August Research Snippet: Hypnosis, Pain, Expectation & Placebo

The Role of Expectation in Hypnosis:
Hypnosis, Imagination & Placebo Pain Relief

James Braid defined hypnotism as focused attention upon an “expectant dominant idea”, to the temporary exclusion (“abstraction”) of other thoughts.  Since that time, researchers have pondered the role of expectation in hypnotic responses.  Most therapists, and even more so stage hypnotists, probably share the common impression that the expectations of clients/subjects are an important factor, shaping how they respond to hypnotic suggestions.  However, human beings have a notable tendency toward “reductionism” and so debates like this tend to involve back-and-forth between all-or-nothing viewpoints, e.g., “hypnotism is all just expectation” versus “expectation doesn’t matter.”  An alternative, middle-way, would be the position that hypnotism is partially determined (“mediated”) by expectation, but not 100% so.  As one of the most prolific researchers in our field Professor Irving Kirsch has famously pointed out in his “response expectancy” theory of hypnosis, this would suggest that hypnotism is fundamentally related to the mechanism underlying the placebo effect, i.e., that hypnosis is a “non-deceptive mega-placebo”.  Again, that is very different from the notion that hypnotism is “just” a placebo, or the naive view that placebo effects are somehow “not real”.  People experience measurable physiological change and symptom remission after being given placebos and the process can be compared to the (anachronistic) concept of “waking suggestion” in the field of hypnotherapy.  Indeed, Braid introduced the concept of “hypnotism” (as opposed to Mesmerism) precisely on the basis of his observations of Victorian quack (“nostrum”) remedies, which modern researchers would consider examples of deceptive placebo remedies.  In other words, Braid saw people physically responding to treatments, such as animal magnetism or wearing “galvanic rings”, whose effects he and other sceptics attributed to expectation and suggestion, and subsequently developed hypnotism as a means of more honestly employing suggestion as an explicit technique in medicine.

In an important new experimental study ‘Response Expectancies: A Psychological Mechanism of Suggested and Placebo Analgesia”, Leonard S. Milling has carried out a very thorough and careful statistical analysis of the extent to which expectation appears to mediate the effect of hypnosis, imagination, and placebo, in the reduction of experimentally-induced pain among a sample of 172 college students (Contemporary Hypnosis, 26(2): 93-110, 2009).  All three interventions reduced pain substantially.  Traditional hypnotism and instructions to “imagine” were nearly equivalent, and both were almost twice as effective as the placebo.  This, and Milling’s other findings, lend additional support to the view that instructions to imagine may often be substituted for a traditional hypnotic induction, a central premise of Barber’s nonstate (“cognitive-behavioural”) theory of hypnosis.

Milling also found strong evidence supporting the role of expectation in mediating pain reduction.  However, the importance of expectation varied depending upon the techniques employed, calculated as follows,

  • Traditional hypnotic induction plus suggestion.  25%
  • Instructions to “imagine” plus suggestion.  29%
  • Placebo (an inert topical lotion).  41%

As Milling concludes, this appears to show that about 25% of the effectiveness of traditional pain-reduction hypnotherapy is due to expectation.  Expectation is an important factor but there may be one or two other factors involved which contribute more to the response, e.g., attention, motivation, imagination, or a trait of hypnotisability, etc.  By comparison, expectation contributed more substantially to the placebo effect, but still less than fifty percent, supporting the view that a cluster of factors contribute to the placebo response and it is not simply reducible to expectation alone, although this may turn out to be the single most important manageable factor involved.  Motivation, role-perception, attention, and other factors may be involved in the placebo response as well and Milling also points to the Pavlovian theory of classical conditioning which has been cited as providing another mechanism by which placebos (and hypnotism) may function.  For instance, a person who has previously received a real medication and experienced its effects may be more likely to respond to a similar-looking placebo because it acts as a reminder (conditioned stimulus) for the associated sense of pain relief (a conditioned response) – independently of the effect of expectation.  So previous experience of a real drug combined with high levels of expectation would probably produce a strong placebo response.  Likewise, tapping into remembered sensations (“sensory recall”) may combine well with expectation in eliciting certain hypnotic responses.

As expectation is a “cognitive” factor, these findings can be interpreted as supporting the view that the effect of hypnotherapy for pain reduction is “cognitively-mediated” in a manner overlapping with CBT interventions, which also stress the role of cognition in shaping the perception of pain.  In other words, although superficially different, hypnotherapy and CBT probably work, to some extent, in a similar manner, at least to some extent. 

As Milling points out, the usual cautions apply insofar as this was experimental pain induced with college students, etc., and therefore only provides an analogy (indirect evidence) for the mechanisms underlying pain relief among genuine therapy clients with genuine medical problems.  (Although, I think most researchers would consider it likely similar factors operate in the clinical setting as well.)

“In sum, this study substantiates that response expectancies are an important mechanism of hypnotic, imaginative and placebo analgesia.  The findings corroborated the view that the effect of hypnosis on pain is partially mediated by response expectancies.  The results also showed that the effect of a placebo on pain was largely, but not completely, mediated by response expectancies.  […] Thus, although the results of this study do not suggest that response expectancies are the final common pathway [as Kirsch has suggested] to pain relief, they do indicate that response expectancies are one of the major psychological mechanisms of suggested and placebo analgesia.”

So, as other studies have shown, the traditional hypnotic induction is probably not essential to hypnotic pain reduction, and client expectation is probably one of the most important factors which we should make use of.  Moreover, Barber, Spanos, Kirsch, and other cognitive-behavioural researchers have already discussed in some detail the possible means by which factors such as expectation may be systematically enhanced in hypnotherapy through methods tested in experimental settings such as role-modelling, manipulation of activating sensations, task-motivational instructions, etc.