Re: differential mechanism/ SUDS/ eye movements

Hello Louise. Sorry for the delay in continuing this discussion, but several things came up the demanded my attention. Here are my comments to Part 1 of your November 11th posts. You mention your participation with Steve Taylor in comparing the effects of EMDR, relaxation, and exposure. I was wondering if you were in any position to share the results from this study. Unfortunately, I was not able to attend the World Congress, and so I missed Dr. Taylor's presentation of the data.

Onto the substantive issues raised in your post. First to the mediator/moderator distinction. I'm going to try to be real careful here, as I do not want to put words in your mouth. I had proposed that you and I may disagree about whether treatment outcome for EMDR and standard exposure therapy are mediated by the same or different mechanisms. To be more specific, my position is that we should assume EMDR and exposure therapy achieve their effects on outcome through the same mechanism(s) until there is evidence to indicate otherwise. You responded by suggesting that we don't disagree on the issue of mediation, and then made the counter-proposal that any additional mechanism associated with eye movements may reflect the operation of a moderating variable.

1. Am I correct in understanding that you and in agreement that treatment outcome in EMDR and exposure therapy are mediated by the same mechanisms? If I am not correct in this, please indicate what the different mechanisms are your reasons for thinking there are different mechanisms.

2. Assuming that I am correct and that we agree that outcome with EMDR and exposure therapy are mediated by the same mechanisms, please say more about your proposal of the operation of a moderator variable. What client-types do you think would most benefit from the addition of eye movements? Based on the EMPIRICAL response you provided to my question about whether there were any THEORETICAL reasons to think that a dual task would enhance memory processing, I am inclined to infer that you would predict that adding eye movements (or some other "dual task") would result in better outcome for highly dissociative individuals. Am I correct in this inference? If so, what is the underlying theoretical rationale for this idea? Stated differently, could you please elaborate on what you think the moderating variable is, what other variable or relationship is being moderated, why the moderator has its effect, and indicate the predicted pattern of outcome? What kind of study would constitute a fair test of your hypothesis? Are there any client-types you would predict for whom eye movements would be harmful?

Thanks for the three references (DePrince & Freyd, 1999, Kavanaugh et al., 2001; and van den Hout et al., 2001). Although I haven't had a chance to collect them yet, I look forward to doing so, and seeing whether they can answer the questions I posed about the similarity/differences between the notions of dual attention and distraction, and whether there are any theoretical reasons to think that that a dual task would enhance memory processing rather than detract from it.

Finally, onto your comments about why there has been resistance to my proposal that, in the absence of data to showing that any of the unique elements of EMDR add to treatment outcome, it is most parsimonious to assume that they achieve their effects through the same mechanisms.
Now, I'm not sure if the comments you provided represent your own position, or if they are your observations of what other people have said. However, before addressing your specific comments, let me better lay out my position.

In comparing any two therapies, each therapy can be broken down into two components. Component A represents common features that are shared by both therapies while Component B represents the specifics of the treatment that distinguish one therapy from another. In addition to the features of therapy that are frequently subsumed under the rubric of "common factors" (e.g., meeting with an empathic and knowledgeable professional, a chance to express one's feelings and concerns, etc), EMDR and exposure therapy also share the common feature of imaginal exposure to the trauma memory. Thus, imaginal exposure is part of Component A for both treatments. What distinguishes the two treatments are such things such as the specific manner in which imaginal exposure is implemented (e.g., brief vs. long exposure, repeating the trauma memory vs. allowing the patient to "free associate" the content, the use of eye movements/dual task vs. not doing so, etc).

We know from prior research that both exposure therapy and EMDR work. Prior research also tells us that, thus far, there is no evidence that any of the unique features of EMDR improve outcome (e.g., no good evidence that eye movements are associated with better outcome than no
eye movements, no evidence that utilizing installation trials is any better than just adding more desensitization trials, etc.). We can represent this state of affairs by saying that:

(1) A + B(EMDR) works
and that
(2) A + B(exposure therapy) works.

In our pursuit of trying to identify the effective ingredients, then, it is most parsimonious to assume that the effective component of treatment lies with Component A. Until there is evidence for the additional effects of specific elements of Component B, these features are best viewed as just bells and whistles.

You suggest that "some of the resistance to, and confusion about this conclusion stems from understanding of exposure protocols, and clinical observations of the free association element in EMDR." Again, I'm not sure whether you are stating this simply as an observation that some people feel this way (I've heard and read similar things), or whether it represents your own position. In either case, I hope that my extended description above makes it clear that this argument boils down to asserting (without any evidence to back it up) that some Component B elements of EMDR are active ingredients, rather than bells and whistles. While this MAY be true, the burden of proof falls on whoever is making this claim to show that specific Component B elements of EMDR are effective. The burden of proof can NEVER fall on the skeptic who suggests that the Component B elements are inert for the simple reason that it is logically impossible to prove that something doesn't work.

Furthermore, given that the assumption of a single set of active ingredients (those associated with Component A) is more parsimonious than assuming two sets of active ingredients (those associated with Component B), I restate my original position. Until there is evidence for any of the unique features of EMDR, it is most parsimonious to assume EMDR and exposure work through the same mechanisms. The simple fact that procedural differences can be identified does not mean the procedural differences are responsible for recruiting different mechanisms.

Let's take, for example, the assertion that EMDR differs from exposure therapy by allowing free-association to occur. Does this mean that free association is related to the efficacy of EMDR? Of course not. We would need to conduct the appropriate study to determine if this is true: Randomly assign subjects to standard EMDR (permitting variable content), "fixed" EMDR (always focussed on the index trauma memory), and a control group (e.g., waitlist). There are at least three possible outcomes of interest. (1) The free association condition is superior to the fixed content condition, which would provide good evidence for a specific feature of EMDR. (2) No difference between the two EMDR variations, which would not support the hypothesis and leave us where were prior to the study. (3) Fixed EMDR is superior to standard EMDR. This last possibility, were it to be observed, would be particularly devastating, as it would directly refute the hypothesis that free association improves treatment outcome.

Finally, I'd like to provide a historical perspective on your comment regarding the possible evolution of the concept of exposure therapy if it were to encompass free association. I would argue this represents returning to an earlier notion of how broad a concept exposure therapy is, rather than an evolution of the concept.

Dollard and Miller (1950) very clearly re-interpreted the Freudian idea of Free Association into a general exposure therapy model (please excuse the archaic use of the masculine pronoun to represent men and women):

"The first of the new conditions of learning imposed on the patient is that of free association. Under the rule of free association the patient is required to say everything that comes to mind immediately...He is not to reject any thought whatever...If imageal responses occur, he is to describe them since they may be the cues to which emotional responses are attached… Furthermore, the patient must *volunteer* information, i.e., he must take the risk of saying what he must say in the face of whatever hampering forces there are. Only thus can extinction of fear occur... The goal of the benign conditions we have been discussing [non-punitive responses from the therapist, no matter what the patient says] is 'freedom of thought,' that is to make the patient free within his or her own mind to consider every possible alternative course of action... How is freedom of thought achieved in the therapeutic situation? The answer to this question from the standpoint of behavior theory is clear. The patient must pronounce the forbidden sentences *while being afraid*...Punishment does not occur and the fear attached to the formerly forbidden sentences is not reinforced...Gradually his fear of talking about matters formerly forbidden is extinguished.... The extinction effects which are first attached to talking out loud generalize swiftly to 'talking without voice' (thinking). At this point, the patient has achieved his valuable freedom to think." (p. 241, 242, 249, 250, emphasis in the original).

Skinner (1953) uses similar language while interpreting talk therapy from his operant perspective (again, please excuse the sexist language):

"The commonest current technique of psychotherapy is due to Sigmund Freud...[I]t may be described simply in this way: the therapist constitutes himself a nonpunishing audience...As the therapist gradually establishes himself as a nonpunishing audience, behavior which has
hitherto been repressed begins to appear in the repertoire of the patient. For example, the patient may recall a previously forgotten episode in which he was punished... The patient may also begin to describe current tendencies to behave in punishable ways... Nonverbal behavior which has previously been punished may also begin to appear...If such behavior has been wholly repressed, it may at first reach only the covert level; the individual may begin to behave verbally or nonverbally 'to himself' - as in fantasying punished behavior. The behavior may later be brought to the overt level. The patient may also begin to exhibit strong emotions: he may have a good cry, make a violent display of temper, or be 'hysterically' silly. If in the face of such behavior, the therapist is successful in maintaining his position as a nonpunisher, the process of reducing the effect of punishment is accelerated. The appearance of previously punished behavior in the presence of a nonpunishing audience makes possible the extinction of some of the side effects of punishment [previously identified to include escape, passive resistance, fear, anxiety, anger, and depression, drug addiction]. This is the principle result of such therapy. Stimuli which are automatically generated by the patient's own behavior become less and less aversive and less and less likely to generate emotional reactions. The patient feels less wrong, less guilty, or less sinful. As a direct consequence he is less likely to exhibit the various forms of operant behavior which, as we have seen, provide escape from such self-generated stimulation." (370, 371).

Alan Wilson and colleagues (Wilson & Smith, 1968a, b; Boudewyns & Wilson, 1972) developed and evaluated a form of systematic desensitization that replaced the use of a hierarchy with free association. According to Wilson and Smith (1968b, as cited in Boudewyns & Wilson, 1972, p. 260):

"Learning theory would predict that a patient's associations would begin by being on stimulus gradients with, but far down the gradients from, the anxiety foci (Dollard & Miller, 1950, pp. 51-53). The generalization gradients would, of course, presumably be verbally mediated (Osgood, 1953, pp. 695-699). Deconditioning of anxiety to distant associations should generalize to the central conflicts, thus allowing the patient to move closer to the areas of greatest conflict as therapy progressed. In other words, the patient would be expected to build and work on one or more anxiety hierarchies spontaneously in each treatment session."

Boudewyns and Smith (1972) continue with, "The advantage of this technique over the more structured systematic desensitization may lie in the flexibility it affords the therapist in dealing with anxiety which is not clearly situation specific, that is, when the stimuli that produce the anxiety are not easily identified." (p. 260).

To evaluate the efficacy of this free association desensitization, Boudewyns and Smith compared it with implosive therapy and milieu therapy among a group of VA inpatients. Thus, all participants received the milieu therapy and two groups received additional individual therapy comprised of either free association desensitization or implosive therapy. The results suggest both add-on treatment added to outcome and implosive therapy was more effective than free association desensitization, there are several limits to this based on contemporary standards. For example, the sample is not well described, the primary outcome measure was the MMPI (not very specific and not very sensitive to change), and the data for milieu therapy group was collected after the data were collected for the other two conditions. However, important to this conversation, the data for the two experimental treatment conditions provide no support for the hypothesis that free association resulted in better outcome than intensive and focussed exposure.

As may be seen, there is precedent for viewing free association in terms of exposure therapy although it may not be considered to be optimal under most conditions.

References

Boudwyns, P. A., & Wilson, A. E. (1972). Implosive therapy and desensitization therapy using free association in the treatment of inpatients. Journal of Abnormal Psychology, 79, 259-268.

Dollard, J., & Miller, N. E. (1950). Personality and psychotherapy. New York: McGraw-Hill.

Skinner, B. F. (1953). Science and human behavior. New York: Macmillan.

Wilson, A. E., & Smith, F. J. (1968a). Counterconditioning therapy using free association: A pilot study. Journal of Abnormal Psychology, 73, 474-478.

Wilson, A. E., & Smith, F. J. (1968a). Counterconditioning therapy using free association: A case study. Proceedings of the American Psychological Association, 3, 529-530 (summary).

Replies:

There are no replies to this message.

| Behavior OnLine Home Page | Disclaimer |

Copyright © 1996-2004 Behavior OnLine, Inc. All rights reserved.