Explaining the Placebo Effect

Alfred A Barrios, Ph.D.

The following is an excerpt from Dr. Barrios' book, Understanding Hypnosis: Theory, Scope and Potential.

In discussing the broad implications of the definition of hypnotic induction, it was stated that the theory could also be used to explain 'the hypnotic effects (placebo effect) of psychotherapists and doctors of medicine' (Barrios, 2001: 171). The question is how? The section of the theory on prestige helps throw some light on this question:

[T]he statements, commands or suggestions of a person with prestige tend to be questioned less; that is, such a person evokes a greater inhibitory set to begin with. In general, people have previously been conditioned to accept at face value the statements of someone who is an authority in his field. That is, an inhibitory set which inhibits contradictory stimuli has been previously conditioned (in much the same way as in the hypnotic induction process). This is so because what the authority says has usually turned out to be true. (Barrios, 2001: 181)

The placebo when given by a doctor or person of authority works in the same way as hypnotic suggestion, for the person is in a heightened state of belief. For example, when the doctor gives a patient an injection 'to kill the pain', he is essentially giving the sug¬gestion 'this is going to ease your pain'. The actual pain relief occurs even if the injection is an inert saline solution because of two factors associated with suggestion. First, there is the cognitive stimulus 'pain relief' with its associated endorphin (the body's natural pain killing substance) release into the bloodstream. And second, the inhibitory set of the suggestion is evoked that would inhibit anything that might interfere with the cognitive stimulus, such as any doubts about the doctor's skills, or doubts about the painkiller's effectiveness, or even the sensory pain stimulus itself.

As another example, when the doctor gives the patient any medicine or treatment that he says will cure the patient, the cognitive stimulus 'healing' is evoked with its attendant immune associated response (e.g. release of t-cells, macrophages, etc.).

The next question that needs to be answered is from whence do the cognitive stimuli 'pain relief' or 'healing' derive their meaning: i.e. how did the words or thought 'pain relief' come to be associated with endorphin secretion or how did the word or thought 'healing' come to be associated with the immune response? I would say the answer is: through a process of higher-order classical conditioning. As Pavlov (1960: 407) so aptly put it: 'Speech, on account of the whole preceding life of the adult, is connected up with all the internal and external stimuli which can reach the cortex, signaling all of them and replacing all of them, and therefore can call forth all those reactions of the organism which are normally determined by the actual stimuli themselves.'

In other words, at some point in a person's life, the words or thought 'pain relief' were associated with the body's own natural pain relieving endorphin secretion response; and the word or thought 'healing' was associated with the body's own natural healing response while the person was experiencing the same.

There, of course, is another way that a placebo response can occur. This would be more from a form of first-order classical conditioning. For instance, when a person or animal is injected a number of different times with a pain killing medication, the stimuli associated with the injection (e.g. the syringe, the person giving the injection, etc.) are the conditioned stimuli (the CS). The pain relief (the UCR) produced by the actual pain¬killer, let's say morphine (the UCS), becomes associated with the CS such that the CS can eventually produce a conditioned response (CR) of pain relief. This CR can then also be looked upon as a placebo - in this case produced via first order conditioning. I believe this is what is behind the conditioning explanation of the placebo response of such researchers as Gliedman, Gantt and Teitelbaum, 1957; Hernstein, 1962; Knowles, 1963; and Wickramasekera, 1980.

I believe the above two-fold (first-order and higher-order conditioning) explanation may help throw some light on the questions raised in the section on placebos in Kirsch's 1985 paper on response expectancies. This should help eliminate the apparent clash between the 'conditioning' and the 'response expectancy' explanation of placebos if we can look upon the terms 'response expectancy' and 'belief' as being similar as I have previously discussed, and see that conditioning is also a factor in the 'expectancy' placebo, although higher-order as opposed to first-order.

One other area that should also be cleared up by the above higher-order conditioning explanation of placebos is the question raised by Kirsch: how can one explain placebos in terms of conditioning when placebos often exhibit functional autonomy? As put by Kirsch:

A second interesting finding of the Montgomery (1995) study is that instead of extinguishing, the placebo effect increased over the course of 10 extinction trials. This is inconsistent with classical conditioning, models of placebo-effects, but is consistent with clinical data indicating that placebo effects can be remarkably persistent. (Kirsch, 1997: p 75)

However, one can see from the previous section 'Helping towards a more comprehensive theory of learning', how one can establish some fairly strong functionally autonomous responses via the conditioning power of the belief or response expectancy aspect of placebos.

Barrios AA (2001) A theory of hypnosis based on principles of conditioning and inhibition. Contemporary Hypnosis 18: 163-203.

Gliedman LH, Gantt, WH, Teitelbaum, HA (1957) Some implications of conditional reflex studies for placebo research, American Journal of Psychiatry 113: 1103-07. Pavlov I (1960) Conditioned Reflexes. New York, NY: Dover.

Hernstein R (1962) Placebo effect in the rat. Science 138: 677-8.

Kirsch I (1985) Response expectancy as a determinant of experience and behavior. American Psychologist 40: 1189-1202.

Kirsch I (1997a) Response expectancy theory and application: A decennial review. Applied & Preventative Psychology 6: 69-79.

Knowles JB (1963) Conditioning and the placebo effects of decaffeinated coffee on simple reaction time in habitual coffee drinkers. Behavior Research and Therapy 1: 151-7.

Montgomery GH (1995) Mechanisms of placebo analgesia: Expectancy theory and classical conditioning. Unpublished doctoral dissertation, University of Connecticut, Storrs.

Wickramasekera I (1980) A conditioned response model of the placebo effect: Predictions from the model. Biofeedback and Self-Regulation 5: 5-18.