The Rest of the Story

Every once in a while I take some ideas out of the tumbler before they are polished, and inflict them on the world, complete with grit and rough surfaces.

Those who read the Ann Arbor News, and perhaps some other sources, may have heard about this study.  One of the neuroscientists at the University of Michigan did a brain scan study on the placebo effect, showing that there are visible changes in the functioning of the brain when a person has a positive response to a placebo.  (For an interesting tangent, see also this study concerning gender differences in pain perception.)

U-M neuroscientist Jon-Kar Zubieta, M.D., Ph.D.

[...] The study provides the first direct evidence that the brain's own pain-fighting chemicals, called endorphins, play a role in the phenomenon known as the placebo effect — and that this response corresponds with a reduction in feelings of pain.

Previous studies at U-M and elsewhere have shown that the brain reacts physically when a person is given a sham pain treatment, which they believe will help them.

The study provides the first direct evidence that the brain's own pain-fighting chemicals, called endorphins, play a role in the phenomenon known as the placebo effect — and that this response corresponds with a reduction in feelings of pain.

Previous studies at U-M and elsewhere have shown that the brain reacts physically when a person is given a sham pain treatment, which they believe will help them.

But the new study is the first to pinpoint a specific brain chemistry mechanism for a pain-related placebo effect. It may help explain why so many people say they get relief from therapies and remedies with no actual physical benefit. And, it may lead to better use of cognitive, or psychological, therapy for people with chronic pain.

The results will be published in the August 24 issue of the Journal of Neuroscience by a team from the U-M Molecular and Behavioral Neurosciences Institute (MBNI). The research was funded by the National Institutes of Health.

“This deals another serious blow to the idea that the placebo effect is a purely psychological, not physical, phenomenon,” says lead author Jon-Kar Zubieta, M.D., Ph.D., associate professor of psychiatry and radiology at the U-M Medical School and associate research scientist at MBNI. “We were able to see that the endorphin system was activated in pain-related areas of the brain, and that activity increased when someone was told they were receiving a medicine to ease their pain. They then reported feeling less pain. The mind-body connection is quite clear.” [...]

Naturally, it would be a matter of great practical interest to learn how the placebo effect works, so that it can be harnessed for therapeutic effects.  For example, it would be great if we could learn to use the placebo effect to reduce pain, reliably, without drugs.  

However, it seems to me that there could be another, perhaps more important, practical application to this type of research.  One possibility is this: One of the biggest problems in drug discovery is the problem of separating active drugs from those with no effect.  When studying conditions with a high placebo response rate, the placebo effect can be quite a problem.  It can lead to the unnecessary discarding of drugs that have a definite positive effect on a small subgroup of patients.  If we had some objective means of discerning when a study participant was having a placebo response, we might be able to salvage some of the new molecular entities that otherwise could not be demonstrated to separate from placebo.  

But wait!  There's more!  In studies of the effectiveness of psychotherapy, the placebo effect is especially difficult to factor out of the analysis.  We know that some studies have shown that patients placed on a waiting list for psychotherapy show improvement that appears to be spontaneous; that is, they show improvement that cannot be attributed to a therapeutic intervention.  They do not even receive a placebo.  Why do these patients improve?  Could the improvement be like a placebo response, in some obscure way, or is there some other factor at work?

There are two things about the wait-list patients that are different than people with the same problems, who have not been placed on a waiting list.  One is that the patients on the waiting list have a positive expectation of change.  That would be similar to the effect of a placebo.  Another difference is that the waiting-list patients have taken an important step toward recovery: they have made a decision to do something about their problems.  

These musings suggest two new lines of research.  For one, would it be possible to demonstrate that there are neurochemical changes in the brains of people who have a positive expectation of improvement, even when no placebo is given?  For another, would it be possible to detect neurochemical changes in the brains of people who have made a decision to try to get some help?  

Personally, I think the latter question is more interesting.  Psychotherapists have known for decades that there is something powerful about the act of making a decision.  Indeed, they devote a great deal of effort to coaxing people to make firm, unequivocal decisions to make changes in their lives.  Could it be that it is not so much the changes the patient make, that are therapeutic, but the decision itself?  That is an interesting hypothesis, although it would be difficult to test.

Is it possible to detect brain changes in response to psychotherapy?  Yes, for example, this study shows that patients with obsessive-compulsive disorder, who are treated with and who respond to psychotherapy, have brain changes that are smaller than, but similar to, changes that are seen in those who respond to medication.  

Caudate glucose metabolic rate changes with both drug and behavior therapy for obsessive-compulsive disorder.

Baxter LR Jr, Schwartz JM, Bergman KS, Szuba MP, Guze BH, Mazziotta JC, Alazraki A, Selin CE, Ferng HK, Munford P, et al.

Department of Psychiatry, UCLA School of Medicine.
Arch Gen Psychiatry. 1992 Sep;49(9):681-9.

We used positron emission tomography to investigate local cerebral metabolic rates for glucose (LCMRG1c) in patients with obsessive-compulsive disorder before and after treatment with either fluoxetine hydrochloride or behavior therapy. After treatment, LCMRG1c in the head of the right caudate nucleus, divided by that in the ipsilateral hemisphere (Cd/hem), was decreased significantly compared with pretreatment values in responders to both drug and behavior therapy. These decreases in responders were also significantly greater than right Cd/hem changes in nonresponders and normal controls, in both of whom values did not change from baseline. Percentage change in obsessive-compulsive disorder symptom ratings correlated significantly with the percent of right Cd/hem change with drug therapy and there was a trend to significance for this same correlation with behavior therapy. By lumping all responders to either treatment, right orbital cortex/hem was significantly correlated with ipsilateral Cd/hem and thalamus/hem before treatment but not after, and the differences before and after treatment were significant. A similar pattern was noted in the left hemisphere. A brain circuit involving these brain regions may mediate obsessive-compulsive disorder symptoms.

This demonstrates, at least, that it is possible to use PET scans to see changes in brain function that occur as a result of psychotherapy.  We do not know how generalizable this finding is, so we do not know if it would be possible to see such changes in conditions that are not as well defined as OCD.  

Even if the changes are great enough to be discerned by PET scans, I cannot think of a study protocol that would enable us to separate all the pertinent variables and test the hypothesis cleanly.  So maybe this is not such a great idea.  But it is interesting to think about.