate cortex, the anterior insula, prefrontal cortex and amygdala are engaged during placebo analgesia.
Pharmacogenomics and the placebome
There is emerging data indicating that certain genetic polymorphisms may increase the likelihood of a response to placebo, which could have implications in randomized clinical trial design and therapeutic care more generally, as noted by Kaptchuk and colleagues in Trends in Molecular Medicine.
“Our team was the first team to find genetic markers of placebo, we’re at the forefront of the genetic stuff,” says Kaptchuk.
In 2018, Kathryn T. Hall, Joseph Loscalzo and Kaptchuk published an article in ACS Chemical Neuroscience on the genomic effects on placebo response, which they had coined “the placebome” in an earlier paper published in Trends in Molecular Medicine.
According to Hall, first author of the study, “Placebo studies tend to be multidisciplinary, and few geneticists or computational biologists are interested in studying the effects of inert interventions. Even fewer funding sources are interested in taking the risk to answer the question of who responds to placebo.”
To do robust genetic studies Hall emphasizes that you need a lot of participants: “Genetic effects in behavioral outcomes are often small and combinatorial, that is multiple genetic products interact to influence these effects. So early work in this field has relied on secondary analysis of randomized placebo-controlled trials for the genome-wide association studies and candidate gene analyses of smaller studies.”
The ACS Chemical Neuroscience study sought to investigate whether genetic variation in the molecules mediating signaling pathways – known to play a role in an individual’s response to placebo – could cause variability in response. Hall et al. conducted a comprehensive literature review and developed an algorithm to investigate genes/proteins that may be associated with the placebo response. From this they created a subset of proteins enriched for neurotransmitters, hormones and cytokines that could be mapped to pain, autonomic and immune placebo response pathways. Kaptchuk and colleagues discovered that a number of genes/proteins were known drug targets.
According to the authors, 15 of the drug categories, “including analgesics, appetite suppressants and antidepressives” mapped significantly closely to the placebome. It remains unclear how genetic variation in placebo response pathways might affect clinical studies, but additional research on the placebome appears to support the hypothesis that it can modify treatment response.
Hall discusses this further, “Gene‒drug interactions can mask the true effects – beneficial or harmful – of a drug. To evaluate the outcomes of a clinical trial, we rely on averages. We look at the combined effect in the drug treatment arm and compare it to the same in the placebo arm. If there are large subpopulations that respond to placebo but not drug, and if as we are finding another subset related to the same genetic locus that has the inverse response, i.e., responds to drug but not placebo, when we average the outcomes in each arm, these effects are masked.”
Although according to Hall, this is observed time and time again in the literature, few researchers are interested in examining this further. The reason for this, she believes, is that there is a powerful mindset among clinical trialists that their trial will be the one that is successful, “after all they have a compelling and demonstrated mechanism of action, safety and demonstrated benefits in small Phase 1 or even Phase 2 studies.”
“By the time they get to the Phase 3 or even large Phase 2 stage they are not expecting to fail. Why would they do the trial if they really believed they would lose hundreds of thousands – millions of dollars? Then when the trial is completed, the results unmasked, it’s too late,” Hall adds.
She continues, “Bottomline is, all too often, it’s the placebo that beats them and it’s too late at the end of the trial. Secondary analyses to dissect what happened are frowned upon, and for good reason. Post hoc analyses or hindsight is only useful if you can go back and examine your findings further.
In a separate publication, Hall and colleagues demonstrated that certain polymorphisms in catechol-O-methyltransferase are associated with a higher likelihood to respond to placebo treatment in IBS. She has also conducted secondary analyses of several large-scale randomized trials that included Childhood Asthma Management Program (asthma), Women's Genome Health Study (cardiovascular disease and cancer).
“We have in every case observed gene‒drug/placebo interactions that mask the true effects of drugs. However, we are not able to rerun these trials to look at placebo effects so replication of these findings which is critical to moving them forward is slow to happen.,” Hall concludes.
Shifting peoples’ perception of placebo
With more evidence surfacing around the underlying mechanisms of placebo response, Kaptchuk believes doctors are becoming more comfortable with the idea. However, studies in this area are still marginal in the medical community because not many groups receive funding.
“We live in a pharmacocentric medical world, placebo effects really are still tainted by the notion of deception and trickery. It’s been my job for the last 10 years to do everything I can to change that,” states Kaptchuk. In an effort to alter people’s perceptions, Kaptchuk has conducted a number of “honest” open-label placebo studies, whereby placebo is fully disclosed to the research participants.
"We live in a pharmacocentric medical world, placebo effects really are still tainted by the notion of deception and trickery,"
Earlier on in his career, Kaptchuk investigated the use of placebos in double-blind settings and in laboratory-deceptive situations, but with each published paper he began feeling increasingly unsettled: “I was doing something that was not ethically possible in clinical care, meaning giving placebos by concealment, either in a double-blind or deception-like scenario.”
In 2010, Kaptchuk switched gears. He didn't believe in the conventional wisdom that the placebo effect is a conscious process., whereby patients think they’re receiving an active drug, but in fact they are assigned a placebo, and that’s the reason they begin to feel better.
“My colleague Dr. Anthony Lembo said, ‘this is the nuttiest idea you’ve ever had ‒ but let's do it’,” recalls Kaptchuk. By giving participants honest placebos (informed consent was obtained from each participant) and comparing their data to a no-treatment control, it would be possible to see if an honest placebo works, and the no-treatment controls for regression to the means (a statistical phenomenon that can make natural variation in repeated data appear to be real change) and spontaneous remission.
Investigating the effects of open-label placebo
In a PLOS ONE paper, published December 2010, Kaptchuk and colleagues shared results from a single-center, randomized and controlled three-week study designed to investigate if non-deceptive, non-concealed administration of placebo (open-label placebo) was superior to a no-treatment control, in 80 patients with IBS.
The authors found that “patients given open-label placebo in the context of a supportive patient‒practitioner relationship and a persuasive rationale had clinically meaningful symptom improvement that was significantly better than a no-treatment control group.”
Since this IBS trial, which is believed to be the first randomized controlled trial to compare an open-label placebo with a no-treatment control, Kaptchuk has led at least seven more non-deceptive placebo studies.
“My goal is to harness placebo effects honestly and make them a therapeutic option for many functional illnesses that are defined by self-appraisal, such as chronic pain, irritable bowel syndrome, chronic fatigue and menopausal hot flashes,” says Kaptchuk.
He continues, “I'm not sure if it'll ever be adopted, but patients will try it for sure. For example, those patients suffering from unrelenting discomfort that doctors can't find any reason for, disorders of hypersensitization.”
“We need to conduct more trials, with a greater number or participants, that are longer in duration to generate more data. But for doctors that's a little uncomfortable, they were taught to prescribe medications that have effects beyond placebos,” says Kaptchuk.
Kaptchuk and his team are accumulating evidence that suggests that placebos can be used honestly and effectively in these types of patients, with no side effects.
In 2021, Kaptchuk and colleagues published the results of a larger six-week IBS study (N= 262) in PAIN. "IBS is one of the most common reasons for healthcare consultations and absenteeism from work or school. Effective treatment options for IBS are limited, and we hypothesized it may be possible to ethically harness the placebo effect for clinical benefit,” said Anthony J. Lembo, MD, professor of medicine at Beth Israel Deaconess Medical Center, and corresponding author of the study.
“[…] sixty nine percent of participants who received open-label placebo reporting a clinically meaningful improvement in their symptoms," Lembo noted in a related press release.
In addition, those that received the open-label placebo experienced improvements that were significantly greater than those reported by individuals assigned to the no-treatment control arm.
There was no difference in the degree of symptom improvement between open-label placebo and double-blind placebo arms. “The fact that there was no difference between these two groups is staggering, no one can easily explain that ‒ but what a powerful statement about the potential of healing,” notes Kaptchuk.
Negative expectations and the nocebo effect
Factors that promote beneficial or “positive” placebo effects also have the potential to cause adverse effects, known as nocebo effects. “My colleagues refer to nocebo as the evil twin, but I see it more as the cousin,” Kaptchuk says. “In randomized control trials, whatever side effects you observe in the active drug arm (with some exceptions) you'll get in the placebo arm to a degree. Some events like heart attack you won’t get in the placebo arm, but common symptoms like fatigue, headache, nausea, dizziness, malaise, achiness, etc. can occur in both.”
Kaptchuk notes various reasons why nocebo effects may occur, for example, it could be because of placebo, but it could also be from misattribution. He elaborates, “A person may experience regular headaches, every week for example, they begin taking a new drug, and they get a headache. They aren’t going to think it's one of their usual headaches, they are going to assume it’s from the drug.”
As highlighted by Kaptchuk, individuals could be attributing side effects to drugs that are actually caused by anticipation of negative effects or heightened attentiveness to “normal” day-to-day discomforts: “It's background noise that gets interpreted as placebo, that’s why nocebo effects are very large.”
Another key consideration when it comes to nocebo, according to Kaptchuk, is mechanism. “It's not that the patients expect to get better; we know this from neuroimaging studies done by my team, which have been replicated by other research teams. Instead, when you look at the activation of nocebo effects in the brain and assess the engagement of that phenomena, it includes the hippocampus, which never happens in placebo effects.”
"The hippocampus, among other things, is involved in anxiety, and that anxiety can manifest as headaches, skin rashes etc. But when it comes to nocebo effects in double-blind clinical trials, for any individual in the research study, most nocebo effects and the effects of the active drug feel the same. With a good drug, the group nocebo effects are less than the group effect of the drug. Nocebo and placebo effects feel identical to drugs.”
“My colleagues refer to nocebo as the evil twin, but I see it more as the cousin,”
In January 2022, Kaptchuk and team published findings in JAMA Network Open on the nocebo effect in randomized and placebo-controlled COVID-19 vaccine trials. The researchers compared the incidences of adverse events (AEs) reported in both the placebo and vaccine groups across 12 publications, for a total of 45,380 trial participants (22,578 placebo recipients and 22,802 vaccine recipients).
Systemic AEs were reported by 35% of participants after receiving their first dose of a placebo, with headache (19.6%) and fatigue (16.7%) most commonly reported as minor symptoms. Local AEs (e.g., pain or swelling at the site of administration) were reported in 16.7% of those receiving placebo.
Forty-six percent of participants receiving a vaccine experienced one systemic AE. Following a second dose of placebo, 32% of participants reported a systemic AE and 12% reported a local AE. For those receiving a COVID-19 vaccine, 61% reported a systemic AE following a second dose and 73% reported a local AE.
Based on these numbers, the researchers believe that 76% of systemic AEs after the first dose and 52% of systemic AEs after the second dose were actually nocebo responses.
“For every 100 side effects in the COVID-19 vaccine arm, 76 of the same side effects were seen in the placebo arm, meaning that a lot of those side effects were actually nocebo responses,” says Kaptchuk. He notes that while they were able to identify nocebo responses, they were unable to determine whether these effects were the result of anxiety or due to misattribution.
Placebo effects and the Bayesian brain
“Most of my colleagues in placebo research say that placebo effect is caused by expectation. You expect to get better, it’s a self-fulfilling prophecy, you get better. That is true when you do deceptive experiments on healthy normals. But the evidence for whether expectation creates placebo effects in clinical populations over time is confined to one or two experiments in psychiatry,” he notes.
Kaptchuk explains that the problem with mind‒body concepts is that there's no obvious pathway that goes from the mind to the body and vice versa. While there is mounting evidence that neurotransmitters and genetics play a role in placebo responses, the exact process remains elusive. “How does taking a placebo pill for low back pain provide relief in the back in some people some of the time?” he questions.
One theory that could address this question has to do with the mechanisms that are treated by placebo, which are related to central sensitization, meaning, they don't get rid of hard pathophysiology, they get rid of the amplification of symptoms by the nervous system.
“It’s a very complicated idea and isn’t easy to explain, despite having written several papers on it,” admits Kaptchuk. To help explain the theory, he shares a scenario: “A person experiences back pain for three weeks ‒ a sprained back let’s say. But that injury should heal... except it doesn’t and they continue to experience pain. The patient visits multiple doctors, and they’re told ‘it’s all in your head!’. Being told that there's no biological reason for the pain doesn't make them better, if anything it will make it worse.”
The person’s pain isn’t not real, they are experiencing pain – it’s just that places in their somatic cortex are hypersensitive despite the original physical injury being healed, which according to Kaptchuk, is why doctors are unable to find underlying pathophysiology or “root cause” of the pain.
“A bottom-up symptom perception of back pain, and especially its unhealthy amplification, happens through neural distributive systems involving central sensitization (neural hypersensitivity), and in the case of chronic pain creates perceptions of symptoms, sometimes severe, when there is no disease in the low back.”
Kaptchuk continues, “The amplification the patient is experiencing has a very clear theoretical model, it's called Bayesian brain prediction coding,” says Kaptchuk. “When you give a placebo, you actually turn down that amplification in some people some of the time. The pathway that causes the amplification is the same pathway the placebo uses.”
When a person feels symptom improvement in their back from a sugar pill, the identical Bayesian system is operating but in the opposite direction and sometimes automatically causes the brain to reduce the sensation of pain.
According to Kaptchuk, it’s all to do with symptom application processing. “The Bayesian brain is a very complicated system; it’s related to the way the brain computes ‒ it's not a conscious process.”
For those interested in learning more about how Bayesian models offer a different picture of how the brain perceives symptoms and relief, Kaptchuk highlights two recent publications in PAIN and BMJ.
Realizing the power of placebo
In Kaptchuk’s opinion, the big question is, “Why do we have a prejudice, a bias, against a way of healing that seems to have potential value?”
He provides me with an example to demonstrate his point: “In 1095, Pope Urban II declared that Jewish doctors were not allowed to touch Christian patients or see them. And it wasn’t because they weren't good doctors – in fact his own doctor was secretly a Jewish doctor ‒ it was because at the time they believed it was the work of the devil. And what you have here is a way of healing, but you could say, it's the work of the biomedicine’s devil.”
From our conversation, I believe that it’s Kaptchuk’s multidisciplinary career and experience of alternative medicine that allows him to see the potential of placebo effects beyond that of others, uniquely positioning him as a leader in the field.
He closes our interview by saying,
“I've been a student of placebo for years, and I wrestle with it, but I know there's something there, and my job is to ultimately take it from the margins to a central question of medicine and healing.”
Professor Ted J. Kaptchuk was speaking with Laura Elizabeth Lansdowne, Managing Editor at Technology Networks.