Of Kids And Cats: Placebo By Proxy

Treatment responses in (small) children are difficult to assess since descriptions of symptoms and symptom changes are not easily available without the use of language, e.g. in the case of pain and emotions. The same holds true for therapies in animals, be it domestic animals or otherwise.

In all these cases, the judgments rely on observation and interpretation of behaviors through proxies (parents, relatives, caregivers) on the one hand and animal owners on the other. And of course, drug trials in both children and animals are required to control for specific and non-specific effects, e.g. by using placebos in randomized controlled trials.

Whether such a treatment can produce a pronounced placebo effect as known from studies in adults, i.e. a symptom improvement that is related to the expectation of a treatment rather than the compound applied, is even more difficult to answer since in such cases, both the expectation of the patient (child, animal) as well as the proxy counts and contributed to the effect of a placebo provided. And more importantly, placebos could also operate indirectly by producing changes in how the proxies themselves behave toward the patient, which in turn provoked behavioral and symptomatic changes. This concept of “placebo by proxy” (PbP) has recently received attention from a methodological point of view (1).

If you don’t believe that animals have treatment expectations in a medical setting, go and observe animal behaviors (and that of their holders) in a veterinary outpatient practice with plenty of sick animals of different breed in the same waiting room — no barking, strolling around, playing, or fighting  — or in the doctor’s office. And not only newborn babies respond to touch and voice with calming down and stopping crying. 

Poor database so far

Surprisingly little has been published on the PbP concept over the last 70 years — our database identifies only about 7 papers — out of more than 4000 in total on the placebo effect/response (2), and this has not much increased since we last reviewed the literature (3). This is the more surprising since the PbP concept related not only to kids and cats but basically to patients of all ages, as we will show below. But let’s look at the literature first:

In their basic concept, Grelotti and Kaptchuk (1) argue that the frequent over-prescription of antibiotics specifically in children occur because of parents’ concerns and wishes (4). Proxies’ influence on (placebo) responsiveness may also be responsible for differences in expectancy reports seen between doctor and patient-reported outcomes, especially in depression (5). Whalley and Hyland (6) were the first to investigate whether the efficacy of a presumed placebo — homeopathic — remedy (Bach flower) would be affected by parents’ beliefs and emotions, after the exclusion of any direct effect of physicians’ interaction with either the children or the parents using a telephone system. The authors noted the improvement of tantrum symptoms (frequency and severity) to be associated with parents’ mood but cannot exclude this to be mediated by altered parents behaviors towards the child.

A study investigated parents’ anxiety on the one hand and their expectation towards acupuncture on the other on children post-operative symptom reports after tonsillectomy or appendectomy, in which acupuncture (versus treatment as usual) was used as a therapy of postoperative morbidities (7). The authors did not find an association between the children’s symptoms and parents’ expectations/concerns, but call for caution because of unknown confounders and limitations. In a survey conducted within the same study (8), the authors found that parents expectations changed over time in the course of the study and that such changes to the positive were reflected in better (lesser) post-operative symptoms in the children.

In a placebo-controlled study of the neuroprotective effect of early high-dose recombinant human erythropoietin, a biologic drug, in very pre-term-born babies (9), it was investigated whether parent’s belief that the babies had received the drug versus that it had received placebo made a difference for the long-term outcome (intelligence quotient, IQ at age 5). This was the case: After controlling for a number of potential confounding variables (social status, head circumference, and sepsis), children of parents assuming that their baby had received the drug had a small but significantly better long-term outcome (IQ) than if their parents believed it had received placebo, but this effect was too small to be of clinical relevance.

Recording of treatment (and placebo) effects is challenging not only in the very young, but also in children and adults with intellectual disabilities, e.g. Prader-Willi’s, Down’s, or William’s syndrome. A meta-analysis of placebo-controlled drug trials (10) on patients intellectual abilities in these patients revealed high placebo response rates both for proxy-observed as well as for objective outcome measures. Higher IQ predicted higher placebo responses, as did younger age. The authors assumed that PbP effects are driving these effects. Similar observations have been made in other conditions where proxy ratings are essential for evaluation of therapy outcome, e.g. in Attention-Deficit-Hyperactivity Disorder (ADHD) in children (11).

Back to the cats: A meta-analysis of 5 placebo-controlled trials of analgesics in 96 cats (12) revealed a 50 to 70% placebo response on subjective measures of treatment success by the cats’ owners, while only 10 to 50% of placebo success rates were noted with objective measures. Similar discrepancies between symptom and improvement ratings by owners and veterinarians were noted earlier (13). No predictors of the placebo response were found among the animal characteristics, while owners’ characteristics were never studied.

A systematic approach to PbP

We believe that the PbP concept as illustrated above represents a subset of an interactional system in which placebo responses are generated by increasing complexity of the network, (see Figure, different grey shades may reflect different communication intensities).

A: This is the idealized medical situation in contemporary medicine, and least in westernized social systems where the (adult) patient individually communicates with the doctor and reports all relevant events in his/her medical history and environment (including family). Nearly the entire placebo literature (2) assumes this constellation as the basis for placebo effects.

B: The PbP concept that is illustrated here is a reflecting a deficit of the patient to directly communicate with the doctor, either verbally (babies, animals) or intellectually (in case of respective disability), and therefore the proxy reports instead, based on either observation of the patient behavior or on (limited or special) communication strategies.

C: Instead of exclusively communicating with the proxy, doctors will normally rely on additional information retrieved directly from the patient, either from observations or from “objective” measures of symptoms, performance, disease markers etc. This may generate conflicting information, e.g. higher placebo effects from proxy reports than from measures (12).

D: While model C may mirror normality compared to A, the social environment of a patient usually contains more than just one proxy, and usually also proxies with variable proximity to the patient, from family (parents, children, siblings) to relatives and friends/peers/ colleagues. Proximity determines how much they may be involved in the medical history and its reporting, and how much the doctor may be aware of this social network and its influence on disease reporting, management and efficacy. Social network theories would be needed to establish their influence on treatment, placebo or otherwise.

E: It is not only possible but most likely that one or more of the members of a social network of a given patient has also a relevant disease that is communicated about, and eventually also the same or a similar disease than the index patient. In all these cases, observed (or experienced) treatments and their efficacy become contributors to the treatment effect of the index patient, either via social observation or explicit or implicit learning, vice versa.

Where should placebo research go

As indicated above, placebo research has vastly limited its focus on the individualized doctor-patient relationship (A) and therefore, contains an anachronistic moment: because this is what we long for — a doctor caring for just for me — does not make it reality; we have seen a similar ignorance towards contemporary e-health and m-health in placebo research (14). The PbP concept (B), on the other hand, only substitutes a communication deficit on the patient´s side but maintains this social isolation of the doctor-patient interaction. Model C finally has been referred to in placebo research so far only with specific medical conditions where proxy ratings are explicitly needed. It is therefore not surprising that searching for individual predictors of the placebo effect, e.g. sex, age (15), personality (16) or gene outfit has become fashionable and state of the art.

The complex social models for placebo effects, as in D and E, are only about to become apparent, with research looking at the effects of social observation (17), at the gene-environment interactions of pain and placebo analgesia using twin studies (18), and with exploration of the implicit learning mechanisms of placebo effects between children and their parents (19), and in relation to other models (20). Placebo research still has to go a long way to cover these elements, and to integrate social psychology into its armory in the future.

This is part 9 of a series covering “placebo” provided by Paul Enck and Sibylle Klosterhalfen from the Tübingen University Hospital. Continuous updates on placebo research can be found at www.jips.online.


  1. Grelotti DJ, Kaptchuk TJ. Placebo by proxy. Bmj. 2011;343:d4345.
  2. Enck P, Horing B, Broelz E, Weimer K. Knowledge Gaps in Placebo Research: With Special Reference to Neurobiology. International review of neurobiology. 2018;139:85-106.
  3. Weimer K, Gulewitsch MD, Schlarb AA, Schwille-Kiuntke J, Klosterhalfen S, Enck P. Placebo effects in children: a review. Pediatric research. 2013;74(1):96-102.
  4. Mangione-Smith R, McGlynn EA, Elliott MN, Krogstad P, Brook RH. The relationship between perceived parental expectations and pediatrician antimicrobial prescribing behavior. Pediatrics. 1999;103(4 Pt 1):711-8.
  5. Rief W, Nestoriuc Y, Weiss S, Welzel E, Barsky AJ, Hofmann SG. Meta-analysis of the placebo response in antidepressant trials. Journal of affective disorders. 2009;118(1-3):1-8.
  6. Whalley B, Hyland ME. Placebo by proxy: the effect of parents’ beliefs on therapy for children’s temper tantrums. Journal of behavioral medicine. 2013;36(4):341-6.
  7. Liodden I, Sandvik L, Norheim AJ. Placebo by Proxy – the Influence of Parental Anxiety and Expectancy on Postoperative Morbidities in Children. Forschende Komplementarmedizin. 2015;22(6):381-7.
  8. Liodden I, Pripp AH, Norheim AJ. Placebo by proxy expectations toward acupuncture change over time: a survey comparing parental expectations to acupuncture pre- and postoperatively. BMC complementary and alternative medicine. 2018;18(1):183.
  9. Burkart TL, Kraus A, Koller B, Natalucci G, Latal B, Fauchere JC, et al. Placebo by Proxy in Neonatal Randomized Controlled Trials: Does It Matter? Children (Basel). 2017;4(6).
  10. Curie A, Yang K, Kirsch I, Gollub RL, des Portes V, Kaptchuk TJ, et al. Placebo Responses in Genetically Determined Intellectual Disability: A Meta-Analysis. PloS one. 2015;10(7):e0133316.
  11. Weimer K, Colloca L, Enck P. Placebo effects in psychiatry: mediators and moderators. The Lancet Psychiatry. 2015;2(3):246-57.
  12. Gruen ME, Dorman DC, Lascelles BDX. Caregiver placebo effect in analgesic clinical trials for cats with naturally occurring degenerative joint disease-associated pain. Vet Rec. 2017;180(19):473.
  13. Conzemius MG, Evans RB. Caregiver placebo effect for dogs with lameness from osteoarthritis. Journal of the American Veterinary Medical Association. 2012;241(10):1314-9.
  14. Enck P, Klosterhalfen S, Weimer K. Unsolved, Forgotten, and Ignored Features of the Placebo Response in Medicine. Clinical therapeutics. 2017;39(3):458-68.
  15. Weimer K, Colloca L, Enck P. Age and sex as moderators of the placebo response – an evaluation of systematic reviews and meta-analyses across medicine. Gerontology. 2015;61(2):97-108.
  16. Horing B, Weimer K, Muth ER, Enck P. Prediction of placebo responses: a systematic review of the literature. Frontiers in psychology. 2014;5:1079.
  17. Colloca L, Benedetti F. Placebo analgesia induced by social observational learning. Pain. 2009;144(1-2):28-34.
  18. Weimer K, Mönnikes N, Stengel A, P E. Genetics, shared, or non-shared environment? An experimental twin study on placebo analgesia. Psychosomatic medicine. 2017;79(2):A-120.
  19. Weimer K, Horing N, Colloca L, Gulewitsch MD, Schlarb AA, P E. Social learning of placebo effects in children and their parents – a feasability study. Psychosomatic medicine. 2014;2014:A-29.
  20. Weimer K, Mazurak N, Wolf C, Meyer R, Gulewitsch MD, P E. Social learning of placebo effects through observation in children: Unfamiliar models are more effective than the mother. Psychosomatic medicine. 2017;79:A120.