Placebo Power Hype
The placebo effect is fascinating, but its “power” isn’t all it’s cracked up to be
A placebo is relief from belief: people often feel better simply because they believe they have been treated. More precisely, it is the appearance or illusion of a treatment effect that is not actually attributable to a biological treatment mechanism, which can occur in many ways.
Placebo effects can be downright bizarre, and they can seem almost miraculous in the right circumstances. Human beings will get great placebo from any treatment that seems impressive in any way: more invasive, novel, unpleasant… anything that suggests potency. And if we don’t think we’re getting strong medicine, then even strong medicine won’t work as well.1 Expectations account for most treatment success stories from both patients and professionals. It’s all legitimately interesting biology and psychology — and you can read an overview of placebo like this on almost every health blog on the Internet. Now for the bucket of cold water available only here (and a handful of other places):
Placebo gets more love than it deserves.
Placebo is not “powerful,” as famously claimed by Dr. Henry K. Beecher in 1955.2 Placebo is not a magical mind-over-matter phenomenon, or even a good consolation prize when treatment is otherwise ineffective. Many medical problems are entirely immune to positive thinking and expectations (try treating tuberculosis with a sugar pill and see how many Nobel Prizes you pick up). The power of belief is strictly limited and accounts for only some of what we think of as “the” placebo effect.3 There are no mentally-mediated healing miracles. But there is an awful lot of ideologically motivated hype about placebo …
Putting placebo on a pedestal
There is an annoying trend in alternative medicine: the aggrandization of “the power of placebo” as a justification for therapy that can’t beat a placebo. Therapies that perform no better than placebo are now predictably spun as being “as good as placebo,” as though placebo is the new gold standard to meet. Acupuncture is the most glaring example of this,4 but we’ve seen a lot of it with homeopathy as well.
The absurdity of this inflation of the placebo currency becomes clear with a simple comparison: would you take a drug that had been proven to be “as good as a placebo”? Not a chance! We expect real medicine to be significantly more effective than fake medicine. If Big Pharma tried to get away with touting a drug that way, they’d be … well, I guess they’d be even more hated and criticized than they are already are.
Placebo has been hijacked and re-branded for its public relations value to alternative medicine. If your treatment isn’t evidence-based, no worries: you can still sell the power of placebo!
We have a word for medical treatments that only work if you believe that they will, and it rhymes with “gazebo.”
Book Review, Unlearn Your Pain [Schubiner], by Scott Alexander
And it’s not just alternative medicine that hypes placebo
Placebo often gets hyped as a defensive reaction to negative trial results in general, by any kind of health care professional wondering if evidence-based medicine might just put them out of business.
For example, in 2001, the infamous Philadelphia Panel on rehabilitation interventions showed that many popular treatments cannot beat a placebo,5 an evidence-based bummer of epic proportions. Physical therapist Carol Davis found this “shocking in many ways” and wrote a letter to the editors of Physical Therapy.6 Like so many others, before and since, her reaction was to blunt the bad science news by putting placebo on a pedestal and moving the goalposts of science, suggesting that controlled trials aren’t so great. The editors’ reply was perfect:
If Dr. Davis believes that our future as a profession lies in our ability to produce placebo effects, perhaps she misses the point. Her view taken to its logical conclusion would not mean that we could, as she said, “reduce the musculoskeletal curriculum by two thirds,” but rather that we could possibly eliminate this aspect of the curriculum in its entirety as we become not physical therapists but rather practitioners of “placebo enhancement.” As a curriculum coordinator, Dr Davis should know that this role is not what sets us apart from other practitioners and is not seen as our raison d’être in any practice act or in any document that describes our practice. I believe Dr Davis’ views to be unwise and reckless and, most importantly, potentially injurious to those patients who expect us to have some basis in science for our practice.
That was a strong condemnation. It’s a bad idea — “Unwise and reckless … potentially injurious … ”! — to try to move away from controlled trials as the gold standard for figuring out what works. Let’s go where the evidence takes us. Let’s abandon what doesn’t work better than placebo, and find out what does.
In any case, it’s clear that treatments are still ineffective in spite of very high patient satisfaction with physical therapy.7 There’s no question that having faith in the therapy and liking your therapist can drive some placebo, but it’s clearly not enough.
This video presents a perspective on placebo that doesn’t worship the “power” of placebo, but looks instead at the paradox of placebo: if placebo is so powerful, why isn’t everyone cured? Because the placebo effect is basically a dysfunctional mistake.8 How refreshing!
Not one mind-over-matter effect, but miscellaneous illusions
It’s always called “the” placebo effect, but placebo is not just one thing. The term is almost hopelessly imprecise. It labels a whole category of phenomena, many of them clinically meaningless.9 We truly need to learn and use more precise terminology. It’s not just word-nerd nitpickery! Speaking only of “placebo” is about as useful as saying only “entertainment” when you might mean casino or a broadway show or Netflix.
There are many ways for the appearance or illusion of a treatment effect to occur, “a wide variety of mechanistically different and likely experimentally distinct factors, including but not limited to bias, natural history, analysis errors, and regression to the mean.”10 Some of those illusions are biologically interesting, but many are not. For instance, a placebo effect may occur due to illusions and distortions in data collection and reporting — almost embarrassingly trivial and boring factors, like regression to the mean or comparing endpoints to baseline instead of to inert treatment.11 Paperwork placebo!
Another example: being better is not equivalent with actually feeling better, and feeling better is not the same as saying you feel better. People can and do say things that don’t accurately represent their internal experience. As Dr. House would say, “Everybody lies.” A classic example is exaggeration of benefit to please authority figures, like doctors. Placation placebo!
These complexities of human psychology and behaviour are why researchers often look at placebo effects as an annoyance: they are irritating distractions in research, and difficult to eradicate. There are myriad ways in which both treatment and placebo effects may not be what they seem. A placebo might be an interesting mind-powered effect on biology … or it might just be a statistical mirage! And that is what the evidence shows: that placebos do not have important clinical effects in general.12 Here’s a good example …
You’ve got the
power research artifact!
A 2011 paper about massage for low back pain contained a profound example of just how boring and trivial placebo effects can be. That’s right: boring and trivial. Not “powerful.”
Some patients got free massage therapy and their results were compared to patients who got no free massage. The massaged patients seemed to have significantly less pain after 10 weeks. But there was a serious flaw: the deprived patients knew they weren’t getting free massage. And that was almost certainly a disappointment to them, what is cutely referred to as a “frustrebo” effect.13 And that probably slanted the results.
The authors acknowledged that this flaw may actually have made massage “seem more superior than it really is” in comparison to the artificially negative results of the unmassaged patients. Scientific data that creates a false impression of reality is called a research artifact and this is a whopper, nasty and misleading. Lesser glitches have been the downfall of much bigger science. This kind of thing is exactly “Why Most Published Research Findings Are False.”14
What we ended up with here was an appearance of a treatment effect in the difference between what may be genuinely therapeutic results on one side of the gap, and completely meaningless bias on the other. This is a perfect example of a placebo effect that is partly due not to the power of the mind, but to the power of an unfair comparison.
Just because a published paper presents a statistically significant result does not mean it necessarily has a biologically meaningful effect.
Science Left Behind: Feel-Good Fallacies and the Rise of the Anti-Scientific Left, Alex Berezow & Hank Campbell
It is possible that every stitch of apparent massage benefit in the Cherkin study was attributable to real changes in the physiology of those patients, but that is unlikely. It was probably a complicated mix of real benefit and research artifacts.
At least some of the benefits were due to the limited power of mind over body, stimulated by the inherently pleasing nature of massage as an experience and resulting in genuinely improved symptoms (subjective experience) — and we should certainly not throw that baby out with the bathwater. And some of the apparent benefits were probably due to simply being happier and more confident (reporting artifact). And then some more of the apparent benefits were almost certainly due to the fact that not being massaged just looked rather bad by comparison, causing patients to actually feel a little worse (subjective experience) and report their experience more negatively (reporting artifact).
The appearance of a positive effect of massage here is greatly watered down by the unfair comparison, and definitely not just the “power” of mind over matter. The point is that it would be absurd to conclude from this data that massage “works” by the power of placebo.
Evidence that placebo isn't all that powerful
When placebo is tested directly, it doesn't do all that well. In 2021, Strijkers et al reviewed the magnitude of effect that placebo had in trials that compared placebo, treatment, and no treatment for chronic unexplained back pain.15 They found 21 such trials, pooled all their data, crunched the numbers… and found that the placebos did more than doing nothing. But how much more? Not that much! "Probably not clinically relevant." Less than one point on a 10-point pain scale is barely noticeable.
And some of that difference can almost certainly be attributed to research artifacts… not the the mind asserting itself over pain.
Such results are generally typical of the literature on placebo power.
So what is a myth in medicine? It is not everything that is false, but rather concepts that are based on something akin to a story. It is a good story that meniscal tears cause pain and that when they are removed, the pain goes away as well. Showing that pain relief is mainly due to placebo effects means destroying a good story, which might be as unpopular as depriving some colleagues of a good business.
Mythbusting in Orthopedics challenges our desire for meaning, by Per Aspenberg, 547
Placebo actually still involves deception!
One of the greatest placebo hype examples to date was a 2010 study which purportedly proved that the power of mind over body is so stupendous that a placebo — a fake medicine, a sugar pill — works even when you know it’s a placebo,16 a so-called “open-label” placebo. This study is partly responsible for the notoreity of the lead author, Ted Kaptchuk. The results have been uncritically reported and passed along by pretty much everyone ever since, including a many people who should probably know better. ScienceBasedMedicine.org published the only critical analysis I’m aware of, by Dr. David Gorski, and he nails it:
Even though they did tell their subjects that the sugar pills they were being given were inert, the investigators also used suggestion to convince their subjects that these pills could nonetheless induce powerful “mind-body” effects. In other words, the investigators did the very thing they claimed they weren’t doing; they deceived their subjects to induce placebo effects.
In other words, can you “impress” patients into feeling better? Hey, maybe! Is it a non-deceptive placebo if you do it by saying “this fake medicine is really awesome”? No. That’s just a regular placebo — very old school. There’s really nothing new here. It’s not placebo “without deception,” but placebo with deception about the value of placebo itself.
As I predicted would happen,17 Kaptchuk continued to tell this story with another study in 2016, this time concerning back pain.18 Dr. Gorski once again did the Yeoman’s work of deconstructing the placebo propaganda, with a new article, “The revenge of the son of the myth of ‘placebos without deception’,” in which he makes it clear that this new study is even more ridiculous than the last one:
Right off the bat the investigators recruited patients who were interested in “mind-body” effects and then told them that placebo effects could produce powerful “mind-body self-healing processes” in rigorous clinical testing. There was the deception, because even the most generous and sympathetic characterization of the rigorous research existing on placebo effects would not justify such a description. As I said at the time, not only did Kaptchuk et al deceive their subjects to trigger placebo effects, whether they realized or would admit that that’s what they did or not, but they might very well have specifically attracted patients more prone to believing in the power of “mind-body” interactions. Yes, patients were informed that they were receiving a placebo, but that knowledge was tainted by what the investigators told them about what the placebo pills could do.
Ted Kaptchuk is transparently aggrandizing placebo, expanding its definition in a way that would (just coincidentally!) make it seem more magical and awesome — thus reinforcing his own story. “This fake medicine is really awesome as proved by studies like mine showing that this fake medicine is really awesome as proved by studies like mine showing … ”
This is essentially propaganda research, producing results to be used to try to shore up alternative medicine’s credibility as it gets backed further into a scientific corner. As study after study shows that treatments like homeopathy and acupuncture are no better than a placebo — a slam dunk fail — many beleaguered proponents have stopped trying to deny it and have pivoted to spinning it as good news:
[They] start arguing that they “work” by “harnessing the power of placebo” effects to “induce natural healing,” misrepresenting placebo effects as the power to use one’s mind to heal oneself. Never mind that this argument never flies with conventional science-based medicine and represents a flagrant double standard in which CAM is held to a much lower standard of evidence. Again, this is about belief, not evidence. Never mind that lying to patients to invoke placebo effects is the resurrection of medical paternalism.
There’s now evidence supporting the rebuttals: a 2017 study nicely demonstrated that open-label placebos only work if you give the patient some reason to have faith in them. If you don’t pump them up with talk about the potency of placebo, they fail as expected.19 “Placebos with a plausible rationale are more effective than without a rationale,” the authors concluded.
Honest placebo: now longer lasting?
In 2021, Carvalho et al, including Ted “Placebos Are Powerful” Kaptchuk, published a paper with the rather extraordinary punchline that honest placebo did not just help back pain patients … but continued to do so for five years.20 Despite a glaring limitation, they declared that their data “suggest that reductions in pain and disability after open-label placebo may be long lasting.”
Studies like this often stand alone, never replicated nor contradicted by other research, or not for a long time. But we're in luck! Just a year later, in 2022, Kleine-Borgmann published their own long-term follow-up on a very similar experiment … finding no lasting effect.21
I think the only thing here that’s long lasting here is the hype. 😜
Is the placebo effect getting stronger? The Tuttle kerfuffle
In 2015 Tuttle et al. reported on a bizarre thing: placebo response in trials of pain-killers seems to be growing.22 This finding got a lot of attention, and added to placebo hype and mystique.
But let’s be very clear that this study raised more questions than it answered, and probably the only implication you can take to the bank is that “it is yet more evidence that placebo effects are complicated and are largely due to artifacts in the way clinical trials are designed and executed” (Dr. Steven Novella). And Dr. David Colquhoun seems to think the whole business is a tempest in a teapot (and he’s as authoritative on the subject of placebo as anyone has ever been):
I find the arguments in this piece quite baffling. If a treatment does not perform better than a dummy placebo, then it doesn’t work. The trial is not thwarted by the placebo response: it has answered the question that was asked, even if the answer is a disappointing one. It is also not right to refer to responses that are observed in people who are given a dummy pill as a placebo response. The changes seen in the dummy group are a combination of placebo response and regression to the mean (roughly. the get-better-anyway effect). There is an increasing body of evidence that the latter is more important than the former. Placebo effects, though real, are generally too small in size to be of noticeable benefit to patients.
Placebo is interesting, but not nearly as interesting as most people seem to think these days, and as a general rule they are nowhere near powerful or consistent or “real” enough to justify deliberate prescription (lying to patients), or as a reason to continue using treatments that are otherwise proven to be ineffective or barely effective. Placebo, in short, is not therapy. Kaptchuck et al. — and many others — exaggerated both the certainty and the therapeutic potential of placebo to his study subjects, so it’s hardly surprising they responded more strongly. Placebo effects are actually quite erratic, small, transient, and conspicuously absent entirely for many pathologies. Try telling patients that about their sham medication and see how well it works.
Justified, rational reassurance is still the only ethical way to persuade/condition a patient into experiencing some interesting mind-over-matter effects — which is practically the mission statement for this website. I am devoted to explaining common pain problems, not fluffing up the value of unproven therapies. Dr. Harriet Hall in Benedetti on Placebos (an excellent read):
Prescribing placebos is uniformly rejected by medical ethicists: instead, we can put our increasing knowledge of placebo neurophysiology to good use without lying to patients.
And Drs. Silvernail and Ingram, physical therapists (Ingram et al):
We need to work harder to understand and communicate the basic science of our profession. This focus should naturally cause us to consider prior plausibility more strongly when deciding which interventions and concepts to include in our therapeutic process.… We feel strongly that our patients deserve scientifically defensible care that is more than just artfully delivered placebo.
Backfirin’ placebos! How the placebo effect can actually make back pain worse
Placebo is not only not as potent as popularly imagined, but it can also backfire quite badly.
Reminder: even researchers with a clear pro-acupuncture agenda have prominently conceded that acupuncture is no better than a placebo, but then go right ahead and daftly and paradoxically recommended it anyway … you know, for the sake of a good placebo effect.23 Not only did Berman et al recommend it, they advised doctors to send patients to a “properly trained” acupuncturist.
Properly trained how, exactly? In placebo delivery? At the point of a needle. We live in an age of such vigorously defended patient rights and robust anti-paternalism that it’s ethically verboten for doctors to prescribe so much as a sugar pill. And that’s (mostly) a good thing. But these pro-acupuncture doctors think it’s okay to send you to a “properly trained” acupuncturist for $1000 worth of placebo-inducing ritualistic needling?24
What could possibly go wrong?
It’s not just about the pointless expense.25
When I was a massage therapist, I routinely saw significant harm done by acupuncture and other ineffective therapies. Far from enjoying a robust mind-over-matter placebo effect, most patients seemed to believe all the more in their back pain as an unassailable affliction that “even acupuncture” couldn’t help.
More tragic than simply wasting time and money on a treatment that doesn’t work is that so many patients conclude not that the treatment was ineffective but that acupuncture was defeated … defeated by an unusually serious case of back pain. Patients are strongly predisposed to anxious assumptions that their problem is “really bad,” and the failure of acupuncture confirms it. The acupuncturist is given the benefit of the doubt, while their back pain is elevated to the status of a fiercer enemy. A nice trap.
How’s that for a “placebo”? In fact, that’s a kind of nocebo, the anti-placebo, where perception and belief result in a negative effect instead of a positive one: instead of relief from belief, it’s grief from belief. Everything placebo can do, nocebo can do … in the other direction. But you never hear freelance alt-med therapists talking up the “power of nocebo” or worrying about giving their patient anxiety-producing notions about what might or might not be wrong with them. They should be, but they don’t.
The scientific evidence is overwhelming that emotional and psychological factors are of real importance in low back pain (and many other kinds of chronic pain). The pain is not “all in your head,” but it is affected by what’s in your head. The despair that sets in when a minor placebo effect wears off is really problematic, significantly exacerbating people’s fear that they are “screwed.” Thanks, acupuncture! Thanks a bunch.
Is it okay to pay for a placebo?
It seems to me that placebo treatments ought to be paid with a placebo payment.
Even when people begin to accept that a treatment is probably bunk — homeopathy is a good example — they often still protest that they are happy to pay for a placebo. As long as it works, who cares how? And placebo can work! So why not? This is basically the medicinal version of Pascal’s wager (erring on the side of faith). If the stuff actually works, great. If not, at least there’s a little placebo consolation prize.
I have no problem with people paying for a placebo as long as their eyes are wide open. But … the wider your eyes get, the less likely you are to benefit from a placebo (despite the “open label” hype, see above). And there are very strict limits to what placebo can do even at its best … and paying for things is never completely harmless, because that money could be used for other things … and trying to use a placebo gets a lot more problematic when something really medically serious comes up.
Where I have the least objection to “paying for a placebo” is for treatments in the grey zone: unknown efficacy but some plausibility and low risk. I’ve tried many such treatments, knowing full well that any effect I enjoy is probably just placebo (or regression to the mean, or natural recovery)… but it might be an actual effect, and I’m willing to pay a little for that chance. I’m gambling on getting a genuine benefit, with a bit of placebo as a consolation prize. So, for me, the plausibility has to be there.
It ain’t there with homeopathy. And lots of other things where people are banking on getting at least a placebo.
Placebo minimization: what if we sell it with a wink?
Many professionals are in the middle of the process of accepting that treatment X is mostly or entirely a placebo delivery mechanism. Sometimes that process is painful. They know (or almost know) that it’s just “therapy theatre” and they know (or almost know) that it’s unethical to sell a confirmed placebo, but they aren’t quite ready to let go of it yet. These tormented souls sometimes come to me looking for an endorsement of a salvage strategy: What if it’s undersold? What if it’s not presented as a “cure”? What if we just tell patients it’s a minor adjunct to the real therapy, like exercise? Would it be kosher then? Would it ethical?
Sometimes they ask, and sometimes they just skip right to angrily insisting that it is okay.
Placebo minimization is a slippery slope to a silly place. What if we wink when we offer it? What if we cross our fingers behinds our back? Slide any further and what’s the point? Here’s where that ends: “This doesn’t work but we’re going to do it anyway, because it’s just what we do here.”
Trying to make ineffective treatments acceptable by reframing them as unimportant is a bad compromise: it’s still on shaky ethical ground, and yet it undermines the very placebo effect that is supposedly worth exploiting. It’s expectation control when the entire point of the treatment is to generate expectations. It’s still therapy theatre, but now it’s a much less impressive show. It’s kind of like saying that faith healing is cool as long as it’s in a smaller tent with a smaller, quieter crowd.
But actions speak louder than words, and patients tend to assume that there’s a good reason for doing something regardless of what is said about it. So it’s actually difficult to undermine the placebo verbally unless you go all the way, and the weak attempt to do so is mostly just a bit of half-assed ass covering.
Sensation-enhanced placebo: the basis for most therapies for pain
Classic placebo is “relief from belief,” and the belief can be (greatly) enhanced by novel sensations, and this is the basis for many popular treatments, or even entire professions.
There are many ways to enhance placebo, and one way is to impress the patient by making the treatment seem more impressive. Surgery is the most dramatic example of this — indeed, it is the ultimate placebo26 — but there are many other fairly obvious examples, like the high-tech razzle dazzle of high-intensity ultrasound, the laboratory sophistication of platelet-rich plasma injections, or the exotic cultural richness of acupuncture.
Sensation enhancement is sneakier.
Every placebo involves some degree of fundamentally misleading exaggeration to make the patient believe that it’s better than it actually is. There are many ways to achieve this: risks, costs, size, intensity, technology and much more can all impress the patient with the seriousness and legitimacy of the treatment. But what could possibly be better than for the patient to actually feel the “power” of the treatment? Even painfully?
Treatments that produce unusual/intense sensations are a particularly good way to impress patients. If you take it far enough, it seems less like a placebo and more like something important must be happening to the body above and beyond “just” sensation. The stronger and stranger the sensation, the easier it is to convince the patient that something therapeutic must be happening. Sensation is extremely persuasive, especially relieving sensations.
The rabbit hole goes even deeper, because pain is a form of sensation, and it can actually be changed by interactions with other sensations. But if it boils down to convincing the patient’s brain to dial down the pain, it’s still a placebo at heart — just a rather clever and fancy one.
Here are some examples of treatments that have a large component of sensation-enhanced placebo (in many cases, that’s probably all they are):
- joint-popping in chiropractic spinal adjustments
- “good pain” in massage therapy while digging at “muscle knots.”
- tingling, vibrating sensations produced by transcutaneous electrical nerve stimulation (TENS)
- warming with ultrasound (which adds that technological element)
- ointments and balms like Tiger Balm15 or RUB A535 and Deep Heat contain a chemical irritant or “rubefacient” that feels hot basically because it gives you a mild chemical burn
- “provocation therapy” like Graston Technique (intense scraping of tissue with tools)
- elastic therapy taping, like we saw a lot of at the 2012 summer Olympics
- acupuncture, which is probably one of the most impressive examples of “therapy theatre,” and relies heavily on the weird sensations caused by needling (the human nervous system really sits up and takes notice when it’s treated like a pin cushion, unsurprisingly)
Are animals immune to placebo? Of course not!
Homeopaths and acupuncturists often insist that their potions and punctures cannot be just causing a placebo because they (supposedly) work on animals, and animals are “immune to placebo,” ergo homeopathy and acupuncture must work.
The argument is based on yet more misunderstandings about the nature of placebo, and it is also a little bit insulting to the intelligence of everyone concerned … including the dogs, cats and horses (who need real care and medicine, by the way).
The first problem is with the reporting. Who says that acupuncture and homeopathy cures animals? The people who really want it to be true — they are the pretty much the only ones claiming this, and they have a huge bias.
There is lots of room for bias and placebo in the human observer of the animal. (Remember, placebo includes all non-specific effects and artifacts of investigation, not just mind-over-symptom effects.) People know almost nothing about the animal’s experience, and nothing at all that doesn’t come through the filter of (intense) human interpretation. Anecdotes about animal healing, if anything, are more likely to be warped by wishful thinking than personal reports, not less.
But … who says animals can’t be directly influenced by human hope and expectations and feelings? Who says they aren’t spectacularly sensitive to human non-verbal cues? The horse Clever Hans was so sensitive to non-verbal cues that he could get the right answer to any simple math question simply by paying attention to the reactions of humans to his hoof taps.27 Clever indeed! Or have you ever watched The Dog Whisperer? It’s absolutely amazing how dogs are like mirrors for our souls. I have no doubt at all that a dog might very well enjoy an “expectation effect” from a hopeful human taking charge and confidently giving a remedy. This is not a stretch.
This kind of thing is why comparative psychologists mostly test animals in isolation, without interacting with them. If they aren’t careful, animals will modify their behaviour in response to human expectations.
Here’s a good rule for life: give animals some credit! And give humans less.
Placebo enthusiasm cannot be curbed
Enthusiasm for the power and mystique of placebo is like the mole in Whack-A-Mole: curb it over here, and it just pops up somewhere else! I get pushback whenever I share anything on social media about how placebo isn’t actually all that exotic or potent. Someone invariably responds with their favourite “is so powerful!” example … none of which, so far, have persuaded me to renounce my skepticism and start worshipping the power of placebo.
Trying to debunk the power of placebo is like trying to debunk ice cream. People are determined to be in awe of placebo. And I get it: we all badly want a bit of magic to be left in this world (please), and placebo seems like a decent candidate. So people have latched onto it, drawn to something a wee bit magical — but unfortunately that’s mostly just based on some misleading old hype. It’s essentially a form of superstition, a manifestation of the belief that things are stranger than they actually are.
And it’s always easy to avoid messy, boring truths.
If there’s a way to convince people placebo is not all it’s cracked up to be, it’s education, of course. Superstition is just the misplaced awe of the ignorant. We must learn enough to be amazed by the genuine marvels of science. We must be nerds! We must be fascinated by that which requires lots of hard work to understand, and would otherwise be opaque and boring.
Did you find this article useful? Interesting? Maybe notice how there’s not much content like this on the internet? That’s because it’s crazy hard to make it pay. Please support (very) independent science journalism with a donation. See the donation page for more information & options.
About Paul Ingraham
I am a science writer in Vancouver, Canada. I was a Registered Massage Therapist for a decade and the assistant editor of ScienceBasedMedicine.org for several years. I’ve had many injuries as a runner and ultimate player, and I’ve been a chronic pain patient myself since 2015. Full bio. See you on Facebook or Twitter., or subscribe:
What’s new in this article?
Ten updates have been logged for this article since publication (2013). All PainScience.com updates are logged to show a long term commitment to quality, accuracy, and currency. more
When’s the last time you read a blog post and found a list of many changes made to that page since publication? Like good footnotes, this sets PainScience.com apart from other health websites and blogs. Although footnotes are more useful, the update logs are important. They are “fine print,” but more meaningful than most of the comments that most Internet pages waste pixels on.
I log any change to articles that might be of interest to a keen reader. Complete update logging of all noteworthy improvements to all articles started in 2016. Prior to that, I only logged major updates for the most popular and controversial articles.
See the What’s New? page for updates to all recent site updates.
May 19, 2023 — Added a new conclusion: “Placebo enthusiasm cannot be curbed.”
2022 — Added a small section, “Honest placebo: now longer lasting?”
2022 — Added some classic citations, most notably to Beecher and Hróbjartsson.
2017 — Science update: cited Locher et al, a major blow to claims that open-label (non-deceptive) placebos work.
2017 — New section: “Are animals are immune to placebo? Of course not.”
2016 — New section: “Sensation-enhanced placebo: the basis for most therapies for pain.”
2016 — New section: “Placebo minimization: what if we sell it with a wink?”
2016 — Science update: coverage of another round of nonsense about “placebos without deception.”
2016 — New section: “Is it okay to pay for a placebo?”
2016 — New section: “Is the placebo effect getting stronger? The Tuttle kerfuffle”
2013 — Publication.
- Amanzio M, Pollo A, Maggi G, Benedetti F. Response variability to analgesics: a role for non-specific activation of endogenous opioids. Pain. 2001 Feb;90(3):205–15. PubMed 11207392 ❐
This study tested the effects of pain-killing drugs with and without the patients knowing they were being medicated. They also blocked neurotransmitters to stop the brain from doing any of its own pain-killing, which we know can happen when people believe they are being medicated (placebo effect). The results showed that the effects of drugs are strongly affected by awareness. “We found that the hidden injections [of pain killer] were significantly less effective and less variable compared with open injections.”
On the one hand, this seems profound and seems to show the “power” of placebo and the mind. On the other, it’s merely a good demonstration of what we already know: that meaning and expectation are an important factors in every therapeutic equation.
- Beecher HK. The powerful placebo. J Am Med Assoc. 1955 Dec;159(17):1602–6. PubMed 13271123 ❐
This is the 1955 paper that launched the Legend of Placebo. Dr. Henry Beecher reported that 35% of 1000 patients were “satisfactorily” treated with a placebo alone. His conclusion catapulted placebo to lasting fame… and wasn't really questioned for a long time.
Kienle and Kiene published a strong criticism of Beecher's findings in 1996, but no one took much notice. Much more prominently, Hróbjartsson and Gøtzsche reported in 2001 in the New England Journal of Medicine that they “found little evidence in general that placebos had powerful clinical effects” and concluded that there is “no justification for the use of placebos” outside clinical trials.
The topic has been hotly debated ever since, but few experts still believe that placebo is “powerful.”
For much more about Beecher and how this all started, see The Legend of the Wartime Placebo
- Hróbjartsson A, Gøtzsche PC. Is the placebo powerless? Update of a systematic review with 52 new randomized trials comparing placebo with no treatment. J Intern Med. 2004 Aug;256(2):91–100. PubMed 15257721 ❐ “We found no evidence of a generally large effect of placebo interventions. A possible small effect on patient-reported continuous outcomes, especially pain, could not be clearly distinguished from bias.”
- There are several prominent cases of this, but the most infamous example is a 2010 paper in the New England Journal of Medicine (see Berman et al). This and other examples are discussed in the introduction of my review of acupuncture for pain.
- Philadelphia Panel. Philadelphia Panel evidence-based clinical practice guidelines on selected rehabilitation interventions: overview and methodology. Phys Ther. 2001;81(10):1629–1640.
“Philadelphia Panel evidence-based clinical practice guidelines on selected rehabilitation interventions for low back pain”
“Philadelphia Panel evidence-based clinical practice guidelines on selected rehabilitation interventions for knee pain”
“Philadelphia Panel evidence-based clinical practice guidelines on selected rehabilitation interventions for neck pain”
“Philadelphia Panel evidence-based clinical practice guidelines on selected rehabilitation interventions for shoulder pain”
- Davis CM. More questions than answers. Phys Ther. 2002 Mar;82(3):289–290. PainSci Bibliography 56054 ❐
- Hush JM, Cameron K, Mackey M. Patient Satisfaction With Musculoskeletal Physical Therapy Care: A Systematic Review. Phys Ther. 2010 Nov. PubMed 21071504 ❐
Quite a bit of research has been done on physiotherapy and its efficacy, most of it quite discouraging. But what about the patients? Do they like it? Does it work well for them? This study attempted to calculate what the patient satisfaction level was for those who received physical therapy care.
A review of the literature was undertaken from several databases. A search of 3,790 studies allowed for a thorough study of 15 that met the criteria.
The researchers concluded that “patients are highly satisfied with musculoskeletal physical therapy care” and found that “the interpersonal attributes of the therapist and the process of care are key determinants of patient satisfaction.” Given that, it’s a bit odd that the authors thought it was “unexpected” that how well treatment worked was “infrequently and inconsistently associated with patient satisfaction.”
I’ve always considered it a given that how a patient feels about a treatment has almost nothing to do with how well it worked (independently of placebo), but this study is the first time I’ve seen some good hard evidence of it. “Satisfaction is not the same thing as effectiveness” (Long).
- Humphrey N, Skoyles J. The evolutionary psychology of healing: a human success story. Curr Biol. 2012 Sep;22(17):R695–8. PubMed 22975000 ❐
Refreshingly, this is a perspective on placebo that doesn’t worship the “power” of placebo, but looks instead at the paradox of placebo: if placebo is so ding dang powerful, why isn’t everyone cured? Perhaps because the placebo effect is basically a dysfunctional mistake.
There’s a video about this (Why does the placebo effect work? 6:01), and here’s some excellent extra reading from Todd Hargrove, who did a nice job interpreting and explaining. The additional perspective is most welcome.
- ScienceBasedMedicine.org [Internet]. Brissonnet J. Placebo, Are You There?; 2015 Mar 12 [cited 22 May 26]. PainSci Bibliography 54158 ❐
Harriet Hall translated this French article on placebo for ScienceBasedMedicine.org, calling it “the best explanation of placebo that I had ever read.” (I was quite involved editorially, and in particular spent a bunch of time on producing shinier, translated diagrams.) I’m not quite sure it’s the best placebo explanation I’ve ever read, but on the other hand I can’t point to a better one, and it is certainly chock-a-block with historical context, effectively cited science, and little gems of clarity on this tricky topic.
The gist of the article is that “placebo” is an imprecise term and we need to distinguish placebo objects (essential for research), the effect of placebo (“that doesn’t exist,” because placebo objects inert), and contextual effects (which go beyond a response to a placebo object but are not “powerful”). Brissonnet concludes:
So, then! Placebo, are you there? The placebo object is certainly there! It will be irreplaceable for the foreseeable future in carrying out the controlled clinical studies that are essential to medical research. As for the effect of the placebo, that doesn’t exist. As for the effect “called” placebo, if its existence is undeniable albeit limited, it would be better to simply name it “contextual effect” in order to better understand its true nature and to make its magical connotations disappear.
- Ingram T, Silvernail J, Benz LN, Flynn TW. A cautionary note on endorsing the placebo effect. J Orthop Sports Phys Ther. 2013 Nov;43(11):849–51. PubMed 24175623 ❐ PainSci Bibliography 54098 ❐
- Ingraham. Statistical Significance Abuse: A lot of research makes scientific evidence seem more “significant” than it is. ❐ PainScience.com. 4270 words.
- Hróbjartsson A, Gøtzsche PC. Placebo interventions for all clinical conditions. Cochrane Database Syst Rev. 2010;(1):CD003974. PubMed 20091554 ❐
The nugget of this large review of placebo effects: “We did not find that placebo interventions have important clinical effects in general.” And, particularly pertinent to this website, “The effect on pain varied, even among trials with low risk of bias, from negligible to clinically important.” This is consistent with my position that placebo is far from “powerful,” but may have a strong impact on pain in ideal circumstances. As Dr. Steven Novella summarized it, “In other words, the best research we have strongly suggests that placebo effects are illusions, not real physiological effects.”
- Power and Hopayian describe the frustrebo effect in great detail in a paper for the Journal of the Royal Society of Medicine.
- Ingraham. Ioannidis: Making Medical Science Look Bad Since 2005: A famous and excellent scientific paper … with an alarmingly misleading title. ❐ PainScience.com. 3272 words.
- Strijkers RHW, Schreijenberg M, Gerger H, Koes BW, Chiarotto A. Effectiveness of placebo interventions for patients with nonspecific low back pain: a systematic review and meta-analysis. Pain. 2021 Dec;162(12):2792–2804. PubMed 33769366 ❐
- Kaptchuk TJ, Friedlander E, Kelley JM, et al. Placebos without deception: a randomized controlled trial in irritable bowel syndrome. PLoS One. 2010;5(12):e15591. PubMed 21203519 ❐ PainSci Bibliography 55214 ❐
- I wrote this in 2015: “A 2015 article advertised his plans to study open-label placebo for cancer. The 2010 study was well-crafted to produce a placebo effect, just using a different kind of deception. And he will do the same with the next study, with completely predictable results. And it will be a big deal, mark my words … because cancer. And because everyone loves the “power” of placebo!
- Carvalho C, Caetano JM, Cunha L, et al. Open-label placebo treatment in chronic low back pain: a randomized controlled trial. Pain. 2016 Oct. PubMed 27755279 ❐
- Locher C, Frey Nascimento A, Kirsch I, et al. Is the rationale more important than deception? A randomized controlled trial of open-label placebo analgesia. Pain. 2017 Dec;158(12):2320–2328. PubMed 28708766 ❐
They compared four groups of about 40 patients each, with one (control) group getting nothing at all, and the other three all receiving a placebo cream presented in a different way:
- a deceptive placebo: “this cream is medicine”
- open-label without rationale: “this cream is a placebo”
- open-label with rationale: “this cream is placebo, but it will help you, because placebo is potent”
Subjects were tested for heat-pain tolerance. By objective measures, everyone was the same, but there were differences when measured subjectively. Groups 1 and 3 were both given a reason to believe the cream would work, and they both experienced the same reduced intensity and unpleasantness; group 2 got no benefit. Thus the authors conclude that “placebos with a plausible rationale are more effective than without a rationale.” Translation: expectation of efficacy is the active ingredient, whether it is achieved with deception or exaggeration of the power of placebo.
- Carvalho C, Pais M, Cunha L, et al. Open-label placebo for chronic low back pain: a 5-year follow-up. Pain. 2021 05;162(5):1521–1527. PubMed 33259459 ❐
This paper arrives at the rather extraordinary punchline that an honest placebo (“this is a placebo we’re giving you”) not only helped back pain patients, but continued to do so for five years. Despite a rather glaring limitation, they believe that “our data suggest that reductions in pain and disability after open-label placebo may be long lasting.” For contrast and more detail, see Kleine-Borgmann et al, a similar study that did not find a long-term benefit.
It is the official position of the Salamander that this study continues a tradition of overhyped BS about open-label placebo.
- Kleine-Borgmann J, Dietz TN, Schmidt K, Bingel U. No long-term effects after a three-week open-label placebo treatment for chronic low back pain: a three-year follow-up of a randomized controlled trial. Pain. 2022 Aug. PubMed 35947884 ❐
In 2019, Kleine-Borgmann reported that prescribing an open-label (non-deceptive) placebo for back pain was more helpful than typical care. In this follow-up three years later, they report that the benefit did not last: “our follow-up data do not support the previously suggested assumption that a three-week open-label placebo treatment has long-term effects.” Not really a surprise.
As the authors note, “research on nondeceptive placebos is still in its early stages, and studies on long-term effects are almost nonexistent.” They present their findings primarily in contrast to the only other data of this kind: Carvalho et al’s startling conclusion that the benefits of an non-deceptive placebo were sustained for five years. However, as they acknowledged, there was a significant methodological limitation. As pointedly summarized by Kleine-Borgmann et al.:
“…because of a predefined crossover from treatment-as-usual to open-label placebo, in which all patients received OLP at the end of the study, follow-up was performed as an observational analysis without a control group, preventing causal attribution of continued improvements to OLP treatment.”
So the sensational long-term results Carvalho et al. reported were probably just an artifact — and now those results are contradicted by this study.
- Tuttle AH, Tohyama S, Ramsay T, et al. Increasing placebo responses over time in U.S. clinical trials of neuropathic pain. Pain. 2015 Dec;156(12):2616–26. PubMed 26307858 ❐
- Berman BM, Langevin HH, Witt CM, Dubner R. Acupuncture for Chronic Low Back Pain. N Engl J Med. 2010 Jul 29;(363):454–461. PubMed 20818865 ❐ PainSci Bibliography 54942 ❐
- The New England Journal of Medicine does not actually have a great reputation for editorial rigour. (The Last Psychiatrist recently snarked at it, “NEJM: where peer review= spell check”!) This bizarre article, in such a prominent journal, attracted the attention of critics at Science-Based Medicine, of course: both Drs. Crislip and Novella wrote about it this quite brilliantly. (Dr. Crislip’s post is funny.)
- Though goodness knows that’s enough of a problem right there. One of the classic perks of placebo is that sugar pills are cheap — in a world full of impoverished patients, an expensive placebo is a bad idea right out of the gate.
- Harris I. Surgery: The ultimate placebo. NewSouth Publishing; 2016.
This excellent book by an orthopedic surgeon explores the shameful history of untested surgeries in considerable detail. It’s fascinating and mostly easy enough reading even for patients.
- Clever Hans is the canonical example of animal sensitivity to subtle human cues: the horse who seemed to be able to do arithmetic and other intellectual tasks. What he was actually doing was much more remarkable in some ways: “responding directly to involuntary cues in the body language of the human trainer, who had the faculties to solve each problem,” the “Clever Hans effect.” And that effect actually wasn’t just limited to his trainer, either — he could usually get the answer by watching the reactions of any person who was watching.