The Placebo Phenomenon | Hacker News Reader

The Placebo Phenomenon

harvardmagazine.com

55 pointspaulsutter39 comments

From the article: "Kaptchuk . . . doesn’t argue that you can simply 'think yourself better.' 'Sham treatment won’t shrink tumors or cure viruses,' he says."

That's an honest statement and an important point, because many publications with Kaptchuk's name on them are overinterpreted to reach the conclusion that there is a "real" placebo effect that can actually cure disease. In fact, there is no strong evidence of that at all.

This issue has come up often enough on Hacker News that I have gathered some links that are helpful for understanding what placebo effects are all about. Some of these online links cite quite a few useful scholarly publications.

http://www.sciencebasedmedicine.org/index.php/michael-specte...

"In other words, the best research we have strongly suggests that placebo effects are illusions, not real physiological effects. The possible exception to this are the subjective symptoms of pain and nausea, where the placebo effects are highly variable and may be due to subjective reporting."

Numerous press releases on the Web point to publications co-authored by Ted Kaptchuk, the main person profiled in the article kindly submitted here. Although Kaptchuk has the academic title of an associate professor of medicine, he has no medical training or credentials or clinical experience in independently verified patient care.

http://www.sciencebasedmedicine.org/index.php/dummy-medicine...

The article submitted here today, as befits an article from Harvard Magazine, is pretty good about giving the point of view of researchers who disagree with Kaptchuk, and is good about giving his updated opinion on issues he has changed his mind about. The statements found in some earlier articles posted to Hacker News, such as "Recent research demonstrates that placebo effects are genuine psychobiological phenomenon [sic] attributable to the overall therapeutic context, and that placebo effects can be robust in both laboratory and clinical settings" are untrue.

http://theness.com/neurologicablog/index.php/the-rise-and-fa...

"Despite the spin of the authors – these results put placebo medicine into crystal clear perspective, and I think they are generalizable and consistent with other placebo studies. For objective physiological outcomes, there is no significant placebo effect. Placebos are no better than no treatment at all."

http://www.ncbi.nlm.nih.gov/pubmed/20091554

"We did not find that placebo interventions have important clinical effects in general. However, in certain settings placebo interventions can influence patient-reported outcomes, especially pain and nausea, though it is difficult to distinguish patient-reported effects of placebo from biased reporting. The effect on pain varied, even among trials with low risk of bias, from negligible to clinically important. Variations in the effect of placebo were partly explained by variations in how trials were conducted and how patients were informed."

Fabrizio Benedetti, a co-author of one of the most cited papers who is also a medical doctor, sums up his view this way: "I am a doctor, it is true, but I am mainly a neurophysiologist, so I use the placebo response as a model to understand how our brain works. I am not sure that in the future it will have a clinical application."

http://www.brainsciencepodcast.com/storage/transcripts/year-...

To sum up, despite claims to the contrary that are often covered by the lay press, the best-considered view among medical practitioners with clinical experience is that the placebo response has no ethical clinical application.

http://www.skepdic.com/placebo.html

http://www.sciencebasedmedicine.org/index.php/revisiting-dan...

And of course see LISP hacker and Google director of research Peter Norvig's article "Warning Signs in Experimental Design and Interpretation" on how to interpret scientific research.

http://norvig.com/experiment-design.html

Isn't focusing on the placebo effect ignoring the broader and more important question of the impact of state of mind on the body's ability to fight disease and heal itself? Once that question is well understood I imagine the understanding of notions like the placebo effect will simply fall out.

A rather deep research and clinical practice on working with the mind to cure the body has been conducted by John E Sarno.

The Mindbody Prescription [1] and The Divided Mind [2] are two of his books.

My interest in this started with the motivation to get some relief for the RSI-like symptoms and it's been part of my recovery for a few weeks together with the improved ergonomics.

[1] http://www.amazon.com/The-Mindbody-Prescription-Healing-Body...

[2] http://www.amazon.com/The-Divided-Mind-Epidemic-Disorders/dp...

> To sum up, despite claims to the contrary that are often covered by the lay press, the best-considered view among medical practitioners with clinical experience is that the placebo response has no ethical clinical application.

Of course I agree with this view, but when interpreted in its widest sense, it argues against the legitimacy of religion. It also occurs to me that few will notice the connection.

On a more practical level, it argues against most psychological treatments, which research shows cannot reliably be distinguished from a placebo effect. Few will notice this connection either.

This research invalidated the only reason I had for not minding homeopathy. It was at least healing people by the placebo effect.

Now it doesn't even do that. Good riddance.

Huh, your extensive research doesn't square with much of what I learned as a neuroscience undergrad. We read a few papers that demonstrated non-trivial physiological responses due to 'sham' treatments.

In particular, placebo treatments have a well established impact on pain by causing release of endogenous opioids (see http://webusers.astro.umn.edu/~larry/CLASS/NOTHING/READINGS/... and http://pharmacology.ucsd.edu/graduate/courseinfo/placebartic...).

More generally, it would be shocking if beliefs and perceptions didn't have (both beneficial and harmful) medical effects. Mental states have physical consequences. For example, stressful thoughts cause measurable differences in levels of various hormones and neurotransmitters. Over time, particular patterns of thinking can even cause changes in the morphology of the brain. The strong assertion of a mind/body split seems simplistic and outdated, there's simply too much evidence of a complex entangled relationship between the two.

But the effects are still rather limited to nausea, pain and the systems influenced by stress response (esp. some immune system functions). You have to look at the absolute change in the body. Which is rather limited. Stress in general can be reduced by more effective means (CBT, relax. tech.) than sham treatments.

For practical application there are other OTC substances you can take (such as adaptogens/anxiolytics - e.g. Ashwaganda & l-theanine) which will have a placebo effect (obviously differing on method of deliver (white coat vs. self, injection vs. pill) and additional effects.

> From the article: "Kaptchuk . . . doesn’t argue that you can simply 'think yourself better.' 'Sham treatment won’t shrink tumors

http://www.ncbi.nlm.nih.gov/pubmed/22287549

> process by which stress may potentiate carcinogenesis and how reducing body stress may prevent cancer growth and progression. The opioid peptide β-endorphin plays a critical role in bringing the stress axis to a state of homeostasis

http://www.scienceagogo.com/news/20050725232410data_trunc_sy...

> A study from the University of Michigan (U-M) has provided the first direct evidence that endorphins - the brain's own pain-fighting chemicals - do play a role in the phenomenon known as the placebo effect

Sure throws a spanner in the works.

You should read you sources. These are not hard data. It's guesswork.

I get very uneasy reading articles that start to pitch "value of caring" and someone who "didn't doubt the value of acupuncture" against "western medicine" and "western doctors" who are portrayed as stubborn and close-minded.

Agreed. "Western medicine" only exists as half of a false dichotomy.

If the doctors so labeled really were stubborn and close-minded, then we would still perform blood-lettings and chew tree mark (instead of just the active ingredient). Instead, these doctors are more open to new information; they're just willing to also incorporate the evidence against efficacy so often ignored by "alternative medicine" practitioners.

That's not 100% true. It's only really the case in patent protected medicines.

New drugs undergo testing, to get FDA approval. New surgical treatments don't. New (and old) surgical treatments and old drugs aren't lucrative enough to warrant real testing. Surgeons will claim it's for ethical reasons, or because they just know better, or that they are artists who can't be constrained by the laws of statistics, but it's simply a lack of regulations and / or incentives.

That's not 100% true.

Correct. That's why I expressed it as a relative position instead of making any such absolute claims.

My theory: many disorders are stress-related. Receiving care - even if it's a sham - relieves stress.

Going deeper: we human animals have limited physiological resources. Are they channelled into healing or surviving? If you are in a dangerous environment, then it's survival first, healing second. Therefore, a sense of safety will promote healing.

A placebo provides just that.

If the placebo effect is superstition, as most comments in this thread seem to imply, how come pharmaceutical companies are finding it increasingly hard to beat?

http://www.wired.com/medtech/drugs/magazine/17-09/ff_placebo...

Basically, the "effect" is the margin of error under which variations in self-healing, self-reporting, and researcher bias obscure actual effect.

As we address more and more of the "big" pharmaceutical effects, that only leaves the very minor ones or big ones that only apply to increasingly thinner edge cases.

If the human body was software, one might say that the pharmaceutical companies have debugged it to the point that they have little left to find other than kilobyte-sized memory leaks. Beyond that, the vast majority of the blue-screening issues remain PEBKAC errors.

On a tangent, I think some of these massive open online courses going around nowadays might act as a form of what Rory Sutherland terms "placebo education". Education where you don't necessarily learn things but that gives you the impression that you did, resulting in extra confidence leading to extra success later on ;-)

If you continue this line of thinking, putting items on todo lists and checking them off is a "placebo completion". Getting a fancy title instead of a promotion or raise is a "placebo advancement".

A placebromotion, if you will.

Not to be confused with the hypophone, bromotion, wherein the new title is accompanied by a round of high-fives and concludes with a keg stand.

Along these lines: placebo buttons, e.g. walk buttons at crosswalks, elevator door close buttons

http://en.wikipedia.org/wiki/Placebo_button

> Companies spend millions of dollars and often decades testing drugs; every drug must outperform placebos if it is to be marketed. “If we can identify people who have a low predisposition for placebo response, drug companies can preselect for them,” says Winkler. “This could seriously reduce the size, cost, and duration of clinical trials…bringing cheaper drugs to the market years earlier than before.”

Wat? Is that a typo? Shouldn't it be "preselect against them"? .. or better still, include predisposition to placebo response as a variable?

I think what he means to convey there is that a person with a low predisposition would be affected by the actual content in the drug than just taking any pill.

Including predisposition to placebo as a variable would just increase the number of experiments needed. Instead, if you could eliminate a variable from the system by preselecting for people for whom the variable doesn't apply, it would reduce the time to experiment

> Instead, if you could eliminate a variable from the system by preselecting for people for whom the variable doesn't apply, it would reduce the time to experiment

... if the goal is to reduce the time to prove that your drugs work. That goal is different from making drugs that work.

For an analogy - in order to prove that I'm smart, I must compete with those others who're known to be smart, not with those who've (by some agreed on measure) demonstrated a lack of the required aptitude.

Even this article keeps emphasizing throughout that Kaptchuk is at Harvard, he got praise from scholars, etc.

So the real proof of a drug's worthiness ought probably to be that it works over and above the placebo effect shown for those who respond very well to placebos. That would provide more data to help evaluate the risk of a drug to a patient, given placebo predisposition can indeed be measured. That would be important particularly if the drug has significant unwanted side effects.

edit: And fwiw, preselecting for those who would respond well to placebos would also reduce the time to experiment .. but, I think, it would do so in the right way.

The placebo/nocebo effect appears to be simply an example of the confirmation bias, a couple of other psychological biases and various other known statistical effects.

> False impressions of placebo effects can be produced in various ways. Spontaneous improvement, fluctuation of symptoms, regression to the mean, additional treatment, conditional switching of placebo treatment, scaling bias, irrelevant response variables, answers of politeness, experimental subordination, conditioned answers, neurotic or psychotic misjudgment, psychosomatic phenomena, misquotation, etc. These factors are still prevalent in modern placebo literature. The placebo topic seems to invite sloppy methodological thinking. Therefore awareness of Beecher's mistakes and misinterpretations is essential for an appropriate interpretation of current placebo literature.

Beecher retrospectively attributed the improvements in the placebo groups to effects of the placebo administration. However, on the basis of the published data, in all of these trials the reported outcome in the placebo groups can be fully, plausibly, and easily explained withoutpresuming any therapeutic placebo effect. The published data of these trials make it quite obvious that there were a variety of reasons for the reported results, such as spontaneous improvements, additional treatments, methodological artifacts, etc. In some of the original trial publications even the authors themselves had explicitly written that there were no placebo effects.

Source: The powerful placebo effect: fact or fiction?, GS Kienle, H Kiene, Journal of clinical epidemiology, 1997 (http://ukpmc.ac.uk/abstract/MED/9449934/reload=0;jsessionid=...)

Placebo interventions were again not found to have important clinical effects in general but may influence patient-reported outcomes in some situations

Source: http://en.wikipedia.org/wiki/Placebo#Clinical_utility

The effect, apparently, doesn't exist.

> Placebo interventions were again not found to have important clinical effects in general but may influence patient-reported outcomes in some situations

But doesn't this matter?

As it says in the article of course, feeling better when you're not actually better can be dangerous for potentially lethal ailments; but in diseases that don't kill you, feeling better means a lot.

Placebo in a nutshell: inject someone in pain with morphine. do it repeatedly over time. once, without telling them, inject them with saline instead. their brain will respond as if it was morphine and continue to block the pain.

The whole "people believing X so X happens" is a perverse co-opting of the placebo title.

And, just for fun, if you give a patient naloxone (opiate blocker) before giving them the saline, their brains won't interpret it as morphine. naloxone blocks placebo response. in case you missed that, read it again: naloxone blocks placebo response. (and once more: placebo means a very narrow thing. it means your body responds as if it biologically received the medicine it had been receiving -- it does not mean any new age "think your way to better health" mumbo jumbo).

> inject someone in pain with morphine. do it repeatedly over time. once, without telling them, inject them with saline instead. their brain will respond as if it was morphine and continue to block the pain.

I must admit I have a very hard time believing this exact claim as written. Do you have any sources?

A talky explanation is included in http://www.youtube.com/watch?v=gQS0tdIbJ0w

As for other sources, just type random related words into a search engine until something reliable pops out. You'll mostly find pop science articles including the word placebo though.

> I must admit I have a very hard time believing this exact claim as written. Do you have any sources?

As it happens, one of the original placebo observations took place in a battlefield hospital that had run out of morphine. A physician began injecting saline, hoping the switch wouldn't cause too much psychological stress. To his surprise, many of his patients continued to report pain relief. Source:

http://www.brainfacts.org/sensing-thinking-behaving/mood/art...

"While treating wounded American soldiers during World War II, Henry Beecher ran out of pain-killing morphine. Desperate, he decided to continue telling the soldiers that he was giving them morphine, although he was actually infusing them with a saline solution. Amazingly, 40 percent of the soldiers reported that the saline treatment eased their pain."

That's not a controlled study, plus the effect claimed is much weaker than what I was addressing. I can believe that placebo provides some pain relief--I was addressing the claim that they provided identical levels of relief. All the studies I could find on PubMed comparing morphine to placebo for treating various kinds of pain have morphine as the undisputed winner.

It's not morphine or salt water. It's morphine->morphine->morphine->morphine->salt water.

Pharmaceutical studies of only using drug or not-drug have different characteristics.

I got that from your original post. Unfortunately, I wasn't able to find any studies with that experimental setup, hence my earlier question to you.

> I can believe that placebo provides some pain relief--I was addressing the claim that they provided identical levels of relief.

I doubt that any serious researcher will make that claim. Such studies are, after all, based on self-reporting, notoriously unreliable and difficult to accurately quantify.

This is not to say that psychologists won't make claims like that. But psychological science is only peripherally related to the other kind of science -- the real kind.

>That's not a controlled study

Just wondering, how could you get rid of the placebo effect in the placebo experiment?

Fair point. I was using "controlled" as a shorthand for the usual qualities (double blind, randomized, sufficiently many test subjects, etc) you'd associate with good experimental design.

> ... how could you get rid of the placebo effect in the placebo experiment?

That's easy -- give one group a real placebo, give the other an imitation.

Some related studies:

'Partial antagonism of placebo analgesia by naloxone' (http://www.ncbi.nlm.nih.gov/pubmed/6308540) -- no actual morphine was ever adminstered, only a regimen of saline injections, whose placebo effect is apparently achieved by the release of endogenous opioids.

'A comprehensive review of the placebo effect'(http://pharmacology.ucsd.edu/graduate/courseinfo/placebartic...): "placebos have actual biological effects on the brain and body and are more than response biases."

In case you missed it, endocannabinoids (CB1 agonism, in particular) appear to account for the other half of placebo analgesia:

http://www.nature.com/nm/journal/v17/n10/full/nm.2435.html

Reply on news.ycombinator.com