Don't Believe Everything You Think

A look at the placebo effect in health

Placebo is the “most effective medication known to science, subjected to more clinical trials than any other medicament yet nearly always does better than anticipated. The range of susceptible conditions appears to be limitless.” (from Ernst 1995)

I’ve been absolutely enthralled by the placebo effect for as long as I can remember, and I am surely not alone. It may be one of the most nuanced and fascinating parts of the human condition. There are countless stories of the placebo effect and most people probably have their own. Beyond amazement though, beneath that, there is a much to learn and even more to be conflicted about.

I remember early in some of my clinical practice noticing that some of my colleagues were doing a more advanced and thorough job than the boss, but the boss was having similar (if not better) results and patients’ perception of the boss’ work was much higher. Similarly I DISTINCTLY remember the boss using the phrase; “that’ll work great” when giving interventions to patients. To me this was madness after all, how could the boss know? Well it turned out the success was more common than not (a mental tally of mine was >90%), so I tried it. To a mixture of my dismay and surprise, it worked!
(Note:*I’ve intentionally altered some of the details in this story for privacy reasons)

Some readers are probably lost as to my emotional conflict, but this is the challenge of placebo in the heath setting: the overlay is ethics and how you see your role. Does the outcome justify the means (a utilitarian approach)? Or is complete honesty and complete, strict, adherence to guidelines as dictated by the research the obligation? To be clear, I don’t know if there’s a good answer, I certainly spent enough time in medical ethics classes and thinking outside of them to argue the point from both sides in a compelling way (I will spare you).

People Generally Misunderstand the Placebo Effect

When we are first introduced to the placebo effect it tends to be as “a sugar pill”. We are told in drug trials people are given sugar pills and the effect of the drug in question is the improvement beyond that of the sugar pill. As an introduction this is appropriately basic and informative but it doesn’t take long in contemplation or research to realise there’s much more to what people call “The Placebo Effect.” As it is commonly understood and used, “The Placebo Effect” has come to mean any effect that isn’t ‘real’. This is quite problematic and in my opinion, in much the same way that thinking of the mind as distinct from the body is problematic (as I covered here).

Readers may appreciate, some of the so called ‘placebo’ effects may indeed be what’s actually called the ‘natural history of the disease’ aka it gets better without intervention (see figure below). So, ideally trials of interventions would have ‘non-intervention’ controls, not just placebo controls. In doing so, though, we end up with swiftly escalating complexity not to mention costs, and research is already tough and expensive enough as it is (I wrote a little about some of this and why personal experiments may be the future here). In the performance world, this may be changes not associated to the intervention, growing teenagers come to mind (maturation tends to improve certain things like force outputs but perhaps impairs some aspects of coordination for a period, thus impacting skill).

From Hafliðadóttir et al

Beyond this though the catch all term of ‘Placebo’ actually encompasses much nuance and aspects that can (and thus should, for clarity and understanding) be named.

Interestingly, if examining these with an open mind they can teach us a lot about improving healthcare (on a related note here’s my piece on a healthier hospitals) and to be honest, this applies to a coaching/performance setting too. Looking at the below diagram, the psychosocial factors factors specifically speak to this point.

Understanding cultural context for individuals being treated (or coached) is crucial, these inform and shape their expectations which are hugely impactful in trust and belief (you know, the bedrock on which relationships are built). Similarly their context will have informed their conditioning and expectations. To give an example, using dry needling or acupuncture is much more likely to be seen as beneficial by someone from an Eastern culture than a Western culture (in a performance or healthcare setting). The argument that this should be dismissed as it “doesn’t work” or is “placebo” (based on currently available evidence) is not unreasonable, though may be a little closed minded when considering the role of the coach or healer. It dismisses the human interaction as part of the benefit of either situation (hello being replaced by AI).

From Benedetti et al

To use a more subtle example, that may help better prove the point, let’s use the example of two doctors (they could be coaches in a different setting but I’m already failing to keep this piece brief, so use your imagination to create a similar scenario).

Doctor 1 sees you, with your fever, aches, cough and general feeling of being hit by a bus. They look at you briefly, and tell you it’s likely viral and to go home and use paracetamol (aka acetaminophen/tylenol) for the aches. You’re out of the office in a few minutes, they’ve also spent a large portion of it looking at the computer.

Doctor 2 sees you, examines you, listening to your chest, looking in your throat, takes your temperature then has a discussion about why it is likely a self-limiting viral illness, unlikely bacterial and why antibiotics are not warranted. They explain what to do, what may mean you need to seek care again and some potentials for symptomatic relief. They ask about your concerns and ensure they’re all addressed. You’re with the doctor for what feels like a good amount of time (it’s probably not much longer to be honest, interestingly this is something that doctors misunderstand often based on what research I’ve seen).

It’s arguable that both of these are evidence-based and very reasonable ways to manage the situation (let’s ignore systemic pressures of patient loads, billing structures etc). It’s also possible that both see resolution of symptoms in a similar time frame (let’s call that 48hrs or so). But which in which scenario do you think the patient is more confident and do you think this plays a role in healing?

As sometimes occurs, someone else’s writing summarises your thinking better than you can. Kaptchuk’s writing on placebo in the New England Journal of Medicine is great and this portion is perhaps the crux of it:

“In a broad sense, placebo effects are improvements in patients’ symptoms that are attributable to their participation in the therapeutic encounter, with its rituals, symbols, and interactions. These effects are distinct from those of discrete therapies and are precipitated by the contextual or environmental cues that surround medical interventions, both those that are fake and lacking in inherent therapeutic power and those with demonstrated efficacy. This diverse collection of signs and behaviors includes identifiable health care paraphernalia and settings, emotional and cognitive engagement with clinicians, empathic and intimate witnessing, and the laying on of hands.”

So, we end up with some texture to what is the catch all of “The Placebo Effect”. We have natural course of the disease, then sociocultural aspects of healing, the therapeutic effect of touch and finally the ‘other’ bucket (for the record there is always an ‘other’ or ‘miscellaneous’ category when creating categories - this helped a lot in medical school).
We also end up at the point where many readers probably think part of the placebo effect is actually just good medical practice. Now you understand the ethical dilemma…

How Effective are Placebos?

The painfully unsatisfying but technically correct answer is “it depends”.

The more helpful answer is, there are many factors that modify this, including factors covered above (in the figures) such as genetics. There certainly seems to be some conditions where ‘placebos’ seem to work more effectively, including pain and mental health disorders. This looks to be as a result of some brain chemistry and the way the mind (not the brain) interplays with physiology.

That said, to try give a quantification to this, to make things more satisfying, the only numbers I could find is the relatively consistent placebo response rate of 35-40% in antidepressant trials (Furukawa, 2016). Hafliðadóttir reported that around half of the treatment effects in randomised clinical trials are attributable to contextual effects rather than interventions (this speaks to the texture in what people are broadly terming ‘The Placebo Effect’). When it comes to performance Lindheimer and colleagues suggest that “the placebo effect is approximately half of the observed psychological benefits of exercise training”. Interestingly this made me think of one of my favourite studies where hotel maids told that their work was exercise effectively saw the benefits of exercise whereas those who weren’t told this (the control group) saw no benefits despite the same activity (I am not convinced this is true placebo but regardless it’s fascinating and speaks to the impact of belief and buy-in in training).

What is a real ‘noggin scratcher’ is the efficacy of placebo’s if people know they have taken a placebo. This gets very ‘meta’, and quickly, but in essence, in some drug trials patients are asked whether they thought they took the placebo or the active drug. Which also means there’s research on the efficacy of drugs when people are told they’re placebo and placebos when people are told they’re placebos.

“A recent study of episodic migraine demonstrated that when patients took rizatriptan (10 mg) that was labeled “placebo” (a treatment that theoretically had “pure pharmacologic effects”), the outcomes did not differ from those in patients given placebos deceptively labeled “rizatriptan” (pure expectation effect). However, when ritzatriptan was correctly labeled “rizatriptan,” its analgesic effect increased by 50%.” (from Kaptchuk 2015)

---

Does Placebo Pill Colour Impact Efficacy?

The short answer is YES!

Broadly:

• Yellow seems to be associated with being an antidepressant/stimulant

• Blue is associated with tranquillising/anxiolytic effects

• Red seems to be associated with arousal/anxiety provoking effects

• Green, Red, Blue and Yellow are associated with tranquillising/anxiolytic effects

• White seems to have neutral effects

All of this said, there is some suggestion that effectiveness of the placebo depends on the patient's colour preference rather than attributing colours to any effects which may be why the research on pill colour is a little mixed and contentious.

This is particularly interesting to me, as it fundamentally boils down to ‘brand’ and ‘branding’. That is to say, certain colours, fonts etc will tell a story and play a role in this. This is perhaps why white does not have an attributable effect. This is even more interesting in the context of generic vs none generic drugs (especially in the USA where drug brand names are more commonly used).

Note; it is for the above reason that I am passionate about the need for everyone to understand and use storytelling more to help augment impact (and why I think blanket criticism and dismissal of marketing is narrow sighted).

---

The Role of Genetics in Placebo aka “The Placebome”

This is an area of the literature very much in its infancy, but it is certainly interesting enough to warrant inclusion and helps to clarify the picture on the placebo effect somewhat.

Below is a table from the literature showing some genetic variants (known as single nucleotide polymorphisms aka SNPs) that are associated with greater susceptibility to the placebo effect.

From Hall et al

The thing that stood out to me about these is that the genes all code for parts of the neurotransmitter pathways. Suggesting, perhaps unsurprisingly, that there is a huge component of neuropsychobiology in the placebo effect. This also tracks with the fact that the placebo effect is strongest in conditions with large psychobiological components such as pain and certain mental health disorders (no this isn’t to say these don’t exist, but more so that they have components of psychology as part of them).

What Alzheimer’s Can Teach us About the Placebo Effect?

The history of medical understanding and ‘research’ is quite sordid in parts (read human rights infringements) but occasionally a ‘natural experiment’ presents itself. This effectively boils down to something happening that gives us insights into a situation we could never test/research (or haven’t for some reason). An example may be someone going blind via a genetic disease and as a result tracking the brain via fMRI or similar to see that the brain devotes resources previously for vision to other senses.

Alzheimers, as cruel and sad as it is, presents a type of natural experiment for the placebo effect. In some cases of Alzheimers, the function of the prefrontal cortex is impaired. This region is responsible for executive function which includes impulse control, emotional regulation, problem solving, understanding consequences and the list goes on. Interestingly, this region is seen to be activated by placebo-induced expectation of benefit for example pain reduction. Moreover when the function of this region is impaired, responses to placebos are reduced or totally lacking. Thus providing insight into the role the mind, expectations, conditioning and the likes play a role in the placebo effect.

Nocebo

The flip side of the placebo coin, is the “Nocebo Effect”. In short, being told of the potential or likely negative impact of an intervention can indeed induce a negative impact. This is probably the reason that every single drug or medicine on the market can cause headaches, it’s hard to believe that every single one has a mechanism to do this.

This is where the ethics get REALLY murky. If the nocebo effect exists, how much detail of risk is reasonable to give to balance informed consent and potential risk of the nocebo effect?

What About Wearables?

The very nature of wearable data means they are rife for potential erroneous reporting, compounded further by product focussed composite scores (all be they well meaning). A level of knowledge about things like appropriate use, personality type, measured vs estimated data points etc (all covered here) may help inoculate against this but it still has the potential for misinterpretation. Who cares I hear you ask? Well, whilst there isn’t much research, there is some that is suggestive that erroneous sleep data specifically can impair performance the following day. So mindful use is pertinent, which includes checking in with subjective feeling before checking data and perhaps not checking data on days where performance is crucial.

Take Aways

• There are many societal, cultural and contextual factors that contribute to what most people call placebo. These should inform and constitute appropriate interactions in all settings as much as we can (health, performance, business, whatever).

• The effectiveness of placebo is more significant in issues with more psychological overlays (it’s a neuropsychobiology effect). This also means certain genetics may play a role, particularly those involving neurotransmitters.

• The ethics of intentionally deceiving folks even if it is to their benefit is a very difficult discussion and where folks end up is probably based on their own morals and ethics. This may be less so in situations of performance rather than healthcare.

Make sure you are subscribed and check out next week’s newsletter, where I will dig further into placebo through a more performance lens.

References

1. Ernst E, Resch KL. Concept of true and perceived placebo effects. BMJ. 1995 Aug 26;311(7004):551-3. doi: 10.1136/bmj.311.7004.551. PMID: 7663213; PMCID: PMC2550609.

2. Furukawa TA, Cipriani A, Atkinson LZ, Leucht S, Ogawa Y, Takeshima N, Hayasaka Y, Chaimani A, Salanti G. Placebo response rates in antidepressant trials: a systematic review of published and unpublished double-blind randomised controlled studies. Lancet Psychiatry. 2016 Nov;3(11):1059-1066. doi: 10.1016/S2215-0366(16)30307-8. Epub 2016 Oct 7. PMID: 27726982.

3. Hafliðadóttir, S.H., Juhl, C.B., Nielsen, S.M. et al. Placebo response and effect in randomized clinical trials: meta-research with focus on contextual effects. Trials 22, 493 (2021). https://doi.org/10.1186/s13063-021-05454-8

4. Lindheimer JB, O'Connor PJ, Dishman RK. Quantifying the placebo effect in psychological outcomes of exercise training: a meta-analysis of randomized trials. Sports Med. 2015 May;45(5):693-711. doi: 10.1007/s40279-015-0303-1. PMID: 25762083.

5. Crum AJ, Langer EJ. Mind-set matters: exercise and the placebo effect. Psychol Sci. 2007 Feb;18(2):165-71. doi: 10.1111/j.1467-9280.2007.01867.x. PMID: 17425538.

6. Kaptchuk TJ, Miller FG. Placebo Effects in Medicine. N Engl J Med. 2015 Jul 2;373(1):8-9. doi: 10.1056/NEJMp1504023. PMID: 26132938.

7. JACOBS, KEITH W.; NORDAN, FRANCES M. (1979). CLASSIFICATION OF PLACEBO DRUGS: EFFECT OF COLOR. Perceptual and Motor Skills, 49(2), 367–372. doi:10.2466/pms.1979.49.2.367 

8. Kaptchuk TJ, Miller FG. Placebo Effects in Medicine. N Engl J Med. 2015 Jul 2;373(1):8-9. doi: 10.1056/NEJMp1504023. PMID: 26132938.

9. Hall KT, Loscalzo J, Kaptchuk TJ. Genetics and the placebo effect: the placebome. Trends Mol Med. 2015 May;21(5):285-94. doi: 10.1016/j.molmed.2015.02.009. Epub 2015 Apr 14. PMID: 25883069; PMCID: PMC4573548.  

10. Benedetti, F., Carlino, E. & Pollo, A. How Placebos Change the Patient's Brain. Neuropsychopharmacol 36, 339–354 (2011).

11. Draganich C, Erdal K. Placebo sleep affects cognitive functioning. J Exp Psychol Learn Mem Cogn. 2014 May;40(3):857-64. doi: 10.1037/a0035546. Epub 2014 Jan 13. PMID: 24417326.

12. Gavriloff D, Sheaves B, Juss A, Espie CA, Miller CB, Kyle SD. Sham sleep feedback delivered via actigraphy biases daytime symptom reports in people with insomnia: Implications for insomnia disorder and wearable devices. J Sleep Res. 2018 Dec;27(6):e12726. doi: 10.1111/jsr.12726. Epub 2018 Jul 10. PMID: 29989248.