Pain and the Placebo Effect

Pain and the Placebo Effect

Reprinted from "Frontiers of Pain" with the kind permission of Science Press


Many forms of placebo have been used in trials for pain relief. The most effective responses have been to injections and sophisticated looking equipment, but the response rate varies very widely depending on the circumstances and patient expectations. Also, the doctor's beliefs and expectations are of surprising importance, showing through even a double blind procedure. Every treatment, whether active or placebo, has a psychological effect on its recipient and this effect is strongly conditioned by the behaviour of the therapist.


In recent years there has been a growing interest in the role of psychological factors in the experience of pain, along with the development of several new psychologically-based therapies for pain-related problems (Wall and Melzack, 1989). This has been accompanied by a parallel increase in awareness of the importance of psychological influences on patients' responses to the more traditional physical forms of treatment. Perhaps the clearest example of such an influence is the placebo effect.

A placebo has been defined as "any therapy or component of therapy that is deliberately used for its nonspecific psychological or psychophysiological effect, or that is used for its presumed specific effect, but is without specific activity for the condition being treated" (Shapiro and Morris, 1978). Whether or not one subscribes to the deliberate use of placebos in routine medical practice, the findings of research in this field are of relevance to all clinical work in so far as the placebo effect can be seen to exemplify the psychological impact of medical treatments in general. This article summarises the current state of knowledge concerning the range and extent of placebo effects, with special reference to the treatment of pain.

Symptom Relief

Placebos have been administered under many different guises. Although the stereotype is the sugar pill, survey findings suggest that a 5% saline injection - for the purposes of pain relief is probably the most common form of administration (Gray and Flynn, 1981). Placebo surgery has been reported (Cobb et al., 1959, Diamond et al., 1960) in which the patient receives nothing more than an operation scar. Placebo responses have also been induced from exposing patients to technically sophisticated equipment, e.g. X-rays, (Schwitzgebel and Traugott, 1968) and by other treatment procedures which mimic a purportedly active procedure (e.g. placebo TENS - Langley and Sheppeard, 1987; sham acupuncture - Vincent and Richardson, 1986).

Placebo induced symptom relief has been documented in a wide range of disorders, including: allergies, angina pectoris, asthma, cancer, cerebral infarction, diabetes, enuresis, epilepsy, insomnia, migraine, multiple sclerosis, neurosis, parkinsonism, psychosis, skin diseases, ulcers and warts (Hass et al., 1959; Horningfeld, 1964; Totman, 1976; White et al., 1985). Indeed Ross and Olson (1982) conclude that there seems to be almost no limit to the range of disorders for which placebos have produced some degree of symptom reduction.

Response rates vary widely depending on the condition being treated (Ross and Olson, 1982). Where placebo analgesia is concerned, it is commonly reported that 35% of patients will respond to placebo administration with a degree of relief comparable to that which might be expected from a clinical dose of morphine (Beecher, 1955; Evans, 1974). However the level of response does not appear to be fixed, and may depend both on the particular measures used to assess pain (e.g. subjective vs objective) and on the context and manner of administration of the placebo. Thus, for example, laboratory studies of experimentally induced pain typically yield lower placebo response rates (15% - Evans, 1974), whilst research focusing directly on the mechanisms of placebo analgesia has typically found much higher rates (e.g. 55% Langley et al., 1984; 62% - Grevert et al., 1983).

Placebo Side Effects

Adverse reactions have frequently been noted following placebo administration. Common side-effects include nausea, drowsiness, sweating, concentration difficulties and skin rashes (Gowdey, 1983; Haegerstam et al., 1982). A small number of patients report placebo induced symptom-worsening (referred to as the nocebo effect by Kissel and Barrucand, 1974). In addition Boleloucky (1971) reports the case of a patient who apparently showed all the classic signs of dependence on a placebo which had been previously substituted for her morphine injections.

In the hope of maximising the placebo effect, researchers have attempted to identify the characteristics of physical treatments which best enhance their psychological impact. The results of such endeavours have been mixed. Although many claims have been made about the ideal size, apparent dosage level, and colour of the maximally effective pill, the overall result has been a string of inconsistent and unreplicated findings (Richardson, 1989).

Clearer findings emerge from comparisons of the effectiveness of different treatment modalities (e.g. pill vs injection). It seems that treatments that are perceived as more "serious" or "major", in some respect, are associated with greater placebo responsiveness. For example, placebo injections have generally been found to be more effective than inert pills (Carne, 1961; Grenfell et al., 1961). Moreover, unusually high improvement rates have been found following placebo surgery (85% in angina pectoris patients - Cobb et al., 1959). Placebo procedures which make use of apparently sophisticated technical equipment (e.g. subliminal pulse therapy Langley and Sheppeard 1987) have also generated high positive response rates where pain is concerned.

The fact that only a proportion of patients respond to placebos has led to the suggestion that placebo responders may differ in some significant way from non-responders. Responders have been claimed to be more likely to be: emotionally dependent (Lasagna et al., 1954); extrovert (Campbell and Rosenbaum 1967); neurotic (Garttner, 1961); and suggestible (McGlashan et al., 1969). Once again however, research findings in this field are generally inconsistent (Richardson, 1989) and it appears that placebo responsiveness per se may not be an enduring trait (Liberman, 1964). Thus the idea of a "typical placebo responder" may in itself be misplaced. The suggestion that a positive placebo response can be interpreted as evidence that a patient's pain is "imaginary" or psychogenic also fails to find support in the research literature (Evans, 1985).

Therapist's Behavour

Therapist variables have been shown to influence placebo responsiveness. For example, the manner in which the treatment is administered has an effect on therapeutic outcome. Placebos administered with greater confidence (Uhlenhuth et al., 1959) and concern (Grevert et al., 1983) produce stronger positive responses. Moreover, placebos administered by high status therapists appear to have greater effects than those given by those of lowlier status (Shapiro and Morris, 1978).

The idea that a placebo pill may only be important in so far as it symbolises the healing powers of the doctor was explored by Park and Covi (1965). In their small uncontrolled pilot study, psychiatric patients were explicitly told that the medication was an inert sugar-pill but that the doctors believed it would help their condition. Results showed that 13 of the 14 participants still derived clinically significant benefits from the placebo.

On the other hand doctors' beliefs and expectations may influence patient responses in far less obvious ways than those investigated by Park and Covi. There is evidence for example that the influence of the doctor's expectations on the patient's response to a treatment may even bypass the expected constraints of the double blind control procedure. In a study investigating the role of the endogenous opioids in placebo analgesia Gracely et al., (1985) found that, despite the use of a double blind procedure, physician expectations still exerted differential effects on patient responses to injections of fentanyl, naloxone, or placebo.

Overall, then, the evidence appears to suggest that the therapist's behaviour has subtle, yet powerful influences on patient responses to placebo. What is as yet unclear is which particular aspects of the therapist's behaviour influence which specific aspects of the patient's response and via what psychological, or other, processes.


The precise mechanisms whereby placebos produce their pain-relieving effects have as yet to be delineated. Endorphin mediation was implicated by early research which found placebo analgesia to be naloxone-reversible (Levine et al, 1978). More recent studies however suggest that this can at best provide only a partial explanation (Grevert and Goldstein, 1985). Classical conditioning processes have been implicated by Voudouris et al. (1985,1989) in a series of laboratory studies examining conditioned placebo responses to a bogus analgesic cream. Again, however, the available evidence in this area is too sparse to enable firm conclusions to be drawn. Other accounts of placebo phenomena have invoked expectancy effects, anxiety reduction, cognitive dissonance, misattribution and labelling processes, transference phenomena, and a host of other hypothetical mediating processes (cf Richardson, 1989).

What remains clear is that whatever psychologically mediated events take place in the mind or body of the placebo-responsive patient, a major determinant of the process of placebogenesis is the behaviour of the therapist. Every treatment - active or inert - can be expected to have at least a psychological impact on its recipient (i.e. a placebo response). This implies then that the patient who fails totally to respond to our therapies, is in part failing to respond to us.

1. Beecher H K (1955) The powerful placebo. Journal of the American Medical Association. 159:1602-6

2. Boleoucky Z (1971) A contribution to the problems of placebo dependence: case report. Activitas Nervosa Superior. 13:190-1

3. Campbell J, Rosenbaum P (1967) Placebo effect and symptom relief in psychotherapy. Archives of General Psychiatry. 16:364-8

4. Carne S (1961) The action of chorionic gonadotrophin in the obese. Lancet. ii: 1282-4

5. Cobb L A, Thomas G I, Dillard D H et al (1959) An evaluation of internal mammary-artery ligation by a double blind techinque. New England Journal of Medicine. 260: 1115-8

6. Diamond E G, Kittle C F, Cockett J E (1960) Comparison of internal mammary-artery ligation and sham operation for angina pectoris. The American Journal of Cardiology. 4: 483-6

7. Evans F J (1974) The placebo response in pain reduction. Advances in Neurology. 4:289-96

8. Evans F J (1985) Expectancy, therapeutic instructions and the placebo response. In: White L, Tursky B, and Schwartz G E, eds. Placebo: Theory, Research and Mechanisms. The Guildford Press, New York: 215-28

9. Gartner M A (1961) Selected personality differences between placebo reactors and non-reactors. Journal of American Osteopathic Association. 60:377-8

10. Gowdey C W (1983) A guide to the pharmacology of placebos. Canadian Medical Association Journal. 128: April 15:921-5

11. Gray G, Flynn P (1981) A survey of placebo use in general hospital. General Hospital Psychiatry. 3:199-203

12. Gracely R H, Dubner R, Deeter W R, Wolskee P J (1985) Clinical expectations influence placebo analgesia. Lancet. i: 43

13. Grenfell R, Briggs A H, Holland W C (1961) A double-blind study of the treatment of hypertension. Journal of the American Medical Association. 176:124-67

14. Grevert P, Albert L H, Goldstein A (1983) Partial antagonism of placebo analgesia by naloxone. Pain. 16:129-43

15. Grevert P, Goldstein A (1985) Placebo Analgesia, Naloxone, and endogenous Opioids. In: White L, Tursky B, and Schwartz G E, eds. Placebo: Theory, Research and Mechanisms. The Guildford Press, New York.

16. Haas A, Fink H, Hartfelder G (1959) Das Placebo problem. Fortschritte der Arzneimittelforschung. 1:279-454

17. Haegerstarn G, Huitfeldt B, Nilsson B Set al (1982) Placebo in clinical drug trials - a multi-disciplinary review. Methods and Findings in Experimental and Clinical Pharmacology. 4(4): 261-78

18. Horningfeld G (1964) Nonspecific factors in treatment: 1. Review of placebo reactions and placebo reactors. Diseases of the Nervous System. 25:145-56

19. Kissel P, Barrucand D (1974) Placebos et effet-placebo en medecine. Masson, Paris

20. Langley G B, Sheppeard H (1987) Transcutaneous electrical nerve stimulation (TNS) and its relationship to placebo therapy: a review. New Zealand Medical Journal. 100:215-7

21. Langley G B, Sheppeard H, Johnson M, Wigley R D (1984) The analgesic effects of transcutaneous nerve stimulation and placebo in chronic pain patients. Rheumatology International. 2:1-5

22. Lasagna L, Mosteller F, Von Felsinger J, Beecher H (1954) A study of the placebo response. American Journal of Medicine. 16:770-9

23. Levine J D, Gordon N C, Fields H L (1978) The mechanisms of placebo analgesia. Lancet. 2:654-7

24. Liberman R (1964) An experimental study of the placebo response under three different situations of pain. Journal of Psychiatric Research. 2:233-46

25. McGlashan T H, Evans F J, Orne M T (1969) The nature of hypnotic analgesia and placebo response to experimental pain. Psychosomatic Medicine. 31:227-46

26. Park L C, Covi L (1965) Nonblind placebo trial. Archives of General Psychiatry. 12:336-45

27. Richardson P H (1989) Placebos: their effectiveness and modes of action. In: Broome A K, ed. Health Psychology: Processes and Applications. Chapman and Hall, London: 34-56

28. Ross M, Olson J M (1982) Placebo effects in medical research and practice. In: Biser J R, ed. Social Psychology and Behavioural Medicine. John-Wiley and Sons, Chichester: 441-58

29. Schwitzgebel R, Traugott M (1968) Initial note on the placebo effect of machines. Behavioural Medicine. 13: 267-73

30. Shapiro A K, Morris L A (1978) The placebo effect in medical and psychological therapies. In: Bergin A E, and Garfield S, eds. Handbook of Psychotherapy and Behavioural Change (2nd edn). John Wiley and Sons, New York: 369-410

31. Totman W A (1976) Cognitive dissonance in the placebo treatment of insomnia - a pilot experiment. British Journal of Medical Psychology. 49:393-400

32. Uhlenhuth E H, Canter A, Newstadt J O, Payson H E (1959) The symptomatic relief of anxiety with meprobamate, phenobarbital and placebo. American Journal of Psychiatry. 115:905-10

33. Vincent C A, Richardson P H (1986) The evaluation of therapeutic acupuncture: concepts and methods. Pain 22:1-13

34. Voudouris N J, Peck C, Coleman G (1985) Conditioned Placebo Responses. Journal of personality and Social Psychology. 48 (I): 47-53

35. Voudouris N J, Peck C, Coleman G (1989) Conditioned Response of the placebo phenomena: further support. Pain. 38:109-16

36. Wall P, Melzack R (1989) A Textbook of Pain (2nd edn). Churchill Livingstone, Edinburgh

37. White L, Tursky B, and Schwartz G E (1985) In: White L, Tursky B, and Schwartz G E, eds. Placebo: Theory, Research and Mechanisms. The Guildford Press, New York: 3-8

The British Medical Acupuncture Society.


By P.H. Richardson

Share this with your friends