Sep 27 2011

Well, It Worked for Me!

Understanding the various aspects of the placebo effect is now a priority for proponents of science-based medicine. Now that for many modalities the evidence is in and is largely negative, proponents are exploiting the general lack of understanding of placebo effects to claim that their modality “works” as a placebo. Even skeptics may have a hard time understanding some of the counter-intuitive aspects of placebo effects.

Here, for example, is a question from reader PharmD28:

My question/issue is regarding debating someone about an intervention that has not been proven effective yet they clearly tell you that it is effective for them.

Take acupuncture. I was talking to a nurse practitioner colleague just today about acupuncture as one of the MD’s at my facility does it and has done it for her. She basically conceded that there is not strong evidence that it works, but she has had it done for her headache 5 times with “very good” results”. She said one of the five times it made her nautious, but this was “expected” for the first treatment and subsequently it eliminated her headache. She told me it did nothing for her back pain or for her TMJ. I have no explanation for as to why this intervention worked in this case for “headache”, but I find myself in that instance without good rebuttle for such, except to thinking to myself that yeah, placebo works too for subjective outcomes.

This is a common question – if a treatment “works”, even though it is just a placebo effect, isn’t that still a worthwhile effect.

The answer is – it depends, but mostly no. The problem is in the assumption that because one is feeling better the treatment worked. This is the post hoc ergo propter hoc logical fallacy. We do not know what the subject’s headaches would have been like had they not received acupuncture. It’s possible they were destined to improve in any case.

Simple regression to the mean explains why this is likely. People will tend to seek treatment when their symptoms are at their worst, which means they are likely, by chance alone, to regress to the mean of the distribution of symptoms – or return to a less severe state, which will be interpreted as improvement.

This is further compounded by confirmation bias. This case provides an excellent example of this. The acupuncture did not work for the back pain or TMJ (really TMJ syndrome, “TMJ” just stands for temporo-mandibular joint). So she tried acupuncture for a variety of symptoms, and one improved (while on one occasional developing nausea, which could have been a side effect or just a worsening of the headache).

In other words – we have very noisy data, with some improvement, some worsening, and some unchanged. It does not make scientific sense to pick out only the positive effects from this distribution of data and declare that the acupuncture “worked” in those instances.

This is exactly like having an alleged psychic guess cards, and perform no better than chance but declare that for those random hits they did make her psychic power was working. You have to look at all the data to see if there was an effect.

This principle applies to medical interventions as well – you have to look systematically at all the data to see if there is an effect. When you do this – acupuncture does not work. Saying “well it worked for me” is exactly like saying that the psychic powers worked whenever they randomly hit, even though the overall pattern was negative (consistent with random guessing).

Another layer of randomness to the data which is then ripe for cherry picking and confirmation bias is trying multiple therapies for the same problem (in addition to the same therapy for multiple problems). For example, someone might take medication, acupuncture, chiropractic, and homeopathic remedies at the same time for their headaches, and if they improve credit one or more of the alternative treatments. Or they may try them in sequence, and whichever one they took when their symptoms improved on their own gets the credit due to the post hoc fallacy.

We intuitively ignore the failed treatments – the misses – and commit the lottery fallacy by asking the wrong question: what are the odds of my headache getting better shortly after taking treatment X. But the real question is – what are the odds of my headache getting better at any time, and that I would have recently tried some treatment.

There are also psychological factors in play. When people try an unconventional treatment, perhaps out of desperation or just the hope for relief, they may feel vulnerable to criticism or a bit defensive for trying something unorthodox and even a bit bizarre. There is therefore a huge incentive to justify their decision by concluding that the treatment worked – to show all the skeptics that they were right all along.

Then, mixed in with all of this, is a genuine improvement in mood, and therefore symptoms, from the positive attention of the practitioner (if there is one – i.e. you’re not taking an over-the-counter remedy), or just from the hope that relief is on the way and the feeling that you are doing something about your health and your symptoms. This is a genuine, but non-specific, psychological effect of receiving treatment and taking steps to have some control over your situation.

What is distressing to those of us who are trying to promote science-based medicine is that this latter factor is often treated as if it is the entire placebo effect, or at least a majority. The evidence, however, suggests that it is an extreme minority of the effect.

A recent study with asthma, for example, shows that the placebo effect for objective measurements of asthma severity was essentially zero. There was a substantial effect for subjective outcomes. So subjects reported feeling better even when objective measures showed they were no better. This sounds an awful lot like confirmation bias and other psychological factor, like expense/risk justification and the optimism bias.


Placebo effects are largely an illusion of various well-known psychological factors and errors in perception, memory, and cognition – confirmation bias, regression to the mean, post-hoc fallacy, optimism bias, risk justification, suggestibility, expectation bias, and failure to account for multiple variables. There are also variable (depending on the symptoms being treated) and subjective effects from improved mood and outlook.

Concluding from all of this that a treatment “works”, when a treatment appears to be followed by improved symptoms, is like concluding that an alleged psychic’s power “works” whenever their random guessing hits. This is why anecdotal experience is as worthless in determining if a treatment works as is taking the subjective experience of a target of a cold reading in determining if a psychic’s power is genuine.

Yet, even for many skeptics, the latter is more intuitive than the former. It is hard to shake the sense that if someone feels better than the treatment must have “worked” in some way.

43 responses so far

43 Responses to “Well, It Worked for Me!”

  1. titmouseon 27 Sep 2011 at 10:19 am

    Comparing “it worked for me” to a psychic’s “hits” is very apt.

    I am going to shamelessly steal ur meme.

  2. banyanon 27 Sep 2011 at 10:30 am

    You didn’t steal the meme, it just reproduced onto you.

  3. nybgruson 27 Sep 2011 at 10:33 am

    How fortuitious for me that you post on the topic Dr. Novella. For a fair while now, spurred on by SBM and NeuroLogica and RI, I have been intensely interested in the placebo effect. And Dr. Hall posted on it over at SBM. I’ve posted up my thoughts there and a question and would love to hear from those over here their thoughts (and if you find the time, yours as well of course).

    The comments are here.

    Sorry for those that already read over there regularly but for those that don’t, I really am interesting in gaining more insight into this topic.

  4. Adam Slagellon 27 Sep 2011 at 10:51 am

    Guess it’s not only me that runs into this all the time. 🙂

  5. SARAon 27 Sep 2011 at 12:33 pm

    I have felt the psychological factor myself when reporting to my doctor the results of any given depression med. I feel this sense of expectation that is should work, and that I will somehow be letting everyone down if I report no help. Its a very subtle feeling and recognizing it makes me feel even more confused about my own sense of what I am actually feeling.

    My point is that any subjective self reported evidence is useless.
    And mental illness therapies are mostly based on self reported evidence.

    Nor have I found anyone who is willing to forego any “alternative med” treatment in which they have personally experienced the placebo effect.

    Even knowing that the flaw of the placebo is in the self reported process, doesn’t change the actual issue with the person who is convinced that they are feeling better.

    So, I have personally given up arguing against any therapy that is not actually detrimental to the person using it. If someone wants to skip the chemo and homeopathy, I’m all over it with everything I’ve got. If someone wants to get acupuncture, I skip the argument.
    Chiropractors are the hardest. They are very well established in the minds of many people as credible and useful therapists. The fact that they could make things worse is ignored and I have yet to convince anyone to skip an appt. A larger campaign against the well established credibility will have to happen before any one on one conversation is likely to have an effect.

  6. PharmD28on 27 Sep 2011 at 12:59 pm

    I feel like a celebrity 🙂

    As it relates to bias – the nurse practitioner, her collaborative provider is non-other than guess who…the MD that dabbles in accupuncture. I believe he also does a bit of the electro-stimulation (spelling?) version as well, which I understand really IS NOT “accupuncture”.

    I see this issue come up so darn much now it makes my head spin. Most of my newer pharmacy students are much more skeptical I think…but I have had conversations with a few here and there where it was obvious that they were much more willing to lapse and have another standard of evidence for some “alternative therapies”…even some colleagues of mine are a bit of “CAM apologists” to some degree…they hear me being harsh on CAM and their reaction is that “I am being too black and white”….or “science cannot figure out everything”…or “it worked for him/her”….”ever hear of the placebo effect? explain that?!”….

    This placebo effects (plural) conversation is so critical I think. As I have read through old posts on the SBM site on this topic (you can you to topics and see alot of the old articles on the SBM site) – I read the comments and find quite alot of interesting conversation about the role of placebo,
    ?physiological/neural/hormonal effects?, use in certain diseases such as IBS, fibromyalgia, etc…

    Most of it does not impress me very much though…I stand largely unconvinced that “the placebo effect” is meaningful in medicine besides recognizing it as an artifact that has little or no clinical utility….

    @SARA – that is exactly how I feel…I will say though that on fb I have had some success turning people away from chiros for pediatric stuff (like to avoid t-tubes)….I pissed off alot of opinionated parents, lol…in the end, my friend took my advice 😀 – score! Such victories are relatively rare compared to our pointing it out and it being disregarded though….once people decided for themselves, for them to take your advice would mean admitting that their original basis for taking (in most cases they assumed it to be an informed decision, perhaps even an evidence based one) was bull-crap….enter the ego….

  7. yogzototon 27 Sep 2011 at 1:29 pm

    I just wanted to add one key factor that heightens the cherry picking: Considering several outcomes for your condition, both objective and subjective.

    People rarely define their condition by only a single element but a bunch. Lending a thought from @SARA, consider depression: Depending on your situation, you may focus on your conscious mood, how you behave with friends and colleagues, your energy level, your sleep pattern, your appetite, your weight, your complexion, etc. Feel free to attribute any positive or negative change in any of the outcomes to a treatment “that works” – or not.

    It’s easy to see how likely it is to be misled by random variations – or to under- or overestimate true effects – in complex conditions.

  8. sonicon 27 Sep 2011 at 3:12 pm

    I don’t think you are being pedantic. I wish words were better defined. Often what seems like an argument (or agreement) is really two people saying different things using the same word.

    I think the body heals the body in most cases of injury or illness. I’m thinking of colds, flu, cuts and scrapes- even broken bones heal themselves (a proper setting will insure usefulness later).
    A placebo is useful to the extent that it encourages these mechanisms.
    So a person feeling hopeful and less anxious about a situation is brought on by brain changes that can also bring about changes in the body that can activate the body’s healing responses.

    But as Benedetti and others have shown different placebo can bring about different responses from different individuals.
    So if I were going to give a placebo to you, I’d make it an injection and I would make it sting a bit.
    Someone else might go for a ‘laying on of hands’. (Am I right- laying on of hands wouldn’t work for you, but an injection that stung might?)
    In any case– the suggestion/ritual would be valuable to the extent that the body response was appropriate for healing.
    And this is why “It works for me,” isn’t always as silly as it might sound.

  9. Steven Novellaon 27 Sep 2011 at 3:56 pm

    Sonic – but the evidence shows that placebo effects are only significant for subjective outcomes, not objective one, which suggests bias and illusion, not a real physiological effect. There is also no convincing evidence that psychological inputs can alter the bodies built in healing mechanisms.

    The exceptions are pain, for which psychological factors can have physiological modification (like endorphin release), and anxiety, to the extent that anxiety is an exacerbating factor for specific illnesses, like heart attacks.

    But outside of those specific situations, there does not appear to be any general physiological placebo effect. It is a mistake to generalize from the pain literature, and assume mechanisms that are likely not there.

  10. PharmD28on 27 Sep 2011 at 4:06 pm

    I heard this study brought up on NPR this year…

    Here is another “well, it worked for me!” example…although to be fair this was a study in 80 IBS patients…interesting that their knowledge that it was placebo still resulted in beneficial effect….Its very funny that when “she ran out of placebo” she then went and bought some more placebo (I mean herbal pills) at a health foods store….kinda ironic lol.

    I wonder if we had 80 skeptics with IBS do the same thing would it matter? I refuse (or at least am very skeptical) to believe that if I had IBS or back pain that taking a placebo would do jack or that taking a placebo with my knowledge would reduce my stress/pain? I think you would have to be open to mystical possibilities of placebo in order for the ritual to work much (as you take it you are saying, “well, maybe it will work, some way some how, who knows”)? Lets do a study of the response to placebo looking at say atheists vs. those with significantly more religiosity. That would be interesting….but perhaps a placebo effect would still be apparent….who knows…my ramblings….food for thought.

  11. sonicon 27 Sep 2011 at 4:35 pm

    Dr. N–
    According to Benedetti these effects are also seen for the endocrine system, the immune system and the cardiovascular system as well. He suggests that the notion that it’s all about pain is incorrect. (I’m reading page 17 of the transcript–)

    Am I misreading?

  12. sonicon 27 Sep 2011 at 4:47 pm

    Such studies have been done- here is an article about one example-

    I don’t know if a follow up on this was done, but similar studies indicate that there are a number of factors that go into who would like what type of placebo.

    In your case I’m guessing if the placebo gave you slight nausea it would work better than otherwise. Just a guess. (Google ‘active placebo’ for more on that).

    Personally it would be all right with me if I could be healed of any and all diseases with a passing of a wand over my head. Better yet– a pretty girl’s kiss is the cure for any and all disease I ever get.
    But then I’d just want to get sick again… hmmm 😉

  13. PharmD28on 27 Sep 2011 at 5:03 pm


    “But then I’d just want to get sick again”

    daily prophylaxis then would be in order…dare I say Q4 Hours ATC and PRN 😉

  14. bob_plotkinon 27 Sep 2011 at 5:28 pm

    An aspect of placebo effect that I do not see talked about is that the effect only occurs when the patient believes that that the treatment works. I am amused when I see suggestions that one should use the placebo effect as the main course of treatment.

    If a treatment were advertised as such, (“sticking needles in you has no known medical benefit, but you should fell better regardless”) – then I am not sure there would be a placebo effect.

    There is only a placebo effect because there is a claim (regardless of true efficacy) that a procedure works. Without these claims – false or not – there would be NO placebo effect!

    I suspect if you ran a study with “a new form of” acupuncture designed to cause “mild discomfort” and followed the identical protocol for healing methods, the numbers would reflect a similar negative placebo effect.

  15. Steven Novellaon 27 Sep 2011 at 9:11 pm

    sonic – those all relate to stress hormones. Reducing stress by relieving anxiety over being ill or feeling helpless reduces stress hormone levels. The clinical effect of this depends on what outcome you are interested in. Does not appear to affect cancer survival, asthma, any most serious illnesses. Appears to affect cardiac survival (very sensitive to stress levels).

    I think what has happened is basic science studies looking at biochemical markers find that placebo effects are actually doing something biological (increasing endorphins, reducing stress hormones and all the downstream effects of this) and some people see this as a shocking result with wide ranging implications.

    I see it as a predictable result with narrow implications.

    When you look at the clinical literature you find that, except for those outcomes that directly respond to endorphins or stress hormones (essentially mental effects, which does include when the mental effects are the end points themselves, like mood) the placebo effect is almost entirely an illusion.

    When you look at placebo effects for mental states and their downstream effects, they are modest, short lived, and inconsistent – but real.

    This is hardly a revolution in medicine, as health care providers figured out centuries ago the benefits of bed-side manner.

  16. nybgruson 27 Sep 2011 at 9:11 pm

    Thank you all for the commentary. I have been commenting over at the SBM thread and don’t feel the need and question the propriety of just copying and pasting everything over here.

    But to address a few specific points here:

    @Adam: Fabulous post. Well said.

    @SARA and PharmD: Yes, there is a cost benefit we must do in individual cases and sometimes it just doesn’t pan out. As I’ve said before, despite my firm stance and sometimes fierce rhetoric I would never dream of slapping the echinacea out of a patient’s hand and calling him/her an idiot.

    @yogzotot: Yes, I think you are quite right. If I am sick and have stomach pains, myalgias, headache, a runny nose, and a cough and I take something for it and my headache and runny nose abate, I would undoubtedly consider that “getting better.” With a bit of post hoc ergo propter hoc combined with recall bias later on, it would be easy enough to attribute those minor and incomplete improvement to echinacea instead of regression to mean or natural fluctuation and then feel confident saying “Echinacea “worked” for my flu so I’ll take it again.”

    @sonic: Thanks for not thinking I’m being pedantic. We often don’t agree, but unlike some of the other posters (ahem, trolls) who have come through here I still enjoy chatting with you.

    I would disagree with a point you make though:

    A placebo is useful to the extent that it encourages these mechanisms.

    From all my reading on the topic the placebo response cannot and does not encourage self-healing mechanisms. In certain specific cases the placebo effect can allow for processes inhibiting those mechanisms to be removed (i.e. stress response and peptic ulcer disease). Perhaps a minor quibble and not at all what you meant, but I am a bit on the pendantic side, after all 🙂 I would be interested to see a placebo study on something that could demonstrate the negative aspects of stress relief (since I posit that as the primary mechanism for the few objective changes we can find in psychogenic placebo effects). I’m hard pressed at the moment to think of a good example, but perhaps something like starvation states would be useful. Monitor blood glucose levels in subjects being starved for 24-72+ hours. Placebo effects, if possible to induce, should diminish catecholamine levels and counteregulatory hormones, thus decreasing the ability for gluconeogenesis and causing a drop in BGL (which would be bad in a starvation state). I’m not sure that would be a good study but it was the only thing that could really come to my head at the moment.

    So if I were going to give a placebo to you, I’d make it an injection and I would make it sting a bit.
    Someone else might go for a ‘laying on of hands’. (Am I right- laying on of hands wouldn’t work for you, but an injection that stung might?)
    In any case– the suggestion/ritual would be valuable to the extent that the body response was appropriate for healing.

    Not unreasonable assumptions, but I would venture to say incorrect in my case. I would want to know what is going on in both cases and since I full well know that more invasive procedures are supposed to elicit a stronger placebo response I think I’d be more or less immune to such things. At least, I’d like to think so.

    And this is why “It works for me,” isn’t always as silly as it might sound.

    Read Adam’s blog post linked above. He makes a very good case for why “It works for me” is indeed always silly.

    As for the endocrine effects you are seeing re: Benedetti I dont think you are misreading. But you must realize that not only is the endocrine system directly innervated by autonomic nerves, but that the pituitary also control endocrine function. Thus, these neurophysiologic changes can lead to outcomes there and thus have downstream effects on the endocrine system. But there is no direct correlation for disease processes there, which is what Dr. Novella was referring to.

    And yes, a pretty girl’s kiss pretty much always makes things better for me 😀

  17. nybgruson 27 Sep 2011 at 9:16 pm

    and it seems while I was writing, Dr. Novella answered sonic’s question saying pretty much what I was going for but in more detail. In re-reading what I wrote, I was not clear and closed off the downstream effects of disease processes that would be affected by such hormonal influences – obviously cardiac disease is very much dependent on blood pressure and vessel lumen diameter, both of which are exquisitely sensitive to catecholamines, which are directly inlfuenced by neurophysiologic changes. But osteoarthritis is not, and thus there is nothing the placebo effect can do for that, except mitigate pain perception (but not actual nociception).

    So indeed, this is not revolutionary at all.

  18. tmac57on 27 Sep 2011 at 9:48 pm

    nybgrus,did you listen to or read the interview with DR. Benedetti?I came away from it thinking that either I have a poor understanding of what placebos do,or maybe his research is off base. His findings on Parkinson’s seems to indicate that their are objective improvements to motor function and dopamine response in the brain due to placebo. He did say that he didn’t see placebos as being useful for clinical application though.

  19. nybgruson 27 Sep 2011 at 10:07 pm

    actually I haven’t yet. I have to go to the morgue this afternoon to observe an autopsy, and the bus ride out is reasonably long so I was going to listen to it then. I’m happy to provide my thoughts on it after I return.

    As for the objective improvements in motor function with Parkinson’s, that doesn’t surprise me in the slightest. Parkinson’s is a depletion of the dopamine generating neurons in the substantia nigra affecting the ability for intentional movement – hence the mask-like facies and paucity of movement and festinating gait. Parkinson’s treatments are all based around increasing the amount of dopamine available to bring back function. The psychogenic portion of the placebo response (I’m trying out that phrase for size, based on the conversations I’ve been having between here and SBM) could very reasonably up-regulate the stimulation of the dopaminergic neurons, either directly through relevant neural pathway restructuring or indirectly via a global increase in dopamine release via reward pathways actived by the psychogenic placebo response (not sure which, but either seems plausible to me). Thus, there would be objective improvements in movement – but not from placebo, from the psychogenic component down tangible and specific pathways.

    Does that make sense to you tmac or am I not getting my point across well? I ask because I am trying to hone in on accurate and clear descriptions of placebo.

  20. tmac57on 27 Sep 2011 at 10:22 pm

    It’s a bit over my head but I get the gist of it. Apparently it is commonly believed that the placebo response is only effective for subjective things such as pain,range of motion,and psychological complaints and such. Dr. Benedetti’s research seems to go beyond those types of things.(And I did realize that it was not the placebo itself that was causing the response,I just worded my question sloppily.)
    Thanks for the input.It’s a good interview if you can get through the heavy accent that the good Dr. has 🙂

  21. nybgruson 27 Sep 2011 at 10:50 pm

    That’s not quite what I was going for. I have to run to the morgue now, but I will respond and clarify after I get back. I’ll also have listened to the Benedetti podcast by then.

    Ciao for now!

  22. nybgruson 28 Sep 2011 at 3:39 am

    Well, back from the morgue (insert dark humor here).

    I listened to Benedetti’s interview and found it very interesting. Honestly, it was pretty much exactly what I have thought and been saying all along – though with a bit more refinement.

    The only real difference is the nomenclature I have been using. He likes to refer to the placebo effect as the real effects and the other biases, reporting error, and artifact as NOT the placebo effect. I have been calling what he calls placebo effect the “psychogenic placebo effect” and calling the placebo effect all those things he says are not the placebo effect. I think this is because of our different frame of view. He even states that placebo effect is fundamentally different from a clinical vs neuroscience POV. So I guess I am still not fully resolved on the nomenclature here, but at least I am edified in my stance and understanding of the issue.

    One thing that I wish he’d gotten more into and that I don’t know enough about is his commentary on the psychogenic placebo effect on immune response and cytokines. From what I know, that can only be indirectly via cortisol from adrenal stimulation so I am not sure if there is something I don’t know if that is all he is referring to.


    To clarify the whole Parkinson’s thing (and this is probably going to get long, so bear with me):

    Deep in the brain there are nuclei (clusters of neurons) that interact with each other to allow for movement to happen. There are actually two circuits and both of them use dopamine. In one circuit, when dopamine comes in it stops movement from happening. In the other one, when dopamine comes in, it stops the circuit that stops the movement from happening. In other words, it lets the movement happen (like stopping the guy from lowering the dam across the river lets water flow). When you want to move, your cortex (higher brain) send a signal down to these deep parts (called the basal ganglia). If the circuits are stuffed up then you get problems. In Parkinson’s since it is the double negative circuit you get less movement. Huntingtons on the other hand affects the single negative circuit, so then you get constant moving (the chorea or “dancing” of Huntingtons).

    In Parkinsons, the part of the brain that is the double negative (stops the stopping) is affected. The cells in there that release dopamine die off, so you don’t have enough to adequately stop the stopping and so you get decreased movement. When treating Parkinsons we give medicines that either make your cells produce more dopamine, make the dopamine hang around longer, or directly stimulate the dopamine receptors (so substituting dopamine, basically). That gives more dopamine effect, so that double negative circuit can work better and you can have improved movement.

    When you utilize a psychogenic placebo, what you are doing is stimulating the reward centers of your brain (one of the four defined mechanisms Benedetti talks about). The reward center includes the nucleus accumbens which shoots of tons of dopamine. That stimulates things downstream and ends up causing those cells in the Parkinson’s circuit to release more dopamine themselves, which ends up having the same effect as giving the active drug.

    So to me, that all not only makes sense but I expect that to happen. Benedetti talks about how the research is pretty knew that the psychogenic placebo effects change deep in the brain and at a single neuron level. That is pretty new, and very important to have verified, but once again not even remotely surprising to me.

    The brain is like a gigantic massively parallel binary computer chip. The deep structures interact with each other constantly well below the level of our consciousness. But these parts obviously have to communicate with each other somehow. So having a cortical level input (psychogenic placebo) has a clearly defined mechanism for affected deep brain structures.

    This also ties in with the classical conditioning mechanism that Benedetti was talking about. The way you learn new things is by making new connections in that massively parallel chip in your head. The more times you do something, the stronger and more numerous those new connections get (that’s called Hebbian learning). So if you give a drug that has a molecular effect in the brain at a deep level, it will begin to induce Hebbian learning and literally physically remodel the neural circuits in your brain. Now, you are also giving this drugs in a context – the patient knows you are giving a drug. So now, after you have made this new deep neural pathway AND created the cortical connection to it via the drug giving ritual, when you give the placebo the cortical pathway will active the newly formed deep pathway and elicit the same response. That is why if you just start out with placebo you can’t get that deep response – there is no pathway there for it to work (this is in reference to the growth hormone (GH) study he talks about). It is also why if you administer that same active drug but the patient has no idea, and then you give a placebo you will get no response. The pathway is there, but the cortical link to it isn’t so you can’t activate it via that mechanism.

    So things that can have some pathway from the cortical input to the final outcome can indeed be amenable to psychogenic placebo. The ulcer one is a great example of a really long pathway but one that exists nonetheless. I’d be interested to know how much the response diminishes the further it gets away from the initial cortical stimulation. This is also why Benedetti notes that conscious psychogenic placebo can only work on conscious physiological outcomes (anxiety, stress, cortisol, pain, reward…) and unconcious pyschogenic placebo can only work on unconcious outcomes (growth hormone release, etc) unless there is a handy bridge already in place or classical conditioning is used to create one. Hence the Parkinsons effetcs via the reward system pathway in the limbic system. The limbic system (emotional centers) is a deep and very primitive part of our brain and it interfaces with everything and uses dopamine so I think you can see how that connects together quite logically.

    So yeah, it is commonly believed that the placebo is an inert thing that can only elicit subjective improvements (BTW, range of motion is not necessarily subjective and is often quite objective). So when we find psychogenic placebo effects that have actual mechanisms and actually change physiology, it is easy to conflate the two and thus you get CAMsters claiming their woo “works” because placebo is an active and effective treatment.

    But, as was said in the podcast as well, placebo can only effect changes for a very specific type of pathology. So things like cancer, infection, trauma, asthma, etc are not amenable to placebo effects except tangentially and with small effect sizes via stress reduction and cortisol levels, etc.

    Lastly I’ll add that the whole single neuron stuff just makes sense – any brain response is from an aggregate of neurons which means at some point a single neuron must be firing. His study showing that isn’t revolutionary or even all that dramatic, IMO, but very important since it demonstrates strongly that the psychogenic placebo response is mediated via neural pathway restructuring and Hebbian learning processes.

    I know that was long winded, but hopefully that gave a pretty clear explanation for you Tmac57. I apologize that my earlier more brief one was so full of technojargon. I don’t know what your background is and I assumed you were hip to the lingo. Please let me know if there is anything I can clarify. Part of why I like writing on these fora is that I get to practice my communication skills and ways of explaining things to people of different backgrounds – something I believe is vitally useful as a physician.

  23. PharmD28on 28 Sep 2011 at 9:40 am

    @nybgrus – yeah, I liked the “long winded” summary you have there. I read the transcript last night…and it did not really change my view…although it did raise my level of consciousness a bit about how complex the brain is.

    This was the most interesting and telling part for me of the interview from Benedetti:

    “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. This is a very important point—a translational research: can we use placebo in routine clinical practice? Well, sometimes it works; but that’s not the important problem. The important thing right now is to understand how our brain works.

    And I would say the
    placebo response is a fascinating phenomenon, because it is a sort of melting pot of concepts, of ideas for neuroscience. So, if you use a placebo response, you can understand a lot of brain functions, like anxiety, like social learning, classical conditioning, reward mechanisms, and so forth. So, the clinical application is—I
    think in English you say it is a ‘different kettle of fish.’”

  24. tmac57on 28 Sep 2011 at 10:36 am

    nybgrus- Thanks so much for taking the time to write such a clear explanation. I appreciate the clear detail,because I am just an interested layman with no medical background. I have always had an interest in medical science,and try to get at least a general understanding of where the current state of knowledge is,and try to keep my friends and family informed.Most of them do not spend much time reading about science,and tend to fall prey to misinformation,so I do quite a bit of “Well actually the current research says…” kind of thing. So thanks again,your communication
    skills are excellent!

  25. locutusbrgon 28 Sep 2011 at 10:52 am

    After all of the highly descriptive comments here I feel a little intimidated.
    I would say that the “Placebo Effect in Anxiety”, certainly has plausibility and has some scientific support.
    I think it is actually maladaptive for anxiety. Like the OCD sufferer who temporarily relieves his anxiety by hand washing. In the long term it fails to relieve the anxiety and slowly worsens the condition. Allowing a patient to continue to utilize a placebo for a medical condition, worsened by anxiety, is like handing a OCD hand-washer a bottle of Purell. In my opinion, a waste of time even though it might have clinical effect.

  26. sonicon 28 Sep 2011 at 12:39 pm

    Dr. N-
    I don’t find it surprising that placebo effects are actually doing something biological. I agree that some seem to find that amazing and have gotten all excited and are overstating the importance of that.
    I also think that the implications are limited- perhaps less limited than you, but placebo is obviously not a cure-all.
    What I note is that the interest and amount of study and information on this topic is growing rapidly.
    I think there may well be more surprises ahead in this area–

  27. sonicon 28 Sep 2011 at 12:40 pm

    It turns out that prophylaxis might not work in this case.
    I think many more years of study are needed to be sure though… 🙂

    Mr. Pedantic– who knew?
    I read your “long winded” explanation and I agree with much of what you are saying. I’m not sure the fact that only certain pathologies are effected is that big a problem. You could say the same about antibiotics. Yet we consider them important…
    I’m guessing that what you know about the pretty girl’s kiss is on the order of “It works for me.”
    But that just proves it is a silly idea, huh? 🙂

  28. PharmD28on 28 Sep 2011 at 1:05 pm

    “I’m not sure the fact that only certain pathologies are affected is that big a problem. You could say the same about antibiotics. Yet we consider them important…”

    I do not want to misunderstand you’re point – clarify please? That statement simply does not compute in my mind.

  29. sonicon 28 Sep 2011 at 1:44 pm

    nybrus said-
    “But, as was said in the podcast as well, placebo can only effect changes for a very specific type of pathology. So things like cancer, infection, trauma, asthma, etc are not amenable to placebo effects except tangentially and with small effect sizes via stress reduction and cortisol levels, etc.”

    I’d agree that these placebo effects are limited and mostly important in certain situations. But that is true of antibiotics as well.

    I was thinking nybrus was minimizing the importance of the effect by pointing out it has limits. Perhaps I misunderstood nybrus.

  30. PharmD28on 28 Sep 2011 at 2:16 pm

    “But that is true of antibiotics as well” – that “the effect of antibiotics are limited and mostly important in certain situations”?

    Thats the part I am unclear about what you are saying….here. Yes antibiotics are useful in targeted situations sorta – but levofloxacin (for example) can be used for how many types of infection?, and the effect is VERY robust as compared to what would amount to “clinically” an irrelevant effect on infection by any sort of “placebo effects”.

    “I was thinking nybrus was minimizing the importance of the effect by pointing out it has limits.”

    Personally I am all for minimizing the importance of the placebo at the clinical practice level…there is little there anyway to require minimizing….one only need build up the limited scope of the data in this arena in order to feel the need to minimize it….at least that is how I see it. I think that Dr. benedetti, Harriet Hall, and Dr. Novella all would agree with this general point….that on the clinical level “placebo does not work”.

  31. nybgruson 28 Sep 2011 at 7:47 pm

    @Tmac57: You are very welcome. I am all about education, especially patient and lay-person education. It empowers people and always leads to better outcomes and a better therapeutic relationship. I hope to keep that going as much as possible in my future practice of medicine.

    @locutusbrg: (BTW, love your handle) – I agree. That is a good reason why the psychogenic placebo effects cannot and should be used in isolation. CAM is indeed often employing psychogenic placebo effects, but that is all they are offering. So I like your analogy – CAM use for a pathology is like giving an OCDer a bottle of purell. Just as a physician would never just give them the purell and say they were done, so should a physician never prescribe a CAM.

    @sonic and PharmD:

    PharmD I actually disagree with you slightly. Sonic misinterpreted me – I am not trying to downplay the role of psychogenic placebo effects by saying they are only narrowly applicable. While not really clear I think his analogy to antibiotics is reasonably apt. We need to recognize those places where psychogenic placebo effects will be most useful and those where it wouldn’t be useful at all. However, the part I think Sonic missed and you got right is that at a clinical level the distinction becomes moot (and this is where his abx analogy breaks down).

    As health care providers we must always use the therapeutic ritual to our advantage in all settings. In some the psychogenic placebo effects will come into play and have a measurable and useful objective improvement in the disease process itself. In areas where it wouldn’t there would still be a better quality of life and a stronger and better therapeutic relationship which, IMO, would lead to better outcomes down the road because the patient would be better empowered and feel better about being more active in their own health maintainence.

    So in the context of CAM – yes, the placebo effects are worthless because that is the entirety of what CAM has to offer and the “it worked for me” argument is self deluded and not useful for those same reasons. But in the context of actual medical practice, the utility of the psychogenic placebo effects is quite important. It is a small effect size to be sure, but consistently practiced over large populations that adds up to big positive changes in health outcomes.

    I guess what I am saying, which we all already knew (or should have at least), is that bedside manner is important. And now we know better exactly why that is and also can use that to change and add ways to interact with patients to maximize the utility of it. As Dr. Campbell said in the podcast – it is important to tell the patient they are receiving a pain med and what it will do. And now we have an understanding of the neurophysiology of why that is so.

  32. nybgruson 28 Sep 2011 at 7:48 pm

    whoops that’s “cannot and should NOT be used…..” but I hope that was pretty obvious

  33. sonicon 28 Sep 2011 at 8:03 pm

    An analogy– like antibiotics the placebo effects are limited to usefulness.
    Nothing more- nothing less.
    Like all analogies it is flawed. Not only that but in this case I’m putting antibiotic in the same sentence as placebo and this could be seen as making placebo more important by association. Perhaps the association is distracting as it is question begging.

    What Dr. Benedetti said (page 25 of transcript)–
    “If you give a hidden injection of morphine—which means that the patient does
    not know that anything is being given; this means that the patient does not expect
    anything—the effect of morphine is reduced….
    So, we have to enhance the expectations of the patients in order to enhance the effect of the therapy—the treatments that we give in routine medical practice. ”

    Sounds like he advocates using the placebo effects to advantage.

    Dr. N. talks of ‘bed-side manner’. That is another way of saying reassuring the patient and encouraging the patient and other actions known to involve placebo responses.

    it happens if you want it to or not– might as well use it to advantage…

  34. sonicon 28 Sep 2011 at 8:05 pm

    I shoulda read yours first.
    I think you are making a good distinction– placebo only versus placebo to enhance actual medical intervention.
    Very good distinction.

  35. nybgruson 28 Sep 2011 at 9:02 pm

    indeed. I think the most fundamental mistake made by so called “holistic” types is that the reductionism of modern scientific medicine means that we will act on just those reduced factors in isolation. I think this is probably reinforced by some who do in fact do this. I would call those people bad doctors.

    But reducing things down to better understand how they work is fundamentally different from divorcing them out of the treatment of a patient. That is why I hate that old adage that people seem to think is so profound “Treat the patient, not the lab value.” No! You must always treat the patient AND the lab value. The sCAMsters are claiming we divorce the two and ignore the patient. But they are at the same time divorcing the two and ignoring the medicine. Neither way is right.

    So yes, use those psychogenic placebo effects in concert with actual medicine and yield much better results. I have seen this in my own work in the ER. Talking to patients, explaining how and why something will work, giving them an expectation of benefit, and being kind and empowering them with knowledge is so vastly useful that I can’t imagine treating patients without doing so.

    An anecdote from my experience I love to relate:

    Last year I did a month long surgical elective. Mostly general surg but I covered trauma call and did some ortho as well. In this case I was doing call and we got called over to evaluate a pt with abdo pain as to whether she needed her gallbladder removed. The relevant studies all demonstrated clearly that was not the cause of her pain and that it was in fact a gastroparesis (her stomach just wasn’t churning and moving like it should). The surgeon came in, briefly told the pt and family that and called it a day. The pt and family were in distress – the pt was getting intense waves of pain every 10-15 minutes and couldn’t understand why it wasn’t the gallbladder or what it all really meant. While the surgeon was dictating, the mother of the pt came over and snagged me and asked me if she could ask the surgeon a question. I said he was busy at the moment, but if she wanted to hang out for 5 minutes I would flag him down after he was done. In the meantime, I was happy to answer whatever I could, making it very clear that I was just a medical student and not a doctor. She asked if the gallbladder tests could be wrong. I explained to her how the test worked and why they couldn’t be. She then asked why her daughter got pain every 10-15 minutes. I then explained the migratory motor complex and how it causes the intestines to contract in a wave from top to bottom every 10-15 mins as a way of keeping our guts clean. I also explained that guts can’t feel pain like the rest of our body – they only hurt when the get stretched. So since her stomach was paralyzed the MMC caused that pain. I explained it in reasonably basic terms, but I did use the technical jargon from time to time, explaining what it was each time. After just 5 minutes, she was very relieved and said she not only understood why the results were what they were and why her daughter was having that pain, but also thanked me for taking the time and not treating her like some stupid person who wouldn’t be able to understand.

    I think that most people not only can but want to hear some technojargon and have it explained to them to really understand what is going on. But we are taught in med school never to use big words around patients and to make matters worse many doctors don’t know how to describe something without using our latin based obfuscatory language that is pounded into our heads all through med school. So they are left with no middle ground – either really technical language or so basic as to not really explain much and make the patient feel like we are talking down to them.

    The middle ground is hard to find, but I think very important and I advocate for it at every single chance I get. I’ve even had my clinical comm skills tutors chastise me for using big words in role plays (yes, we do those) and I’ve stopped the session and explained clearly my rationale and advocated people follow suit. Writing constantly on these fora helps me develop those skills (as I’d said to Tmac57) as does leading my weekly tutorials for 1st year students. Teaching effectively is a skill that must be practiced – not a gift that some people just don’t have.

    One thing I do need to work on though, is being less long winded! haha

  36. PharmD28on 28 Sep 2011 at 10:14 pm

    Very good points!

    I work in primary care, and the pharmacy residents and students that cycle by me…I have to work very hard to get them to stop from saying “lipid”, “hypertension”, “any change in your vitamin K intake?, “hypoglycemia”, and other such things….

    Sometimes you really do wonder how much of what we say, really sinks at all….

    I make it a very very important point to speak in the simplest of terms in every respect within my practice, and I am glad to hear others strive to do the same!

    One thing I definitely value about this placebo discussion is the whole bedside manner, therapeutic ritual piece. For the psychogenic aspect perhaps in its smaller way, but also just that how many times do we see COPD’ers having an exacerbation because they mistakingly take their scheduled dry powder inhalers PRN….or how many seizure patients have a seizure because they are non-compliant…etc..etc…etc…that therapeutic ritual is quite important…if it is one thing the CAM community does criticize us right in one aspect it is that we are failing in this regard many times – although their criticism is from an ivory tower, not in the trenches of the difficult realities of modern medicine.

  37. PharmD28on 28 Sep 2011 at 10:25 pm

    “Sounds like he advocates using the placebo effects to advantage”

    yes, to his advantage for piloting new ideas for research purposes, and “understanding brain pathways”…but as he said in the interview….the utility in actual clinical practice is far from clear at this point…this finding, while interesting (especially the GH part, boggled my mind) is still a far cry from useful in clinical practice I believe….but I DEFINITELY agree that clearly the bedside manner and all that jazz is important as part of “the placebo effect”….and I do find some of this stuff interesting on this basic science level…but on the clinical level I will have to see much more….

    You all rock on here…learn something new every day and new perspectives – lovin’ it 😀 – good night.

  38. BillyJoe7on 29 Sep 2011 at 6:18 am


    Thanks for the low down on the podcast.
    (I still haven’t had time to listen to it – I’ve just finished training for and completing my annual 50 km walk over the Dandenongs, and I’m now in training for the 14km City2Sea run in 6 weeks time).

    I’ve listened many times to Ben Goldacre expounding on placebos but he always leaves me feeling confused. The problem, I now realise, is that he doesn’t ever get around to explaining the underlying neurophysiology so that all the examples he gives make sense.
    So thanks for filling in the vacuum.

    Regarding bedside manner:
    I do not have any medical problems and therefore rarely attend for medical advice. However I recently had to attend a dentist urgently because one of my front teeth suddenly broke off close to the gum margin leaving a gaping hole in my smile, which subsequently disappeared. Because of my previous alienating experiences with dentists (standoffish and rude), I hadn’t seen a dentist in about fifteen years. However, my experience with this particular dentist was very different. She actually made eye contact and smiled! She seemed genuinely interested in me as a person with a broken tooth rather than just as a case of a broken tooth. And she explained everything she was doing – the topical anaesthetic to numb the pain of the injected anaesthetic (which I would still feel a little bit); and the slightly more painful injection into the hard palate etc etc. I was so impressed that I decided to let her correct fifteen years of dental neglect. She made me feel so at ease with her helpful explanations and her reassurances that I actually look forward to attending my appointments!

    She was also the only dentist I’ve ever attended who seemed to actually enjoy her work.
    I have a feeling you’re going to be just like her.

  39. nybgruson 29 Sep 2011 at 6:49 am

    First off, thanks for the kind words billyjoe. And I am glad I could lend some insight to the mechanisms – I found them interesting myself. And congrats on the fitness goals. Last year I actually had more time and spent a solid 12 hours per weak doing intense training including a weekly 18km run. I used to have a 6-pack. Now, not so much 🙁

    And I’m glad you liked your dentist. It really makes such a big difference when they seem like they enjoy what they are doing. My GF was never a huge fan, until I hooked her up with my dentist. He is so friendly and actually asks you about how your day has been, anything new since last time, etc. It makes a huge difference.

    I find that when I speak with colleagues I refer to patients as their condition i.e. “the appy in 4” – but when I am actually with a patient, that is quite different. My step father, who is a critical care doc and someone whom I admire told me his philosophy:

    “There’s two kinds of doctors. Those who know what their patients do for a living, and those who don’t.”

    I aim to be the former.

    BTW, I often explain the lidocaine by saying it makes all the nerves fire once and then get frozen, which is why it burns and stings at first for just a little bit and then it goes completely numb. I find that is particularly helpful with children. They really have much more capacity to understand than we tend to give credit for.

  40. tmac57on 29 Sep 2011 at 12:17 pm


    …I hadn’t seen a dentist in about fifteen years.

    You know what they say:
    “Your teeth are the only problem that will go away if you ignore them.”

  41. DLCon 30 Sep 2011 at 2:14 am

    Much of “herbalism” is placebo effect. The Herbalist tells you “take this stuff, it’ll make you feel good” and so, you take the stuff and, if it doesn’t kill you, you probably regress to the mean and feel somewhat better.

  42. reedonlyon 01 Oct 2011 at 9:06 am

    It’s time to hit at the root of the problem. Stop calling it the “placebo effect,” because effect implies effectiveness. Called it what it is – the placebo illusion.

    Just because the magician appeared to make the elephant vanish into thin air doesn’t mean the elephant actually vanished into thin air. Just because a modality appears to work doesn’t mean it actually works.

    Science, or an understanding of how magicians work, can help find the elephant.

  43. nybgruson 01 Oct 2011 at 5:15 pm


    That is precisely the distinction I am trying to make very clear and easy. There is and actual, neurophysiological, biochemical change that occurs during placebo administrations. There are defined mechanisms for it that have been demonstrated empirically. And they do make for real changes in certain disease processes. However, IMO, that is not a placebo effect – that is the effect of the active psychological intervention going on concurrently with placebo administration (active in the same way that CBT is an active therapy). That is why I have been trying to push for using “psychogenic placebo effects” to refer to those real neurophysiological changes and “placebo effect” to refer to the illusion of change via reporting error, bias, regression to the mean, etc.

Trackback URI | Comments RSS

Leave a Reply

You must be logged in to post a comment.