Mar 11 2008

The Price of Placebos

Placebos are inactive treatments used in clinical trials as a comparison for the active treatment being studied. This is necessary because there is a host of artifacts and effects that create the appearance of a response to a treatment above and beyond any real physiological response – collectively called “the” placebo effect. To further illustrate how squirrelly the placebo effect can be, Waber et. al. recently published a study (you may need a subscription to access this link – here is a link to a NYTimes article discussing the study) in which they found that more expensive placebos had a greater effect on pain reduction than less expensive placebos.

The researchers looked at 82 subjects who were receiving a placebo they were told was a pain reliever and their pain tolerance was assessed using small electrical shocks (a standard procedure). Half of the subjects were told the placebo was a medication that cost $2.50 per pill and the other that it cost $0.10 per pill – 85.4% of the subjects getting the expensive placebo experienced pain reduction, while only 61% of the cheap placebo group experienced pain reduction.

While many news outlets reports these results as surprising, they actually make perfect sense. The study authors interpreted the findings as indicative of a well-known marketing effect – that of quality expectations. We expect higher quality from more expensive products, or lower quality from cheap products. Therefore those subjects who were told their placebo cost $2.50 per pill had higher expectations of efficacy, perhaps resulting in the release of more pain-fighting endorphins. The authors speculate, reasonably, that this effect may explain why many people believe that expensive brand name drugs are more effective than their cheaper, but chemically identical, generic alternatives.

However, while I think this explanation is plausible and probably a factor, it reflects the common assumption that the placebo effect is chiefly a mind-over-matter phenomenon, while downplaying or ignoring the many other factors that make up the placebo effect. One such factor is the purely psychological effect known as investment justification. This results from our tendency to justify to ourselves investments of time, energy, money, or other resources by convincing ourselves that the investment was justified by a proportional benefit. So if we sink time, money, and hope into a treatment we want to believe it was worth it, therefore we convince ourselves that we are better, even if there has not been any real physical improvement.

Of course this, like any psychological factor, is not absolute. There are also many ways to reduce the cognitive dissonance resulting from believing in a treatment that does not work other than just self-deception that it did work. For example we could come to the belief that the treatment was worth a try, that the chance of benefit was worth the investment. Or that the investment was justified because of the hope that it brought, if nothing else. Or we could conclude that we were deceived by those offering the treatment or making the claims for it, and therefore we can blame our loss of investment on someone else and see ourselves as a victim.

This is a very interesting study that further illuminates the nature of the placebo effect. I would like to see it replicated with more subjects, while also exploring other aspects of this effect – for example would it also be seen for outcomes other than pain. But also the lesson we should take away from studies such as this is that the placebo effect is actually a complex array of multiple effects. Some of those effects are psychological, which themselves may be complex and subtle and yet have a profound effect on the apparent response to a treatment. Placebo control is a critical part of our medical science technology, and therefore we are wise to understand the nature of the various placebo effects as much as possible.

42 responses so far

42 thoughts on “The Price of Placebos”

  1. Amon1492 says:


    Can the reverse of this “placebo effect” have a detrimental effect on health in society? (i.e. We know a product/behavior is bad for us because we have been informed that it is. Consuming this product or performing this behavior potentially does even more harm than otherwise due to our psychological bias against it.)

    Fast food is known to be typically unhealthy; high in saturated fats and cholesterol leading to obesity and other various diseases. If the public is aware this food is not the optimal source of our daily caloric intake, however due to rising prices and relatively cheap and vast accessibility – Americans still routinely chow down on enormous amounts of this unhealthy fare.

    Could the consumption of fast food do even more harm because we know it is bad for us, yet we are compelled to partake in order to accommodate our lives on the “run”? Or maybe less so for someone who believes eating fast food can do no harm so long as it is not their major staple of caloric intake?

    We would not have to limit our selection to fast food. Perhaps smoking could be examined.

    Is there any way to even test this? Or are their too many variables at work to come to any conclusive evidence one way or the other?

  2. Amon,

    That’s an interesting ideas but I think you would need to formulate a more specific prediction before designing a way to test it.

    Also – I don’t agree with your premise (I know it was just and example) regarding fast food. This is actually a big misconception. Fast foods are actually quite healthful in that they are good nutrition and usually contain many food groups. They are typically no worse in saturated fats than other similar foods that we eat.

    The problem with fast food is that they tend to be high in calories – big portions cheap. It is easy to overeat with fast food. But as long as you keep the calories at a reasonable level, there is nothing unhealthy about fast foods.

  3. Amon1492 says:

    Thanks for the fast response!

    I am surprised to hear fast foods saturated fats are no worse than ordinary foods. Definitely something for me to keep in mind next time I am at burger place for a quick snack (keep it small stupid).

    I suppose my original question boils down to whether negative bias against a product can lead to a more negative impact (or less positive impact) than ordinarily, had there not been a negative bias at all.

    If psychological mindset appears to have a subtle effect of some magnitude – perhaps positive thinking all around might be warranted and beneficial.

  4. Tom Nielsen says:

    Amon, about the fast food being unhealthy subject, you can also learn more about the subject in the 10 min long Skeptoid podcast, which you can find here:

    (sorry, Steven, for redirecting readers to other podcasts than the brilliant SGU podcast)

  5. Roy Niles says:

    Dr. Novella said in part: “There are also many ways to reduce the cognitive dissonance resulting from believing in a treatment that does not work other than just self-deception that it did work.”

    Perhaps this is quibbling a bit, but it’s not accurate to call it self-deception if the deception was actually accomplished by inference from a trusted source that the treatment could be expected to work.

    Self-deception ordinarily requires one part of a person’s mind to know something that for one reason or another is kept from or deliberately ignored by another part of that same mind.

    One would have to at least suspect the possibility of a placebo being involved to give rise to self-deception, or even cognitive dissonance.
    Someone who is led to expect a certain result is more likely to interpret actual results as successful than not, because the expectations alone influence any such assessment.

  6. EdSG says:

    That’s a very good episode Tom, I like Brian Dunning format too: short and to the point, leaving you to think about it and investigate more on your own.
    And The Skeptics’ Guide to The Universe gives you information AND entertainment, that’s just awesome.

    What I find most fascinating it’s the misunderstanding of the placebo effect; me being a layman without scientific training, I always thought the “mind over matter” was a fact, then comes Dr. Novella and explains it so simple that I had to slap my forehead for not really thinking it through.

    The article that made me understand it more was precisely the first one Dr. Novella pointed at above: ” – collectively called “the” placebo effect.” :

    Anyway, thanks for making this information a little more understandable for us laymen, Dr. Novella.

  7. pec says:

    “Fast foods are actually quite healthful in that they are good nutrition and usually contain many food groups. ”

    I have NEVER heard anyone say that before! Most fast food has no unprocessed ingredients and therefore it CANNOT be “quite healthful.”

    We have evolved in the natural environment and therefore our food should be as natural as possible. Whole grains are better than refined grains. Fruit and vegetables are better than soda and candy. Didn’t your mother tell you anything?

  8. ellazimm says:

    pec, what specifically is wrong with processed food? If you are objecting to additives then which ones? Reason before reaction.

    Dr Steve, I was wondering if an attempt had ever been made to blind patients to the fact that they had even been given a potential treatment? I can see there is an ethical issue here but I can’t imagine that someone hasn’t thought to test a new treatment on people who didn’t even know they were getting it. I do not want to come across as a “mad scientist” type but I was just trying to think of a way to avoid having to deal with placebo issues.

  9. Nitpicking says:

    Pec, you’ve gone from claiming to be a scientist to spouting nearly-meaningless stuff like the idea that all processed foods must be unhealthy since they didn’t exist thousands of years ago. Can you say “non sequitur”?

  10. ellazimm – you are referring to so-called triple blinding – blinding subjects to the fact that they are subjects. This is unethical because it precludes informed consent, so your question will go unanswered.

  11. pec is making the naturalistic fallacy (he is a textbook of logical fallacies)

    Also – fast food is not equivalent to “soda and candy”.

    Listen to the skeptoid episode if you want to hear someone else say it – and no worries, Brian Dunning is a skeptical colleague and we are in fact currently collaborating on a project.

  12. pec says:

    If you believe in evolution, then you must be able to see that natural selection has acted on our species, and our species’ ancestors, for quite a long time. Are you discounting the importance of natural selection?

    Our ancestors did not have access to plentiful refined sugar, for example, so evolution has not prepared us with ways of dealing with it. It is very common for Americans, and anyone on the Western fast-food diet, to develop insulin-resistance.

    Refined grains — found in all fast food — are also a problem, for the same reason. They are not found in nature, and therefore natural selection has not prepared us to deal with them.

    Natural food is extremely complex and cannot be fabricated in laboratores. But that is what the fast-food makers think they can do. If you take all the natural nutrients out of a substance, and then pour some back in, you are not creating healthy food. That is because nutrition is still not well understood.

    This is not a logical fallacy. It is obvious and it is common sense.

    Until recently most fast food contained trans fat, which we now know contributes to heart disease. We don’t know what harmful substances fast food might still contain. Should we keep eating fast food 10 years until science discovers it contains more harmful substances, or should we eat natural food instead?

    Our diets can’t be perfectly natural, because of the way most food is grown. But we can use a little common sense and follow the obvious advice we have all been given all our lives.

    I can hardly believe I am having this conversation! Mainstream medicine advocates eating fruit, vegetables and whole grains, and avoiding refined sugar and flour.

    Novella is deep in his fantasy world.

  13. Skeptical Cat says:

    Hey pec, did you ever consider the fact that we are omnivores, and that maybe “evolution prepared us” to be able to eat and be healthy eating a lot of different things? Even things that we have never encountered before?

    “That is because nutrition is still not well understood.”

    So why do you lecture us on what we should and should not be eating? This is indeed a logical fallacy: argument from ignorance. “Hey, we don’t know what’s bad for us, so you should avoid these foods because they are bad for you!”

    “It is obvious and it is common sense.”

    Sounds like dogma to me.

  14. petrucio says:

    “This is not a logical fallacy. It is obvious and it is common sense.”

    As once was the Flat Earth and Geocentrism. Being ‘obvious’ and common sense has nothing to do with being truth. ANOTHER logical fallacy right there!

  15. Steve Page says:

    In moderation, they’re not nearly as bad for you as you think. In excess, there’s no doubt that the consumption of fast food is dangerous, but unless you’re Morgan Spurlock or Mr Creosote, you don’t have to panic.

    If you are Mr Creosote, I have a wafer thin mint here with your name on it.

  16. petrucio says:

    I’ll believe what proper studies tell us.

    Many so called ‘experts’ don’t know squat. The vast majority of Nutritionists would tell you that fast food is bad, but they are just following the common sense and the naturalistic fallacy (nutritionists are specially fond of that one), not weighting the appropriate evidence.

    My fast food meals have tomatoes, onions, lettuce, pickles. And I don’t eat the French fries nor the soda. They are one of the most balanced and low-calories meals I eat. They’d only be bad for me if there’s some unknown ‘ghost in the meal’ that most people seems to think there is in anything ‘unnatural’.

  17. Roy Niles says:

    Pec believes we hit the fast food slippery slope when man first put raw meat on the fire.

  18. Evidence certainly does support eating fruits, vegetables, and whole grains. I never said otherwise. It also supports eating a varied diet.

    My point was that fast foods are not “candy” and they are not poison. They are also not as bad as many people falsely assume.

  19. daedalus2u says:

    Humans are unable to achieve good health eating a “natural” diet, i.e. one eating things only “found in nature”. Humans require their food to be cooked. There have been no recorded instances of hunter gatherers not utilizing fire for cooking.

    When humans attempt to eat an uncooked diet while “hunting and gathering” at a supermarket (i.e. with essentially zero acquisition metabolic cost) they lose weight and women experience amenorrhea. They become infertile. Children become malnourished and die of malnutrition.

    Cooking is an absolute requirement for modern humans. Cooking our food has become “natural”.

  20. pec says:

    “Evidence certainly does support eating fruits, vegetables, and whole grains. I never said otherwise. It also supports eating a varied diet.”

    You said:

    “Fast foods are actually quite healthful in that they are good nutrition and usually contain many food groups.”

    They don’t contain fruits, vegetables, or whole grains. They don’t contain anything that isn’t processed or created in a laboratory.

    The evidence supports NOT eating fast food. The health crisis in America is additional evidence. It’s quite sick that an MD would call fast food “healthful,” and he probably says that to his patients. If there’s a shriveled up shred of lettuce and some ketchup on the cheeseburger, that counts as vegetables.

    But it’s OK, because you can just prescribe statins for your patients. Actually, it works out well — more money for the fast food industry, the drug industry, and for the MDs.

  21. pec says:

    Ok Skadorwa, I forgot about Subway. It’s the only fast food I will go along with occasionally. I am not obsessed with nutrition and I just try to be reasonable.

    Aside from Subway I really don’t know any others. Inexpensive restaurants in general are not all that healthy (btw “healthful” and “healthy” have the same meaning and are both adjectives, so we don’t need to say “healthful” and sound odd).

    Subway does have a variety of vegetables that are not all shriveled up, and it is fast.

    Oh another one, where I live, is Panera, and it’s great. It has really good salad.

    But I do not think Dr. Novella was talking about Subway or Panera. The most famous fast food is McDonald’s and Burger King, Dunkin’ Donuts, places like that.

    You don’t have to be anti-nature just because you’re a scientific materialist. As I said, if you believe in evolution and natural selection you must admit that “natural” food, the kind of substances we evolved to use, is better for us. Food that is conjured up in laboratories, or processed to death, has unpredictable and often harmful effects.

  22. Roy Niles says:

    daedalus2u: I think you sort of missed the point of my comment about putting raw meat on the fire.
    And you said: “Humans require their food to be cooked. There have been no recorded instances of hunter gatherers not utilizing fire for cooking.”
    Actually, there are few recorded instances of hunter gatherers making mention of what they did at all, let alone whether they always cooked their food. They haven’t said much about what they did before they discovered how to make fire, for example.
    Also there are a few recorded instances of more modern humans eating sashimi and other raw fish, and many instances of eating dried meat -also of some hunting societies letting meat “hang” a bit to soften it up for chewing.
    Also raw ham was considered a bit of a delicacy until trichinosis became a recognized threat.
    I could go on about steak tartare, etc., but it’s time for a little diced liver, blubber and roe.

  23. Nitpicking says:

    Pec is engaging in the classic crank argument, almost certainly without realizing it. Pec, think about it and you may realize that your whole argument boils down to “My intuition is more valuable than the accumulated knowledge of all the world’s experts.”

  24. pec – you are the one making unjustified absolute statement and assumptions. You have abandoned the mantle of “reasonable” long ago.

    You are taking my comments out of context – you are the one who said fast food was “candy” and now you are saying it is all processed. This is simply not true.

    And having lettuce, pickles, and tomato on a sandwhich is not negligible – it contributes to a varied diet, and they are not processed. And yes, I include subway, and panera etc in fast food. You can also get salads at most fast food restaurants. But a hamburger is a good nutritious meal also, and it is not unhealthy if you eat it occasionally as part of a varied diet. Nothing is good if you eat it every day or to excess.

    Of course if you eat large amounts of french fries and wash it down with a chocolate shake, that is not a very healthful diet.

    Regarding a “natural” diet, of course it is better to eat food that we are evolutionarily adapted to eat. No one said otherwise (another straw man), but that can be taken to an extreme – like saying that food should be as natural as possible, and that any processing is bad. Just about everything we eat – the fruits, vegetables, and meats – have been cultivated and domesticated in the last 10-15 thousand years. It is not clear if this has been enough time to evolve to adapt to it. In our “evolutionary milieu” we likely ate very differently than after the agricultural revolution – perhaps gorging on a kill then not eating for a couple of days, eating fat because it is calorie dense, etc. My point is that you have to be careful not to over apply the notion of what is “natural” for humans – that is where you get into the naturalistic fallacy.

  25. daedalus2u says:

    This is the article I was refering to

    Human ancestors started using fire well before the Homo sapiens speciation even occurred. In this article they argue that modern humans are incapable of surviving without fire or other heat for cooking. I brought it up to make the point that “natural” is not synonymous with “being able to sustain life”.

    My point wasn’t that hunter gathers always cooked their food, but that no hunter gatherers, or any other human society has ever been found that does not cook some of their food.

  26. Roy Niles says:

    daedalus2u: Not only was that NOT what you said was your point, but you still don’t get that the reference to fire was part of a joke – the point of that being that humans (or perhaps I should have said genus homo) did eat raw meat at one time.

    And Homo erectus reportedly started the use of for cooking purpose. Cooked food is softer than noncooked food and would have been easier for young children and the elderly to chew. The powerful jaws of the earliest genus homo species were now no longer needed and the result was an evolution of our face and cranial structure.

    Now if you want to restrict the definition of “human society” to that which occurred AFTER these homo erectus people found fire useful, then of course you kill the whole point of the joke.

  27. Roy Niles says:

    I see there were some typos earlier, but that was due to my thinking too much in a naturalistic and therefor fallacious manner.

    I was trying in some way to see if pec’s naturalistic ruminations were not entirely out of line. I checked with the ever-present and usually reliable Wikipedia, according to which the naturalistic fallacy is often claimed to be a formal fallacy (hinting that it might not be as claimed?).
    It was described and named by British philosopher G. E. Moore in his 1903 book Principia Ethica. Moore stated that a naturalistic fallacy was committed whenever a philosopher attempts to prove a claim about ethics by appealing to a definition of the term “good” in terms of one or more natural properties.
    He explains that the fallacy involved is an instance of a more general type of fallacy, which he leaves unnamed, but which some might call the “definitional fallacy”. The fallacy is committed whenever a statement to the effect that some object has a simple indefinable property is misunderstood as a definition that gives the meaning of the simple indefinable property.
    Now the question is, who’s at fault when such a misunderstanding occurs? Clearly we must conclude from extensive evidence, and perhaps game theory pattern analysis (or was that something or other Bayesian) that this is clearly a matter of cognitive dissonance on pec’s part, brought on by the also naturalistic view that self-deception had survival value for the tribe if not for some of those who were sacrificed in the belief that the tribal stew pot was in need of a certain je ne sais qua for those long-neclected taste-buds. That was after the onset of the homo sapiens of course.

  28. Roy Niles says:

    What are the odds that sesquipedalian fallacies are self-replicating, and therefor take on lives of their own?

  29. daijiyobu says:


    Yes, memes [per ]:

    “* meme is: an idea that, like a gene, can replicate and evolve.

    *a unit of cultural information that represents a basic idea that can be transferred from one individual to another, and subjected to mutation, crossover and adaptation.

    * a cultural unit (an idea or value or pattern of behavior) that is passed from one generation to another by nongenetic means (as by imitation); ‘memes are the cultural counterpart of genes.'”

    By the way, per sesquipedalian, the Times [see ] has an article on its larger cousin,




  30. DavidCT says:

    It is interesting that the press found this modification of expectations by price would alter perception of pain relief.

    When this sort of study is done with wine, people percieve the wines presented as more expensive taste better. Psycologically this is the same sort of study but the results are no surprise when dealing with testing something non-medical. I guess we expect to be decieved by comercial products.

  31. pec says:

    [My point is that you have to be careful not to over apply the notion of what is “natural” for humans – that is where you get into the naturalistic fallacy.]

    I am not talking about going to extremes. We can’t possibly know exactly what our prehistoric ancestors ate, especially since it varied tremendously depending on the environment.

    But we can easily see that the current food industry is creating a lot of stuff that does not deserve the name “food.” If it has long chemical names on its ingredient list, we know our prehistoric ancestors never ate it.

    And the food industry can be very devious. I sometimes have trouble finding plain old natural yogurt, because there are so many “low fat” and “no fat” varieties. The average consumer will assume these are healthier and will pay more for them. But the “low fat, no fat” yogurt is actually much less healthy than regular yogurt, because it often contains corn starch and refined sugar. They take out the harmless fat and replace it with substances that, although “natural,” are probably more damaging to health than any other natural food.

  32. Amon1492 says:


    Thanks for the response. Your response concerning a “nocebo” is very near what I was questioning.

    Are the patients given a nocebo told by the physician that such a drug will do them harm? If not, where do they obtain the negative psychological bias toward the nocebo?

  33. daedalus2u says:

    Roy, I did get that your reference was a joke. The physiology of placebos is something near and dear to my heart, so to me it is no joking matter.

    The point I was trying to make was that in the absence of cooking, humans are unable to achieve good health. May foods can and are consumed uncooked. A diet consisting solely of uncooked foods is unhealthy and humans living in the wild could not sustain a society without cooking. That group would go extinct. The naturalistic fad of eating only uncooked food is completely non-physiologic for humans.

    There is no cooked food “in nature”. Humans have evolved such that they require cooked food to sustain themselves. It isn’t just jawbones that have evolved, our entire physiology has evolved. Cooked food is easier to digest, meaning less resources are needed to sustain a digestive system so that more can be devoted to supporting the gigantic brain that humans have.

    This completely disproves pec’s naturalistic fallacious reasoning. His latest example of yogurt is another. Milk products from other mammals can only be obtained after those mammals have been domesticated. Domestication of animals occurred fairly recently in human prehistory, perhaps 10-20k years ago? Only then could yogurt have been produced. What is “natural” about yogurt? Mammals lactate to provide milk to their offspring, who consume it as it is produced. Only stored milk ferments into yogurt, only non-human milk is stored. Plant starch is a more “natural” food for humans than is non-human milk fat. pec decided that milk fat is good and starch is bad, and then generated a wrong naturalistic fantasy to justify his wrong beliefs.

    Refined sugar is sugar that has had relatively small quantities of impurities removed. Unrefined sugar and refined sugar are more than 95% identical. The nutritive properties of sugar depend on the amount of sucrose present, not small quantities of ill defined impurities.

  34. jim says:

    I thought what was interesting about the wine study quoted above was that it was done with an MRI scanner. the taste parts of the brain reacted exactly the same to the “different” wines as they were same, but the pleasure centre of the brain reacted more to the more expensive wine.

  35. daedalus2u says:

    Amon1492, the study I was referring to was research where the “harm” was being told the inert substance would exacerbate nausea. This was within the bounds of the research study. It would be unethical for a physician to give a nocebo (as such) to a patient. However being told of potential adverse side effects of drugs may cause them to function as nocebos.

    Many anti-medication advocates amplify the seriousness of potential side effects and so cause them to act as nocebos. Being told that any treatment is useless and harmful will likely reduce its placebo effect and may cause it to have a nocebo effect. I think this is the main reason that many physicians don’t “bad mouth” CAM as much as is scientifically justified. To “bad mouth” CAM would be to destroy its placebo effect (the only mechanism by which it works). A physicians primary duty is to the welfare of his patient, not to “science based medicine”. A physician such as Dr Novella can denounce CAM as useless quackery “wholesale”, but if one of his particular patients used some CAM modality and perceived to derive great benefit from it, I think it would be unethical to “burst the patients bubble” and cause them to lose the placebo benefit they were deriving. How one walks along that very fine line is not something I know how to do.

    Quacks are under no ethical restraints, so driving people away even from useful treatments causes them no difficulty.

Leave a Reply