EMDR: Do bi-lateral eye movements actually add anything to treatment?

by Michael D. Anestis, M.S.

Several months ago, I wrote an article discussing eye movement desensitization and reprocessing (EMDR).  The article focused on a special issue of the Journal of Anxiety Disorders in which several empirical studies investigated the degree to which EMDR produces reliably strong results and whether or not the treatment actually differs from traditional exposure techniques (e.g., do bi-lateral eye movements actually add anything to therapy?).  The conclusion of those studies and of the article I wrote was that EMDR produces reliably strong results in the treatment of post-traumatic stress disorder (PTSD), but that the evidence did not substantiate claims that bi-lateral eye movements play an important role in that process.  As such, although few people argue against the idea that EMDR can be helpful in treating PTSD, there are many who believe - based upon the evidence available - that EMDR simply represents another form of exposure, which we already know to be effective in the treatment of PTSD (although exposure is not the only important component of treatment).  If you are unfamiliar with EMDR, I recommend reading the original PBB article for a description of the treatment. 

The response to the PBB article was, to put it mildly, surprising.  I received a number of rather offensive and unprofessional emails that, quite frankly, left me substantially less inclined to spend time covering this topic.  That being said, my hope is that this time around, individuals who disagree with my conclusions and those of the researchers we discuss on this page will choose to engage in a civil conversation on the topic.  Readers will absolutely benefit from an open and honest discussion of all the available research on this topic and from the presentation of multiple discussions.  There is no need for anyone to be hostile.

A common complaint in the responses I received was that the research I covered was 10 years old.  This complaint, however, was a bit baffling.  First of all, age of the study has no bearing on the methodological quality of the study.  In other words, good research does not stop being good research ten years after it is published.  If it does, we are all in trouble, as all of our research builds upon work at least that old.  Secondly, I acknowledged the age of the research in my original article, directed readers to the APA Division 12 website for a summary of more recent findings, and based my decision to stop there on the fact that research in the past 10 years has not changed any of the conclusions discussed in my article.  In other words, just as was the case in 1999, the research indicates that EMDR can be very useful in the treatment of PTSD, but that the importance of bi-lateral eye movements remains in question. 

All that being said, I was inspired the other day to finally follow-up on the original EMDR article when a reader in the comment section of that article posted a reference to a study that claimed to provide  evidence that bi-lateral eye movements are, in fact, very important.  I greatly appreciated the effort on the part of the reader to further the conversation through the discussion of specific studies. My curiosity was piqued and I thought it might be a worthwhile cause to review this study and others for readers and to add more clarity to the question at hand. 

First, it seems worthwhile to point readers towards a more recent meta-analysis on this general topic.  In 2006, Guenter Seidler and Frank Wagner published a summary of studies comparing EMDR to trauma-focused cognitive behavioral therapy (CBT) for the treatment of PTSD in the journal Psychological Medicine.  Much like the meta-analysis discussed previously on PBB, the authors found that the two treatments produced equivalent results for PTSD patients.  Also like the earlier meta-analysis, the authors concluded "what remains unclear is the contribution of the eye movement component in EMDR to treatment outcome."  In other words, we once again were presented with evidence that EMDR is very useful in treating PTSD, but were still left without evidence supporting the claim that eye movements rather than exposure explained these results.

Fortunately, researchers have attempted to answer this question more directly in recent years.  The article referenced in the comment section of the earliest EMDR article was published in 2008 in the Journal of Anxiety Disorders by Christopher Lee and Peter Drummond.  The authors randomly assigned 48 undergraduates to either an eye movement or eye stationary condition and to two types of therapist instructions: reliving an unpleasant memory or distancing oneself from the experience.  Their goal was to determine whether eye movements, type of therapist instruction, or the interaction of both impacted the degree to which participants were distressed by their memory and vividly recalled the memory.  What they found was that, contrary to hypotheses (and prior work), eye movements had no impact on vividness, but they did result in less distress in response to the distressing memory.  They also found that individuals in the eye movement-distancing condition experienced less vividness.

Sounds promising, right?  When I simply read the abstract, I thought so too.  A closer look at the data, however, reveals some fatal flaws in this study.  First of all, the sample was comprised of undergraduates, none of whom met criteria for PTSD or any other mental illness and none of whom exhibited elevations on symptom measures of any mental illness.  In other words, in a study looking to examine the impact of eye movements on responses to traumatic memories, the authors utilized a sample in which none of the participants reported a history of trauma or symptoms of PTSD.  The problems did not stop there either.  The measure of distress utilized in this study was Subjective Units of Distress (SUDs), a common measure of emotional response in which a participant is asked to rate his or her distress from 0 to 10 (or some other scale).  Participants in this study were told to only consider traumatic events that resulted in a SUDs of 6 or less.  Any event more distressing than that was not allowed and any participant who experienced elevations of greater than 6 during the study were excluded from the analyses. 

There are several problems with using this type of sample and exclusion criteria in a study like this.  The authors were attempting to determine whether eye movements differ from traditional exposure; however, in order for exposure to work, a client must actually experience anxiety.  The process of exposure involves encountering anxiety in response to a feared stimulus.  By repeating this process over and over again, the fear response dampens and eventually extinguishes so that the individual no longer experiences fear when in the presence of or thinking about the stimulus.  In a study in which participants are not allowed to feel more than mild to moderate anxiety, exposure is handicapped and a comparison between eye movements and exposure is impossible.   The result is a study in which, rather than determining if eye movements differ from exposure in the treatment of PTSD, the authors examined whether eye movements impact response to unpleasant memories in a sample of undergraduates with no reported history of trauma or mental illness.  Given that the results did not consistently conform to the authors hypotheses even in this sample, it seems off base to look at this study and conclude that it supports that idea that the mechanism of change in EMDR is different from exposure in the treatment of PTSD.

Now, despite these flaws, this study did have some legitimate value.  The authors used some strong methodology (e.g., random assignment) and some sophisticated statistical analyses.  If the same study was conducted on a clinical sample of individuals diagnosed with PTSD through structured clinical interviews and the authors were able to demonstrate that the actual symptoms of PTSD were reduced more in an eye movement condition than in a no eye movement condition, the implications of the data would be much more substantial.

The Lee and Drummond (2008) study is not the only attempt that has been made to address this question.  In 2008, Raymond Gunter and Glen Bodner published results from three studies examining the impact of eye movements on unpleasant memories in Behaviour Research and Therapy.  In study 1, which was conducted on 36 undergraduates, none of whom met criteria for any mental illness or exhibited elevations on any measure of traumatic experiences, the authors found that eye movements reduced vividness of unpleasant memories when the participants were asked to keep the unpleasant memory in their mind while engaging in the eye movements.  They also found that eye movements resulted in increased arousal in response to the unpleasant memory.  Some individuals might argue that this means that eye movements cause the participant to focus more intensely on the unpleasant memory, thereby increasing the potential impact of exposure; however, higher arousal scores were not associated with larger decreases in ratings from pre- to post-intervention, so that explanation simply has no empirical support.  

In study 2, conducted on 36 undergraduates, none of whom met criteria for any mental illness or exhibited elevations on any measure of traumatic experiences, the authors found that eye movements reduced the vividness of unpleasant memories significantly more than a no eye movement control condition.  Contrary to expectations, however, horizontal and vertical eye movements produced equivalent results.  The latter finding, of course, does not indicate that eye movements have no utility, but rather that the type of eye movement might be irrelevant.  As was the case in study 1, eye movements resulted in greater arousal in response to unpleasant memories.  At one week follow-up, the benefits of eye movements were substantially reduced, but still greater than those experienced in the no eye movement condition. 

In study 3, conducted on 72 undergraduates, none of whom met criteria for any mental illness or exhibited elevations on any measure of traumatic experiences, the authors reported several interesting findings.  Participants were randomly assigned to engage in one of three tasks while keeping an unpleasant memory in their mind: eye movements, listening to a simple speech recording, or copying drawings of complex geographic shapes.  The authors found that the drawing condition produced greater benefits than did the eye movement condition and the speech condition and that the latter two did not differ from one another.  All three conditions resulted in reduced vividness, emotionality, and completeness of unpleasant memories and the authors argued that the results, in general, support the idea that, by taxing an individuals central-executive component of their working memory, these tasks produce important therapeutic benefits.

First of all, the authors should be applauded for attempting to empirically test the utility of eye movements in treatment.  Now, that being said, once again it is unclear how these findings actually support the contention that eye movements add important components to treatment that differentiate EMDR from exposure in the treatment of PTSD.  None of the participants in any of these studies had PTSD.  They did not even exhibit elevations on any of the symptoms of PTSD.  They were all healthy undergraduates and, in the first study we reviewed, they were not even allowed to experience substantial anxiety.  Certainly, there is some evidence that eye movements reduce the vividness of unpleasant memories in healthy samples of individuals who are not thinking about events that trigger clinical responses (although this was not consistently supported in these studies), but is that the goal of treatment in PTSD?  PTSD is characterized by a vast array of debilitating symptoms.  Until research can indicate that, in samples of individuals who actually meet criteria for the disorder, eye movements produce a distinct impact on these symptoms relative to exposure alone, it makes absolutely no sense to assume that they do. 

Again, as I have said before, my argument is not that EMDR fails to produce results in the treatment of PTSD.  My argument is that the research thus far does not support the contention that eye movements add significant impact to treatment relative to traditional exposure.  Perhaps they do, but until there is actual evidence supporting this contention, debate on the matter will continue and scientifically-minded individuals will persist in their questioning of the phenomenon. If you know of studies that directly address this question, feel free to send them my way.  The goal of PBB is to provide readers with digestible summaries of clinical research so that they can make informed decisions on mental health treatment and better understand the nature of mental illness.  We do not campaign for or against particular forms of treatment.  Data drive our articles and we are more than happy to cover a wide range of empirical studies.

If you would like to learn more about the treatment of PTSD, we recommend the following items, all of which are available through our online store of scientifically-based psychological resources:

• Prolonged Exposure Therapy for PTSD: Emotional Processing of Traumatic Experiences Therapist Guide by Edna Foa, Elizabeth Hembree, and Barbara Rothbaum

• Cognitive Processing Therapy for Rape Victims: A Treatment Manual by Patricia Resick and Monica Schnicke

• Prolonged Exposure Therapy for Adolescents with PTSD Emotional Processing of Traumatic Experiences, Therapist Guide by Edna Foa, Kelly Chrestman, and Eva Gilboa-Schechtman

Mike Anestis is a doctoral candidate in the clinical psychology department at Florida State University