Edited by: Allan V. Kalueff, ZENEREI Institute, USA; Guangdong Ocean University, China; St. Petersburg State University, Russia
Reviewed by: Carsten T. Wotjak, Max-Planck-Institute of Psychiatry, Germany; Jason Moser, Michigan State University, USA
*Correspondence: Olivier A. Coubard
This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution and reproduction in other forums is permitted, provided the original author(s) or licensor are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.
Since the seminal report by Shapiro that bilateral stimulation induces cognitive and emotional changes, 26 years of basic and clinical research have examined the effects of Eye Movement Desensitization and Reprocessing (EMDR) in anxiety disorders, particularly in post-traumatic stress disorder (PTSD). The present article aims at better understanding EMDR neural mechanism. I first review procedural aspects of EMDR protocol and theoretical hypothesis about EMDR effects, and develop the reasons why the scientific community is still divided about EMDR. I then slide from psychology to physiology describing eye movements/emotion interaction from the physiological viewpoint, and introduce theoretical and technical tools used in movement research to re-examine EMDR neural mechanism. Using a recent physiological model for the neuropsychological architecture of motor and cognitive control, the Threshold Interval Modulation with Early Release-Rate of rIse Deviation with Early Release (TIMER-RIDER)—model, I explore how attentional control and bilateral stimulation may participate to EMDR effects. These effects may be obtained by two processes acting in parallel: (i) activity level enhancement of attentional control component; and (ii) bilateral stimulation in any sensorimotor modality, both resulting in lower inhibition enabling dysfunctional information to be processed and anxiety to be reduced. The TIMER-RIDER model offers quantitative predictions about EMDR effects for future research about its underlying physiological mechanisms.
In 1987, Francine Shapiro observed on herself that rhythmic eye movements may cause cognitive and emotional changes. From that observation, she developed a Cognitive and Behavioral Therapy (CBT)-inspired protocol called Eye Movement Desensitization and Reprocessing (EMDR; initially EMD then EMDR) to treat anxiety disorders, particularly Post-Traumatic Stress Disorder (PTSD; Shapiro,
After a brief overview of procedural aspects of EMDR and theoretical hypotheses, I will describe how the scientific community is divided about EMDR. Then, I will present theoretical and technical tools currently used in movement research, and explore how such quantitative tools may help to put forth our understanding of EMDR neural mechanism.
EMDR is a complex protocol in the sense that it involves several components: behavioral, cognitive, emotional and physical (Shapiro,
First, EMDR effects may be attributed to exposure, which induces desensitization on its own (Lohr et al.,
Second, switching between exposure and metacognition reinforces self-mastery and self-efficacy (Bandura,
Third, identify physical sensations associated with emotion (e.g., stomach pain associated with fear) and quantify the level of disturbance on a subjective scale—the subjective units of discomfort scale (SUDS; Wolpe,
Fourth, cognitive reprocessing is a central aspect of EMDR protocol. The patient is invited to verbalize the negative cognition he/she may covertly express about him/herself, which acts as dysfunctional schemes. As a matter of fact, he/she becomes aware of potential irrational character of such schemes and of their deleterious impact in daily life. During the EMDR protocol, the patient is then invited to replace this negative cognition by a new positive one, which is reinforced using another subjective scale—the validity of cognition (VOC)—to quantify his/her belief in such cognition. This positive cognition acts as new narrative content of self-perception and self-esteem, which facilitates the therapeutic process (Beck,
Fifth, one aspect of PTSD is that traumatic memory information is dissociated and disseminated into fragments. Here, we define traumatic memory as a blend of multi-sensory images (visual, auditory, somatosensory, etc.), negative cognition, negative emotion, and unpleasant physical sensations. Those fragments behave as labile and dysfunctional information until they can be reconnected—in other words integrated. The goal of EMDR is specifically to integrate these fragments (van der Kolk,
Sixth, EMDR is not only compatible with CBT but also with free association of psychoanalysis (Wachtel,
Seventh, EMDR is in line with meditative practice (Kabat-Zinn,
Eight, a final important property of EMDR protocol is bilateral stimulation. Originally, bilateral stimulation has been obtained through pursuit eye movements in response to the therapist fingers’ movement or using
Besides potential EMDR effects inherent in its procedural aspects, several hypotheses have been advanced to account for the mechanisms underlying EMDR, such as orienting response (Armstrong and Vaughan,
In orienting response, it has been suggested that, before intervention, access to traumatic memory is only available through a configuration of automatic physiological states (Shapiro,
Behavioral models have also suggested that orienting response acts as a learning that breaks with avoidance or flight behaviors (Armstrong and Vaughan,
In neurobiological models, left-right eye movements would act as REM known to occur during the paradoxical sleep phase and hypothesized to collect episodic recent events and to integrate them into semantic memory (Bergmann,
It has been suggested that EMDR may act as hypnosis or some sort of suggestibility. However hypnosis has been rejected to explain EMDR effects for at least three reasons. First EMDR has shown significant results in PTSD, which is not the case of hypnosis (Shapiro,
The best EMDR effects are produced by the intermediate situation in which the patient is half in the past and half in the present, in other words not too distracted and not too capable of negative association. Distraction is an important feature of attentional control (Parasuraman,
A final issue that deserves discussion is the hypothesis according to which EMDR may act as hemispheric synchronization. Under normal conditions, neurological balance is achieved by excitatory and inhibitory mechanisms (Pavlov,
Such hypothesis is partially based on hemispheric asymmetry in emotional processing. The left hemispatial bias in perceptual face processing is reduced vs. enhanced in high-depressed vs. high-anxious participants, compared to low-depressed and low-anxious participants, respectively (Heller et al.,
Since the original report by Shapiro (
The argument to support EMDR enthusiastically as therapeutic intervention in anxiety disorders can be analyzed according to three levels.
At level 1, qualitative reports and uncontrolled studies (i.e., without control group) have suggested that EMDR might have interesting therapeutic results (e.g., Marquis,
At level 2, other studies have showed that EMDR intervention in the experimental group leads to better outcomes than in the control group without any treatment (e.g., patients on a waiting list). Specifically in PTSD and trauma, several reports have documented that patients benefiting from EMDR as compared to patients without treatment, yield significantly lower SUDS (e.g., Wilson et al.,
Considering other anxiety disorders, EMDR has shown better results than no treatment in spider phobia as evidenced by lower SUDS and higher VOC (Bates et al.,
These early studies in EMDR history have had the merit to point out the importance of treatment duration by showing that 5–12 sessions of EMDR are more efficient than only two, as well as the absence of correlation between subjective, psychometric and physiological measures (Cahill et al.,
At level 3, EMDR intervention has been found to be superior to non validated treatments. Non validated treatments refer to interventions which are not well defined in their content, and/or not standardized in their methods, and/or not validated by Random Clinical Trials (RCT). Image Habituation Training (IHT; Vaughan and Tarrier,
Taken together, there are arguments in the literature to suggest that EMDR might be a useful strategy for treating PTSD/trauma and eventually other anxiety disorders, when it is compared to a control group without treatment or a control group benefiting from other non validated treatments. However such studies present some limitations in their methodology and the potential impact of their results, which I now discuss.
In intervention studies, the changes observed between pre- and post-test periods may be attributed to multiple factors: (1) passage of time, which involves healthy aging, natural remission or spontaneous recovery; (2) test-retest effect, which includes learning effect (the fact of repeating the assessment improves performance) and regression to the mean (the tendency of extreme scores in pre-test period to move away from extremes in post-test period); (3) placebo effect, which refers to volition and expectancy (the will and expectation that things will be better improves performance, particularly for subjective measures such as SUDS and VOC), and for which it is worth recalling that it yields on its own to effect size of 30%–40%; (4) trainer effect, which involves dimensions like the quality of relationship between therapist and patient, empathy, healing, and in the extreme the guru effect (a particular therapist showing particular influence that is nevertheless not reproducible with other practitioner); (5) treatment effect without specificity, when the intervention is either complex (like EMDR) or not well defined or when the control group is not specific (i.e., waiting list, no treatment or non validated treatment for targeted disorder); and (6) treatment effect with genuine specificity, when the treatment is compared to previously validated treatment for targeted disorder (like CBT for PTSD) and when the assessment uses standardized measurements (Gastright,
Given these methodological aspects, the argument to exhibit scepticism about EMDR treatment can be assessed on five levels.
At level 1, qualitative reports and uncontrolled studies (see “Half the Scientific Community is Enthusiastic About EMDR” Section, level 1) do not rule out effects to be due to factors 1–5.
At level 2, studies comparing EMDR to no treatment (see “Half the Scientific Community is Enthusiastic About EMDR” Section, level 2) neutralize factors 1–2 but not factors 3–5.
At level 3, studies comparing EMDR to non validated treatments (see “Half the Scientific Community is Enthusiastic About EMDR” Section, level 3) control factors 1–4 but not factor 5.
At level 4, one way to conclude about EMDR effectiveness and specificity is to directly compare this treatment to another validated treatment in a Randomized Clinical Trial (RCT). According to Cahill et al. (
Since 1999, further RCTs have shown that EMDR is either as effective as CBT or CBT variants such as CBT-Prolonged Exposure (PE) in adults (Leiner et al.,
In other than PTSD/trauma anxiety disorders, EMDR has been demonstrated to be less effective that Participant Modeling (PM; Bandura et al.,
At level 5, a final way to study EMDR effectiveness is to isolate and examine each of its procedural aspects and their potential effect in a so-called dismantling study (Cahill et al.,
Taken together, the superiority of EMDR as compared to other validated treatments in PTSD/trauma and other anxiety disorders still remain to be demonstrated. A few studies have evidenced that EMDR is as effective as CBT or CBT variant validated treatments, but there is little evidence that EMDR and its added-value (bilateral stimulation) are more effective than other validated CBT or CBT variants in anxiety disorders.
The main reason why the scientific community is divided about EMDR is that its underlying neural mechanism is unknown. As elegantly described by Servan-Schreiber (
A first issue that is worth mentioning about the enthusiasts is ethical. Many authors are biased in their intentions by the fact that they do business with EMDR, should it be by practicing, supervising, training other practitioners, or selling products for bilateral stimulation. Even when endorsing an academic position and/or not declaring their conflict of interest, they can often be seen using the term “client” instead of “patient”, leaving the reader dubitative about the objectivity of their findings and proposals.
The second type of error by enthusiasts is methodological. In “the Scientific Community is Divided About EMDR” Section, I have reviewed the different aspects of the study designs that need improvement. As previously recommended by Cahill et al. (
The third error by the enthusiasts is theoretical. Most explanations to account for EMDR effects (see “Theoretical Hypotheses” Section) are—as is characteristic of certain kinds of psychological theorizing—purely verbal. As such, they do not allow researchers to make predictions about quantitative outcomes that can be linked to underlying physiological mechanisms.
Shortcomings also emerge on the side of sceptics. The first error is historical. Whereas Shapiro made the clinical discovery of EMDR, it is unfair to expect from her that she also unveil its underlying physiological mechanisms. Clinical psychology and basic research in physiology are two different jobs, which are each time consuming and based on specific knowledge and skills, thus usually made by different people. As mentioned above for asepsis and microbes, it usually takes years or decades for a clinical discovery to be elucidated in its intimate mechanisms. Iproniazide initially used as antituberculous drug was found in 1958 by Nathan Kline to have antidepressant properties. But it took years for basic researchers to demonstrate the inhibition of tritiated-noradrenaline uptake by imipramine (Glowinski and Axelrod,
The second defect of the skeptics is methodological, by using military veteran stress as a gold standard of PTSD. Contrary to received ideas, I suggest that military posttraumatic stress disorder is not a good model of PTSD, explaining contrasting results between veterans (e.g., Boudewyns et al.,
A final drawback that needs to be mentioned on the skeptical side is to deny that eye movements may have anything to see with emotion, which leads me to the development on the underlying physiological mechanisms of EMDR.
From the physiological viewpoint, there is no doubt that eye movements can induce emotional shifts, and that reversely emotional shifts can be accompanied by eye movements as motor consequences. Anyone can make this observation and, on that note, writers have as usual preceded scientists. For example, the French writer Guy de Maupassant (
The reason why eye movements and emotional shifts can be associated as cause or effect is simply that they share common neural circuitry. In a recent article (Coubard,
To conclude this section, the critical question in EMDR is not to ask whether eye movements and emotion are linked since it is indeed a physiological fact, but rather to determine how a model of attention and bilateral stimulation might explain the complexity of EMDR, and more specifically the way how attention and bilateral stimulation may systematically and effectively enable the processing of trauma-related cognition and emotion.
To address this issue, I suggest that EMDR may gain in being re-examined in integrative models. The Threshold Interval Modulation with Early Release-Rate of rIse Deviation with Early Release (TIMER-RIDER) model, a new framework of motor and cognitive control (see “Towards a Physiological Framework of Motor and Cognitive Control” Section) might be useful to achieve this goal. Before entering the core of this model, I first introduce the reader to the LATER model (Carpenter,
Though eye movements are only one aspect of bilateral stimulation in EMDR, it is worth focusing on their properties given the rich information they provide about brain functioning. To explore the visual world, humans make a variety of eye movements. Version or conjugate eye movements refer to movements of the two eyes in the same direction (to the left or to the right, up or down, or in any oblique direction). In vergence or disconjugate eye movements, the eyes move in the opposite direction: they converge at close or diverge at far. Both version and vergence eye movements can be either step (they are saccades in direction and step vergence in depth, respectively) to jump from one area of the visual scene to another one, or smooth (pursuit and vergence) to follow a target moving in direction or in depth. Besides intentional movements, the eyes also move in a reflex way to stabilize images on the retina during head movements (vestibulo-ocular reflex) or to track a moving visual pattern (optokinetic reflex). Even when the eyes fixate a single point and seem stationary, they are never at rest and continuously move through micromovements—tremor, drifts and microsaccades—which are crucial for vision (Coubard,
Every (little or large) movement offers researchers a variety of behavioral parameters for their research purpose: reaction time or latency, duration, amplitude, gain, peak velocity.
Saccades are extraordinary movements as their stereotyped behavior provides reliable measures subjected to detailed and quantitative analysis, while their underlying neural circuitry forms a well-known system that offers insight into brain functioning (Carpenter,
On a traditional frequency histogram (see Figure
Taken together with the fact that SRT is surprisingly long and variable from trial to trial, Carpenter (
More interestingly for our purpose, two types of changes can occur in LATER model, which provides researchers direct information about underlying central motor control or decisional mechanisms (Carpenter and Williams,
Researchers have here a tool to test any type of intervention, pharmacological (Michell et al.,
I recently introduced a new framework to bridge attention and decision models, which may be useful to re-examine EMDR underlying mechanisms—TIMER-RIDER model (Coubard,
Theoretically, attentional control has been described in temporal or top-down theories as a supervisory system gathering multiple cognitive processes (e.g., Norman and Shallice,
TIMER-RIDER is a framework that aims at bridging attention and decision by emphasizing two processes, inhibition and decision, and introducing two modulators for regulating these processes, so-called TIMER and RIDER (Coubard,
In this model, excitatory mechanisms for movement activation are embodied in LATER units. Considering the visuomotor modality, visual information from the retina of both eyes is processed to build a percept of stimulus location in space. As soon as the visual stimulus jumps from one location to another one, the LATER unit devoted to saccade will be activated (see Figure
Inhibitory mechanisms for movement suppression take the form of a global network lodging two modulators to regulate both the inhibition process and the LATER units (see Figure
To regulate movement, the attention-inhibition lodges two modulators: TIMER and RIDER. TIMER stands for
Importantly, the rate of rise of TIMER (T) and RIDER (R) signals is supposed to be linearly and inversely correlated to the Inhibition Process (I). In other words, the increasing activity of TIMER and of RIDER causes a direct decreasing activity of the inhibition process. Reversely, a higher inhibition process causes the reduction of TIMER and of RIDER activities (Coubard,
Thus, TIMER and RIDER have the ability to modulate the subject attentional state, leading to attentional fluctuations over time depending on their level of activity. We suggest that such attentional state may be influenced by internal or external contingencies, such as emotional disorder or therapeutic intervention. The TIMER-RIDER model also formalizes some distinction between express vs. early movements (see Figure
In summary, I propose that TIMER-RIDER model may be seen as physiological framework of motor and attentional control. Indeed, Claude Bernard was already writing from 1866 that “the brain exerts on reflexive movements a certain influence that we have characterized already, by calling it moderative action. But there is also a directive action.”
How such a model reflects physiological reality in the human brain and how it may contribute to our understanding of anxiety disorders and of EMDR mechanism are the issues we now discuss.
The eyes move thanks to oculomotor muscles, which are innervated by motor neurons, which are themselves innervated by burst neurons in the brainstem. Different burst neurons control the subtypes of eye movements. As example, burst neurons for horizontal saccades are located in the paramedian pontine reticular formation (PPRF; Fuchs et al.,
For inhibitory mechanisms, the attention-inhibition network involves different areas from the brainstem to the prefrontal cortex. Premotor neurons are under the inhibitory control of omnipause neurons (OPN) located in the pontine raphe (Scudder et al.,
How TIMER and RIDER modulation of LATER units may be implemented in the brain has been revealed by an fMRI study examining the neural correlate of LATER model (Domenech and Dreher,
In terms of TIMER-RIDER model, this suggests that the TIMER signal (T) for modulating distance to decision threshold approximates ACC activity, while the RIDER signal (R) for modulating decision gain fits rDLPFC activity (see Figure
Let’s now examine how TIMER-RIDER framework helps to understand further PTSD and its treatment by EMDR.
It is well established that emotion and attention are tightly linked such that acting on one influences the other (for a review, see Brosch et al.,
Emotional bypass occurs whenever emotion overcomes attention, and takes different forms from mindless emoting and hypocontrol (Greenberg and Safran,
Cognitive stifling occurs whenever attention overcomes emotion, and takes various forms from feeless thinking and hypercontrol (Greenberg and Safran,
In this context, the goal of EMDR in PTSD is to restore some balance between these two tendencies, over-emotion and over-attention, to lead to the state of harmonious flow of cognition, emotion and physical sensations (Servan-Schreiber,
My first suggestion is that EMDR may boost the TIMER modulator and ACC activity (see Figures
Functional brain imaging studies are compatible with this suggestion. Indeed, a meta-analysis of neuroimaging of emotional processing in PTSD and other anxiety disorders has evidenced hypoactivation in the dorsal and rostral ACC and ventromedial prefrontal cortex-structures linked to the regulation of emotion (Etkin and Wager,
Whether EMDR may also boost the RIDER modulator and rDLPFC activity is more hypothetical. As previously mentioned, bilateral stimulation may help to modify to generate imbalance in automatisms (Shapiro,
An easy way to test my prediction will be to measure SRT before and after an EMDR procedure. If my suggestion is correct, EMDR by boosting TIMER rather than RIDER should result in a swiveling rather than a shift of SRT recinormal distribution (see Figure
Finally, TIMER-RIDER is also compatible with an effect of bilateral stimulation on its own, particularly of overt or covert eye movements
The different aspects of EMDR procedure as well as hypotheses and theories accounting for its beneficial effects have put emphasis on attentional control in the process of targeting the emotional deficit. Indeed, EMDR seems to restore some balance in attention and emotion between two tendencies of anxiety disorders, over-emotion vs. over-attention. In the framework of movement models such as the TIMER-RIDER—model, EMDR might restore some balance between two types of responses, uncontrolled vs. over-controlled movements. Such effect may be obtained by two processes acting in parallel: activity level enhancement of a specific component of attention control, presumably energizing or TIMER modulator; and bilateral stimulation in any sensorimotor modality. The two processes result in lower inhibition and higher distractibility enabling dysfunctional information to be integrated. The TIMER-RIDER model, by integrating theories of EMDR, eye movement neurophysiological findings, and functional brain imaging of PTSD and of EMDR intervention, may be a useful integrative model to study attentional and/or emotional disorders, such as anxiety disorders.
OAC confirms being the sole contributor of this work and approved it for publication.
The author declares that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.
OAC thanks Pr. C. Birkan-Berz (Université Sorbonne Nouvelle, Paris) for her help to translate French literature of the 19th century.
1The original text in French is “C’est toujours par hasard que tout commence. La science ne vient qu’après et elle raisonne sur ce que le hasard a montré.”
2The original text in French is the epitaph.
3The original text in French is “[L]e cerveau exerce sur les mouvements réflexes une certaine influence que nous avons caractérisée déjà, en l’appelant action modératrice. Mais il y a aussi une action directrice.”