Subtle cues transmit placebo effects

We have known for a while that different doctors can produce different effects using the same substance, or even placebo, such that otherwise effective treatments might become ineffective or placebo becomes effective. Chang and colleagues now clarify that such differential effects are likely transmitted by subtle facial cues, using a placebo–pain model.

Placebo research has morphed from a fringe to a mainstream topic over the last years. Placebo effects might be the main reason why a lot of conventional and unconventional pharmacological and psychotherapeutic treatments work1,2. Especially since the first brain imaging studies have shown that placebo-effects are not only not imaginary but are associated with changes in major neurotransmitter systems in the brain3,4,5, even sceptics have started to understand that the measurable effects of molecules on receptors are only half the therapeutic effect in most diseases except in comatose, unconscious or demented patients. Research has subsequently clarified that expectancy effects and conditioning are responsible for such effects. Even open placebo administration, i.e., placebo provided with full knowledge of patients that they are being given placebo, shows statistically and clinically significant effects6. We also know that individual experimenters produce different results in experimental studies, whether with humans7 or with animals8. But we do not know exactly how such effects are transmitted. This is where a new study in Nature Human Behaviour steps in. Chang and colleagues at Dartmouth College in New Hampshire have tackled this question with an elegant experiment.

The authors used a placebo–pain model with thermal stimulation9. In such an experiment, participants are stimulated thermally via small thermal electrodes placed at the forearm. The electrodes are heated, which produces an unpleasant painful sensation. In a placebo conditioning trial, a placebo cream is then applied, which the participants believe to be a potent pain medication, while at the same time the temperature is lowered, in this study from 47 °C to 43 °C. Thereby an objective reduction of pain is coupled with the application of an otherwise ineffective cream. This leads to the conditioned belief that the cream is actually an effective treatment for thermal pain.

To study how such placebo effects are then transmitted from providers to their patients, the authors constructed experimental doctor–patient dyads from their pool of undergraduate student participants by randomizing one member of a pair to act as the ‘doctor’ and one to act as the ‘patient’. The experimental doctors first underwent the conditioning phase, learning about the ‘effectiveness’ of the placebo cream, presented with the fancy name “Thermedol”, and were then told to apply the Thermedol cream and a control cream—both were in fact identical in content—in similar situations to the experimental patient. The patient was, of course, kept blind as to the cream and the experimental procedures.

With this experimental setting the authors copied the single-blind clinical situation in which a real doctor is convinced that he or she is using an effective treatment and a real patient is being given such a treatment without knowing about its real effectiveness.

The ‘doctors’ found the treatment effective, and not very surprisingly, their ‘patients’ found it similarly effective, when compared to the control cream that was supposed to be inert. The amount of pain relief the patients experienced and the degree to which they believed in the effectiveness of the treatment were nicely matched to those of their doctors, and the doctors were rated more empathic in the Thermedol condition compared with the control.

But how did this happen? This is the decisive question on which the authors have now lifted the veil, at least a little. They mounted cameras on the patients’ heads to analyse the facial expressions of the doctor. Indeed, they saw that during the control trials, the doctors showed stronger facial expressions of pain, which correlated with patients’ experiences and their own facial expressions. The experimental doctors, apparently convinced through their own experience, expressed more pain-like features in their faces while administering the control cream than while administering the Thermedol cream, the alleged pain-relieving effect of which they had learned during their own conditioning. Thus—often suspected but now shown—it is subtle facial cues that seem to be picked up by the patients. Whether this just increases the patients’ own confidence, conveys information about what to expect, makes the doctors more empathic in the supposedly real treatment or has some other effect, we do not know. But we do know now that subtle cues are being transmitted by doctors and read by patients.

The study is strong in itself, as it replicates its own results three times, with three different student cohorts. The authors have safeguarded their results against various potential confounders and problems. Importantly, they kept experimenters that interacted with the dyads blind as to what the ‘true’ and the false cream were. This lends credibility to their view that it was indeed the doctors, and not the experimenters, that conveyed the effect.

But it should not be overlooked that this is an experimental, and by the same token, artificial setting. The ‘doctors’ and ‘patients’ were neither doctors nor patients, but students. It would be interesting to now study the clinical field: how do effective doctors communicate, perhaps compared to less effective ones? Are such effects transported via subtle cues, like facial expressions, as suggested by this study, or by verbal information or a mix? What happens when patients’ and doctors’ expectations differ? Or when doctors’ verbally expressed information is dissonant with their implicitly held views? Will, then, the implicit views, perhaps conveyed by subtle facial cues, trump verbal information? The clinical field is likely a little bit messier than the lab. Many interventions work very well in clinical experiments of controlled clinical trials but are pretty useless in clinical practice, and interventions that are less than optimal in clinical experiments have created a nimbus around so-called wonder drugs that is mostly a postmodern myth10. So what, apart from chemical constituents, makes an effective drug? An effective doctor who believes in the drugs? A doctor who claims to believe and knows how to hide his or her true feeling? Or is it not at all possible to hide one’s true opinions in the presence of patients, because they will decode the doctor’s subtle facial cues anyway? A lot remains to be done, but this study will surely be a landmark because it started to disentangle the question of how placebo effects might be transmitted. The answer is: socially, very likely through subtle facial cues.


  1. 1.

    Blease, C. & Kirsch, I. Psychol. Conscious. 3, 105–198 (2016).

    Google Scholar 

  2. 2.

    Colloca, L., Flaten, M.A. & Meissner, K., eds. Placebo and Pain: From Bench to Bedside. (Elsevier-Academic Press, 2013).

  3. 3.

    de la Fuente-Fernández, R. et al. Science 293, 1164–1166 (2001).

    Article  Google Scholar 

  4. 4.

    Petrovic, P., Kalso, E., Petersson, K. M. & Ingvar, M. Science 295, 1737–1740 (2002).

    CAS  Article  Google Scholar 

  5. 5.

    Wager, T. D. et al. Science 303, 1162–1167 (2004).

    CAS  Article  Google Scholar 

  6. 6.

    Kam-Hansen, S. et al. Sci. Transl. Med. 6, 218ra5 (2014).

    Article  Google Scholar 

  7. 7.

    Uhlenhuth, E. H. et al. Psychopharmacology (Berl.) 9, 392–418 (1966).

    CAS  Article  Google Scholar 

  8. 8.

    Rosenwein, R.E. in The Social Psychology of Science (eds W.R. Shadish, Jr. & S. Fuller) 262–285 (Guilford Press, 1994).

  9. 9.

    Chen, P.-H.A. et al. Socially transmitted placebo effects. Nature Human Behaviour (2019).

  10. 10.

    Gøtzsche, P.C. Deadly Psychiatry and Organised Denial. (People’s Press, 2015).

Download references

Author information



Corresponding author

Correspondence to Harald Walach.

Ethics declarations

Competing interests

The author declares no competing interests.

Rights and permissions

Reprints and Permissions

About this article

Verify currency and authenticity via CrossMark

Cite this article

Walach, H. Subtle cues transmit placebo effects. Nat Hum Behav 3, 1246–1247 (2019).

Download citation


Quick links

Nature Briefing

Sign up for the Nature Briefing newsletter — what matters in science, free to your inbox daily.

Get the most important science stories of the day, free in your inbox. Sign up for Nature Briefing